Centaur SP 4230

31276-*front matter 3/7/05 3:18 PM Page 1
Virginia P. Dawson and Mark D. Bowles
The NASA History Series
National Aeronautics and Space Administration

Office of External Relations
Washington, DC 20546
2004
NASA SP-2004-4230
31276-*front matter 3/7/05 3:18 PM Page 2
Library of Congress Cataloging-in-Publication Data

Dawson, Virginia P. (Virginia Parker)
Taming liquid hydrogen : the Centaur upper stage rocket, 1958-2002 / Virginia P. Dawson,
Mark D. Bowles. p. cm. (NASA-SP-2004-4230)
Includes bibliographical references and index.
1. Centaur rocketHistory. 2. Hydrogen as fuelResearchUnited StatesHistory. 3.
Liquid propellant rocketsResearchUnited StatesHistory. I. Bowles, Mark D. II. Title.
TL783.4.D39 2004
621.4356dc21
2004042092
31276-*front matter 3/7/05 3:18 PM Page i
Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iii
Chapter 1: Centaurs Origins in Atlas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Chapter 2: Marshalls Unruly Wards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Chapter 3: Abes Baby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Chapter 4: Heavy Lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103
Chapter 5: The Giant Titan-Centaur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139
Chapter 6: Centaur Reborn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167
Chapter 7: Eclipsed by Tragedy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
Chapter 8: Like a Phoenix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221
List of Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259
The NASA History Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267
31276-*front matter 3/7/05 3:18 PM Page ii
Introduction iii
Introduction
During its maiden voyage in May 1962, a Centaur upper stage rocket, mated to an Atlas
booster, exploded 54 seconds after launch, engulfing the rocket in a huge fireball. In ve s t i g a t i o n
re vealed that Centaurs light, stainless-steel tank had split open, spilling its liquid-hyd rogen fuel
down its sides, where the flame of the rocket exhaust immediately ignited it. Coming less than
a year after President Kennedy had made landing human beings on the Moon a national
priority, the loss of Centaur was re g a rded as a serious setback for the National Ae ronautics and
Space Administration (NASA). During the failure investigation, Homer Newell, Di rector of
Space Sciences, ruefully declared: Taming liquid hyd rogen to the point where expensive oper-
ational space missions can be committed to it has turned out to be more difficult than anyone
supposed at the outset.1
After this failure, Centaur critics, led by Wernher von Braun, mounted a campaign to cancel
the program. In addition to the unknowns associated with liquid hydrogen, he objected to the
unusual design of Centaur. Like the Atlas rocket, Centaur depended on pressure to keep its paper-
thin, stainless-steel shell from collapsing. It was literally inflated with its propellants like a football
or balloon and needed no internal structure to give it added strength and stability. The so-called
pressure-stabilized structure of Centaur, coupled with the light weight of its high-energy cryo-
genic propellants, made Centaur lighter and more powerful than upper stages that used
conventional fuel. But, the critics argued, it would never become the reliable rocket that the
United States needed.
Others, especially military proponents of Centaur, believed that accepting the challenge of
developing liquid-hydrogen technology was an important risk to take. Herbert York, Chief
Scientist for the Advanced Research Projects Agency (ARPA), had urged NASA Administrator T.
Keith Glennan to expedite the program in 1959 because Centaur was the only vehicle that has
the capability of meeting our payload requirements for high altitude orbits.2 Six months after
1
U.S. House of Representatives Subcommittee on Space Sciences of the Committee on Science and Astronautics, The Centaur
Program, 87th Congress, 15 and 18 May 1962 (hereafter cited as Centaur Program, 1962 Hearings), 11.
2
Herbert York to T. Keith Glennan, 19 September 1959, Centaur files, NASA Historical Reference Collection.
iv Taming Liquid Hydrogen
Centaurs aborted first flight, President John F. Kennedy demanded to know what NASA hoped
to achieve with Centaur. Administrator James Webbs response was unequivocal. With the Apollo
program already committed to using liquid hydrogen in the upper stages of the giant Saturn
vehicle, Centaur would prove the fuels feasibility.3 The Agency was also counting on Atlas-
Centaur to launch Surveyor, a robotic spacecraft with a mission to determine whether the Moons
surface was hard enough to land future spacecraft with human beings aboard.
Despite criticism and early technical failures, the taming of liquid hydrogen proved to
be one of NASAs most significant technical accomplishments. Centaur not only succeeded
in demonstrating the feasibility of liquid hydrogen as a rocket fuel, but it also went on to a
brilliant career as an upper stage for a series of spectacular planetary missions in the 1970s.
Iro n i c a l l y, this success did little to ensure the future of the Centaur rocket. Once the Shuttle
became operational in the early 1980s, all expendable launch vehicles like Centaur we re
slated for termination. Centaur advocates fought to keep the program alive. They won
funding for the redesign of Centaur as an upper stage for the Shuttle, spent nearly $1 billion
integrate them, and then witnessed the cancellation of the program within weeks of the first
scheduled flight of Shuttle/Centaur.
Miraculously, Centaur survived into the commercial era of the 1990s and is still flying as the
upper stage for the Atlas. Although unthinkable at the height of the Cold War, the idea of priva-
tizing the delivery of launch vehicle services gained currency in the early 1980s because it
dovetailed with the free-enterprise, small-government ideology of the Reagan administration.
Now NASA is just a customeralbeit a favored oneof a new service that rocket manufacturers
offer to a variety of customers, including foreign governments. A competitor of the European
Ariane rocket and the Boeing Delta, Centaur continues as the upper stage for a redesigned
Lockheed Martin Atlas.
Centaurs importance in the history of ro c k e t ry has escaped most historians of the space
program. One reason is that upper stage rockets compose the murky middle phases of spaceflight
that rarely make news headlines. They do not create the dramatic plumes of fire that the public sees
on the launch pads at Kennedy Space Center. Nor do they carry space probes all the way to the
distant planets. The booster or first stage of a multistage rocket uses the brute force of its large rocket
engines to propel the stages and payload through the atmosphere. Once the booster has used up its
propellants, it is jettisoned rather than burdening the remaining stages with the weight of a spent
rocket. At this point, Centaur takes over. The role of the upper stage is briefusually about 10
minutes of glory between the shutdown of the booster engines and the release of the spacecraft or
satellite. Centaurs final task is simply to get out of the way, while the payloadthe spacecraftis
just beginning its (often multiyear) journey.
3
Transcript of Presidential Meeting in the Cabinet Room of the White House, 21 November 1962, 19, tape no. 63, John F.
Kennedy Library Pre s i d e n ts Office files, NASA Historical Re f e rence Collection.
Introduction v
Despite this neglect, Centaurs importance to the space program, and satellite communications
in particular, is unquestionable. Centaur serves as the critical link between its booster stage (Atlas or
Titan) and the missions payload (satellite or spacecraft). The sole objective of the Centaur is to add
the extra speed needed to guide the payload into a desired orbit and to orient it before separa-
tion. The more accurately Centaur does its job of positioning, the less need there is for the
payload to use its own fuel to make up for any inaccuracies in trajectory. With this extra fuel, a
spacecraft bound for a planet has more maneuvering capability both as it travels tow a rd its desti-
nation and once it gets there. If the payload is a satellite, accurate positioning allows it to stay in
orbit for a longer period of time. The ability of Centaur to restart its engines in space allow s
mission designers greater flexibility in accommodating the relative positions of the moving and
rotating Earth and moving payload targets, whether planetary or lunar.
Though never identified with the dream of landing human beings on the Moon, or the
product of a massive military crash program like the Atlas, Centaur has enjoyed an unusually long
and sometimes controversial career as an upper stage rocket. As our title suggests, Centaur is espe-
cially notable because of its role in the development of liquid hydrogen as a rocket fuel.
Hydrogena light and extremely powerful rocket propellanthas the lowest molecular weight of
any known substance and burns with extreme intensity (5,500F). In combination with an
oxidizer such as liquid oxygen, liquid hydrogen yields the highest specific impulse, or efficiency in
relation to the amount of propellant consumed, of any known rocket propellant.
Because liquid oxygen and liquid hydrogen are both cryogenicgases that can be liquefied
only at extremely low temperaturesthey pose enormous technical challenges. Liquid hydrogen
must be stored at minus 423F and handled with extreme care. To keep it from evaporating or
boiling off, rockets fuelled with liquid hydrogen must be carefully insulated from all sources of
heat, such as rocket engine exhaust and air friction during flight through the atmosphere. Once
the vehicle reaches space, it must be protected from the radiant heat of the Sun. When liquid
hydrogen absorbs heat, it expands rapidly; thus, venting is necessary to prevent the tank from
exploding. Metals exposed to the extreme cold of liquid hydrogen become brittle. Moreover,
liquid hydrogen can leak through minute pores in welded seams. Solving all these problems
required an enormous amount of technical expertise in rocket and aircraft fuels cultivated over a
decade by researchers at the National Advisory Committee for Aeronautics (NACA) Lewis Flight
Propulsion Laboratory in Cleveland.
Today, liquid hyd rogen is the signature fuel of the American space program and is used by other
countries in the business of launching satellites. In addition to the Atlas, Boeings Delta III and Delta
IV now have liquid-oxygen/liquid-hyd rogen upper stages. This propellant combination is also
burned in the main engine of the Space Shuttle. One of the significant challenges for the European
Space Agency was to develop a liquid-hyd rogen stage for the Ariane rocket in the 1970s. The Sov i e t
Union did not even test a liquid-hyd rogen upper stage until the mid-1980s. The Russians are now
designing their Angara launch vehicle family with liquid-hyd rogen upper stages. Lack of Soviet
vi Taming Liquid Hydrogen
liquid-hydrogen technology proved a serious handicap in the race of the two superpowers to the
Moon.4 Taming liquid hydrogen is one of the significant technical achievements of twentieth-
century American rocketry.

From our study of Centaur history, we have identified three themes. The first is Centaurs remarkable
s u rv i val in the face of three decades of attempts to cancel the program. The acceptance of liquid
h yd rogen as a rocket fuel demonstrates the truism that success or failure of an innovation never depends
on technology alone. Thomas Hughes, a prominent historian of technology, has argued that large-scale
technology organizations, whether public or private, often stifle innovation. But innovations sometimes
win acceptance for reasons that are external to the technology itself.5 Other historians have expanded on
this theme, contending that technology has more to do with people, their values, and the external chal-
lenges they face than with the intrinsic superiority of one system over another.6 We argue that liquid
hydrogen as a rocket fuel would never have been adopted by NASA without the strong advocacy of Abe
Silverstein and others who came out of a background in liquid-fuels research at Lewis Research Center.
Initially, during the development phase of Centaur at Marshall Space Flight Center (discussed
in the first three chapters by Virginia Dawson), problems that put Centaurs surv i val in jeopardy
included the novelty of its fuel, its controversial design, and poor management of the program. After
the first test of the rocket failed, the program narrowly missed cancellation. The transfer of the
Centaur program in 1962 to Lewis Re s e a rch Center in Cleveland, Ohio, represented a decision by
Headquarters to keep innova t i ve aerospace companies like General Dynamics in business. At the
same time, former Vanguard personnel at Cape Canaveral took over what was then called
unmanned launch operations, allowing them to develop independently of the von Braun team that
managed the Apollo program on the other side of the Banana River.
In the 1970s, the proven technology of Centaur was almost cast aside during the second
phase of its history when new social and organizational priorities dictated a change in policy.
After a new reusable Space Transportation System (STS) received funding in 1972, expendable
launch vehicles we re considered obsolete. Iro n i c a l l y, the upper stages most significant contri-
butions to NASA occurred while the nation awaited the Shuttle. Centaurs service as an upper
stage for Atlas-Centaur and Titan-Centaur planetary missions is discussed in chapters 4 and 5
by Ma rk Bow l e s .
4
This point is made at various points by Asif A. Siddiqi, Challenge to Apollo: The Soviet Union and the Space Race, 19451974
(Washington, DC: NASA SP-2000-4408).
5
Thomas P. Hughes, American Genesis: A Ce n t u ry of In vention and Technological Enthusiasm (London: Penguin Books, 1989)
and Ne t w o rks of Power: Electrification in Western Society, 18801930, (Ba l t i m o re, MD: The Johns Hopkins Un i versity Pre s s ,
1983).
6
See essays by Wiebe Bijker, Thomas P. Hughes, and Trevor J. Pinch, eds., The Social Construction of Technological Systems: New
Directions in the Sociology and History of Technology (Cambridge, MA: MIT Press, 1987).
Introduction vii
Sh o rtly before the Shuttle began flying, the Eu ropean Ariane rocket became available for
launching communications satellites, ending a monopoly NASA had enjoyed since the 1960s.
NASA recognized that in order for the Space Shuttle to become the all-purpose vehicle envisioned
by NASA, the Shuttle required an upper stage to replicate the important job of positioning satellites
and spacecraft (already performed by Atlas-Centaur). The modification of Centaur to fit in the Shuttles
cargo bay won a reprieve for the Centaur program in the 1980s. Howe ve r, objections to the radical
nature of its design and liquid-hydrogen fuel threatened to scuttle the program.
The loss of Challenger in 1986 proved another turning point in Centaurs tortuous career.
Chapter 8, by Virginia Dawson, describes how Atlas-Centaur again rose like a phoenix from the
ashes of Challenger. With NASA no longer controlling its fate, Centaur was reborn in the Atlas as
a commercial launch vehicle. Commercialization of Atlas-Centaur in the 1990s provides an
example of technology transfer from the government to the private sector of the economya
policy promoted since NASAs founding. It also challenges the theory that without the stimulus
of the government to pay for innovation, companies like General Dynamics (later acquired by
Martin Marietta and now part of Lockheed Ma rtin) would have no incentive to develop new
technology. The Atlas rocket now faces the challenge of a highly competitive international
marketplace and declining demand for communications satellite launches.
Another theme, closely related to the first, is how NASAs changing tolerance for risk
influenced Centaur. How a rd Mc Cu rdy pointed out in his insightful study of NASA culture
that risk tolerance diminished as NASA mature d .7 Yet even in the 1960s, most of NASAs tech-
nical decisions focused on minimizing risk. For example, Administrator T. Keith Glennan had
p ressed Atlas into service for the Me rc u ry program because it was a known technology. He
thought the Agency should use already-developed missiles for launch vehicles, extending the
state of the art as little as necessary. 8 Centaur was an anomaly because of its novel fuel.
Glennan strongly supported its development because liquid hyd rogen promised a leap in
performance urgently needed to trump the Russians in space. Gl e n n a ns successor, Ja m e s
We b b, also became an advocate of Centaur, willing to stand up to critics like von Braun and
p owe rful members of the U.S. House of Re p re s e n t a t i ves Committee on Science and
A s t ronautics, which controlled NASAs purse strings.
Tolerance for risk enters the Centaur story again in the 1980s. Chapters 6 and 7, by Ma rk
B owles, present a critical examination of one of the most interesting episodes in NASA launch
vehicle history. The Shuttle/Centaur saga shows that the fate of technology can often rest upon
an unplanned-for contingencythe level of social acceptance of risk. Despite the enormous
advantages of sending the Galileo space probe to Jupiter on Shuttle/Centaur, the pro g r a m
e n c o u n t e red strong resistance from engineers at Johnson Space Center and the astronaut corps
7
Howard E. McCurdy, Inside NASA: High Technology and Organizational Change in the U.S. Space Program (Baltimore: The
Johns Hopkins University Press, 1993), 150.
viii Taming Liquid Hydrogen
located there. The astronauts thought that a liquid-hyd rogen rocket in the Shuttles cargo bay put
their lives at risk, despite the fact that the main engine of the Shuttle burned liquid hyd rogen
carried in a large external tank. Johnson Space Center engineers also questioned the capability of
the Lewis Centaur team to safely integrate Shuttle/Centaur into their vehicle. The Challenger
disaster forced NASA to redefine the Agencys risk limits and led to Shuttle/Centaurs cancellation.
Risk also entered the commercialization story when General Dynamics saved Atlas-Centaur
by pledging company funds to underwrite its manufacture and marketing. Commercialization
ended the companys dependence on the cost-plus-fixed-fee government contract. Ge n e r a l
Dynamics/Astronautics staked the future of the company on its ability to upgrade its rocket and
launch pad and sell launch services to the communications satellite industry. The gamble paid off,
butironicallynot for General Dynamics, but for its former rival and new owner, Martin Marietta.
A third theme of the book is the collaboration between government engineers and their
industry counterparts. Few books, with the exception of Joan Lisa Brombergs NASA and the Space
Industry, have explored this re l a t i o n s h i p. Bromberg points out that Glennan did not approve of
allowing the government to control the creative aspects of design while farming out to industry
only the repetitive and straight production items.9 His successor, James Webb, had also steered
the Agency away from developing an exc l u s i vely in-house technical capability. He envisioned
government-industry collaboration in the management of the large-scale technology re q u i red for
the space program. These birds are going to fly, Webb told President Kennedy in 1962, not
by what you put on the schedule or the amount of money you put in it, but the way this thing
[NASA] is run. He thought Centaur could validate the capacity of the government to run a
program like this in partnership with industry.10 This was a depart u re from the top-down in-
house Army model of working with contractors favo red by Wernher von Braun. The relationship
b e t ween General Dynamics and NASA under Marshall Space Flight Center (MSFC) manage-
ment had been adversarial, but after the transfer to Cleveland, the relationship gradually became
m o re collegial. Members of the NASA-industry Centaur team believed that they we re on the
front line of the Cold War and that Centaur held one of the keys to winning back the countrys
lost prestige. Lewis engineers demanded that contractors meet specifications, but they also
worked closely with their industry counterparts to improve the vehicle. Despite the continual
t h reat of cancellation, the NASA-industry team introduced a new Teledyne avionics system and
d e veloped a re vo l u t i o n a ry steering program to adjust for upper atmosphere winds encountered
immediately after liftoff.
As NASA matured, the government-industry partnership became the foundation for
Centaurs commercialization, described in chapter 8. Technology transfer from the government to
8
J. D. Hunley, ed., The Birth of NASA: The Diary of T. Keith Glennan (Washington, DC: NASA SP-4105, 1993), 13.
9
Quoted from Glennans diary by Joan Lisa Bromberg, NASA and the Space Industry (Baltimore: The Johns Hopkins University
Press, 1999), 40.
Introduction ix
the private sector of the economy was realized in the 1990s. No longer held back by the govern-
ment from improving Atlas-Centaur, General Dynamics made its launch vehicle comparable to
the European Ariane rocket in power and versatility. Ironically, although most of the upgrades to
the vehicle were made to the Centaur upper stage, the entire vehicle is now referred to simply as
the Atlas. Thus Centaur, still a vital part of the vehicle, has become invisible.
NASA Lewis (now Glenn) Research Center engineers and procurement officers helped to
u n r a vel the gove r n m e n ts complex contractual relationship with General Dynamics.
Commercialization liberated General Dynamics from a culture of government dependency. Today,
although human bonds formed over three decades are still strong, NASA Glenn no longer
manages Centaur. The limited oversight the government retains for NASA missions is currently
carried out by Kennedy Space Center.

Few secondary sources recognize the role Centaur played in the controversies surrounding the
acceptance of liquid hydrogen and its positive contributions to the unmanned space program. For
example, in On Mars, Edward and Linda Ezell dismiss Centaur as a genuine troublemaker because
of the delays it caused for the Surveyor and Mariner programs.11 In Clayton R. Koppess critical study
of the Jet Propulsion Laboratorys early unpiloted projects, JPL and the American Space Program, At l a s -
Agena and Atlas-Centaur receive scant notice,12 while a major new encyclopedia of Americas greatest
space programs leaves out the Centaur program altogether.13 An exception is the late John Sloop,
whose engagement with high-energy liquid rocket propellants in the 1950s shaped his career at the
Lewis Flight Propulsion Laboratory in Cleveland, one of three laboratories for the NACA. After
NASAs formation, he became Abe Silversteins technical assistant at Headquarters. Sloop recognized
the pivotal role of liquid hydrogen in the American space program, and upon re t i rement, he
published Liquid Hyd rogen as a Propulsion Fuel, 19451959, a detailed and insightful study that
concludes in 1959, before NASA took over Centaur program management from the Air Fo rc e .14
His interv i ews (now part of the NASA Hi s t o ry Collection) with former General Dynamics
employees, notably Krafft Ehricke and Deane Davis, as well as other papers in the NASA
Historical Collection, provided an indispensable starting point for our study. Another source that
provided essential background in understanding Centaurs significance was Roger E. Bi l s t e i ns
Stages to Sa t u rn.15 Se veral superb papers by Richard Ma rtin, formerly of General Dynamics,
contributed to our understanding of the unique stru c t u re of the Centaur.16 Articles by G. R.
10
Transcript of Presidential Meeting, 20 November 1962, 45.
11
Edward Clinton Ezell and Linda Neuman Ezell, On Mars (Washington, DC: NASA SP-4212, 1984), 47.
12
Clayton R. Koppes, JPL and the American Space Program (New Haven: Yale University Press, 1982).
13
Frank N. Magill and Russell R. Tobias, eds., USA in Space, three volumes (Pasadena, CA: Salem Press, Inc., 1996).
x Taming Liquid Hydrogen
Richards and Joel W. Powell, and Joseph Green and Fuller C. Jones are among the few specifi-
cally focused on Centaurs origins and early history.17 Our understanding of the development of
Pratt & Whitneys RL10 engine was enhanced by work of Joel Tucker and Dick Mulready.18
Finally, the work by John Krige on the Eu ropean Space Agency provided valuable backgro u n d
for our chapter on commercialization.19
One of the themes in Virginia Da w s o ns earlier book, Engines and In n ovation,20 a history of the
NASA Lewis Re s e a rch Center, was the tension between the re s e a rch side of the laboratory and the
real-time demands of managing the Agena and Centaur missions for NASA. The management of
Centaur is arguably Lewis Re s e a rch Centers most important contribution to the space program. In
this book, we emphasize how the research culture and test facilities of the Center materially
contributed to Centaur development. If, as some of the re v i ewers of our manuscript fromGe n e r a l
Dynamics have commented, we may have overstated the Lewis role, we can only offer the hope that
a well-documented history of General Dynamics/Astronautics21 can redress any imbalance that we
may have created. Although we we re fortunate to be able to interv i ew some former General
Dynamics employees, we we re disappointed that ve ry few company documents can be found in
publicly accessible arc h i ves.
This history of Centaur is the result of an association that began nine years ago when NASA
history first brought the authors together. We collaborated on an essay about NASAs Advanced
Turboprop Project in a book on Collier trophy winners.22 Since that time, we have coauthored a
variety of projects that have come to us through our company, History Enterprises, Inc. We have
14
John L. Sloop, Liquid Hydrogen as a Propulsion Fuel 19451959 (Washington, DC: NASA SP-4404, 1978). A native of North
Carolina, Sloop earned a B.S. in electrical engineering from the University of Michigan in 1939. Hired by the NACA in 1941,
he moved to Cleveland in 1942 and remained at Lewis until his move to Washington, DC, in 1958.
15
Roger E. Bilstein, Stages to Saturn (Washington, DC: NASA SP-4206, 1980).
16
See Richard E. Martin, The Atlas and Centaur Steel Balloon Tanks: A Legacy of Karel Bossart, 40th Congress of the
International Astronautical Federation, Malaga, Spain, 814 October 1989, IAA-89-738, Cot 7-13m, reprint; A Brief History
of the Atlas Rocket Vehicle, Part III, QuestThe History of Spaceflight Quarterly 8 (2001): 48; Atlas II and IIA Analyses and
Environments Validation, Acta Astronautica 35 (1995): 771791.
17
G. R. Richards and Joel W. Powell, The Centaur Vehicle, Jo u rnal of the British In t e r p l a n e t a ry Society 42 (1989):
99120; Joseph Green and Fuller C. Jones, The Bugs That Live at -423, Analog Science Fiction/Science Fa c t 80 (1968):
841.
18
See Joel E. Tucker, The History of the RL 10 Upper-Stage Rocket Engine, 19561980, in History of Liquid Rocket Engine
Development in the United States, 19551980, ed. Stephen E. Doyle, AAS History Series (San Diego: AAS, 1992), vol. 13,
123151. Also, Dick Mulready, Advanced Engine Development at Pratt & Whitney: The Inside Story of Eight Special Projects,
19461971 (Warrrendale, PA: Society of Automotive Engineers, Inc., 2001), chapters 34.
19
See J. Krige, A. Russo, and L. Sebesta, A History of the Eu ropean Space Ag e n c y, 19581987 (ESA SP-1235 April 2000),
vol. 2.
20
Virginia P. Dawson, Engines and Innovation: Lewis Laboratory and American Propulsion Technology (Washington, DC: NASA
SP-4306, 1991).
21
A company history by Bill Yenne, Into the Sunset: The Convair Story (Lyme, CT: Greenwich Publishing Group, 1995), could
be used as a starting point.
Introduction xi
enjoyed ferreting out the technology involved in developing and managing the Centaur rocket and
have tried to achieve a balance among the technology, the scientific goals of the missions them-
selves, and the political aspects of the program. Our first essays on Centaur were published in To
Reach the High Frontier, edited by Roger D. Launius and Dennis R. Jenkins.23
We bring very different perspectives to NASA history. Mark Bowles grew up reveling in the
achievements of the planetary probes of the 1970s. As a boy, he shared with his father the deep
sense of awe in the mysterious dark universe surrounding Earths tiny blue oasis of life. He recalls
how they would gaze up at the Moon during the Apollo missions and imagine the activities of the
astronauts.
Virginia Dawsons engagement with NASA history came through the encouragement she
received from the late Professor Melvin Kranzberg, founder of the field of the history of tech-
nology, who served as chair of the NASA History Advisory Committee for many years. In his
foreword to This New Ocean: A History of Project Mercury, he pointed out that the legislation creating
the National Aeronautics and Space Administration in 1958 included not only the charge to
expand human knowledge of phenomena in the atmosphere and space, but also to provide for
the widest possible dissemination of information about its activities. NASA wisely interpreted
this mandate to include responsibility for documenting the epochal progress of which it is the
focus, he wrote. The result has been the development of a historical program by NASA as
unprecedented as the task of extending mans mobility beyond his planet.24 Dr. Kranzberg
strongly believed that professional historians working outside academe could make valuable
contributions to the history of twentieth-century institutions. He was particularly impressed that
NASA gave authors unrestricted access to participants and unclassified sources and encouraged
them to examine Agency history critically.
We enjoyed the assistance and encouragement of many current and former NASA people.
First, we would especially like to thank Joe Nieberding, who enthusiastically took us under his wing
in the early stages of the book and pushed the project along. He opened doors to interv i ews with
people at Glenn Re s e a rch Center, Kennedy Space Center, the Jet Propulsion Laboratory, and
Lockheed Ma rtin, as well as former employees of General Dynamics in San Diego. We would also
like to thank Me yer Reshotko for helping us get started on this project. J. Cary Nettles, Roger
Lynch, and Richard Ma rtin gave us copies of key photos. We are greatly indebted to the people who
patiently re v i ewed our draft of the manuscript. These included Joe Ni e b e rding, Frank Spurlock,
22
Mark D. Bowles and Virginia P. Dawson, The Advanced Turboprop Project: Radical Innovation in a Conservative
Environment, in From Engineering Science to Big Science: The NACA and NASA Collier Trophy Research Project Winners, ed. Pam
Mack (Washington, DC: NASA SP-4219, 1998), 321343.
Virginia P. Dawson, Taming Liquid Hydrogen: The Centaur Saga, 334356, and Mark D. Bowles, Eclipsed by Tragedy:
23
The Fated Mating of the Shuttle and Centaur, 415442, To Reach the High Frontier: A History of U.S. Launch Vehicles, ed. Roger
D. Launius and Dennis Jenkins (Lexington: The University Press of Kentucky, 2002).
24
Loyd S. Swenson, Jr., James M. Grimwood, and Charles C. Alexander, This New Ocean: A History of Project Mercury
(Washington, DC: NASA SP-4201, 1966), V.
xii Taming Liquid Hydrogen
R i c h a rd Ma rtin, Larry Ross, Ed w a rd Bock, and John Neilon. History colleagues Jo h n

Hunley and Dwayne Day not only read our manuscript with extraord i n a ry care and insight,
but also shared with us documents and articles important to our story. Alan Lovelace, Marty
Wi n k l e r, Ron Eve rett, Cary Nettles, Art Zimmerman, Len Perry, and Del Tischler re v i ewe d
individual chapters. All their comments proved exceedingly helpful. However, we take
responsibility for our interpretation of events and any errors of fact that might have gone
u n c o r rected. We are also ve ry grateful to the many people we interviewed, either by tele-
phone or in person. They are listed at the end of the book. Our understanding was also
enhanced less formally by many others too numerous to cite here. Many of the people we
interviewed spoke of the extraord i n a ry dedication of St e ve Szabo. Because of his untimely
death, we were not able to interview him, but we hope that we have captured some of the
Centaur team spirit that he helped to create.
As colleagues in NASA history will appreciate, finding documentation for government-funded
studies is always a challenge. For this history, we were fortunate to have the assistance of Kevin
Coleman, in charge of the history program and records management at Glenn; he, along with
Deborah Demaline, proved to be a judicious and support i ve ally. We used Glenn records stored in the
famous World War II bunkers at Plum Brook and at the National Archives and Record Center in
Chicago. Bonita Smith made certain we had access to Centaur material in the NASA Glenn Archives.
Janice Nay and Lynn Patterson provided excellent transcripts of our oral interv i ews. Ro b e rt Arrighi of
Hi s t o ryEnterprises assisted with re s e a rch.
During the writing of Engines and Innovation, Virginia Dawson and historian of science
colleague Craig Waff discovered a trove of records of the Launch Vehicles Division stored in the
vault of the Development Engineering Building at Glenn. In looking for Centaur records more
than ten years later, we found that some boxes containing correspondence fortuitously were left
behind in the move to Kennedy Space Center. Mark Bowles was able to use some of this material
to document the Shuttle/Centaur story. The contents of these boxes, catalogued by Galen Wilson,
are referred to as NASA GRC Records to distinguish them from the NASA GRC Archives. The
ELV Resource Library, managed by Boeing at Kennedy Space Center, loaned us a selected set of
historical Lewis documents. Robert Bradley made available Krafft Ehrickes early Air Force reports,
now located in the archives of the San Diego Aerospace Museum. We also used Centaur records
in the NASA Historical Reference Collection in the NASA History Office in Washington, District
of Columbia, many of which were assembled through the efforts of the late John Sloop. We are
grateful for the assistance and encouragement we received from Roger Launius, Steve Dick, Steve
Garber, Louise Alstork, and Jane Odom in the NASA History Office and Mike Wright at Marshall
Space Flight Center.
Last, but certainly not least, many thanks are due to the professional graphic designers,
editors, and print specialists who made this book and accompanying CD physical realities. In
the NASA He a d q u a rters Printing and Design Office, Douglas Ortiz, Joel Vendette, and James
Introduction xiii
Gitlin expertly handled the layout of the book; Lisa Jirousek carefully edited the book; and
Jeffrey McLean took care of the printing process. Special thanks are also due to Kristin Jansen
of NASA Glenn Re s e a rch Center, who devised an innova t i ve format for the CD that accompa-
nies the printed text.
31276-chapter 1 3/8/05 10:49 AM Page 1
Centaurs Origins in Atlas 1
Chapter 1
Centaurs Origins in Atlas

Going back to the old ideas of Oberth, I said I have a relatively dense
first stage. And it so happens that the second stage, because it is less dense,
fits just beautifully on the first stage. Now all we have to do is remove the
neck from Atlas and make it cylindrical all the way and we have a 10-
foot-diameter base. For a second stage, thats just beautiful.
Krafft Ehricke, Centaurs designer
In early 1956, the development of Atlas preoccupied Krafft Ehricke. Known for his enormous
capacity for work, he spent 18-hour days driven by the urgency of producing the countrys first
intercontinental ballistic missile (ICBM). At the same time, Ehricke envisioned more peaceful
applications for Atlas. Why not add a second stage to produce a two-stage rocket capable of placing
communications or weather satellites into orbit, sending instrumented space probes to the Moon
and the planets, or installing an orbiting platform around Earth? He even calculated how the Atlas
itself could become a satellite when launched into low-Earth orbit.
This was the kind of fantasy that Ehricke had entertained ever since his youth in Weimar,
Germany. Mesmerized by the 1929 Fritz Lang film Frau im Mond (Woman in the Moon), which he
saw more than a dozen times, he said, I felt, My God, it really must be possible to get to the
Moon, which for an 11-year-old boy is a kind of revelation.1 The son of two dentists, Ehricke
tinkered with their dental apparatus and chemical compounds while he steeped himself in the ideas
of the early rocket pioneers. He recalled the inspiration for Centaur: Going back to the old ideas
of Oberth, I said I have a relatively dense first stage. And it so happens that the second stage,
1
Interview with Krafft Ehricke by John Sloop, 26 April 1974, NASA Historical Reference Collection, NASA History
Office, Washington, DC. On Ehrickes biography and ideas, see Marsha Freeman, How We Got to the Moon: the Story of
the German Space Pioneers (Washington, DC: 21st Century Associates, 1993), 292339; and Krafft Ehrickes
Extraterrestrial Imperative: A Memoir, in History of Rocketry and Astronautics, eds. Donald C. Elder and Christophe
Rothmund (San Diego: American Astronautical Society, 2001), AAS History Series, vol. 23, 163222. See also Al
Vinzant, History of Centaur, Exclusive (GD Space Systems Division Newsletter), August 1988; and Daniel A. Heald,
LH2 Technology Was Pioneered on Centaur 30 Years Ago, 43rd Congress of the International Astronautical
Federation, August 1992, Washington, DC.
31276-chapter 1 3/8/05 10:49 AM Page 2
2 Taming Liquid Hydrogen
Krafft Ehricke, Centaurs designer and first program director (left), discusses advanced spacecraft designs with James Dempsey,
president of General Dynamics/Astronautics, early 1960s. (Courtesy of Lockheed Martin)
because it is less dense, fits just beautifully on the first stage. Now all we have to do is remove the
neck from Atlas and make it cylindrical all the way and we have a 10-foot diameter base. For a
second stage, thats just beautiful.2 Because the two-stage rocket he envisioned for spaceflight
needed a very powerful upper stage, he considered high-energy propellant combinations like liquid
hydrogen with fluorine as the oxidizer, hydrazine with fluorine, methane with ozone, and liquid
hydrogen with liquid oxygen.3 These studies smacked of science fiction in 1956, and Ehricke
expected James Dempsey, his boss at the Convair Division of General Dynamics, to put an end to
them. Instead, Dempsey urged him to continue.4
Ehricke championed the idea of space travel. He chaired the Space Flight Committee for the
American Rocket Society, which comprised some of the nations most prominent missile
2
Interview with Krafft Ehricke by John Sloop, 26 April 1974, NASA Historical Reference Collection.
3
Al Vinzant, History of Centaur.
4
Shirley Thomas, Men of Space, vol. 1 (Philadelphia: Chilton Company, 1960), 2.
31276-chapter 1 3/8/05 10:49 AM Page 3
Diagram: Atlas-Centaur. (NASA E2283)
designers, including Karel Charlie Bossart of Convair, Wernher von Braun of the Army
Ballistic Missile Agency, Milton Rosen of the Naval Research Laboratory, and Hubertus
Strughold of Randolph Air Force Base. Shortly before the Soviet Union launched Sputnik, the
worlds first artificial satellite, the committee finished a report that they then submitted to
President Eisenhower less than two weeks after Sputniks launch. The report proposed a national
31276-chapter 1 3/8/05 10:49 AM Page 4
space agency to benefit all mankind.5 Ehricke thought the new space age would usher in a new
age of discovery and international cooperation. He envisioned space travel as a serious, practical
and worthwhile effortnot at some future time, but right now, in this century and in this age
of ours. He thought space travel appealed to mans highest aspirations and would promote inter-
national peace and goodwill.6
In response to Sputnik, James Dempsey asked Ehricke to proceed with a design for an upper
stage for the Atlas missile, transforming the weapon into a two-stage spacecraft for travel to the outer
reaches of the solar system. He was encouraged to move offsite with an elite team of the companys
engineers that included Atlas veterans Charlie Bossart, William Patterson, Howard Dunholter, Frank
Dore, William Radcliffe, James Crooks, and Hans Friedrich, a German expatriate who had designed
the autopilot for the German V-2 rocket. After debating the merits of various propellant combina-
tions, the group agreed that liquid hydrogen with the oxidizer liquid oxygen was the best propellant
for the high-energy upper stage Ehricke had in mind. Ehricke may have named Centaur for the
nearest star after the Sun, the Alpha Centauri, which might one day be approached by the vehicle
he envisioned.7 It has also been speculated that Ehricke chose the name because the mythic man-
beast with the body of a horse and the torso and head of a man reflected the hybrid nature of the
Atlas-Centaur combination, with Centaur carrying the guidance and control for both vehicles.
Despite the challenges of producing, storing, and handling liquid hydrogen, its distinct advantages
as a rocket propellant presented the rocket team at General Dynamics with a tantalizing opportunity
at a time when U.S. launch vehicle capability was well below that of the Soviet Union. Burning liquid
hydrogen with liquid oxygen produces the highest specific impulse or thrust of any known rocket fuel.
Theoretically, a liquid-hydrogen-powered rocket can lift approximately 35 to 40 percent more payload
per pound of liftoff weight than conventional rocket fuels.8
Hydrogen had remained a laboratory curiosity until the late nineteenth century, when scientists
driven by the promise of industrial applications for gases began to investigate how to convert gases to
liquids. The British scientist Michael Faraday succeeded in liquefying chlorine, ammonia, and carbon
dioxide. Oxygen and hydrogen proved more difficult to liquefy because they are cryogenicgases that
can be converted to liquids only at extremely low temperatures. In the 1880s, a method was found for
liquefying oxygen. Then, in 1898, Sir James Dewar succeeded in liquefying hydrogen for the first
5
ARS Urges National Space Flight Program, Astronautics 2 (December 1957): 1828.
6
Krafft A. Ehricke, The Anthropology of Astronautics, Astronautics 2 (November 1957) 28.
7
The Alpha Centauri suggestion is based on information from Harold Dunholter, former chief development engineer
for General Dynamics. See Helen T. Wells, Susan H. Whiteley, and Carrie E. Karegeannes, Origin of NASA Names
(NASA SP-4402, 1976), 1112.
8
Centaur, a pamphlet prepared by the Convair Aerospace Division of General Dynamics, dated May 1973, provided the
35-percent figure. Generally, NASA press releases use 40 percent. Centaur files, NASA Glenn History Office.
31276-chapter 1 3/8/05 10:49 AM Page 5
time. Dewar described the principle of regenerative cooling, a method that takes advantage of the
extreme cold of liquid hydrogen to cool hot metal surfaces during combustion. Dewar also invented
vacuum containers for safely storing liquid hydrogen.9
Liquid oxygen found applications in the 1920s as an oxidizer for alcohol or kerosene fuel, but
early rocket pioneers like Konstantin Tsiolkovskiy and Robert Goddard avoided liquid-hydrogen fuel
because of its well-known dangers.10 Hermann Oberth had calculated that liquid hydrogen was espe-
cially appropriate for an upper stage because in the near-vacuum of space, less pressure is required
to keep the sides of the rocket rigid.11 However, like Goddard and Tsiolkovskiy, Oberth had
avoided testing an oxygen/hydrogen rocket, presumably because of hydrogens volatility. The
tragic explosion of the hydrogen-gas-filled Hindenberg dirigible in 1937 seems to have reinforced
the general prejudice against hydrogen.
Although Ehricke understood the theoretical advantages of liquid hydrogen, he had
grossly underestimated the difficulties of developing a liquid-hydrogen rocket. A rocket
enthusiast during his student years in Berlin, he had filed two rocket patents, helped organize
the German Society for Space Research in 1938, and written a series of articles for the
Societys journal Space.12 But he had avoided testing a liquid-hydrogen/liquid-oxygen rocket
because of the fear that the mixture might explode prematurely.13 Graduate studies in nuclear
physics at Humboldt University under Werner Heisenberg were interrupted by the war.
Ehricke served at the Russian front as a tank driver before being recruited for work on the V-
2 at Peenemnde. There he came under the spell of Walter Thiel, Wernher von Brauns rocket
engine expert. Thiel inspired Ehricke to imagine rocket engines with millions of pounds of
thrust. Learning of Heisenbergs attempts to build a nuclear reactor, Ehricke and Thiel
dreamed of nuclear rocket propulsion with hydrogen as the working fluid.14
Thiel gave careful attention to the possibility of developing a liquid-hydrogen rocket. In a
significant memorandum written in 1937, Thiel made a survey of existing knowledge of the prac-
9
John D. Clark, Ignition! An Informal History of Liquid Rocket Propellants (New Brunswick, NJ: Rutgers University Press,
1972), 45.
10
See Exploring the Unknown: Selected Documents in the History of the U.S. Civil Space Program, vol. I (NASA SP-4218,
1995), 6768; John Sloop, Liquid Hydrogen as a Propulsion Fuel, 19451959 (Washington, DC: NASA SP-4404, 1978),
259.
11
In discussing hydrogen fuel, Hermann Oberth wrote, The thinner the outside air, the smaller are the forces which
tend to cause the rocket to buckle or collapse and the smaller is the inside pressure necessary for rigid filling. See Ways
to Spaceflight, tr. Wege zur Raumschiffahrt, originally published in 1929 (Washington, DC: NASA TT F-622, 1972), 343.
12
Marsha Freeman, Krafft Ehrickes Extraterrestrial Imperative, 164.
13
14
31276-chapter 1 3/8/05 10:49 AM Page 6
tical possibilities for improving liquid rockets.15 Thiel drew attention to the very large energy gain
that could be obtained from using high-energy propellants such as liquid hydrogen/liquid oxygen
and stated that a research effort appears definitely rewarding. At the same time, he tempered his
enthusiasm for liquid hydrogen by pointing out the strong obstacles to the development of a
liquid-hydrogen rocket, such as the need for insulated tanks and ducts to compensate for the
extremely low temperatures of liquid hydrogen/liquid oxygen. He pointed out that very little was
known about the behavior of metals at extremely low temperatures, particularly the permeability
of light metals to liquid hydrogen. This influential rocket expert was killed in an Allied bombing
raid on Peenemnde, but his ideas probably took root in the fertile mind of Krafft Ehricke. After
Peenemnde, the possibilities of liquid hydrogen continued to intrigue him.
In 1945, Ehricke managed to obtain a contract from the Department of the Army and joined
the German rocket group at White Sands, New Mexico. Between 1947 and 1950, he worked on
ramjet and rocket systems at Fort Bliss, Texas. Because there was little information to guide rocket
designers in the selection of propellants, the Journal of the American Rocket Society devoted an entire
issue to the subject in 1947. A Journal paper by Maurice Zucrow argued that, in addition to thrust,
propellant density must also be considered in assessing rocket performance, since it can affect
vehicle size and weight. In a chart, Zucrow showed that the liquid-oxygen/liquid-hydrogen combi-
nation produced greater thrust but required a larger tank because the low density of hydrogen took
up more volume.16
Ehricke recalled being impressed by a paper by JPL researcher Richard Canright that urged
more research on high-energy liquid propellants, especially liquid hydrogen with the oxidizer
hydrazine.17 In a 1974 interview, Ehricke said, I had again run into hydrogen, from Tsiolkovskiy
through von Hoefft to Oberth to Thiel and through my nuclear investigation. So I said, Its too
often that it has looked good. I think we ought to do it. But von Braun thought they should
stick with the denser propellants they already knew how to handle.18
15
Walter Thiel, Memorandum of the Practical Possibilities of Further Development of the Liquid Rockets and a Survey
of the Tasks to be Assigned to Research, 13 March 1937, tr. D. K. Huzel, folder CT-168000-01, Thiel, Walter,
National Air and Space Museum Archives. Thanks to historian John Hunley for sending us a copy of this document.
16
Ehricke says the paper was published in Journal of the American Rocket Society in 1947 and authored by Jet Propulsion
Laboratory rocket researcher Richard Canright. There is no paper by Canright in the journal between 1947 and 1953.
We believe that the paper referred to here is by Maurice Zucrow in the 1947 issue devoted to liquid propellants: Liquid
Propellant Rocket Power Plants, no. 72 (December 1947): 2844.
17
Interview with Krafft Ehricke by John Sloop, 26 April 1974, NASA Historical Reference Collection. A review of the
Engineering Index revealed that Richard Canright published Problems of Combustion: Liquid Propellant Rocket
Motors, in Chemical Engineering Progress 46 (1950): 228232. There is, however, no indication of a paper on a compar-
ison of rocket propellants.
18
Interview with Krafft Ehricke by John Sloop, 26 April 1974, NASA Historical Reference Collection. See also John
Sloop, Liquid Hydrogen, 4748.
31276-chapter 1 3/8/05 10:49 AM Page 7
In 1950, Ehricke moved with the German team to Huntsville, Alabama, where he joined the
Guided Missile Development Group at the Redstone Arsenal. He headed the Gas Dynamics Section
but chafed under von Brauns authority and his conservative approach to rocket design.19 In a rocket,
where every excess pound reduces the size of the possible payload and increases cost, Ehricke paid close
attention to propellant weights. In a paper comparing propellants and working fluids for rocket propul-
sion, he compared the weight of propellants, their specific impulse, and their density. He argued that
for ascent in a gravity field, heavy or medium propellants were preferable. However, once a staged rocket
reached space, the performance of light, less dense propellants like ozone-methane, hydrogen-oxygen,
and oxygen-hydrazine increased. They beat the heavier propellants because in zero gravity, their high
specific impulse and low density offset the weight of the structures needed to contain them. He
suggested that liquid-hydrogen upper stages needed investigation.20
At the end of his contract with the U.S. Army in 1952, Ehricke moved to a job at the Bell Aircraft
Corporation in Buffalo under Walter Dornberger, another German expatriate who had worked on the
V-2. In 1954, Ehricke joined Convair in San Diego as a design specialist in the same year that it became
a division of General Dynamics. He headed the Preliminary Design and Systems Analysis group. He
later served as the director of the Centaur program for General Dynamics/Astronautics from 1958 until
January 1962, when he became the director of Advanced Studies.21
The Thin-Skinned Atlas
At Convair, Ehricke found an innovative engineering culture that had evolved under the tech-
nical direction of Charlie Bossart, an unassuming Belgian migr. The father of the Atlas was born
in Antwerp. He had earned a degree in mining engineering from the Free University of Brussels in
1925. During his last year at the university, an optional course in aeronautical engineering earned him
a scholarship to study aeronautical engineering at MIT. After receiving a masters degree in 1927, he
returned to Belgium for military service, then emigrated to the United States in 1930. He worked for
Sikorsky at General Aviation in aircraft structures before moving to Downey, California, to work for
Vultee Aircraft in 1941. The company became Consolidated Vultee, or Convair, in 1943.
Bossarts background in structures gave him a different perspective from which to approach rocket
design. He conceived the innovative pressure-stabilized design of the Atlas while working on an early
Air Force missile project in 1946 called the MX-774. Like many postwar rocket designers, Bossarts
19
Interview with Krafft Ehricke by John Sloop, 26 April 1974, NASA Historical Reference Collection. See also John
Sloops Liquid Hydrogen as a Propulsion Fuel: 19451959, 191195.
20
Krafft A. Ehricke, Comparison of Propellants and Working Fluids for Rocket Propulsion, American Rocket Society 23
(1953): 287296, 300.
Biography introducing article in Astronautics 2 (1957): 26. See also interview with Krafft Ehricke by John Sloop, 26
21
April 1974, NASA Historical Reference Collection. In 1957, Convair became the Convair/Astronautics Division of
General Dynamics; in 1961, the name was changed to General Dynamics/Astronautics, a Division of the General
Dynamics Corporation.
31276-chapter 1 3/8/05 10:49 AM Page 8
group had taken the V-2 as their starting point. However, unlike the rocket team that worked under
von Braun at White Sands, they were free of Peenemndes traditions and prejudices.
Bossarts design dispensed with the heavy rings and stringers of the V-2. Ring and stringers
were metal bands connected by longitudinal metal rods that reinforced the inside walls of the
rocket. This rigidity provided an extra margin of safety, but it also added unnecessary weight. The
originality of the MX-774 lay in its lack of the internal buttressing. Pressure alone kept the sides
of the rockets tank rigid when inflated like a balloon with its fuel. Its stainless-steel walls were no
thicker than a dime. Although it was not an entirely new idea, since Hermann Oberth in Germany
had discussed its theoretical possibility, Bossart was the first to demonstrate the feasibility of a
pressure-stabilized rocket design. Bossart always attributed his confidence in the thin-skinned
design to his background in aircraft structures. He wrote:
You see most of my previous experience was in structures. So the first thing I did was to
decide what kind of a structure were going to use for this missile. Well, we knew that we
had to have a certain pressure in the tank to maintain the required net positive suction
head for the pump. So the first question was, how thin does the skin have to be to resist
that pressure? With the skin thickness thus arrived at and assuming the tank to be under
pressure, how much stiffening would you have to add to make this thing capable of
taking the compression and the bending anticipated in flight? And lo-and-behold, we
didnt need to add any.22
Even after the Air Force canceled the MX-774 project in 1947, Bossart continued to push
development of the rocket with missionary zeal. His tenacity paid off in the early 1950s after tech-
nical advances allowed thermonuclear warheads to be scaled down to fit inside the nose of a
missile. The rocket now won the backing of two influential Air Force officials, Trevor Gardner,
Assistant Secretary of the Air Force for Research and Development, and Lieutenant Colonel
Bernard Schriever.23 Concerned that the United States was in a race with the Soviet Union in
ICBM development, Gardner and Schriever advocated setting up a crash program comparable in
scale and funding to the Manhattan project.
22
Interview with K. J. Bossart by John Sloop, 27 April 1974, NASA Historical Reference Collection, Washington, DC. For a
superb article on Bossart and the balloon tank, see Richard E. Martin, The Atlas and Centaur Steel Balloon Tanks: A Legacy
of Karel Bossart, reprint by General Dynamics Corp., 40th International Astronautical Congress paper, IAA-89-738, 1989.
23
See Frank L. Winter, Rockets into Space (Cambridge, MA: Harvard University Press, 1990); Edmund Beard, Developing
the ICBM: A Study in Bureaucratic Politics (NY: Columbia University Press, 1976); Jacob Neufeld, Ballistic Missiles in the
United States Air Force, 19451960 (Washington, DC: Office of Air Force History, 1990); John Clayton Lonnquest, The
Face of Atlas: General Bernard Schriever and the Development of the Atlas Intercontinental Ballistic Missile, 19531960,
Ph.D. dissertation, Duke University, 1996. See also Davis Dyer, Necessity Is the Mother of Convention: Developing the
ICBM, 19541958, Business and Economic History 22 (1993): 194209; and TRW: Pioneering Technology and Innovation
since 1900 (Boston: Harvard Business School Press, 1998), 167194. A review of Atlas development to the present can be
found in Dennis R. Jenkins, Stage-and-a-Half, The Atlas Launch Vehicle, in To Reach the High Frontier: A History of
U.S. Launch Vehicles, eds. Roger D. Launius and Dennis R. Jenkins (Lexington: The University of Kentucky Press, 2002),
70102.
31276-chapter 1 3/8/05 10:49 AM Page 9
Charlie Bossart, designer of Atlas, holds up a sketch of the missile, early 1960s. (Courtesy of Lockheed Martin)
31276-chapter 1 3/8/05 10:49 AM Page 10
A blue ribbon committee of scientific, industrial, and academic leaders headed by Princeton
physicist John von Neumann endorsed the development of a long-range ballistic missile. The
committee, unconvinced that the Air Force had sufficient technical depth to manage the project,
recommended that the Ramo-Wooldridge Corporation assume the responsibility for technical
direction of the program, reporting directly to Schriever. Prior to approval of the design in January
1955, Ramo-Wooldridge put pressure on Bossart to abandon the pressure-stabilized design
because of the tendency of the bulkhead to collapse if the tanks were not pressurized correctly. To
Bossarts relief, Ramo-Wooldridge agreed that development was too far along to change to the type
of reinforced structure favored by most rocket designers.24
Engineers at Convair were intensely proud of the originality of the Atlas design, which they
liked to call the free worlds first ICBM. However, not all rocket experts celebrated the success
of Atlas. The group at Huntsville looked askance at such lightweight structural innovations as
Bossarts thin-wall, pressurized tanks for the Atlas ICBM, which they jokingly referred to as
blimp or inflated competition.25
During Atlas development, the company built launch complexes at Cape Canaveral and static
firing test facilities at Sycamore Canyon, northeast of San Diego. It also carried out a research
program at Edwards Air Force Base near Boron, California. Originally, the group working on Atlas
was based in the Lindbergh Field plant located next to the San Diego airport. As it grew from about
20 employees in 1951 to 200 in 1955, the company began construction on a large plant about ten
miles north of downtown San Diego at Kearny Mesa. Manufacture of the technically demanding
stainless steel tanks for Atlas (and later the even more stringent requirements for the Centaur
tanks) initially was located at Kearny Mesa, but several years later was moved to Plant 19, owned
by the Air Force, near the Lindbergh Field plant.
In 1957, to strengthen management, General Dynamics Convair set up a separate
Astronautics Division under James Dempsey, an Alabamian with a flair for handling both
administrative and technical problems. A graduate of West Point, Dempsey had served in
World War II. After the war, he earned a masters degree in aeronautical engineering at the Air
Force Guided Missile School at the University of Michigan, served at the Pentagon and Patrick
Air Force Base near Cape Canaveral, then accepted a position with the Convair Corporation
in 1953.
24
Interview with K. J. Bossart by John Sloop, 27 April 1974, NASA Historical Reference Collection. On the Air Force
approach to contracting, see also Joan Lisa Brombergs chapter, Legacies, in NASA and the Space Industry (Baltimore:
The Johns Hopkins University Press, 1999), 1644.
25
John Sloop, Liquid Hydrogen, 208. See also G. R. Richards and Joel W. Powell, The Centaur Vehicle, Journal of the
British Interplanetary Society 42 (1989): 99120. Missile concepts pioneered on Atlas included the stage and a half
launch concept. After firing all three engines at liftoff, two were jettisoned after 140 seconds, leaving the main engine to
power the vehicle toward its target. Bossart introduced the idea of a separable nose cone to save weight and facilitate
reentry. Other innovations included a radio-inertial guidance system, which increased the accuracy of the booster,
swiveling rocket motors, and small vernier motors for greater accuracy and velocity control.
31276-chapter 1 3/8/05 10:49 AM Page 11
With national security riding on the successful development of the Atlas, pressure on the
Astronautics Division was intense. Atlas failed during the first test on 11 June 1957, although the
controversial balloon structure was validated during its dramatic tumble earthward. Despite the
extreme lateral stresses placed on the tank, it did not break up until it was destroyed by the range
safety officer. The launch of Sputnik the following October struck a devastating blow to the
countrys sense of security.
Looking back on NASAs early years, Administrator T. Keith Glennan admitted, In truth, we
lacked a rocket-powered launch vehicle that could come anywhere near the one possessed by the
Soviets. And it would take years to achieve such a system, no matter how much money we spent.26
Without vehicles with adequate thrust, the space program could achieve none of its goals, which
included an ambitious program of planetary exploration. Moreover, the inability to launch even a
relatively light satellite seemed to indicate that the United States lacked a sufficiently powerful missile
to retaliate in the event of attack by the Soviet Union.
In the politically charged post-Sputnik atmosphere, ARPA Director Roy Johnson asked Jim
Dempsey whether the company was capable of launching anything resembling a spacecraft.
Recalling Ehrickes early studies of spacecraft, Dempsey bravely replied, Sure, the whole Atlas.27
This led to a top-secret crash program called Project SCORE (Signal Communication by Orbiting
Relay Equipment). Assured that the Atlas would not fail, Glennan convinced President Eisenhower
to use Project SCORE as a not-so-subtle proxy to demonstrate the countrys military prowess.
Glennan noted in his diary: The recent successes of the Atlas launches would permit him [the
President] to portray dramaticallypossibly by television or a demonstration for the pressthe
effectiveness of the Atlas as a nuclear warhead carrier.28 The Atlas, launched from Cape Canaveral
on 18 December 1958 with a pretaped Christmas message of peace on Earth from President
Eisenhower, carried the largest American payload placed in orbit up to that time.29
The increasing reliability of the Atlas encouraged NASA officials to press it into service for
the Mercury program, on the theory that to catch up with the Soviet Union, NASA should use
known technologies, extending the state of the art as little as necessary.30 Atlas would launch
Mercury astronauts John Glenn, M. Scott Carpenter, Walter Schirra, and L. Gordon Cooper, but
without an upper stage, it was consigned to low-Earth orbit.
26
See Glennans comments on this issue in The Birth of NASA: The Diary of T. Keith Glennan, ed. J. D. Hunley
(Washington, DC: NASA SP-4105, 1993), 23.
27
William H. Patterson, The Evolution of a New Technology Concept in the USA (Americas First ICBMthe Atlas),
typescript, no date, 19. Copy courtesy of Charles Wilson.
28
J. D. Hunley, ed., The Birth of NASA, 24.
29
Deane Davis, The Talking Satellite: A Reminiscence of Project Score, Journal of the British Interplanetary Society 52
(1999): 239259. See also John L. Chapman, Atlas: The Story of a Missile (New York: Harper & Brothers, 1960),
152153.
30
J. D. Hunley, ed., The Birth of NASA, 13.
31276-chapter 1 3/8/05 10:49 AM Page 12
Abe Silverstein, Chief of Research at Lewis Flight Propulsion Laboratory, 1950s. (C1963-63846)
For Dempsey, Centaur was the key to securing the companys place in the emerging aerospace
industry. Though the Atlas missile became operational in 1959, it had severe limitations as a weapons
system. The company anticipated its replacement in the countrys arsenal with the Minuteman solid-
propellant missile and the much larger Titan II. Atlas, fitted out with a new high-energy upper stage,
later proved a commercial success for the company. However, development of Centaur into a
dependable launch vehicle would be neither inexpensive nor easy. The difficulties of developing a
liquid-hydrogen-fueled upper stage would prove a shocking revelation to the talented engineering
team at General Dynamics.
The General Dynamics proposal, A Satellite and Space Development Program, was ready
two months after Sputnik. The group pitched the idea to Dempsey, who immediately packed
Ehricke and William Patterson off on a sales mission to Washington, District of Columbia. They
31276-chapter 1 3/8/05 10:49 AM Page 13
submitted Krafft Ehrickes first Centaur proposal to the Air Force in December 1957. The
company asked for $15 million to begin work on the high-energy upper stage.31
Although the Air Force did not accept the first proposal, Ehricke called on Abe Silverstein at
NACA Headquarters during the following June. Well known in the nations aerospace community,
Silverstein was among the nations most knowledgeable experts in aircraft and rocket propulsion. Born
in 1908 in Terre Haute, Indiana, Silverstein had graduated from Rose Polytechnic Institute in 1929
with a degree in mechanical engineering. After more than a decade in wind-tunnel design and oper-
ations at NACAs Langley Laboratory, he transferred to the new engine research laboratory in
Cleveland during World War II. As Chief of Research for the Lewis Flight Propulsion Laboratory in
Cleveland in the 1950s, he aggressively supported the work of the rocket section and put pressure on
NACA Headquarters for greater research funding for the study of rocket propellants.
Silverstein shared Ehrickes enthusiasm for liquid hydrogen, but his hands were tied.
Congress had not yet passed the Space Act setting up the National Aeronautics and Space
Administration. He suggested that ARPA might be receptive. Ehricke presented the proposal for
a liquid-hydrogen upper stage to Herbert York, who was chief scientist for ARPA at that time.
Because of the well-known difficulties of developing a hydrogen pump, Ehricke had specified a
pressure-fed system with four 7,500-pound-thrust oxygen/hydrogen engines to be designed by
the Rocketdyne Division of North American Aviation.32 At this meeting, he was surprised and
delighted to learn of the Air Forces prior funding for the development of a liquid-hydrogen
airplane for high-altitude reconnaissance.33 At the ARPA meeting, Ehricke was told that the Air
Force was so strongly committed to the development of liquid hydrogen that large liquefiers to
convert hydrogen gas to liquid hydrogen had already been built in various parts of the country.
Early Liquid-Hydrogen Initiatives
Ehricke and the design team at General Dynamics were unaware of earlier liquid-hydrogen
initiatives. Centaur brought together liquid-hydrogen research and development by academia,
the military, and NACA. The development of a national capability in liquid hydrogen began in
the 1920s with the pioneering research of William Giauque and his student Herrick L. Johnston
at the University of California at Berkeley. Their research contributed to the discovery of
deuterium, or heavy hydrogen, in 1931 by Nobel Prize winner Harold Urey. When Johnston
took a position at The Ohio State University (OSU) in the late 1920s, he continued this work
31
Patterson, 20.
32
Centaur Technical Handbook and Log, General Dynamics, 12 March 1963, Old RL10 Records, NASA Glenn
Research Center (GRC) Records.
33
31276-chapter 1 3/8/05 10:49 AM Page 14
in cryogenics. The hydrogen liquefier, or cryostat, he built at Ohio State provided a model for
other cryostats built in the United States after World War II.34 With new methods to convert
hydrogen gas to liquid hydrogen, postwar rocket designers began to consider the tantalizing
possibilities of liquid hydrogen. Not only could liquid hydrogen be stored in a smaller space
than hydrogen gas, but keeping hydrogen in its liquid state until just prior to ignition signifi-
cantly reduced the dangers of handling and storage.
Liquid-hydrogen research found early support from the Department of Defense. The Air Force
sponsored a series of liquid-hydrogen rocket experiments by Johnston and his students in which they
applied the principle of regenerative cooling by circulating liquid hydrogen in passages that surround
the thrust chamber prior to injection and firing of the rocket.35 They also discovered that ball bear-
ings cooled by liquid hydrogen did not require lubricationa phenomenon rediscovered by
engineers at Pratt & Whitney during the development of the RL10 engine.
In the late 1940s, experiments at the Aerojet Engineering Corporation in Azusa, California,
contributed to solving problems associated with liquid-hydrogen supply. With help from Johnstons
team at OSU, Aerojet built a hydrogen liquefier.36 When Aerojet lost Navy support, the work was
continued by the Jet Propulsion Laboratory in Pasadena, where a liquid-hydrogen rocket was tested
for the first time.37 The Navy also contracted with Martin Marietta and North American Aviation
to design a hydrogen/oxygen rocket called the High Altitude Test Vehicle (HATV). This design
featured pressure-stabilized tanks with a common bulkhead separating the two propellants. None
of these projects enjoyed high visibility or lavish funding.38
In the early 1950s, after President Harry Truman authorized the development of a hydrogen
bomb, generous support for basic and applied cryogenics research suddenly became available to
academic, government, and industry researchers.39 In 1952, the U.S. National Bureau of Standards,
under the auspices of the Atomic Energy Commission, set up a Cryogenics Engineering Laboratory
in Boulder, Colorado, to produce liquid hydrogen in larger quantities. The discovery in 1953 of a
catalyst to take free hydrogen, a mixture of one isotope of regular hydrogen and one of deuterium,
and convert it to a para-hydrogen molecule before liquefaction greatly speeded up the conversion
process.40 The Boulder laboratory established a data center and began to publish the results of exper-
34
John Sloop, Liquid Hydrogen, 1315.
35
36
See H. L. Coplen, Large-Scale Production and Handling of Liquid Hydrogen, Journal of the American Rocket Society
22 (NovemberDecember 1952): 309322, 338.
37
38
39
See Richard Rhodes, Dark Sun: The Making of the Hydrogen Bomb (New York: Simon & Schuster, 1995), 487489.
40
See Joseph Green and Fuller C. Jones, The Bugs That Live at -423, Analog Science Fiction/Science Fact, 80 (January
1968): 14.
31276-chapter 1 3/8/05 10:49 AM Page 15
John Sloop presents scope of NACA rocket research, 1947. (NASA G19785)
imental research conducted there. However, production capacity from all hydrogen liquefiers in the
United States was less than one-half ton per day as late as 1956.
An effort by NACA at the Lewis Flight Propulsion Laboratory in Cleveland was among the
important early liquid-hydrogen initiatives that contributed to a national capability in high-energy
fuels. Members of the Rocket section of the laboratory studied the performance of various liquid
propellants and chamber and nozzle cooling.41 After study of German technical papers, as well as
reports produced by Aerojet and the Jet Propulsion Laboratory, they decided to narrow their inves-
tigation to one of the least studied areas, that of high-energy liquid propellants. An important
1948 theoretical study by Paul Ordin and Riley Miller concluded that liquid hydrogen held
promise as a rocket fuel because of its high specific impulse.42
Because of the laboratorys lack of rocket test facilities and its inability to purchase liquid
hydrogen in sufficiently large quantities, the group focused on then-exotic chemicals such as
41
Combustion Research Program, NACA Lewis Laboratory, Active as of September 1949, NASA Lewis Research
Center Records, RG 255, Box 295, NARA, Chicago, IL. The Rocket Section was part of the Combustion Branch of the
Fuels and Lubricants Division.
42
R. O. Miller and P. M. Ordin, Theoretical Performance of Rocket Propellants Containing Hydrogen, Nitrogen, and
Oxygen, NACA RM E8A30 (1948), cited in John Sloop, Liquid Hydrogen, 80.
31276-chapter 1 3/8/05 10:49 AM Page 16
hydrazine with oxidizers like hydrogen peroxide, chlorine trifluoride, liquid oxygen, nitrogen
tetroxide, and liquid fluorine.43 Although a small research effort, this work on rocket fuels laid the
basis for the Cleveland laboratorys rocket expertise.
By 1954, researchers had completed dozens of reports on various rocket fuels and compo-
nents. They were experimenting with full-scale rocket engines and beginning to focus on
hydrogen/oxygen as a promising propellant combination in a 5,000-pound rocket, but
funding was inadequate to carry out extensive testing.
In late 1955, the Cleveland laboratory seized the opportunity to increase its expertise in the
storage, handling, and firing of liquid hydrogen by contributing to an Air Force project to develop
a more advanced spy plane than Lockheeds U-2, which flew at very high altitudes over the Soviet
Union to avoid getting shot down. The low density of the atmosphere at high altitudes limited the
performance of the U-2s engine, which used conventional kerosene-based fuel. Because hydrogen
burned readily at extremely low pressures, it seemed a promising fuel for high-altitude reconnais-
sance aircraft. To investigate the possibilities of liquid-hydrogen fuel, Silverstein and NACA
research engineer Eldon Hall produced a classified NACA technical report later published as
Liquid Hydrogen as a Jet Fuel for High-Altitude Aircraft.44 The report recommended a pressure-
stabilized structure for the design of an airplane fuel tank filled with liquid hydrogen. Whether
the Silverstein-Hall report stimulated the Air Forces keen interest in liquid hydrogen or simply
reinforced it is unclear. Silverstein won funding for a $1-million liquid-hydrogen test program at
Lewis called Project Bee. The Air Force supplied a B-57 bomber and two Curtiss Wright J-35
engines for this research. The laboratory was given one year to determine whether the aircraft
could be adapted to fly with liquid-hydrogen fuel.
Project Bee gave Lewis engineers experience with problems such as the insulation,
instrumentation, and pumping of an engine run on liquid hydrogen. They determined that
even when the fuel tank was insulated, heat leaks caused the temperature of liquid hydrogen
to rise. When it reached the boiling point, the liquid fuel changed to hydrogen gas, causing
the pressure in the tank to rise. Without proper venting, the fuel tank exploded. Too much
venting, however, reduced the fuel available for completing the flight. They also discovered
that helium gas prevented the insulation from freezing to the exterior wall of the tank.45
During the first flight over Lake Erie in December 1956, the pilot switched from JP-4 fuel
to liquid hydrogen twice, but the aircraft failed to maintain its speed. During a subsequent
43
44
NACA RM E55 C28a, 15 April 1955. See John Sloop, Liquid Hydrogen, 98102.
See Experimental Study of Foam-Insulated Liquefied Gas Tanks, by Thaine W. Reynolds and Solomon Weiss, NACA
45
RM-E56K08A.
31276-chapter 1 3/8/05 10:49 AM Page 17
flight, the engine operated for 20 minutes on liquid hydrogen at an altitude of 49,500 feet
and a speed of Mach 0.72.46 By the mid-1950s, the Lewis laboratory had become one of the
acknowledged leaders in cryogenic fuels research, but it lacked adequate facilities for rocket
testing.47
Centaur and Project Suntan
Project Suntan, the code name for a top-secret, high-altitude spy plane, directly contributed
to the development of Centaur.48 During Project Suntan, the Air Force funded the construction
of hydrogen-liquefaction facilities in Trenton, New Jersey (later canceled); Painesville, Ohio; and
Bakersfield, California. Two additional liquefiers were built adjacent to a new Pratt & Whitney
test center in West Palm Beach, Florida. Together, these plants increased the capacity of the United
States to produce liquid hydrogen from 500 pounds a day in 1956 to about 68,000 pounds a day
by 1959. This liquid-hydrogen infrastructure contributed to the national capability to develop a
liquid-hydrogen rocket.
Another important connection between Suntan and Centaur was its Air Force management.
John Seaberg, later one of Centaurs champions, headed Project Suntan. A veteran of the U.S.
Army Air Corps, Seaberg had worked for Chance Vought Aircraft for five years as an aerody-
namicist until the start of the Korean War, when he was recalled to active duty at
Wright-Patterson Air Force Base in Dayton, Ohio. Seaberg conceived and directed the U-2
program. A strong proponent of liquid hydrogen, Seaberg obtained funding for an advanced
spy plane designed by Clarence (Kelly) Johnson of Lockheed in 1956. Pratt & Whitney
Aircraft Engine Company received a contract to design Suntans liquid-hydrogen turbojet
engine. Pratt & Whitneys design team, headed by Richard Coar, completed a prototype for
the innovative 304 turbojet engine in just sixteen months.
In conventional turbojet engines, a small fraction of the propellants is burned in a separate
combustion chamber to begin the combustion process. This requires a separate propellant control
and ignition system, which adds complexity and weight. The Pratt & Whitney engine had an
expander cycle in which liquid hydrogen flowed through tubes into the combustion chamber
walls on the way to the turbine. In the process, the cryogenic propellant, now gaseous, picked up
46
Interview with Paul Ordin by John Sloop, 30 May 1974, NASA Historical Reference Collection. See also John L.
Sloop, NACA High Energy Rocket Propellant Research in the Fifties, AIAA 8th Annual Meeting, 28 October 1971,
NASA Historical Reference Collection.
47
Richard Canright, Rocket Propellants: Views of an Airframe Manufacturer, 31 July 1957, transmitted to members of the
NACA Subcommittee on Rocket Engines with a cover letter by Benson E. Gammon, 21 August 1957, quoted in Sloop, Liquid
Hydrogen, 85.
48
See John Sloop, Liquid Hydrogen, on project Suntan, 141166, and General Discussion by Col. N. C. Appold, Lt.
Col. J. D. Seaberg, Maj. A. J. Gardner, Capt. J. R. Brill, transcribed February 1959, 15, John Sloop Papers, NASA
Historical Reference Collection.
31276-chapter 1 3/8/05 10:49 AM Page 18
heat and expanded, providing energy to turn the turbine and turbopumps; at the same time, it
cooled the thrust chamber walls.
Despite the elegance of the Pratt & Whitney design for the engine, the design team at
Lockheed began to doubt that the proposed altitude and range for the new airplane could be
achieved. Kelly Johnson convinced Donald Quarles, Deputy Secretary of Defense, to cancel the
project. Nevertheless, Lieutenant Colonel Seaberg continued to champion liquid hydrogen. In
June 1958, he convinced ARPA officials to fund a project first known simply as the high-energy
upper stage. Thus, when Ehricke coincidentally presented his plans for a liquid-hydrogen upper
stage for Atlas, the proposal won immediate support. John Sloop commented, Two days after
the ARPA order, Pratt & Whitney conducted the tenth and final series of tests with the
hydrogen fueled 304 turbojet engine. Suntan became a thing of the past and Centaur, a
hydrogen-oxygen rocket stage on top of Atlas, rose as its replacement. All the plant equipment,
and technology of Suntan could now be brought to bear in assuring that Centaur would
succeed.49
Project management was assigned to a special projects office at Wright Field under Colonel
Norman Appold until Abe Silverstein, in charge of the Office of Space Flight Programs for the
newly formed space agency, insisted that NASA take over the program. However, the Air Force
officers assigned to project continued to manage it under NASAs direction. This arrangement
later became a source of friction between the Air Force and Marshall Space Flight Center.
Ehricke learned of Pratt & Whitneys liquid-hydrogen engine at his first meeting with
ARPA officials in the summer of 1958. The innovative expander cycle of the engine proved
a revelation to Ehricke and his colleagues at General Dynamics. Kenneth Newton, Director
of Launch Vehicle Programs, recalled the excitement that the first set of Pratt & Whitney
drawings generated at the company. We were so enthralled with their simplicity, he said.50
The expander cycle (also referred to as a bootstrap cycle) of the Suntan engine would serve
as the basis for the design of the Centaur RL10 engine (originally designated LR115). It was
one of the keys to the striking efficiency and reliability of the RL10. Dick Mulready, second
in command of the RL10 project, wrote:
It is hard to over-emphasize the significance of the cycle choice for the RL10. The
singular physical properties of pure hydrogen, which the cycle exploits, allow very large
margins of safety in design. The simple cycle completely eliminates many complex ancil-
lary systems. This fact and the large design margins that are possible at low temperature
lead to unmatched reliability. Starting is accomplished by opening the propellant supply
49
John Sloop, Liquid Hydrogen, 196. See also Roger Bilstein, Stages to Saturn, 135. The first public announcement of the
project was made by General Bernard Schriever in May 1959. See R. B. Scott, Technology and Uses of Liquid Hydrogen
(New York: The Macmillan Company, 1964), 3.
50
Interview with the Aerospace Division of Convair/General Dynamics by John Sloop, 29 April 1974, NASA Historical
Reference Collection.
31276-chapter 1 3/8/05 10:49 AM Page 19
valves and admitting hydrogen into the thrust chamber cooling tubes. The sensible heat
residual in the tubes, even after many hours of coasting in space, is enough to spin up the
turbine.51
The unusual design of the RL10 engine, based on the unique properties of liquid hydrogen,
prompted Ehricke to propose a bold new design for Centaur to capitalize on the low density and
extreme cold of liquid hydrogen. Ehricke had first envisioned a more complicated design for
Centaur consisting of two large tanksone for liquid hydrogen and the other for liquid oxygen.
Each had a smaller tank within the larger one. The inner tanksintended to supply fuel to restart
the engines in space after a coast periodadded weight and complexity.52
Ehricke simplified the design for Centaur. By adopting Pratt & Whitneys hydrogen pump,
two of the four proposed engines could be eliminated. Rocketdyne, the countrys preeminent
designer of rocket engines, was outraged that the Air Force had awarded a sole-source contract to
Pratt & Whitney, a company without prior experience in the design of a rocket engine.53 The deci-
sion was based on Pratt & Whitneys previous experience with liquid hydrogen and the
propellants ready availability at the Florida test site.
Next, Ehricke eliminated the unnecessary complexity of the prior design. The new Centaur
was remarkably like the thin-walled, pressure-stabilized Atlas in structure. Centaurs two propel-
lants were separated by a double-walled bulkhead, consisting of two thin metal skins separated
by a quarter-inch layer of fiberglass insulation. The internal pressure of the two liquids kept the
outer wall of the tank rigid and pushed against the bulkhead to keep it from collapsing. Prior
to tanking with its cryogenic propellants, the bulkhead was backfilled with dry gaseous
nitrogen. When loaded with liquid oxygen on one side of the bulkhead and liquid hydrogen on
the other, the liquid hydrogen froze out the nitrogen to produce a near-perfect vacuum within
the hollow bulkhead. This thermal barrier greatly reduced heat transfer between the liquid
oxygen (-299F) and the much colder liquid hydrogen (-423F). Pratt & Whitneys expander
cycle and Ehrickes double-walled bulkhead design were both logical extensions of the special
properties of liquid hydrogen. Ehricke said:
Using this design, we achieve a very high mass ratio for a vehicle of this size with low-
density propellants. The membrane bulkhead design uses cryogenic vacuum formation as
the logical utilization of an existing extreme temperature environment; just like Pratt &
Dick Mulready, Advanced Engine Development at Pratt & Whitney, 60. See also Joel E. Tucker, The History of the RL10
51
Upper-Stage Rocket Engine, 126.

52
Reference Collection. Ehricke also mentioned the earlier design in his 1962 congressional testimony. Neither the original
proposal nor a blueprint of the design is available to confirm inclusion of a pressure-stabilized structure.
53
Robert S. Kraemer, Beyond the Moon: A Golden Age of Planetary Exploration 19711978 (Washington: Smithsonian
Institution Press, 2000), 25.
31276-chapter 1 3/8/05 10:49 AM Page 20
Whitneys bootstrap method of engine start [i.e., use of residual heat in the thrust
chamber cooling tubes to spin the turbine] was an ingenious and logical utilization of the
special physical characteristics of liquid hydrogen.54
The double-walled bulkhead, created late one night with the help of the companys thermody-
namicist, William Mitchell, was arguably the most fundamental part of the design. We had that
integral bulkhead before we even knew how many engines we had on the thing or how long the thing
was going to be. That was the kind of invention that sprung the whole thing, Deane Davis, Centaur
project engineer, recalled.55 The bulkhead restricted heat transfer between the relatively warmer liquid
oxygen and the liquid hydrogen. The double-walled bulkhead would not only be the most innovative
feature of Centaur, but also the most technically demanding. Engineers would later discover that
liquid hydrogen leaked through minute pores in the welds of the bulkhead.56
Another innovative feature involved putting the thrust structure, which distributes the
concentrated thrust load, inside the liquid-oxygen tank. This saved weight by shortening the
vehicle and reducing bending loads. The exterior of the liquid-hydrogen tank was protected by
four quarter-section fiberglass insulation panels. The insulation shielded the liquid hydrogen
from aerodynamic heating, minimizing fuel boil-off. The panels remained bolted together
during the early portion of flight. As soon as the aerodynamic heating rate was no longer crit-
ical, shaped explosive charges were fired to separate them from the vehicle. The design also
included explosive bolts to separate the two halves of the nose fairing. Small nitrogen gas
thrusters would propel the two sections of the fairing away from the vehicle.57
To make sure that nothing interfered with the General Dynamics focus on Atlas, ARPA
funded Centaur as an experimental program without a mission assigned to it. Ehrickes new
proposal, accepted in August 1958, had modest goals. Considered a bridge between Atlas-Agena
and much larger future boosters, Centaur was intended to prove the feasibility of a liquid-
hydrogen/liquid-oxygen rocket. The guidance system, test facilities, and launch complex were not
included in the bare-bones contract. Development was proposed to take twenty-five months, with
the first flight to occur in January 1961. Cost was projected at $36 million for six vehicles.
To keep down costs and prevent interference with the Atlas program, ARPA insisted that
General Dynamics use Atlas hardware wherever possible.58 A member of the design team recalled,
U.S. House of Representatives Subcommittee on Space Sciences of the Committee on Science and Astronautics, The
54
Centaur Program, 87th Congress, 15 and 18 May 1962, 67.

55
Interview with the Aerospace Division of Convair/General Dynamics by John Sloop, 29 April 1974.
Richard E. Martin, The Atlas and Centaur Steel Balloon Tanks: A Legacy of Karel Bossart, reprint by General
56
Dynamics Corp., 40th International Astronautical Congress paper, IAA-89-738, 1989, 11.
Description from Centaur Capability Handbook, GD/A-BTD64-119-1, 1 October 1964, Kennedy Space Center ELV
57
Resource Library.
58
Centaur Program, 1962 Hearings, 66.
31276-chapter 1 3/8/05 10:49 AM Page 21
Centaur is raised into the J Tower for testing at Point Loma, early 1960s. (Courtesy of Lockheed Martin)
31276-chapter 1 3/8/05 10:49 AM Page 22
One of the ground rules to us was to use Atlas tooling wherever possible, and so really the
Centaur was, and still is, bits and pieces of Atlas put together.59 These design constraints some-
times interfered with achieving an optimal design.
Although a NASA project, Centaur continued to be managed by the Air Force. Ehrickes
monthly reports indicated the growing complexity of the project. Program costs increased to $42
million with the addition of a guidance system and construction of additional test facilities. By
June 1959, the company had awarded a subcontract for the inertial guidance system to
Minneapolis-Honeywell. In July, General Dynamics signed a contract for construction of a launch
complex, designated Pad 36, at Cape Canaveral. In August, it completed construction of a flow
test facility at Point Loma, located on a peninsula on the Pacific coast north of San Diego. A static
test facility in Sycamore Canyon was activated in September shortly before the first test. The
Centaur team then initiated liquid-hydrogen tanking tests at Point Loma.
Glennan took steps to prevent any disruption of Centaur development when it was trans-
ferred from ARPA to NASA in July 1959. He urged that the project office be located in
Washington, but he was overruled by officials of the Department of Defense who insisted
that Centaur management be coordinated by the Air Force from its Ballistic Missile Division
at Inglewood, near Los Angeles, California.
Because the United States had no launch vehicle powerful enough to lift the heavy satellites
needed for a global military communications network, Centaur became one of the top priorities of
the military. Herbert York urged Glennan to expedite the program because it was needed for
launching the Advent series of communications satellites assigned to Centaur in December 1959.60
A secret five-year $174-million research and development program, Advent included ten
planned launches of communications satellites for the Army, Air Force, and Navy. ARPA was
counting on the 1,200-pound satellites to provide instant point-to-point transmissions to
receiving stations within the line-of-sight beam of the satellite. The mission required
restarting the Centaur engines in space. After insertion into a nearly 1-hour parking orbit,
Centaur would position the satellite for an elliptical 5-hour coast orbit. The final restart
would then place the spacecraft in a 24-hour circular orbit. Since Centaur was not yet ready,
the first three satellites were supposed to be launched with the far less powerful Agena B as
the upper stage for Atlas. The Air Force planned to launch the remaining seven Advent satel-
59
60
Herbert York to T. Keith Glennan, 19 September 1959, Centaur files, NASA Historical Reference Collection. On
Advent, see David N. Spires and Rick W. Sturdevant, From Advent to Milstar: The U.S. Air Force and the Challenges
of Military Satellite Communications, in Beyond the Ionosphere: Fifty years of Satellite Communication, ed. Andrew J.
Butrica (Washington, DC: NASA SP-4217, 1997), 65.
31276-chapter 1 3/8/05 10:49 AM Page 23
lites in 1963 on Atlas-Centaur. Centaurs development problems made meeting the demands
of this schedule impossible.61
By December 1959, General Dynamics had completed a metal prototype and moved it to the
final assembly area, where a mockup of the electrical and pumping systems could be added. At the
same time, the company began a flight program at Edwards Air Force Base to investigate the
behavior of liquid hydrogen under zero-gravity conditions using an Air Force KC-135. Ehricke opti-
mistically predicted that the first Centaur flight article would be delivered in January 1961.62
Centaur and Saturn
In addition to the assignment of Advent to Centaur, another key decision of the early space
program took place in December 1959. Six months before the official transfer of the von Braun
team to NASA, Abe Silverstein chaired a committee to evaluate the Saturn vehicle and settle the
question of propellants for Saturns upper stages.63 Silverstein occupied one of NASAs most impor-
tant posts during NASAs formative years. His Office of Space Flight Programs was responsible for
both human and robotic exploration of space. This office set the agenda for planetary exploration
and completed preliminary planning for a lunar landing well before it became a goal of the
Kennedy administration.
Silverstein had been determined to prove the competence of former NACA engineers to run
the new agency when he accepted the job at NACA Headquarters in 1957.64 He personally
recruited many of the NASA Headquarters staff, bringing a large contingent from Lewis Research
Center in Cleveland. Because the inadequacy of the nations launch vehicles was a source of
concern at all levels of government, Silverstein championed Centaur and called it the kind of
thing upon which our whole future technology, I think, reststhat is, the development of an early
capability with these high-energy propellants.65 Based on Silversteins advocacy, Congress author-
61
Military Space Projects: Report of Progress for JuneJuly 1960, Office of the Director of Defense Research and
Engineering, Office of the Staff Secretary: Records of Paul T. Carroll et al., 19521961, Department of Defense
Subseries, Box 9, Missiles and Satellites, Military Space Projects [JuneAugust 1960], Dwight D. Eisenhower Library.
The authors thank Dwayne Day for providing a copy of this document.
62
Centaur Progress Reports, 31 December 1958 through 31 July 1960, San Diego Aerospace Museum Archives.
63
See John Sloop, Liquid Hydrogen, 230243. Other members included Abraham Hyatt, NASA; George P. Sutton,
ARPA; T. C. Muse, Office of the Director of Defense Research and Engineering (ODDR&E); Norman C. Appold, U.S.
Air Force; and Wernher von Braun, Army Ballistic Missile Agency (ABMA). Eldon Hall, Silversteins close colleague at
NASA, served as secretary. See Report to the Administrator on Saturn Development Plan by Saturn Vehicle Team, 15
December 1959, Silverstein file, NASA Historical Reference Collection. The text of the report is partially reproduced in
John M. Logsdon et al., Exploring the Unknown, Volume IV: Accessing Space, document 1-29, 116119.
64
Homer E. Newell, Beyond the Atmosphere: Early Years of Space Science (NASA SP-4211, 1980), 100.
65
U.S. House of Representatives, 1960 NASA Authorization, Hearings Before the Committee on Science and
Astronautics, 86th Congress, H.R. 6512, April and May 1959, 391.
31276-chapter 1 3/8/05 10:49 AM Page 24
ized additional funding and increased the contract with General Dynamics from six to ten vehi-
clesthis being considered an absolute minimum to prove out a vehicle design upon which so
much of the national space program was beginning to depend.66
Because of Silversteins strong faith in liquid hydrogen, he insisted that Saturn include at
least one liquid-hydrogen upper stage. The team, headed by Wernher von Braun, was adamantly
opposed to liquid hydrogen. At Kummersdorf in the 1930s, von Braun had witnessed the
demonstration of a small liquid-hydrogen engine designed by Walter Thiel and noted that the
greatest impression he retained was of the numerous line leaks and difficulties of handling
liquid hydrogen.67
At a series of key meetings of the Silverstein committee, von Braun urged the use of conven-
tional kerosene-based fuels. In Silversteins view, The von Braun team was apparently willing to
take on the difficulties of the 1.5 million-pound thrust-booster stage rather than the hazards which
they contemplated in the use of hydrogen as fuel.68 Silverstein used data prepared by NACA
colleague Eldon Hall to beat down opposition to liquid hydrogen. For a week, the NASA team
and the von Braun group debated, until von Braun, to the astonishment of the other members
of his team, capitulated. The Saturn Launch Vehicle task team was unanimous in its endorse-
ment of liquid hydrogen/liquid oxygen in all Saturn upper stages. Their report concluded,
Current success in the Centaur engine program substantiates the choice of hydrogen and oxygen
for the high-energy propellants.69 To Oswald Lange, project director for Saturn at Marshall, the
recommendations of the Silverstein committee for liquid-hydrogen upper stages represented a
major milestone in the Saturn program.70 The Saturn C-1 would use conventional fuel in its first
stage but have liquid-hydrogen upper stages. The second stage (designated S-IV) was to use four
RL10 engines uprated to 20,000 pounds thrust each. The proposed C-1s third stage would be the
Centaur itself with two RL10 engines, each with 15,000 pounds of thrust. Silverstein said in a
1977 interview that he did not know exactly what led von Braun to accept liquid hydrogen, but
in hindsight it had proved to be the correct action. I believe that the decision to go with
66
Centaur Program, 1962 hearings, 7. According to NASA Pocket Statistics for December 1961, research and develop-
ment (R&D) funding for Centaur increased from $4 million in FY 1959 to $36.64 million in FY 1960 and to $62.58
million in FY 1961.
67
68
Abe Silverstein, How It All Began, speech at Kennedy Space Center Writers Conference, Cocoa Beach, Florida, 13
September 1977, quoted by Virginia Dawson, Engines and Innovation, 167.
69
Report on Saturn Development Plan B Saturn Vehicle Team, 15 December 1959, Silverstein file, NASA Historical
70
Oswald H. Lange, Development of the Saturn Space Carrier Vehicle, in Ernst Stuhlinger et al., Astronautical
Engineering and Science from Peenemnde to Planetary Space (New York: McGraw-Hill Book Company, Inc., 1963), 5.
See also the letter from Abraham Hyatt to Thomas O. Paine, 24 November 1969, Silverstein family papers. NASA Glenn
History Office.
31276-chapter 1 3/8/05 10:49 AM Page 25
hydrogen-oxygen in the upper stages of the Saturn V was the significant technical decision that
enabled the United States to achieve the first manned lunar landing, he said. The Russian effort
to accomplish this mission without high-energy upper stages was doomed to failure.71
The decision of the Silverstein committee directly influenced the fortunes of Centaur. Centaur
moved into NASAs mainstream as a stepping-stone to a piloted lunar landing. However, the von
Braun team, although convinced of the superiority of liquid-hydrogen upper stages, was never fully
committed to Centaur. Referring to the new designation of Centaur as the final stage (or S-V stage)
for the Saturn C-1, Krafft Ehricke joked that on top of Saturns horrible tower, at least there would
be one decent vehicle. He referred to the Saturn C-1 as the Centaur-tipped Saturn, to which
von Braun retorted, Centaur-tipped Saturn, Hell!72
Von Braun favored large, heavy structures to ensure reliability. His conservative design
philosophy, which Ehricke derided as comparable to the Brooklyn Bridge in its conservatism,
mitigated against the use of liquid hydrogen which, more than conventional fuels, depended
upon very light structures to help offset the handicap of low density.73 Though von Braun
agreed to Saturn stages powered by liquid hydrogen, he apparently did not approve of
Centaurs balloon structure. His opposition may have influenced the NASA decision to award
the contract for the S-IV second stage to the Douglas Aircraft Company, ostensibly to provide
some semblance of competition among aerospace companies.74 The S-IV second stage, with
its cluster of six RL10 engines, produced an additional 90,000 pounds of thrust to the
1,296,000 pounds of thrust generated by the first stage. The choice of Douglas was controver-
sial enough to merit an investigation by the GAO, since Douglas had no prior experience with
liquid hydrogen.75 Douglas avoided the balloon structure by adopting a honeycomb design
similar to that of the Douglas Thor missile.
Caught Between DOD and NASA
From July 1960 to early 1962, the Air Force continued to manage both the Centaur and the Agena
programs, reporting to Marshall Space Flight Center. D. L. Forsythe, the Agena program chief,
71
See note 68.
72
73
John Sloop, Liquid Hydrogen, 208. See also G. R. Richards and Joel W. Powell, The Centaur Vehicle, Journal of the
British Interplanetary Society 42 (1989): 99120.
74
J. D. Hunley, The Birth of NASA, 129. See also Ray A. Williamson, The Biggest of Them All, Reconsidering the
Saturn V, in To Reach the High Frontier, 304.
75
Ray Williamson, Access to Space: Steps to the Saturn V, in John M. Logsdon et al., Exploring the Unknown, Volume
IV, Accessing Space (Washington, DC: NASA SP-4407): 131.
31276-chapter 1 3/8/05 10:49 AM Page 26
lamented the lack of Marshall attentiveness to both Agena and Centaur, which he said received only
a few crumbs which have fallen from the banquet table of thought and effort at MSFC.76
NASA had acquired the Agena rocket, manufactured by the Lockheed Missiles and Space
Company, in early 1960. The engine, manufactured by Bell Aerospace Company, burned unsym-
metrical dimethylhydrazine with inhibited red fuming nitric acid as the oxidizer. As a second
stage for Atlas, Agena could launch only about 500 to 750 pounds into an Earth-escape trajec-
tory. Although the Air Force needed Agena to launch the early Advent satellites, the Agena
program proved an unwanted distraction to Marshall engineers whose energies were consumed
by Saturn development.
Although Marshall had little interest in Centaur, the RL10 engine figured prominently in its
plans because von Braun was counting on an uprated version of the engine for the Saturn S-IV
liquid-hydrogen upper stage. In February 1960, he began to lobby to take over RL10 engine
management from the Air Force. This made sense to A. O. Tischler, NASAs Chief of Liquid Fuel
Rocket Engines, because the Air Force seemed incapable of keeping development costs under
control. The RL10 engine already had a project overrun of about $10 million. The danger in
allowing Marshall to take over the RL10 engine contract was that the engine for Centaur might be
neglected in the interest of the hydrogen engine for the Saturn stage. He urged that Headquarters
management be strengthened in order to minimize project favoritism.77
Major General Don Richard Ostrander, Director of NASAs launch vehicle programs between
1959 and 1961, had similar concerns. He told von Braun that he was concerned that the high cost
of RL10 engine development would draw funds away from other parts of the Saturn program
because of the Pratt & Whitney reputation for high costs. But he admitted that Pratt & Whitney
produced engines of unusually high quality.78 Immediately after Marshall won its bid to take over
RL10 development, it contracted with Pratt & Whitney for a more powerful version of the RL10
called the LR-119. The Douglas S-IV stage specified a four-engine cluster producing 70,000
pounds of thrust. Later, the configuration was changed to a six-engine cluster producing 90,000
pounds of thrust.79
Centaurs prominence within NASA increased when it was assigned two important plan-
etary and lunar exploration programs, Mariner in 1959 and Surveyor in 1961. Engineers at
the Jet Propulsion Laboratory in Pasadena, California, who managed both Surveyor and
Mariner, were skeptical of Centaurs vaunted advantages over more conventional propulsion
76
Edward Clinton Ezell and Linda Neuman Ezell, On Mars: Exploration of the Red Planet (NASA SP-4212, 1984), 47.
77
A. O. Tischler to Abe Hyatt, 5 February 1960, John Sloop papers, NASA Historical Reference Collection.
78
Don R. Ostrander to Wernher von Braun, 20 June 1960, John Sloop papers, NASA Historical Reference Collection.
79
See Joel E. Tucker, The History of the RL 10 Upper-Stage rocket Engine, 19561980, published in History of Liquid
Rocket Engine Development in the United States, 19551980, ed. Stephen E. Doyle, AAS History Series, vol. 13 (San
Diego: AAS, 1992), 133. See also Mulready, 78; Bilstein, 131140, 188190.
31276-chapter 1 3/8/05 10:49 AM Page 27
systems. Their opposition to Centaur would grow as Centaurs development problems became
increasingly apparent.
NASA set up a Centaur Project Technical Team under Colonel Don Heaton to facilitate
technical support of NASA missions, although Centaur still remained under Air Force manage-
ment because of its key role in the Advent Program for the Department of Defense. When
Milton Rosen took over as NASAs Chief of Rocket Vehicle Development, pressure on Ehricke
increased. Rosen, who had served as project engineer for the Vanguard program, was not a fan
of Centaur. He demanded that General Dynamics develop Vega as an upper stage for the Atlas.
The Vega program, later canceled, siphoned off personnel and funds needed for the more tech-
nically challenging Centaur.80
General Dynamics began to build a flight-qualified Centaur, with the first launch now
projected for June 1961. Construction was initiated on the Centaur launch complex at the
Atlantic Missile Range at Cape Canaveral. A test program at the Tullahoma Vacuum Test Facility
at Arnold Engineering Development Center, additional nose-fairing development, and a zero-
gravity flight-test program with Aerobee rockets at Wallops Flight Research Station all added to
the increasing cost and complexity of the program.
NASA relied on Lewis Research Centers expertise in problems associated with the behavior
of hydrogen in zero-gravity. The urgency of the program made it necessary to place fifty-four
people at Lewis on a 48-hour week for a period of six months. The Center pressed the Propulsion
Systems Laboratory into service for tank insulation and pressurization tests. Testing brought the
disquieting knowledge that Centaurs intermediate bulkhead would collapse without the main-
tenance of a 3-pound pressure differential.
The Advent requirement that Centaurs engines be restarted twice after two coast periods
posed questions for which there were not yet answers. During both periods, the longitudinal axis
of the vehicle would be pointed at the Sun with the oxygen tank exposed to the Suns heat. Also,
depending on the time of year, Centaur might encounter other problems, such as radiation from
the Earth. How would the propellants behave in the extreme environment of space, particularly
when Centaur tanks were only partially filled? Would liquid hydrogen continue to completely wet
the tank walls under near-zero-gravity conditions? Would the pressure buildup cause excessive
venting, resulting in the loss of sufficient fuel for the second burn? Everyone involved in the space
programboth NASA engineers and the people working under Krafft Ehricke at General
Dynamicswere attempting to pierce the veil of ignorance surrounding the problems to be encoun-
tered in the low gravity of outer space.
One of the obstacles to piercing that veil was the design constraint of using Atlas tooling
rather than designing Centaur from scratch. Fred Merino, a scientist at General Dynamics with a
background in thermodynamics, recalls going to the design group headed by Daniel Heald.
80
Interview with Donald Lesney (former General Dynamics employee) by Virginia Dawson, 5 June 2000.
31276-chapter 1 3/8/05 10:49 AM Page 28
Merino asked the group to design a special valve system for venting hydrogen gas during the two
coast periods. They refused until Merino insisted that without proper venting the rocket would
fail. He recalled:
The designer responsible for some of this hardware said, You cant do it because the
ground rules are we use off-the-shelf hardware and they said that means we are going to
take this Atlas valve, scale it down a little bit, mount the same valve on the liquid
hydrogen tank as the liquid oxygen tank, and were going to command the valve to open
when you have to vent. Thats all were going to do. And I said, You will never fly a
mission that way. Its going to fail. The guy said, Those are my marching orders. So I
had to write a report explaining why upper management had to alter their design guide-
lines. They agreed when they heard why you cant do this.81
Denny Huber, an engineer hired by General Dynamics in 1960, recalled that the same bare-
bones approach to Centaur tooling meant they had to use the Atlas liquid-oxygen propellant loading
system. Although the new hydrogen system they had designed worked beautifully, the old oxygen
system was inadequate. They had to wait until a second pad was built to redesign the system. Then
Pad 36A was retrofitted with the new technology.82
Another question centered on how to insulate the Centaur liquid-hydrogen tank. Insulation
was necessary to keep the liquid hydrogen from vaporizing on the ground and during launch. The
insulation was applied in four segments running the length of the tank and was held against the
tank by metal bands. At about ten times the speed of sound, explosive bolts on the bands were
supposed to fire, allowing the insulation to fall away from the vehicle.83 Lewis engineers were
skeptical that the proposed insulation would work. A March 1961 memo by Lewis engineer
Vern Gray stated, The troubles with the insulation are delaying the test program, are poten-
tially hazardous, and illustrate Convairs apparent indifference to the peculiar problems of
hydrogen tank insulation. The memo described how hard it was to cajole test data out of the
company. Gray observed on a visit to San Diego that the application of the insulation was
uneven, leaving lumps. During tests, the insulation split open when the extreme cold of the tank
wall came into contact with the air. The loss of insulation increased the rate of hydrogen boil-
off, raising pressure in the tank. There was danger that the tank might rupture, spilling volatile
liquid hydrogen. The effort to fix the problem with adhesive tape appalled Gray. He commented,
81
Interview with Frederick Merino (former General Dynamics employee) by Virginia Dawson, 5 June 2000.
82
E-mail communication from Denny Huber to Ed Bock, 30 March 2002.
83
Memo from Irving Pinkel to Colonel D. H. Heaton, 25 January 1960, regarding Meeting on Centaur insulation panel
flutter, 22 January 1960, RG 255, NARA Box 254.
31276-chapter 1 3/8/05 10:49 AM Page 29
Apparently, Convair has not grasped the importance of sealing off atmospheric air from a
hydrogen tank nor the knowledge of how to do it.84
Representatives from the National Bureau of Standards Laboratory in Boulder, Colorado; Linde
Air Products Company (with extensive knowledge of the handling of liquid oxygen); and NASA all
voiced their frustration at repeated refusals to accept suggestions for improvement. They suggested
changing the type of insulation or applying an external seal coating, and purging beneath the shell
with helium gas in order to prevent air from contaminating the insulation. The memo from Gray
concluded, Nothing being done now gives any real basis for optimism.85
As the costs of the program mounted and the launch date slipped, Centaurs troubles, both tech-
nical and managerial, became increasingly serious. Ehricke recalled his frustration with Pratt &
Whitney and the Jet Propulsion Laboratory (JPL). Both organizations blamed General Dynamics for
the delays in the Surveyor Program when, in fact, each had serious development problems of its
own.86 Up to this point, Marshall had left day-to-day management of the Centaur to the Air Force.
But as Centaur problems multiplied, Headquarters demanded greater accountability. Advent, an
important Department of Defense program, and two high-profile NASA missions, Mariner and
Surveyor, depended on Centaur. Even more important, Centaur was considered essential to proving
the feasibility of liquid-hydrogen upper stages for the Saturn C-1. Headquarters demanded that the
von Braun team assert greater control over one of NASAs most innovative programs.
84
Memo from Vernon H. Gray to Acting Director, 17 March 1961, RG 255, NARA 254.
85
Ibid.
86
31276-chapter 1 3/8/05 10:49 AM Page 30
31276-chapter 2 3/8/05 12:11 PM Page 31
Marshalls Unruly Wards 31
Chapter 2
Marshalls Unruly Wards

To the Huntsville people the methods and philosophies of the Atlas and
Centaur were as mysterious as the dark side of the Moon.
Deane Davis, General Dynamics
The Centaur had serious technical and managerial problems that proved an unwanted distrac-
tion from the development of the Saturn C-1 at Marshall Space Flight Center. Administrator Keith
Glennan aptly captured Huntsvilles attitude when he remarked, Saturn was a dream; Centaur was
a job.1 As the Centaur schedule slipped, tensions between Marshall and NASA Headquarters
surfaced. In January 1962, Headquarters insisted that von Braun become more involved in the
troubled program. The explosion of the first Centaur in May 1962, a scathing indictment of the
program by the House Subcommittee on Space Sciences, and bad press led to the transfer of
Centaur from Huntsville to Cleveland in the fall of 1962. Implicit in the Centaur failure hearings
was the tension between different philosophies of design and management of large-scale technology
projects.
Huntsvilles Indifference
As early as March 1961, Headquarters expressed increasing impatience with von Brauns
indifference to the fate of Centaur. Don Ostrander, Director of NASAs launch vehicle programs,
began to pressure von Braun to exercise greater oversight of the program. He presented him with
a long confidential memo that almost certainly raised the hackles of the Huntsville team. It
included a management checklist. Ostrander called Centaur the most urgent task before us at
this time. It was needed not only for Surveyor and Mariner, but also for its hardware contri-
bution to Saturn. Ostrander demanded that Centaur be supported fully by the best talent
available at MSFC. He instructed von Braun not only to submit informal monthly progress
reports to Headquarters, but also to deliver them in person. He wanted regular meetings set up.
These were to be attended by the senior NASA oversight personnel at Pratt & Whitney and
General Dynamics: Streamlined, effective management of the project must be considered essen-
1
Oran Nicks, Far Travelers (Washington, DC: NASA SP-480, 1985), 90.
31276-chapter 2 3/8/05 12:11 PM Page 32
tial, he wrote, and any organization changes within NASA or within contractors plants
deemed necessary to accomplish this must be immediately implemented.2 It appears that von
Braun had no intention of assigning his best people to monitor General Dynamics.
Ostrander also thought that Pratt & Whitney needed better supervision. He wanted a test
program set up at Lewis Research Center to help solve engine ignition problems. In short, he
thought that Marshall must provide more active management of all aspects of the program,
including a full-time project manager. To facilitate better management, the Air Force transferred
all Centaur contracts (with the exception of those with Pratt & Whitney) to Marshall in June
1961. Von Braun named Hans Hueter Director of the Light and Medium Vehicle Office. Frances
Evans took over as Centaur program manager.
These measures did not improve the performance of General Dynamics. In September 1961,
the Centaur launch schedule continued to slip, while program costs rose to $100 million. In a
confidential memo, Abe Silverstein informed Associate Administrator Robert Seamans that
Centaur had become an emergency of major proportions.3 Silverstein emphasized that NASA
had a tremendous investment to protect. He prophetically warned, More serious problems,
such as explosion on the pad or a fundamental vehicle or propulsion problem, could virtually wipe
out the robotic lunar and planetary program until 1965, except for the makeshift effort with Atlas-
Agenas. Silverstein thought one of the main problems was that the Department of Defense
Advent satellite seemed to be taking priority over NASA missions. Because Advent weight require-
ments were increasing, while the lunar and planetary requirements were holding steady, NASA
was saddled with an improvement program which we dont really need, one which will probably
lower the reliability of the early NASA vehicles. He recommended that all changes to Surveyor be
halted as soon as an injection weight of 2,500 pounds could be guaranteed. He also thought the
RL10 engine allocations to Saturn should be reviewed to see whether they were interfering with
Centaur. In November 1961, Homer Newell became head of a new Office of Space Sciences and
Applications with responsibility for both Centaur and Surveyor. Newell, however, had limited
time to coordinate the two extremely demanding programs.
Oran Nicks, Director of NASAs Lunar and Planetary Programs, was fully aware of the
magnitude of Centaurs problems. He was dubious that Centaur would be ready in time to fly
the Venus probe planned for August 1962. By the spring of 1961, he began to consider Atlas-
Agena as an alternative. After discussing the possibility with people at JPL, he returned to
Washington determined to find out whether the development of Centaur was as problematic as
he had heard from various sources, including Donald Heaton. A meeting attended by
2
Don R. Ostrander to Wernher von Braun, 20 March 1961, Centaur files. See also Ostrander to von Braun, 29 March 1961,
John Sloop Papers, NASA Historical Reference Collection.
3
Abe Silverstein to Associate Administrator, 6 September 1961, papers of John Sloop, NASA Historical Reference Collection.
31276-chapter 2 3/8/05 12:11 PM Page 33
Administrator James Webb, Silverstein, von Braun, Nicks, Ostrander, and Edgar Cortright,
Assistant Director of Lunar and Planetary Programs, confirmed his worst fears. Nicks recalled,
The meeting produced a formal position by Wernher von Braun that the future of Centaur was
totally uncertain. We left the meeting with the clear understanding that our Centaur-based plan-
etary missions were postponed indefinitely.4
Originally, the payload capability of Centaur was set at 2,700 to 2,800 pounds for lunar
missions. Based on the projected weight of Centaur, JPLs contract with Hughes for Surveyor
called for a spacecraft weight of 2,500 pounds with a science payload of approximately 340
pounds. In the spring of 1962, Marshall managers informed JPL that Centaur could not lift this
weight and asked for a reduction in payload weight to 2,100 pounds. This weight change had a
demoralizing effect on JPL and Hughes engineers because they had to redesign the Surveyor
mission to accommodate fewer scientific experiments.
Whether Surveyor was even necessary for the Apollo program was called into question in
1962. Nicks recalls in his personal memoir Far Travelers that Max Faget, one of the leading engi-
neers on the Apollo program, gained the ear of Congressman Joseph Karth (D-Minnesota), the
powerful head of the Subcommittee on Space Sciences of the U.S. House of Representatives
Committee on Science and Astronautics. Nicks claims that Faget flatly declared that not only was
Surveyor superfluous as a prototype for the Lunar Excursion Module, but even the data Surveyor
was supposed to provide on the topography of the Moon could be obtained by other means.
Fagets meddling with a powerful Congressman raised a furor at Headquarters. Webb promptly
set the record straight about Surveyors importance to Apollo, Nicks relates, but Apollo engineers
continued to believe they would never have data they needed from Surveyor. Although they were
careful not to speak openly of doubts about Ranger and Surveyor, Nicks believes they were
convinced that they could not depend on these questionable programs.5
Apollo designers may have refrained from directly attacking Centaur, but Karth felt no such
compunction. Congressional hearings in March tackled the issue of the escalating costs of NASAs
programs. Congress wanted to terminate the Centaur and Surveyor programs. Karth thought
NASA should admit its mistake in developing Centaur. He was especially critical of the fuel-feed
system used in the RL10 engine. But NASA and the Air Force presented a unified front in defense
of Centaur.6 A decade later, policy analyst Erasmus Kloman speculated that although scientists had
probably been fairly sure of the composition of the Moons surface by the time Surveyor was
launched, it provided a forcing mechanism for the development of Centaur as part of the space
agencys long-range launch vehicle program.7 NASA evidently thought the space program needed
liquid-hydrogen fuel and that Centaur would prove its feasibility.
4
Oran Nicks, Far Travelers, 17.
5
Ibid, 9192.
6
Brian Duff, NASA Officials Defend Centaur, San Diego Union (20 March 1962).
31276-chapter 2 3/8/05 12:11 PM Page 34
By December, the Centaur program was in crisis. With Atlas-Centaur already installed on the
launch pad at the Cape, an extremely serious new problem had surfaced. An ambitious test program
run at the companys Sycamore Canyon facility revealed leaks in the intermediate bulkhead that sepa-
rated the liquid-hydrogen and liquid-oxygen tanks. Hydrogen was seeping through imperceptible holes
in the welds. These holes seemed to open up only after the liquid hydrogen was loaded into the tank.
This hydrogen leakage was a serious problem because it destroyed the vacuum between the oxygen and
hydrogen tanks and increased the transfer of heat across the bulkhead. The leaks called into question
the basic design of the rocket. General Dynamics asked some of its best engineers to tackle the problem,
including Atlas designer Charlie Bossart.
Headquarters sent John Sloop, then NASAs Deputy Director of Launch Vehicles and
Propulsion Programs, to investigate exactly what was happening at the company. Sloop, a strong
advocate of high-energy fuels, had worked in the rocket section at Lewis Research Center until
called to Washington by Silverstein. In a confidential memo to Homer Newell, written in
December 1961 after a visit to General Dynamics, Sloop revealed that the company considered
the problem of leaks across the bulkhead so serious that the integral tank design might have to be
scrapped. He reported that the company thought that it might still be feasible to fly the Surveyor
and Mariner missions on Centaur because the hydrogen leaked so slowly. But because of the
longer coast time of the Advent mission, liquid-hydrogen leakage might jeopardize the mission.
[General Dynamics/Astronautics] believes that the only safe way to meet all Centaur missions is
to drop the integral tank design and to go separate fuel and oxidizer tanks,8 Sloop noted. This
was a blow to the defenders of the integral tank design, particularly Charlie Bossart. The memo
noted that Marshall had already requested General Dynamics to begin a design study using sepa-
rate tanks. Sloop argued against abandoning the innovative tank design, urging a redesign of the
bulkhead with greater wall thickness at the points of welding. Whether this would provide a real
solution to the problem, however, was not yet clear.
Sloops second report, filed a few days later, contained an indictment of General Dynamics
management structure. At this time General Dynamics had a matrix structure with all the projects
drawing support from the operating divisions. Most of its 32,500 employees worked on the Atlas
project, which commanded the lions share of the company resources. Ehricke was one of five
program directors. He had only twenty-seven people assigned to the project, only five of whom
reported directly to him. Deane Davis, the Centaur Project Engineer, had two that he directly
supervised. Although the men assigned to Centaur had specialized knowledge of Centaur compo-
nents and subsystems, only Ehricke and Davis were concerned with overall performance and
7
Erasmus Kloman, Unmanned Space Project Management: Surveyor and Lunar Orbiter (NASA SP-4901, 1972), 10.
8
John Sloop, Memorandum for Director of Space Sciences, 18 December 1961, John Sloop papers, NASA Historical
31276-chapter 2 3/8/05 12:11 PM Page 35
systems. Sloop observed, Mr. Ehricke is doing a Herculean task but Centaur is too big for one
man. What is badly needed is strength in depth; namely, give Mr. Ehricke a larger staff directly
under his control.9 The small team of five Marshall engineers who were monitoring the contract
in the Resident Office were powerless to effect the sweeping organizational changes that were
needed to make Centaur a viable program.
At last, NASA acted to exert more direct control over the Centaur program. Marshall Space
Flight Center set up a program office in Huntsville to provide greater oversight. In response to
pressure from NASA, General Dynamics/Astronautics President James Dempsey directed Grant
Hansen to replace Krafft Ehricke as head of the Centaur program. Krafft Ehricke became Director
of Advanced Studies, charged with the companys long-range plans and applications for Centaur
hardware. Dempsey promoted Hansen to vice-president reporting directly to him, bypassing the
chief engineer of the company, Mort Rosenbaum, who had been at loggerheads with Ehricke. In
discussing the transition in leadership, Hansen explained, And then finally came sort of a show-
down and, as Dempsey explained it to me, they felt that Krafft was a tremendously imaginative,
creative idea man, a hell of a good engineer, but that he wasnt enough of an S.O.B. to manage a
program like this. He had functional department people who would tell him to get lost, and he
would be willing to do just that, and they needed somebody who wouldnt.10 Hansen proved to be
a skilled manager who possessed the authority and resources that Ehricke had lacked. The early
months of his tenure would be marked by a major reorganization of the project, the failure of the
first flight, and a highly critical congressional investigation. Undeterred, Hansen headed Centaur
through the end of the Surveyor program.
Born in Bancroft, Idaho, in 1921, Hansen had been brought up in California, where he
attended San Bernardino High School and Junior College. He had served in the Navy during
World War II as an electronics technician and engineer. The G.I. Bill made it possible for him
to earn a B.S. in electrical engineering at Illinois Institute of Technology. He did graduate work
at the California Institute of Technology while he worked in missiles and space systems engi-
neering with the Douglas Aircraft Company from 1948 to 1960. At Douglas, he designed the
Nike launch control and test equipment and supervised test firings at White Sands Proving
Ground. He had responsibility for electrical systems for Nike Ajax, Nike Hercules, Honest
John, Sparrow, MB-1, Thor, Thor-Delta, Nike Zeus, and Skybolt missiles. Moving from
Douglas to General Dynamics in May 1960, Hansen became Chief Design Engineer, in charge
of all design and laboratory testing, and also became the production-engineering liaison on the
Atlas and Centaur programs.
9
John Sloop, Memorandum for Director of Space Sciences, 20 December 1961, John Sloop papers, NASA Historical Reference
Collection.
John Sloop interview with members of the Aerospace Division Convair/General Dynamics, 29 April 1974, NASA Historical
10
31276-chapter 2 3/8/05 12:11 PM Page 36
Hansen had known that Centaur was in trouble, but the promotion took him by surprise.
He read it in a memo that simply showed up in his mailbox.11 His appointment as vice-president
in charge of the Centaur program coincided with the change in management structure from a
matrix to a project-type organization. In announcing the change to employees, a company
newsletter stated, In arriving at the Project Centaur decision, two principal factors were
considered by management: the future of the company demands to a large degree on upgrading
the Centaur effort to an exceptional level of efficiency, and also the company must utilize
Ehrickes unique creative talents to the fullest degree possible for originating new products, just as
he originated the Centaur.12 Hansens integrity and calm leadership style earned the respect of
those with whom he worked, both inside the company and within NASA.
The reorganization, called projectization, gave Hansen more direct control and freedom to
build an organization from the ground up. We had some very good engineers, he said. I could
hand-pick the best ones to be part of the Centaur project, and I did that. So we had people like
Red Lightbown and many others who were the best engineers that Convair had.13 General
Dynamics assigned 1,000 employees to the project. They took over a building vacated by one of
the Atlas groups. To assist Hansen in implementing the change to a project structure, the company
hired a consulting firm, Robert Heller and Associates.
In a letter to Marshall program director Hans Hueter, Jim Dempsey emphasized how the
reorganization would improve Centaur operations: Such an organization change cannot of itself
automatically solve all the problems, but can serve as an effective tool for improving attention to
the program, facilitating effective internal communications, clarifying responsibility and authority,
speeding up program activity, and identifying team relationships and appropriate communication
channels with the Marshall Space Flight Center and the General Dynamics/NASA office.
Dempsey assured Hueter that the new launch date would not be delayed by this reorganization.14
In an appended memo to Hueter, Dempsey assured him that Atlas expertise would be avail-
able to Centaur. The Atlas launch team at the Atlantic Missile Range would not only be
responsible for Atlas, but would also act as a subcontractor for Centaur. Dempsey further assured
Hueter that engineers with Atlas knowledge and experience were being transferred to Centaur.
In addition, Hansen would participate in Dempseys daily staff meeting. Atlas reports would be
distributed to Centaur personnel, and Centaur personnel would participate in the Atlas engi-
neering change board reviews. Dempsey acknowledged that one of the weaknesses of a project
11
Interview with Grant Hansen (former employee of General Dynamics) by Virginia Dawson, 6 June 2000.
12
FYI: Reorganization for Project Centaur, 29 January 1962, John Sloop papers, NASA Historical Reference Collection.
13
Interview with Grant Hansen by Virginia Dawson, 6 June 2000.
14
J. R. Dempsey to Hans Hueter, 12 February 1962, John Sloop papers, NASA Historical Reference Collection.
31276-chapter 2 3/8/05 12:11 PM Page 37
organization, as opposed to one that was organized in functional groups, was its isolation within
the company. He assured Hueter that Centaur top management would make a special effort to
stress a cross feed of learning. He wanted Marshall representatives to have maximum visi-
bility into [General Dynamics/Astronautics] Centaur operations so that they have every
opportunity to aid in ferreting problems.15
To implement more rational schedule planning and control, NASA insisted that General
Dynamics adopt Program Evaluation and Review Technique (PERT) and Companion Cost
reporting. PERT, a time-/cost-management tool first used by the Navy to develop the Polaris
missile, was adopted by NASA in late 1961 to impose discipline on the management of large proj-
ects by developing a master schedule for every task of the program and tracking progress. Each
department head was responsible for accomplishing a specified task within an allotted timeframe.
Dempsey assured Hueter, We have adopted the attitude that any schedule slippages not directly
assignable to a[n] unforeseeable technical development problem are indicators of lack of compe-
tence on the part of either the Schedule Planning and Control personnel or the responsible
function line managers.16
Dempsey expected everyone in the organization to think in terms of systems engineering.
Each part must be designed, each procedure written, each test planned and accomplished with full
recognition of how that effort fits into the total picture and with concern for the implications upon
all other elements of the system.17 To strengthen systems engineering at General Dynamics, Deane
Davis was given more control. As Deputy Program Director for Technical Control, he was respon-
sible for all technical planning and integration. At the same time, design engineering under William
Radcliffe was broadened to include system design engineering. Centaur design personnel drawn
from design groups throughout the company were organized into nine specialized design groups,
such as propulsion, structures, and guidance. Hansen believed that the move from matrix to project
management was just in time. Without this reorganization, the program would have been canceled.
I knew that I had to get the problem solved, and get it working successfully, or NASA was going
to slam the door on us, he said.18
But the change from matrix to project management without any real effort to control the
program by Marshall could not solve all of Centaurs problems. In Hansens view, von Braun just
sort of pushed it into a backroom and let it be handled by Hans Hueter. Hueter had served as
Chief Test Engineer at Peenemnde, emigrating to the United States with the other members of
15
Ibid.
16
Ibid.
17
Attachment to letter to H. Hueter from J. R. Dempsey, 12 February 1962, John Sloop papers, NASA Historical Reference
Collection.
18
31276-chapter 2 3/8/05 12:11 PM Page 38
the von Braun team in 1945. His German background did not prepare him for the entirely
different engineering philosophy at General Dynamics. The two cultures inevitably clashed. From
Deane Daviss perspective:
To the Huntsville people the methods and philosophies of the Atlas and Centaur were as
mysterious as the dark side of the Moon. On the Astronautics side, still riding the crest
of their brilliant success with the Atlas, and with no reason to believe that the Centaur
would prove to be any different, the new Centaur people resented what they felt was
undue interference in their established manner of creating, producing, and operating
their product. After all, they said repeatedly, the Atlas and Centaur were their inven-
tions.19
Unlike most of the Germans on the project, Hueter was actually loved and respected by his
unruly wards at Astronautics. Hueter had the impossible task of reconciling what Davis consid-
ered to be two diametrically opposed design and administrative approaches. Shortly after the
program was transferred to Marshall, the Convair plant in San Diego was deluged with
Huntsville personnel attempting to understand their new creature. Apparently, whenever the
Germans came to San Diego, Atlas designer Charlie Bossart always tried to keep a low profile.
During one visit, the Germans took a tour of the plant and visited the Sycamore Canyon hot-firing
testing facility. Then they all sat down for a late-day briefing. Trouble began when Bossart and Willie
Mrazek, von Brauns Structural Section Chief, began to argue in loud whispers. Their voices began
to rise over the subject of the rockets structure. Charlie was trying to explain its merits to a disbe-
lieving Mrazek and from experience I knew that the only solution was to get those two gentlemen
separated from the briefings so they could have at it.20
Bossart led Mrazek out into the factory yard, where a Centaur tank stood gleaming in the
sunlight. Mrazek asked, Whats inside it? To which Bossart responded, Nitrogen. Nitrogen
was used for pressurization until the rocket was filled with its liquid-hydrogen/liquid-oxygen
propellants just prior to launch. Without pressurization, the thin skin of the rocket would first
wrinkle, then collapse. Mrazek, familiar with the solid, reinforced walls of rockets designed by the
von Braun team, was perplexed by Bossarts insistence that nitrogen, kept at the relatively low pres-
sure of 8 to 10 pounds per square inch, was sufficient to keep the tank rigid.
He also questioned how such a thin, unbuttressed structure could be strong enough to carry a
rocket aloft. To quell Mrazeks doubts, Bossart invited him to take a sledge hammer and give the tank
a whack. Failing to put even the slightest dent in the tank, he tried again, this time giving the side
of the tank a glancing blow that caused the sledge hammer to fly out of his hand, knocking his glasses
19
Deane Davis, Seeing Is Believing, or, How the Atlas Rocket Hit Back, Spaceflight 25 (5 May 1983): 196198.
20
Ibid.
31276-chapter 2 3/8/05 12:11 PM Page 39
Turbine
Thrust Chamser
Expander Cycle
Pumps
Diagram of RL10 engine: Liquid oxygen and liquid hydrogen are pumped through tubes toward the engines thrust chamber. Liquid
hydrogen cools the walls of the thrust chamber while it picks up heat and becomes gaseous. The engines unique expander cycle, in
which a small amount of gaseous hydrogen in the exhaust stream flows backward to drive the pumps of turbine, can be seen in the
center of the diagram. (Courtesy of Pratt & Whitney)
off, but again leaving the surface unscathed. Although this test may have proved the strength of the
balloon structure, it did nothing to endear General Dynamics to Mrazek or win the von Braun
groups faith in the ability of Centaur to lift an expensive spacecraft into space.21
Shortly after the reorganization of the program, the Aerophysics Group at General Dynamics
began to tackle the aerodynamics of the integration of the two rockets. This was a major undertaking
that Don Lesney realized would be of enormous importance to the success of Centaur. In addition to
a big interstage adapter to connect Atlas with Centaur, the vehicle carried a 10-foot-diameter payload
fairing. What concerned Lesney was whether the middle of the vehicle was strong enough to with-
stand the aerodynamic stresses caused by flying through the wind. The vehicle was double the length
of the original Atlas ICBM. Questions such as how thick to make the skin on Atlas and how much
21
Ibid.
31276-chapter 2 3/8/05 12:11 PM Page 40
pressurization was necessary to keep it structurally intact could only be solved through testing. Lesney
recalled that because his Huntsville counterparts never returned his calls, he was unable to get the go-
ahead for wind-tunnel studies.22 Management of RL10 engine development by the von Braun group
produced similar tensions, although it saw the potential of the RL10 as the propulsion system for one
of Saturns upper stages. In contrast, it seemed to want little to do with the upper stage launch vehicle
for which the RL10 was originally designed.
Troubles at Pratt & Whitney
Pratt & Whitneys engineers had developed confidence in handling liquid hydrogen fuel
during Project Suntan. They relished the prospect of contributing this new expertise to the space
program. However, Pratt & Whitney was a turbojet engine manufacturer and had no prior expe-
rience in designing a rocket engine. The company had signed a contract with the Air Force in
mid-October 1958 to develop the RL10 enginean engine whose innovative design and
outstanding reliability would earn the designation of historic engineering landmark from the
American Society of Mechanical Engineers in 1979. Since the 304 turbojet engine they had
designed for Suntan was air-breathing, the first hurdle was the choice of oxidizer. Originally, the
companys engineers favored fluorine because of its superior efficiency, but fluorine proved so
corrosive and volatile that they settled on liquid oxygen. The turbopump designed for the 304
dictated the RL10s original 15,000-pound thrust rating.23
Initially, Pratt & Whitney engineers did not try to adapt the unique expander cycle of the
304 to the RL10. Early designs, based on the conventional gas generator cycle, required a sepa-
rate combustion chamber to start the turbopumps. A suggestion from Perry Pratt, the
companys chief engineer, led the design team back to the elegant simplicity of the 304. Instead
of fighting against the inherent properties of liquid hydrogen, the expander cycle made them
intrinsic to the design.24
The engine was designed and built at the companys factory in East Hartford, Connecticut.
Testing took place at the companys new Florida Research and Development Center. The RL10 test
team reveled in the freedom from interference it enjoyed in this remote swampland near West Palm
Beach. By July 1959, they had completed 230 successful firings in a horizontal test stand. The RL10
started, stopped, and restarted with surprising reliability. By November, the technical problems
appeared to be nearly solvednot so the escalating costs of developing the new engine.
Prior to contracting with NASA, Pratt & Whitney government contracts had always been
fixed-price. With this type of contract, any cost overruns had to be assumed by the company. The
22
Interview with Don Lesney, 5 June 2000.
Joel E. Tucker, The History of the RL 10 Upper-Stage Rocket Engine, 19561980, published in History of Liquid Rocket Engine
23
Development in the United States, 19551980, ed. Stephen E. Doyle, AAS History Series, vol. 13 (San Diego: AAS, 1992), 123151.
24
Mulready, Advanced Engine Development, 60.
31276-chapter 2 3/8/05 12:11 PM Page 41
NASA engineers Ali Mansour and Ned Hannum review test results of Pratt & Whitneys RL10 engine at Lewis Research Center,
1963. (NASA C-64329)
31276-chapter 2 3/8/05 12:11 PM Page 42
cost-plus-fixed-fee contract for the RL10 allowed the company to bill for additional development
costs when they ran into unforeseen problems. Initially, Mulready apparently neglected to submit
all the design changes to NASA, simply addressing the problems as they arose. We had not learned
to wait and put changes into the contract. It was noticed that Convair, which was an old hand at
the cost-plus business, tended to keep its head down and point left in times of trouble.25 Once
Pratt & Whitney began to bill for design changes, costs escalated. Mulready recalled how after a
dressing-down at NASA Headquarters, he returned to Florida determined to cut costs. Confident
after one hundred successful test firings that the engine was extremely reliable, he canceled test
equipment for the ignition systema decision he later regretted. Only one engineer challenged
him, claiming that data revealed that the engine was not lighting on the same spark every time.26
Mulready dismissed this objection as ludicrous.
Meanwhile, Lewis Research Center became deeply involved in the development of the RL10
engine. Lewis had previously developed a regeneratively cooled fuel/oxidizer injector that served
as the basis for the engines injector design. Pratt & Whitney engineers added a textured surface
of special aluminum mesh called Rigi-Mesh that facilitated cooling.27 NASA Headquarters not
only asked Lewis engineers to assist Pratt & Whitney in troubleshooting, but also wanted them to
develop an in-house competency that would be useful in monitoring the work of the contractor.
In March 1960, the Center initiated a 10-month test program with three RL10 engines. In addi-
tion to providing a check on contractor performance, this testing yielded a complete altitude
evaluation of the propulsion system.28 By November, Lewis engineers had completed two
successful firings of the engine in the altitude wind tunnel. In recognizing the contributions of
Lewis Research Center, Ostrander wrote, The Lewis staff assistance, which is based on an exten-
sive technical base and on unique experience in handling these propellants in similar equipment,
has been and is invaluable.29
After a series of tests of the engine in the horizontal position, Pratt & Whitney initiated
tests of two engines in October 1960 in a new vertical test stand called E-5 at the companys
Florida test center. The engine would actually be flown in the vertical position, so these new
tests were especially important. The first test went as planned, but during the second test, the
engine exploded. Investigation suggested that the explosion could be attributed to a human or
procedural error. Nevertheless, the mishap caused the Centaur launch schedule to slip to 20
25
Mulready, 76.
26
Ibid.
27
Mulready, 70. See also Sloop, Liquid Hydrogen, 191193; Bilstein, Stages to Saturn, 134140; Tucker, History of RL 10, 132.
Centaur Propulsion System Testing, Contract NAS9-2691, Pratt & Whitney, 7 December 1962, Vol. 1, Box 59, Old RL10
28
Records (Goette files), DEB vault, NASA Glenn.
29
General Don R. Ostrander to Eugene Mangeniello, 23 June 1961, Box 254, RG 255, NARA, Chicago, IL.
31276-chapter 2 3/8/05 12:11 PM Page 43
December 1961. At this point, the engine also had thrust-control problems, in addition to
hydrogen leakage rates in excess of specifications. Pratt & Whitney fixed the damaged test
stand and resumed testing on 12 January 1961.
After the engine exploded a second time during an important demonstration to Air Force
and NASA officials in the spring of 1961, Mulready immediately ordered the igniter test rig he
had previously canceled. Testing determined that in the horizontal position, a small amount of
liquid oxygen, previously unnoticed, flowed back toward the spark, mixing with the hydrogen
gas. In the horizontal position, this facilitated ignition, but in the vertical position, gravity cut
down the backflow of oxygen, thereby preventing the firing of the engine. The installation of an
oxygen feed line to the igniter solved the problem. On 24 April 1961, two RL10 engines success-
fully completed their first full-duration test on the E-5 test stand.
Von Braun used the two engine explosions as his cue to take a more assertive role in manage-
ment of the contractor. (He called this penetrating the contractor.) Marshall engine specialist
Leonard C. Bostwick insisted on regular meetings and a formal development plansomething the
fiercely independent Pratt & Whitney engineers found disruptive and inappropriate. Mulready
wrote, The rapid growth of the RL10 technology, which was totally new in so many dimensions,
could not have happened with the formality that NASA sought.30
But working with NASA proved to be a positive experience for Pratt & Whitney in many
ways. Although its engineers were enormously competent turbojet designers, they benefited
from the rocket expertise of the Marshall and Lewis teams. For example, at Pratt & Whitney,
turbojet engines were always given a manual countdown. NASA insisted on automated test
stands. The gentle nature of the RL10, which operates similarly to a jet engine, was the only
reason that Pratt & Whitney was able to use the old system as long as it did, Mulready
admitted.31
In November of 1961, the Air Force contracts with Pratt & Whitney were transferred to
Marshall. By May of 1962, the RL10 engine had completed over 700 hot firings. It proved to be
so reliable that Marshall chose to power the Saturn S-IV stage with six RL10 engines. When called
to testify at the Centaur hearings in May 1962, Bruce Torel asserted proudly that all of the devel-
opment problems of the free worlds first liquid hydrogen rocket engine had been solved.32 He
was correct, although Pratt & Whitney would have to wait more than a year for General Dynamics
to solve many of Centaurs technical problems before the RL10 would fly.
30
Mulready, 83.
31
Ibid.
32
Centaur Program, 1962 Hearings, 111
31276-chapter 2 3/8/05 12:11 PM Page 44
Launch Operations at the Cape
At Cape Canaveral, launch operations for unpiloted vehicles were located at the Atlantic
Missile Range east of the Banana River. This area under Air Force control originally contained one
launch pad for Atlas-Centaur, referred to as Pad 36A. In addition to the first pad, a second pad,
36B, was completed in the mid-1960s. A Central Control Building, situated a safe distance from
the launch sites, housed the Air Force range safety officers. They were responsible for all range
support, including destroying a vehicle if it suddenly veered off course.
Initially, Marshall managed Atlas-Agena, Atlas-Centaur, and Thor-Agena launch operations
at the Cape. This included all robotic scientific and applications missions except Delta, which was
under the control of Goddard Space Flight Center. Goddard personnel at the Cape were all former
Vanguard rocket people. General Dynamics was responsible for integrating Centaur with Atlas,
while JPL was responsible for integrating the spacecraft with the launch vehicle.
Building AE, previously a staging area for Air Force missiles, played an important role in
prelaunch operations. Oran Nicks had the former hangar air-conditioned. Nicks remarked,
While prelaunch checkout facilities steadily became less ramshackle, it was 1964 before we could
begin to treat our interplanetary travelers with the care they deserved.33 A cleanroom was built
onto the rear of the building that was later used for the installation of Centaur shrouds.
Everyone at the Cape during the early NASA years was learning. James Womack, one of the first
NASA staff at the Cape, recalled that he was familiar with more conventional propellants but had
never worked with liquid hydrogen. Because the Cape operations staff knew little about liquid
hydrogen, they attended the required safety lectures. They were cautioned that since hydrogen
produces no visible flames during combustion, they should take a broom along every time they went
near a hydrogen line. If the broom did not burst into flames, it was considered safe to proceed. They
quickly realized that this precaution was not necessary; although liquid hydrogen required attentive-
ness and respect, it was considerably less volatile than they had been led to believe.34
Roger Lynch was one of the first Convair employees assigned to Cape Canaveral in early 1961.
A graduate of the U.S. Merchant Marine Academy who had joined General Dynamics in 1956,
Lynchs first job at the company was at Sycamore Canyon. He was asked by Ken Newton to go to the
Cape to ready the Atlas-Agena launch site in preparation for the launch of the Ranger series of Moon
shots. He was transferred to Pad 36 in anticipation of the launch of the first Atlas-Centaur.
During that time, Lynch observed the early maneuverings of the Marshall Space Flight Center
engineers as they began to make their presence felt at the Cape. They were in the process of
installing the large Saturn bureaucracy on the other side of the Banana River (at the site later to
be known as Kennedy Space Center) under Kurt Debus. Relations with the Air Force and
33
34
Interview with Jim Womack by Virginia Dawson, 11 November 1999.
31276-chapter 2 3/8/05 12:11 PM Page 45
Wernher von Braun anxiously monitors the launch of a Pioneer 4 spececraft for a lunar flyby using a Juno II, 1959.
(NASA 9131491)
Marshall were strained because the Air Force was determined to maintain control of robotic oper-
ations. General Dynamics managed a launch site under the distant supervision of Marshall
personnel. Initially, the Marshall people did not take much interest in the preparations at Pad 36.
As the date of the first launch approached, Roger Lynch became aware of the contempt of
Marshall engineers for the pressure-stabilized design of Atlas and Centaur.
We had come out of a very successful major ballistic weapons system program with tech-
nology that was foreign to Marshall. Our monocoque structure was considered high risk.
They took exception to all aspects of the vehicle design. It was different in every way
possible from what they were starting to design for Saturn. I was in a meeting at
Huntsville when it was admitted that even Saturn, for all of its structure, would collapse
in flight without internal tank pressure. Over drinks, they admitted that it made little
sense to fly all that structure. The real reason for their design was a passion for vertical
check-out.35
35
Letter from Roger C. Lynch to Virginia Dawson, dated 4 June 2000.
31276-chapter 2 3/8/05 12:11 PM Page 46
Lynch observed that Marshall engineers were equally conservative with respect to electrical
systems. They preferred old-fashioned relays and vacuum tubes. Their junction boxes were so
heavy that he thought they belonged on a battleship, not a spaceship. General Dynamics and
Honeywell designers were pushing the state of the art with much lighter solid-state electronics,
although components often proved unreliable.
The von Braun team was also critical of the ground support equipment used by General
Dynamics. At this time, Atlas and Centaur were considered two distinct systems with a
minimum interface. Servicing the two systems was more complicated than it would have been
if an integrated system had been designed from scratch. A whole new vacuum technology that
included the design of large valves and transfer lines had to be developed for liquid hydrogen.
General Dynamics engineer Denny Huber recalled that the design of the vacuum insulation
posed quality-control problems and limited the number of suppliers willing to bid on vacuum-
system hardware. If a potential supplier had no capability to handle liquid hydrogen, his design
approach was at risk.36
Because it was the first time the two vehicles had been mated together, the General Dynamics
launch team was inundated with fierce and unpredictable design changes.37 This was intolerable
to the Huntsville team, which worked from more detailed specifications and left little room for
adjustments by the contractor at the launch site. Von Braun may have been referring to the chaotic
situation at Pad 36 when he observed that it was better to build a rocket in the factory than on
the launch pad.38
The Failure of F-1, the First Atlas-Centaur Launch
Liquid-hydrogen leakage, in addition to guidance and engine problems, kept Atlas-Centaur

sitting like a lone sentinel over the Capes broad expanse of beach from October 1961 through the
following spring. Finally, it was ready. On 8 May, F-1 lifted off into the clear Florida skies for its
maiden voyage. As an eyewitness in the crowded blockhouse reported, While the elated engineers
were shaking hands and congratulating each other the vehicle reached plus 54 seconds into the
mission, and a sudden yell of Missile blew up! Missile blew up! came from one of the observers
at the periscopes.39 Two seconds later, the vehicle was completely obscured by a massive fireball
fueled by thousands of gallons of liquid hydrogen and kerosene fuel. NASA quickly determined
that Centaur was the cause. Aerodynamic pressure on the cover protecting the insulation (called
36
E-mail communication from Denny Huber to Ed Bock, 30 March 2002.
37
Letter from Roger C. Lynch to Virginia Dawson, dated 4 June 2000.
Andrew J. Dunar and Stephen P. Waring, Power to Explore: A History of Marshall Space Flight Center 19601990 (Washington,
38
DC: NASA SP-4313, 1999), 3951.
39
Joseph Green and Fuller C. Jones, 841.
31276-chapter 2 3/8/05 12:11 PM Page 47
the weather shield at that time) had caused it to burst, ripping away the insulation and exposing
the walls of the fuel tank to the heat of the atmosphere. Pressure buildup in the tank from the
boiling off of the liquid hydrogen caused the fuel tank to rupture, spilling volatile liquid hydrogen
down the sides of the rocket, where it was ignited by a spark from the engine. The loss of Centaur
was a serious setback for NASA.
The aborted first flight of Centaur put NASA on the defensive. The House Subcommittee on
Space Sciences, headed by Joseph Karth (D-Minnesota), called for a probe. Karth questioned
whether Centaur was a key to the nations future rocket propulsion needs or simply an increasingly
expensive national liability. The troubled Centaur program had implications for national security.
Four and one-half years after Sputnik, American rocket propulsion technology still lagged behind
that of the Soviet Union.40 Centaur had been billed as an important new propulsion technology. Its
failure reinforced the perception of American technical inferiority. Now, with Centaur considered
the key to a soft landing on the Moon, the subcommittee needed reassurance that the space
program was still on track. Spokesmen for General Dynamics, Pratt & Whitney, and Marshall
Space Flight Center knew that this was by no means a routine investigation. The fate of the
Centaur program hung in the balance.
Given the enormous technical and managerial challenges involved in developing this new
technology, the failure of F-1 seemed almost inevitable. However, the timing was unfortunate.
The Soviet Union had already succeeded in crash-landing two lunar vehicles on the Moon, while
five attempts by the United States Ranger vehicles had all failed. Both Ranger and Surveyor were
needed to provide scientific data on radiation, magnetic fields surrounding the Moon, and the
Moons surface topography. The Ranger failures, compounded by the aborted flight of F-1, forced
NASA to consider whether the beleaguered Centaur program should be canceled.
Wernher von Braun and Hans Hueter admitted that Marshall had inadequately supervised
General Dynamics, but they blamed the laissez-faire approach of the Air Force, which had allowed
the program to evolve without adequate oversight. Von Braun thought that in-house competence
by the government and complete control over contractors was the only way to avoid development
problems. He stated:
I think the general lesson we have learned from Centaur management is probably that
one should never underestimate the magnitude of a program where so many new and
unproven ideas are tried out and I think we will always get in difficulties, as a
Government agency, unless we build up a competence in the Government that we can
really stay on top of the problems right from the outset and learn how to identify poten-
tial problem areas before we have explosions and fires and setbacks.41
40
Walter LaFeber, America, Russia, and the Cold War, 19451996, 8th edition (New York: The McGraw-Hill Companies, Inc.,
1997), p. 192.
41
31276-chapter 2 3/8/05 12:11 PM Page 48
Marshall had taken over management from the Air Force after many of the important design
decisions had already been made. Von Braun admitted that they had not penetrated the program
sufficiently. The only excuse one can have for it is that it started out as a little exploratory
program and grew and grew and grew into a major program, and it wasnt intended that way from
the outset.42
Marshalls approach to contracting with industry differed radically from that of the Air
Force. Von Braun was used to designing, building, and testing a rocket prototype before turning
it over to an industry contractor for closely supervised production.43 The everything-under-one-
roof approach had prevailed at Peenemnde. It dovetailed with the U.S. Army arsenal tradition
of weapons development. Von Braun believed that to manage contractors properly meant
exerting control over every aspect of testing and manufacture. Marshall furnished its contractors
with detailed design requirements and monitored their operations closely, assigning as much as
10 percent of its staff in the 1960s to resident offices.44 In contrast, the Air Force, because it
lacked extensive in-house expertise, allowed its industry contractors to develop designs and inde-
pendently work out solutions to problems.
But more than management bothered von Braun. He had serious reservations about the basic
design of Centaur. He made it clear that Marshall Space Flight Center would never have tolerated
the balloon structure. He called the approach that General Dynamics had taken imaginative, but
risky. In order to save a few pounds, they have elected to use some rather, shall we say, marginal
solutions where you are bound to buy a few headaches before you get it over with. Ultimately when
you are successful you have a real advanced solution. But von Braun did not believe that
Centaur would ever be successful. Referring specifically to the Centaur pressurized tank, he
said, It is a great weightsaver, but it is also a continuous pain in the neck.45 He pointedly
declared that industry contractors would not have the same design freedom when it came to
Saturn development. We are really making an all-out effort to stick together with the
contractor before major design decisions are even made and have our men argue with his men
as to whether this is really the way to go.46
Krafft Ehricke and Grant Hansen defended the technical decisions that determined
rocket structure and fuel choice. A modest person without the aristocratic pretensions and
confidence of von Braun, Hansen admitted that he was rather frightened when called to
42
Ibid.
43
On the arsenal approach, see Michael J. Neufeld, The Rocket and the Reich: Peenemnde and the Coming of the Ballistic Missile
Era (New York: The Free Press, 1995), 108.
44
Andrew Waring, Power to Explore, 3951.
45
46
Ibid., 59.
31276-chapter 2 3/8/05 12:11 PM Page 49
testify. But he said to himself, All Ive got to do is tell the truth, and convince them that
were understanding the problem and are working it and it should go away.47 Given the cost
overruns and delays, he found the legislators concerns to be justified: When a program is
that far behind schedule and costs that much money and blows up on the launch pad, they
owe it to the public to hold a hearing to find out whats going on, and whether or not these
programs should be continued or not. Asked at the hearing about the difference in design
philosophy between General Dynamics and Huntsville, Hansen agreed that Huntsvilles was
more conservative. In contrast to the 35-percent margin of safety the von Braun team spec-
ified in its contracts, the contract between General Dynamics and the Air Force allowed a
25-percent margin. He explained, We are inclined, I think, to be willing to take a little bit
more of a design gamble to achieve a significant improvement, whereas I think they
[Huntsville] build somewhat more conservatively.48 To designers at General Dynamics, the
pressure-stabilized tank may have been a design gamble, but its advantages made it one
worth taking.
Although no longer in charge of the program, Krafft Ehricke vigorously defended the
design, pointing out that the pressure-stabilized tank was also used on Atlas, a rocket deemed
reliable enough to send Mercury astronaut John Glenn into orbit around Earth. In Centaur, the
pressure-stabilized tank design optimized the unique characteristics of liquid hydrogenits low
density and extreme cold. The result was a dramatic savings in weight, always the driving factor
behind any successful rocket design. Ehricke also strongly defended the use of liquid-hydrogen
fuel. He admitted that hydrogen leakage through minute holes in the welds had proved a
problem, but not an intractable one: We have definitely in this particular case, if I might say
so, gambled at a very low weight and found that we have to correct ourselves.49 At another
point in his testimony, he said, Hydrogen itself has turned out to be less of a culprit than many
thought initially. Hydrogen, like all chemical fluids behaves fine if you know its little idiosyn-
crasies and treat it correctly. But you have to go through a development program such as ours
first.50
To Ehricke, Centaur provided an example of a transition from the development of a
strategic weapon to a launch vehicle for the peaceful exploration of space.51 A launch vehicle
required far greater thrust and reliability than a missile. He pointed out that in addition to the
pioneering work by General Dynamics on liquid hydrogen, the company was breaking new
47
48
49
Ibid., 97.
50
Ibid., 69.
51
Ibid., 6768.
31276-chapter 2 3/8/05 12:11 PM Page 50
ground in important technical areas, such as restarting the vehicle after an extended coast in
zero gravity. Centaur also represented a transition stage in government contracting. Centaurs
management problems had proved that the Air Forces hands-off model of contract manage-
ment was not appropriate for a civilian agency.
Ehricke thought that many of the troubles with Centaur could be attributed to the fact that it
had never been accorded the national priority classification it deserved. Centaur was just as impor-
tant to the national space program as the piloted space programsMercury, Saturn and
Apollowhich all had DX priority (the highest national priority in the procurement of materials).
Indeed, President Kennedys announcement in May 1961 that the nation would put a man on
the Moon within a decade made it imperative that Centaur succeed in sending Surveyor to soft-
land on the Moon. Lack of DX priority had made it more difficult to obtain parts from
contractors in a timely manner. Ehricke said, Like other space programs, Centaur was to grow
as it went along. However, unlike Mercury and Saturn, Centaur could never attain DX priority
although its contributions are as much a cornerstone of our future space capability as those of
a manned space capsule and a high thrust booster.52 The other constraint was the limitation
imposed on Centaur by the requirement that it be designed using off-the-shelf parts that
belonged to Atlas. Although this saved money, it led to design compromises.
After the Centaur hearings, the House Committee on Science and Astronautics issued a report.53
The committee emphasized Centaurs importance to the space program: the new vehicle was needed
to fill the performance gap between the Atlas-Agena and the Saturn class vehicles for space missions.
It had been assigned important programs by both NASA and the Department of Defense. About
twelve operational Centaurs a year had been planned to accomplish these missions. The committee
noted that in addition to mission responsibilities, Centaur had an important role in research and
development because virtually all upper stages of large future space vehicles, both chemical and
nuclear, are presently intended to utilize hydrogen as fuel. Centaur would provide essential informa-
tion on the handling, storage, and firing of liquid hydrogen. In short, Centaur is performing a
technical pioneering function which is considered vital to the future of much of the Nations space
effort.54 That effort included the piloted flight to the Moon as the third stage of the Saturn C-1.
Without liquid-hydrogen upper stages, the weight of the hardware needed to accomplish the Apollo
mission would double. In reviewing the history of the program, the report revealed that in the summer
of 1960, NASA had enhanced the importance of Centaur to the space program when it funded other
projects that used liquid hydrogen, such as the Rover nuclear rocket and the upper stages of the Saturn
52
Ibid., 68.
53
House Committee on Science and Astronautics, Centaur Launch Vehicle Development Program, Report 1959, submitted
by George P. Miller, 87th Congress, 2 July 1962.
54
Centaur Report, 2 July 1962, 2.
31276-chapter 2 3/8/05 12:11 PM Page 51
and Nova vehicles. Given the program importance of Centaur, the committee demanded to know why
it had not been assigned DX priority.
The committee placed much of the blame for Centaurs troubles on General Dynamics. It
concluded that because of General Dynamicss preoccupation with the Atlas ICBM program,
Centaur had not received the companys full attention. Until the reorganization under Grant
Hansen, the company had resisted pressure from NASA to address managerial problems. The
committee was especially critical of the companys failure to anticipate the problem with the
liquid-hydrogen leakage in the intermediate bulkhead until after manufacture and shipment to
the Cape. This fundamental mistake in cryogenics engineering should have been discovered
much earlier through testing. The committee also criticized General Dynamicss carelessness in
the location of test stands. A second Centaur had been damaged when an Atlas on a nearby test
stand exploded.
But the committees indictment was not limited to the contractor. It was also critical of
Headquarters and Marshall for weak and ineffective management. They had allowed personnel
responsible for the program to rely on troubleshooting as difficulties arose, rather than antici-
pating them. The delay of Centaur had compromised important missions like the Air Forces
Advent Program. It was necessary to launch NASAs Mariner payload on the less powerful Agena.
Surveyor, originally to weigh 2,500 pounds, had to be redesigned for a lower weight of 2,100
pounds, at a loss of $20 million to taxpayers.
What particularly galled the committee was the failure to reconcile the differences in design
philosophy between Marshall and General Dynamics, especially as it related to the thickness of
the intermediate bulkhead. The report stated:
The subcommittee would not presume to decide the technical question of what consti-
tutes a proper margin of safety in rocket construction, though the evidence seems to
indicate that a more conservative policy might have saved both time and money in the
Centaur development program. While it may not be possible to draw a final conclusion
to the effect that one design philosophy is right and another is wrong, it is obvious that
only one can prevail. The subcommittee considers it inconsistent for the major industrial
contractor to adhere to one approach while the Government agency responsible for direc-
tion of the same development program adheres to another. Further, the subcommittee
considers it important that industrial contractors follow agency specification, policies,
and recommendations immediately upon becoming aware of them.55
The subcommittee recommended a thorough reevaluation of the program by NASA, including

a resolution of the question of DX priority. Finally, it ordered an investigation of General Dynamicss
billing by the General Accounting Office to find out whether the interests of the Government have
55
Ibid., 13.
31276-chapter 2 3/8/05 12:11 PM Page 52
Centaur Assembly line at General Dynamics, 1962. (Courtesy of Lockheed Martin)
been adequately protected under the contracts.56 Of all NASA programs at that time, only Centaur
and Surveyor had received this level of intense scrutiny and censure.
A Key Decision
Although the House Committee Report stopped short of recommending cancellation of the
Centaur program, pressure was building within NASA to abandon the troubled liquid-hydrogen
upper stage. Opposition to the Centaur program came chiefly from MSFC and JPL. Both Centers
had earned reputations as preeminent authorities in questions related to rocket propulsion and
spacecraft development. Their opinions carried weight with Congress and the American public.
Von Braun and Brian Sparks, Deputy Director of JPL, were working behind the scenes to mount
a campaign for the cancellation of Centaur in favor the Saturn C-1-Agena.
NASA Headquarters refused to give in to this pressure because of the high priority of Centaur
within NASA. A memo from the Deputy Associate Administrator to Robert Seamans unequivocally
stated that Centaur was needed to support Saturn development and carry out NASAs robotic plane-
tary missions. The solution was not cancellation, but new management: For the Centaur to succeed,
56
Ibid., 12.
31276-chapter 2 3/8/05 12:11 PM Page 53
NASA must have a capable project management team to direct its efforts and to direct the contractor.57
As a first step, NASA named Vincent Johnson Centaur program manager in June 1962. Johnson, a
graduate of the University of Minnesota, had served as a physicist at the Naval Ordnance Laboratory
during World War II. After a career in the Navy Bureau of Weapons, he came to NASA in 1960 as
program manager for the Scout, Delta, and Centaur launch vehicles. In 1964, Johnson became director
of the launch vehicle and propulsion programs in the Office of Space Science and Applications, where
he remained until 1967. In Johnson, Centaur gained the undivided attention of a capable Headquarters
executive with a firm grasp of engineering.
Johnson and Homer Newell immediately approached Abe Silverstein to see whether Lewis
Research Center would be willing to take over the troubled program. Silverstein had played a
signal role in shaping NASA under Glennan, but when James Webb took over as NASA
Administrator in 1961, Silversteins influence at Headquarters began to wane. After the decision
to fund the giant Saturn V rocket for a piloted lunar landing, Webb wrestled with the decision of
whom to choose to head the Apollo program. The two top candidates were Silverstein and von
Braun. Webb passed over von Braun, reputedly because of the adamant opposition of his Deputy
Administrator, Hugh Dryden. Put off by Silversteins autocratic management style and his insis-
tence on having complete control of the Apollo program, Webb asked Brainerd Holmes of the
Radio Corporation of America to head the program. Silverstein accepted the post of Director of
Lewis Research Center.58
Silverstein returned to Cleveland in November 1961, shortly before fellow Ohioan John Glenns
historic flight. During the nearly four years Silverstein had been away, Lewis Research Center had
vacillated between whether to go after a significant piece of the space program or continue to focus
on its forteaircraft propulsion research. Many engineers feared that the quality and autonomy of
their research might be compromised by the pressures of managing space missions, while others
fumed that the Center would be left out of exciting developments in the new field of rocket propul-
sion. One engineer warned that JPL was aggressively expanding its rocket activities, particularly in
cryogenics. These programs, he wrote, smack of the same concepts as have been pursued by
[Lewis] with hydrogen-fluorine. JPL is also in a dither to get into development of electrical propul-
sion devices.59
Silverstein ended the debate through a sweeping reorganization. He divided the laboratory
into two parts, making Bruce Lundin Associate Director for Development and Eugene
Manganiello Associate Director for Research. For the next fifteen years, Lundin played a pivotal
role in the management of Center space programs. Lundin had served as head of the Engine
57
Thomas F. Dixon to Robert C. Seamans, Jr., 19 September 1962, Centaur file, NASA Historical Reference Collection.
58
Henry Lambright, Powering Apollo: James E. Webb of NASA (Baltimore: The Johns Hopkins University Press, 1995), 108.
59
Report by Howard Douglass, The NASA Advanced Technology Programs for Liquid and Solid Propellant Rocket Engines,15
January 1962, RG 255, NARA, Box 254.
31276-chapter 2 3/8/05 12:11 PM Page 54
Research Division in charge of testing full-scale turbojet and ramjet engines from 1952 to 1957.
He was enthusiastically behind the effort to win a major piece of the space program. The reor-
ganization was intended to keep research and development entirely separate. By March 1962,
the development side of the laboratory was growing by leaps and bounds. Bruce Lundin set up a
centralized office to handle technical support for four chemical rocket development programs: the
M-1 engine for a proposed giant Nova rocket by Walter Dankhoff, the J-2 engine for Saturn by
Ward Wilcox, the F-1 engine by Fred Wilcox, and the RL10 engine for Centaur by Eugene
Baughman.60
In August 1962, Vince Johnson called Silverstein and Lundin to Headquarters to ask whether
they were willing to take over management of the entire Centaur program. Lundin recalled that
they were not particularly enthusiastic: Abe said, Well someones got to do it, and I said, Well,
well do our best.61 The decision to move Centaur from Marshall proved controversial. There were
technical, organizational, financial, and political problems. Everybody at Lewis thought we were
crazy to take that on because it was such a headache, said Bruce Lundin.62 The task was monu-
mental, and everything that could be wrong was present.
Outsiders shared this opinion. An article in the Washington Evening Star viewed the transfer as an
action that smacked of desperation.63 Journalist William Hines further wrote, In its failure to meet
deadlines, specifications and cost estimates, Centaur has poorly qualified for the title of Space Age
Turkey No. 1. He thought it unlikely that even with a DX priority could Centaur be developed in
time to prove useful for either the robotic space program or the program in planetary sciences.
Because of Centaur program delays, NASA had to send the first Mariner probe on Agena. The
reduction in the amount of weight the launch vehicle could carry made Mariner an engineering
triumph but a scientific starveling. Reflecting the unspoken opinion of many people in the
aerospace community that Centaur would be obsolete before it flies, the journalist favored using
the Saturn C-1 for future scientific missions. He questioned whether another stripped-down
Mariner would be worth flying to Mars. Bad press and skeptical Congressmen would continue to
dog Centaur even after the first Surveyor touched down on the Moon.
Even after Silverstein had agreed to take over the program, von Braun and Sparks continued
to urge cancellation of Centaur. In correspondence with Headquarters,64 von Braun and Sparks
60
Bruce Lundin to Associate Director, Lewis Research Center, Centralization of the Centers Technical Support to Chemical
Rocket Development Program, 7 March 1962, RG 255, NARA, Box 254. The J-2 and the M-1 engine for Nova both used
liquid-hydrogen fuel. Nova was canceled after it was decided to use the less energy-intensive lunar orbit rendezvous (LOR) for
Apollo.
61
Interview with Bruce Lundin by Virginia Dawson, 7 March 2000.
62
Ibid.
63
William Hines, Centaur or SaturnWhich? Washington Evening Star (15 October 1962): 10.
64
Brian Sparks to Homer E. Newell, 21 September 1962, NASA Historical Reference Collection.
31276-chapter 2 3/8/05 12:11 PM Page 55
asked clearly, strongly, and unequivocally that the Saturn C-1-Agena be substituted for Atlas-
Centaur in carrying out unpiloted lunar and planetary missions. Sparks contended that Centaur
had created a deplorable situation for both the lunar and planetary programs. NASA needed the
Saturn C-1-Agena to counter the doggedly determined effort of the Soviets and restore national
pride and international prestige. In a detailed memo appended to his letter, Sparks compared
Centaur to C-1-Agena. Despite the fact that Centaur development was in full swing, while Saturn
C-1 was only in the design stage, Sparks thought that Saturn C-1 personnel would be highly moti-
vated by their association with the piloted flight program. In contrast, the fear of cancellation of
Centaur would demoralize the Centaur team and slow their efforts.
Newell patiently responded that projections for completing the development of Saturn C-
1-Agena were overly optimistic. It had been considered and rejected. He agreed that Centaur
development had been fraught with difficulties, but he expected that the transfer to Lewis
would breathe new life into it, and that it will be able to fill the vital needs of our lunar and
planetary program on a timely basis.65 He thought it unlikely that the untested S-4 stage
could be adapted to the S-1 stage and integrated with Agena before Centaur could be made
flightworthy. During this debate, the transfer to Lewis Research Center in October 1962 went
quietly forward.
If the decision to transfer the program represented an ultimatum for General Dynamics, it
was an equally important turning point for Headquarters. In his detailed analysis of the Surveyor
and Lunar Orbiter programs in 1972, management consultant Erasmus Kloman provided some
useful insights into the change in Centaur management. He speculated that by refusing to cancel
Centaur, Headquarters (heavily staffed by former Lewis Research Center engineers at this time)
asserted its prerogative to determine launch vehicle policy. He wrote:
Headquarters, after carefully reviewing the situation, confirmed its position that the
Centaur concept was both technically feasible and essential to the launch vehicle program
for the space effort. It thus rejected the recommendation of senior management at MSFC
and JPL. Responsibility for Centaur was transferred abruptly then to Lewis Research
Center. This was interpreted as a rebuke to MSFC and a signal to the other Centers that
they could not back out of major development commitments assigned by Headquarters.66
An even more convincing reason for keeping Centaur in NASAs stable of rockets was its
role in the Apollo programboth as the upper stage for Surveyor and as a means to test liquid
hydrogens feasibility in the Saturn liquid-hydrogen/liquid-oxygen upper stages. Only its asso-
ciation with the Apollo program kept the Centaur program alive.
65
Homer E. Newell to Brian Sparks, 23 October 1962, NASA Historical Reference Collection.
66
Erasmus Kloman, Unmanned Space Project Management: Surveyor and Lunar Orbiter (Washington, DC: NASA SP-4901,
1972), 32. See also an unpublished draft, NASA Marshall Space Flight Center archives, chapter 2, 6264.
31276-chapter 2 3/8/05 12:11 PM Page 56
President Kennedy made this clear to NASA Administrator James Webb several times during
a meeting in November 1962 about supplemental appropriations. The meeting was attended by
Webb, top NASA staff, and Jerome Wiesner, Special Assistant to the President. At several points in
this meeting, Kennedy emphasized that beating the Russians to the Moon was the only reason that
he was willing to contemplate the enormous appropriations requested by NASA. Space program
contributions to science were a distinctly lower priority, for which reasonable sums of money could
be spent. But were talking about these fantastic expenditures which wreck our budget, he said,
which could only be justified by the goal of overtaking the Russians, and demonstrate that starting
behind, as we did by a couple of years, by God, we passed them.67
When Kennedy asked Webb what the nation could expect from Centaur, he responded that
Centaur would save 50 percent of the cost of the planetary shots, presumably a reference to the
high cost of developing the Agena D, the upper stage favored by von Braun. This was worth
fighting for,68 he said. But knowing that Kennedy would not accept this as the sole justification
for keeping Centaur in the rocket lineup, Webb immediately added that Centaur would allow the
government to work out problems on a small vehicle before testing the far larger liquid-hydrogen
Douglas upper stages for Saturn.69 Webb did not excuse the poor job of management of the
Centaur program by Marshall or General Dynamicss failure to recognize the magnitude of the
problems associated with liquid hydrogen. However, he was obviously furious that von Braun had
taken the unprecedented step of appealing by letter directly to Congress for the cancellation of
Centaur in favor of the Saturn C-1-Agena.70 Other participants in the meeting emphasized the
absolute necessity for Surveyor as the justification for the Centaur program.
One of the questions that Surveyor was charged with solving was the composition of the
Moons surface. Cornell scientist Thomas Gold had argued that billions of years of bombardment
by asteroids had pulverized the Moons surface into a fine powder about thirty feet deep. He
speculated that the low bearing strength of the surface would cause the Apollo lander to
founder.71 Others questioned this thesis, but lack of actual data increased the risk of sending
astronauts to the Moon. To emphasize the connection between Surveyor and the Apollo
program, Jerome Wiesner told President Kennedy,
We dont know a damn thing about the surface of the Moon. And were making the
wildest guesses about how were going to land on the Moon and we could get a terrible
67
Transcript of Presidential Meeting in the Cabinet Room of the White House, 21 November 1962, John F. Kennedy Library,
Presidents Office files, tape no. 63, NASA Historical Reference Collection, page 31 of transcript.
68
Transcript, 19.
69
Ibid.
70
Ibid, 18. This letter has not been found.
71
Robert S. Kraemer, Beyond the Moon: A Golden Age of Planetary Exploration 19711978 (Washington: Smithsonian Institution
Press, 2000), 29.
31276-chapter 2 3/8/05 12:11 PM Page 57
disaster from putting something down on the surface of the Moon thats very different
than we think it is. And the scientific programs that find us that information have to have
the highest priority. But they are associated with the lunar program. The scientific
programs that arent associated with the lunar program can have any priority we please to
give em.72
Despite the problems associated with the development of Centaur, the need to solve the ques-
tion of the composition of the lunar surface gave Centaur the priority it needed to avoid
cancellation. Making Centaur an inextricable part of the lunar landing effort was a calculated risk
on the part of NASA Headquarters. Webb refused to yield to pressure exerted by Huntsville and
JPL for cancellation. Now he was depending on Abe Silverstein, with all the resources of a research
laboratory at his disposal, to prove that it was the right decision. Silversteins stubborn advocacy
of liquid hydrogen would now be put to the test.
72
Transcript, 29.
31276-chapter 2 3/8/05 12:11 PM Page 58
31276-chapter 3 3/8/05 11:54 AM Page 59
Abes Baby 59
Chapter 3
Abes Baby
These guys had a capacity, an ability, an energy that is just not
describable . . . . They gave up their lives, their families, they gave up
everything. They had total commitment. It was an era, a period of
time, that was absolute, total commitment.
Harlan Simon, Lewis Research Center
At Headquarters, Abe Silverstein had experienced the risks and challenges of managing large-
scale projects. He seemed to have no doubt that the NACA-trained engineers at Lewis could learn
how to monitor General Dynamics and deal with the prickly payload specialists at JPLespecially
since Lewis already had a unique expertise in liquid-hydrogen technology and a Director who
would let nothing stand in the way of Centaur development. A down-to-earth midwesterner who
did not suffer fools gladly, Silverstein was known for his keen engineering insight. In contrast with
the tepid efforts at Marshall, NASA under Silverstein would move heaven and Earth to turn
Centaur into a reliable launch vehicle. Harlan Simon, a young procurement officer, called Centaur
Abes Baby, pointing out that both dedication and fear of Silversteins famous temper drove the
program: You just didnt dare let him down.1
Management of Centaur would prove both a trial and a triumph for Lewis Research Center.
Silverstein announced that he would personally direct the project with J. Cary Nettles as acting
Centaur project manager. This was a temporary arrangement until permanent leadership could be
established. Initially, forty-one Lewis personnel were assigned to the project. Nettles, a graduate of
Louisiana State University with a degree in electrical engineering, had been among the contingent of
NACA engineers transferred from Langley to Lewis in 1941. As chief of the Flight Projects Branch,
he was already serving as the Lewis representative on a NASA project to study the effects of zero
gravity on liquid hydrogen at Wallops Island, Virginia. Nettles immediately uncovered a chaotic
contracting situation. A large number of changes, not only in the contract with General Dynamics,
1
Virginia P. Dawson, Engines and Innovation: Lewis Laboratory and American Propulsion Technology (Washington, DC: NASA
SP-4306, 1991), 192.
31276-chapter 3 3/8/05 11:54 AM Page 60
AC-2, suspended by an overhead crane, is readied at the General Dynamics factory for shipment to Cape Canaveral in 1964.
(Courtesy of Lockheed Martin)
but also with Minneapolis-Honeywell for the guidance, had driven up the cost of the program.2 One
of the priorities was to get the escalating costs of the program under control.
A terse telegram from NASA Headquarters to Silverstein announced that the transfer of Centaur
from MSFC to Lewis would occur between 8 October 1962 and 1 January 1963. About forty
Marshall engineers assisted in the changeover. They were not inclined to linger in Cleveland; they
hastened back to the milder climate of Huntsville before the first snows flew. As Centaur documen-
tation began to accumulate, a skeleton project office was set up in the hangar.
NASA set an ambitious target date of late 1964 for the first Surveyor mission. Space
Technology Laboratories (STL) agreed to conduct an independent guidance analysis. This was
apparently at the insistence of JPL, which did not trust the competence of either General
Dynamics or Lewis in this area.3 Many of the engineers at STL, a spinoff of the Ramo-Wooldridge
Company, had served as technical monitors for Atlas. Silverstein pressed NASA to transfer
management of the RL10 engine to Lewis at this time. Von Braun refused, presumably because of
2
Communication from J. Cary Nettles to Virginia Dawson, June 1999.
3
Interview with Richard Martin, 5 June 2000.
31276-chapter 3 3/8/05 11:54 AM Page 61
Abes Baby 61
the RL10 engines role in the development of the S-IV liquid-hydrogen upper stages for the Saturn
C-1. Whether Atlas procurement would continue to be managed by the Air Force as government-
furnished equipment also proved contentious.4 Silverstein placed Ed Baehr in charge of an Atlas
office at Lewis.
Under Silverstein, for the first time, the government would take a strong role in managing
Centaur. Silverstein announced that the Research & Development phase for Centaur would use
eight test vehicles. This phase was expected to last through 1964 (it was not actually completed
until October 1966). Only after they had proven the reliability of Centaur would Silverstein
permit a series of eight Surveyor spacecraft to be launched. To reduce technical complexity and
interface problems, he decided that the first seven Surveyors would carry essentially the same
payload with different experiment packages. This decision meant that many of the scientific exper-
iments planned for Surveyor had to be sacrificed.
Three months later, in December 1962, Lewis Research Center took over Agena, another
troubled launch vehicle program previously managed by Marshall. The transfer was precipitated
by the failure of the fifth Ranger spacecraft to reach the Moon. Headed by Seymour Himmel at
Lewis, the Agena program never received the same attention as Centaur, but its management
contributed to the growing launch vehicle expertise among NASA staff. Not only did the program
launch Atlas-Agenas from the Eastern Test Range, but it also launched Thor-Agenas from
Vandenberg Air Force Base in California for polar and Sun-synchronous orbits. Atlas-Agena
missions included the first closeup pictures of the Moon, Mars, Venus, early communications
satellites such as Echo, and large weather satellites like Nimbus.
Atlas-Agena B could place about 5,000 pounds in Earth orbit and lift 750 pounds to escape
velocity. Atlas-Agena B launched the last series of Ranger spacecraft, which succeeded in sending
back medium- and high-resolution photographs of the Moons surface. Between 1964 and 1968,
Atlas-Agena D launched twenty payloads for NASA and the Department of Defense, keeping the
countrys lunar and planetary science program alive while the nation waited for the more powerful
Centaur.5
In February 1963, the General Accounting Office (GAO) completed its investigation of the
Centaur program. The report, based on an internal NASA Headquarters investigation by
William Fleming, was extremely critical. Fleming had worked at Lewis prior to accepting a post
at Headquarters under Silverstein. His report reflected criticism of General Dynamics that came
directly from the experience of various NASA staff, including what they perceived as the
contractors significant lapses leading up to the failure of F-1. The Fleming report concluded
that both NASA and General Dynamics should have predicted the rupture of the weather shield
4
Thomas Dixon to Robert Seamans, Jr., 4 October 1962, John Sloop papers, NASA Historical Reference Collection.
5
Lewis News (18 August 1967, 15 October 1965, 24 November 1967). NASA launched thirty-eight Agenas between 1961 and
1967.
31276-chapter 3 3/8/05 11:54 AM Page 62
from available test data. The GAOs report called this not merely an indication of error or
failure to anticipate a difficulty but rather a critical dereliction in management, since the
problem was perceived, tests undertaken to resolve the problem, and a report on test results [by
Ames Research Center] furnished but not completely utilized.6 GAO qualified this sharp crit-
icism by noting that on average, twenty flights were usually needed to establish a 50-percent
reliability for a new launch vehicle. Less significant technical lapses, such as the failure of
General Dynamics to provide an independent flight abort system for both Atlas and Centaur,
were also noted. Additionally, the GAO accused General Dynamics of improper billing prac-
tices, including failure to provide adequate documentation for fees.
The GAO was equally critical of what it called Marshalls sloppy method of contracting.
Marshall had allowed General Dynamics to set unrealistic launch dates and had failed to monitor
how much weight Centaur could lift. The GAO concluded, The most penetrating business crit-
icism related to the substantial number of uncompleted contractual actions and generally poor
documentation which, though originally attributable to poor AF [Air Force] management, was
perpetuated under Marshall management and is only now being remedied by Lewis.7
Building Confidence in Centaur
Silverstein insisted on a system-by-system review to help bring Lewis engineers up to speed

and allow problems to surface. They found problems in every area. Their first task was to build
an effective working relationship with General Dynamics engineers. In February 1963,
Silverstein appointed David Gabriel manager of the program. The project was first located in
the basement of the administration building, where Silverstein could check on it frequently.
Under Gabriel, he assigned three assistant managers: Edmund Jonash, in charge of test engi-
neering, structures, and vehicle propulsion; Cary Nettles, oversight of guidance, reliability,
quality assurance, and ground support equipment; and Ronald Rovenger, head of field opera-
tions. Rovenger, one of the few Marshall staffers to transfer with Centaur, headed the NASA
field office at General Dynamics. Under Lewis management, his office increased to forty
NASA engineers. Gabriel and Jonash, like Nettles, were old NACA hands. Gabriel, a native of
Lakewood, Ohio, had graduated from the University of Akron in 1943. He was already
familiar with Centaur problems through zero-gravity test work. At this time, he headed the
Advanced Development and Evaluation Division, which was providing technical support for a
joint Atomic Energy Commission-NASA nuclear rocket program called Project Rover.8 Jonash,
a native of Kenton, Ohio, with a degree from MIT in chemical engineering, had participated
6
Memo from Nathaniel H. Karol to Thomas W. Bugher, Acting Chief, Contract Management Branch, GAO Exit Conference
at NASA Headquarters, 7 February 1963. Box 9, NASA Glenn History Office Archives.
7
Ibid.
31276-chapter 3 3/8/05 11:54 AM Page 63
Abes Baby 63
Meeting of the Centaur team at General Dynamics, early 1964. Standing, left to right: unknown; Dr. Drew Katalinsky, Director of
Engineering for Centaur; Grant Hansen, General Dynamics Vice-president and Centaur project head. Seated, left to right: Ed
Jonash, Dave Gabriel, and Cary Nettles, leaders of the team from Lewis Research Center; Major Joe Heatherly, U.S. Air Force
(USAF); and Ron Rovenger, head of the NASA Resident Office at General Dynamics. (Courtesy of Lockheed Martin)
in testing the RL10 engine in Lewiss Propulsion Systems Laboratory. Because the group spent
much of the week in San Diego, staff meetings were held on Saturdays, the only time most
people could be expected to be in Cleveland.
Lewis staff for the project increased to 150 people. By 1964, Centaur had a budget of $10
million for in-house support. Silverstein assigned more than forty engineers and support staff to
monitor the General Dynamics plant in San Diego and at Minneapolis-Honeywell. Under the
new management, Centaur became a major program for General Dynamics, approaching 50
percent of total effort in the Astronautics Division.9
8
David Gabriel became Centaur Project manager in 1963. In 1965, he left NASA to take a job for Bell Aerosystems. In 1967, he
became Deputy Manager of the Atomic Energy Commission (AEC)-NASA Office Space Nuclear Propulsion Office (SNPO).
9
1966 NASA Authorization, Hearings before the Subcommittee on Space Science and Applications, H.R. 3730, pt. 3, March
1965, 1145.
31276-chapter 3 3/8/05 11:54 AM Page 64
A contingent of young NASA scientists and engineers recruited to work on the space program
reveled in the opportunity to learn in an environment that pushed them intellectually and gave
them responsibility early in their careers. This infusion of raw talent transformed Lewis Research
Center. Lawrence Ross, who became Director of the Center in the 1990s, began as a design and
test engineer for Centaur. He investigated Centaur from stem to stern, learning how to rewire it,
design test equipment, and run tests. This early hands-on experience provided a unique founda-
tion for his NASA career.10 Work on the sloshing of propellants brought Andrew Stofan, another
future Director, into the Launch Vehicles Division. Ross, Stofan, and other recent graduates were
trained and mentored by men who had NACA research backgrounds, as well as extraordinary
technical insight, stamina, and blind faith that Centaur would work. Harlan Simon, a thirty-two-
year-old lawyer at the time he was hired in the Centaur Procurement Office, marveled at the
dedication of Gabriel, Jonash, and William (Russ) Dunbar, who were then in their forties. These
guys had a capacity, an ability, an energy that is just not describable, he recalled. They gave up
their lives, their families, they gave up everything. They had total commitment. It was an era, a
period of time, that was absolute, total commitment.11
In December 1962, Centaur received DX priority because of its connection with the Apollo
program. This made keeping contractors to strict deadlines much easier, since the program had
national priority in terms of supplies and personnel. The most pressing concern was the need to
untangle the complex contractual relationship between the government and General Dynamics.
NASA inherited Centaur cost-plus-fixed-fee contracts from the Air Force. This included six Air
Force contracts and one NASA contract; the latter contained 120 changes with 47 in progress at
the time of the transfer. The process of making necessary NASA-directed changes needed to be
addressed.12 Each design change required the approval of the Air Force System Project Office. The
Air Force resisted changes to Centaur because they often had an impact on the Atlas design, which
the Air Force considered operational. The Air Force complicated relations between General
Dynamics and Lewis, creating bottlenecks in Centaur development and causing friction between
the two government agencies. Ultimately, when the Air Force Centaur contracts were turned over
to Lewis, NASA was allowed to purchase a bare bones Atlas from the Air Force so that the Air
Force could still control Atlas manufacturing. Atlas could then be modified to support Centaur
requirements.
10
Interview with Larry Ross by Mark Bowles, 29 March 2000.
11
Interview with Harlan Simon by Virginia Dawson, 20 March 1985.
12
Thomas F. Dixon, Deputy Associate Administrator to Robert Seamans, Associate Administrator, 4 October 1962, files of Cary
Nettles. The GD Centaur contracts had about $98 million in changes, most of which occurred prior to the contracts transfer
to Lewis. Eighty-three percent of the changes were attributed to directed changes and about 17 percent classed as overruns. See
1966 NASA Authorization, Hearings, March 1965, 1145.
31276-chapter 3 3/8/05 11:54 AM Page 65
Abes Baby 65
Because Lewis Research Center had never managed large contracts with industry, it had few
procurement officers. The Center quickly hired about half a dozen lawyers from the Cleveland
Ordinance District, Department of the Army. Nettles insisted that Len Perry, a talented African
American graduate of Cleveland Marshall Law School, handle Centaur contract negotiations with
General Dynamics. Perry worked under Edward Hicks, chief of procurement for Lewis. The
procurement team included Harlan Simon, Richard Campbell, William Monzel, Marvin Clayton,
Richard Proctor, William Brahms, and Robert Hill. Because Silverstein insisted that a new
contract would be necessary, Perry and his counterpart at General Dynamics, T. C. Courington,
began at once to review the contracts.
Silverstein personally conducted Centaur negotiations. He insisted that no Centaur would be
flight-tested until the probability of success approaches that necessary for manned space flight.13
Bruce Lundin recalled that General Dynamics, accustomed to Air Force weapons procurement, at
first objected to being held to a higher standard. I said, Naturally thats impossible, but you wont
get near it unless you require it. This is space flight and that requires perfection at every point.14
The new contract, hammered out over a period of two and a half years, became NAS3-
3232, or 32 squared, as NASA insiders called it. Jim Dempsey arrived in Cleveland to
personally meet with Silverstein as negotiations drew to a close. They ended with a walk in the
garden of Guerin House, a former private home converted to a center for meetings at Lewis.
The two men agreed on the companys fixed fee, or profit of $31,293,524. The new contract,
signed in April 1964, combined the hodgepodge of contracts inherited from the Air Force and
Marshall into one cost-plus-fixed-fee contract for fourteen vehicles and twenty-one test articles.
It included design, fabrication, and modification of ground-support equipment for the General
Dynamics factory, the Atlantic Missile Range Complex 36A, the Point Loma Test Site, the
Sycamore Canyon Test Site, the rocket site at Edwards Air Force Base, and NASA Lewis test
sites at Plum Brook and its space power facility. The estimated cost of the contract was
$321,058,005, plus the fixed fee. This contract covered known items, a specific number of
launch vehicles, and equipment. In addition, Silverstein insisted on a second contract to permit
greater flexibility in dealing with unforeseen problems that were likely to crop up during the
development of new technology under the pressure of a tight deadline. This Sustaining
Engineering and Maintenance (SE&M) contract covered modifications and improvements and
eliminated the need for costly change orders. The SE&M contract covered the costs of the work
of about two hundred General Dynamics engineers. NASA paid the direct costs of technical
13
New NASA Management Concept Aims for Higher Centaur Reliability, Aviation Week & Space Technology (21 January
1963): 37.
14
31276-chapter 3 3/8/05 11:54 AM Page 66
directions from Lewis to General Dynamics, plus an incentive fee that was determined unilater-
ally by Lewis. Silverstein himself made the quarterly evaluation of the contractors performance.
When Bruce Lundin became director of Lewis in 1969, he continued the practice.15
During the time it took to get the contracts straightened out, Lewis and General Dynamics
engineers continued to drive the technical aspects of the program forward. Dave Gabriel used
PERT to force the pace of development. PERT also found favor with Grant Hansen. In an
article in Aerospace Management in 1963, Hansen noted that PERT had made determining the
costs of the programs various subdivisions more realistic.16 Hansen liked the attitude of the new
people in charge at NASA. He recalled that when contract monitors from Marshall had come
out to San Diego, they had made it clear that they were there to check up on them, with the
implication that they could not be trusted to do the job right. Lewis management simply said,
Were here to help you do it.17 General Dynamics engineer Denny Huber noted that when the
Lewis people took over Centaur, they wanted to operate as a team with [General Dynamics]
and not act as Lord and Master over us; what a breath of fresh air.18 By degrees, Lewis engi-
neers and their counterparts at General Dynamics learned to respect each other. As General
Dynamics launch conductor Roger Lynch put it, What they did better than the Marshall guys,
is that they started to identify people they could trust.19
Direct Ascent
Silversteins key technical decision was the abandonment of the requirement for a two-burn
mission for Surveyor. He insisted that first the feasibility of a single-burn missiona straight shot to
the Moonhad to be proven. Only after direct ascent was successful would he agree to attempt a
two-burn mission with a coast period and restart in space. He reasoned that with less technical
complexity, there was less chance of failure. Direct ascent would save weight because it did not
require extra fuel and equipment to settle the propellants during the coast and for attitude control.
The previous summer, while he was considering whether Lewis Research Center should take over
the program, Silverstein had sent Art Zimmerman on a reconnaissance trip to California to discuss
whether direct ascent was feasible. Zimmerman had contacted George Solomon, the head of trajec-
tory analysis at Space Technology Laboratories in Redondo Beach. He called on Victor Clark, a scientist
at JPL, and Frank Anthony, assistant chief engineer of flight mechanics at General Dynamics. None of
15
Interview with Bruce Lundin. These contractor performance reports would be revealing if they could be found. They do not
appear to be part of the NASA Glenn records. Congressman Karth did not think an SE&M contract for Centaur was appropriate,
since $35 million for SE&M was actually for development costs. See 1967 NASA Authorization, Part 3, 583.
16
Philip Geddes, Centaur: How It Was Put Back on Track, Aerospace Management (April 1964): 2429.
17
18
Communication to the authors from Denny Huber, 2 March 2002.
19
Interview with Roger Lynch by Virginia Dawson, 5 June 2000.
31276-chapter 3 3/8/05 11:54 AM Page 67
Abes Baby 67
the three were entirely convinced; it was fortuitous that the Moons orbit happened to be in a place
within its eighteen-year cycle that made direct ascent possible in the mid-1960s.20
The decision by Silverstein to abandon the two-burn requirement initially shocked the rank
and file at General Dynamics, a contractor accustomed to the relatively hands-off management
style of the Air Force. At least two burns had previously been needed for both the Advent mission
and Surveyor. The company had focused on the more difficult approach. Now Silverstein told
General Dynamics that the company had only one requirement on Centaurto show that the
system worked. Deane Davis recalled that people at General Dynamics reacted as though they had
been asked to give up one of the Ten Commandments. It didnt make any difference whether we
were receptive or not, he said. Abe had made up his mind.21
This change to direct ascent also created consternation among scientists at JPL because it
reduced the number of launch windows, or available dates for launch. A January 1963 memo from
Benjamin Milwitzky, chief of the Surveyor program, to Oran Nicks, director of NASAs Lunar and
Planetary programs, strongly recommended that the two-burn option be emphasized in the devel-
opment phase of the program. He worried that if the Centaur program were delayed until the period
between October 1965 and February 1966, Surveyor would have to land in the dark. If these
Surveyors have to land in the dark and survive until lunar daylight before beginning operations on
the Moon, the probability of completing the scientific mission will, most likely, be significantly
reduced. Also, the possibility of obtaining TV pictures on the way down will be eliminated, he
wrote.22 Milwitzky told Nicks that JPL (probably reflecting reservations of Marshall engineers) also
questioned the reliability of the Centaur guidance system because the Minneapolis-Honeywell
hardware had not yet been tested under environmental conditions. Nevertheless, the tone of the
memo was positive. These problems could be solved.23
Although people at JPL and General Dynamics found Silversteins management style to be
heavy-handed, they soon discovered its benefits. Marshall and JPL had wrangled over the change
of weight specifications for Surveyor made necessary by the continuing problems associated with
the development of Centaur. The first order of business was to fix the weight specifications once
and for all. Silverstein asked Hugh Henneberry to head a Performance Trajectory Group (PTG)
made up of people from both Lewis and General Dynamics. Their charge was to make sure that
the weight requirements were carefully negotiated and mutually acceptable. Other members from
20
Interview with Art Zimmerman by Virginia Dawson and Mark Bowles, 5 August 1999.
John Sloop interview with Aerospace Division, Convair/General Dynamics, 29 April 1974, NASA Historical Reference
21
Collection.
22
Memo, Benjamin Milwitzky to Director, Lunar & Planetary Programs, 21 January 1963, Surveyor files, NASA Historical
23
Ibid.
31276-chapter 3 3/8/05 11:54 AM Page 68
Lewis included Zimmerman, Jack Lee, Harold Valentine, Carl Wentworth, Jack Brun, and
Richard Flage. Jan Andrews, Frank Anthony, Robert Foushee, Joseph Garside, F. W. Koester, A.
Rosin, and Richard Wentink participated for General Dynamics. Although Lewis took responsi-
bility for all assumptions and ground rules, the effort was clearly based on the greater experience
and expertise of General Dynamics. Out of this early government-industry collaboration came the
report Estimate of Atlas-Centaur Performance Capabilities for the Surveyor Mission, with
trajectory calculations and performance ground rules for both direct ascent and parking-orbit
ascent.24 The introduction to the document stated:
The philosophy adhered to throughout has been to provide a consistent set of rules
defining a vehicle as it is expected to exist in 1965 when the first operational launch is
anticipated. In many areas, insufficient data are available at present to allow a rigid defi-
nition of vehicle parameters. In spite of this, estimates have been made in all areas so that
a complete configuration is defined. This, together with the assumptions and methods of
the analysis can then form the basis for future changes or re-evaluations as they become
necessary.25
Based on weight and other assumptions agreed to by the group, nominal payload capabilities of
2,257 pounds for a two-burn mission and 2,317 pounds for direct ascent were calculated for
Surveyor missions. The Centaur vehicle for the two-burn mission had to be heavier because of the
need to carry more fuel and equipment. The group suggested that improvements in a number of
areas might allow the payload weights to increase by about 150 pounds. These weights were encour-
aging. In fact, when the missions were flown, the Surveyor model weighed 2,070 pounds, with the
heaviest Surveyor carrying 2,295 pounds.
This preliminary study represented an important first step toward preparing Centaur for
launch and gave Surveyor payload specialists at JPL definite figures on which to base spacecraft
development. It was the first in a series of memoranda, issued monthly by General Dynamics, that
included all changes to the vehicle up to the time of publication. Changes that had occurred the
previous month were printed in red. Since the monthly memorandum always had a centaur on
the front cover, the team began referring to the document as The Horse (a reference to the
mythical Centaura horse with a mans upper torso and head). Roy Roberts, General Dynamicss
principal guidance engineer on Centaur at that time, recalled that The Horse was a very, very
critical document because it documented for a large number of people the environments and the
trajectory that we were flying.26 It was distributed widely to the allied disciplines, such as thermal
Estimate of Atlas-Centaur Performance Capabilities for the Surveyor Mission, 11 February 1963, Kennedy Space Center
24
ELV Resource Library. (The authors thank Art Zimmerman for lending his copy of this report.)
25
Ibid.
26
Interview with Roy Roberts by Virginia Dawson, 21 March 2001.
31276-chapter 3 3/8/05 11:54 AM Page 69
Abes Baby 69
and loads. Under Marshall management, Centaur contractors and NASA managers had lacked
this type of cumulative up-to-date documentation. The Horse reflected Centaurs contin-
uing evolution and kept all participants informed and moving forward as a team. Both
General Dynamics and Lewis engineers could run independent simulations. It was a team
relationship without ever being adversarial, Roberts remarked. We were all in it together
and doing the same thing.27 The effort to approach problems in a collaborative spirit is what
distinguished the program under Lewis management from that of Marshall. What Lewis engi-
neers lacked in training and experience in project management they made up for in
organizational and communication skills.
Although staff at the Jet Propulsion Laboratory had once disparaged NACA-trained engi-
neers as myopically focused on pushing known principles to the next decimal point,
relations between members of the former research laboratory and employees of JPL became
more cordial as Lewis engineers established their competence and won credibility.28 Surveyors
managers at JPL found the weight and performance summary called The Horse especially
helpful. Bill ONeil, a young scientist assigned to the Surveyor trajectory and performance
group at JPL headed by Elliot Cutting, recalled that not only did The Horse prove to be a
model document that really spelled out all the ground rules for everything,29 but it also influ-
enced how JPL managed Hughes Aircraft, the contractor building Surveyor. JPL asked
Hughes to produce a companion document modeled on The Horse that spelled out the
spacecraft performance side. ONeil said, In fact, some people called it the pony. And other
people called it the rider, because it was riding on it [Centaur].
JPL and Lewis established a working group with responsibility for the interface between the
launch vehicle and the payload. The group was charged with resolving problems related to inter-
face design, interface testing, and system operations.30 Once JPL and Lewis reached agreement
on weights, launch times, and mission goals, they provided them to General Dynamics, where
they could be calculated in greater detail. Zimmerman recalled, With our optimization capa-
bility, we could design the trajectory and hand it over to them and show them this is the ideal way
to fly the mission. Then they would take their huge program and tailor it to that flight mission.
It usually came within 8 or 10 pounds of what we proposed.31
27
Ibid.
28
Clayton R. Koppes, The JPL and the American Space Program (New Haven: Yale University Press, 1982), 97.
29
Interview with William ONeil by Virginia Dawson and Mark Bowles, 9 June 2000.
Technical Report 32-1265, Surveyor Project Final Report, quoted by Erasmus Kloman, typescript, 1971, Centaur chapter, 75,
30
MSFC archives.
31
Interview with Art Zimmerman by Virginia Dawson and Mark Bowles, 5 August 1999.
31276-chapter 3 3/8/05 11:54 AM Page 70
Centaur is offloaded from a C-133 transport plane at Lewis Research Center, June 1963. (NASA C-65194)
31276-chapter 3 3/8/05 11:54 AM Page 71
Abes Baby 71
Drawing of the Centaur rocket placed in the Lewis Space Power Chamber, formerly the Altitude Wind Tunnel. (NASA Drawing
7709-EL)
Abes Commandment
Silversteins emphasis on ground testing represented a shift from the relatively hands-off
approach of the Air Force. Greater supervision by NASA initially produced friction, especially
since engineers at General Dynamics were far more experienced in rocket development than their
Lewis counterparts. Grant Hansen, head of Centaur for General Dynamics, questioned his new
Lewis managers willingness to test until everything was absolutely locked down, without any
particular regard for an ultimate end date that had to be met. Deane Davis called testing Abes
commandment. In order to enable General Dynamics to check the interface between Atlas and
Centaur before shipment to the Cape, NASA now agreed to the construction of a ten-story, $6.8-
million facility adjacent to the Kearny Mesa Industrial Park, despite the fact that General
Dynamics already had an impressive array of test facilities.
Silverstein insisted that Centaur also be tested at Lewis. Testing began with components,
followed by engine tests, then tests of the entire system, and, finally, full static tests of the entire
flight configuration. Separation tests, shake tests, structural tests, nose-fairing tests, and insulation-
panel tests all contributed to growing confidence that the Center had the facilities and expertise to
31276-chapter 3 3/8/05 11:54 AM Page 72
monitor the contractor.32 In Silversteins view, even if there were only a one-in-a-thousand chance
of failure of one component, it was better to test it than to risk failure. His technical perfectionism
was legendary. To Silverstein, space vehicles required not statistical accuracy but one hundred
percent accuracy.33 This could only be achieved by extreme diligence and the concept that every
piece of equipment that is taken aboard, every component, must be provenenvironmentally
checked so that it can live in the environment of space, the total environment of space, the vacuum
of space, the temperatures of space. Only after systematic ground testing was completed was he
willing to allow Atlas-Centaur to be launched.
Testing at Lewis not only provided independent verification of contractor performance, but
also contributed new solutions to problems. The loss of F-1 had occurred within a tenth of a
second after hydrogen gas had been vented from an opening at the top of the vehicle. A test
program in the 8-by-6-foot supersonic wind tunnel at Lewis revealed that the hydrogen-venting
system posed a fire hazard during flight. The solution was to design a vent fin or snout on the nose
fairing that extended about 50 inches from the tank, just far enough away to keep the hydrogen
gas from igniting along the hot surface of the vehicle.
As General Dynamics people began to see the concrete results of these test programs, respect
began to replace the adversarial relationship between General Dynamics and the government.34
Testing applied to whole systems, as well as individual components. The laboratory hastily converted
the Altitude Wind Tunnel into a vacuum chamber in order to test a full-scale operational Centaur
vehicle under environmental conditions to simulate spaceflight. Test engineers soaked a Centaur
in the vacuum chamber to test the separation system by firing the retrorockets in the normal and
failure modes. Lewis wind tunnels were also used to test the RL10 engine, while other liquid-
hydrogen tests were carried out in the unique Rocket Engine Test Facility at the Center. Although
needed for no more than 3 minutes, the flight gyros received 1,000 hours of testing, with the
rationale that if the gyros worked for 1,000 hours, they would run several minutes more in space
without a problem. Explosive bolts, which could not be tested, were carefully inspected for defects.
If, as Lewis engineers often pointed out, there were 1,000 single failure modes on Atlas-Centaur,
they were confident that many bugs could be uncovered with adequate testing.35
Silverstein believed that in addition to providing industry with test facilities, it was important
for NASA to build extensive new facilities for rocket tests. Centaur presented special challenges
because its engines had to fire in space. With the Centaur carrying one-of-a-kind multimillion-dollar
payloads, how could engineers be sure that the Centaur engines would start in a near-vacuum while
32
Philip Geddes, Centaur: How It Was Put Back on Track, Aerospace Management (April, 1964): 2429.
33
John Sloop interview with Abe Silverstein, 19 May 1974, NASA Historical Reference Collection.
34
Personal communication to the author from J. Cary Nettles, 17 June 1999. See also Eugene Kloman, Centaur, typescript,
MFSC archives, 6970.
35
31276-chapter 3 3/8/05 11:54 AM Page 73
Abes Baby 73
being subjected at different times to extremely hot and cold temperatures? To answer these questions,
NASA appropriated funds for the construction of a unique facility called the Space Propulsion
Research Facility, or B-2, at the Plum Brook Station, located 55 miles from Lewis Research Center
in Sandusky, Ohio. At the Plum Brooks E Site, the Dynamics Research Test Center, an entire
Centaur test vehicle could be mated to the Atlas, along with a test model of the Surveyor spacecraft.
Important bending mode tests performed by Ted Gerus, head of Dynamics, and Robert P. Miller,
project lead engineer, involved tanking the vehicle with simulated propellants and shaking it hori-
zontally and vertically to test its structural integrity. The Atlas also received axial load tests at 248,000
pounds to simulate the forces on its thin skin during liftoff. It passed with flying colors.
These tests of the steel balloon tank structure provide another example of the cooperation
between General Dynamics and Lewis engineers. David Peery, former head of the Aeronautical
Engineering Department at Penn State University and author of an important textbook on aircraft
structures, had developed a theory for structural strength available after the onset of local skin buck-
ling. At General Dynamics, he showed by analysis that the Atlas-Centaur tank designthought by
many to be too radicalwas actually unnecessarily conservative. According to his theory, it could
withstand vastly greater stresses. As Richard Martin, author of several excellent articles on Atlas,
pointed out, Peery used his famous bulldog tenacity to convince management at General Dynamics
and Lewis to test a full-scale Atlas-Centaur to the point of collapse.36 Martin explained, A bending
moment about 80 percent greater than that at the outset of buckling was achieved before significant
nonlinear deflections occurred, but there were some worrisome local buckles around protuberances
like the liquid oxygen line outlet. Therefore, the analysis was adapted to allow only about a 60
percent increase in the applied load.
Collaboration began at this time between Martin and Gerus, leaders of dynamic analysis
groups at General Dynamics and Lewis, respectively. These groups laid the foundation for an
innovative pitch and yaw program that was later implemented when a new Teledyne digital
computer was introduced after 1973. The Automatic Determination and Dissemination of Just
Updated Steering Terms (ADDJUST) program allowed design loads from flight winds to be
reduced by about 40 percent.
Silverstein insisted on a program review at Lewis once a month. Contractors were expected to
spend two or three days at the laboratory. Vince Johnson, Centaur program manager from
Headquarters, attended these meetings but was careful not to come between General Dynamics
and its new Lewis managers. Johnson had the job of defending the program to Congress and
running interference at Headquarters. In Bruce Lundins view, Johnson played the role of
36
Communication to the authors by Richard Martin, 7 March 2002. See Richard E. Martin, The Atlas and Centaur Steel
Balloon Tanks: A Legacy of Karel Bossart, 40th International Astronautics Congress paper, IAA-89-738, Cot 7-13m, 1989; A
Brief History of the Atlas Rocket Vehicle, Part III, QuestThe History of Spaceflight Quarterly 8 (2001): 48. See also David Peery,
Aircraft Structures (New York: McGraw-Hill, 1949).
31276-chapter 3 3/8/05 11:54 AM Page 74
program manager absolutely perfectly.37 His levelheaded approach to program management was
one of the keys to the programs ultimate success.
Deane Davis, General Dynamics Centaur project manager, recalled that General Dynamics
grudgingly admitted that Lewis people knew somewhat more than they did about liquid
hydrogen. The old Centaur design had required the engines to be prechilled for a short time before
Centaur engines were ignited in space. This wasted some of the precious liquid-hydrogen fuel.
One of the projects important innovations was to chill down the vehicle on the ground, using
liquid helium as a precoolant for the engine. This was a direct transfer of knowledge from Project
Bee, which included precooling the engine. Liquid helium was a relatively rare chemical at that
time, but it was much less difficult to handle than liquid hydrogen. Engineers at General
Dynamics were skeptical that it could be procured in large enough quantities. Within minutes of
leaving a contentious meeting, Silverstein was on the phone ordering a dewar of liquid helium
from the government cryogenics laboratory in Colorado. Liquid helium precooling worked
perfectly.38 General Dynamics and the Lewis engineers also worked together on a new design for
the insulation panelsthe source of failure on the first launch.
During this time, General Dynamics engineers also solved the critical problem of the leaking
of liquid hydrogen through minute pores in the welds of the liquid-hydrogen tank. Tests at General
Dynamics revealed that metal became brittle when exposed to the very low temperatures of liquid
hydrogen. It was found that by adding nickel in the weld area, the stainless steel could be strength-
ened. New techniques for lap-welding the seams, followed by spot welds, provided additional
structural integrity. Factory technicians, already highly skilled builders of the Atlas tank, had to raise
the standard even higher. The vehicle required more than 74,000 spot welds, 360 feet of resistance
seam welds, and 400 feet of heliarc fusion butt welds. The welds were carefully monitored and x-rayed
to make sure there were no imperfections.39 In the most critical area of the designthe interme-
diate bulkheadwhere cracks as small as 1/10,000 of an inch could destroy the vacuum created by
freezing out the dry nitrogen gas in the space inside the double wall, the gores were fusion-welded.40
Advanced welding techniques ensured a high degree of structural integrity without adding exces-
sive weight.
Getting to the bottom of the problem of leaks in the intermediate bulkhead led to the
discovery of another potential problem. Small cameras placed in the hydrogen tank revealed that
as soon as the liquid hydrogen was loaded, the bulkhead mysteriously wrinkled. Everyone took a
look at those pictures and just about fainted, me included, Deane Davis recalled. And of course
37
38
John Sloop interview with Aerospace Division, Convair/General Dynamics, 29 April 1974, NASA Historical Reference
Collection.
39
Irwin Stambler, Centaur, Space/Aeronautics (October 1963): 7475.
40
Ibid.
31276-chapter 3 3/8/05 11:54 AM Page 75
Abes Baby 75
Abe Silverstein got all upset about it.41 Testing determined that when the hydrogen was removed,
the wrinkles disappeared. Through further testing and analysis, the team concluded that the stain-
less steel experienced cryoshock from the very cold liquid hydrogen. The way to avoid cryoshock
was to load the liquid oxygen first, allowing the system to chill down gradually. Then the liquid
hydrogen was slowly loaded. Wrinkles in the bulkhead no longer appeared. At the same time that
they were solving these problems, engineers at General Dynamics and Lewis were building a
formidable expertise in liquid hydrogen that they freely shared with Douglas and Rocketdyne, the
contractors for the Saturn upper stages.42
New Regime at the Cape
Silverstein insisted on a major change in launch operations that had important implications
for the future of robotic operations at the Cape. Under Marshalls management, responsibility for
the rocket had ceased once it reached the launch pad. There, the Air Force director of launch oper-
ations took over. As part of the reorganization to facilitate Centaur development, the Air Force
agreed to allow NASA to manage launch operations. Goddard Space Flight Center Launch
Operations Branch at the Atlantic Missile Range had previously managed only the Delta launches.
The branch, staffed by veterans of the Navy Vanguard Earth satellite program, now assumed
responsibility for all robotic operations. (In 1965, all operations were consolidated under the John
F. Kennedy Space Center, although former Goddard personnel continued to be in charge of
robotic launches).
Headquarters placed Robert Gray in charge of the Goddard staff at the Cape. Gray had grad-
uated from Allegheny College in Meadeville, Pennsylvania, with a degree in physics after World
War II. After working at Bell Aircraft in rocket test instrumentation, he joined the Naval
Research Laboratory in 1956. He served as launch director for Vanguard launches from the
Cape, and when NASA took over, his group became part of Goddard. Gray named John Neilon,
another Vanguard veteran, as his deputy. Neilon, a graduate of St. Anselm College in
Manchester, New Hampshire, had also worked at the Naval Research Laboratory, where he eval-
uated the performance of precision location radars. Neilon would later take over as Director of
Unmanned Launch Operations in 1970, when Gray became Deputy Director of Launch
Operations at Kennedy Space Center. Gray also appointed John Gossett, another veteran of the
Naval Research Laboratory, as chief of the Centaur Operations Division. In this capacity,
Gossett acted as liaison with Lewis for Centaur. As time went on, a standard checkout proce-
dure evolved. Gossett was responsible for checking in hardware shipments, scheduling tests, and
John Sloop interview with members of the Aerospace Division Convair/General Dynamics, 29 April 1974, NASA Historical
41
42
Roger Bilstein, Stages to Saturn (Washington, DC: NASA SP-4206, 1980), 153, 188189.
31276-chapter 3 3/8/05 11:54 AM Page 76
Technicians at Pad 36A gingerly lower Centaur into the interstage adapter on top of Atlas, August 1964. (KSC 64-16746)
31276-chapter 3 3/8/05 11:54 AM Page 77
Abes Baby 77
riding herd on the contractors. He also worked with the range safety officers on documenta-
tion and launch scheduling.43
Because Silverstein strongly believed that the unusual technical challenges of liquid hydrogen
required Lewis engineers to participate in the launch, he sent Bruce Lundin down to the Cape to
make this clear to Kurt Debus, Director of the Launch Operations Center. Debus quickly agreed,
presumably because he had little interest in robotic operations on the Air Force side of the Banana
River.44 The Goddard staff found working with hydrogen challenging. Gray said, We spent years,
really, and dozens of launches before we got all the strange little things that we didnt understand
straightened out on that Centaur stage. Things like prestarts and in-flight ignitions and that sort of
stuff. We had all these chilldowns to run through and all that stuff was very, very critical.45 Preparing
Centaur for launch was like fine-tuning a Jaguar sports car. It had to be tweaked up all the time,
Gray said, and it was not tolerant to any kind of failures, or something not being just right. It had
to be right, period, or else it wasnt going to work. Gray recalled that his launch team had to invent
the launch rules to meet the unique demands of liquid hydrogen.
Mating Centaur with Atlas proved a difficult and delicate maneuver. The first time they gingerly
lowered Centaur to sit atop the Atlas, the mechanical separation latches on the interstage adapter did
not line up with the latch receptacles. These receptacles were large tabs welded to the vehicle with
slots designed to receive the latching mechanisms. Red Lightbown, General Dynamics Senior
Engineering Structures Manager for Atlas, was immediately summoned from San Diego. He discov-
ered basic problems with the separation system and decided that the whole thing should be scrapped.
A new system that used shaped charges was designed. The charge, placed inside a metal container
attached to the interstage adapter, exploded with just enough energy to slice through the adapter,
releasing Centaur from Atlas. To reduce the time it took to separate the two rockets from 45 seconds
to 6, the retrorockets on Atlas were redesigned to provide more power.46
Karl Kachigan, a longtime Centaur chief engineer, ran the launch support team for General
Dynamics. A graduate of Marquette Universitys School of Engineering, Kachigan had made his repu-
tation during Atlas development. He produced a major paper in 1955 that described forced
oscillations of a fluid in a cylindrical tank. Because Atlas had no rings inside the tank to dampen the
oscillations of the liquid propellants, it needed baffles to prevent propellant sloshing. Kachigans
contributions to the problem of sloshing proved to be a major contribution to rocket theory.
During the five or six weeks before a launch, the Cape swarmed with engineers from General
Dynamics and Lewis who worked closely with the Goddard robotic launch operations staff. At
that point, sleepy Cocoa Beach was not the vacation destination it is today. A few inexpensive
43
Interview with John Gossett by Virginia Dawson, 2 July 2002.
44
45
Interview with Robert Gray by Virginia Dawson, 9 November 1999.
46
Letter from Roger C. Lynch to Virginia Dawson, 4 June 2000.
31276-chapter 3 3/8/05 11:54 AM Page 78
motels adorned its expansive beaches. Small restaurants like Fat Boys became hot spots where the
launch teams could grab a meal of chili and hamburgers at any hour.
Kachigans leadership helped shape the early General Dynamics Centaur culture. Several
weeks before a launch, he sent a tiger team to the Cape. The members of the tiger team, key
people from the design and technical groups,47 were to muck through all the paper, look at all
the hardware, and examine all the test procedures and analyze the data. They tried to find the
slightest flaw that might compromise the launch. Kachigan headed investigations when a failure
occurred. Tough, analytical, and fair, he was universally respected for his ability to ferret out prob-
lems and make decisions under pressure.
Because the vehicle could not tolerate the slightest failure of a single component, every system
had to be checked out. Telemetrythe transmission of real-time data such as pressure, velocity,
and surface angular position by radio link from the vehicle to the groundwas an important tool
for assessing whether various systems were functioning properly. Through telemetry, the govern-
ment could independently monitor prelaunch data to make sure that what the contractor told
them was correct.
Once Atlas and Centaur were mated, each was inflated with nitrogen and held rigid with large
suspension cables to prevent collapse. The Atlas-Centaur vehicle was then ready for extensive
prelaunch testing. First, the tiger team ran the propulsion, guidance, and telemetry systems inde-
pendently. Then they simulated a launch without propellants. All this testing culminated in the
so-called tanking test, when liquid hydrogen and liquid oxygen were actually loaded. A simulated
launch of the entire system, called a wet dress rehearsal, verified that the valves were working
correctly, pressures and temperatures were right, and all other systems worked in the low-temper-
ature environment. NASA and General Dynamics carefully reviewed the data at a series of
preflight readiness reviews. Whenever a problem in the data was found, the affected system had to
be taken apart and studied, then put back together and tested.48 As the launch day neared, a tech-
nician performed a coin-tap test, literally tapping the composite payload fairing and insulation
panels of the Centaur to make sure no laminations of the skins had pulled away from the core and
left air pockets. Then, shortly before declaring the vehicle ready, launch director Robert Gray took
his ritual walk down the entire length of the vehicle and visually inspected it for any flaws that
more sophisticated testing might have missed.
On launch day, about thirty engineers huddled in Building AE, where the atmosphere was
thick with cigarette smoke and sweat. A launch erased the distinction between contractor and civil
servant, with everyone focused on the job of checking and rechecking every system. Building AE
housed the instrumentation support for all unpiloted launch vehicles. During the period between
47
Interview with Karl Kachigan by Virginia Dawson, 7 June 2000.
48
Joseph Green and Fuller C. Jones, The Bugs That Live at -423, Analog Science Fiction/Science Fact 80 (January 1968): 30.
31276-chapter 3 3/8/05 11:54 AM Page 79
Abes Baby 79
General Dynamics Launch Conductor Dan Sarokon (foreground) and Roger Lynch, Centaur Operations Manager, prepare for an
Atlas-Centaur countdown, 1964. (Courtesy of Lockheed Martin)
propellant loading and liftoff, all NASA engineers in the AE Building focused on telemetry data
spewed out on a series of strip charts. If any of the data indicated an anomaly, an engineer would
immediately inform the people in the blockhouse, who then might order the launch stopped.
Pressure was especially intense in the blockhouse, where Dan Sarokon, the General Dynamics
launch conductor; Gray; Roger Lynch, Centaur launch operations manager; and representatives
from Honeywell and Pratt & Whitney sat with about sixty other engineers. They checked the
vehicles vital signs on instruments that showed voltages, temperatures, and pressures. Sarokon
pushed the button that began the final countdown. With Atlas-Centaur underway, the director of
the NASA Lewis Launch Vehicles Division, along with other NASA engineers, sat glued to their
monitors in Building AE. Telemetry data proved to be particularly valuable if a launch failed.
Vanguard rocket veteran Skip Mackey, who served as head of the telemetry station for over forty
years, provided the blockhouse with real-time commentary on the data by telephone. All the years
of preparation for a launch came down to the push of a button, one engineer remarked: If it
goes up you are a hero. If it doesnt, youre a goat.49
49
Interview with Joe Nieberding by Virginia Dawson and Mark Bowles, 15 April 1999.
31276-chapter 3 3/8/05 11:54 AM Page 80
Left to right, Krafft Ehricke, the first Centaur Project Manager; Roger Lynch, Centaur Launch Operations Manager; Roger Lewis,
President and Chairman of the Board of General Dynamics; Grant Hansen, the new Centaur Project Manager; and Ed Heineman,
Corporate Director of Technology, wait tensely before the launch of AC-2 inside the blockhouse during a planned countdown hold, 27
November 1963. (Courtesy of Lockheed Martin)
Out of the Doghouse
It took eighteen months of intense preparation for the first test flight of Centaur under
the new Lewis management. To underline the importance of the Centaur connection with
Saturn, President John F. Kennedy visited the launch site as the launch date approached in
November 1963. Several days later, Atlas-Centaur engineers and technicians were devastated
to hear that the President had been assassinated in Dallas. NASA delayed the launch of Atlas-
Centaur 2 (AC-2) one day to allow the Atlas-Centaur launch team to mourn Kennedy and to
refocus on the arduous job of preparing for launch. Because no launch vehicle had ever
required jettisonable insulation panels, there was no knowledge to apply in their design.
Uncertainty over how they would function induced Silverstein to order them bolted to the
Centaur tank in what he called a brute-force fix.
31276-chapter 3 3/8/05 11:54 AM Page 81
Abes Baby 81
First tension (bottom), then cheers (top) erupt in the control room after the first successful launch of AC-2, 27 November 1963.
(Courtesy of Lockheed Martin)
31276-chapter 3 3/8/05 11:54 AM Page 82
On 27 November 1963, the Atlas-Centaur lifted off from Cape Canaveral at 2:03 p.m. Atlas
pushed Centaur to an altitude of 150 miles. The upper stage separated flawlessly from Atlas and
fired its two 15,000-pound-thrust RL10 engines. This first flight proved the compatibility of
Atlas with the upper stage Centaur and also proved that liquid-hydrogen engines could be fired
in space. The same day, Pratt & Whitney Aircraft issued a classified report entitled RL10 Engine
for Advanced Space Missions.50 The RL10, it announced, is a mature engine with proven dura-
bility and reliability. The companys description emphasized the simplicity and controllability of
the operating cycle.
On the heels of the Centaur launch came a test launch of the Saturn S-IV stage. It was
powered by a six-engine cluster of RL10 engines with 90,000 pounds of thrust on 29 January
1964. This liquid-hydrogen upper stage had a more conventional reinforced structure with its
insulation placed inside the rocket tank. The claim that Centaur was the first successful launch of
a liquid-hydrogen upper stage was established just in time.
Although Centaur showed results, Headquarters agonized over whether the United States
could win the race to the Moon. Lift capabilities of the United States still did not yet equal those
of the Soviet Union. Deputy Administrator Seamans asked William Fleming, Director of Program
Review, to compare the ability of the U.S. and U.S.S.R. to land unpiloted spacecraft on the Moon.
Using intelligence reports, he concluded that the total weight the Soviet Union could land on the
Moon in 1964 was 850 pounds, compared to a weight of 750 pounds for Surveyors to be launched
in 1966. An improved Centaur, slated to be flightworthy in 1967, would be able to land a
Surveyor payload weighing 1,000 pounds. He concluded, It is felt that the Surveyor program,
complemented by flights of the Lunar Orbiter beginning late in 1966 and followed by an Apollo
manned landing in 1969 would equal or exceed any program of lunar missions that the Russians
might accomplish over the corresponding period of time.51 Flemings memo led Homer Newell
to strongly recommend the development of a Saturn 1 B-Centaur vehicle to cover the needs of
larger lunar payloads and advanced planetary programs. Newell concluded, Based on Apollo
landings on the Moon in the late 1960s, the sequence of technological and scientific achieve-
ments represented by Ranger, Surveyor, Lunar Orbiter, and Apollo should offset many Soviet
spectaculars.52 However, just in case the Apollo program was delayed into the 1970s, Newell
thought that NASA might consider sending a larger unpiloted lunar spacecraft to explore the
lunar topography and return lunar samples to Earth.
Despite the reassurance he provided to Seamans, Newell was clearly dissatisfied with the progress of
the Surveyor program. To William Pickering, Director of JPL, Newell emphasized the importance of the
50
RL10 Engine for Advanced Space Missions, Pratt & Whitney Aircraft, 27 November 1963, Box 59, Old RL10 Records
(Goette files), DEB Vault, NASA Glenn.
51
William Fleming, Lunar Program Planning, 27 May 1964, John Sloop papers, NASA Historical Reference Collection.
52
Homer E. Newell to Associate Administrator, 3 April 1964, John Sloop papers, NASA Historical Reference Collection.
31276-chapter 3 3/8/05 11:54 AM Page 83
Abes Baby 83
Surveyor program to NASA. Within NASA the Surveyor project is near the top of the overall priority
list. Within the Office of Space Science and Applications, because of our commitment to provide the
needed support to the Apollo mission, as well as because of its scientific value, Surveyor is at the top of
our priority listing.53 He did not think that JPL management was giving the program adequate support.
He wanted closer monitoring of the contract with Hughes Aircraft for the spacecraft and better
communication with Headquarters over Surveyor progress and needs.
During this time, Surveyor earned a well-deserved reputation with Congress as one of the least
orderly and most poorly executed of NASA projects.54 Many of the problems stemmed from poor
planning and lax management. Representative Joseph Karth, a vocal critic of the Centaur, admitted
that Surveyor had not received enough scrutiny when it was conceived in early 1960: We were in
such a sweat to get going that we said, Lets not think about it, lets do it. I was on the space
committee then, and its true, we were impetuous.55 The cost-plus-fixed-fee contract allowed
Hughes to run up huge project costs, charging them as NASA-directed changes without penalty.
This was the first large space contract that NASA let to Hughes Aircraft. Hughes used it to learn the
problems and pitfalls of building a spacecraft.
By 1964, the multiple problems associated with Surveyor development were so serious that
NASA ordered a full review and concluded that the problems were both technical and manage-
rial. Lack of supervision at Hughes, poor oversight by JPL, and neglect by NASA Headquarters
all contributed to the inability of the Surveyor project to meet standards. For example, tests of
the terminal descent guidance control system revealed problems. The vernier motors, designed
by the Reaction Motor Division of Thiokol Chemical Company, were so undependable that
JPL sought an additional contract with Space Technology Laboratories to provide a backup.
This added more costs to the program.56
Under Lewis management, Centaur was equally plagued with a concatenation of problems
that slowed development. AC-2s bolted-down insulation panels had carried instruments that
revealed that the panels were not sturdy enough to withstand the aerodynamic forces Centaur
encountered during flight through the atmosphere. The redesign increased the weight of Centaur,
always a critical factor with respect to the amount of payload the vehicle could carry.
Ground tests revealed that the helium used to chill down the engines and purge the insulation
needed to be released at a higher rate immediately prior to launch. In addition to cooling the
engines prior to starting them in space, helium prevented the insulation from freezing to the skin
of the liquid-hydrogen tank. It was decided that an additional helium tank needed to be carried
53
Homer E. Newell to William Pickering, 13 July 1964, Surveyor files, NASA Historical Reference Collection.
54
H. R. Committee on Science and Astronautics Investigation, 1965. See also Clayton Koppes, The JPL and the American Space
Program, 173.
55
Ibid, 173.
56
Ibid, 173183.
31276-chapter 3 3/8/05 11:54 AM Page 84
on board to continue the helium purge during liftoff.57 The next Centaur (AC-3), launched in
June 1964, jettisoned the insulation panels and nose fairing successfully, but, almost immediately,
a freak failure of the hydraulic system caused the vehicle to spin out of control. Then AC-4,
launched in December 1964, tumbled wildly during the coast phase.
Through a massive effort that involved both analysis and testing, engineers traced the cause
of the AC-4 failure to uncontrolled propellant behavior after venting. Fred Merino of General
Dynamics and Lewis staffers Andy Stofan, Steve Szabo, and Ray Lacovic conducted some of the
first investigations of zero-gravity propellant management. They initiated a test program using the
100-foot drop tower at Lewis for experiments. Later, an elaborate, state-of-the-art, 500-foot zero-
gravity facility was constructed for basic research on liquid hydrogen. Merino marveled at the
openness of Lewis engineers. He often went to Lewis for a month to observe tests. There were
times when I would come up with a problem, he said, and we would have to relate it to the
people in Cleveland and give our rationale for why we thought there was a need for a design
change. And there were many times that they would run tests for us. It was their test, but we
provided input as required to get the results that were needed to effect the design change to meet
a mission requirement.58 With this type of creative collaboration, program personnel made rapid
progress on resolving problems.
The test program focused on the problem of the sloshing of propellants in flightinformation
needed for designing the autopilot controls. Through tests and analysis, scientists discovered that in
zero gravity, liquid hydrogen adheres to the sides of the tank with the ullage area (space where the
gas vapors collect) in its center. Because of the position of the ullage area, a large amount of liquid-
hydrogen fuel had been released, along with hydrogen gas, when AC-4s vent valve opened. As the
liquid flowed overboard, it expanded with a tremendous amount of evaporation and formed a ball
of gas. This ball exerted a force on the vehicle that caused it to tumble. Merino and colleague Clay
Perkins developed a patented vapor sensor to signal to the systems computer whether the valve was
surrounded by liquid or vapor, and they redesigned the vent system. This knowledge was shared with
engineers designing a propellant-management system for the liquid-hydrogen stages for Saturn.
Confidence in liquid hydrogen as a propulsion fuel was building.
The new knowledge of the ullage area also led to greater confidence in managing the propel-
lants during the coast phase. A Reaction Control System (RCS) was designed to provide
acceleration just before venting. The system consisted of a hydrogen peroxide supply bottle, lines,
valves, and small vernier motors called thrusters. Before the tank pressure reached a certain point,
small thrusters in the tank were fired to reposition the propellants in the aft end of the vehicle.
Then a valve could be opened to relieve pressure without danger of loss of liquid propellant. The
57
See Green and Jones, 27.
58
Interview with Fred Merino, 5 June 2000. See also Centaur Project Office, Coast-Phase Propellant and Vehicle Behavior,
paper presented to Low Gravity Cryogenic Fluid Mechanics Ad Hoc Committee Meeting, Convair/General Dynamics, 1618
March 1965.
31276-chapter 3 3/8/05 11:54 AM Page 85
Abes Baby 85
same thrusters were fired in preparation for starting the engine, so that liquid rather than gaseous
hydrogen and oxygen entered the pumps.
Continuing Doubts
Engineers were confident that the launch of AC-5 in March 1965 would succeed because
direct ascent was not as complicated. The checkout went forward without incident. Countdown
was equally flawless. The Atlas engines fired, the hold-down pins retracted, and Atlas-Centaur
lifted off. At 5 feet off the ground, the Atlas engines suddenly shut down and the vehicle dropped
back to Earth, causing an explosion that extensively damaged the pad. An eyewitness provided
this understated report: The burning gas cloud had an expansion velocity of over 3,000
feet/second, and furnished one of the best examples seen at the cape of why fragile human beings
retreat to massively built blockhouses during a launch. No one was injured, other than feelings,
but AC-5 was a mass of blackened rubble, and Pad A had suffered major damage and would
require extensive rebuilding.59 Failure analysis revealed that after liftoff, an improperly installed
instrumentation pressure transducer on Atlas had caused a valve to close, cutting off fuel to the
engines. Although it had nothing to do with Centaur, this explosion secured Centaurs reputa-
tion as the hard luck bird once and for all.60
The NASA Budget Authorization Hearings held the same month gave Representative Joseph
Karth the opportunity to condemn General Dynamics for an inordinately bad job on Centaur.61
Congress was unhappy about the enormous costs of the programnow between $600 and $700
million. The price of the rocket seemed to reflect NASAs dependency on General Dynamics as a
sole-source supplier. Implicit in Karths aggressive questioning was the issue of the unique design
of Centaurthe thin-skinned balloon structure and the intermediate bulkhead, in particular.
Previously, when Centaur had run into troubles, NASA had attributed them to the difficulties of
developing a liquid-hydrogen rocket. Now Marshall had tested the S-IV stage for the Saturn C-1
with the reinforced design favored by von Braun. Marshall engineers had also developed a new
Rocketdyne J-2 engine designated for the second and third stages of Saturn V. This engine burned
liquid hydrogen and produced 150,000 pounds of thrust, compared to the RL10 engines 15,000
pounds.62 Were the foibles of pressure-stabilized design, rather than the difficulties associated with its
59
See Green and Jones, 30.
60
See Green and Jones, 31. See also William Hines, Atlas Blast Laid to Impossible Mishap, Evening Star (5 March 1965).
61
Subcommittee on Space Science and Applications, 1966 NASA Authorization, Hearings, H.R. 3730, March 1965, Part 3,
11431144. NASA had also planned to use Centaur as an upper stage for the Saturn 1-B to be flown on missions after 1968;
see p. 1138. Originally, in addition to the eight development flights for Centaur, seventeen flights of Surveyor were planned, plus
six for Saturn 1-B in support of Voyager. The Saturn 1-B/Centaur program was terminated in 1968.
On the J-2, see Roger Bilstein, Stages to Saturn (Washington, DC: NASA SP-4206, 1980), 140153; Paul N. Fuller and Henry
62
M. Minami, Jr., Rocketdyne Reborn, Space 3 (1987): 5558; W. R. Studhalter, The J-2 Liquid Hydrogen Rocket Engine,
National Aero-Nautical Meeting, Society of Automotive Engineers, April 1963.
31276-chapter 3 3/8/05 11:54 AM Page 86
liquid-hydrogen propellant, holding back Centaur development? Karth thought General Dynamics
was the real problem and cited past problems with the intermediate bulkhead as evidence of a poor
job. This has been a very favored program. It has, if you please, had a DX priority for 3 years, the
highest possible. Tremendous effort has gone into it.63 He thought that NASA should consider an
alternative source, such as Titan III with a high-energy upper stage. At this time, JPL was agitating
for launching Surveyor on the Titan III with an upper stage called the Transtage. Congressman
Weston E. Vivian of Michigan supported the idea of creating some competition among aerospace
companies:
It would seem to me you have a program heavily dependent on Centaur, specifically in

the unmanned spacecraft area. You are planning procurement of several hundred million
dollars more equipment of this variety. You have a single contractor on whom your hopes
rest totally, and whose dollar estimates will control your dollar estimates. I would frankly
feel a competitive program should be initiated, and even though the apparent cost
initially might be higher, I would certainly consider that you should open bids for
procurement of a given number of Centaur vehicles, say one-third of the total you are
planning to buy, from an alternate source.64
In response, Homer Newell reaffirmed NASAs commitment to General Dynamics and the
companys unique expertise in manufacturing the thin-skinned structure: The capability of
building thin-wall pressure stabilized tankage to our extremely high standards has been highly
developed at Convair and is not currently available elsewhere. Because of these considerations,
DOD and NASA have concluded that it is not practicable to consider alternate sources for either
Atlas or Centaur.65
This congressional pressure for more competition was not lost on Martin Marietta-Denver.
The president of the company, I. Nelvin Palley, had already offered its Titan IIIC to NASA as an
alternative to Atlas-Centaur. The company had also raised this possibility with DOD and inter-
ested members of Congress.66 After the first flight of Titan IIIC proved successful in July, the Air
Force offered to carry Surveyor to the Moon on Titans next flight, scheduled for September.67 No
mention was made of the fact that Surveyor was far from being ready.
63
Subcommittee on Space Science and Applications, 1966 NASA Authorization, Hearings, H.R. 3730, March 1965, Part 3,
11431144.
64
1966 NASA Authorization, Hearings, 1140.
65
1966 NASA Authorization, Hearings, 1147.
66
Earl Hilburn to Robert Seamans, Jr., 22 March 1965, Surveyor files, NASA Historical Reference Collection.
Titan 3C Offered to Soft Land Surveyor, Cocoa Tribune (21 June 1965). See also William J. Normyle, Congress May Force
67
NASA Titan 3 Study, Aviation Week & Space Technology (12 July 1965): 30.
31276-chapter 3 3/8/05 11:54 AM Page 87
Abes Baby 87
The spectacular explosion of AC-5 in March 1965 left Pad 34A a mass of blackened rubble. (NASA Glenn Research Center
unprocessed photo)
31276-chapter 3 3/8/05 11:54 AM Page 88
Because of the damage to the launch pad from the explosion of AC-5, launch complex 36B
was rushed into service, making the launch of AC-6 possible on 11 August 1965. This was the
last R&D flight of the direct-ascent program. The Atlas-Centaur operated flawlessly, success-
fully injecting a dummy of the Surveyor spacecraft into a prescribed transfer trajectory and
simulated impact with the target, the so-called paper Moon. AC-6 flight-tested an improved
propellant control system. Clevelands Plain Dealer headline announced, Paper Moon Shot
Gets Rocket Out of Doghouse.68 The article pointed out that the Russians had failed twice that
year in soft-landing attempts, then added that three of the previous five tests of Centaur had
also failed. The feasibility of Centaur was still a matter of opinion. In response to inquiries from
the ever-vigilant Representative Karth, Seamans responded defensively:
The slippages in schedule and the increases in program cost are serious problems; but it
must be realized that this mission is one of the most difficult technical assignments we
have undertaken in the exploration of space. We are continuing to do everything we can
to maximize the chances of successful engineering flights in 1966 and of scientific missions
in 1967. To accomplish these objectives, it may become necessary to reprogram additional
funds in FY 1966 for the Surveyor project.69
The relentless criticism by the press and Congress continued. In March of 1966, the
Committee on Aeronautical and Space Sciences again complained that generous funding for both
Centaur and Surveyor had produced less than spectacular results. Legislators questioned whether
the Apollo programs need for Surveyor had passed. In his confidential testimony before the
committee, Newell responded that NASA still regarded Surveyor as important for verifying the
design concepts used in the Apollo Lunar Excursion Module (LEM). The first Surveyor would
also provide information on landing loads at surface impact. Its TV camera would locate protu-
berances and depressions in the lunar terrain, areas to avoid when selecting an appropriate landing
site.70
What Surveyor could no longer guarantee was the quality of the science it would return.
Narrow Apollo objectives had compromised plans by scientists to use Surveyor to acquire exten-
sive knowledge of lunar geography and geology. Ronald Scott, a scientist at Caltech,
complained that the scientists recognized that Surveyor is now essentially a back-up mission
for Apollo.71 Milwitzky agreed that the massive difficulties in getting the basic Surveyor
system to work had drained both manpower and attention away from the scientific missions
68
Plain Dealer, 12 August 1965.
69
Robert Seamans, Jr., to James Webb, 10 November 1965, John Sloop papers, NASA Historical Reference Collection.
70
Confidential testimony, Homer Newell, 2 March 1966, John Sloop papers, NASA Historical Reference Collection.
71
Ronald Scott to Benjamin Milwitzky, 22 December 1965, Surveyor files, NASA Historical Reference Collection.
31276-chapter 3 3/8/05 11:54 AM Page 89
Abes Baby 89
and changed its basic goals. If Apollo remained on schedule, the role of Surveyor would be
limited to the determination of the Moons surface. Should it turn out to be a surface that would
support a lunar landing, this would be an important contribution in view of the Apollo
programs national importance. If they were to discover that the surface was relatively soft, more
flights of Surveyor might be necessary to provide additional data. Milwitzky speculated that the
Apollo mission might be delayed sufficiently to allow the scientific flights to get to the Moon
first. He reassured Scott that the science missions were an essential part of NASAs lunar explo-
ration program: there is no intent on our part to cancel these missions or let them slowly wither
away on the vine.72
Although Centaur was ready for a direct shot to the Moon, Surveyor was still mired in
development problems. Silverstein decided to launch another R&D flight to test Centaurs
restart capability. Unfortunately, AC-8, launched in April 1966, did nothing to dispel the
concerns of the scientific community and Congress. A leak in the hydrogen peroxide system
used to settle the propellants made it impossible to start one of the Centaur main engines after
the coast phase. This unbalanced thrust sent the vehicle into a tumble. Space Daily dryly
reported, The Centaur still has not demonstrated its parking orbit capabilities, the absolute
requirement to justify the development of the Centaur transportation system.73 Yet to the engi-
neers involved in the redesign of the baffles and vents, the system was finally satisfactory. The
leak in the hydrogen peroxide system had nothing to do with the restart capability. They were
convinced that the restart problem was solved.
To increase its leverage with Pratt & Whitney, Lewis Research Center was finally able to take
control of management of the RL10 engine from Marshall in May. A new office, managed by
William Goette with a staff of four, was set up. The office would be responsible for testing the
more advanced RL10A-3-3, qualified in November 1966.
A Soft Landing on the Moon
At the end of May, Surveyor appeared ready for a first flight, although NASA acknowledged
that a large number of last-minute changes to Surveyor equipment increased the risk of failure.
Although the mission might not achieve its primary objectives, it would yield enough useful data
to permit greater confidence in overcoming some of these uncertainties.74 After a nearly flawless
launch on a bright Memorial Day morning, AC-10 hurtled skyward carrying the first Surveyor.
The Centaur RL10 engines ignited 4 minutes 12 seconds after liftoff. They operated for 7 minutes
72
Benjamin Milwitzky to Ronald Scott, 17 January 1966, Surveyor files, NASA Historical Reference Collection.
73
Space Daily (11 April 1966), Centaur files, NASA Historical Reference Collection.
74
Earl D. Hilburn to Dr. Seamans, Special meeting to ascertain launch readiness of Surveyor, 25 May 1966, Surveyor files,
NASA Historical Reference Collection. For a technical overview of the Surveyor Program, see J. Jason Wentworth, A Survey of
Surveyor, Quest (Winter 1993): 416.
31276-chapter 3 3/8/05 11:54 AM Page 90
The Fourteen Launches of the Atlas-Centaur for the Surveyor Program

(19631968)
Launch Date Mission Vehicle Payload Objective
27 November 1963 R&D AC-2 No payload Demonstrate separation.

(one-burn)
30 June 1964 R&D AC-3 No payload Demonstrate jettison of

(one-burn) insulation panels and nose
fairing. Failed.
11 December 1964 R&D AC-4 2,070 lb Demonstrate coast-phase

(two-burn) propellant control.
2 March 1965 R&D AC-5 1,411 lb Demonstrate operational

(one-burn) readiness. Failed.
11 August 1965 R&D AC-6 2,100 lb Demonstrate operational

(one-burn) readiness in direct ascent.
7 April 1966 R&D AC-8 1,730 lb Demonstrate coast-phase

(one-burn) propellant control. Failed.
30 May 1966 Surveyor 1 AC-10 2,193 lb First operational Atlas-Centaur

(one-burn) flight. First controlled
unpiloted landing on the
Moon. First pictures from the
Moon.
20 September 1966 Surveyor 2 AC-7 2,204 lb Second soft landing on the

Moon. Postlanding TV survey.
26 October 1966 R&D AC-9 1,740 lb First engine restart of

(two-burn LH2/LO2 engines. Centaur
simulated operational for two-burn
lunar transfer) missions.
17 April 1967 Surveyor 3 AC-12 2,281 lb Perform soft landing within

(two-burn) Apollo landing zone.
Manipulate lunar surface
with soil sampler.
14 July 1967 Surveyor 4 AC-11 2,295 lb Perform soft landing at Sinus

(two-burn) Medii. Conduct vernier
engine experiment.
8 September 1967 Surveyor 5 AC-13 2,217 lb Perform soft landing on Mare

(two-burn) Tranquillitatis. Determine
chemical elements in soil.
7 November 1967 Surveyor 6 AC-14 2,220 lb Perform soft landing at

(two-burn) Sinus Medii.
7 January 1968 Surveyor 7 AC-15 2,289 lb Perform soft landing at Tycho.

(two-burn) Last Surveyor mission.
31276-chapter 3 3/8/05 11:54 AM Page 91
Abes Baby 91
General Dynamics launch team members at Kennedy Space Center (left to right): Tom Henry, Roger Lynch, and Deane Davis hear
the good news of the successful touchdown of Surveyor 1 on the Moon, June 1966. (Courtesy of Lockheed Martin)
18 seconds before engine cutoff and release of the spacecraft. Then engineers and scientists 3,000
miles away at JPL in Pasadena, California, took over. They monitored data in the JPL Space Flight
Operations Facility (SFOF), where there were banks of computers for analyzing trajectories and
telemetry data and sending commands to the spacecraft. Their connection to the Deep Space
Information Facility kept JPL engineers in radio contact with tracking stations around the world.
The next day, the spacecraft made what was referred to as a mid-course correction to place it within
9 miles of its target. Cary Nettles recalled that what the Surveyor actually performed was not a mid-
course correction, but reverse and forward firings to dump extra fuel before the craft landedthe
result of a guidance miscalculation by JPL. He said:
The tracking stations were indicating that the Centaur guidance had performed perfectly
and the spacecraft was right on target. This situation came as a complete surprise to JPL
and they needed to do something to get rid of extra fuel they had loaded on for their own
estimate of what the guidance was going to do.75
75
Letter from J. Cary Nettles to Virginia Dawson, 15 June 2002.
31276-chapter 3 3/8/05 11:54 AM Page 92
No one had expected both Centaur and Surveyor to operate flawlessly on the first try. The
team at JPL scrambled to coordinate the landing as the spacecraft neared the Moon after trav-
eling 230,000 miles in 63 hours. Its braking rockets slowed it from nearly 6,000 miles per hour
to a bare 3.5 miles per hour. Then, 14 feet from the lunar surface, the rockets cut off and the
vehicle dropped to the Moons surface. All three footpads hit the surface at almost the same
time and penetrated to a depth of a few inches. The spacecraft rebounded once before coming
to rest about 35 miles north of the crater Flamsteed in Oceanus Procellarum on 2 June 1966.
After 40 minutes of suspense, jubilant cheers sounded at JPL, Hughes, General Dynamics,
Lewis Research Center, and Kennedy Space Center when the first pictures were flashed back.
Silversteins decision to go with direct ascent was vindicated.
Surveyor operated for two lunar days (about twenty-eight Earth days) before its batteries went
dead. During this time, Surveyor transmitted about 15,000 striking pictures of the solar system,
including the Suns corona out to at least three or four solar radii and other celestial bodies
including Sirius, Canopus, Jupiter, and Gemini. As the lunar night closed in, Surveyor recorded
the rate of temperature decay. In their final positioning, Surveyors solar panels faced west so that
the craft would cast a long shadow at lunar dawn for the Apollo astronauts to see during their
forthcoming mission.
The performance of Centaur made the Surveyor mission possible. Placed in its trajectory by
Centaur with near-pinpoint accuracy, Surveyor hit the Moon surprisingly close to its target. A wire
from Deputy Administrator Seamans congratulated the Lewis team. Referring to the transfer of
Centaur to Cleveland three and a half years earlier, Seamans wrote, The achievement of the
Centaur organization, both government and contractor, is particularly striking when the technical
difficulty and complexity of the Centaur development is properly understood. At times, some of
these difficulties must have appeared to be nearly insurmountable. However, the outstanding tech-
nical competence of the people and the dedication which they have given to the Centaur project
seems to have reduced all these difficulties to manageable size.76
The United States had achieved its first soft landing of a spacecraft on the Moon, but the
Soviet Unions Luna IX had already beaten Surveyor to the Moon by over three months. Far
simpler mechanically than Surveyor, Luna IX had a crushable structure that was damaged on
impact. Thus, NASA could claim that Surveyors landing was the worlds first controlled soft
landing. From the few pictures that Luna IX sent back, scientists were relatively sure that a space-
craft would not sink into lunar dust, but Surveyors thousands of pictures were sharper and clearer.
Gone were doubts about whether the surface could support the weight of the Lunar Excursion
Module. Because Surveyor was intended to demonstrate the soundness of the basic landing tech-
76
Lewis News (10 June 1966); The Orbit (24 June 1966) (Pratt & Whitney Aircraft, Florida Research and Development Center).
31276-chapter 3 3/8/05 11:54 AM Page 93
Abes Baby 93
nique the LEM would use, astronauts Deke Slayton, Edwin (Buzz) Aldrin, Thomas Stafford, and
Eugene Cernan witnessed the first Surveyors touchdown from JPLs control room.
The next launchAC-7 carrying Surveyor 2, an engineering test modelwas launched from
the Cape in September 1966. This time, Surveyors vernier engines failed, causing it to tumble out
of control. It crashed on the lunar surface at a velocity of 8,737 feet per second. Despite this
setback, Centaurs increasing reliability enabled NASA to justify an order of five additional
Centaurs from General Dynamics on 23 September 1966. This time, the new contract for
$15,565,331 was fixed-price, a controversial but astute way to control Centaurs costs. For their
superhuman effort in turning the troubled rocket into a reliable upper stage, Cary Nettles and Ed
Jonash (who took over as head of the Launch Vehicles Division when David Gabriel left) received
NASAs Distinguished Service Medal in early October as they readied AC-9 for launch.
The next Centaur, launched on 26 October 1966 from Complex 36B, definitively proved
Centaurs capability to restart its liquid-hydrogen engines in space. The vehicle carried a 1,750-
pound dummy model of the Surveyor spacecraft. Centaur was able to compensate automatically
for the shutdown of Atlas sustainer engine 6 seconds early. Boosted into a 90-nautical-mile
parking orbit by its first burn, AC-9 coasted in a circular orbit under low-gravity conditions for
24 minutes. The main engines of Centaur ignited, propelling the vehicle into the proper simu-
lated lunar intercept trajectory. Although the Agena upper stage had proved its restart capability
five years earlier, the ability to restart liquid-hydrogen engines in space was hailed as an
outstanding achievement. With a coast period, Centaur launches were no longer restricted to
the summer months but could take place throughout the year. The restart capability also
increased the amount of available launch time each day.
The flight also verified the reliability of the inertial guidance system to control an indirect-
ascent mission to the Moon. Homer Newell wrote to Abe Silverstein to congratulate the Center
on this achievement: Your Center may take justifiable pride in the aggressive manner in which
this most difficult task was undertaken and the professional manner in which it has been
completed.77 Edgar Cortrights congratulatory letter stated:
Recalling the situation at the time you undertook to manage Centaur, I cannot think of
a single other group in the country which would have accepted that responsibility under
such dire circumstances. I can still recall your personal response to me when I asked you
to consider itwhich was, in effect, Someone has to do it. Having watched Centaur
development fairly closely, I am acutely aware of the personal sacrifice made by many
Lewis employees, not to mention their families, in completing this development.
Although it has been tough, I cant help but feel that most of them will look back on these
77
Homer E. Newell to Abe Silverstein, 10 November 1966, John Sloop papers, NASA Historical Reference Collection.
31276-chapter 3 3/8/05 11:54 AM Page 94
years in the way many of us look back on our military servicenamely, they hope they
never have to do it again, but they wouldnt have missed it.78
Newspapers jumped on the bandwagon. A powerful Atlas-Centaur rocket shattered a long-

time scientific jinx today and handed the United States new muscle to lift heavy payloads to the
Moon and planets, reported the Chicago Tribune on October 27. NASA hailed Centaur as the
worlds first operational rocket using liquid hydrogen and liquid oxygen for propellants.79
Behind the scenes, however, analysis of the data from this flight revealed some disquieting
facts. Besides the premature shutdown of the Atlas sustainer engine, one of the Pratt & Whitney
engines had experienced an unexplained drop in performance. Although it had not affected the
outcome of the mission, it was a bad sign. Another problem was that the hot exhaust from
Centaurs nozzles was impinging on the turbine pumps, causing them to warm more than they
should. As a precaution, protective shields were added to prevent damage from exhaust gases.
Surveyor III (AC-12), launched in April 1967, marked the first operational use of Centaurs
restart capabilityonce for initial boost after Atlas separation and again to launch the spacecraft
into a lunar trajectory after a short parking orbit that lasted approximately 25 minutes. After AC-
12 released Surveyor, the spacecraft received a command to lock onto the Sun and Canopus, the
largest star in the southern hemisphere. Surveyor coasted until it reached the point for its mid-
course correction. Then one of the tracking stations sent the command to orient the spacecraft
into its landing position. It landed only 3 miles from its planned targeta point on a moving
object 237,000 miles away from Earth. The spacecraft took several unplanned, very large hops
before coming to rest at an angle inside a lunar crater strewn with rocks on the Ocean of Storms.
Surveyor III sent back thousands of pictures, including a picture of its own footprints. In addi-
tion to a camera, Surveyor carried a soil scratcher to scoop up lunar soil and break lunar rocks.80
By examining photographs of the digging operations of its tiny scoop, scientists were able to
calculate the bearing strength of the Moons surface.
Surveyor V (AC-13), launched in September 1967 on an improved Atlas, landed on the Sea
of Tranquility, 18 miles from its target. Experiments using an alpha scattering spectrometer, a
device that bombarded the surface of the Moon with subatomic particles, proved that the lunar
surface consisted of volcanic rock basalt with high titanium content. This was similar to soils
found on Earth. Surveyor V sent back 18,006 pictures of terrain under consideration for the
landing site for Apollo. Gone were any doubts that a controlled landing on the Moon could be
accomplished. Surveyor VI (AC-14) explored another potential landing site and confirmed the
78
Edgar Cortright to Abe Silverstein, 28 October 1966, and Chicago Tribune article, both reproduced in Centaurs Tenth
Anniversary (Cleveland, OH: limited edition, 1972).
79
1968 NASA Authorization, Hearings, March 1967.
80
Lunar Landmarks, Surveyor Project Newsletter (24 April 1967), NASA Historical Reference Collection.
31276-chapter 3 3/8/05 11:54 AM Page 95
Abes Baby 95
Apollo astronaut Charles Conrad, Jr., examines Surveyor III (landed 19 April 1967) with the Lunar Module 600 feet away on the
Ocean of Storms, 1969. (NASA AS12_48_7136)
composition of the lunar soil. Then the vernier engines were ignited and the spacecraft flew to
another location about 8 feet away.
By the seventh and final launch of Surveyor (AC-15) in January 1968, NASA had already
selected all the landing sites for Apollo. Since Apollo support was superfluous, NASA decided to
make the last Surveyor a completely scientific mission. The Surveyor team selected a site near
Crater Tycho for the spacecrafts long arm to scratch and dig. Geologists hailed the results. Despite
its many troubles, the Surveyor program, managed by JPL, proved to be a triumph for NASA and
a highlight in the careers of the people involved in this ambitious undertaking. In reflecting on
the significance of Surveyor, Oran Nicks wrote:
It was as recently as 1966 that Surveyor 1a robot that, on its first flight, flawlessly made
a radar-controlled landing on another body in the solar system, and turned its television
eye to stare about with the insatiable curiosity of its creatorsshowed that men setting
foot on the Moon would not sink in a quick sand of Moondust. Since then we have
learned that our expendable machines can be asked to perform tasks of remarkable deli-
31276-chapter 3 3/8/05 11:54 AM Page 96
cacy, and that our astronauts can be sent off with higher confidence of safe return than
the hardy sailors who had first crossed the unknown oceans.81
Beyond demonstrating that the surface of the Moon was hard enough for the astronauts to
land safely, Surveyor proved the usefulness of robotic spacecraft to perform complex tasks. Flight
controllers learned how to repair spacecraft by remote control. Surveyor also served as a prototype
for more advanced spacecraft, such as the Viking Mars lander.82
Even more important, but far less heralded, the missions upper stage proved the feasibility of
liquid-hydrogen fuel for transporting delicate spacecraft to a designated point in space. The
RL10 engine was the first liquid-hydrogen rocket engine to be successfully restarted in space
(October 1966). It proved that with a coast and restart, a spacecraft could be positioned for an
optimal trajectory to another body in the solar system. With knowledge gained from Centaur,
the designers of Saturns J-2 engine for the upper stages of the Saturn V could move forward
with greater confidence. Among the most significant contributions to basic knowledge of liquid
hydrogen were the solutions to propellant sloshing and the venting of hydrogen gas in zero
gravity.
Silversteins advocacy of liquid hydrogen and his leadership proved to be crucial not only for
the success of the Surveyor program, but also for the development of the Saturn rocket. Referring
to the first launch of a Saturn V on the Apollo 4 mission on 9 November 1967, von Braun cred-
ited Silverstein with pioneering work in liquid hydrogen technology [that] paved the way to
todays success.83 This was not an overstatement because Silversteins unflagging advocacy had
won von Brauns support for liquid hydrogen upper stages for the Saturn V. When Apollo 4
demonstrated the restart capability of the J-2, this decision was vindicated.84
At many points in Stages to Saturn, historian Roger Bilstein refers to the knowledge of liquid
hydrogen that the designers of the Rocketdyne J-2 engine and the Douglas S-IV and S-IV B stages
received through contact with the engineers working on Centaur. Douglas designers were able to
benefit from many of the concepts pioneered on Centaur because NASA required its contractors to
share their know-how with competitors.85 For example, Rocketdyne engineers tried to use a familiar
flat-faced copper injector design for the J-2, but the high operating temperatures of liquid-
hydrogen/liquid-oxygen propellants caused it to burn out. They resisted adopting the porous face
injector design of Pratt & Whitneys RL10 engine until a demonstration of its unique Rigi-Mesh
81
Oran Nicks, ed., This Island Earth (NASA SP-250, 1970), 166167.
82
Jason B. Wentworth, Survey of Surveyor, Quest 2 (Winter 1993): 416.
83
See signed photo in Virginia Dawson, Engines and Innovation, 194.
84
Roger Bilstein, Stages to Saturn, 359.
85
See Roger Bilstein, Stages to Saturn, 188189; and Oswald H. Lange, Development of the Saturn Space Carrier Vehicle,
in Ernst Stuhlinger et al., Astronautical Engineering and Science from Peenemnde to Planetary Space, 21.
31276-chapter 3 3/8/05 11:55 AM Page 97
Abes Baby 97
Surveyors Cold War context is reflected in a cartoon from the Philadelphia Evening Bulletin (22 April 1967): 135. (Courtesy of the
Urban Archives, Temple University, Philadelphia)
31276-chapter 3 3/8/05 11:55 AM Page 98
injector face at Lewis Research Center convinced them. The injector proved to be the solution to
their problem.86 In considering the type and amount of insulation required to maintain tank
temperatures at -432F, S-IVB designers at Douglas consulted with people in the Convair Division
of General Dynamics. However, after thoroughly studying the Centaur external insulation, they
adopted the less technically complex but heavier solution of placing the insulation inside the tank.87
Douglas and Rocketdyne engineers were forewarned of the propensity of liquid hydrogen to leak
through minute pores in metals. They also learned precious lessons on venting and settling propellants
prior to a restart in space. Bilstein wrote:
Although the liquid hydrogen engines were developed and built by two different contrac-
tors, the government managed both programs so that information from one program was
available to subsequent programs. Lewis Research Center, NASAs facility in Cleveland,
represented an interesting intermediary influence, providing a pool of knowledge about
liquid hydrogen technology used by Pratt & Whitney and Rocketdyne alike. Just as early
work at Lewis was a benefit to Pratt & Whitneys RL-10, Rocketdynes later J-2 benefited
from both Pratt & Whitney and Lewis.88
This expertise in liquid-hydrogen technology might have been lost had Webb decided to
yield to pressure for the cancellation of Centaur. In contrast to the American success with liquid
hydrogen, the Russian rocket designer Sergei Korolev had constantly urged the Soviet govern-
ment to support the development of liquid-hydrogen/liquid-oxygen upper stages but never
received adequate funding. Author Asif Siddiqi makes the point several times in Challenge to
Apollo: The Soviet Union and the Space Race that the Soviet Union had neither the technical
expertise nor the manufacturing resources in the early 1960s to develop a liquid-hydrogen rocket.
It also lacked adequate ground-test facilities. Siddiqi considered the failure to develop liquid-
hydrogen upper stages a strategic mistake that cost the Soviet space program much in terms of
capability and efficiency, but Korolev alone did not have the force to single-handedly create a new
industry in the Soviet Union.89 Siddiqi points out that the first ground test of a Soviet liquid-
hydrogen/liquid-oxygen engine was not carried out until April 1967. The Soviet Union was at least
six years behind the United States in developing what he considered a critical area of rocket engine
technology.90
86
See Roger Bilstein, Stages to Saturn,145.
87
Ibid., 173.
88
Ibid., 153.
Asif Siddiqi, Challenge to Apollo, 318. A similar point is made by Tom D. Crouch in Aiming for the Stars: The Dreamers and
89
Doers of the Space Age (Washington, DC: Smithsonian Institution Press, 1999), 179180.
90
Siddiqi, Challenge to Apollo, 548. Siddiqi notes on p. 840 that the actual launch of the Soviet liquid-hydrogen/liquid-oxygen
rocket did not occur until May of 1987.
31276-chapter 3 3/8/05 11:55 AM Page 99
Abes Baby 99
Left to right: Abe Silverstein; Fred Crawford, President of TRW; and former NASA Administrators T. Keith Glennan and James
Webb at the dedication of the Glennan Space Engineering Building at Case Institute of Technology, January 1969. (Courtesy of
Rebman Photo Service, Inc.)
On the tenth anniversary of the transfer of Centaur from Huntsville to Cleveland, Grant
Hansen reminded Bruce Lundin, then Director of Lewis Research Center, of how close Centaur
had come to being canceled. He regarded the transfer to Lewis as the key decision in saving the
program. Reveling in the role they had played in pioneering the development of the worlds first
liquid-hydrogen rocket, he wrote, Were it not for its being overshadowed by the Moon landing
program, I believe that the Centaur story would have been the prime technical achievement of the
last decade.91 Homer Newell linked Centaur with the end of the anxiety over Americas launch
vehicle capability. He wrote, By 1966when Centaur became fully operationalthe United
States could at last launch spacecraft for just about any space mission . . . . Although the debate
over whether the United States could or could not match Russian launch capability still arose
occasionally, the subject no longer had the importance that it once did.92
91
Letter from Grant Hansen to Bruce Lundin, 7 October 1972, reproduced in Centaurs Tenth Anniversary (Cleveland, OH:
limited edition, 1972).
92
Newell, Beyond the Atmosphere, 140.
31276-chapter 3 3/8/05 11:55 AM Page 100
By the late 1960s, the Civil Rights movement and President Lyndon Johnsons War on Poverty
had turned national attention away from the imperatives of a space race with the Soviet Union. The
war in Vietnam sapped the energies of the Johnson administration, leading to the Presidents deci-
sion not to run for another term. At the same time, James Webb, NASAs capable Administrator,
lost President Johnsons ear. He had kept alive Kennedys dream of landing human beings on the
Moon, but he resigned before the historic flight of Apollo 11 in August 1969.
Krafft Ehricke still firmly believed in industrialization of the Moon and piloted flight to the
planets. However, in the face of environmentalists and zero-growth advocates of the early 1970s,
his enthusiasm for space tourism, orbiting hotels, and zero-gravity sports may have seemed more
eccentric than prophetic. He continued to promote space travel as a means of transcending the
petty divisions among human beings on Earth, but with the imminent cancellation of Centaur
and diminishing support for NASAs nuclear propulsion program (which he thought held the key
to distant travels in space), he sensed a new kind of disillusion that eroded confidence in what
he referred to as mans extraterrestrial imperative.93
Ironically, although Centaurs close identification with the Apollo program had initially saved it
from cancellation, the successful conclusion of the Surveyor program increased its vulnerability.
Centaur had won funding during a time when the urgency of the space race with Soviet Union had
provided the NASA with ample funding for facilities and personnel to support the development of a
radical technology. Centaur had proven not only the feasibility of liquid hydrogen, but also the
viability of the concept of government partnership with industry in developing the space agencys
technical competence. However, declining budgets for the space program and lack of unity among
scientists over long-term planning threw the planetary program temporarily into limbo.94 With no
future orders for Atlas-Centaur, General Dynamics prepared to shut down its factory. On 6
February 1968, W. L. Gorton, Division Vice-President and General Manager of the Pratt &
Whitney Development Center, informed Abe Silverstein at Lewis Research Center that within
months, they would complete delivery of all RL10 engines on order. Key RL10 engine personnel
would have to be reassigned and production space turned over to other projects. He lamented that
the nation was about to lose a precious technical capability in liquid hydrogen. Although the
special manufacturing and testing facilities can be retained and stored in whatever fashion is most
practical, there is no way to mothball the know-how, the production team, and the factory space,
he wrote. Previous experience with other engines has demonstrated that once the production
team is dispersed, it is most difficult and costly to assemble and train a new team with the required
expertise.95
93
Quoted by Marsha Freeman, How We Got to the Moon, 314.
94
Robert S. Kraemer, Beyond the Moon: A Golden Age of Planetary Exploration: 19711978 (Washington: Smithsonian Institution
Press, 2000), xiv.
95
W. L. Gorton to Abe Silverstein, 6 February 1968, Box 59, Old RL10 Records (Goette files), DEB Vault, NASA Glenn.
31276-chapter 3 3/8/05 11:55 AM Page 101
Abes Baby 101
A Changing Vision of Space Travel
Had the Centaur rocket become superfluous? The Apollo years in the 1960s had focused
single-mindedly on landing Americans on the lunar surface before their Soviet counterparts.
In an era of nearly limitless funds for Cold War projects, the notion of a reusable rocket was
not given a high priority.96 However, once the Moon missions were successfully completed,
simply discarding rocket stages that cost millions of dollars to manufacture seemed extravagant.
Faced with tightening budgets, NASA leaders decided that seemingly more economical
ferrying systems and reusability were the most important goals for the future. Apollo 11 astro-
naut Michael Collins recalled that in the late 1960s, reusability implied a technological
maturity, a feeling that space was here to stay, that launching ships into Kennedys new ocean
would become routine. Collins concluded that for an aeronautical engineer, the path to
reusability was glorious.97
As the 1960s came to a close, NASA began to officially investigate designs for a reusable space
vehicle called the Space Transportation System (STS) or Space Shuttle.98 In 1969, NASA awarded
contracts for the first phase of the project. General Dynamics, Lockheed, McDonnell Douglas,
and North American Rockwell each received $500,000 to conduct feasibility studies.99 By
September, NASA had pooled these recommendations together and appraised the possible
designs. The approved plan was to develop a reusable-rocket transportation system that was
capable of routine access to space.100 When the next phase began in July 1970, reusability was
the primary goal, along with a piloted orbiter capable of at least one hundred missions. In early
1972, after a meeting with NASA Administrator James Fletcher and his deputy, George Low,
President Richard Nixon officially announced the plan to dramatically change the American
initiative in space.101 The main rationale for the Shuttle was the presumed low cost of a vehicle
that could be used many times compared to that of a launch vehicle that was used only once.
96
Stuart W. Leslie, The Cold War and American Science: The Military-Industrial-Academic Complex at MIT and Stanford (New
York: Columbia University Press, 1993).
97
Michael Collins, Liftoff: The Story of Americas Adventure in Space (New York: Grove Press, 1988), 202.
98
T. A. Heppenheimer, The Space Shuttle Decision: NASAs Search for a Reusable Space Vehicle (Washington, DC: NASA, 1999).
99
David Baker, Evolution of the Space Shuttle, Part I, Spaceflight 15 (June 1973): 202210.
100
Grey, Enterprise, 59. Space Task group, The Post-Apollo Space Program: Directions for the Future, September 1969, as
found in Exploring the Unknown, vol. I, 525543.
Henry C. Dethloff, The Space Shuttles First Flight: STS-1, in From Engineering Science to Big Science: The NACA and NASA Collier
101
Trophy Winners, ed. Pamela E. Mack (Washington, DC: Government Printing Office, 1998), 286. See also White House Press Secretary,
The White House, Statement by the President, 5 January 1972, Richard M. Nixon Presidential Files, NASA Historical Reference
Collection, NASA Headquarters, Washington, DC, as found in Nixon Approves the Space Shuttle, Roger D. Launius, NASA: A
History of the U.S. Civil Space Program (Malabar, Florida: Krieger Publishing Company, 1994), 232.
31276-chapter 3 3/8/05 11:55 AM Page 102
What NASA did not count on was the exorbitant operational costs of sending people into space,
as well as the extremely long time it took to turn the Shuttle around.102
As plans for the Shuttle were finalized, General Dynamics and Lewis engineers reconciled
themselves to the idea that the Centaur Program would be canceled at some time in the not-too-
distant future. Centaur won a reprieve after the transfer of Atlas-Agena missions to Atlas-Centaur
in January 1969.103 Then Atlas-Centaur became the designated launch vehicle for NASAs ambi-
tious planetary program of the 1970s while the nation awaited the Shuttle. In the next decade, a
Centaur-launched spacecraft visited every large body in the solar systemthe Sun, Earths Moon,
Mercury, Venus, Mars, Jupiter, Saturn, Uranus, Neptune, and beyond.
102
See Introduction by Roger Launius to To Reach the High Frontier, 14.
103
Agena continued as an Air Force single and upper stage until 1985.
31276-chapter 4 3/9/05 9:27 AM Page 103
Heavy Lift 103
Chapter 4
Heavy Lift
It was one of those rare moments in time. In the brief span of eight years,
from 1971 to 1978, Americans launched a fleet of robot spacecraft on
paths to the far corners of the solar system . . . undertaking what must
surely be the greatest burst of exploration in the history of mankind.
Robert S. Kraemer, NASA Office of Space Science
The 1970s became one of the most significant decades in the history of space science, a
branch of knowledge that received impetus after the launch of Sputnik in October 1957. In 1958,
Eisenhowers Science Advisory Committee defined space science in a pamphlet entitled
Introduction to Outer Space as knowledge of the Earth, solar system, and the universe.1 In his
book describing NASAs early space program, Homer Newell made clear how space science
differed from the work of earlier earthbound astronomers who had trained their telescopes on the
heavens. The space program gave scientists new instruments and a new vantage point from which
to see the heavens, a vantage point free from many of the optical distortions produced by Earths
atmosphere. Newell called attention to the important post-Sputnik role that rockets and space-
craft played in the pursuit of new knowledge of the solar system. For Newell, space science
comprised those scientific investigations made possible or significantly aided by rockets, satellites,
and space probes.2 NASAs space science program had an enormous impact on the sciences,
particularly the disciplines of geophysics, astronomy, physics, and geology. Robert Kraemer, a
former director of NASA planetary programs, recalled, It was one of those rare moments in time.
In the brief span of eight years, from 1971 to 1978, Americans launched a fleet of robot space-
craft on paths to the far corners of the solar system . . . undertaking what must surely be the
greatest burst of exploration in the history of mankind.3
1
Homer Newell, Beyond the Atmosphere: Early Years of Space Science (Washington, DC: NASA, 1980), xiii.
2
Ibid., 11.
3
Press, 2000).
31276-chapter 4 3/9/05 9:27 AM Page 104
During the 1970s, Centaur technology advanced dramatically. The most significant technical
innovation during this period was a new Teledyne guidance and control system first tested in
1973. With more advanced avionics, mission planners had more flexibility in the design of
missions. The new guidance and control system made possible the development of a revolu-
tionary approach to programming vehicle flightpaths through the winds encountered during
liftoff. At the same time, a new rocket propulsion test facility at Lewiss Plum Brook Station
in Sandusky, Ohio, permitted more sophisticated ground-testing of Centaurs restart capa-
bility. The missions themselves during this period provided opportunities for testing new
hardware and software in flight.
Like the spaceflights that sent human beings to the Moon, missions that contributed to new
knowledge of the universe had propaganda value in demonstrating Americas technical and scien-
tific prowess. Mars and Venus were highly sought-after prizes for whichever nation could be the
first to explore them.4 J. N. James, a JPL engineer during this period, commented, Although the
official U.S. position at that time was that we were not competing with the USSR, my fellow team
members and I felt otherwise.5 The race to the planets put pressure on NASA to take risks. With
the benefit of hindsight, Newell wrote that because the Agency was exposed directly to the
outside pressures to match or surpass the Soviet achievement in space, NASA moved more rapidly
with the development of observatory-class satellites and the larger deep-space probes than the
scientists would have required.6 These pressures probably contributed to early failures of missions
dedicated to the advance of NASAs space science program.
In the 1960s, NASAs long-range plan was to send flyby missions to the closest planets to
Earth. These missions would pave the way for more complex orbiting missions, eventually leading
to landing on a planet. In 1960, JPL concluded that Centaur would make the ideal upper stage
for Mariner but opted for the less powerful Atlas-Agena when Centaur development problems threat-
ened cancellation of the program. The lack of reliable launch vehicles bedeviled the American
space endeavor from the beginning, wrote historians Clinton and Linda Ezell. While this rela-
tionship between launch vehicle and spacecraft was apparent in any space project, it had an
especially negative effect on Mariner.7 With the American planetary program held back by the
lack of a sufficiently powerful and reliable upper stage, the Soviet Union launched the first probe
to Mars in 1962. Relief over the fact that the Russian probe never reached the planet probably did
4
James Schefter, The Race: The Uncensored Story of How America Beat Russia to the Moon (New York: Doubleday, 1999); William
H. Schauer, The Politics of Space: A Comparison of the Soviet and American Space Programs (New York: Holmes & Meier,
Publishers, 1976).
5
J. N. James, The First Mission to Mars, in Mars: Past, Present, and Future, ed. E. Brian Pritchard, vol. 145: Progress in
Astronautics and Aeronautics, 1992, 30.
6
Homer Newell, Beyond the Atmosphere, 97.
7
Edward Clinton Ezell and Linda Neuman Ezell, On Mars: Exploration of the Red Planet, 19581978 (Washington: NASA SP-
4212, 1984), 25.
31276-chapter 4 3/9/05 9:27 AM Page 105
Heavy Lift 105
little to calm NASAs sense of urgency. Mariner 1 attempted a Venus flyby in 1962, but range
safety officials had to destroy it when it veered off course. A Mars flyby with Mariner 3 in 1964
went silent 9 hours after launch. The second attempt at missions to each of these planets proved
more successful. Mariner 2 became the first spacecraft to encounter another planet when it
reached Venus in 1962, and Mariner 4 provided the first close view of the mysterious red planet
in 1964.
However, mission specialists in the 1970s needed the heavy lift capability of Centaur. Robert
Gray, Director of Unmanned Launch Operations at Kennedy Space Center, captured the essence
of this era in two wordsheavy science.8 The high thrust capability of Centaur, achieved
through the use of liquid-hydrogen/liquid-oxygen propellants, made it the upper stage of choice
for JPLs scientific missions in the 1970s. As Centaur proved its reliability, JPL personnel formed
strong working relationships with their counterparts at General Dynamics and Lewis Research
Center. Joe Nieberding attributed this change in perception to the increasing competence of Lewis
Research Center Centaur managers. They recognized that we knew what we were doing, he said.
We really depended on each other to do our jobs.9 Mission designer Charles Kohlhase of JPL
chose Centaur because of the vehicles enormous performance capability.10 Norm Haynes of JPL
recalled that in 1971, Atlas-Centaur was the only way we could get a spacecraft into Mars orbit
. . . we started out with a big mass and we started shaving a little in the spacecraft, but we got most
of it out of the launch vehicle.11
Atlas-Centaur sent very heavy orbiting observatories to circle Earth and record extensive data
about the celestial sky. It also sent a variety of spacecraft on trips throughout the solar system.12 Titan-
Centaur, to be discussed in the following chapter, launched six missions to the Sun and the inner
and outer planets. Two decades later, a redesigned Titan-Centaur launched the Cassini mission to
Saturn. NASAs ambitious probes to the outer reaches of the solar system helped win the prestige race
with the Soviet Union. Because the Russians used conventional fuel in their upper stages, their craft
were heavier and could not provide the same weight-carrying capacity as Centaur.
8
Interview with Robert Gray by Virginia Dawson, 9 November 1999.
9
10
Interview with Charles Kohlhase by Virginia Dawson, 8 June 2000.
11
Interview with Norm Haynes by Virginia Dawson, 9 June 2000.
12
The nine planetary launches, together with the six launches of Titan III-Centaur discussed in the following chapter, constitute
almost the entire history of Centaur launched probes to uncover the mysteries of our solar system. (The first of the seven Titan
III-Centaur launches was a mass simulator payload.) There were two later Atlas-Centaur launches: the Atlas-Centaur IIAS launch
of the Solar Heliospheric Observatory (SOHO) for NASA by the Atlas commercial program in December 1995 (AC-121) and
the Titan IV-Centaur launch of Cassini for NASA in October 1997 (TC-21). SOHO was put into a LaGrange orbit to study
the Sun, and Cassini was a science mission to the outer planets.
31276-chapter 4 3/9/05 9:27 AM Page 106
In addition to its service to NASAs space science program, Atlas-Centaur became the
preferred launch vehicle for the International Telecommunications Satellite Organization
(Intelsat). During the 1970s, Atlas-Centaur launched nineteen commercial satellites (with two
failures) and the first of seven Fleet Satellite Communications (FLTSATCOM) satellites for the
military. These satellites were the spaceborne portion of a global Department of Defense
communications system. The decade proved exceptionally busy for the Launch Vehicles
Division at Lewis, headed between 1971 and 1974 by Dan Shramo. As the Lewis role in launch
vehicles expanded in 1974, Center Director Bruce Lundin created a Launch Vehicles
Directorate under Andrew Stofan. Henry Slone became manager of Atlas-Centaur, with Paul
Winslow replacing Stofan as project manager of Titan-Centaur.
A Growing Expertise
A massive downsizing of Lewis Research Center in the early 1970s could not dampen the
enthusiasm of the members of the Launch Vehicles Division for their bird. No longer Abes baby,
Centaur belonged to the new generation of men and women who had experienced the tribula-
tions of developing Centaur. The young engineers whom NASA had recruited in the early 1960s
were just reaching their stride. Not part of the laboratorys research tradition, they had matured
with the Centaur program, shaped by the constant pressure of a launch date dictated by the posi-
tion of Earth in relation to the other planets. From their common focus and physical isolation
from the rest of the laboratory evolved a distinctive launch vehicles culture within Lewis. Joe
Nieberding, a graduate of local John Carroll University with a major in physics, had spent his
entire life in Cleveland before he started work at NASA Lewis in the early 1960s. Within his first
month in the Launch Vehicles Division, he had seen both the Atlantic and Pacific Oceans for the
first time.13
People in the Launch Vehicles Division grew used to shuttling back and forth across the
country. They attended meetings, scrutinized the fabrication of Centaur tanks in San Diego,
monitored tests at Sycamore Canyon and Point Loma, negotiated with the payload specialists at
JPL, set up shake tests at the Lewis Plum Brook facility, and flew down to Cape Canaveral to
prepare for launches. Some of the older research men at Lewis resented the privileges and high
salaries of the new recruits in the Launch Vehicles Division, but they wanted nothing to do with
the rough-and-tumble world of Centaur. The Development Engineering Building (DEB)
completed in the mid-1960s provided office space for several hundred engineers outside the labo-
ratorys main gate. The location of the Launch Vehicles Division in the DEB emphasized its
separation from the research side of the laboratory.
Because of the importance of the interface between the payload and the launch vehicle,
Centaur engineers became thoroughly familiar with the science of the missions. Most often, they
13
31276-chapter 4 3/9/05 9:27 AM Page 107
Heavy Lift 107
worked with mission planners at JPL in Pasadena, California. They also managed missions for
Goddard Space Flight Center, Ames Research Center, and Marshall Space Flight Center during
this period. Lewis and General Dynamics launch vehicle engineers were invited to the key science
meetings so that they could better understand the Centaur payload requirements. Andrew Stofan
said, We attended all of their major reviews, so we were part of each science spacecraft and, of
course, that meant we had to have a lot of people in a lot of places, because we were part of every
science project being put on the launch vehicle. Scientists reciprocated by taking a hands-on
approach to Centaurs technology. They dont sit back up in the ivory tower, Tom Shaw of JPL
remarked. Theyre out there kicking tires regularly.14
Mission planning began two or three years before the actual launch. The Centaur team liked
to joke that, almost without exception, the payload always started out heavier than the Centaur
could lift. Mission interface, handled by a group working under Edwin Muckley of Lewis, tackled
questions such as how much can the spacecraft weigh? How much power does it need from the
rocket? What is the target? What is the mission? How fast does the rocket have to go? How much
heat can the spacecraft take? How much vibration can it stand when the rocket lifts off?
Another important way in which the Lewis launch vehicle team influenced a mission was through
trajectory analysis. The team began its work by defining the key variables. Some of these were constant
among all missions. For example, Earth always rotated and revolved around the Sun with near-
constant speeds because of its almost circular orbit. Other variables were mission-specific. Different
planets had vastly different trajectories. Once the destination was known, other variables could be
defined, such as the weight of Centaur, which changed with the differing fuel requirements, and the
weight of the payload, which depended on the type of science instruments to be flown.15
Because Centaur was launched from a revolving, rotating platform to a revolving, rotating
target, trajectory calculations changed depending upon the time of launch. A two-day launch
opportunity required a trajectory for each possible minute of liftoff. Members of the Centaur
trajectory team at Lewis pioneered many of the advanced trajectory calculation techniques needed
to handle planetary mission design.
Lewis Omer (Frank) Spurlock, working under Fred Teren in the Lewis Performance Trajectory
Group, became an expert in calculating trajectories. Spurlock had graduated from the University of
New Mexico and won a Woodrow Wilson fellowship to Western Reserve University (later renamed
Case Western Reserve University) to study physics. Just twenty-one when he started at Lewis in 1961,
he traded graduate work for the chance to contribute to the space program at a time when knowledge
of how to calculate the trajectory of a spacecraft was in its infancy.16
14
Interview with Tom Shaw by Virginia Dawson, 10 November 1999.
Richard T. Mittauer, Mariner Mars 1971 Launches, 30 April 1971, Box AC-15 to AC-24, Division Atlas/Centaur Project
15
Office, NASA GRC Records.
16
31276-chapter 4 3/9/05 9:27 AM Page 108
Designing missions for expendable launch vehicles involves generating a mathematical model
that takes into account launch vehicle performance capabilities and trajectory characteristics over a
range of mission requirements and launch periods. For this task, Spurlock started out with an IBM
computer with only 8 kilobytes of memory. He used an analog computer for the atmospheric
portion of the trajectory. He then manually completed the upper stage calculations with the tiny
digital computer. Lewiss acquisition of an IBM 7090 computer with 32 kilobytes of memory greatly
increased Spurlocks computational capability. At first, Spurlock turned the trajectory calculations
for each mission over to Harry Dempster at General Dynamics to work up into particular payload
specifications. Then Spurlock developed an optimization computer code known as DUKSUP.
(Pronounced duck soup, it was not an acronym but a reference to the ease with which it could
be applied.) DUKSUP provided an alternate, very accurate means of determining a mission trajec-
tory. It allowed mission planners to model various mission scenarios. The analytic techniques used
in DUKSUP depended on an accurate vehicle model and the power of the calculus of variations.
Early versions of the code were used on Surveyor, Mariners 6 through 10, Pioneers 10 and 11,
Viking, Voyager, Helios, Pioneer Venus, the High-Energy Astronomy Observatory (HEAO),
FLTSATCOM, INTELSAT, the Orbiting Astronomical Observatory (OAO), and Applications
Technology Satellites (ATS). DUKSUP also gave government engineers an analytic tool to eval-
uate contractor data.17
After Surveyor
The first assignment for Centaur after Surveyor was to launch two Department of Defense
satellites needed for meteorological and space environment tests. Although Atlas-Agena had
launched the previous satellites in this series, Centaur was chosen for ATS 4 and 5 because they
were too heavy for Agena to lift.18 Some key modifications were required to adapt Centaur tech-
nology for this new mission. Previous Centaurs were capable of a 25-minute coast period, but
the ATS needed a 60-minute coast. For attitude control and propellant settling during the long
coast period, engineers increased the supply of hydrogen peroxide. To compensate for the
extreme cold Centaur would encounter during the coast period, they developed better thermal
control sensors. However, the most significant change was to design a new satellite adapter to
replace the one that had attached Surveyor to Centaur. The modifications begun on 8 April were ready
17
L. R. Balkanyi and O. F. Spurlock, DUKSUP: A High Thrust Trajectory Optimization Code, AIAA 93-1127, Irvine, CA,
1619 February 1993; and an interview with Frank Spurlock by Virginia Dawson, 6 April 1999. Much of the documentation
of its early applications has been lost.
18
Donald H. Martin, Communication Satellites, 4th edition (California: The Aerospace Press, 2000), 18. For further informa-
tion on ATS, see R. H. Pickard, The Applications Technology Satellite, in Proceedings of the 16th International Astronomical
Congress (1965), vol. 4: Meteorological and Communications Satellites (1966); and Paul J. McCeney, Applications Technology
Satellite Program, Acta Astronautica 5 (MarchApril 1978): 299325.
31276-chapter 4 3/9/05 9:27 AM Page 109
Heavy Lift 109
by 24 July 1968. In all, there were thirty significant changes. The Centaur Project Office at Lewis
Research Center compiled the following summary of technical modifications to Centaur in prepara-
tion for post-Surveyor missions.19
Centaur Technical Modifications After Surveyor Missions Were Completed

System Change from Surveyor Reason for Change
Mechanical Systems, Provide a dual-bottle H 2O2 Provide an additional H2O2

Airborne, Peroxide supply system and the supply to cover the extended
System necessary support structure. coast period.
Mechanical Systems, Provide for improved thermal Required to compensate for

Airborne, Peroxide control of the H 2O2 bottles the possible temperature
System and the boost pump extremes resulting from the
overspeed sensors. extended coast period.
Mechanical Systems, Requalify the 3-pound Required to meet the extended

Airborne, Peroxide thrust H 2O2 motors for coast-period requirement.
System 65-minute operation.
Mechanical Systems, Redesign the vent valve friction General design

Airborne, Pneumatics devices and controller bellows. improvement.
Mechanical Systems, Requalify the pneumatic Required as a result of

Airborne, Pneumatics regulators to 45F. possible lower temperature
environment during the
extended coast period.
Mechanical Systems, Propellant utilization Difficulties with subcontractor.

Airborne, Propellants electronics package to be
built and tested by GD/CC
instead of subcontractor.
Mechanical Systems, Add two access ports on the + Provide access to the ATS
Airborne, Structural, and y axes. electrical connectors.
Nose Fairings
19
Presentation of Launch Vehicle System in Support of the ATS-D Mission, 16 April 1968, Glenn Research Center, DEB
Archives, Box AC-15 to AC-24, Division Atlas/Centaur Project Office.
31276-chapter 4 3/9/05 9:27 AM Page 110
Mechanical Systems, Add two electrical disconnect Support spacecraft safe/arm

Airborne, Structural, support arms. and power supply cabling
Nose Fairings and withdraw the disconnect
cards from the spacecraft
during nose-fairing jettison.
Mechanical Systems, Add spacecraft T&C antenna Prevent the spacecraft

Airborne, Structural, ramps and rub surfaces. antennas from 1) being
Nose Fairings abraded and 2) hanging up
on the fairing longerons
during nose-fairing jettison.
Mechanical Systems, Delete fiberglass air Make the ATS duct

Airborne, Structural, conditioning ducts from the configuration common with
Nose Fairings conical fairing and add that required for Mariner 69
removable metal ducts. and allow removal of the
duct for cleaning.
Mechanical Systems, Relief cuts made on thermal Provide sufficient clearance

Airborne, Structural, bulkhead at nose-fairing between the thermal bulkhead
Nose Fairings split line. and transition adapter during
nose-fairing jettison.
Mechanical Systems, Delete Surveyor-peculiar Meet ATS requirements to

Airborne, Structural, TV lights and add ATS TV check out spacecraft cameras.
Nose Fairings target lights.
Mechanical Systems, Delete Surveyor forward Required to join the ATS

Airborne, Structural, payload adapter and add an (HAC) adapter with the
Payload Adapter ATS-peculiar transition adapter. Centaur field joint.
Mechanical Systems, Modify the electrical interface Accommodate the ATS

Airborne, Structural, island. destructor leads and the
Payload Adapter payload separation and
instrumentation connectors.
Mechanical Systems, Install new air conditioning Conform to the design of the
Payload Adapter ducts in the transition new adapter and satisfy the
adapter. ATS thermal requirements.
31276-chapter 4 3/9/05 9:27 AM Page 111
Heavy Lift 111
Electrical System, Delete Surveyor-peculiar Not required to support

Airborne harnessing. ATS mission.
Electrical System, Provide a pyro harness from Meet ATS requirement for
Airborne the pyro relay package to the Centaur to provide spacecraft
adapter interface island. separation electrical power.
Electrical System, Add harnesses between Meet ATS requirements for

Airborne Centaur umbilical and 1) spacecraft safe/arm and
1) the spacecraft and power supply circuits and
2) TV lights on the nose 2) nose-fairing TV lights.
fairing.
Electrical System, Install a larger capacity Meet need for additional

Airborne (150 ampere-hour) battery Centaur electrical power
with a voltage monitor required because of the
resistor. extended coast and
retromaneuver periods.
Electrical System, Delete Surveyor-peculiar Primarily provide for retrieval

Instrumentation and instrumentation and add of data associated with the
Telemetry Systems instrumentation for additional extended coast. Also provide
Centaur data. for analysis of the spacecraft
adapter interface vibrations
during flight.
Electrical System, Add a second telepak. Provide for increased number

Instrumentation and of measurements.
Telemetry Systems
Electrical System, Provide for two ATS Meet ATS requirement.

Instrumentation and acceleration measurements.
Telemetry Systems
Guidance and Autopilot, Remove pitch program Effect system simplification.

Atlas from Atlas programmer and
place in Centaur computer.
31276-chapter 4 3/9/05 9:27 AM Page 112
Guidance and Autopilot, Revise the programmer Reduce the aerodynamic

Atlas sequence to delay the nose- heating input to the spacecraft
fairing jettison event. and establish commonality
with AC-19 and AC-20.
Guidance and Autopilot, Redesign rate gyro package Provide increased vehicle to
Centaur stability margin. provide additional levels and
switching for rate and position
gains.
Guidance and Provide nonlinear position Enable the accomplishment

Autopilot, Centaur gains during the reorientation of a large angle reorientation
interval. maneuver in a limited time
interval and with a high
degree of accuracy.
Guidance and Relocate the Spin Motor Provide room for the additional
Autopilot, Centaur Rotation Detector external components required in the
to the rate gyro package. rate gyro package to effect
the above changes.
Guidance and Delete Surveyor-peculiar Required primarily to

Autopilot, Centaur functions from the implement the reorientation
programmer and add and retromaneuver sequences.
the ATS discretes.
Guidance and Revise the timer logic to Enable more rapid

Autopilot, Centaur reference spacecraft separation retromaneuver due to
to the Main Engine elimination of fixed
Cut Off event. time delays for burn-time
dispersions.
Guidance and Revise programmer switching Compensate for anticipated

Autopilot, Centaur to effect a longer prestart higher temperatures resulting
sequence. from long coast.
31276-chapter 4 3/9/05 9:27 AM Page 113
Heavy Lift 113
Since satellite technology was still in the experimental stage, the main goal for ATS 4 and
5 (AC-17 and -18) was to investigate how to maintain satellite stability while in orbit.20 Both
launches were disappointing. In August 1968, Centaur separated from the Atlas and coasted
successfully for just over an hour, but its engines failed to restart. A year later, Centaur success-
fully placed the satellite into transfer orbit after a 25-minute coast, then developed an
unplanned-for spin that compromised its communications experiments.21 Because of its poor
performance, Centaur was abandoned for the final ATS launch in 1974 in favor of the Titan
IIIC vehicle.
After this inauspicious beginning, the Centaur launch of a revolutionary, very heavy Orbiting
Astronomical Observatory (OAO), managed by Goddard Space Flight Center, was more
successful.22 The great weight of OAO was an unavoidable attribute because it required extremely
sophisticated optical instrumentation. Considerable weight was also taken up by basic equipment
such as power supplies, temperature control, and tracking and telemetering equipment. The first
OAO, launched by Atlas-Agena D in 1966, had failed.23
OAO-2, launched in December 1968, weighed 4,450 pounds.24 Its main objective was to make
unprecedented, precise telescopic observations 480 miles above Earth, where the atmosphere
produced fewer optical distortions. Because the payload was too large for the Atlas-Centaur nose
fairing, engineers used the Atlas-Agena fairing and connected the OAO to the front of Centaur
with the Agena adapter.25 Four days after a flawless Atlas-Centaur launch, the OAO onboard
experiments immediately began showing the presence of star fields, returning thousands of unique
images and mapping the northern skies.26
An even heavier and more complex scientific satellite, OAO-3, was launched in November 1970
but never reached orbit. When the Centaur computer attempted to jettison the nose fairing around
20
Daniel R. Glover, NASA Experimental Communications Satellites, 19581995, in Beyond the Ionosphere: Fifty Years of
Satellite Communication (Washington, DC: NASA, 1997), 5660.
21
Roy K. Hackbarth, Atlas-Centaur AC-18 Performance Evaluation, NASA TM X-2383, Box AC-15 to AC-24, Division
Atlas/Centaur Project Office, NASA GRC Records.
22
AC-16 Centaur Flight Evaluation Report, Glenn Research Center, DEB Archives, Box AC-15 to AC-24, Division Atlas/Centaur
Project Office. See also Mike Reynolds, The Orbiting Astronomical Observatories, in USA in Space, eds. Frank N. Magill and Russell
R. Tobias (Pasadena, CA: Salem Press, Inc., 1996), 471; and Homer Newell, Beyond the Atmosphere, 145.
23
The Orbiting Astronomical Observatory, Sky and Telescope (December 1962): 339340.
24
Atlas/Centuar-16 Orbiting Astronomical Observatory-2 Final Field Report, 17 October 1969, Box AC-15 to AC-24,
Division Atlas/Centaur Project Office, NASA GRC Records.
25
AC-16 Centaur Flight Evaluation Report, Glenn Research Center, DEB Archives, Box AC-15 to AC-24, Division
Atlas/Centaur Project Office. See also G. R. Richards and Joel W. Powell, The Centaur Vehicle, British Interplanetary Society
42 (1 March 1989): 108.
26
Arthur Code quotation found in OAO-B Launch, General Release, Box AC-15 to AC-24, Division Atlas/Centaur Project
Office, NASA GRC Records. Arthur D. Code, ed., The Scientific Results from the Orbiting Astronomical Observatory
(Washington, DC: NASA SP-310, 1972).
31276-chapter 4 3/9/05 9:27 AM Page 114
Atlas-Centaur parking orbit mission delivering a spacecraft to synchronous apogee transfer. (Courtesy of Lockheed Martin. See
Centaur: Mission Planners Guide, File 010216, NASA Historical Reference Collection.)
31276-chapter 4 3/9/05 9:27 AM Page 115
Heavy Lift 115

31276-chapter 4 3/9/05 9:27 AM Page 116
Atlas-Centaur Launches of Scientific Satellites
Date Mission Vehicle Payload Result

Weight
10 August 1968 ATS-4 AC-17 130 lb* Centaur failed
7 December 1968 OAO-2 AC-16 4,447 lb Success
12 August 1969 ATS-5 AC-18 130 lb* Success
30 November 1970 OAO-3 AC-21 4,698 lb Centaur failed
21 August 1972 OAO-4 AC-22 4,914 lb Success
12 August 1977 HEAO-1 AC-45 5,549 lb Success
13 November 1978 HEAO-2 AC-52 6,866 lb Success
20 September 1979 HEAO-3 AC-53 6,321 lb Success
*The actual weights carried by Centaur to transfer orbit were approximately double these numbers.
Atlas-Centaur Launches of Planetary Science Missions
Date Mission Vehicle Payload Result

Weight
24 February 1969 Mariner 6 AC-20 850 lb Mars flyby
27 March 1969 Mariner 7 AC-19 848 lb Mars flyby
9 May 1971 Mariner 8 AC-24 2,192 lb Mars orbiter; Centaur failed
30 May 1971 Mariner 9 AC-23 2,201 lb Mars orbiter
2 March 1972 Pioneer 10 AC-27 568 lb Jupiter and the outer planets
5 April 1973 Pioneer 11 AC-30 562 lb Jupiter and the outer planets
3 November 1973 Mariner 10 AC-34 1,108 lb Venus and Mercury flyby
20 May 1978 Pioneer Venus AC-50 1,222 lb Venus orbiter
8 August 1978 Pioneer Venus AC-51 1,948 lb Venus Multiprobe mission

31276-chapter 4 3/9/05 9:27 AM Page 117
Heavy Lift 117
the payload, one of the sixteen explosive bolts failed to release, and the rocket and satellite plunged to
Earth together.27 Because of this failure, the Centaur fairing was significantly modified. Full-scale
jettison tests, under the direction of William Prati, were conducted in the Altitude Wind Tunnel at
Lewis until the team had confidence in the new design.
No unforeseen glitches occurred when Atlas-Centaur (AC-22) lifted Copernicus, the fourth
and final spacecraft in the OAO series, into orbit on 21 August 1972. OAO-4 carried telescopes
designed by scientists at Princeton and University College of London. At 4,900 pounds, it was
NASAs heaviest satellite, and it continued to send data back to Earth for nine years.28 Despite
the two failures of the OAO program, the second and fourth satellites were widely judged to be
technical and scientific successes, and that bolstered NASAs confidence that a space telescope
much larger than the OAO series could be built.29 They marked the beginning of above-the-
atmosphere observatories that now include the Hubble telescope.30
First Centaur Interplanetary Missions: Mariners 6 Through 9
Although the Orbiting Astronomical Observatories had provided unprecedented obser-

vations of the sky, this achievement paled in comparison with sending a spacecraft to the
vicinity of a planet. Earlier Mariner spacecraft had avoided using Centaur, but the next gener-
ation needed the additional weight that only Centaur could carry. By using Centaur,
Mariners 6 through 9 carried heavy atmospheric and topographical instruments needed for
gathering data in preparation for a future Mars landing.31 Because failures were so common
in the early space program, NASA built two identical Mariner spacecraft for each mission and
launched them in pairs.32
27
AC-21 Flight Evaluation Report, Box AC-15 to AC-24, Division Atlas/Centaur Project Office, NASA GRC Records.
28
Raymond N. Watts, Jr., An Astronomy Satellite Named Copernicus, Sky and Telescope (October 1972): 231232, 235.
Lyman Spitzer, Jr., Searching Between the Stars (New Haven: Yale University Press, 1982).
29
Smith, The Space Telescope, 44.
30
Although no comprehensive history of the OAO exists, several other articles from Sky and Telescope document some of its
achievements. For further information, see Observing a Comet from Space, Sky and Telescope 39 (March 1970): 143; The
Orbiting Astronomical Observatory, Sky and Telescope 24 (December 1962): 339340; Watts, More About the OAOs, Sky
and Telescope 28 (August 1964): 7879; and Orbiting Astronomical Observatory, Sky and Telescope 28 (August 1964): 7879.
31
Project Mariner Mars 1969, 14 February 1969, Box AC-15 to AC-24, Division Atlas/Centaur Project Office, NASA GRC
Records.
32
Mariner 6 was 271 pounds heavier than Mariner 4 and 399 pounds heavier than Mariner 2. For further information on
Mariners 6 and 7, see James Wilson, Two Over Mars: Mariner VI and Mariner VII, FebruaryAugust 1969 (Washington, DC:
NASA EP-90 Government Printing Office, 1971); Clark R. Chapman, Planets of Rock and Ice (New York: Charles Scribners
Sons, 1982); Scientific and Technical Information Division, Office of Technology Utilization, Mariner Mars 1969: A Preliminary
Report (Springfield, VA: NASA SP-225 Clearinghouse, Department of Commerce, 1969).
31276-chapter 4 3/9/05 9:27 AM Page 118
Mariner 6 (AC-20) was nearly marred by a catastrophic accident ten days before its scheduled
liftoff. Atlas-Centaur and its Mariner payload were sitting on the launch pad at the Cape when
Atlas suddenly began to lose pressurization. A faulty relay switch had caused one of the main valves
to open. As the Atlas began to slowly crumple, two General Dynamics mechanics ran under the
twelve-story assembly with its deflating booster to seal off the pressure manually by closing a valve.
The Lewis News reported the heroic story:
Ignoring the shrill scream of the evacuation alarm, Billy McClure hastily removed the
liquid oxygen prevalve locking bolts to the Atlas vehicle on Pad 36A at the Cape. As he
did this, he heard the sound of metal buckling, and looking upward he saw the Atlas
begin to lean under the weight of the Centaur and Mariner spacecraft as the pressure
leaked out of the vehicle. At the same time Charles Beverlin had run under the Atlas and
squirmed up into the Atlas thrust section to close the sustainer engine prevalve. Seconds
later both mechanics had cleared the area, and the vehicle began to right itself as the tank
pressure built up inside.33
Without their quick action, the mission would have been lost. They received a rare award
given by NASA, the Exceptional Bravery Medal.34 Liftoff in February 1969 went as scheduled
with a replacement Atlas. It flew by Mars in July at an altitude of 2,170 miles.35 Total flight time
was 156 days.
Mariner 7 (AC-19), launched in March, had almost identical mission objectives to those of
Mariner 6, except that it was to fly by the Martian southern polar cap.36 The relative locations of
Earth at launch and Mars at encounter allowed Mariner 7 to use a direct-ascent trajectory. Together,
the two spacecraft photographed twenty times more surface area of the planet than Mariner 4. This
included photographing the surface of the planet, studying the Martian atmosphere, and also
performing spectrometer studies to determine whether the conditions for life were present on
Mars. The atmospheric and surface data gathered by the spacecraft led scientists to conclude that
life probably did not exist on Mars unless it was microbial.37 In discussing the significance of these
missions, Stuart Collins wrote, Mariners 6 and 7, launched by the more powerful Atlas-Centaur
33
Mechanics Risk Lives on Launch Pad, Lewis News, Box Awards Folders Through 1979, Folder Exceptional Bravery (NASA),
NASA GRC Records.
34
Abe Silverstein to NASA Headquarters, 15 April 1969, Box Awards Folders Through 1979, Folder Exceptional Bravery
(NASA), NASA GRC Records.
35
Atlas-Centaur 20 Mariner Mars-6 Final Field Report, 19 January 1970, Box AC-15 to AC-24, Division Atlas/Centaur
Project Office, NASA GRC Records.
36
Atlas-Centaur-19 Mariner Mars-7 Final Field Report, 19 January 1970, Box AC-15 to AC-24, Division Atlas/Centaur
37
Henry W. Norris, Mariner 6 and 7, in USA in Space, 366369.
31276-chapter 4 3/9/05 9:27 AM Page 119
Heavy Lift 119
Centaur D-1. (Courtesy of Lockheed Martin. See Centaur: Mission Planners Guide, File 010216, NASA Historical Reference
Collection.)
31276-chapter 4 3/9/05 9:27 AM Page 120
rocket, were to greatly increase and improve the observations made by Mariner 4 and by Earth-
based investigators. Thus scientists and engineers began to design a more ambitious exploratory
mission and to build larger, more sophisticated spacecraft to complete this mission.38
What distinguished the two additional Mariner missions launched in 1971 from earlier
trips to Mars was that each spacecraft was intended to orbit Mars for a period of 90 days, not
merely perform a flyby of the planet. The weight-carrying capability of Centaur was again crit-
ical to the success of these missions. Each of the new Mariners weighed 2,197 pounds, more
than twice the weight of Mariners 6 and 7. When Mariner 8 (AC-24) lifted off on 9 May 1971,
everything went as planned until moments after Centaur and Mariner separated from Atlas, 255
seconds into the flight. Robert Kraemer remembers that, as he sat next to Dan Schneiderman
at the consoles in the JPL Mission Control Center at the Cape, he gave the thumbs-up sign to
Schneiderman as soon as the Centaur engines ignited. Turning back to the console, he noticed
that the pitch recorder was showing unstable oscillations about the Centaur pitch axis.39 Then
Centaur began tumbling end over end and plunged, with the spacecraft, into the Atlantic
Ocean 350 miles north of Puerto Rico.
Immediately, there was speculation that the guidance system was the culprit. Bill ONeil, JPL
navigation chief for the spacecraft, quickly discounted the possibility that the entire guidance
system was at fault. He suspected that the problem was merely a defective autopilot since it had
occurred just at the point when the system was supposed to turn on.40 Investigation occurred at
breakneck speed under the leadership of George Low, NASAs Deputy Administrator. Time was of
the essence because if the next Mariner did not launch within a month, it would be two years before
the next opportunity. Ed Ziemba, an engineer on the Lewis Centaur team, and Chet Norris,
Kennedy Space Centers lead guidance engineer on Centaur, determined that ONeil was right. An
integrated circuit in the pitch rate gyro preamplifier had failed, not the entire guidance and control
system. This tiny, $5 integrated circuit, located in the autopilot system, had caused the failure of
the $70-million rocket.41
With the malfunction corrected, Mariner 9 (AC-23) lifted off twenty-one days after the
Mariner 8 loss. Centaur operated flawlessly. Four months later, it began a Martian orbitthe first
time a spacecraft had orbited another planet. Engineers were able to assign some of the tasks orig-
inally scheduled for the ill-fated Mariner 8, such as mapping the Martian surface, to Mariner 9.
38
Stewart A. Collins, The Mariner 6 and 7 Pictures of Mars (Washington, DC: NASA SP-263 Government Printing Office,
1971), 6.
39
Press, 2000), 50.
40
Interview with William ONeil by Virginia Dawson, 9 June 2000.
41
Interview with Ed Ziemba by Virginia Dawson, 19 May 1999. See also Atlas-Centaur 24 Mariner Mars-8 Final Field Report, 26
February 1973, Box AC-15 to AC-24, Division Atlas/Centaur Project Office, NASA GRC Records.
31276-chapter 4 3/9/05 9:27 AM Page 121
Heavy Lift 121
Assembly of OAO shroud in the Space Power Facility at Plum Brook. (GRC C1965-1458)
31276-chapter 4 3/9/05 9:27 AM Page 122
A technician makes final checks on Mariner 9, which was launched by Atlas-Centaur 23 to Mars on 30 May 1971. (NASA
KSC_71P_0366)
However, during orbit around Mars, the spacecraft encountered a massive dust storm in the
Noachis region, and all of its high-resolution cameras simply returned pictures of dust.42 A month
later, after the storm cleared, Mariner returned striking images of Martian volcanoes, canyons, polar
caps, and the satellites Phobos and Deimos.43 Journalist Clyde Curry Smith reported that Mariner
revolutionized the understanding of a very old New Mars.44
Despite the successful investigation into what had gone wrong with Mariner 8 and the brilliant
achievements of Mariner 9, Congress remained extremely critical of NASAs management of
Centaur. In addition to the loss of Mariner 8, Centaur had caused the loss of an ATS in August
42
43
For further information on Mariners 8 and 9, see Exploring Space with a Camera, ed. Edgar M. Cortright (Washington, DC:
NASA SP-168 Government Printing Office, 1968); William K. Hartman and Odell Raper, The New Mars: The Discoveries of
Mariner 9 (Washington, DC: NASA SP-337 Government Printing Office, 1974); Patrick Moore and Charles A. Cross, Mars
(New York: Crown Publishers, 1973); and NASA Scientific and Technical Information Office, Mars as Viewed by Mariner 9: A
Pictorial Presentation by the Mariner 9 Television Team and the Planetology Program Principal Investigators (Washington, DC:
NASA SP-329 Government Printing Office, 1976).
44
Clyde Curry Smith, Mariner 8 & 9, in USA in Space, 370374.
31276-chapter 4 3/9/05 9:27 AM Page 123
Heavy Lift 123
1968 and an OAO in November 1970. In June 1971, during congressional hearings into these
failures, NASA Deputy Administrator George Low admitted that while every effort was made to
thoroughly check Centaur systems before launch, it was impossible to guarantee launch success,
no matter what level of caution. These losses, he said, served as harsh reminders that space explo-
ration is still a very difficult business, and, in spite of dedicated personnel and intense attention
to details, failures must be expected occasionally.45
The hearing reminded the Centaur team that they were still under the watchful eye of the
House Committee on Science and Astronautics and needed to improve Centaur performance.
Members not only were critical of the three recent failures, but also questioned the overall
success rate of Atlas-Centaur. Congressman Joseph Karth stated that an 80-percent success rate
was totally unacceptable for a vehicle on which we have spent about half a billion dollars or
more.46 He thought NASA should consider a substitute launch vehicle. John Naugle, Associate
Administrator for Space Science and Applications, responded, I think it would be an unmit-
igated disaster too if we stopped the ongoing program.47 Karth warned NASA that although
the Committee had strongly supported the launch vehicle and the planetary space program,
future support was not a foregone conclusion. He said, Unless the reliability numbers
change, Dr. Naugle, I dont think we are going to be as kind in the future.48 Despite this
scrutiny, Atlas-Centaur continued to serve as the launch vehicle of choice for NASAs
program of planetary exploration. Indeed, the power of the Centaur stage made more elabo-
rate missions possible.
To Jupiter and the Solar System: Pioneer 10
While interplanetary explorations of the early 1970s were limited to the nearest neighbors on
either side of EarthVenus and Marsscientists eagerly sought knowledge about more distant
planets. Fortuitously, in the late 1970s, the largest planets in the solar system lined up on one side
of the Sun, a phenomenon that occurs once every 176 years. A carefully timed spacecraft sent from
Earth could take advantage of this configuration to make an unprecedented tour of the planets in
a single voyage. Although the first nine spacecraft of the Pioneer series did not use the Centaur
45
George Low statement, Review of Recent Launch Failures, Hearings Before the Subcommittee on NASA Oversight, 1517
June 1971, 8.
46
Joseph Karth statement, ibid., 26.
47
John E. Naugle statement, ibid., 61.
48
Joseph Karth statement, ibid., 62.
31276-chapter 4 3/9/05 9:27 AM Page 124
Atlas-Centaur in flight just prior to Centaur second stages separation from the Atlas first stage. (NASA Glenn Research Center
unprocessed photo)
upper stage, Pioneers 10 and 11 needed the greater Centaur thrust capability to reach Jupiter and
other points in the outer solar system.49
Pioneer 10 (AC-27) arrived at Pad 36A in December 1971. One month later, TRW
shipped the Pioneer spacecraft to the Cape, and engineers began integrating the payload with
the spacecraft. High upper altitude winds and a loss of facility electrical power caused the first
delay, and continued unacceptable upper altitude winds caused two further postponements.
Then, after last-minute tanking difficulties almost aborted the launch, the spacecraft lifted
off on 2 March 1972, the beginning of one the longest voyages ever taken by a humanmade
object.50
Pioneer 10 was unusual because, in addition to its second Centaur stage, it had a spin-stabi-
lized Delta third stage with a solid rocket propellant motor (Thiokol TE-M-364-4), which
49
Roger D. Launius, NASA: A History of the U.S. Civil Space Program (Florida: Krieger Publishing Company, 1994), 102. See
also AC-27 Launch Operations Pioneer F Mission, Box AC-25 to AC-32, Division Atlas/Centaur Project Office, NASA GRC
Records.
50
AC-27 Flight Data Report, Launch Vehicles Division, Lewis Research Center, 1 June 1972, Box AC-25 to AC-32, Division
31276-chapter 4 3/9/05 9:27 AM Page 125
Heavy Lift 125
allowed it to achieve an extremely high velocity of 32,000 miles per hour.51 The Delta stage was
mounted on a spin table at the top of Centaur. Thruster rockets spun this craft at 60 revolutions
per minute, and explosive bolts separated it from Centaur after a 7-minute Centaur burn. To
prevent Deltas reflected motor exhaust from damaging Pioneer, Centaur performed a retroma-
neuver to move it 25 feet from the Delta stage.52 Delta ignition lasted for 44 seconds. Then
another explosive bolt firing took place, with compressed springs pushing the Pioneer spacecraft
away from the Delta stage. Throughout all of these complex maneuvers, the Atlas, Centaur, and
Delta stages performed almost flawlessly. The Pioneer spacecraft entered into a successful Earth-
escape orbit that required only a minor course correction. With Pioneer 10s successful launch
bolstering the confidence of the launch vehicle team, they planned to replace the Atlas-Centaur
computer system with one that was more advanced for Pioneer 11.
The Improved D-1A Centaur and Pioneer 11
NASA tested the improved Centaur D-1A with its new Teledyne avionics on Pioneer 11
(AC-30), one of the highest profile payloads of the United States. Although NASA had origi-
nally planned to make an Intelsat satellite the first test of the new avionics system, Comsat
officials had objected. Everyone involved knew the riska failure would be a significant
setback to the space program. Pioneer 11 had the same destination as Pioneer 10. It was again
carried aloft by an Atlas, Centaur, and Delta vehicle combination. This redundancy was espe-
cially important because of the ambitious goals of the mission and the possibility that one of
the three might fail.
Development of the D-1A improved Centaur began in 1968 with the goals of lowering
costs associated with Centaur launches and increasing rocket reliability. There were twenty
engineers assigned to the project, headed by Russ Dunbar. Long associated with the Centaur
program, Vincent Johnson, now Deputy Associate Administrator for Space Science and
Applications, stated to Congress that the reason for improving the Centaur was that many
of its features were becoming outdated. He said, It was recognized that the technology on
which certain key systems of the Centaur was based was becoming relatively ancient.53 Nearly
$40 million was spent in creating the new Centaur, with $8 million in development costs for
a new Teledyne computer.
The improved Centaur was tested by NASA in its new Spacecraft Propulsion Research Facility,
or B-2, at the Plum Brook Station in Sandusky, Ohio. This unique facility consisted of a huge stain-
51
Pioneer 10 and AC-27 Press Kit, Box AC-25 to AC-32, Division Atlas/Centaur Project Office, NASA GRC Records.
Atlas/Centaur 27 Pioneer 10 Flash Flight report, Centaur Operations Branch KSC-ULO, 20 March 1972, Box AC-25 to
52
AC-32, Division Atlas/Centaur Project Office, NASA GRC Records.
53
Vincent L. Johnson statement, Review of Recent Launch Failures, Hearings Before the Subcommittee on NASA Oversight,
1517 June 1971, 83.
31276-chapter 4 3/9/05 9:27 AM Page 126
less steel vacuum chamber 38 feet in diameter and 55 feet talllarge enough to put an entire
Centaur upper stage in the chamber, tank it with liquid hydrogen and liquid oxygen, and fire its
engines. The B-2 facility could simulate the environmental extremes encountered at heights of
approximately 100 to 125 miles and test Centaurs ability to restart its engines in the vacuum of
space. The issues that particularly concerned engine restart included the necessary chilldown of
the engine; making sure that liquid hydrogen, not hydrogen gas, entered the boost pumps; and
ensuring that proper pressures in the engine propellant inlets were maintained. The liquid-
nitrogen-cooled walls of the B-2 facility simulated the temperatures of 320F found in space, while
quartz lamp thermal simulators mimicked the intense heat of the Sun.
Checkout testing of the new facility began in October 1969 with the first successful hot
firing of Centaur engines in the B-2 facility on 18 December 1970. Centaur was mounted
vertically in the chamber with its nozzles facing down. After ignition, the engines fired through
a water-cooled exhaust diffuser and into a deep spray chamber. Thousands of gallons of water were
sprayed into the hot gas to cool it and turn it into steam. An exhaust system released the steam
into the air while the heated water was returned to the spray chamber. Glen Hennings, chief of
the rocket systems division at Plum Brook Station, recalled, All systems performed properly in
the test and we achieved most of our basic objectives.54 Further testing of Centaur engines in the
B-2 facility revealed that the RL10 engines could be run with a pressurized propellant feed
system, thus eliminating the need for boost pumps.55
The most important element of the improved Centaur was its new Teledyne computer. James
Patterson and Don Garman of NASA, along with David Geyer of General Dynamics, played key roles
in the ultimate success of the new computer. In contrast to the serial drum-type Librascope computer
it replaced, the new Teledyne computer was a true digital computer that could be reprogrammed
and updated with relative ease. It had a 16,384-word memory (24-bit)five times the memory of
the earlier Librascope computer. The new software gave Centaur complete control of Atlas for the
first time. The new computer had fifteen different software modules to manage functions such as
navigation, guidance, autopilot, propellant utilization, attitude control, sequencing, telemetry, and
data management for both vehicles.56 Previously, separate mechanical or electronic units controlled
54
Glen Hennings, quoted in First Test for New Facility: Centaur Hot-Fires in B2, Lewis News (2 January 1970).
55
See Steven V. Szabo, Centaur Space Vehicle Pressurized Propellant Feed System Tests, NASA TN D-6876, October 1972.
56
A. B. Yanke, AC-30 Preliminary Flight Analysis Report GDCA-HAB73-019, Box AC-25 to AC-32, Division
Atlas/Centaur Project Office, NASA GRC Records. Some of the more important modules included Navigation (NAV), which
provided position, velocity, and acceleration data to the guidance system; Powered Guidance (PGUID), which provided data
to optimize trajectory and determine engine cutoff time; Coast Guidance (CGUID), which determined the next main engine
start; Powered Autopilot (PAUTO), which maintained Centaurs stability during main engine firings; Coast Phase Autopilot
(CAUTO), which kept Centaur stable during the parking orbit; Propellant Utilization (PU), which maintained the desired
ratio of liquid hydrogen and liquid oxygen in the tanks; and the Computer Controlled Vent & Pressurization System
(CCVAPS), which used redundant tank-pressure sensors to maintain liquid-hydrogen and liquid-oxygen tank pressures. See
Titan IIIE/Centaur D-1T Systems Summary, September 1973, Joe Nieberding Personal Collection.
31276-chapter 4 3/9/05 9:27 AM Page 127
Heavy Lift 127
Cross section of the B-2 Facility at Plum Brook shows the Centaur rocket mounted for test. (CD-93-64366)
31276-chapter 4 3/9/05 9:27 AM Page 128
functions such as propellant management and tank pressurization. Not only did the advanced
computer system provide more integrated and customizable control over Atlas-Centaur, but it also
allowed mission specialists greater flexibility in the design of a mission.57 Instead of the redesigning
hardware, the computer could be reprogrammed to prepare the Centaur for different tasks.58
The Teledyne guidance and navigation systems made possible one of the most significant
contributions to Centaur D1-A capabilities. Because the computer could be programmed just
before launch, a joint government-industry Centaur team created a new system to compensate
for the winds the launch vehicle would encounter in its flight through the atmosphere. Winds
are dangerous for launch vehicles because they create side forces that can either damage the
vehicle or knock it off course. For the earliest Surveyor flights, Atlas-Centaur had used seasonal
pitch and yaw programs based on historical data from sounding balloons. These early programs
served as rough estimates of the winds that a vehicle might encounter on a typical launch day in
summer, fall, winter, or spring. Frequently, weather balloons released on launch day revealed
wind profiles quite different from the seasonal expectations. In 1966, the number of programs
available expanded from four seasonal pitch and yaw programs to ten. Between 1966 and 1973,
twenty-one Atlas-Centaur vehicles were launched using this protocol.
However, despite the power of the vehicle, it was still susceptible to upper atmosphere winds. If
high jetstream winds on launch day did not somewhat match one of the ten predesigned pitch and yaw
programs, then the entire launch had to be scrubbed or delayed until the winds cooperated. Delaying
a mission was costly. What was needed was an approach that tailored the trajectory to actual real-time
wind data on the launch day itself. Lewis engineers recalled that after Pioneer 10 was scrubbed twice in
1972 because of wind problems, Lewis Center Director Bruce Lundin demanded that the team come
up with a solution by the time they launched Pioneer 11 thirteen months later. Engineers from the
General Dynamics Aerodynamics Group, headed by Don Lesney, and a Lewis team, under Joe
Nieberding, developed a revolutionary new real-time wind program called ADDJUST (Automatic
Determination and Dissemination of Just Updated Steering Terms).59 The ADDJUST systeman
acronym coined by Frank Anthonywas incorporated into the D-1A Centaur launched in 1973.60 It
cost just $250,000 (roughly the same amount for one scrubbed launch).
The ADDJUST system dramatically increased the number of days available for launch. Prior
to ADDJUST, about 43 percent of the launch dates between December and March were unavail-
able because of wind characteristics. After ADDJUST, even after allowing for uncertainties in the
57
58
Don Savage and Ann Hutchison, Pioneer 11 to End Operations After Epic Career, NASA press release, 29 September 1995.
59
Interview with Frank Anthony by Virginia Dawson, 6 June 2000.
60
Concept Review ADDJUST System to Design Booster Steering Programs During Preflight Launch Operations, 7 July 1972,
Joe Nieberding Personal Collection.
31276-chapter 4 3/9/05 9:27 AM Page 129
Heavy Lift 129
Mariner Venus/Mercury 10 is prepared for encapsulation on 21 September 1973 at Kennedy Space Center. (NASA
KSC_73P_0541)
31276-chapter 4 3/9/05 9:27 AM Page 130
actual winds encountered in flight, this percentage decreased to only 5 percent of launch days.
The rest of the year, from April to November, Centaur achieved nearly 100-percent launch
availability with respect to upper level winds. The ADDJUST system was used on all seven
Titan-Centaur launches and has continued to play a vital role on all Atlas-Centaur launches
since that time. After 1973, a launch with a Centaur upper stage would rarely be scrubbed due
to an upper level wind problem.61
The process of implementing ADDJUST began 125 minutes before a scheduled liftoff. At this
point, the Air Force released specially designed weather balloons into the Florida skies near the
Eastern Test Range. As the balloons floated higher into the sky, they recorded wind velocity and
direction at all altitudes, up to about 80,000 feet in about 60 minutes of rise time. After 20 minutes,
these balloons began to transmit the data on the wind patterns to a ground-based computer. Five
minutes later, using 300-baud modems, computers at General Dynamics in San Diego and Lewis
Research Center in Cleveland received the data. Using two Fortran programs, their computers
processed the data for 10 minutes. Then the General Dynamics computer transmitted pitch and yaw
data to the Eastern Test Range in Florida, while the Lewis Research Center computer validated the
approach. Ten minutes later, General Dynamics sent Lewis Research Center the current pitch and
yaw data, and their computer in Cleveland validated the information for the next 10 minutes. After
all the data were confirmed, they were sent to the onboard Centaur computer 15 minutes before the
scheduled liftoff. To ensure that the data were accurate, Centaur transmitted the same data back to
the computer at the Cape for verification.
All the Centaur engineers felt the anxiety of getting the improved Centaur ready on time.
During tests of the new computer, its many glitches raised serious questions about reliability.
Lewis engineer Ed Ziemba recalled that every time Teledyne delivered a computer, General
Dynamics would test it and it would fail.62 Up to 15 minutes before the launch of Pioneer 11, the
computer was still giving the engineers trouble. Ground-based computers began spewing out
strange data. Ziemba, who was assigned to the blockhouse, agonized over whether to postpone the
launch. Because he was sure that the engineers had thoroughly checked out the new flight control
system, he gave the go-ahead. His decision was vindicated by the perfect launch of Pioneer 11 on
5 April 1973. Atlas-Centaur (AC-30) cut through the Florida winds effortlessly with the
ADDJUST system and placed its payload in its proper trajectory.
Pioneers 10 and 11 gathered unprecedented knowledge of the solar system. Pioneer 10 became
the first spacecraft to travel through the asteroid belt, the first to make direct observations of Jupiter,
and the first to venture outside the solar system. At a distance of 7 billion miles from Earth, it long
held the distinction as the most remote human artifact. Pioneer 11, originally on a redundant path,
adjusted its course to make observations of Jupiter as well as the first close observations of Saturn in
61
ADDJUST Overview, Lewis Research Center, undated, Joe Nieberding Personal Collection.
62
Interview with Ed Ziemba by Virginia Dawson, 19 May 1999.
31276-chapter 4 3/9/05 9:27 AM Page 131
Heavy Lift 131
1979. By 1995, the signal it sent back to Earth was so weak that scientists could only receive about 2
hours of data per month. In 1996, all communications fell silent and Pioneer 11 now drifts as a ghost
ship headed for the Aquila constellation that it will reach in four million years.63
The First Two-Burn Planetary Mission: Mariner 10
Centaurs association with the Mariner program came to an end in 1973 with the
Venus/Mercury launch (AC-34). Mariner 10s objective was a flyby of both Venus and Mercury
NASAs first two-planet mission. The complexity of this mission required that the new Centaur
D1 use two burns in the launch of its interplanetary spacecraft. Although the two burns had been
tested during the proof flights for Surveyor and used to launch Intelsat IV missions, Mariner 10
was the first planetary mission that required Centaur to restart its engines in space.
Deciding whether to use direct ascent or two burns depended on the geometry of Earths
location and the ultimate destination for the spacecraft. If Earth and the destination planet
were aligned in a way that allowed the voyage to be made with just one propulsive push from
the Centaur, direct ascent was selected. However, sometimes Earth and the destination planet
were aligned in such a way that it was impossible to make a straight shot.
Prior to 1973, scientists had avoided the risk of the second burn and accepted the limitations
of the direct ascent. The four previous Mariner launches to Venus and Mars were accomplished
without the two-burn option. However, the alignment of the two innermost planets with Earth for
Mariner 10 necessitated the more complicated mission profile. Although NASA could have used
the more powerful Titan-Centaur for a direct ascent to Mercury, this option was more costly. Also,
the greater speed of Titan-Centaur as it passed by the planets would have limited the time during
which Mariner 10 could acquire data. Atlas-Centaur offered the best alternative. However, many
scientists objected that using two burns might jeopardize the mission.64
Despite the complexity of the mission, when NASA launched Mariner 10 in November 1973,
there were no unexpected problems or countdown holds. The Atlas phase ended 254 seconds into
the launch. Centaur separated, and 10 seconds later, its engines fired. After 12 seconds, it jetti-
soned the nose fairing; 297 seconds later, the engines cut off and the rocket went into parking
mode in a near-circular orbit of Earth. During its 1,540-second parking orbit, Centaur cruised
into position, moving one-third of the way around Earth. The second burn began within 2
seconds of the predicted start. This sent Centaur, with its precious Mariner payload, on a course
63
For information on the Pioneer 10 and 11 voyages, see William R. Corliss, The Interplanetary Pioneers, three volumes
(Washington, DC: NASA SP-278 Government Printing Office, 1972); Richard O. Fimmel, William Swindel, and Eric Burgess,
Pioneer Odyssey (Washington, DC: NASA SP-396 Government Printing Office, 1977); Elizabeth A. Muenger, Searching the
Horizon: A History of the Ames Research Center, 19401976 (Washington, DC: NASA SP-4304 Government Printing Office,
1985); Mark Washburn, Distant Encounter: The Exploration of Jupiter and Saturn (San Diego: Harcourt Brace Jovanovich, 1983);
Tom Gehrels, ed., Jupiter (Tuscon: University of Arizona Press, 1976); Glenn S. Orton, Pioneer 10, in USA in Space, 490494;
and Manfred N. Wirth, Pioneer 11, in USA in Space, 495499.
64
31276-chapter 4 3/9/05 9:27 AM Page 132
Atlas-Centaur 34 undergoing a tanking test on NASA Complex 36B at Cape Kennedy. This launch was the first dual-planet flight,
Mariner Venus/Mercury 10. 19 September 1973. (NASA KSC_73P_0537)
31276-chapter 4 3/9/05 9:27 AM Page 133
Heavy Lift 133
opposite to the direction of Earths orbital motion around the Sun. Velocity at cutoff was greater
than statistical predictions because of a drifting Centaur gyro, but fortunately, Centaurs computer
programming could compensate for this error and successfully place Mariner on its complicated
trajectory. After 133 secondsjust as it entered an Earth-escape orbitthe Centaur main engines
shut down for the final time. Centaur and Mariner stayed connected for another 95 seconds
before they separated. After 515 seconds, Centaur performed a maneuver to get out of the way of
the spacecraft.65 In addition to requiring two burns, the Mariner trajectory was significant for
another reason: it was the first time that a spacecraft used the gravitational attraction of one planet
to catapult itself to anothera feat requiring exceptional navigational accuracy and multiple
trajectory corrections to be performed by the trajectory analysis team.66
Mariner 10 reached Venus in February 1974. One month later, it made its first of three flybys
of Mercury. After this first approach, Mariner 10 went into a solar orbit that permitted it to make
two more encounters with the planet. It made its second Mercury encounter in September 1974,
passing over the sunlit side of the planet and the southern polar region. Mariner 10 made its final
pass by Mercury in March 1975making its closest approach to the planet at 203 miles. It went
silent eight days later when the fuel was depleted. Mariner 10 remains the only spacecraft to have
visited Mercury. It made the first examination of the interplanetary space between Venus and
Mercury and returned the first close observations of the cloud cover over Venus.67
A Journey to the Mysterious Planet: Pioneer Venus
While the previous two Pioneer spacecraft took long voyages away from the Sun and out into
deep space, two additional Pioneer probes in the late 1970s were launched toward the second planet
from the Sun. Venus, long known as the mysterious planet, was most like Earth, with a similar
mass, radius, and density. However, telescopic observations from Earth and data from the Mariner
10 probe suggested some striking differences. Temperatures on Venus approached 850 degrees, and
its magnetic field was 1/3,000 of Earths. Its rotational motion was 1/243 of Earths and in the
opposite direction. Finally, Venus was thought to be extremely arid, with a water content of
1/10,000 of Earths biosphere. How could a planet so similar to Earth have surface conditions so
AC-34 Atlas-Centaur Flight Evaluation Report, Box AC-33 to AC-41, Division Atlas/Centaur Project Office, NASA GRC
65
Records.
66
Pioneer used gravity-assist as well, but although it was launched first, it was the second spacecraft to use the gravity-assist tech-
nique. See Mariner Venus/Mercury 1973, Launch Operations and Flight Events, Box AC-33 to AC-41, Division
67
For further information on Mariner 10, see James A. Dunne and Eric Burgess, The Voyage of Mariner 10: Mission to Venus and
Mercury (Washington, DC: NASA SP-424 Government Printing Office, 1978); Clark R. Chapman, Planets of Rock and Ice: from
Mercury to the Moons of Saturn (New York: Charles Scribners Sons, 1982); Bruce C. Murray and Eric Burgess, Flight to Mercury
(New York: Columbia University Press, 1976); Robert G. Strom, Mercury, the Elusive Planet (Washington DC: Smithsonian
Institution Press, 1987); and Donna Pivirotto, Mariner 10, in USA in Space, 375379.
31276-chapter 4 3/9/05 9:27 AM Page 134
radically different? Space scientists wanted to make a comparative study of Earth and Venus in the
hopes of not only solving these tantalizing questions, but also shedding new light upon the evolu-
tion of our own planet.68
By the mid 1970s, the United States and the Soviet Union had sent a total of eight missions
to study the planet. These included three Mariner flyby missions from the United States and five
more complex missions by the Soviet Union with its Venera craft. The final two craft in this series
(Venera 7 and 8 in 1971 and 1972, respectively) each landed on the surface of the planet and
conducted measurements. Scientists in the United States believed that they were slipping behind
their Soviet counterparts and sought to catch up. Although Pioneer Venus did not beat Soviet
accomplishments, it served as the first United States mission to send both an orbiter and an atmos-
pheric probe on separate launches.
NASA had originally chosen the Delta launch vehicle for Pioneer Venus. However, as
scientists began designing their experiments, they found themselves severely constrained by
Deltas weight capabilities. Either they could miniaturize and reduce the scope of their experiments,
or they could spend an additional $10 million on an Atlas-Centaur launch vehicle. They soon
realized that the funds saved by not having to miniaturize their experiments would offset
Centaurs higher price tag.69
Atlas-Centaur (AC-50) launched the Pioneer Venus orbiter in May 1978.70 Over the
next fourteen years, the orbiter successfully carried out seventeen scientific experiments. It
maintained an orbit around the planet until fuel depletion in August 1992 and disintegrated
upon entering the planets atmosphere. While Pioneer Venus yielded significant scientific
results about the planet, these results were overshadowed by the excitement generated by a
complicated Venus-bound spacecraft called the Multiprobe launched two and a half months
later. Like the Pioneer Venus orbiter, the Multiprobe trajectory required a two-burn ascent
with similar launch times and procedures. When the first orbiter arrived at Venus in
December 1978, the Multiprobe was only five days behind.
The Multiprobe carried four atmospheric-entry probes to gather data on the Venusian
clouds, atmospheric structure, and circulation pattern of the atmosphere. One of these
probes managed to send back radio signals for an hour after impact with the surface of the
68
Atlas-Centaur D-1A AC-50 Launch Operations and Flight Events, Box AC-50 to AC-57, Division Atlas/Centaur Project
Richard O. Fimmel, Lawrence Colin, and Eric Burgess, Pioneer Venus (Washington, DC: NASA SP-461 Government Printing
69
Office, 1983), 26.
O. Frank Spurlock, Trajectory and Performance, as found in Pioneer Venus Orbiter Mission Flight Data Report, Box AC-
70
50 to AC-57, Division Atlas/Centaur Project Office, NASA GRC Records.

31276-chapter 4 3/9/05 9:27 AM Page 135
Heavy Lift 135
The High Energy Astronomy Observatory (HEAO-3) encapsulation in 1979. It was launched by Atlas-Centaur 53 on 20 September
1979. (NASA 8003540)
31276-chapter 4 3/9/05 9:27 AM Page 136
planet.71 The Pioneer Venus Orbiter and Multiprobe missions attracted significant press
coverage, as well as kudos from NASA Headquarters. In a letter to Larry Ross, NASA
Administrator Robert A. Frosch recognized the overwhelming success of Pioneer Venus activ-
ities and contributions of the Lewis team. The program represented the professionalism and
dedication of many groups of people at NASA, in universities, and in industry.72
In Search of Cosmic Rays: The High-Energy Astronomy Observatory
The earlier OAO series of heavy science satellites were followed by a new version of the orbiting
satellite observatory called the High-Energy Astronomy Observatory (HEAO), managed by
Marshall Space Flight Center. When the HEAO program began in 1970, NASA planned to launch
the new satellites with two Titan IIIC boosters. In 1973, TRW won a contract to build the space-
craft, but budget cuts threatened cancellation of the program. Marshall was forced to scale back the
program and reduced the weight of the satellites from 18,000 to 6,000 pounds so that they could
be carried aloft by Atlas-Centaur. At that time, the satellites represented the heaviest mission
attempted by an Atlas-Centaur.
The objective of the new class of HEAO spacecraft was to locate, identify, and analyze celestial
high-energy radiation sources.73 HEAO spacecraft were designed to survey the sky for x rays, inves-
tigate the shape and structure of these x-ray sources, measure gamma-ray flux and determine source
locations, and increase the understanding of various cosmic phenomena.74 With a series of three
launches planned (all aboard the improved Centaur D1 with a single burn), scientists hoped that
the HEAO would help them begin to uncover answers to the most intriguing mysteries of the
universepulsars, black holes, neutron stars, and supernovae.75
The HEAO program represented the first use of new in-flight Centaur retargeting software. This
feature could be used if Centaur lost power in flight. The software automatically changed the aiming
point for the vehicle to a lower orbit. Although this compromised the flight, it prevented complete
mission failure. No HEAO required the retargeting software because all three Centaurs performed as
expected; nevertheless, the advance was an important software feature for future missions.
71
For further information on the Pioneer Venus missions, see Eric Burgess, Venus: An Errant Twin (New York: Columbia
University Press, 1985); D. M. Hunten et al., eds., Venus (Tuscon: University of Arizona Press, 1983); and Robert J. Paradowski,
Pioneer Venus 1 and Pioneer Venus 2, in USA in Space, 500508.
72
Robert Frosch to Larry Ross, 10 April 1979, Box AC-50 to AC-57, Division Atlas/Centaur Project Office, NASA GRC
Records.
73
J. R. Brown, RL10A-33 Engine Performance Summary for Atlas Centaur Flight AC-45, Box AC-42 to AC-49, Division
Edwin Muckley, HEAO Mission Objectives, from the Director of Launch Vehicless report to the Director, Box AC-42 to
74
AC-49, Division Atlas/Centaur Project Office, NASA GRC Records.
75
Atlas-Centaur D-1A AC-52, Box AC-50 to AC-57, Division Atlas/Centaur Project Office, NASA GRC Records.
31276-chapter 4 3/9/05 9:27 AM Page 137
Heavy Lift 137
The mission of the first HEAO, launched on an Atlas-Centaur (AC-45) on 12 August 1977, was
to map the sky broadly for x and gamma rays. The countdown was interrupted for electrical storms
and the brief failure of a payload ground computer. Then shrimp boats encroached upon the range
safety area in the Atlantic Ocean. Once the computer was fixed, the storms over, and the shrimp
boaters banished, liftoff occurred with 13 minutes to spare. Centaur completed its single-burn ascent
and placed the 5,581-pound HEAO-1 in its designated circular orbit 276 miles above Earth.76 After
separation from the HEAO, Centaur began a deflection turn 11 seconds later to minimize any
possible chance of contaminating the HEAO. This turn was a 90-degree pitch and yaw maneuver in
which Centaur pointed its engines towards the Sun. At the same time, Centaur engaged in a 45-degree
roll. When Centaur finished repositioning, the propellant tanks settling and venting sequence was
initiated. This procedure further increased the distance between Centaur and the HEAO. The final
Centaur blowdown phase began 74 minutes after launch and continued for 305 seconds, lowering
it below the HEAO orbit.77 (During blowdown, any residual propellants are vented from the Centaur
tanks before the vehicle reenters the atmosphere.) HEAO-1 conducted nearly three full celestial
surveys and discovered 1,500 new sources of x rays in the sky.
NASA launched the second spacecraft in this series, HEAO-2, in November 1978. The
Atlas-Centaur (AC-52) launch sequence was similar to that of HEAO-1, but after payload
separation, Centaur carried out two experiments designed to improve its own future perform-
ance. The first experiment occurred 106 minutes into the flight. During the first orbital pass over
Florida, Centaur began a special spin-up experiment to demonstrate its ability to spin a commer-
cial spacecraft to 12 degrees per second before payload separation.
The second experiment occurred 109 minutes into the flight. Using newly implemented
Centaur tumble-recovery software made possible by the new Teledyne computer, the vehicle simu-
lated a severe tumble of 329 degrees of pitch and 584 degrees of roll. The new software allowed a
controlled steering maneuver to realign Centaur. This was the first NASA vehicle to have tumble-
recovery software encoded into its onboard computer software.78
The operational objectives of HEAO-2 were similar to those of the first craft in the series.
Again it searched the universe for x, gamma, and other cosmic rays and mapped their locations.
However, the HEAO-2 was larger and contained much more sophisticated observational equipment
than its predecessor. At 6,866 pounds, it became the heaviest object ever lifted into space by
an Atlas-Centaur, and it orbited Earth at 333 miles. The centerpiece of all the instrumentation
76
This was the first mission in which Centaur delayed its usual separation time because engineers wanted telemetry coverage over
Ascension Island. AC-45 Preliminary Analysis report, 1 September 1977, Box AC-42 to AC-49, Division Atlas/Centaur
77
Atlas Centaur Flight Evaluation AC-45, Box AC-42 to AC-49, Division Atlas/Centaur Project Office, NASA GRC Records.
78
Atlas-Centaur Postflight Analysis AC-52, Box AC-50 to AC-57, Division Atlas/Centaur Project Office, NASA GRC Records.
31276-chapter 4 3/9/05 9:27 AM Page 138
was a large x-ray telescope that had the ability to isolate and point directly at the most inter-
esting stellar emanations that HEAO-1 identified in the sky.79 HEAO-2 made over five
thousand specific observations and also discovered that both Earth and Jupiter emit x rays.
HEAO-3 contributed to a better understanding of interstellar matter by detecting cosmic-ray
particles and gamma-ray photons.
The final launch in the HEAO series occurred on 20 September 1979 with Atlas-Centaur
(AC-53). At 6,321 pounds, HEAO-3 was actually significantly lighter than HEAO-2. Its
objectives were similar to those of HEAO-1.80 It was the last mission of Atlas-Centaur to
launch a low-Earth-orbiting satellite until the commercial era of the 1980s.
Despite all of their mission complexity, almost all of the Atlas-Centaur missions succeeded
during the busy decade of the 1970s. The team lost one Orbiting Astronomical Observatory and
one Mariner mission, caused by the malfunction of a tiny integrated circuit in the Centaur
autopilot. Launches of two of nineteen Intelsat commercial satellites also failed. These setbacks
did nothing to lesson the resolve or growing confidence among the government-industry
Centaur team. They continued a tradition of innovation on Centaur with the upgrade to its
computer system. Centaur reliability was greatly enhanced by the ability to ground-test its restart
capability in a new vacuum facility completed at Lewis Research Centers Plum Brook Station in
the early 1970s.
The early successes of the Atlas-Centaur planetary program paved the way for even more
ambitious missions in the service of heavy science. The Titan-Centaur program, with voyages to
the Sun, Mars, and the outer planets, continued the legacy of the Atlas-Centaur program. Titan
III-Centaur sent spacecraft to probe deeper into the secrets of Earths planetary neighborhood
and enabled scientists to learn more about the evolution of Earth itself. Questions about the
future of expendable rockets seemed not to dampen the enthusiasm of the engineers and scien-
tists involved in the planetary science program for developing an even more powerful launch
vehiclethe giant Titan-Centaur.
79
Atlas/Centaur D-1A AC-52 HEAO-B Flight Data Report, February 1979, Box AC-50 to AC-57, Division Atlas/Centaur
80
Atlas-Centaur D-1A AC-53 HEAO-C Mission, Box AC-50 to AC-57, Division Atlas/Centaur Project Office, NASA GRC
Records. For further information about the HEAO satellites, see James Cornell, Wendell Johnson, and Carroll Dailey, eds., High
Energy Astronomy Observatory (Washington, DC: NASA EP-167 Government Printing Office, 1980); Wallace H. Tucker, The
Star Splitters: The High Energy Astronomy Observatory (Washington, DC: NASA SP-466 Government Printing Office, 1984);
Dave Dooling, Window on Violent, Cataclysmic Universe Closes, Space World (May 1982): 1617; and Dooling, The High-
Energy Astronomical Observatory, in USA in Space, 253256.
31276-chapter 5 3/9/05 9:48 AM Page 139
The Giant Titan-Centaur 139
Chapter 5
The Giant Titan-Centaur

But when Viking came alongit was the first Mars landerit was
too heavy for the Atlas-Centaur so we had to go with a bigger booster,
Titan . . . . All that effort spent to integrate the Centaur to the Titan,
[and] we only launched it seven times, because by the time we launched
Voyager the Shuttle was on the horizon and it was going to displace
expendable launch vehicles.
Joe Nieberding, Lewis Research Center
While Atlas-Centaur was proving its heavy lift capability, a giant launch vehicle that deliv-
ered even more power emerged in 1974Centaur coupled with a Titan IIIE booster.
Titan-Centaur was capable of lifting unprecedented heavy science missions. The heaviest
Atlas-Centaur planetary launch was the 2,201-pound Mariner 9 spacecraft; Titan-Centaur
would more than triple this capability with the launch of the 7,767-pound Viking probe.
Wernher von Braun, once skeptical that a liquid-hydrogen upper stage was even feasible, show-
ered the new Titan-Centaur with high praise shortly before its first launch. In Popular Science,
he announced the flightworthiness of a new generation of heavyweight interplanetary spacecraft.1
Standing 160 feet tall, the Titan-Centaur launch vehicle would give scientists the capability to learn
more secrets about the solar system than ever before. With a hint of regret, von Braun further wrote
that the Titan has the prospect of becoming NASAs largest launch vehicle in active use since our
two remaining Saturn Vs are unassigned and mothballed. Wernher von Brauns praise was justi-
fied. Bigger and more powerful than Atlas-Centaur, Titan III-Centaur proved to be one of the most
powerful and flexible launch vehicle systems ever designed.
Titan-Centaur launched key science missions to the Sun, Mars, and Jupiter with the Helios,
Viking, and Voyager probes. Ironically, despite its dramatic impact on the planetary space
program, NASA gave this Titan III-Centaur a life of only seven launches. Even before the Titan-
1
Wernher von Braun, Our Biggest Interplanetary Rocket, Popular Science (July 1974): 62.
31276-chapter 5 3/9/05 9:48 AM Page 140
The Seven Launches of Titan-Centaur for the Space Program

(19741977)
Launch Date Mission Vehicle Payload Objective
11 February 1974 Proof Flight TC-1 7,985 lb Simulated Viking launch.

Sphinx spacecraft. Failed.
10 December 1974 Helios 1 TC-2 1,093 lb Sent Helios 1 to the Sun.
9 September 1975 Viking 2 TC-3 7,767 lb Sent Viking 2 to Mars.
20 August 1975 Viking 1 TC-4 7,764 lb Sent Viking 1 to Mars.
15 January 1976 Helios 2 TC-5 1,102 lb Sent Helios 2 to the Sun.
20 August 1977 Voyager 2 TC-7 2,168 lb Sent Voyager 2 to Jupiter.
5 September 1977 Voyager 1 TC-6 2,171 lb Sent Voyager 1 to Jupiter.
Centaur had a chance to prove itself, its grave was already being prepared. As the new launch
vehicle rolled out to the launch pad for its inaugural flight, the expendable launch vehicle was
marked as a dying breed. John Noble Wilford wrote in the New York Times, The mammoth
rocket, the Titan III-Centaur, is expected to be the last new launching vehicle to be developed by
the National Aeronautics and Space Administration until the advent of the reusable space
shuttle.2 Despite its uncertain future, Centaur, mated so elegantly and powerfully with Titan,
would make possible the acquisition of significant new knowledge of the solar system.
Integrating Titan and Centaur
In the early 1960s, no one had predicted that Titan and Centaur would mate. NASAs
long-range launch vehicle plan was to continue to use Atlas-Centaur until a reusable launch
system or a nuclear-powered upper stage could be developed. The escalation of the Vietnam
conflict and President Johnsons War on Poverty disrupted these plans. Congress drastically
reduced expenditures for the civilian space program, and the development of a reusable launch
vehicle was put on hold. Because NASA needed a launch vehicle more powerful than Atlas-
Centaur to send heavier unpiloted planetary probes like Viking and Voyager into space in the
1970s, NASA began in 1967 to consider the possibilities of mating a Centaur upper stage with
the giant Titan III.3
2
John Noble Wilford, Rocket for Exploration of Planets Rolled to Pad for Proof Flight, New York Times (3 October
1973).
3
Linda Neuman Ezell, NASA Historical Data Book, Volume III (Washington, DC: NASA, SP-4012, 1988), 38.
31276-chapter 5 3/9/05 9:48 AM Page 141
Titan, an Air Force launch vehicle developed in the mid-1950s as an ICBM by the Glenn L.
Martin Company (later Martin Marietta Corporation), was a behemoth. But the tons of liquid-
oxygen oxidizer and RP-1 kerosene it required meant that it could not be readied for flight rapidly
enough in the event of a military emergency. To solve the propellant problem, the Martin
Company developed a new two-stage version called the Titan II that used storable, hypergolic
propellants. These propellants did not require the elaborate tanking procedures of Titan I. Titan
IIs power promised an unprecedented level of performance and cost-effectiveness as a launch
vehicle for NASA. Titan IIs first space mission to lift a two-person Gemini capsule was consid-
ered a triumph.4
The Titan III program began in 1962 when the Department of Defense authorized the
development of a launch vehicle for heavy military satellites. Titan III had two solid rocket strap-
ons for additional thrust. The Air Force tested the ability of Titan IIIA and IIIC to lift
13,000-pound payloads into low-Earth orbit in 1964. One of the most significant additions to
Titan was an upper stage rocket called Transtage intended to carry multiple satellites into
different orbits on a single mission. Because Transtage was plagued with development problems,
NASA chose Titan-Centaur as an upper stage for its heavy science missions.
Lewis engineers confronted the daunting prospect of integrating Centaur with Titan III.
Andrew Stofan, former head of the Propellant Systems Section, managed the Titan-Centaur
project office at Lewis. His counterpart at General Dynamics, Russ Thomas, led the devel-
opment of the Centaur D-1T and managed the industry team supporting the seven Titan
III-Centaur D-1T planetary launches. John Neilon, who headed Unmanned Launch
Operations at Kennedy Space Center between 1970 and 1975, recalled, It was a distinct
break from our way of doing business on Atlas-Centaur and Delta.5 All three organizations
had to learn how to deal with a new contractor, Martin Marietta, as well as the Air Force,
which owned both Titan and the launch pad.
Titan was launched from the Integrate, Transfer and Launch (ITL), its own launch facility.
The ITL consisted of the Vertical Integration Building (VIB), the Solid Motor Assembly Building
(SMAB), and two launch pads, 40 and 41. Pad 41 was modified for Titan-Centaur. All launches
for the Titan-Centaur were from Launch Complex 41, owned and controlled by the Air Force.
Prior to reaching the launch pad, the vehicles went through several check stages. The Titan
was assembled and tested at Martin Mariettas Denver, Colorado, plant. The Titan core vehicle was
then sent by plane to Florida. United Technologies built the solid rocket motors in Sunnyvale,
4
Frank Winter, Rockets Into Space (Cambridge: Harvard University Press, 1990), 9193. See also Robert L. Perry, The Atlas,
Thor, Titan, and Minuteman, in The History of Rocket Technology, ed. Eugene M. Emme (Detroit: Wayne State University Press,
1964), 160; and Barton C. Hacker and James M. Grimwood, On the Shoulders of Titans: A History of Project Gemini
(Washington, DC: NASA SP-4203, 1977).
5
Communication from John Neilon to the authors, 7 July 2002.
31276-chapter 5 3/9/05 9:48 AM Page 142
California, and shipped them to Florida by rail. The Aerospace Division of General Dynamics
assembled the Centaur in San Diego, California, with final preparations taking place at the Kearny
Mesa facility. From there, Centaur was taken to Miramar Naval Air Station, with a final flight on
a C-5A plane to Kennedy Air Force Station. At Kennedy, Titan was moved to the VIB, where it
was enclosed in a service structure. After checkout, a railroad locomotive pulled the core vehicle,
consisting of the two Titan stages, on tracks to the SMAB, where the two solid rocket motors were
attached to the Titan core vehicle. The Titan-Centaur was moved by rail one last time to Complex
41 to be mated with its payload.6
Since Pad 41 did not have a blockhouse, checkout and launch functions were carried out from
the Launch Control Center in the VIB, located several miles from the launch site. Neilon wrote,
Distance, rather than concrete provided safety. The same distance was also something of an
inconvenience; one did not simply walk out to the pad as one could on Complex 36, but we got
used to it. In summary, being different didnt mean wrong or inferior, just different. We all worked
together and made things work.7
When erected on the launch pad, Titan-Centaur was an impressive sight. Together, the two
vehicles stood 160 feet tall and were 10 feet in diameter, with a 14-foot-diameter shroud.
Flanking the Titan on either side were the two 10-foot Solid Rocket Motors (SRMs) built by
the Chemical Systems Division of United Technologies. The SRMs were capable of 2.4 million
pounds of thrust. Each weighed 500,000 pounds and stood 85 feet tall. When the SRMs
(considered stage 0) burned out, they were jettisoned and the Titans first and then its second
stage took over. These were called the core vehicle. Both burned liquid Aerozene-50 fuel and
red fuming nitric acid (RFNA) as an oxidizer. The first stage, 73 feet long and 10 feet in diam-
eter, provided 470,000 pounds of thrust. The 23-foot second stage delivered 100,000 pounds
of thrust. Aerojet manufactured both core vehicle engines.
The Centaur D-1T (stage 3) was 23 feet in length and weighed 39,000 pounds, including its
liquid-oxygen and -hydrogen propellants. Its two Pratt & Whitney main engines each provided
15,000 pounds of thrust.8 Many features of the Centaur D-1T for Titan were the same as those
of the Centaur D-lA for Atlas. For example, Centaur remained the brains because of its Teledyne
Digital Computer Unit with its 16-K, twenty-four-bit capacity. The computer had a twenty-five-
term instruction set, plus additional input and output instructions. The Titan version offered
greater computational reserve memory and input/output functions.9
6
Titan IIIE/Centaur D-1T Systems Summary, September 1973, Joe Nieberding Personal Collection.
7
Ibid.
8
John F. Kennedy Space Center, Titan/Centaur, press release, 20 August 1977.
9
Titan IIIE/Centaur D-1T Systems Summary, September 1973, Joe Nieberding Personal Collection. Unlike Atlas, which was
a one-and-a-half-stage booster, Titan was a two-stage rocket (stage 1 and stage 2) with two solid rocket motor strap-ons, called
SRMs (stage 0). Centaur became the third stage and the Delta TE-364 the fourth stage.
31276-chapter 5 3/9/05 9:48 AM Page 143
The D-1T Centaur had an updated electrical system and a totally redesigned shroud assembly
called the Centaur Standard Shroud (CSS), manufactured by Lockheed Missiles and Space
Company.10 It required extensive testing in the Space Power Facility at NASAs Plum Brook Station.
The shroud enclosed both Centaur and the payload, serving as environmental protection for the
payload on the ground and during the first few minutes after launch. It consisted of two halves
wrapped around Centaur and held in place by eight compressed springs mounted in pairs. The shroud
had a venting system that minimized the differences in structural pressures between the vehicle and
spacecraft during flight through the atmosphere.11 Once the vehicle left the atmosphere, the springs
forced the halves to separate and the shroud was jettisoned.12
The Titan-Centaur used a thermal radiation shield to reduce the heat that Centaurs liquid-
hydrogen fuel absorbed while it coasted in space. The radiation shield enabled Titan-Centaur to
coast in space for longer periods than Atlas-Centaur, which had a maximum coast period of about
30 minutes. Titan-Centaur could coast for over 5 hours, providing improved synchronous-orbit
capability and extended launch windows. Although Titan-Centaur payloads never needed these
extended coast periods, these capabilities were tested on the two Helios missions. Technical infor-
mation learned from these important experiments was later used in the development of Titan
IV-Centaur and the commercial version of Centaur in the 1990s.
Integrating Centaur with Titan proved to be a significant technical challenge. When Seymour
Himmel, then Associate Director for Rockets and Vehicles at Lewis, took charge of the integration of
the two vehicles, he found that the procedure was flawed. There was a lack of clear and effective tech-
nical communication between the two major aerospace companies involved. He directed that a
common database and set of definitions be established, documented, and maintained as the governing
Interface Control Drawings (ICD) for the integration activity. This was no minor task because the
two companies had differing cultures and design processes.13
In addition, he directed that complete structural and structural-dynamics analyses be conducted
for the combined vehicle. This effort underscored the nature of the integration problems because the
two companies did not use a common set of definitions for such analyses. Once a dictionary was
developed and documented, the process went forward much more smoothly.14
10
S. V. Szabo, Jr., and L. J. Ross, Titan/Centaur 1 Post-Flight Evaluation Report, April 1975, Box TC-1 TC-2 TC-3 Records,
11
K. A. Adams, Helios B Flight Data Report, Box TC-5, TC-6 Records, Division Atlas/Centaur Project Office, NASA GRC
Records.
12
Centaur Operations Division, Kennedy Space Center, Titan-Centaur-2 Helios 1 Field Report, 6, Box TC-1 TC-2 TC-3
Records, Division Atlas/Centaur Project Office, NASA GRC Records.
13
Interview with Seymour Himmel by Virginia Dawson, 1 March 2000, and his letter to her, dated 14 December 2002.
14
Interview with Seymour Himmel by Virginia Dawson, 1 March 2000.
31276-chapter 5 3/9/05 9:48 AM Page 144
An early problem was the difference between the diameters of the Titan core and the Centaur
shroud; the latter was four feet wider. The solution was to taper the Centaur shroud at the junction
point between the two rockets. This created a bulge, which gave the launch vehicle its unique
hammerhead shape. A more difficult problem resulted from the different temperatures of the two
rockets. Heat transfer from Titans extremely hot engines might compromise the very cold temper-
atures needed to keep Centaurs cryogenic propellants from becoming gaseous during launch. John
Gossett, Centaur Operations division chief at the Cape, told Craig Covault of Aviation Week that
to design a system with just the right amount of insulation without sacrificing weight required a
major design effort. Installing insulation in the shroud and bulkhead between the stages satisfied
the different thermal requirements.15
In addition to the technical challenges, dealing with the different cultures of the Atlas and
Titan programs proved a challenge for Lewis project managers. Air Force personnel were rotated
after a few years, a practice that prevented them from forming close personal relationships with
their counterparts at Martin Marietta. In contrast, years of working together had produced a
seamless relationship between the Lewis and General Dynamics people.16
Karl Kachigan, chief engineer of General Dynamics, recalled that the main problem was Air
Force procedures that accounted for Martin Mariettas glacial slowness in responding to requests.17
In addition to contending with the problem of protecting the proprietary technology of two rival
aerospace companies, Lewis engineers also found it necessary to persuade Martin Marietta to use
one of the innovations pioneered by General Dynamicsthe ADDJUST wind program, devel-
oped shortly before the first Titan-Centaur launch. Once installed, the new wind technique
worked flawlessly for every Titan-Centaur launch.
Before a launch, in addition to the prescribed launch checkout procedures, another set of
unofficial preparations typically occurred. These were various superstitions, traditions, and rituals
that NASA engineers brought with them to the Cape. For example, going back to Surveyor days,
the launch vehicle team from Lewis Research Center all had to eat at Ramons at Cocoa Beach,
and everyone had to order the Caesar salad. Joe Nieberding recalled, I never liked Caesar salad,
but who was I to say, being new to the program at the time. We had a great success record, and
we just didnt want to change anything.18 Since Titan was an Air Force rocket, Air Force rituals
now entered the mix. Key personnel from the mission would gather on the beach the night before
a launch. A quiet, somber, reverent mood more befitting a sance room than an engineering/mili-
tary gathering would settle over the group. A designated individual took a long knife and buried
15
Craig Covault, Titan 3E to Fill Gap in Space Boosters, Aviation Week and Space Technology (17 September 1973): 96; and
Craig Covault, First Titan 3, Aviation Week and Space Technology (4 February 1974): 5255.
16
Interview with Mike Benik by Virginia Dawson, 9 November 1999.
17
Interview with Karl Kachigan by Virginia Dawson, 7 June 2000.
18
31276-chapter 5 3/9/05 9:48 AM Page 145
it in the sand with the knife pointing in the launch direction. These were things you had to do,
Nieberding recalled. Its hard to believe, when you look back.19 Other engineers had more
private superstitions and rituals they performed before each launch. General Dynamics chief engi-
neer Karl Kachigan admitted, Whenever the launch was at T-minus 18 seconds, I would always
cross my fingers. One of the fellows took a picture of that once and he gave it to me. He said, We
dont have a chief engineer, weve got a witch doctor!20
An Inauspicious Beginning: The Proof Flight
Before the first missions to the Sun and Mars could be flown, a Proof Flight, designated as Titan-
Centaur 1 (TC-1), was necessary to demonstrate the flightworthiness of the vehicle. The Proof Flight
would have the same trajectory as the Viking mission to Mars that was scheduled to be launched in
1975. The Proof Flight was to carry the Viking Dynamic Simulator (VDS), a model of the Viking
spacecraft which was not intended to actually separate from Centaur. Its role was simply to deter-
mine flight loads to which the real Viking would be subjected during its launch.
Although the Proof Flight was not originally planned to carry any payload, Lewis engineers
responsible for the launch persuaded their colleagues at the Center to send an experimental
package on it. Known as the Space Plasma High Voltage Interaction Experiment, or SPHINX, it
was designed to measure how space plasmas interacted with high-voltage surfaces on the space-
craft. Lewis engineers planned to use this information to design better high-voltage systems to
operate in space environments. The year-long SPHINX mission would gather sample data from
many different types of plasma particles.21
The Proof Flight was a disaster.22 After a normal launch countdown on 11 February 1974
(except for a 45-minute hold resulting from questionable data from the boosters hydraulic
systems), the first Titan-Centaur lifted off at 9:48 a.m. Titan jettisoned its shroud and Centaur
separated, but at this point, the Centaur engines failed to start. Without power, Centaur went
into a freefall and was destroyed by Range Safety 748 seconds after liftoff.23 While the loss was
devastating for the members of the launch team, it was even more tragic for those involved in
the Lewis SPHINX program. They were looking for a ride, Joe Nieberding recalled. And we
said, We have this perfectly good rocket. It will be a whole Lewis thing with a Lewis payload
19
Ibid.
20
Virginia Dawson interview with Karl Kachigan, 7 June 2000.
Proof Flight, Launch Operations and Flight Events, Box TC-1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office,
21
NASA GRC Records.
S. V. Szabo, Jr., and L. J. Ross, Titan/Centaur 1 Post-Flight Evaluation Report, April 1975, Box TC-1 TC-2 TC-3 Records,
22
23
Kennedy Space Center report, Titan/Centaur-1 Proof Flight (SPHINX) Field Report, Box TC-1 TC-2 TC-3 Records,
31276-chapter 5 3/9/05 9:48 AM Page 146
and a Lewis rocket. They said, Great! We put them in the water and they never forgave us.24
Despite the failure, many of the evaluation criteria for the Proof Flight were met.25 The Proof
Flight demonstrated the structural integrity of the new Centaur and its capability to jettison the
Centaur Standard Shroud.26
The reasons for the failure of the Centaur engines were not immediately clear even after a
detailed investigation revealed that the problem had occurred in the boost pumps.27 The Failure
Review Board, chaired by Seymour Himmel, could not pinpoint the exact cause of the failure
of the Centaur boost pumps. The Board concluded that the two most probable causes were
either 1) freezing of the hydrogen-peroxide-generated steam as it expanded through the boost-
pump turbine or 2) ingestion of a foreign object that jammed the pump, preventing it from
rotating. No obvious source of such an object could be identified, since no change had been
made in the way the Centaur propellant tanks were built. Tests conducted at the B-2 facility at
the Plum Brook Station proved inconclusive. The best corrective action that could be devised
was a verification procedure to make sure that the boost pumps were free to rotate just before
liftoff. Subsequent launches were successful despite the fact that the root cause of the Proof
Flight failure had not been indisputably determined.
It took four years to solve the mystery. At General Dynamics, an employee that everyone called
the little old winemaker had been, for a long time, the only one to install the clip used to anchor
the propellant-utilization probe to the wall of the Centaur oxygen tanks. When he retired, the new
person assigned the task had great difficulty in installing the clip using the materials and procedures
specified in the shop instructions. Frequently, the clip fell off, but after several trials, he apparently
succeeded in securing it. This difficulty was reported to the little old winemaker, who confirmed
the difficulty of anchoring the clip to the wall of the tank. He had concluded that the length of the
rivet specified in the drawings was too short, so he had always obtained and used a longer rivet that
worked much better. Unfortunately, he had never reported the problem or his solution. Since the
new mechanic had installed the probe for the proof flight, it was concluded that the clip had fallen
off and jammed the boost pump.28 What was ironic was that none of the new Titan-Centaur
components were at fault and the failure was part of the old, proven Centaur. Even though the first
Titan-Centaur failed, both industry and government engineers learned an enormous amount
24
25
Craig Covault, Centaur Failure Stirs Minimal Concern, Aviation Week and Space Technology (18 February 1974): 20.
A. B. Yanke, TC-1 Flight Data Preliminary Analysis, Box TC-1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office,
26
NASA GRC Records.
27
General Dynamics report, Titan/Centaur Flight Evaluation TC-1, Box TC-1 TC-2 TC-3 Records, Division Atlas/Centaur
28
Interview with Seymour Himmel by Virginia Dawson, 1 March 2000.
31276-chapter 5 3/9/05 9:48 AM Page 147
about the new rocket system. The next six Titan-Centaur launches were flawless. NASAs most
ambitious probes to the Sun, Mars, Jupiter, Saturn, Uranus, and Neptune were hailed as impor-
tant contributions to space science.
A German Partnership To Explore the Sun
In 1969, the United States and the Federal Republic of Germany (West Germany) agreed on
their first joint space project to send two probes to explore the mysteries of the Sun, the star that has
dominated human interest in the skies for millennia. These probes would provide unprecedented
observations of the Sun, reveal important data about the relationship between the Sun and Earth,
and also gather key data about the solar wind, magnetic and electric fields, cosmic rays, and cosmic
dust.29 The spacecraft would also test Einsteins theory of general relativity.30 The mission was called
Helios, named in honor of the ancient mythological Greek god of the Sun.
Originally, the Germans had planned to use Atlas-Centaur, but as the design evolved, glass
plates and mirrors had to be added to dissipate the intense heat generated by the Sun. These addi-
tions brought the crafts weight close to the limits of what the Atlas-Centaur could lift. The
Americans convinced the Germans that by using the untested Titan-Centaur, they would not have
to compromise the scientific objectives of the mission by eliminating valuable scientific equip-
ment. The Americans also promised the Germans that the second Helios flight would occur after
the Viking Proof Flight. Reassured that any system flaws would be worked out by the time Helios
was launched, the Germans agreed to use the untried vehicle. They were willing to tolerate these
risks because of the payload weight opportunities that Titan-Centaur offered.31
Lewis Research Center managed the launch vehicle, Goddard Space Flight Center directed
the overall mission, and Kennedy Space Center managed operations at the launch site. In
Germany, the Federal Ministry for Research and Technology (Bundesminister fr Forschung und
Technologie) provided the spacecraft. The Gesellschaft fr Weltraumforschung controlled all
the technical facets of their construction. It also ensured that the prime contractor for the space-
craft, Messerschmidt-Boelkow-Blohm, as well as the other German contractors, worked
together to deliver the probes to NASA on schedule. Each Helios was a relatively short, sixteen-
sided, cylindrical spacecraft with two solar arrays attached to its front and aft ends. Each Helios
carried seven German scientific experiments. The United States provided three additional
experiments on each probe, along with tracking data from NASAs Deep Space Network (DSN).
29
See Karl Hufbauer, Exploring the Sun: Solar Science Since Galileo (Baltimore and London: The Johns Hopkins University Press,
1991); and Titan-Centaur D-1T, TC-2, Launch Operations and Flight Events, Box TC-1 TC-2 TC-3 Records, Division
30
Glenn Research Center report, Titan-Centaur D-1T TC-5 Launch Operations and Flight Events, Box TC-5, TC-6 Records,
31
Press, 2000), 87.
31276-chapter 5 3/9/05 9:48 AM Page 148
A Helios spacecraft prototype is encapsulated in its payload fairing at Kennedy Space Center on 27 September 1974. This work was
in preparation for a future Titan-Centaur flight. (NASA KSC_74P_0222)
31276-chapter 5 3/9/05 9:48 AM Page 149
Most importantly, the Americans supplied the launch vehicles for these missionstwo Titan
IIIE/Centaur/Delta TE-364 rockets.
As the launch approached, however, tension grew in the relationship between the Germans and
the Americans. After the Viking Proof Flight failed, the Germans feared that NASA was simply using
their Helios mission to test the Viking launch profile before launching the expensive Viking space-
craft.32 Specifically, they objected to using their mission to test a two-burn capability needed for
Viking when Helios did not require it. If the restart were not achieved, this high-profile interna-
tional mission would be stranded in low-Earth orbit.
Other concerns focused on the Centaur Teledyne digital computer. A circuitry problem within
the computer was discovered in July 1974. During qualification testing, engineers found that 1 in
every 10,000 microelectronic modular assemblies failed. Since Centaur contained 2,400 modules,
this became a very serious concern. There was a possibility that the Helios mission, as well as a
scheduled Atlas-Centaur Intelsat launch, would be delayed. Vibrations caused chips to fall off the
computer circuit board.33 Teledyne worked around the clock and found a solution to the problem.
By means of a different bonding technique, the chips remained securely fastened.
Despite the controversy over the two-burns (which NASA refused to give up) and the concerns
over the Teledyne computer chips, the launch went forward. During the first launch attempt on 8
December 1974, a malfunction was detected in the Centaur liquid-hydrogen pump. When the
cause of the problem was not resolved during the standard 10-minute delay, engineers postponed
the launch. They later discovered that the problem lay with a faulty transducer assembly, which they
easily replaced. Two days later, the giant vehicle lifted off.
The first Titan core burn lasted 258 seconds, at which point the stage was shut down and
jettisoned. One second later, the Titan second stage engines started. The next major launch
event was the jettisoning of the Centaur Standard Shroud (CSS), which occurred 318 seconds
after liftoff. At 469 seconds into the launch, the Titan second stage engines shut down. Four
seconds later, Centaur separated from Titan. At this point, both German and American engi-
neers collectively held their breaths because this was the exact moment at which the Proof
Flight had failed. Eleven seconds later, Centaur passed the critical test when its main engines
started. This first burn lasted for 100 seconds. At this time, the Centaur, Delta TE-364, and
Helios were traveling at 25,600 feet per second. The main engines then shut down on schedule
and Centaur began a nearly 22-minute parking orbit.
Toward the end of the parking orbit, Centaurs thrusters reoriented the vehicle in preparation
for the second burn. This burn also occurred on schedule and lasted for 273 seconds, giving the
rocket a velocity of 37,400 feet per second. The second main engine cutoff occurred at 36 minutes
32
33
Daniel Shramo, quoted in Craig Covault, Centaur Computer Trouble May Delay Helios, Intelsat, Aviation Week &
Space Technology (5 August 1974): 48.
31276-chapter 5 3/9/05 9:48 AM Page 150
after liftoff, and the engine remained silent for 72 seconds after the end of the second Centaur
burn. Then the Delta TE-364, mated to the Helios probe, separated from Centaur. The Delta TE-
364 motor started 42 seconds later and burned for 44 seconds. Helios separated after 72 seconds
and journeyed alone the rest of the way to the Sun.34 The successful launch meant a great deal to
the Titan-Centaur teams. Writing in Aviation Week and Space Technology, Craig Covault said,
Launch of the Helios solar probe from Cape Canaveral last week is as much a success for NASAs
Titan Centaur program as it is for the West German team that built the Helios spacecraft.35
With Helios safely on a trajectory to the Sun, Centaur began an additional set of experiments
to improve future performance. After the Delta TE-364 separated, Centaur began a 1-hour coast
perioddouble the length of the previous successful parking orbit. During this period, Centaur
slowly turned itself to point toward Earth so that ground engineers could pick up more data from
the rocket. To assist in this procedure, data transmission was switched to two high-gain, narrow-
beam antennas. At the end of the 1-hour coast, Centaur began its third engine start in space with
a burn of 11 seconds. This was followed by another coast period and included several 180 rolls
of the rocket (called thermal maneuvers, which kept one side of the rocket from overheating), a
vent sequence, and other thermal control operations. After 3 hours, the Centaur engines had
their fourth and final burn, which lasted 47 seconds. This final coast phase lasted almost 27
minutes and included a boost-pump experiment to determine the rockets capability to recover
from a severe cavitation condition (when the pump sucks gas rather than liquid propellant), as
well as a hydrogen peroxide depletion experiment. As a result of these post-Helios experiments,
Lewis Research Center engineers concluded, All of the Centaur systems performed satisfactorily,
the design parameters for zero gravity coasting were verified, and no significant problems were
encountered.36
The post-Helios experiments proved that Centaur was capable not only of four-burn
missions, but also of vastly extended parking orbits. This capability was accomplished through
six key modifications to the basic Centaur design: a new three-layered aluminized Mylar radi-
ation shield that dramatically reduced the hydrogen tank sidewall heating, an improved tank
vent control system that enabled the propellant tanks to vent only when necessary, the addi-
tion of new purges to keep key lines free of liquid, the ability to engage in a 180 thermal
roll every 28 minutes to keep a uniform temperature on the rocket, the capability to fire
hydrogen peroxide axial thrusters only to collect propellants during a tank vent or engine
34
K. R. Krebs, TC-2 Centaur Stage Post Flight Data Review at LeRC, 28 and 29 January 1975, General Dynamics, Box TC-
1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office, NASA GRC Records.
35
Craig Covault, Helios Launch Verifies Titan Centaur, Aviation Week and Space Technology (16 December 1974).
36
A. B. Yanke, TC-2 Centaur Stage Post Flight Data Review at LeRC, 28 and 29 January 1975, General Dynamics, Box TC-
1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office, NASA GRC Records.
31276-chapter 5 3/9/05 9:48 AM Page 151
restart, and, finally, the implementation of improved tank-pressurization techniques. These

experiments established a major new capability for future missions.37 Engineers at Lewis
predicted that more than 7 hours of coasting would have been possible before tank venting was
necessary.
The Helios probe contributed significantly to humankinds knowledge of the Sun. Just
over three months after liftoff, the spacecraft reached its closest point to the Sun, 28.7 million
miles, on 15 March 1975. Data from Helios were sent back to scientists at the German
Control Center in Oberpfaffenhofen, West Germany. All of the scientific instruments func-
tioned normally and sent back information from the closest point to the Sun to which any
craft had ever voyaged. The Helios craft produced new information about the solar wind, the
emission of a stream of plasma from the Sun, which distorts the magnetic field and produces
a shock wave similar to a wind. Study of the solar wind is important because every second, the
Sun expels roughly one million tons of ionized particles at great speeds. This wind is propelled
throughout the solar system, greatly affecting Earths environment. Helios voyaged to the outer
solar corona, where these particles picked up their final, massive accelerations. Scientists hoped
that data gathered at this point would be very revealing about the nature of these particles.
Scientists and engineers on the Viking program also celebrated the success of Helios. Because
Mars launch windows open every two years, any launch malfunction would have delayed for two
years the planned Mars voyage that was scheduled to take place nine months after Helios. Walter
Jakobowski, Viking program manager, said that while watching the Helios launch, We were
holding our breath. That launch vehicle had to work for us. We were very pleased with the overall
performance.38 Ironically, after the launch of the second Viking, a fire on the launch pad threat-
ened to jeopardize the second Helios flight. With help from the Air Force, the launch facility was
repaired in time for the launch of Helios 2 (TC-5) on 15 January 1976.
This was the fifth Titan-Centaur launch, with two successful Viking launches occurring
between the two Helios launches. The launch vehicle and launch sequence for Helios 2 were
nearly identical to those of Helios 1. In April 1976, this probe passed 2 million miles closer to
the Sun than Helios 1. As a result of this close passage, the Helios craft was subjected to 10
percent more heat than Helios 1 and continued to operate through temperatures of over 700F,
the melting point of lead. Terry Terhune, Engineering Chief of the Centaur Operations
Division at Kennedy Space Center, called Helios 2 a perfect mission with flawless operation of
Raymond F. Lacovic, Centaur Zero Gravity Coast and Engine Restart Demonstration on the TC-2 Extended Mission, Lewis
37
Research Center, October 1975, Box TC-1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office, NASA GRC Records.
38
Walter Jakobowski, quoted in Victor K. McElheny, 76 Mars Landing Gains With Launching of Titan, New York Times (18
December 1974): 15.
31276-chapter 5 3/9/05 9:48 AM Page 152
all ten experiments (seven German and three American) and absolutely no problems with the
any of the spacecrafts systems.39
As with the previous voyage, Lewis engineers were able to perform several post-Helios flight
maneuvers. Centaur had about 4,000 pounds of fuel remaining after Helios separated. This
massive fuel supply enabled Centaur to make an unprecedented five additional engine restarts.
These experiments demonstrated a high-altitude, synchronous-orbit injection capability with a
parking orbit that lasted over 5 hours between the second coast and the third engine start. Centaur
also tested a very short coast of 5 minutes.40
The German engineers who originally designed the Helios probes had set a specific definition
of what results would constitute a success for the mission. They had decided that if the craft
returned data from its launch date to the end of the first perihelion passage, plus a mission exten-
sion of eighteen months, they would consider it a success. But the unexpected happened. They
greatly underestimated the abilities of their own craft. Helios 1 had a lifespan of eleven years, and
Helios 2 did not die until 3 March 1990, fourteen years after its initial launch.
The decade-long gathering of data from the Helios probes continues to actively shape scien-
tific knowledge of the Sun. In 1984, at the ten-year anniversary of the program, the Germans
invited all of the American scientists who had participated in the project to Germany for an all-
expenses-paid trip. American scientist Al Opp summarized the significance of Helios. He said
that Helios had given the scientific world a detailed, close-in view of the Sun over vastly
differing solar conditions. It has enabled scientists to observe cosmic rays coming into the solar
system from our galaxy, and when combined with results from deep space probes and earth-
orbiting satellites, has given a detailed picture of the structure of the solar system and of the
characteristics of low energy galactic cosmic rays.41
Interestingly, the number of scientists working on the Helios project has increased with time.
Although several of the scientists who designed and ran the first experiments are still active in the
program, there are many more scientists now studying the Sun with Helios data than were ever
originally involved with the project. This expansion was made possible through the Helios inter-
national data centers, which provide information to the wider space science community. Rainer
Schwenn and Eckart Marsch, two of the scientists working on Helios today, wrote that these
missions shaped our current perception of the Sun and the inner heliosphere.42 Equally signifi-
39
C. A. Terhune, Titan/Centaur-5, Helios 2 KSC Field Report, 30 June 1976, Box TC-5, TC-6 Records, Division
40
Raymond Lacovic, Propellant Management Report for the TC-5 Extended Mission, September 1977, Box TC-5, TC-6
Records, Division Atlas/Centaur Project Office, NASA GRC Records.
41
Herbert Porsche, ed., 10 Years Helios (Munich: Wenschow Franzis Druck GmbH, 1984), 9.
42
Rainer Schwenn and Eckart Marsch, eds., Physics of the Inner Heliosphere I: Large-Scale Phenomena (London: Springer-Verlag,
1990), 6.
31276-chapter 5 3/9/05 9:48 AM Page 153
Titan-Centaur 2 at Complex 41, Cape Canaveral Air Force Station, with Helios 1 as its payload. The rocket lifted off on 10
December 1974 and was the first successful launch of a Titan-Centaur. (NASA KSC-74P-0298)
31276-chapter 5 3/9/05 9:48 AM Page 154
cant was the Helios contribution to the international approach to space science. Schwenn and
Marsch asserted, It should be recognized in retrospect that the Helios mission at the time of its
conception represented an unprecedented challenge to the . . . network facilities available for
national and international collaboration.43 The significance of Helios remains not only the
knowledge of the inner solar system it reaped, but also the international cooperation it fostered.
The Search for Extraterrestrial Life: Viking
The two Viking spacecraft launched in 1975 were massive spacecrafteach included an
orbiter and robotic lander. So large were these spacecraftover 7,600 poundsthat they
dwarfed the 2,200-pound Mariner 9, the largest planetary craft yet launched by Atlas-Centaur.
With four scientific experiments packed into each orbiter and nine more aboard the landers, the
Viking missions represented the heaviest science ever undertaken in outer space. Centaur would
once again prove its heavy lift capability and make possible an unprecedented soft landing on
the surface of another planet.
Everyone involved with the Viking program knew the expense and prestige associated with it.
In the late 1960s, NASA had planned to launch Viking on a Saturn V, but after Saturn production
was canceled, NASA selected Titan-Centaur for this important mission.44 In Beyond the Moon,
Robert Kraemer described the repercussions for NASA should the mission fail. Viking was so
expensive that a mission failure would shake the entire agency down to its toes, and even a modest
cost overrun, say, just 10 percent, could cause the cancellation of several smaller science projects,
he wrote. That would generate a science revolt against future planetary missions that could last for
years.45 With Vikings cost nearing $1 billion, engineers at Lewis knew that the future of NASAs
space science program was riding on their vehicle.
Mars has been one of the most compelling objects in the sky throughout history. Because of
its strange red color, ancient astronomical observers associated it with fire and blood and named
it Mars after the Roman god of war. In 1608, Galileo had discovered the surface features of the
planet, and in 1659, Christian Huygens made the first sketches of the planets dark region, now
known as Syrtis Major. With the ability to pinpoint a location on the surface, he was able to prove
that Mars rotated like Earth. By the late nineteenth century, Giovanni Schiaparelli and Percival
Lowell had published theories that Mars was inhabited by intelligent creatures; they used what
appeared to be unnaturally straight lines of canals as evidence. But by the twentieth century scien-
tists were becoming increasingly skeptical of finding life on Mars. The Mariner flyby missions in
the 1960s and also the Mariner 9 orbiter in the early 1970s had indicated that Mars was uninhab-
43
Ibid.
44
45
Press, 2000), 136.
31276-chapter 5 3/9/05 9:48 AM Page 155
ited, yet they had also revealed a planet that was both dynamic and evolving. This new knowledge
made Mars a compelling location for further scientific study. The goal of the Viking missions was
to learn more about the planet Mars with special emphasis on the search for extraterrestrial life.46
The Viking missions were the third and fourth launches with Titan-Centaur. Viking 1 on TC-
4 was launched on 20 August 1975, nine days later than originally scheduled. Within 2 minutes
of launch on 11 August, one of the forty-eight valves in the thrust vector control system of one of
the two solid rocket boosters of the Titan indicated that it was leaking its propellant. The launch
was scrubbed because, as Seymour Himmel said, You dont commit to launch unless you under-
stand what has happened.47 The errant valve was removed and replaced. It was determined that
the source of the leak was the pintle, a conical steel plug that acts to seal the valve outlet when
fluid is not supposed to flow. Himmel directed Pratt & Whitney to conduct an x-ray examination
of the pintle. The x ray revealed a string-like contaminant in the pintle from base to tip. Having
ascertained the nature of the failure and that the replacement valves pintle did not come from the
same production batch, Himmel approved the launch.
With the launch rescheduled for 14 August, technicians now discovered that the Viking
orbiters batteries had dropped from 37 to 9 volts. A rotary switch that was supposed to be turned
on 7 minutes after liftoff had been mistakenly left in the on position for days. To fix this
problem, the entire Viking spacecraft had to be removed for detailed troubleshooting. During this
time, the Lewis launch team was also in the midst of managing other launches as well. An Atlas-
Centaur Intelsat launch was scheduled sixteen days after Viking. Our focus was intense, Larry
Ross recalled. We were launching a lot of Atlas-Centaurs right in the middle of the Voyager
Viking program on Titan-Centaurs, and that was daunting. And the penalty for missing the plan-
etary window is severe. The time and money are extraordinary.48
Finally, on 20 August, Titan-Centaur (TC-4), carrying the first Viking spacecraft, lifted off at
the opening of its 71-minute launch window.49 The vehicle functioned flawlessly through the
various intricacies of liftoff. Centaur separated from Titan at the 473-second mark. Centaur then
started its main engines 11 seconds later, and they burned for 127 seconds. Centaur, still carrying
the heavy Viking payload, then coasted in space for 918 seconds. When this parking mode ended,
its main engines started a second burn that lasted for 316 seconds, from roughly the 25th to the
30th minute of the launch. At 2,066 seconds after the launch, Viking separated from Centaur on
46
Viking Encounter Press Kit, Box TC-1 TC-2 TC-3 Records, Division Atlas/Centaur Project Office, NASA GRC Records.
47
Interview with Seymour Himmel by Virginia Dawson, 1 March 2000; and Himmel letter to interviewer dated 14 December
2002.
48
Interview with Larry Ross by Mark Bowles, 29 February 2000.
49
TC-4 Centaur Stage Flight Data Preliminary Analysis, 11 September 1975, Box TC-1 TC-2 TC-3 Records, Division
31276-chapter 5 3/9/05 9:48 AM Page 156
a perfect trajectory to Mars. Twenty days later, the second Viking (TC-3) was launched 3 minutes
before a potentially mission-canceling thunderstorm. This second launch was virtually identical except
for longer coast phases. The parking mode lasted for 1,092 seconds, 174 seconds longer than the first
Viking, in order to put the spacecraft on an optimal trajectory to Mars.50
On 19 June 1976, Viking 1 first entered into Mars orbit. Although every effort was made to
put the lander on the surface of the planet in time to commemorate the United States July 4th
bicentennial celebration, a safe landing site could not be found in time. Through examination of
the photos taken from orbit, scientists determined that the first site was unsafe. After studying
other options, they agreed on the western slope of Chryse Planitia. In July, Viking 1 touched
down on the Martian surface, with its sister ship landing in September 1976 at Utopia Planitia.51
The mission was scheduled to last 90 days, but it did not officially end until November 1982,
when the first lander (which was the last of the four craft to stop transmitting) made its final
communication.
The contributions of the Viking missions to humankinds knowledge of Mars were made
possible by the ability of Centaur to lift the heavy payload containing the hardware needed for the
planned scientific experiments.52 Although the knowledge gained by these missions did not alter
the fundamental understanding of the red planet established by the earlier flyby missions, the
results were considered significant. Two space scientists wrote, The Viking mission did not revo-
lutionize our ideas about Mars as Mariner 9 had done, but building on the foundation of its
Mariner and Mars precursors it far surpassed them in the variety, quality and quantity of its data
because of four factorsthe number of spacecraft, the number of experiments, the duration of the
mission, and the telemetry rate.53
The orbiters and landers returned thousands of images of the planet covering the entire
Martian surface during all of its seasons. These photos also included both of the Mars moons. Of
all the information returned from the Viking missions, the most controversial type was the biolog-
50
TC-3 Centaur Stage Preliminary Flight Analysis, 26 September 1975, Box TC-1 TC-2 TC-3 Records, Division
51
Robert Godwin, ed., Mars: The NASA Mission Reports (Burlington, Ontario, Canada: Apogee Books, 2000), 175183.
52
For further information on the Viking missions, see Eric Burgess, To The Red Planet (New York: Columbia University Press, 1978);
Robert Zubrin, The Case for Mars (New York: The Free Press, 1996); Michael H. Carr, The Surface of Mars (New Haven: Yale
University Press, 1981); Henry S. F. Cooper, Jr., The Search for Life on Mars: Evolution of an Idea (New York: Holt, Rinehart, and
Winston, 1980); NASA Viking Lander Imaging Team, The Martian Landscape (Washington, DC: NASA SP-425 Government
Printing Office, 1978); Cary R. Spitzer, ed., Viking Orbiter Views of Mars (Washington, DC: NASA SP-441 Government Printing
Office, 1980); Mark Washburn, Mars at Last! (New York: G.P. Putnams Sons, 1977); Gerald A. Soffen and Conway W. Snyder, The
First Viking Mission to Mars, Science (27 August 1976): 759; Philip J. Sakimoto, Viking 1 and 2, in USA in Space, eds. Frank N.
Magill and Russell R. Tobias, vol. 3 (Pasadena, CA: Salem Press, Inc., 1996), 809817.
53
Conway W. Snyder and Vassili I. Moroz, Spacecraft Exploration of Mars, in Mars, eds. Hugh H. Kieffer, Bruce M. Jakosky,
Conway W. Snyder, and Mildred S. Matthews (Tuscon and London: The University of Arizona Press, 1992), 103.
31276-chapter 5 3/9/05 9:48 AM Page 157
ical data.54 The key conclusion came from a gas chromatograph/mass spectrometer experiment
that detected no organic material in the soil.
Although the search for life on Mars had been one of the main reasons that the Vikings were
designed and funded, the controversy continues. Pictorial evidence acquired from the Viking
orbiter resulted in a perplexing mystery about life on Mars and, for some, evidence that it was once
home to an intelligent civilization. On the thirty-fifth pass around the planet, as it was looking for
a landing site for the second lander, the first Viking orbiter traveled 1,000 miles above the
Cydonia region. Photos from this region showed a strange collection of what some called
ruins, including, most ominously, a hollow-eyed, human-like face. While NASA vehemently
discounted any extraterrestrial intelligence behind these findings, others claimed that the
truth was being covered up to prevent mass hysteria on Earth. According to science writer
Richard C. Hoagland, Either these features on Mars are natural and this investigation is a
complete waste of time, or they are artificial and this is one of the most important discoveries
of our entire existence on Earth.55 What did the Viking orbiter really photograph from its
vantage point in the Martian sky? Although the debate continues, the Mars Global Surveyor,
launched in November 1996, suggests that wind erosion and some trick of light and shadow
probably account for the Sphinx-like face.56
Saving the Grand Tour: Voyager
The final two missions for Titan IIIE-Centaur were considered the pinnacle of NASAs plan-
etary efforts of the 1970s. Already the most productive era in the exploration of the solar system,
the Voyager flights to Jupiter, Saturn, and beyond promised to yield the most exciting and signif-
icant information yet about Earths largest planetary neighbors. That these probes would not die,
but would slowly fade away into the darkness of the galaxy in the hopes of one day encountering
intelligent life, also fueled the imagination of popular culture. For example, in the first Star Trek
movie, the crew of the Enterprise confronts a strange alien intelligence known only as VGER. By
films end, the audience learns that the entity is one of the Voyager probes, which has become
self-aware over centuries of seeking knowledge in space, yet which incorrectly names itself
because certain letters on the craft have eroded over time. NASAs own words helped fuel some
of these fantastic ideas. One JPL press kit stated that the Voyagers are able to care for themselves
and perform long, detailed and complex scientific surveys without continual commanding from
the ground.57 Planetary Programs Director Robert Kraemer said that project scientists had
54
Hugh H. Kieffer, Bruce M. Jakosky, and Conway W. Snyder, The Planet Mars: From Antiquity to the Present, Mars, 15.
55
Richard C. Hoagland, The Monuments of Mars: A City on the Edge of Forever (Berkeley: North Atlantic Books, 1987).
56
John Noble Wilford, Mars Beckons: The Mysteries, The Challenges, The Expectations of Our Next Great Adventure in Space (New
York: Alfred A. Knopf, 1990), 111.
57
JPL Press Kit, Voyager, Box TC-5, TC-6, TC-7 Records, Division Atlas/Centaur Project Office, NASA GRC Records.
31276-chapter 5 3/9/05 9:48 AM Page 158
The Centaur stage undergoes tests in the Vertical Integration Building at Kennedy Space Center on 19 October 1976. This Centaur
was mated with a Titan for a Voyager launch one year later. (NASA KSC_76PC_0526)
31276-chapter 5 3/9/05 9:48 AM Page 159
designed a great deal of artificial intelligence into Voyagers brain.58 This type of rhetoric,
combined with the imagination of science fiction, made it a short leap to conceive of Voyagers
becoming self-aware.
Actually, with much less fanfare or fictional speculation, the computer intelligence of the
Centaur rocket literally saved Voyager 1 after a nearly fatal error by Titan. Voyager 2 launched
first because Voyager 1 traveled on a more efficient trajectory. Launched 16 days after Voyager
2, Voyager 1 was scheduled to beat its sister ship to Saturn. (NASA engineers place more
emphasis on arrival date than liftoff order.)
Titan-Centaur (TC-7) looked majestic next to the gantry as it sat poised for the launch of
Voyager 2 in August 1977. One day before liftoff, Bruce Murray, the Director of JPL, and his wife
stood on the gantry, 160 feet above Pad 41, gazing at the spacecraft. In the distance, lightning illu-
minated an ominous sky, and they quickly planned their escape route if the storms came closer.
They were standing next to the largest United States launch vehicle, loaded with 700 tons of
volatile fuel. Murray recalled, As we gingerly descended the winding stairway of the gantry tower
on that glorious Florida afternoon, a quick glimpse of a busy engineer working inside the Centaur
surprised me. It was a reminder that unmanned rockets do indeed fly by the skill and dedication
of humanson the ground. The engineer was checking this giants intricate computer brain . . .
. The lightening drifted harmlessly seaward into a darkening sky.59
During liftoff into the morning skies off Cape Canaveral in Florida on 20 August 1977
exactly two years to the day after the launch of Viking 1the Titan stages operated flawlessly. The
Centaur Standard Shroud was successfully jettisoned 265 seconds after liftoff, and 203 seconds
later, Centaur, carrying the Voyager 2, separated from Titan. The main engines of Centaur started
4 seconds later, burning for a total of 101 seconds. At that point, Centaur coasted into a parking
orbit around Earth and began its longest zero-gravity coast during an operational mission. The
TC-2 and TC-5 postflight maneuvers proved that Centaur was capable of very long zero-gravity
coasts in space. The Voyager flight made use of this capability for a nearly 43-minute coast.
Richard Geye of Lewis Research Center noted, This zero-g mode of coast was selected to
provide maximum Centaur performance to the Voyager missions.60
At the end of the coast, Centaur realigned itself to place it in the exact position desired by
the trajectory engineers for the next phase of the flight. Centaur then began its propellant
settling procedure before the second main engine start. This second burn lasted 339 seconds,
accelerating Centaur and Voyager beyond Earth-escape velocity conditions. After engine
58
Kraemer, Beyond the Moon, 189.
59
Bruce Murray, Journey into Space: The First Three Decades of Space Exploration (New York: W. W. Norton & Company,
1989), 143.
60
R. P. Geye, Introduction, Titan/Centaur D-1T TC-7 Voyager 2 Flight Data Report, Box TC-5, TC-6, TC-7 Records,
31276-chapter 5 3/9/05 9:48 AM Page 160
cutoff, Centaur coasted for another 89 seconds. During this time, the computer inside Voyager
initiated its separation sequence, and Voyager was deployed from Centaur. Fifteen seconds
later, the TE-M-364-4 solid rocket motor (the Propulsion Module) on the Voyager ignited for
45 seconds and gave the spacecraft the additional 6,200-feet-per-second velocity beyond the
46,000- to 48,000-feet-per-second velocity that Voyager needed to inject it into the Jupiter
transfer orbit. This Propulsion Module weighed approximately 2,700 pounds and was jetti-
soned after its engines burned out.
What was left after all of these burn phases was the Voyager Mission Module speeding
toward a rendezvous with Jupiter roughly two years in the future. The Mission Module
contained all of the instruments, communication and data capability, command and control
functions, electrical power, and trajectory-adjustment features that were essential for carrying
out the scientific objectives. Engineers based its design on Mariner Mars 1971 and the Viking
Orbiter.61
While the Voyager 2 spacecraft had suffered some technical problems, Titan-Centaur had
performed perfectly. This was not the case for Voyager 1 (launched after Voyager 2) on Labor
Day, 5 September 1977. Titans final-stage engine shut down too soon. Over the control
rooms radio came the announcement, Solid booster burnout and separation, Titan core igni-
tion . . . Titan burnout and separation. Those in the room immediately knew that there was
a problem because the burnout came much sooner than anticipated. The problem had been
caused by a failure of Titan hardware, which choked off the propellant flow, leaving 1,200
pounds of propellant unburned. Only if Centaur could make up for this failure during the
trajectory-insertion burn could the mission be saved. Bruce Murray, who was watching from
the control room, recalls, This was serious. Titan had underperformed not propelling Centaur
and Voyager fast enough. It was the Centaur computers that recognized the deficiency and began
to correct the problem. In awe of the Centaur brain, Murray said that this was the brilliant
part because it could compensate for any shortfall in the propulsion of the earlier stage.
Centaur extended its own burn and was able to give Voyager the additional velocity it needed.
To the great relief of those in the control room, Centaur had saved the day with only 3.4
seconds of burn time remaining. Murray could only shake his head and say, Wow that was
pretty close.62 Had this flawed Titan rocket been used on Voyager 2, with its more demanding
trajectory, Centaur would not have had enough propellants to perform the corrective
maneuver. Voyager 1 missed burning up in Earths atmosphere only because of the power and
capability of the Centaur rocket. Centaur had saved half of the Grand Tour.63 This occurrence
was never widely publicized because it was the last Titan IIIE-Centaur launch.
R. P. Geye, Voyager Spacecraft, Titan/Centaur D-1T TC-7 Voyager 2 Flight Data Report, Box TC-5, TC-6, TC-7 Records,
61
62
Murray, Journey Into Space, 147.
63
Linda J. Horn, Voyager 1: Jupiter, in USA in Space, 823.
31276-chapter 5 3/9/05 9:48 AM Page 161
The Voyager spacecraft as it is encapsulated within its protective shroud at the Spacecraft Assembly and Encapsulation Facility at
Kennedy Space Center on 2 August 1977. It would later be mated with Titan-Centaur 7 for launch on 20 August 1977. (NASA
KSC_77P_0201)
31276-chapter 5 3/9/05 9:48 AM Page 162
Both Voyager craft were safely on schedule with a mission to gather the most detailed and
scientifically relevant knowledge ever assembled about Jupiter, Saturn, Uranus, Neptune, and the
outer solar system. This was the first so-called Grand Tour of the solar system, and the planets
were aligned in such a way as to make this possible only once every 176 years. With this exact
planetary alignment, the Voyager spacecraft could use planetary assists (a maneuver that uses grav-
itational attraction like a slingshot to assist in propelling spacecraft between planets) to save fuel
and time in reaching the distant destinations. Without this technique, decades would have been
added to the overall flight time.
Thirteen days into its mission, Voyager 1 turned its camera back toward home and took the
first picture of Earth and the Moon in the same frame. Eighteen months later, Voyager 1 encoun-
tered Jupiter, with its closest approach on 5 March 1979. Four months later, Voyager 2 reached
the planet. The Voyagers then employed another slingshot maneuver, using Jupiter itself to
propel them toward Saturn, altering their elliptical orbit around the Sun. This maneuver offi-
cially began the Grand Tour, with a destination beyond the solar system. By the time both
reached Saturn, Voyager 1 was nine months ahead of its sister ship.
At this point, the paths of the Voyagers diverged. Voyager 1 sacrificed future planetary
encounters for a close examination of Saturns largest moon, Titan. Voyager 2 continued its plan-
etary tour and made the first-ever encounter of Uranus on 24 January 1986. Three years later, it
visited Neptune and its moon Triton, officially ending the Grand Tour and marking the beginning
of its interstellar mission. The Voyager missions were a great success, providing the first detailed
visual and scientific studies of the Jovian and Saturnian systems, as well as providing the first close
encounters with the planets of the outer solar system.64 Thus the Titan-Centaur launch combina-
tion once again played an essential role in the planetary space program, and even when Titan
experienced technical difficulties, the power and intelligence of Centaur were able to save the
mission.
However, despite what should have been a time of great celebration for the Centaur team, the
team members faced the termination of the program that they had spent years of their lives in
perfecting. Andy Stofan, reflecting on what he considered to be the shortsightedness of NASA
planners, commented, After the launch of the last Voyager, there were no more Titan-Centaurs
64
For further information on the Voyager spacecraft, see David Morrison, Voyages to Saturn (Washington, DC: NASA SP-451
Government Printing Office, 1982); David Morrison and Jane Samz, Voyage to Jupiter (Washington, DC: NASA SP-439
Government Printing Office, 1980); Margaret Poynter and Arthur L. Lane, Voyager: Story of a Space Mission (New York:
Macmillan, 1981); Henry S. Cooper, Imaging Saturn: The Voyager Flights to Saturn (New York: Holt, Rinehart, and Winston,
1985); Joel Davis, The Interplanetary Odyssey of Voyager 2 (New York: Atheneum Publishers, 1987); Eric Burgess, Far Encounter:
The Neptune System (New York: Columbia University, 1991); Committee on Science, Space, and Technology, Voyager 2 Flyby of
Neptune (Washington, DC: Government Printing Office, 1990); Rick Gore, Voyager Views Jupiters Dazzling Realm, National
Geographic 157 (January 1980): 229; Rick Gore, Uranus: Voyager Visits a Dark Planet, National Geographic 170 (August
1986): 178195; L. Soderblom, The Galilean Moons of Jupiter, Scientific American 242 (January 1980): 88100; and William
J. Kossman, The Voyager Program, in USA in Space, 819.
31276-chapter 5 3/9/05 9:48 AM Page 163
Mating Voyager 2 to Titan-Centaur 7 on 5 August 1977. (NASA KSC_77P_0209)
ordered, and that was the largest lift capability this country had.65 Joe Nieberding lamented, All
that effort spent to integrate the Centaur to the Titan, and we launched it only seven times.66 The
concern was not just that Titan-Centaur was no longer going to fly, but that all expendable rockets
were to be phased out in favor of the Space Shuttle.
Approved in 1972 during an era of drastically reduced NASA budgets, the new reusable launch
system was presented to the Nixon administration and to Congress as a more economical launch
system than expendable rockets. Indeed, space science had never figured prominently in the wider
context of NASA funding priorities. As historian Roger Launius has pointed out, in the 1960s and
1970s, NASAs extremely productive space science missions were unobtrusively folded into the large
Apollo budgets.67 However, in the 1970s, the huge expense of projected planetary missions like
Galileo, Cassini, and Magellan garnered unwanted congressional debate and scrutiny. Of the cancel-
lation of Titan-Centaur, Launius wrote, Because the performance of the shuttle and its
65
Interview with Andrew Stofan by Mark Bowles, 13 April 2000.
66
67
Robert Kraemer, Foreword to Beyond the Moon, xvi.
31276-chapter 5 3/9/05 9:48 AM Page 164
Titan-Centaur 7 lifting off from Complex 41 at Cape Canaveral Air Force Station on 20 August 1977 with Voyager 2 as payload.
(NASA KSC_77PC_0269_01)
31276-chapter 5 3/9/05 9:48 AM Page 165
then-planned upper stage were less than that of the Titan Centaur, the cancellation decreased the size
of the payload that NASA could send to the planets.68
Although the advance of technology often leaves well-designed systems in its wake, the larger
issue was one of potential over-dependence on a single type of launching system. NASA was essen-
tially forsaking two decades of service by reliable expendable rockets for an unproven system that
promised only minor advantages (along with numerous disadvantages) over existing systems. With
astronauts to be unnecessarily involved in the future planetary program, significant safety issues
arose. From the point of view of mission specialists at JPL, the most unfortunate consequence of
the decision was the loss of Centaur, the upper stage upon which they relied to get their heaviest
payload into space. Tom Shaw, a JPL engineer, recalled that in the rush to divest ourselves of
expendable launch vehicles and go put all of our eggs in the Shuttle basket, JPL came close to
getting an official reprimand from NASA Headquarters for its vociferous objection to the phasing
out of the expendable launch vehicles upon which it had relied for the decades most important
scientific missions.69 But the decision had been made.
68
Ibid., xvii.
69
Interview with Lutha (Tom) Shaw by Virginia Dawson, 10 November 1999.
31276-chapter 5 3/9/05 9:48 AM Page 166
Centaur Reborn 167
Chapter 6
Centaur Reborn
At stake is the Shuttles viability as a launch vehicle for the massive new
communications satellites being developed for the latter half of the
decadeas well as the viability of new NASA plans to send spacecraft into
remote parts of the solar system.
M. Mitchell Waldrop, Centaur Wars, Science, 1982
The rebirth of Centaur in the early 1980s stemmed from a radical idea to redesign the
rocket to fit inside the Space Shuttle. Shuttle/Centaur was considered to be the most significant
new program at NASA in the early 1980s. Because the Shuttle was limited to flying in low-
Earth orbit, it needed an upper stage. Shuttle/Centaur promised to solve this Shuttle limitation.
Once in low-Earth orbit (621 miles above Earth), the astronauts could launch Centaur from the
Shuttle bay. Shuttle/Centaur was capable of either delivering its payload to upper-Earth orbit
or placing it on a trajectory to other planets or the far corners of the solar system. These
payloads included communications satellites, Earth-orbiting scientific spacecraft for both
NASA and the Department of Defense, and interplanetary probes.1
Shuttle/Centaur was hailed as the convergence of piloted and robotic spaceflight and the
next step in launch vehicle evolution. In an era of cost cutting and declining space budgets, the
combination of the Shuttle with the worlds most powerful upper stage promised to give the
United States a system of extraordinary power and versatility. Because of these benefits, NASA
Associate Administrator Jesse Moore believed that Shuttle/Centaur would become a very inte-
gral, longtime part of the space shuttle program.2
Despite the hopes and dreams that surrounded the new rocket combination, it proved
extraordinarily controversial throughout the period of development. NASA suspended the
program after the Challenger disaster in 1986, then officially terminated it without
1
General Dynamics report, Centaur/Shuttle Integration Study (Contract NAS 3-16786, 1 August 1973), 2, Drawer 4-C,
NASA Glenn Research Center (GRC) Archives.
2
Robert Locke quotes Jesse W. Moore in General Dynamics Shows Off New Probe-Launching Rocket, San Diego Tribune (14
August 1985).
Shuttle/Centaur ever having flown. Like the original Centaur first conceived in the late 1950s,
Shuttle/Centaur was an innovative concept that was yet to be proven. However, unlike in that
era, when technical decisions were driven by the Cold War, Shuttle/Centaur was developed
during an era with a much lower tolerance for risk.
Today, most people cite concerns over safety as the central reason that Shuttle/Centaur
never made the journey into space. But according to the core group of engineers most respon-
sible for designing and building the new Centaur, all of the safety precautions were addressed
and all necessary emergency redundancy plans were in place. Although the safety concerns were
real, something more powerful was at stake in the decision to abort such an expensive and crit-
ical program. A more accurate explanation for why NASA prevented the launch of
Shuttle/Centaur was that a devastating human tragedy forced the entire space program into an
era of heightened risk sensitivity.
Because Shuttle/Centaur promised to significantly increase its capability to launch key
national security missions for the Air Force, the risks associated with liquid hydrogen had
initially appeared acceptable. However, dtente with Soviet Union and a new Air Force deter-
mination not to depend exclusively on the Shuttle for launching its national security payloads
made the need for Shuttle/Centaur less urgent. After Challenger exploded in January 1986,
taking the lives of the entire crew, NASA decided that launching Centaur from within the
Shuttle was an unacceptable risk.
The Shuttle/Centaur saga suggests that the subjective, yet changing, tolerance for risk influ-
ences technical designs, funding, and management. Everyone involvedpoliticians, industry
contractors, and NASA personnelis willing to assume more risk when national security is at
stake. The Apollo program is an example of a grand dream with significant risk that was unques-
tioned because it was ideologically important for America to beat the Soviet Union to the
Moon. Without this incentive, it is impossible to say whether humans would have flown on top
of a stack of rockets with volatile liquid-hydrogen upper stages. By 1986, with competition
between the superpowers diminishing and the horrific images of a Shuttle exploding off the
Capes coast shown daily on the news, a decreasing risk tolerance spelled the end for
Shuttle/Centaur.
In addition to the changing tolerance for risk, the story of Shuttle/Centaur illustrates the
colliding cultures between the piloted and robotic space program. Although rivalry between
NASA Centers was not unusual, expendable launch vehicles and the piloted space program had
operated in distinctly different spheres since the beginning of the Apollo Program. For the first
time, the manned and unmanned cultures collided forcefully in the contentious relationship
between Lewis Research Center (the home of Centaur) and Johnson Space Center (the home of
astronauts and the Space Shuttle). Although Lewis Research Center and General Dynamics had
formed a collegial relationship with each other over a period of thirty years, neither had any
experience in dealing with the NASA bureaucracy that operated the Shuttle program.
Centaur Reborn 169
Shuttle/Centaur mission patch. (Glenn Research Center unprocessed photo)
Choosing Centaur for the Shuttle
Although the Shuttle promised a revolution, its technical capabilities left some concerned.
Since it could only attain a low-Earth orbit, it needed an upper stage to push some payloads to
a geosynchronous transfer orbitan extremely elliptical orbit with a perigee (closest approach
to Earth) of approximately 125 miles and an apogee (farthest distance from Earth) of about
22,300 miles. One logical choice to fill this requirement was Centaur. But how difficult would
it be to use the old robotic expendable workhorse on a new piloted space vehicle?
Although the liquid-hydrogen upper stages of Saturn V had carried the astronauts to the
Moon, Centaur had never been man-rated. In 1961, the Space Taskforce Group at Langley
had considered the possibilities of using the rocket for piloted missions but decided that
Centaur was not safe to carry astronauts.3 Ten years later, testifying before Congress after several
failures of Atlas-Centaur, NASA Deputy Administrator George Low said that without major
3
James R. Hansen, Spaceflight Revolution: NASA Langley Research Center from Sputnik to Apollo (Washington, DC: NASA SP-
4308, 1995), 283.
and costly changes, the vehicle could not be expected to achieve reliabilities much greater than
about 90 percent.4
However, General Dynamics believed that not only could Centaur satisfy the new upper
stage requirements for the Shuttle, but it also could be made safe enough to be carried in the
cargo bay. By April 1972, the company had completed a compatibility study assessing safety and
performance capabilities for using a liquid-hydrogen rocket with the Space Shuttle. Centaur
had already flown thirty-seven missions and had a flight backlog planned through 1979. The
report concluded, The Centaur can be adapted to Shuttle operation with a minimum of modi-
fication.5
In 1973, the Convair Aerospace Division of General Dynamics issued a more detailed,
eight-month Reusable Centaur Study. This study established the feasibility of modifying the
existing Centaur for upper stage launches from the Shuttle. It analyzed twelve key aspects of the
program, including such important factors as safety, cost, schedule, weight, reliability,
reusability, and flight complexity. The report concluded that on a scale of 0 (low risk) to 10
(high risk), none of these factors were rated higher than 3. The final assessment of the report
was that the Centaur programs are extremely low risk and that no technology breakthroughs
were required to achieve a reusable Centaur.6
A heated controversy developed over whether Centaur was the answer for the Shuttle upper
stage. Should NASA engineers redesign Centaur as General Dynamics proposed, or should an
entirely new upper stage be developed? An author in the journal Science reported, Arcane
though it sounds, the issue ignited a free-for-all between feuding congressional committees, the
aerospace lobby, the Reagan White House, the Air Force, and NASA, with the latter caught in
the middle.7 The way these issues were resolved would possibly shape much of Americas explo-
ration of space for the foreseeable future. These interactions included the growing market for
launching large new communication satellites, as well as NASAs capability to send new probes
further into the solar system.
In the early 1970s, NASA planned to develop what it called a space tug for lifting Shuttle
payloads into higher altitude orbits or boosting spacecraft to other planets.8 In January 1974,
4
George Low, Review of Recent Launch Failures, statement, Hearings before the Subcommittee on NASA Oversight, 1517
June 1971, 8.
5
Compatibility Study of a Cryogenic Upper Stage With Space Shuttle, 17 February 1972, Drawer H-D, NASA GRC Archives; and
Compatibility Study of a Cryogenic Upper Stage With Space Shuttle, Final Report, April 1972, NASA GRC Archives, Drawer H-
D.
6
Reusable Centaur Study, General Dynamics report, Contract NAS 8-30290, 26 September 1973, 222, 233, Drawer 4-C,
NASA GRC Archives.
7
M. Mitchell Waldrop, Centaur Wars, Science 217 (10 September 1982): 10121014.
8
David Barker, Economics of the Space Shuttle, Flight International (29 August 1974): 244246.
Centaur Reborn 171
Artists rendering of Shuttle/Centaur preparing to launch payload. The bay doors are open, and the payload is about to be released by
the Centaur Integrated Support Structure (CISS). (Glenn Research Center unprocessed photo)
NASA awarded a $1.3-million contract to Rocketdyne to design, build, and test liquid-
hydrogen and oxygen pumps and gas generators that would eventually be used in the tug.9
NASA debated the actual design of the tug and whether it would be a modification of an
existing upper stage or an entirely new rocket. Four proposals were evaluated including the Air
Forces Inertial Upper Stage (IUS), a redesigned wide-body Centaur, Transtage, and the Interim
Orbital Transfer Vehicle (IOTV).
Although all four of these stages could lift Shuttle payloads into geosynchronous orbit, each
did so with very different levels of capability. A more powerful upper stage rocket permits the
launch of heavier, more complex spacecraft that help scientists conduct more sophisticated
research. Of all the options, Centaur was the most powerful, enabling spacecraft of up to
13,000 pounds to be sent into orbit. The IOTV, like Centaur, was a cryogenic rocket and was
designed to give a similar performance. The next most powerful was the Transtage, which was
9
Towards the Space Tug, Flight International (3 January 1974): 13.
a storable propellant system capable of lifting spacecraft of up to 8,000 pounds into orbit,
followed by the IUS, which had a 5,000-pound lifting capability.
The Department of Defense (DOD) gave the following assessment of each of these rockets.
They concluded that the IUS would be able to satisfy most defense needs and, with modifica-
tions, could also take on a limited number of basic science missions. The Transtage could satisfy
all of the defense needs but did not have the potential for expanding into other types of
missions. DOD concluded, Shuttle payload limits (65,000 pounds) will limit both the IUS
and Transtage growth such that these systems can never capture a significant portion of the long
term defense needs.10
The IOTV and Centaur were both powerful cryogenic rockets that DOD believed would
not only handle all existing military needs, but could also provide tremendous potential for
more difficult science missions. The problem with the IOTV was that it was a new stage
involving a great deal of development complexity, as well as schedule and cost risks. As a result,
NASA and the Department of Defense concluded, The Centaur is the only vehicle capable of
meeting near term NASA planetary requirements . . . [and] considerable enhancement of DOD
mission capabilities.11
However, the development of the IUS continued. This upper stage eventually began to
launch smaller spacecraft, with a successful first flight on a Titan 34D booster in October 1982.
The operational IUS was 17 feet long and 9 feet in diameter, and it weighed in excess of 32,000
pounds, including 27,000 pounds of solid rocket fuel. But problems occurred on the second
flight when the IUS was spring-launched from the Shuttle on mission STS-6. Designed to carry
a Tracking and Data Relay Satellite System (TDRSS) satellite into orbit, the IUS rocket motor
nozzle accidentally changed its position by 1, causing the IUS and the satellite to tumble into
an incorrect orbit. It took two years to investigate and understand the failure to ensure that it
would not happen again.
Although Centaur development appeared promising at this point, the proponents of the
IUS mounted a vigorous campaign for their rocket. Initially, several key people from NASA
opposed the General Dynamics Centaur plan and agreed with those who sought an alternative
to the Centaur upper stage solution. They argued that the IUS, which was designed by the Air
Force, was a potentially better rocket. The first stage of the two-stage rocket was capable of
launching medium-sized payloads at most. This limitation would be overcome by means of the
addition of a second stage for larger payloads with destinations into deeper space. Specifically,
the Air Force asked NASA to develop an additional stage that could be used for planetary
missions such as a proposed probe to Jupiter called Galileo. NASA made Boeing the prime
contractor for developing the IUS.
10
Ibid.
11
Ibid.
Centaur Reborn 173
Artists rendering of the Shuttle/Centaur. The Centaur and its payload have separated from the Shuttle and are preparing to launch
after the Shuttle flies away to a safe distance. (Glenn Research Center unprocessed photo)
Success for Boeing meant not only building a technically sound rocket, but also attacking
their competitionCentaur and General Dynamics. Boeing made every attempt to demon-
strate that the IUS should replace Centaur for use on the Shuttle. Boeing argued that Centaur
(they derisively referred to it as the fat tank instead of the wide body) was risky in the area
of safety considerations. They also thought it unlikely that Centaur could be ready by the 1985
deadline and claimed that General Dynamics had significantly underestimated the risk associ-
ated with reconfiguring the craft. Boeing contended that its rocket offered a much greater
chance for overall success and argued that while Centaur required a major new development
program, the Boeing alternative was safer because it used solid propellants. The IUS minimized
risks for the crew, was mission-flexible, and was Shuttle-compatible. Boeing concluded, The
modified Centaur cannot satisfy the mission needs.12
The White House and some members of Congress seemed to agree. The Office of
Management and Budget (OMB) was against Centaur because it believed that there were too
many modifications required to make the rocket fit in the Shuttle. If these modifications were
12
Boeing Rebuttal to Shuttle/Centaur, undated, Drawer 4-C, NASA GRC Archives.
to be attempted, OMB also believed that they would be far too costly. Ronnie G. Flippo, a
Democratic Congressman from Huntsville, Alabama, agreed with the White Houses decision
and actively lobbied against Centaur. Flippo was the chairman of the House Subcommittee on
Space Science and Applications, which was the committee that authorized all NASA funds.
Flippo attacked Centaur in a number of letters to his colleagues. His argument was that Centaur
was too expensive (at least $634 million), was of limited usefulness (primarily for just two space
missions), and was an example of the faulty sole-source procurement model (with General
Dynamics being the prime contractor without competitive bids). Lurking strongly in the back-
ground of Flippos argument was local politics. His Alabama district included Marshall Space
Flight Center, which was designated as the Lead Center to manage any future NASA IUS
work.13 However, Centaur did have its supporters in Congress, most notably Bill Lowery, who
was a California Republican. But just as local politics influenced Flippo, so too did they play a
major role in Lowerys advocacy for Centaur. General Dynamics was located within his San
Diego district. Despite Flippos bias, NASA agreed with him over the protests of Lowery, and
they decided to utilize the Air Force-developed IUS.
In his Journey Into Space, Bruce Murray argues that this important subcommittee under
Flippo looked at NASA through an Alabama prism.14 Marshall was desperately looking for
ways to regain greater responsibility in the space program. During the Saturn V era in the
1960s, Huntsville, Alabama, was a prime location for all the major aerospace contractors. When
the Apollo program ended in the 1970s and the charismatic leader of Marshall Space Flight
Center, Wernher von Braun, was stricken with cancer, the Center suddenly faced difficult times.
Its rival, Johnson Space Center, was awarded the Space Shuttle program, and the Marshall engi-
neers were faced with taking direction from another NASA Center. As a result, the mood at
Marshall became more and more defensive, and through Flippo, they vigorously rejected the
Centaur-in-Shuttle idea.
Besides Flippos advocacy, NASA had other reasons for selecting the IUS as the best choice
for the Shuttle. The consensus was that Centaur was too dangerous to fly in the Shuttle along
with humans. Centaur was never conceived as a human-rated machine. Its pressure-stabilized
tank design saved weight, but it was not necessarily seen as compatible with human flight. There
were two main concerns: Centaurs lack of structure reinforcement and the inability of the
Shuttle to land safely if Centaur could not be deployed and remained in the cargo bay. Centaur
was essentially one continuous tank that had no internal reinforcement to give it added stability.
A double-walled bulkhead kept the liquid hydrogen separated from the liquid oxygen. Heat
transfer between the two cryogenic liquids was prevented by a phenomenon known as cryo-
13
M. Mitchell Waldrop, Centaur Wars, Science 217 (10 September 1982): 10121014.
14
Bruce Murray, Journey Into Space: The First Three Decades of Space Exploration (New York: W.W. Norton & Company, 1989):
213.
Centaur Reborn 175
genic pumping that produced a vacuum between the two walls of the bulkhead when the much
colder liquid hydrogen was loaded. The vacuum maintained the liquid hydrogen at -423F and
the liquid oxygen at -297F.
The human-rated Saturn Vs upper stages had a very different design. Each had two sepa-
rate propellant tanks. The entire structure was reinforced internally by a system of ribs. Like
Saturn V, the Shuttle had a sturdy, reinforced structure, and it also carried liquid hydrogen and
liquid oxygen. Critics focused on Centaurs pressure-stabilized tank design and feared that
somehow the liquid hydrogen and liquid oxygen would prematurely come into contact with
each other and explode. Reflecting on the structure, one engineer stated, You would not be
able to sit down and in some rational technical way show that it was dangerous, but intuition
and good engineering would dictate that you dont want that feature.15
The most challenging of the safety-related issues was what to do if the Shuttle mission was
aborted before Centaur was deployed. With human beings involved, the safety issues became a
serious concern. What would happen if one of the crew became sick, or the bay doors would
not open and Centaur was not deployed? The Shuttle could not land with the hazardous liquid
propellants in its cargo hold. The only option was a complicated fuel dump. This was techni-
cally feasible, but the process was complex. On Atlas-Centaur or Titan-Centaur, the worst that
might happen was that it might explode, resulting in a financial loss. Not only would it be
dangerous to try to land with Centaur still in the Shuttle, but also, even with Centaur propel-
lants dumped overboard to lower the weight and the risk, the Shuttle itself might not survive
the landing. The Shuttle is a very poor gliding machine because it loses one mile of altitude for
every three miles of distance and because its landing gear and braking systems were not designed
for the 12 to 15 tons that Centaur would add to its weight.16 These factors initially made the
IUS the upper stage of choice.
However, Boeing was unable to solve key problems with the IUS. Its main limitation was
that it was not powerful enough to launch a payload to a distant planet like Jupiter in a direct
shot. To get to its destination, the payload would have to perform mission design tricks to
accomplish a slingshot maneuver around several planets in order to gain a gravity-assisted speed
boost. Most engineers considered this approach inelegant and too time-consuming. They did
not want to wait additional years for their probes to reach their destinations. Also, new concerns
about the cost of the Inertial Upper Stage booster emerged. Robert A. Frosch, upon leaving his
post as NASA Administrator, argued that the possibilities of readying the IUS to send an orbiter
and probe to Jupiter for its scheduled launch were remote. Frosch decided on 15 January 1981
that the only alternative was Centaur. He argued that by allocating budget resources from 1981
15
16
and 1982, modifications of the Centaur could result in a powerful combination. He

concluded, No other alternative upper stage is available on a reasonable schedule or with
comparable costs. The Shuttle/Centaur would offer both to commercial customers and to
national security interests a highly capable launch vehicle with growth potential.17
Larry Ross, a Lewis Centaur engineer and project manager, agreed with this assessment and
argued that national security interests represented the key reason why the Centaur was chosen
over the IUS. Ross said that the main question was that if NASA married Centaur with the
Shuttle for the planetary program, what does the nation end up with?18 His answer was that
the nation ended up with a very significant capability. Of the two rocketsthe Shuttle/Centaur
and the IUSthe benefits of the former to the nation far outweighed the benefits of the latter.
As a result, Centaur was chosen, both because planetary missions needed the extra boost
Centaur could provide and because it better suited missions controlled by the Department of
Defense and especially the Air Force. Thus, national security interests may have raised the risk
tolerance for the more dangerous rocket.
In 1981, NASA withdrew support for the IUS and instead opted for General Dynamicss
redesigned Centaur. A number of technical factors made the Centaur option more attractive.
First and most important, Centaur was more powerful. It had the capability of delivering the
boost necessary to propel a payload directly to a planet in the deep solar system. Second,
Centaur was a gentler rocket. Solid rockets had a harsh initial thrust that had the potential to
damage their delicate payload. Liquid rockets generated their thrust more slowly, thus reducing
this threat. Finally, a rocket that used liquid propellants had one other tremendous advantage
over one that used solid fuelit could be turned off and on. Once ignited, solid fuel burned
until used up, while the liquid was much more controllable.19 Centaur had already demonstrated
the success of this capability with the various two-burn missions for Atlas-Centaur, in addition
to the multiple restarts performed in the post-mission experiments for the two Titan-Centaur
Helios missions. The only remaining advantage of the IUS over the Centaur was safety. Liquid
hydrogen was a dangerous fuel and presented a significant risk to the rocket and the astronauts
flying with it. NASA decided to take the chance and go with Centaur.
This decision was met with mixed reviews from the Air Force; on the one hand, it was
disappointed in NASAs decision to abandon the IUS in favor of Centaur. On the other hand,
the Air Force branch responsible for developing large military payloads was pleased with the
additional capabilities that Centaur provided. However, the White House Office of
Management and Budget mounted a more vocal protest and threatened to cut funding. Flippo
17
Robert Frosch, Frosch on Centaur, Lewis News (30 January 1981): 1, NASA GRC Archives.
18
19
Waldrop, Centaur Wars, 1012.
Centaur Reborn 177
continued to attack the Centaur decision vigorously, and he launched an intense fight to reverse
NASAs Shuttle/Centaur decision when Congress began its next session in February 1981.20
Eventually, NASA was forced into a compromise, relinquishing its payload role in the
International Solar Polar Mission (saving roughly $50 million) in exchange for the right to keep
Centaur and the Galileo mission to Jupiter. In August 1982, Congress officially gave its
approval to develop the new Centaur high-energy upper stage for the Space Transportation
System. Only time would tell whether this decision was the correct one to make.
While the IUS struggled in its early operational years, NASA was considering another crit-
ical decisionwhich NASA Center would take the responsibility for redesigning Centaur.
Ultimately, the responsibility for managing the Centaur redesign fell upon the shoulders of
NASA Lewis Research Center in Cleveland, Ohio.21 However, before the decision was made,
engineers at Lewis had to withstand a political play by the powerful Johnson and Kennedy
Space Centers to keep the contract out of Lewiss hands and give it to Marshall.
Internal Politics: The Push for Marshall
In the early 1980s, Lewis Research Center was a troubled organization. Over the past
decade, the laboratory had weathered the closing of its nuclear rocket propulsion program and
a devastating Reduction in Force (RIF) that had affected over 700 engineers. At the same time
that it was trying to secure a position in the mainstream of NASAthe piloted space
programit was fighting for its very survival. The Heritage Foundation, a conservative
Republican think tank, recommended cutting federal spending by eliminating what it consid-
ered unnecessary government research and technology. With Lewis Research Center on the
Reagan administrations short list of Centers to be closed, the staff mounted an active campaign
to save the Center.22
Gaining a role in Space Shuttle missions was part of the strategy to reclaim the reputation
of this once-proud laboratory. Throughout the 1970s, the engineers and research scientists at
Lewis had made important contributions to the Shuttle. They were NASAs experts in propul-
sion and energy conversion and had performed basic studies on rocket engine turbomachinery,
reduced gravity phenomena, chemical energy conversion, and flight simulation testing.23 They
hoped that their greatest challenge lay aheadredesigning the most powerful robotic rocket to
fit inside the piloted Shuttle. The difficulty lay not only in the redesign of the rocket, but also
in plans for failure contingencies, such as what would happen if Centaur had to be brought back
20
21
In 1999, this Center was renamed NASA Glenn Research Center at Lewis Field.
22
See Virginia Dawson, Engines and Innovation, 212213.
23
Martin J. Braun, Technology Transfer: Lewis Contributions to Space Shuttle, Lewis News (26 February 1982): 34, NASA
GRC Archives.
to Earth without having been deployed. For example, how could Centaurs fuel be safely
drained? Questions like these were not an issue for other rockets like Redstone, Atlas, or Titan
because they were all first stages with no option not to fire.
In 1979, John Yardley, Associate Administrator for Space Transportation Systems,
instructed Lewis Research Center to determine the feasibility of integrating Centaur with the
Shuttle. By contracting with General Dynamics/Convair, Lewis engineers planned to develop a
design that would minimize modifications to both vehicles. The main area of concern at this
time was the fulfillment of safety requirements imposed on all Shuttle cargo as a result of man-
rating.24 Lewis concluded that a cryogenic upper stage would provide increased mission
flexibility for the Shuttle and would meet the unique safety requirements. This positive finding
was very important for Lewis. The Space Shuttle threatened the expendable launch program,
including its own Centaur program. By proving the feasibility of Shuttle/Centaur, Lewis could
maintain its Centaur work and also significantly enhance its involvement in the space
program.25 Implicit in this preliminary work was the belief among Lewis engineers that if
Shuttle/Centaur were approved, they would win the management of the Centaur conversion.
Lewis Research Center encountered strong opposition from within NASA. The leadership
at Johnson Space Center (JSC), Marshall Space Flight Center (MSFC), and Kennedy Space
Center (KSC) all opposed a Lewis-led Centaur program for Shuttle, and they recommended
that Marshall should take the lead. In a dramatic display of unity and persuasion, the Directors
of these Centers (Chris C. Kraft, Jr., William R. Lucas, and Dick G. Smith, respectively) wrote
a confidential joint letter to Alan M. Lovelace, then-acting NASA Administrator, in January
1981. They marked this letter eyes only, meaning that they did not want their controversial
position to leak out to the wider NASA community. Their argument was simple and straight-
forwardthey did not want Lewis to take any leadership position with Centaur. They believed
that Marshall represented the best choice. They wrote:
The Agencys decision to proceed with the development of the Wide Body Centaur as
a planetary upper stage in lieu of the Inertial Upper Stage presents a most challenging
assignment for the Agency . . . . Recognizing this complexity, it is essential that the
Agency make the proper assignment of the management responsibilities for this
program . . . . Therefore, we believe, when consideration is given to all relevant factors,
it is in the best interest of the program and the Agency that the existing
JSC/MSFC/KSC team be responsible for the Wide Body Centaur program.26
24
Lawrence J. Ross, Launch Vehicles, Lewis News (4 January 1980): 4, NASA GRC Archives.
25
Neil Steinberg, Centaur Star Still Glows Brightly in Agency Plans, Lewis News (15 August 1980): 4, NASA GRC Archives.
26
Chris C. Kraft, Jr., William R. Lucas, and Dick G. Smith to Alan M. Lovelace, 19 January 1981, Larry Ross personal docu-
ment collection.
Centaur Reborn 179
Within this team, the directors believed that Marshall should specifically shoulder the burden
of managing Centaur for the following reasons. First, they argued that Marshall had a unique
capability for developing cryogenic rocket engines and propulsion systems. Also, because a signif-
icant technical challenge for Shuttle/Centaur was in the area of structures and dynamics, they
believed that Marshall had the most experienced team to meet these demands. Second, they
argued that the Shuttle was a complicated piloted system and that only their Centers had the
thorough understanding of the technical and managerial relationships to integrate a new
element like Centaur. As a result, the three Directors concluded, The JSC/MSFC/KSC STS team
clearly recognizes the necessity for utilizing the Centaur system as it now exists . . . and is prepared
to commit to the discipline required to implement this approach.27 They also garnered support
for Marshall from other important NASA leaders such as John Yardley, who was the head of the
Shuttle program at the time.
Despite Marshalls desire to keep its latest recommendation private, some engineers at Lewis
Research Center saw a copy of the confidential letter. Larry Ross, Director of Space Flight Systems
at Lewis, recalled, I was outraged. I never showed [the letter] to the team.28 Lewis engineers
believed that because of their experience in managing the Centaur Program, they should be the
ones to manage the development of Shuttle/Centaur.
In March 1981, Lewis Director John F. McCarthy wrote his own highly persuasive letter to
Lovelace, making the argument for his Center to win the contract. McCarthy listed many reasons
why Lewis was the right choice. First, Lewis had led the program to determine the feasibility of
modifying Centaur for the Shuttle. This experience would be invaluable in following these recom-
mendations and building the new rocket. Second, Lewis engineers had the greatest Centaur
experience among all the NASA Centers. Specifically, their experience with the successful Titan-
Centaur program provided a model on which to base the Shuttle/Centaur program. The effort to
reconfigure the Atlas-Centaur for Titan included many of the same conversion issues that would
arise with Shuttle. Third, Lewis had great experience with mission design and spacecraft integra-
tion. With Atlas-Centaur; Titan-Centaur; Earth-orbital missions; lunar missions with Surveyor;
and Earth-escape missions like Viking, Voyager, and Pioneer, this experience meant, according to
McCarthy, that no other NASA center approaches the level of expertise in this area that Lewis
has attained through the integration of such a wide variety of complex missions. In fact, the
average number of years that each Lewis engineer had on the Centaur team was an impressive thir-
teen. McCarthy concluded, Our staff uniquely possesses both the knowledge and experience and
by virtue of its stability preserves an irreplaceable corporate memory . . . I strongly believe that
Lewis continues to be the best choice for Centaur management.29
27
28
Ibid.
29
John F. McCarthy, Jr., to Alan M. Lovelace, 25 March 1981, Larry Ross personal document collection.
One month later, in May 1981, Lovelace made his decision. In a letter to William R. Lucas,
Director of Marshall, he wrote that he had thoroughly reviewed his proposal for Marshalls
control of Centaur and acknowledged the support given to him by both Johnson and Kennedy.
He wrote, Although I recognize each of the considerations which you mentioned as
contributing to this recommendation, I have concluded that they are outweighed by another set
of factors stemming from the long and continuing management of the Centaur program by the
Lewis Research Center. Primarily, these factors included the Lewis experience and their excel-
lent track record with the rocket. It was clear that Lewis was the right choice. The authors of
the history of Marshall pointed out that Lewis had built the Centaur, and had staked out a role
in advanced propulsion technology that Marshall could not expect to emulate.30 Despite the
significant competition, Lewis won the right to manage the Shuttle/Centaur program.31
Staking the Lewis Reputation on Centaur
Although ecstatic over its victory, the Centaur team knew it was staking its reputation on
the success of Shuttle/Centaur. NASA had a strong political commitment to the
Shuttle/Centaur program, and any failure at Lewis would lead to embarrassment in extremis.
Larry Ross warned that if the Center failed in this vital Agency effort, I would despair of us
ever again being relied on to do important work for NASA!32
The Center quickly began organizing contractors for the daunting task of integrating
Centaur with the Shuttle. Lewis submitted to General Dynamics a request for a proposal to
provide materials, personnel, supplies, and services to create two versions of the new wide-body
Centaurone for NASA and the other for the Air Force.33 By June, Lewis had awarded four
contracts totaling $7,483,000 to four different companies: General Dynamics to develop two
modified Centaur vehicles; Teledyne Industries for a digital computer and several remote multi-
plexor units; Honeywell, Inc., for part of the automatic navigation and guidance system; and
Pratt & Whitney for four RL10A-3-3A rocket engines. All of these contracts were geared toward
two key missionsthe upcoming Galileo probe to Jupiter and the Ulysses (formerly the
30
Andrew J. Dunar and Stephen P. Waring, Power to Explore: A History of the Marshall Space Flight Center, 19601990
(Washington, DC: NASA SP-4313, 1999), 138.
31
Alan Lovelace to William R. Lucas, 27 May 1981, Larry Ross personal document collection.
Larry Ross to Chief Counsel, 7 December 1983, Box 2142, Space Transportation Engineering Division (STED), Project
32
Management Office, Folder General Correspondence, NASA GRC Archives.
33
Technical Evaluation of GDC Proposal, Shuttle/Centaur G Development and Production, 15 December 1983, Box 2142,
Division STED Project Management Office, Folder Centaur G Program, NASA GRC Archives.
Centaur Reborn 181
International Solar Polar) mission to the Sun.34 Despite the start of work and the flow of funds,
uncertainties hovered over the project for over a year and constantly threatened to suspend the
project. Would the funding and support for Shuttle/Centaur continue? These questions could
not yet be answered.
Adding to the uncertainty was a change in leadership at Lewis Research Center. Director
Dr. John F. McCarthy, Jr., announced that he planned to retire in July 1982 and return to MIT
as a professor of aeronautics and astronautics, where he had previously headed the Center for
Space Research.35 Andrew Stofan then became director of Lewis. His experience in launch vehi-
cles would serve him well in the coming years. Stofan began his career as a research engineer at
Lewis in 1958. Four years later, he moved to the first Centaur project office, where he was part
of the Propellant Systems team. He continued to move up the launch vehicle ranks at Lewis,
heading the Titan-Centaur Project Office. He served as director of Launch Vehicles from 1974
to 1978. He was later promoted to Associate Administrator at NASA Headquarters; then he
returned to Lewis as McCarthys successor in 1982.36
Stofans arrival in his new position at Lewis coincided with the major announcement for
the Shuttle/Centaur program. President Ronald Reagan signed an Urgent Supplemental
Appropriations Bill that allocated $80 million for the design, development, and procurement of
the Centaur upper stage. This was a dramatic increase over 1981 funding for this project, which
had amounted to only $20 million. Stofan said that this new funding put Lewis directly into
the mainstream of the Shuttle program.37 In a memo to Lewis employees, he expressed his
enthusiasm over the fact that Congress had decided to continue this project. Stofan believed
that this congressional decision represented the overturning of the last potential threat to the
program and that they could at last expect to carry on from this point without the specter of
precipitous cancellation. Stofan considered the new connection with the high-profile Shuttle
program a major step forward for Lewis.38 Lewis budgets began to increase substantially for
34
Jim Kukowski, Contracts Awarded for Shuttle Launched Upper Stage, 2 June 1981, NASA Headquarters Press Release;
Linda Peterson, NASA Lewis Awards $7.5 Million in Contracts for Shuttle/Centaur Development, 2 June 1981, Lewis
Research Center (LeRC) Press Release, GRC Public Affairs Archives; Four Contracts Awarded for Modifying Centaur, Lewis
News (2 July 1981): 3, NASA GRC Public Affairs Archives.
Mary Fitzpatrick, Lewis Director to Leave NASA to Return to MIT, 2 March 1982, NASA Headquarters Press Release,
35
NASA GRC Public Affairs Archives.
36
William ODonnell, Stofan to Head Lewis Research Center, 19 March 1982, NASA Headquarters Press Release, NASA
GRC Public Affairs Archives.
37
Paul T. Bohn, NASA Lewis Gets Added Funding of $107.4 Million, 27 July 1982, LeRC Press Release, GRC Public Affairs
Archives; Center in Mainstream of Shuttle Program, Lewis News (30 July 1982): 1; Additional Funds Resurrect Important
Programs; Lewis News (13 August 1982): 1, 6, NASA GRC Archives.
Andrew J. Stofan to Lewis employees, 17 September 1982, Box 2142, Division STED Project Management Office, Folder
38
General Correspondence, NASA GRC Archives.

the project. In 1983, the Center received $133 million. In 1984, this sum increased to $159
million, and it increased again to $188 million in 1985.39
Not everyone within Lewis was as enthusiastic over the Centers new management respon-
sibilities. A large Lewis contingent believed that this work did not fall within the mission of the
Center to do basic research. Joe Nieberding recalled, To those folks, Centaur was a sink for a
lot of resources, potentially distracting management attention and NASA resources from the
real Lewis charter: to do research. The research side saw little direct involvement for them in
Centaur, and they were mostly correct.40
Increased funding for Centaur allowed for an increase in the Lewis workforce for the first
time in twenty years. In 1983, 190 new engineers were hired. Stofan said that this hiring
reflected a strong Lewis effort to reinvigorate its professional staff, and the permission to so
expand represents a vote of confidence from Washington in the Centers future.41 The main
area of activity for these new hires was in helping to transform the Centaur launch vehicle into
a Shuttle upper stage.
Lewis formed the Shuttle/Centaur Project Office to serve as the central point of integration
and management for the program. This office developed all Centaur project objectives and eval-
uated the progress of these goals. These goals included defining Centaur redesign requirements,
creating the new vehicle design, determining how best to integrate it with the Shuttle, and
finally developing and producing the new Centaur. This office also controlled the budget for
the program and was responsible for scheduling and coordinating all contractors and internal
work related to the reconfiguration.42 This project office had extensive relationships with NASA
Headquarters; Johnson Space Center; Kennedy Space Center; and outside organizations,
contractors, and other government agencies including the Department of Defense and the
European Space Agency.43
While this office relied upon the expertise of all of its workers, there were several unique
individuals who helped to shape its future. William H. Red Robbins took over as head of the
Shuttle/Centaur project office. He was a longtime NASA engineer with experience in nuclear
propulsion, and this was his first exposure to Centaur. Robbins served as project manager and
was responsible for the administrative functions including budgets and schedules. Steven V.
39
Director Reports Very Good 1984 Budget, Lewis News (11 February 1983): 1; Stofan Charts Labs Future, GRC Archives,
Lewis News (4 November 1983): 1, NASA GRC Archives.
40
Joe Nieberding, correspondence with authors, 7 November 2001.
41
Paul T. Bohn, 190 New Engineer Hires Increasing Lewis Work Force for First Time in Two Decades, 15 July 1983, Lewis
Press Release, NASA GRC Public Affairs Archives.
42
Larry Ross, Space Flight Systems, Lewis News (4 February 1983): 2, NASA GRC Archives.
43
Shuttle/Centaur Project Office, Functional Statements from the NASA Organizational Manual, 1 October 1983, Box 2143,
Division STED Project Management Office, Folder Shuttle/Centaur OrganizationLeRC, NASA GRC Archives.
Centaur Reborn 183
Szabo headed the engineering division and, in 1983, became the new chief of the Lewis Space
Transportation Engineering Division. Another longtime Centaur engineer, Szabo began work
in the original Centaur project office in 1963. In his new position, he led a one-hundred-
member team responsible for keeping Atlas-Centaur operational and for implementing the new
Shuttle/Centaur program.
As the program evolved, Steve Szabo became the critical cog in the development of
Shuttle/Centaur. As chief of the transportation and engineering division, he was directly
responsible for managing the myriad of technical activities required to integrate the Centaur
with the Shuttle. His responsibility included all propulsion, pressurization, structural, electrical
guidance and control, and telemetry hardware systems, as well as the flight software systems and
the structural dynamics, controls, trajectories, and thermal analysis activities. Within the
Shuttle/Centaur project office, Edwin Muckley headed the Mission Integration Office. He had
the complicated job of integrating each of the payloads assigned to Centaur. Frank Spurlock
managed trajectory mission design for Shuttle/Centaur.44 Joe Nieberding took charge of the
Shuttle/Centaur group within the Space Transportation Engineering Division responsible for
development and management of all software and analyses. Spurlock and Nieberding hired
many of the young engineers who later became the division chiefs, branch chiefs, and program
managers.45 With an experienced team of veterans, supplemented by young, bright scientists
and engineers, Shuttle/Centaur moved forward.
Building the New Horse
By the early 1980s, the Centaur boasted fifty-three successful payload launches into space.
Centaur was heralded as reliable. It had sent payloads to destinations like the Moon, Mercury,
Venus, Mars, Jupiter, and Saturn; it had also successfully placed many communications satel-
lites into geosynchronous orbit and several heavy telescopes into Earth orbit. Shuttle/Centaur
promised Centaur a continued and long life of planetary exploration. However, reconfiguring
the premiere upper stage rocket to launch from inside the Shuttle was not an easy task.46
Working from a limited budget and timeframe, the Shuttle/Centaur redesign team struggled to
keep changes to a minimum because each one increased the cost, decreased the reliability, and
lengthened the development period.
The first major milestone event in building the Shuttle/Centaur was a meeting held at
General Dynamics in San Diego. On 30 August 1982, NASA and all Shuttle/Centaur contrac-
44
Charles Owens, Five Space Flight Systems Engineers Promoted at NASA Lewis, 27 July 1983, LeRC Press Release, NASA
45
46
W. F. Rector III and Don Charhut, Centaur for the Shuttle Era, May 1984, Box 2142, Division STED Project Management
Office, Folder Technical Paper Approvals by PM, NASA GRC Archives.
tors met to discuss a System Requirements Review (SRR) for the entire project. Harry O.
Eastman III from the General Dynamics Shuttle/Centaur project office wrote, The purpose of
the SRR is to assure that the Shuttle/Centaur project system level requirements are adequately
stated, and to determine that the elements can indeed meet the program requirements.47 Glynn
Lunney, the manager of the Space Shuttle program at Johnson Space Center, wrote that this
meeting would be of vital importance: It is important that requirements that cannot be met,
are of insufficient detail, or are in error, be addressed at this time.48 With the completion of
this milestone, NASA could commence building the new horse.
There were similarities between the new Centaur and the old version. The propulsion
system was virtually the same. The avionics system was still mounted on the forward equipment
module, and it still included the Teledyne Digital Computer Unit. This was a 16K, 24-bit
computer that controlled guidance and navigation by integrating acceleration data to determine
position and velocity. The Centaur pressure-stabilized tank was still made of the same thin
stainless steel that had served it well for over twenty years. However, the diameter of the
hydrogen tank expanded to accommodate additional fuel for longer planetary voyages. The tank
also required insulation. The forward bulkhead used a two-layer foam blanket and a three-layer
radiation shield.
Despite the similarity of Shuttle/Centaur to Atlas-Centaur, there were some significant
technical changes. Specifically, there were two main redesigns required for the Shuttle to carry
the Centaur. First, Centaur and its payload had to fit into the 60-foot cargo bay. Engineers
developed two different configurations (called versions G and G-prime). The Centaur G was 20
feet long and was capable of supporting a 40-foot payload to accommodate Department of
Defense requirements. By reducing propellant weight to 29,000 pounds, the G version could
carry additional weight. It was designed specifically for missions to place satellites into geosta-
tionary orbit. The United States Air Force funded half of the initial $269-million expense
required to design and develop the Centaur G for use with the Shuttle.49 The other configura-
tion, Centaur G-prime, was nearly 30 feet long and capable of handling payloads of up to 30
feet.50 It was optimized for planetary missions, and the payload space was smaller due to the
47
Harry O. Eastman III, 30 August 1982, unprocessed Shuttle/Centaur records, NASA GRC Records.
48
Glynn S. Lunney, 12 August 1982, unprocessed Shuttle/Centaur records, NASA GRC Records.
49
NASA/DOD Memorandum of Agreement, November 1982, Box 2141, NASA GRC Records. After the cancellation of
Shuttle/Centaur, Centaur G became the upper stage for the Air Force Titan program.
50
Jeweline H. Richardson, The Centaur G-Prime: Meeting Mission Needs Today for Tomorrows Space Environment, May
1983, Box 2142, Division STED Project Management Office, Folder Technical Paper Approvals by PM. See also NASA Lewis
Research Center/USAF Space Division Agreement for the Management of the Shuttle/Centaur Program 10/25/82, Box 2141,
NASA GRC Archives.
Centaur Reborn 185
necessary increase in propellant weight (45,000 lb). NASA funded the G-prime configuration.51
Both versions would fill the entire cargo bay, making it necessary to devote the entire Shuttle
mission to launching them. Overall, the two versions were very similar, with 80 percent of the
design detail being common.52 The second and most important new technical development was
the mechanism needed to support Centaur in the Shuttle and deploy it from the bay.53 The new
deployment device was called the Centaur Integrated Support Structure (CISS), a 15-foot-
diameter aluminum structure that attached the rocket to the Shuttle. This device was of central
importance because it enabled the Centaur to fly in the Shuttle with a limited number of design
modifications. The CISS was located within the cargo bay on rotating support structures. It was
responsible for all of the mechanical, electrical, and fluid interfaces between Centaur and the
Shuttle, as well as most of the safety precautions for the rocket.54 The CISS was also fully
reusable for ten flights.55 The development of the CISS was considered one of the more exten-
sive and challenging tasks of the entire Centaur modification project.56
Along with these two main technical developments, there were hundreds, if not thousands,
of smaller integration issues that engineers had to address. A small sample of these tasks
included internal thermal integration of the three machines (Shuttle, Centaur, and payload)
during launch, deployment, and landing; avionics integration interfaces to monitor tank pres-
sure, power, and computer software between the Shuttle and Centaur; and propulsion
integration, which detailed how the propellants would be loaded and also dumped in case of an
emergency abort. Countless personnel and general support issues also complicated the list of
technical needs. A business-management system was required to control accounting and
contract issues and a project office was set up to manage the entire integration process.57
Estimates from Johnson Space Center indicated that, in total, there were between 500 and
1,500 separate items up for review in the integration process.58
W. F. Rector III and David Charhut, Centaur for the Shuttle Era, May 1984, Box 2142, Division STED Project
51
Management Office, Folder Technical Paper Approvals by PM, NASA GRC Records.
52
Technical Evaluation of GDC Proposal, Shuttle/Centaur G Development and Production, 15 December 1983, Box 2142,
Division STED Project Management Office, Folder Centaur G Program, NASA GRC Records.
53
Harold Hahn (General Dynamics), A New Addition to the Space transportation System, July 1985, Box 2142, Division
STED Project Management Office, Folder Technical Paper Approvals by PM, NASA GRC Records.
54
Omer F. Spurlock (aerospace engineer, Lewis Research Center), Shuttle/CentaurMore Capability for the 1980s, Box 2142,
Division STED Project Management Office, Folder Technical Paper Approvals by PM, NASA GRC Records. This article is the
best technical overview of the entire Shuttle/Centaur.
55
R. Wood and W. Tang, Shuttle/Centaur Prestressed Composite Spherical Gas Storage Tank, July 1985, Box 2142, Division
56
Larry Ross, Space, Lewis News (31 December 1981): 3, NASA GRC Archives.
57
Centaur Integration Task Summary, 16 September 1983, unprocessed Shuttle/Centaur records, NASA GRC Records.
58
Harry O. Eastman, 17 September 1981, Glenn Research Center, unprocessed Shuttle/Centaur records, NASA GRC Records.
Once all of the integration procedures were worked out, the Shuttle/Centaur would be
ready for its first flight. The flight operation and the interactions between the Shuttle, Centaur,
and CISS worked in the following way. The Shuttle launch was itself a standard flight. During
the launch, engineers at Houston planned to monitor the Centaur through a Tracking and Data
Relay Satellite (TDRS) via telemetry links to the ground. Once the Shuttle began its orbit and
a predeployment check, the bay doors opened. The rocket was then rotated 45 out of the
orbiter bay by the CISS to a launch-ready position twenty minutes before it was to be released.
Special springs on the CISS (twelve compressed coil springs called a Super*Zip separation ring)
enabled the Shuttle crew to eject the Centaur from the deployment adapter into free flight at 1
foot per second.59 The Centaur then coasted for 45 minutes prior to its main burn, allowing the
astronauts to take the Shuttle a safe distance away. For planetary missions, only a single burn
was necessary. Once at the required velocity, the final maneuver was to separate the spacecraft
from Centaur. After release, Centaur was capable of maneuvering so that it would not interfere
with the flight of the spacecraft and would also prevent planetary impact.
Safety was always a key concern with Shuttle/Centaur. Howard Bonesteel, Manufacturing
Director of the Space Systems Division at General Dynamics, remarked that the big difference
is the larger fuel tank for the liquid-hydrogen fuel, and a lot of additional safety systems because
its being carried on a manned spacecraft.60 One of the major functions of the CISS was to
address potential problems with the fluids and avionics systems. Increased safety meant an
ability to tolerate multiple failures before some unwanted consequences take place.61 Special
redundancy features were installed to comply with stringent safety considerations. The engi-
neers believed that these would significantly reduce the dangers of losing liquid-hydrogen fuel.
To minimize the risk, a unique propellant fill, drain, and venting system was installed on two
of the Shuttles designated to carry Centaur.62 This system enabled the Shuttle crew to dump the
dangerous propellants in case of an emergency, allowing the Shuttle, Centaur, and its payload
to return to the ground safely. The benefits of Shuttle/Centaur seemed to outweigh the risks.
Larger spacecraft could take longer voyages into the solar system, and much heavier
Department of Defense satellites would be able to reach geosynchronous orbits.63 In short,
59
R. E. Martin (from General Dynamics Space Systems Division), Effects of transient Propellant Dynamics on Deployment of
Large Liquid Stages in Zero-Gravity with Application to Shuttle/Centaur, delivered at the 37th International Astronautical
Congress, October 1986, Innsbruck, Austria, Box 2142, Division STED Project Management Office, Folder Technical Paper
Approvals by PM, NASA GRC Archives. (The separation system was built by Lockheed.)
60
Robert Locke, quoted in General Dynamics Shows Off New Probe-Launching Rocket, San Diego Tribune, 14 August 1985.
61
Centaur for STS, 8 January 1986, Box 2142, Division STED Project Management Office, Folder General Correspondence,
62
NASA GRC Records.
63
AP Press release, General Dynamics, 14 August 1985, Box 2141, NASA GRC Archives.
Centaur Reborn 187
NASA was aware of the risks from the very beginning but decided that the risk was worth the
scientific and national defense rewards.
However, one important question always loomed over the enterprise: was NASA placing
too much of its hope for success on the Shuttle and the ideal of reusability? Launch systems like
Titan-Centaur were taken out of production because they were expendable vehicles. In the new
era of reusability, these old rockets did not satisfy the new vision of spaceflight. However, if the
Shuttles development was halted before completion or the Shuttle was unsuccessful, how would
other payloads, such as robotic planetary missions and satellites, make the voyage into space? In
1981, one reporter for the New York Times speculated that a Shuttle problem would confirm the
worst fears of those who had criticized the agency for its failure to allow for backup systems to
be kept in production while awaiting the shuttle.64 Little did this reporter realize how prophetic
his words were. In five years, before the near launch of the new Centaur, a tragic Shuttle failure
left NASA without a launch vehicle.
64
John Noble Wilford, U.S. to Discontinue New Rocket System, New York Times (18 January 1981).
chapter 7 3/9/05 10:52 AM Page 189
Eclipsed by Tragedy 189
Chapter 7
Eclipsed by Tragedy
It was considered to be probably one of the most hazardous projects that
NASA had ever attempted to fly . . . . The chief of the Shuttle office at the
time was John Young and he called Shuttle/Centaur Death Star.
Rick Hauck, NASA Astronaut
Im convinced to this day we would have made the launch window in

May of 86, but it was a sprint to the finish. It was like the racehorse that
overtakes you at the end. Had not the Shuttle accident occurred, we would
have been ready for launch.
Marty Winkler, General Dynamics
With the infrastructure in place and the redesigns underway, Shuttle/Centaur was ready
to take on a real mission. In the early 1980s, it was designated to launch two of the most signif-
icant explorations of the solar systemone probe to Jupiter and the other to the Sun. But in
preparing for these payloads, significant and tragic problems arose. The first was in managing
Centaur. Project managers and engineers had to overcome significant time constraints because
both of these missions were scheduled to launch in May 1986. This complexity was
compounded by the problems of making a rocket safe enough to ride with humans and
convincing the NASA community that the risks were minimized. The second obstacle was one
from which, ultimately, the Shuttle/Centaur could not recoverthe tragedy of the Challenger
explosion in January 1986. In the wake of this disaster, the liquid-hydrogen propellant that
coursed through Centaurs veins appeared too risky, and ultimately the Shuttle/Centaur was
grounded, eclipsed by a tragedy not of its own making.
The Shuttle tragedy resurrected questions concerning the decision to end the expendable
launch program. The ill-fated Challenger was scheduled to deliver a Tracking and Data Relay
Satellite (TDRS) to a geostationary orbit. Journalist and historian William Burrows pointed
out how tragically unnecessary it was for seven people to lose their lives trying to get a satel-
lite into orbit that could have been sent there on an expendable.1 Centaur, with either an Atlas
1
William E. Burrows, This New Ocean: The Story of the First Space Age (New York: Random House, 1998), 556.
chapter 7 3/9/05 10:52 AM Page 190
or Titan booster, could have delivered this payload without the risk to human life. Both the
risks NASA accepted and those it rejected shaped the future of the Centaur program more than
the actual technical success and capability of the rocket itself.
To the Sun and Jupiter
The destinations for the first two Shuttle/Centaur launches were both extremely impor-
tant in terms of the scientific knowledge they would uncover about our solar system. The Sun
and Jupiter, icons of mystery in the sky for as long as humanity has turned its gaze to the
heavens, would relinquish some of their secrets to the probes to be delivered by Centaur.
Coupled with this importance was a significantly high level of mission complexity for the first
Shuttle/Centaur launch system. With only a six-day launch window between the two launches,
beginning 16 May 1986, there was no margin for error.2
In the 1980s, project Galileo was among NASAs few planetary missions in production.3
Named for the seventeenth-century Italian scientist who discovered Jupiters moons, the
project had its genesis at JPL in the late 1970s. Its objective was to send two spacecraft to
explore the Jovian planetary system. The Galileo team was frustrated by the difficulty in
securing a ride for their probe. They went back and forth, redesigning their probe for the IUS,
then for Centaur. As late as 1982, this decision changed again when the NASA budget
included funding for Galileo but not for Centaur. The IUS again emerged as the launch
vehicle.4 When Centaur was returned to funding status, NASA again abandoned the IUS in
favor of a faster ride for the Galileo probe. Initially, the Galileo team was pleased with this
decision, although they never suspected that in just four years they would again have to search
for an alternate ride.
Equipped with an orbiter and a probe, the Galileo mission would provide unprecedented
observations of the fifth planet as well as the four Galilean satellites. It was also scheduled to
perform a flyby of the asteroid 29 Amphitrite.5 This was a large, 200-kilometer main-belt
2
John R. Casani to William H. Robbins, 9 April 1985, Box 2143, Division STED Project Management Office, Folder Galileo
Correspondence, NASA GRC Archives.
3
For further information on the Galileo probe, see C. M. Yeates, ed., et al., Galileo: Exploration of Jupiters System (Washington,
DC: NASA Government Printing Office, 1985); David Leonard, Free Fall to Jupiter, New Scientist 147 (July 1995): 2629;
C. T. Russell, ed., The Galileo Mission (Boston: Kluwer Academic, 1992); Monish R. Chatterjee, Galileo: Jupiter, USA in Space,
eds. Frank N. Magill and Russell R. Tobias, vol. 3 (Pasadena, CA: Salem Press, Inc., 1996), 191194.
4
John R. Casani to Galileo Review Board, 6 January 1982, Box 2143, Division STED Project Management Office, Folder
Galileo Correspondence, NASA GRC Archives.
5
M. K. Winkler to W. H. Robbins, 23 January 1985, Box 2143, Division STED Project Management Office, Folder Galileo
Correspondence, NASA GRC Archives.
chapter 7 3/9/05 10:52 AM Page 191
asteroid that was the most interesting of several encounter possibilities for Galileo. However,
the asteroid flyby was only considered an optional scientific objective. After launch, engineers
would assess the health of the spacecraft to determine whether this option would be exercised.
This mission would not only gather scientific knowledge, but would also significantly
enhance U.S. scientific prestige.6 Once Galileo reached its final destination, the orbiter would
become the sixth spacecraft to visit the planet and the first to be placed into orbit around
Jupiter.
While orbiting, it would gather scientific data concerning Jupiters atmosphere and the
physical attributes of the satellites. The probe also promised to be the first device to enter the
atmosphere of a planet more distant than Mars.7 Six months after release from the orbiter, it
would descend toward Jupiter and record information on the chemical composition of the
atmosphere, the structure and physical dynamics of the magnetosphere, and the chemical
composition of the Jovian satellites.8 Scientists hoped that this knowledge would provide
greater understanding of the evolution of the solar system, the origin of life, the weather
system on Earth, and additional information about Jupiter itself.9
In November 1982, Lewis Research Center signed a Memorandum of Agreement with JPL
for the Galileo project. Lewis maintained the Shuttle/Centaur project office; JPL held the
Galileo project office. The two Centers maintained close contact to ensure that the new
Centaur could meet all the specifications necessary for a successful flight to Jupiter. JPL was
responsible for managing the entire mission, extending from design concept to mission
completion. To Lewis fell all responsibilities necessary to integrate the Galileo spacecraft with
Centaur and the Space Transportation System.10 This work included coordinating with
Johnson Space Center (JSC) and ensuring that there was engineering compatibility and risk
management between the Shuttle and Galileo.
The second Shuttle/Centaur mission, also scheduled to launch in May 1986, was a joint
venture with the European Space Agency (ESA). Originally called the International Solar Polar
6
James M. Beggs to Jamie L. Whitten, 24 December 1984, Box 2143, Division STED Project Management Office, Folder
Galileo Correspondence, NASA GRC Archives; Jim Kukowski, NASA May Fly by Asteroid with Galileo Spacecraft, NASA
Headquarters Press Release, 27 December 1984, NASA GRC Public Affairs Archives.
7
Charles Redmond, Peter Waller, Test of Jupiter Probe Spacecraft a Success, NASA Headquarters Press Release, 11 August
1983, GRC Public Affairs Archives.
8
Payload Integration Plan Space Transportation System and Galileo Mission, 24 August 1983, Box 2143, Division STED
Project Management Office, Folder Galileo Correspondence, NASA GRC Archives.
9
Environmental Impact Statement for Project Galileo, May 1985, Drawer H-D, NASA GRC Archives.
10
Andrew Stofan and Lew Allen, Memorandum of Agreement Between the Lewis Research Center and the Jet Propulsion
Laboratory for the Galileo Project, November 1982, Box 2142, Division STED Project Management Office, Folder Galileo,
NASA GRC Archives.
chapter 7 3/9/05 10:52 AM Page 192
Mission, the destination for this spacecraft was the Sun and the exploration of the solar envi-
ronment. It was renamed Ulysses in 1984 to allude both to Homers hero and to Dantes desire
to explore an uninhabited world behind the Sun.11 Ironically, in journeying to the Sun, the
probe would first travel to Jupiter (the same destination as Galileo) and then use the gravita-
tional mass of the planet to catapult it out of the ecliptic plane (the plane in which all planets
orbit the Sun). It was necessary to fly out of this plane, with the associated high energy
supplied by the Jupiter gravity-assist, to enable passage over the solar poles. Ulysses was not
designed for a close solar approach. In fact, engineers liked to joke that the closest Ulysses ever
got to the Sun was when it was sitting on the launch pad in Florida. Nevertheless, this mission
would mark the first observations of the solar poles of the Sun. Scientists hoped to uncover
knowledge about the solar wind, the heliospheres magnetic field, the interplanetary magnetic
field, cosmic rays, and cosmic dust.12
Preliminary plans were also being made for Shuttle/Centaur to launch a third mission, the
Venus Radar Mapper, whose name was later changed to Magellan.13 Lewis Research Center hosted
the initial mission integration panel meeting on 8 November 1983.14 For this mission, scheduled
for launch in April 1988, a variety of different upper stage vehicles were examined as launch alter-
natives. These included the Orbital Sciences Corporation TOS/AMS, Astrotech Internationals
Delta Transfer Stage, and Boeings IUS.15 NASA assessed the viability of all these alternatives and
considered trajectory, weights, schedule, and risk factors. The evaluators decided that
Shuttle/Centaur was the best alternative and sent notice to the competitors that NASA has
concluded that your system is not capable of meeting the VRM [Magellan] mission requirements
without imposing unacceptable technical and schedule risks on the program.16 Once again, the
11
For further information on the Ulysses probe, see Craig Covault, European Ulysses Fired to Jupiter, Sun as Discovery Returns
to Space, Aviation Week and Space Technology 133 (15 October 1990): 22; Michael Mecham, After Long Delay, Ulysses Mission
Begins 5-Year Voyage to Expand Solar Data Base, Aviation Week and Space Technology (22 October 1990): 111; David H.
Hathaway, Journey to the Heart of the Sun, Astronomy 23 (1995): 38; Gordon A. Parker, Ulysses: Solar-Polar Mission, in
USA in Space, 785788.
12
Jim Kukowski, Ulysses New Name for International Solar Polar Mission, NASA Headquarters Press Release, 10 September
1984, NASA GRC Public Affairs Archives.
Magellan Mission Design Panel Charter, 7 February 1986, Box 2143, Division STED Project Management Office, Folder
13
VRM Mission, NASA GRC Archives.
Edwin T. Muckley, Action Items from the First Venus Radar Mapper Shuttle/Centaur Integration Meetings, 8 November
14
1983, NASA Archives, Box 2143, Division STED Project Management Office, Folder VRM Mission.
15
Jesse W. Moore to David W. Thompson (Orbital Sciences Corporation CEO), 18 June 1984; Moore to Leonard Rabb
(Astrotech International Corporation Director), 22 June 1984; Moore to H. N. Stuverude (Boeing Aerospace Company Vice
President), 18 July 1984, Box 2143, Division STED Project Management Office, Folder VRM Mission, NASA Archives.
Jesse W. Moore to Robert J. Goss (Astrotech Space Corporation), 15 August 1984, Box 2143, Division STED Project
16
Management Office, Folder VRM Mission, NASA GRC Records.

chapter 7 3/9/05 10:52 AM Page 193
technical merits of Shuttle/Centaur had beaten out competition and won the right to launch
another prestigious planetary mission.
The Department of Defense (DOD) also began scheduling missions for Shuttle/Centaur.17
In 1984, it devised a launch plan for Shuttle/Centaur to put the Milstar satellite into orbit.18
This military communication satellite was designed to operate under the worst conditions,
including jamming, interception, and nuclear attack.19 Plans to use Shuttle/Centaur for this
launch were complex. It proved difficult because the DOD required Centaur to perform
beyond its designed capability. Meyer Reshotko wrote, This mission has some important
requirements which exceed the Centaur G baseline specifications. Each of these requirements
has been requested by the Air Force.20 For example, the Milstar required a direct attachment
to Centaur with explosive bolts used to separate the two. Reshotko said that in order for this
to occur, the Centaur team had not only to redesign the satellite interface, but also to begin
testing to determine what effects the separation shock would have. Further requirements also
included a broader Payload Integration Plan for all DOD payloads destined for launch on
Shuttle/Centaur. These were all classified missions; the underlying security clause in each
signed agreement between the Air Force and NASA stated, There will be no public release of
information pertaining to payload.21 This classification made the job of integration far more
difficult. For example, if engineers at Lewis wanted to make a telephone call to General
Dynamics about the project, they had to go to a safe building with secure communications
lines.
But again, some asked the question, Why not send the deep-space planetary missions
with an expendable launch vehicle such as Titan-Centaur? This launch vehicle had proven
that it was capable of delivering very heavy scientific missions into the deep solar system. The
prowess of Atlas-Centaur as a launch vehicle for communications satellites was unquestioned.
In addition to disdain for all things expendable in a new culture that valued reusability,
another factor weighed against Titan-Centaurits military connection. Marty Winkler said,
They [NASA] decided on Shuttle/Centaur which arguably may have been the wrong deci-
sion, but that was the decision they took. The reason I believe they did not want to go with
17
Shuttle/Centaur Memorandum of Agreement, approved by Colonel Charles H. MacNevin and William H. Robbins, 17 May
1984, NASA GRC Records.
18
Launch Base Test Plan for Milstar Processing at ELS, NASA GRC Records, Box S/C Mission Integration, Division LVPO.
19
Donald H. Martin, Communication Satellites, 4th edition (El Segundo, California: The Aerospace Press, 2000), 208.
20
Meyer Reshotko to Steve Szabo, Summary of Out of Scope Mission Unique Requirements, Box S/C Mission Integration,
Division LVPO, NASA GRC Records.
21
Shuttle/Payload Standard Integration Plan for DOD Deployable/Retrievable-Type Payloads, May 1985, Box S/C Mission
Integration, Division LVPO, NASA GRC Records.
chapter 7 3/9/05 10:52 AM Page 194
the Titan is that the Air Force owned the Titan, and [NASA] would have to go through the
Air Force to do that. Any relationship between the Air Force and NASA, historically, on launch
vehicle matters, had been like oil and water.22 NASA would still work closely with the Air
Force on their military communication satellites, but with Shuttle/Centaur, NASA was at least
able to limit Air Force involvement to the Centaur G configuration. Whatever the merits of
the decision, the Shuttle/Centaur was going into development. The responsibility fell to Lewis
Research Center to pull off what many still considered impossible.
Managing Centaur
With the mission objectives established, the main task became managing Centaur so that
it could become launch-ready for May 1986. Significant funding for the project did not come
until 1983, and the first planetary launches were scheduled for mid-1986. Larry Ross said, It
was a very substantial development program. When we sold it, I dont think we emphasized
how much substance there was to the development. I think we said, its a Centaur, we have
them all over the place, we will just make one that will fit in the cargo bay. It was a relatively
straightforward technical job, but managerially it was extraordinarily demanding.23
The management relations among the various institutions involved in the project were
complex. Although the responsibility for Centaur G was jointly shared between NASA and the
Air Force, the project management lead fell upon the shoulders of Lewis Research Center,
while the Department of Defense and NASA Headquarters provided funding and overall
direction for Shuttle/Centaur. The Air Force assigned Major (later Colonel) William Files as
the Air Force Deputy Shuttle/Centaur Project Manager and assigned a team of half a dozen
other key individuals to become part of the project office in Cleveland. Important interaction
and communication were also maintained with other NASA Centers, including Johnson Space
Center for the Space Shuttle and Kennedy Space Center for the cargo project office. Mission
management included the Jet Propulsion Laboratory for the Galileo project, the Air Force for
DOD projects like Milstar, and the European Space Agency for the Ulysses project. The final
area of interaction for Lewis was with their contractors. These included General Dynamics for
the Centaur vehicle and the CISS, Teledyne for the computer, Honeywell for the guidance
system, and Pratt & Whitney for the engines. William H. Robbins, manager of the
Shuttle/Centaur Project Office, and Vernon Weyers, deputy manager of the Lewis
Shuttle/Centaur Project Office, called this complex management arrangement a challenge
that was a dynamic and demanding assignment for all involved personnel.24
22
Interview with Marty Winkler by Virginia Dawson, 21 March 2001.
23
William H. Robbins and Vernon Weyers, Shuttle/Centaur Project Stages Challenge, Lewis News (22 March 1985): 1, GRC
24
Archives.
chapter 7 3/9/05 10:52 AM Page 195
Lewis project leaders sought out exceptionally qualified individuals and gave them the
support necessary for achieving their goals. To bring together all these individuals from various
locations and backgrounds required the development of a team environment. It was hoped
that this sense of teamwork would unify the project members and direct their focus upon one
key goalintegrating Centaur with the Shuttle. This goal was extremely challenging due to
the inflexible time schedule and the number of firsts that needed to be accomplished. Ross
argued, The bottom line is that we have a very tough job to do in a very short time and we
must succeed.25
To help cultivate a team atmosphere, Ross dug deep into what he called his senior
managers bag of tricks. What emerged were trusted managerial and motivation techniques to
concentrate team attention on May 1986, a date three years in the future. Ross said that time
was the single most important aspect of the motivating theme.26 The schedule was critical for
Shuttle/Centaur because the Jupiter launch had only a twenty-one-day launch window. If this
launch window were missed, the entire mission would have to be delayed over a year.27 The
cost resulting from such a delay was estimated at about $50 million.
Ross relied on a motivational symbol to convey the urgency, focus, and sense of teamwork
required. He wrote, My campaign centers on a graphic which is specially designed to visually
portray the challenge and evoke an active sense of commitment to it.28 The image itself was
an ancient centaur, half-man, half-horse, emerging from the Shuttle, rotating backwards and
aiming his arrow into the sky. Ross insisted that the symbol become ubiquitous throughout
the various laboratories and offices so that all who worked on the project were visually aware
of their goal. Even contractors from industry agreed to display the symbol prominently in their
work environments. Ross confessed, If I am to succeed in this, all team members from our
own [Lewis] engineers to mechanics in the contractors shop will not be able to go through a
workday without seeing the symbol and being reminded, and inspired, by the challenge and
commitment it represents.29
Ross also emblazoned the symbol across a variety of project memorabilia. Campaign-type
buttons were printed, along with drink coasters and notepads. To reinforce the importance of
meeting all deadlines, the team members received a special Centaur pocket calendar. Unlike
most twelve-month calendars, this one covered an unusual twenty-eight-month duration. It
25
Larry Ross correspondence, Box 2142, Division STED Project Management Office, NASA Archives.
26
Ibid.
Vernon J. Weyers to Kevin L. Wohlevar, Cray Research, Inc., 16 August 1985, Box 2141, Division STED Project Management
27
Office, Folder General Correspondence, NASA Archives.
28
Ibid.
29
Ibid.
chapter 7 3/9/05 10:52 AM Page 196
represented the months from January 1984 to May 1986, ending with the acceptable window
of the Galileo launch. William H. Robbins printed his personal reminder on the first page. He
wrote, It is not possible to overstate the importance of fulfilling our commitment to this chal-
lenge . . . . This special calendar is offered as a way of keeping the schedule critical nature of
what you do prominent in your daily planning.30 For the front cover, he combined Rosss
symbol with the trite but effective catchphrase from the popular movie Rocky: Go for it!
Optimism remained strong in 1985 for both the Galileo and Ulysses missions. Fabrication
for the G-prime version was well underway, and the structural testing was being completed at
the Shuttle/Centaur test facility in Sycamore Canyon, California. Slightly behind schedule
were the avionics box and fluid system testing, but this delay did not threaten a major post-
ponement of the program.31 However, despite this technical and managerial success, the safety
concerns would not go away. While it was thought that the safety systems would remove all
doubt about the dangers of using the highly explosive fuel, it was exactly this problem that
would start a bitter inter-Center conflict at NASA and cast the future of Shuttle/Centaur into
doubt.
Centers in Conflict
One of the most significant factors slowing the development of Centaur was the number
of interfaces that Lewis handled on a daily basis. The interfaces for Shuttle/Centaur included
mission management, the Air Force, customers, systems management at NASA Headquarters,
element and cargo management at other NASA Centers, and finally the industrial contractors.
Engineers knew that the number of interfaces they encountered during a job increased the
overall complexity by geometric proportions. This complexity was further compounded
because most of these interfaces were authoritative, meaning that they had the power to
influence or to stop the progress of the overall mission. Marty Winkler, who took over as head
of the Shuttle/Centaur program at General Dynamics, recalled how difficult this program was.
He said, In the beginning of 1983 I took over the Shuttle/Centaur program, and spent an
extraordinary long and painful three yearsvery, very pressure filled. There were 70-hour
work weeks for three years. I was not a pleasant person to be around. And a lot of it was caused
by the internecine warfare [between] the NASA Centers.32 Specifically, Winkler said, The
people at Johnson Space Center hated the idea. They fought it all the way. It was miserable.
Of all the relationships, Lewis engineers believed that the most difficult and dominant
one was with Johnson Space Center. As Larry Ross lamented, Johnson never wanted this
30
Ibid.
31
Harold Robbins and Vernon Weyers, Shuttle/Centaur Project Stages Challenge, Lewis News (22 March 1985): 1, GRC
Archives.
32
chapter 7 3/9/05 10:52 AM Page 197
program.33 Red Robbins agreed, There was an attitude problem at Johnson. They didnt want
to fly Centaur. They didnt want a big hydrogen tank sitting behind the astronauts because
they thought the risk was too great.34 Many of the astronauts who were based at Johnson
Space Center felt the same way. General Dynamics aerospace engineer Edward Bock said, The
astronauts never became comfortable with the vehicle . . . . There were astronauts that just flat
out said, I will not fly the Shuttle with that thing in the payload bay.35 Tom Shaw, who
managed launch vehicle integration at JPL, also commented on the reluctance of astronauts to
fly with Centaur. He said, The people that sit up there in the cockpit looked back over their
shoulder and listened to the gurgling and the rumbling of the vent valve opening and closing,
and had some reservations. The astronaut office did not feel very kindly towards the use of that
big liquid-hydrogen/liquid-oxygen stage in the Shuttle bay.36 But as was often pointed out,
the Shuttle itself was powered with liquid hydrogen and oxygen. Shaw noted that the astro-
nauts were sitting astride twenty-five Centaurs worth of liquid oxygen and liquid hydrogen
in that external tank. So its just a question of whether its between their legs or behind their
back is the way I view it.37
Why was the Johnson-Lewis relationship strained? The crux of the problem was the
understandably cautious attitude Johnson had toward its astronauts. As a Center, Johnson
essentially owned the human space program. They were the ones who took the awesome
responsibility of risking human life every time they sent a mission into space. Because of this
responsibility, crew safety was the first priority. When they faced the need to put a dangerous
rocket into their Shuttle, they became highly cautious and conservative. Furthermore, when
the decision was made to allow Lewis Research Center to manage the building of this new
rocket, and not Marshall as Johnson had recommended, their reluctance increased further. It
was Marshall that had always been a member of the manned spaceflight team. Lewis was
always on the outside. And when the decision was made to put Lewis squarely in the middle
of what was the most significant new space program in the early 1980s, the manned space
establishment did not welcome them with open arms.
Johnson Space Center and Lewis Research Center attempted to work together as a team
to make Shuttle/Centaur flight-ready, and there were some signs that the earlier disagreements
would be put behind them. There was some positive interaction between the two Centers. As
the project got underway, Andrew Stofan wrote, cooperation between both centers continues
33
34
Interview with William H. Robbins by Mark Bowles, 21 March 2000.
35
Interview with Edward Bock by Virginia Dawson, 22 March 2001.
36
37
Ibid.
chapter 7 3/9/05 10:52 AM Page 198
to be excellent.38 However, under the strain of the short timeframe and high technical
complexity of the Shuttle/Centaur program, this relationship was to be severely tested.
Specifically, the Centers came into conflict over what was, predictably, the most contentious
issuesafety. The dangers inherent in allowing a human crew to carry a liquid-hydrogen
rocket into space were significant. Although the safety of the crew and Shuttle was the
number-one priority for both Centers, there was often disagreement over the tradeoffs between
safety and mission success. Each Center had a different perception of risk.
The first question was how to integrate Centaur with the Shuttle. Would it be considered
an element or a payload? Elements were large, conceptual chunks of a launch vehicle, like
the Solid Rocket Boosters, the External Tank, and the orbiter itself. They were the responsi-
bility of Johnson Space Center, where all the safety, reliability, and quality-assurance
requirements were developed and managed. A 1981 Memorandum of Agreement between
Johnson and Lewis identified Centaur as a Level III Space Shuttle element.39
Initially, Lewis engineers and managers believed that it was in their best interest to reclas-
sify the Centaur as a payload because there was less bureaucracy and documentation associated
with this status. With Centaur classified as an element, Johnson would have had more direct
control of the project and the budget, and Lewis engineers would have had to take direct
orders from them. Classifying Centaur as a payload allowed Lewis to remain more
autonomous; the payload handbook guided the specifications. During the monthly safety
reviews, Johnsons payload people merely determined if the Lewis work was up to their speci-
fications. With time always a critical factor, Lewis engineers reasoned that the less energy spent
in navigating Johnson bureaucracy, the more time they would have for ensuring that Centaur
was flight-ready. Red Robbins, head of the Shuttle/Centaur Program at Lewis, argued, I had
so much trouble with JSC the last thing I wanted them to do is muck around in my budget,
so I picked the payload option, and it probably can be argued forever whether it was better to
be payload or element.40 Though initially pleased with the new integration arrangements, he
later regretted this decision, as did other engineers.41 Nieberding concluded, I really believe
that had we been an element from the beginning, and worked with the Johnson engineering
staff side by side for four years so they could get some confidence in our vehicle, as opposed
38
Andrew Stofan correspondence, Box 2142, Division STED Project Management Office, NASA GRC Records.
39
Shuttle/Centaur Project Plan, Drawer H-D, NASA GRC Archives.
40
41
William H. Robbins to Johnson Space Center, 16 August 1983, Box 2142; Mission Operations Directorate Systems Division
Operational Role for Centaur Missions, 27 March 1984, NASA Archives, Box 2142, Division STED Project Management
Office, Folder Johnson Correspondence 1984; Glynn S. Lunney to Lewis Research Center, 15 June 1984, Box 2142, Division
STED Project Management Office, Folder Johnson Correspondence 1984, NASA GRC Records.
chapter 7 3/9/05 10:52 AM Page 199
to treating it as a payload . . . I think we would have had a lot more success.42 Edward Bock,
a member of the Shuttle/Centaur program office at General Dynamics between 1983 and
1986, said that the Shuttle payload criteria applied to satellites and spacecraft that were rela-
tively inert while they were in the cargo bay of the Shuttle:
When these rules were applied to a high energy upper stage such as Centaur, they
resulted in design solutions that were asinine in their complexity and cost. The
resulting reliability of these complex design solutions was poor. As these designs
matured, this became obvious to all parties involved, and criteria compromises had to
be reached to provide rational design solutions. This cost time and money, and the
resulting compromises were more akin to element design criteria. These payload
criteria could not have been reasonably applied to any of the other Shuttle elements,
which were designed to different standards. Centaur was an element of the Space
System, and should have been designed that way from the start.43
After Centaur was redesignated as a payload in 1983, responsibility for performance and
reliability for Centaur and CISS shifted to Lewis, with Johnson maintaining the power to over-
ride any Lewis decision. Johnson Space Shuttle Manager Glynn S. Lunney (who himself had
worked at Lewis Research Center until 1958) explained that this decision was made because
the Centaur project has demonstrated reliability and performance history and the imposition
of Shuttle [element] requirements beyond those required to assure interface compatibility and
safety is neither cost effective nor consistent with the agreed responsibilities.44
With payload status, Centaur was treated like any other piece of cargo, but, in fact, it was
a living, breathing rocket, full of hydrogen.45 Because this payload was going to be launched
from within the Shuttle, it imposed a new reality upon Johnson payload-integration experts.
When deployed, the payload influenced the trajectory of the craft and the amount of propel-
lant Centaur required to adjust its course. Johnson Space Center engineers were responsible
for this deployment decision, but Lewis had to negotiate with the Johnson payload experts.
Frustrations and problems quickly mounted.
Over the next few years, the correspondence between Robbins and Lunney revealed the
conflict that surrounded Shuttle/Centaur. Although Robbins initially indicated to Lunney that
42
43
Edward Bock communication to authors, 11 March 2002.
44
Glynn S. Lunney to NASA Headquarters, 14 April 1983, Box 2142, Division STED Project Management Office, Folder
Johnson Correspondence 1983, NASA GRC Archives; Henry C. Dethloff, Suddenly Tomorrow Came . . . A History of the Johnson
Space Center (Washington, DC: NASA SP-4307, 1993), 278.
45
chapter 7 3/9/05 10:52 AM Page 200
he was pleased with Johnson Space Center internal roles and responsibilities and the plan to
organize the entire integration, the relationship soon deteriorated.46 Robbins always main-
tained his great respect and admiration for Lunney, whom he described as a no nonsense, lets
get the job done without the politics kind of guy.47 But these feelings never extended to the
rest of Lunneys team at Johnson.
Complicating the payload-versus-element question was the decision by Johnson Space
Center to add further requirements for Shuttle/Centaur that went beyond previous Shuttle
payloads. One of the most onerous was a fluid systems interface and the special requirements
for dumping the liquid-hydrogen propellant if the mission had to be aborted. In late 1984,
Johnson released a Centaur Safety Waiver (WACR-1A) that requested a redesign of the fill,
drain, and dump vent system. The problem it wanted addressed was the placement of the
liquid-hydrogen and liquid-oxygen vent line inlets below the propellant liquid level in case of
a mission abort. The result, according to Lunney, was potentially catastrophic to the Orbiter
and Crew.48
Lewis engineers were greatly angered by this new request and believed these safety waivers
to be the fundamental problem all the way through the Shuttle/Centaur program.49 As early
as October 1982, just as the project was commencing, during a Shuttle/Centaur meeting at
Johnson, Lunney and others expressed concern about the safety aspects of combining the
Centaur project with the Space Shuttle program.50 At that meeting, Robbins gave a report to
ease the concerns about safety at Johnson. Again and again throughout the project, the safety
issue came to the forefront of discussions. Thus, for the latest safety waiver in 1984 concerning
the redesign of the propellant-dumping system, Robbins wrote to Lunney that although this
request was technically feasible and might have been a reasonably straightforward design
task a year ago, now it would have a serious impact on the program.51 Robbins contended that
the redesign would delay Centaur by eight weeks, which jeopardized the fixed launch window
46
William H. Robbins to Glynn S. Lunney, 24 November 1982, Box 2142, Division STED Project Management Office, Folder
Johnson Correspondence 1982, NASA GRC Records.
47
48
Glynn S. Lunney to William H. Robbins, 10 October 1984, Box 2142, Division STED Project Management Office, Folder
49
50
Minutes for Shuttle/Centaur Board, 28 September 1982, from Melvin E. Dell, concurrence of G. S. Lunney, Glenn Research
Center, unprocessed Shuttle/Centaur records, NASA GRC Records.
51
William H. Robbins to Glynn S. Lunney, 25 October 1984, Box 2142, Division STED Project Management Office, Folder
chapter 7 3/9/05 10:52 AM Page 201
of the next mission. Robbins concluded, I am not in a position to add risk to the mission
launch schedule by accepting and implementing this new design requirement.52
Lewis engineers also believed that engineers at Johnson were unresponsive to their needs.
The attempt to transition Centaur to a piloted rating was complex, and the engineers required
a vast amount of knowledge about the Shuttle. Lewis engineers often felt as if their requests
were ignored. As Larry Ross said, They almost deliberately debilitated the program by not
being responsive.53 Although this lack of responsiveness was partly due to the strained rela-
tionship, it was also caused by the fact that this was an extremely difficult time at Johnson.
The effort to prepare the early Shuttles for flight was immediate and enormous, and often the
requests from engineers working on a payload not scheduled until 1986 took a back seat. Faced
with a lack of information, Lewis engineers relied on their judgment to solve a problem.
Often, when the Lewis engineers brought their solutions up for review at Johnson, they were
rejected. The result was frequent requests for design changes throughout the project.
One such disagreement concerned vehicle testing. Johnson wanted to perform their
harness testing on the vehicle itself, while Lewis wanted to test it before installation was
complete. Robbins argued, Our rationale for this position is based on over twenty years of
experience on Centaur where this method has been successfully proven and used.54 Robbins
believed that performing the tests after installation would incur an additional cost of $500,000
and delay the project by two weeks. He was not willing to jeopardize the mission by failing to
meet the schedule.
In February 1985, Lunney articulated a problem he saw developing between the two
Centers.55 He argued that the Lewis team believed that the problem stemmed from the
changing flight requirements at JSC. Others attacked Lewis engineers for their tardiness in
identifying problems with the rocket. Lunney proposed a solution of a one-week dedicated
effort between the Centers to iron out their differences. The idea was accepted, and Larry Ross
wrote, Its a good idea, and Im sure it will go a long way in achieving closure on the nagging
STS/Centaur safety issues.56
52
William H. Robbins to Glynn S. Lunney, 25 October 1984, Box 2142, Division STED Project Management Office, Folder
53
54
William H. Robbins to Glynn S. Lunney, 7 December 1984, Box 2142, Division STED Project Management Office, Folder
55
Glynn S. Lunney to Lewis Research Center, 7 February 1985, Box 2142, Division STED Project Management Office, Folder
Larry J. Ross to Glynn S. Lunney, 21 February 1985, Box 2142, Division STED Project Management Office, Folder Johnson
56
Correspondence 1985, NASA GRC Records.

chapter 7 3/9/05 10:52 AM Page 202
The result was the development of a Centaur flight decisions philosophy in January
1986. This was a set of mission actions by which the Centaur would be jettisoned from the
Shuttle if it encountered any critical system failures such as loss of power from the Shuttle to
Centaur, any propellant tank leaks, or any leaks in the helium tanks. Detaching Centaur in
these (and other) cases would help preserve the lives of the crew and protect the costly Shuttle
itself.57 These safety issues were still unresolved when NASA experienced one of its most tragic
and devastating catastrophes in its history, and one that would spell the end of the
Shuttle/Centaur program: the loss of Challenger in January 1986. To provide a better under-
standing of the events leading up to this calamitous event, the following account will trace the
yearlong countdown to ready Shuttle/Centaur for launch. Built in San Diego, it left the hands
of the contractor and journeyed from west to east to the launch pad in Florida, where it sat
the day of the Challenger disaster.
T Minus One Year: The Countdown Begins
In May 1985, one year before the first launch of Shuttle/Centaur, the astronaut crews were
named for the flight. Challenger was the designated Shuttle vehicle to launch the Ulysses
probe. The commander was to be Frederick (Rick) H. Hauck, on his third flight after previ-
ously piloting a mission in 1983 and commanding another in 1984. His crew included pilot
Roy D. Bridges and mission specialists David C. Hilmers and J. Mike Lounge.58 Pilots were
second in command and had the responsibility of controlling and operating the Shuttle. The
mission specialists coordinated all Shuttle onboard operations, experiments, and spacewalks.
That same month, Johnson Space Center announced the Atlantis crew that would launch
the Galileo probe. The commanding officer would be David M. Walker, a captain in the Navy
who had flown the Shuttle the year before. The pilot was to be Ronald J. Grabe, a lieutenant
colonel in the Air Force. The two mission specialists were James Ox Van Hoften and
Norman E. Thagard.59 These were high-profile missions, and everyone expected the assigned
crew members to become highly visible as soon as their scheduled flights are completed.60
Rick Hauck was a key member of the team because he was the Astronaut Office project
officer for Shuttle/Centaur integration. Hauck was a distinguished pilot and commander.
After earning a masters degree in nuclear engineering from MIT, he became a U.S. Navy test
57
Arnold D. Aldrich, Manager, National Space Transportation System, Johnson Space Center, 13 January 1986, Box 2142,
Division STED Project Management Office, Folder Johnson Correspondence 1986, NASA GRC Records.
58
Charles Redmond and Steve Nesbitt, NASA Names Astronaut Crews for Ulysses, Galileo Missions, NASA Headquarters
Press Release, 31 May 1985, NASA GRC Public Affairs Archives.
59
Charles Redmond and Steve Nesbitt, NASA Names Two Space Shuttle Crews, NASA Headquarters Press Release, 19
September 1985, NASA GRC Public Affairs Archives.
Ken Nus, General Dynamics Space Program, general memo, 4 June 1985, Box 2142, Division STED Project Management
60
Office, Folder Johnson Correspondence 1985, NASA GRC Records.

chapter 7 3/9/05 10:52 AM Page 203
Shuttle/Centaur rollout at General Dynamics, August 1985. (Courtesy of the San Diego Aerospace Museum.)
chapter 7 3/9/05 10:52 AM Page 204
pilot. In 1978, NASA began training him as an astronaut for the Shuttle. He served as pilot
on STS-7 in June 1983 and commander of STS-51A in November 1984. Throughout the
preparation of Shuttle/Centaur, the astronaut crews were heavily involved in many of the deci-
sions regarding the launch, particularly software and procedural development. Because of the
risk factors associated with Shuttle/Centaur, Jesse Moore, head of human spaceflight,
frequently invited Hauck and Walker to attend senior management meetings where key devel-
opment issues were discussed. While it was unusual to have astronauts attend these meetings,
Moore felt that it was essential for the astronauts to be involved as much as possible. Hauck
said, There was a tremendous amount of focus within the crew on working major issues that
probably wasnt typical of many of the other Shuttle missions.61 These issues included
concerns surrounding pushing the Shuttle engines to the 109-percent power level (the
previous high was 104 percent). Another problem was that the Shuttle had to orbit at the
lowest altitude possible, at 105 miles, because of its limited lift capability. Finally, there were
safety issues that the crew had to understand for an abort. If the Shuttle had to return to Earth
with Centaur still inside, the center of gravity for the spacecraft would be further aft than ever
before, making landing particularly difficult.
Kennedy Space Center was busy making plans for the Galileo and Ulysses launches. The
two Shuttles would be housed at Complex 39 on launch pads A and B. To complicate matters,
there were only six days scheduled between the launches, with each launch having only a 1-
hour window. Officials at Kennedy felt pressure from both the media and the public to
duplicate the success of the Viking and Voyager missions and to add to the dramatic images
and scientific data these spacecraft returned. A Kennedy briefing warned, Remember the
publics response to . . . pictures of Saturn and Jupiter. They will be looking for more. They
and the world will be watching.62 Lewis Research Center also had the pressure of getting two
Centaurs ready for launch. With ten months to go before liftoff, they began their much publi-
cized Centaur Countdown. They reported that while much remains to be done, all activities
support the crucial schedule.63
When scientists and engineers at General Dynamics triumphantly unveiled
Shuttle/Centaur (SC-1) in San Diego in August 1985, such excitement had rarely been seen
since the glory days of the Apollo program. The theme from Star Wars accompanied the
applause of more than 300 officials from the company, the Air Force, and NASA. Craig
Thompson, a General Dynamics operations representative, said, I almost had tears in my eyes
61
Interview with Rick Hauck by Mark Bowles, 23 August 2001.
Charter Planetary Panel Galileo and Ulysses Mission, Kennedy Space Center, Box 2142, Division STED Project
62
Management Office, Folder Kennedy Correspondence, NASA GRC Records.
63
Vernon Weyers, Centaur Countdown, Lewis News (9 August 1985): 2, GRC Records.
chapter 7 3/9/05 10:52 AM Page 205
when they rolled that thing out.64 Alan Lovelace proclaimed it the future of space probes in
our generation.65 The vehicle then underwent three weeks of extensive checking to make sure
that all specifications and requirements were met. After passing all of the tests, SC-1 was flown
to Cape Canaveral, where the crucial CISS was waiting for its mate. The CISS had been flown
to the Cape two months earlier. It was placed on a converted Atlas-Centaur launch pad, where
it received its final assembly and checks.
In September 1985, with eight months to go, Vernon Weyers, Deputy Manager of the
Lewis Shuttle/Centaur Project Office, wrote, All activities are in high gear in preparation for
the dual planetary launches in May/June 1986.66 There were some minor delays in assembling
and mating the SC-1 with the CISS-1. There was a shortage of parts that were supplied by
General Dynamics; this shortage resulted in about a one-month delay. However, Weyers was
confident that this was not a problem because there was a two-month period built into the
schedule for such common delays. The only trouble was that this luxury was not available for
the next launch. The SC-2 and CISS-2 were right behind the first vehicle and did not have
this schedule slack, so future delays could not be tolerated.
A month later, progress was still on track, and Larry Ross reported, A lot of hard work
remains, but all of the tasks required to launch the first Shuttle/Centaur seven months from
today are doable.67 General Dynamics sent the missing parts to Cape Canaveral, and the SC-
1 and the CISS-1 were successfully mated together. A test rotation and separation of the
Centaur was also completed successfully. Engineers hoped that this was the last time that this
test would be performed before the real separation took place in orbit above Earth. SC-2 and
CISS-2 were continuing their final inspections at General Dynamics in San Diego and would
be transported across the United States in November. To reduce the difficulty in housing both
the SC-1 and SC-2 at Kennedy Space Center (KSC) at the same time, the Air Force made their
Shuttle Payload Integration Facility available between November and December. This gave
NASA the ability to process both vehicles simultaneously. Other good news included the
report of a 109-percent engine thrust level for the Galileo mission (100 percent was consid-
ered the optimum safety level, and the previous high was 104 percent). This thrust capability
made it possible to launch a heavier Centaur that carried more propellant. With more energy,
there were more available launch days. Marshall engineers were actually displeased by this
because they did not want to operate their engines at this risky thrust level. A higher thrust
64
Michael L. Norris, quoted in New Booster Rolled Out in San Diego, The Los Angeles Times (14 August 1985).
65
Vernon Weyers, Centaur Countdown, Lewis News (6 September 1985): 3, GRC Archives.
66
Vernon Weyers, Centaur Countdown, Lewis News (4 October 1985): 3, GRC Archives.
67
Vernon Weyers, Centaur Countdown, Lewis News (1 November 1985): 5, GRC Archives.
chapter 7 3/9/05 10:52 AM Page 206
level pushed the engines past safe performance levels. The higher the accepted percentage level,
the more powerful the thrustand the higher the risk.
With six months to go, the SC-1 and the CISS-1 were still undergoing testing at the Space
Center Eastern Test Range at Kennedy Space Center. Special attention was given to leak checks
on the hydraulic system. A few minor leaks were detected, but all were tracked down and
repaired. The next key test was the cryogenic cold flow test that was held in mid-November.
Two successful tests were performed for both the liquid-oxygen and liquid-hydrogen systems.
SC-1 and CISS-1 also passed the first two phases of the Design Certification Review. The final
stage was the presentation of the certification to NASA Associate Administrator Jesse Moore.68
While the first Shuttle/Centaur was being tested, its twin arrived in Florida. The CISS-2 was
the first component to show up, and it waited for its mate. The SC-2 was still in California,
where General Dynamics was performing propellant tank insulation and avionics system
checks.
In January 1986, Centaur encountered two significant technical problems. During a test
of the mounts that attached the Propellant Level Indicating System inside the liquid-oxygen
tank, the mounts failed. Engineers quickly redesigned them, then fabricated and installed
them on the SC-1. The new mounts passed all subsequent tests. The second main problem
occurred during the liquid-oxygen and liquid-hydrogen cold flow tests. Specifically, the valves
on the CISS-1, which controlled the fill, drain, dump, and engine feed functions, experienced
erratic operations. Using a simulated system, engineers located this problem as well and devel-
oped a redesign. New tests were scheduled, and 15 January became a do or die hurdle for the
program.69 Both propellant tanks were completely filled and pressurized at 110 percent of
flight pressure, and all of the major system procedures were tested. Vernon Weyers recounted,
Nobody close to the program dared to predict the success which was achieved . . . the entire
avionics system performed flawlessly.70 Robbins was equally pleased with the performance and
concurred that a very important milestone was completed.71 Ross said, The tanking was
absolutely perfect.72
The astronauts were not as enthusiastic. In early January 1986, they became concerned
about whether the abort of a Shuttle/Centaur mission could be safely orchestrated. They ques-
tioned whether there was adequate helium pressure to activate the dump valves if something
went wrong. They worried that there were not adequate backup systems to ensure their safety,
68
Vernon Weyers, Centaur Countdown, Lewis News (29 November 1985): 4, GRC Archives.
69
Vernon Weyers, Centaur Countdown, Lewis News (10 January 1986): 3, GRC Archives.
70
Vernon Weyers, Centaur Countdown, Lewis News (7 February 1986): 3, GRC Archives.
71
W. H. Robbins, Shuttle/Centaur Weekly Status Report, 1317 January 1986, Box 2143, Division STED Project
Management Office, Folder Status Reports, NASA GRC Records.
72
chapter 7 3/9/05 10:52 AM Page 207
and they began to refer pejoratively to Centaur as Death Star.73 Astronaut Rick Hauck said,
This was a big issue for us. Very big. Hauck and John Young, the head of the Shuttle office,
took up the safety issue with the Johnson Space Center Configuration Control Board. They
were surprised when the Board ruled the system acceptable. Hauck recalled, I went back to
my crew that day and said, Things have changed at NASA. We are willing to take risks that
we didnt take before and if any of you want to resign from this flight I will support you.74
However, when given a chance to withdraw, with no questions asked, from the crew that was
assigned to fly Shuttle/Centaur, none of the selected crew accepted the offer. Three weeks later,
however, the Challenger disaster radically altered the perception of risk.
On 28 January, Challenger lifted off from Cape Canaveral; minutes later, it exploded over
the Atlantic Ocean, thereby killing the crew. This was the worst disaster in NASA history up
until that time, and the first time that any American astronaut was ever lost in flight.
Immediately, the Shuttle/Centaur program was in jeopardy. Despite years of work, the rocket
quickly began to fall out of favor with both Johnson Space Center and NASA Headquarters.
Fighting for Survival
The Centaur team at Lewis was devastated. Since Shuttle/Centaur was so close to the date
of its first launch, many of the Lewis engineers working on the project were already at the Cape
and witnessed the Challenger explosion with their own eyes. They immediately realized that
Shuttle/Centaur, along with Galileo and Ulysses missions, was in jeopardy. When Red Robbins
returned to Cleveland the next Monday morning, he went directly to the personnel office and
tendered his resignation. Already close to retirement, Robbins knew that at best,
Shuttle/Centaur would be years from launching, and he could not endure the wait.
Although the Shuttle/Centaur program was not immediately canceled, its future was very
uncertain. On 20 February 1986, Jesse Moore sent out the order to postpone the Galileo and
Ulysses launches.75 The very earliest that the two missions could be rescheduled was thirteen
months away. This was the time when Jupiter would be on the direct opposite side of the Sun
from Earth. Even the Ulysses mission was dependent upon the location of Jupiter because it
was going to receive a gravity-assist boost from the planet to catapult its way to the Sun. In the
meantime, the Galileo probe was moved to the Vertical Processing Facility at KSC, where it
was mated with the Centaur. Engineers continued to perform compatibility and separation
tests in the hopes that the missions might still get a green light.76
73
74
Ibid.
75
Jesse Moore, Galileo and Ulysses Launch delay, 20 February 1986, Box 2143, Division STED Project Management Office,
Folder Status Reports, NASA GRC Records.
76
Galileo, Centaur Milestone Tests are Underway, Lewis News (16 May 1986): 2, NASA GRC Archives.
chapter 7 3/9/05 10:52 AM Page 208
Acting NASA Administrator William R. Graham said that his decision to postpone did
not mean that the next Space Shuttle launch would be delayed until after the Galileo and
Ulysses launch opportunity the next May. The Shuttle might be back in operation earlier than
thirteen months, but he said that two key factors forced his hand to delay these missions: 1)
key personnel required to ensure the safe and successful launch of either Galileo or Ulysses
were preoccupied with the timely analysis of causes of the 51-L accident and 2) the conse-
quences of the accident had significantly eroded the schedule margins for launch-site
processing required prior to the first flight of the Shuttle/Centaur upper stage.77
Despite guarded optimism, even if an additional thirteen months of testing could be
performed, the hopes for an eventual Shuttle/Centaur launch were beginning to fade. One
reason for concern was the issue of weight and the ongoing disagreement between Lewis and
Johnson. The Lewis research team had three key weight factors to consider: the weight of
Centaur, the weight of its payload (Galileo), and the weight of the Centaur propellant. When
Johnson committed to a 65,000-pound lift capability for the Shuttle, Lewis knew that they
were slightly over the weight allowance. However, the actual lift capability of the Shuttle was
never close to that commitment. To compensate, Centaur was able to decrease the amount of
propellant carried in its tanks. This tradeoff meant that the launch window shortened because
Centaur would have insufficient fuel to hit the planetary target on those days that required a
higher spacecraft velocity at separation.
Early in the integration process, Lewis engineers struggled with the weight problem
because Johnson engineers kept decreasing their allowance for the Centaur weight.
Approximately two weeks before the Challenger launch, Lewis trajectory engineer Joe
Nieberding made a presentation to Jesse Moore and to Lewis, KSC, JSC, and MSFC top
management sounding the alarm that the number of viable launch days had decreased to fewer
than six for Galileo, far too few to ensure a reasonable probability of launch. In the presenta-
tion, he demonstrated that using 109-percent Shuttle engine thrust (instead of the previously
high level of 104 percent) was the only potentially viable way of restoring a reasonable number
of launch days. In fact, Jesse Moore had approved the higher thrust level on the spot, over the
strenuous objections of JSC and MSFC. So when Challenger exploded, Lewis engineers knew
that any redundancies added to make the Shuttle safer would also make it heavier. It would be
able to carry even less than before. Although in no way related to the failure, the fact that
Challenger exploded only a few seconds after its engines reached 104 percent of its nominal
thrust contributed to the perception that pushing the Shuttle engines to 109 percent on the
77
James F. Kukowski, NASA Postpones Galileo, Ulysses, Astro-1 Launches, NASA Headquarters Press Release, 10 February
1986, GRC Public Affairs Archives; William R. Graham, Galileo and Ulysses Missions Postponed for Year, Lewis News (21
February 1986): 6, NASA GRC Archives.
chapter 7 3/9/05 10:52 AM Page 209
Galileo mission would be unsafe.78 With the Shuttle getting heavier and able to lift less weight
when it returned to flight, and with JSC and MSFC now vehemently opposed to pushing the
Shuttle engines above a 104-percent thrust level, Lewis engineers realized that the Shuttle
would never be able to lift Centaur and Galileo.79
Another issue was safety as it related to Centaurs fuel. Following the Shuttle disaster,
Lewis held meetings on the Centaur safety program in May 1986. The attendees included
representatives from General Dynamics, JSC, KSC, the Air Force, TRW, Boeing, Lockheed,
Martin Marietta, and Analex, an aerospace industry service provider located in Brookpark,
Ohio. The goal of the meeting was to prove that despite the risk presented by liquid hydrogen,
the Centaur rocket was safe. Engineers described the entire safety organization for the Centaur
at Lewis, reviewed recent hazard reports, and presented a revised abort/contingency plan.80 But
the effort was not enough to convince opponents.
Even more serious, from the point of view of engineers at Lewis, was the realization that
increased redundancy in Centaur and Shuttle systems meant that the Shuttle would no longer
be able to lift Centaur. They knew that the Shuttle engines would not be allowed to push
beyond the 100-percent thrust level, significantly less than the 109-percent level needed.
Another problem that contributed to the growing pessimism over the future of Centaur
concerned General Dynamics. Despite a long and successful relationship, Lewis management
was coming to believe that General Dynamics no longer possessed the requisite skills to build
and staff Centaur effectively. The astronauts were especially critical.81 When astronaut David
Walker went out to the plant, he was surprised at the conditions he found. Compared to the
Rockwell facility, where the Shuttle was meticulously assembled in a clean, white-glove envi-
ronment, there seemed to be less attention to manufacturing a quality product at General
Dynamics. Larry Ross recalled one incident when there was a lapse in common sense on the
part of workers moving a Centaur from the factory to a test site. They took a Centaur G and
ran it into a bridge going up to Sycamore Canyon, he said. It was dumb. They hadnt meas-
ured the height. He felt that some of the fault lay in Lewiss management. The lesson was
that we should have fired some people at General Dynamics. That was a mistake.82 Vernon
Weyers also was aware of the problems at General Dynamics. He visited the plant once or
twice a month over the four-year Shuttle/Centaur program. He suspected that drugs were a
problem. He recalled, I personally was out in the shop by myself at General Dynamics one
78
79
80
William E. Klein, Minutes for Shuttle/Centaur G Safety Certification Process Briefing, 2 June 1986, Box 2142, Division
81
chapter 7 3/9/05 10:52 AM Page 210
day where they were working on a big piece of structure. I looked through the access hole, and
an engineer was sound asleep. It was very frustrating at times, because on the technical side,
the production side, they did not seem to get their act together . . . . They just did not perform
well technically.83
On 22 May 1986, Headquarters began to address the problems surrounding
Shuttle/Centaur and consider what future action they should take. Rick Hauck gave an influ-
ential briefing citing risks to the Shuttle crew as the reason why the program should be
canceled. He concluded that Shuttle/Centaur, even after satisfactory accomplishment of
currently identified safety-related fixes, poses additional risk to crew and Shuttle safety.
Attempts to integrate an unmanned upper stage in the Shuttle have resulted in compromises,
which caused undue risk to orbiter and crew. The ability to reduce these risks to an acceptable
level is questionable.84 An independent study of safety concerns, conducted by the House
Committee on Appropriations and chaired by Edward P. Boland and Congressman William
Green, concluded that it was in the best interest of NASA to terminate the Shuttle/Centaur
program.
On 19 June, NASA Administrator Dr. James C. Fletcher gave the order to terminate the
program.85 After the country had spent nearly one billion dollars to transform Centaur for the
Shuttle, Lewis Research Center received this heartbreaking order: You are directed to termi-
nate the Shuttle/Centaur upper stage program. 86 Fletcher stated, Although the
Shuttle/Centaur decision was very difficult to make, it is the proper thing to do and this is the
time to do it.87 Astronaut and future NASA Administrator Richard H. Truly was directed to
examine other alternatives for the Shuttle/Centaur planetary and scientific payloads.
The result was catastrophic for the space program. Suddenly, the Galileo probe again lost
its ride into space. Not only was it forced to wait until a solid rocket upper stage could be built,
but also, the lack of power meant that the entire mission would take much longer. With a
Centaur launch, Galileo would have arrived at Jupiter in two years. When Galileo finally did
launch in October 1989 with the less powerful Boeing-built Inertial Upper Stage, it took six
years to arrive in the Jovian atmosphere.88
82
83
Interview with Vernon Weyers by Mark Bowles, 8 April 2000.
84
85
John M. Logsdon, Return to Flight: Richard H. Truly and the Recovery from the Challenger Accident, in From Engineering
Science to Space Science, 360.
86
Shuttle/Centaur Termination Status NASA Headquarters, 4 September 1986, Box 2144, NASA GRC Records.
87
Sarah G. Keegan, NASA Terminates Development of Shuttle/Centaur Upper Stage, NASA Headquarters Press Release, 19
June 1986, NASA GRC Public Affairs Archives.
88
Barbara Selby, Leon Perry, and Terry Eddleman, Upper Stage Selected for Planetary Missions, NASA Headquarters Press
Release, 26 November 1986, NASA GRC Public Affairs Archives.
chapter 7 3/9/05 10:52 AM Page 211
The loss of Centaur also dramatically changed the navigational course that Galileo took.
With Centaur, the Galileo flightpath was to have been direct. With the solid IUS, Galileo was
required to take a much more circuitous route with the help of gravitational assists. Galileo
actually flew by Venus once and Earth twice, each time using the planets gravity to effectively
steal some energy and direct it to the velocity of the craft. Although this reduced the dura-
tion of the flight, the six years to arrive in the Jovian atmosphere still took three times as long
as what Centaur was capable of delivering.
Not only did the Centaur cancellation result in nearly a decades delay of Galileo for data
return, but it also significantly jeopardized mission success. In April 1991, Galileo attempted
to open the large high-gain antenna that was the primary mechanism by which the craft trans-
mitted Jupiter data to Earth.89 Soon after the attempt, telemetry revealed to the JPL engineers
that something had gone wrong. The motors only partially opened the antenna, and then they
stalled. Over a period of weeks, more than a hundred experts at JPL analyzed the problem and
concluded that the antenna failed because of excessive friction between antenna pins and
sockets. Ironically, JPL engineers blamed this problem on the cancellation of Centaur. In
1985, Galileo and the antenna were shipped by truck from JPL to Kennedy. After the
Challenger explosion, Galileo was shipped back across the country in 1986, and then back
again in 1989, when it was finally launched. The vibration that the antenna experienced
during the cross-country trips by truck loosened the pin lubricant and caused the main
antenna to fail.
Although some important scientific observations were lost, the Galileo mission remained
a success when engineers reconverted the low-gain antenna.90 In October 1991, Galileo
performed a flyby operation of the 8-mile-diameter asteroid Gaspra. When Galileo passed by
at a distance of 1,000 miles, it was the first time a spacecraft had provided direct observations
of an asteroid.91
Like Galileo, the Ulysses mission also suffered from the cancellation of the
Shuttle/Centaur program. D. Eaton was the Ulysses project manager at the European Space
Agency. Upon hearing news of the suspension of the Centaur program, he and his team reacted
with deep disappointment. Despite the wasted money and the delay to the Ulysses program,
Eaton said that the overriding thought is the bitter blow this must be to all at Lewis and
Paula Cleggett-Haleim, Galileo Antenna Deployment Studied by NASA, NASA Headquarters Press Release, 19 April 1991,
89
90
New Telecommunications Strategy Aimed at Maximizing Return from Galileos Low-Gain Antenna,
http://cass.jsc.nasa.gov/publications/newsletters/lpib/lpib76/gal76.html.
91
Paula Cleggett-Haleim and Jim Wilson, Galileo To Set Course for Encounter with Asteroid Gaspra, NASA Headquarters
Press Release, 28 June 1991, NASA GRC Public Affairs Archives.
chapter 7 3/9/05 10:52 AM Page 212
General Dynamics who worked intensely long hours to achieve an impossible schedule only to
have the goal posts taken away and the game abandoned at the last minute.92 Like Galileo,
Ulysses also eventually made it into space. In May 1991, the spacecraft took its long journey
to Jupiter for an eventual slingshot to the Sun. While at the planet, it performed physics inves-
tigations such as taking measurements of the Jovian magnetosphere.93
To engineers at Lewis Research Center and JPL, however, the abandonment of the
program was not in the national interest. Nieberding said, To this day, the country is hurting
for not having a liquid-hydrogen upper stage coupled with the Shuttle or with an equivalent
expendable launch vehicle capability.94 Tom Shaw at JPL recalled, June 19, 1986, was a dark
day as far as I was concerned. I felt that we suffered a truly significant setback in national capa-
bility, and specifically my own narrow purview of planetary exploration, where we need
high-energy vehicles to get there with any kind of significant payload . . . . You learn how to
use a much, much smaller spacecraft.95
Beyond their disappointment over the cancellation of Shuttle/Centaur, Lewis Research
Center engineers pointed out that many missions carried by the Shuttle in the 1980s,
including the commercial satellite payload carried by the ill-fated Challenger, should have been
assigned to an expendable launch vehicle such as Atlas-Centaur. Red Robbins observed, I
think they sold their soul to the devil to fund the Shuttle . . . the expendable launch vehicle
was a good animal and not nearly as expensive as the Shuttle program. He said that while the
expensive Saturn V could deliver approximately 250,000 pounds of payload to low-Earth
orbit, the Shuttle, which eventually cost as much, could deliver the same weight into space.
The problem was that 200,000 pounds of that weight was the Shuttle itself. Thus, with the
Shuttle payload of only 50,000 pounds, the cost per pound to fly into low-Earth orbit with
the Shuttle is up by a factor of 5. You dont even need a rocket scientist to tell you that.96
Terminating the Program
Ends can come for many different reasons. Technologies can evolve to the point where
older designs are obsolete. Mission objectives can change, making previous areas of exploration
unnecessary. Costs can escalate to the point where funding no longer allows the continuation
D. Eaton to Larry Ross, 11 July 1986, Box 2143, Division STED Project Management Office, Folder Galileo Correspondence,
92
NASA GRC Records.
Paula Cleggett-Haleim and Robert MacMillin, Ulysses To Begin Jupiter Physics Investigations, NASA Headquarters Press
93
Release, 28 May 1991, NASA GRC Public Affairs Archives.
94
95
96
Interview with Red Robbins by Mark Bowles, 21 March 2000.
chapter 7 3/9/05 10:52 AM Page 213
of a project. Managerial and technical errors can jeopardize a mission. Although these are the
most common reasons for premature endings to NASA projects, Shuttle/Centaur did not
encounter any of these fates. The technology was still cutting-edge; planetary missions were
still needed; the funding remained intact; and no uncorrectable errors were made in either the
design or the management of the program. Shuttle/Centaur met with one of the most devas-
tating and uncontrollable problemsfear. After the Challenger explosion, a liquid-hydrogen
rocket inside the bay doors of the Shuttle just looked too dangerous and too risky for NASA
to continue with the program.
In June 1986, NASA sent out termination letters to their own Centers and to contractors
like General Dynamics, Honeywell, Analex, Teledyne, and Pratt & Whitney. Their order was,
You are hereby directed to take those necessary steps for an orderly shutdown and cessation
of all work.97 Each of these organizations had to slowly bring Centaur work to a halt, resulting
in over 200 stop-work orders.98 With great remorse, Andy Stofan wrote, Lewis is proceeding
with the orderly shutdown of the terminated Shuttle/Centaur program.99 Throughout July,
closeout meetings were held with all the contractors. Most of the work of the contractors was
completed by 30 September 1986; all work was to be finished by the end of the year. Each
contractor was expected to set a completion date for itself that would maximize the potential
use of its work for the government. Because Centaur might be used again as an expendable
launch vehicle, all contractors were also instructed to assess the impact that the end of the
Shuttle/Centaur program would have on future capabilities.100
The remaining resources from the Shuttle/Centaur program included documentation,
technology, and people. The documentation was the easiest resource to relocate. Fortunately,
in February 1986, NASA created the Centaur Engineering Data Center where all engineering
paperwork associated with the project was sent. Besides increasing the speed in which infor-
mation could be accessed, it was also designed to relieve engineers of filing chores and
promised to free up work space.101 This information included official files stored in personnel
offices, the mailroom, contractor sites, and storage facilities.102 The Data Center received key
97
Termination letter 23 June 1986, Shuttle/Centaur Termination Status NASA Headquarters, Box 2144, NASA GRC Records.
R. E. Jumont to W. H. Robbins, 16 July 1985, Box 2142, Division STED Project Management Office, Folder General
98
Dynamics Correspondence, NASA GRC Records.
99
Andrew J. Stofan memo to R. Truly, 8 July 1986, Shuttle/Centaur Termination Status NASA Headquarters, Box 2144,
NASA GRC Records.
Vernon J. Weyers to distribution, 1 July 1986, Box 2142, Division STED Project Management Office, Folder General
100
Correspondence, NASA GRC Records.
James O. Rogers, Jr., to distribution, 20 February 1986, Box 2142, Division STED Project Management Office, Folder
101
General Correspondence, NASA GRC Records.
102
Thomas S. Banus to distribution, 11 June 1986, NASA Archives, Box 2142, Division STED Project Management Office,
Folder General Correspondence; Thomas S. Banus to distribution, 2 July 1986, Box 2142, Division STED Project Management
Office, Folder General Correspondence, NASA GRC Records.
chapter 7 3/9/05 10:52 AM Page 214
documents, and the information was entered into a database and then sent to a vault. Many
materials were then destroyed as engineers cleaned out their offices. Saved materials included
minutes of all reviews of the project, significant correspondence, and all engineering drawings.
Between October and November 1986, all the major contractors and NASA Centers began the
process of turning over their documentation.
By the middle of 1987, most of the contractors had issued final reports on various aspects
of the Shuttle/Centaur development. In May 1987, Pratt & Whitney released a final report on
the RL10 Ignition Limits Test for Shuttle Centaur. Kennedy Space Center and Lewis
Research Center summarized testing done to determine instability in the CISS and the
Pneumatic Activated Valve Control System. United Technologies issued a final report called
Shuttle Centaur Engine Cooldown Evaluation. These three reports were representative of
the type of work done to resolve technological questions that still surrounded the project.103
Preservation of the technologies associated with the Shuttle/Centaur project was a more
difficult problem. Suggestions on how to use the technology were made by engineers who
could not accept the possibility that their years of work had been for naught. For example,
Floyd Smith, the chief of the structures and facilities branch at Lewis, believed that the govern-
ment should review all potential hardware uses and then store the equipment in a suitable
location. He hoped that it might still be possible to use one of the new Centaur vehicles for a
later NASA mission. Even if NASA decided never to use the new rocket configuration, Smith
wanted other possible spinoffs from their work. He thought that one application might
support other payload mounting structures in the Shuttle.104
The Air Force also had to decide what to do with its Shuttle/Centaur hardware. Because
the Air Force was a full partner in the development costs, it was entitled to any of the residual
hardware that might be used for other craft, like Titan-Centaur.105 In June 1986, their official
termination orders stated, Through appropriate coordination with NASA Headquarters and
in view of the needs of the DOD, you are to implement the shutdown in a manner which is
most advantageous to the U.S. Government.106 To take greatest advantage of the work already
accomplished, NASA asked for additional funds to enable its contractors to take their hard-
ware development to a suitable stopping point. Thus, the completed technology would have a
greater chance of being used on another spacecraft.
103
Final reports found in Box S/C Records, consignor Alan Willoughby, NASA GRC Records.
104
R. C. Edwards for Floyd Z. Smith to Acting Manager, Shuttle/Centaur Project Office, 21 July 1986, Box 2142, Division
STED Project Management Office, Folder General Correspondence, NASA GRC Records.
C. Thomas Newman to Administrator, 29 September 1986, Box 2142, Division STED Project Management Office, Folder
105
Air Force Correspondence, NASA GRC Records.
Vernon J. Weyers, Acting Manager Shuttle/Centaur Project Office, to Headquarters Space Division, 26 June 1986, Box 2142,
106
Division STED Project Management Office, Folder Air Force Correspondence, NASA GRC Records.
chapter 7 3/9/05 10:52 AM Page 215
In the process of terminating the program, Lewis Research Center received a number of
accolades from the Air Force. Larry Ross received a letter of appreciation from Colonel David
Raspet at the Department of Defense in recognition of the exceptional performance of his
laboratory. Raspet said that because of the efforts of Lewis, despite the cancellation of the
program, We were on course with Centaur G and succeeded in laying foundations for
Titan/Centaur.107 In December 1986, Brigadier General Nathan J. Lindsay said that the
government had decided that it was in the governments best interest to buy back all of the
flight hardware from NASA. He believed that the termination has been difficult for all of
us,108 but he offered his thanks to Lewis Research Center for helping to make the process as
smooth as possible. Ross appreciated the accolades given to him and his organization by the
Air Force and wrote, Dispositioning these assets is a painful process, but I find consolation in
the potential for their use on other flight vehicles.109
Centaur G-prime lived on when it was mated, with some modifications, to the Air Force
Titan IV booster, first launched in 1994. In 1992, NASA chose Titan IVB/Centaur G-prime
as the launch vehicle for the Cassini mission to Saturn launched in 1997. This $3.4-billion
mission, a project managed jointly by NASA, the European Space Agency (ESA), and the
Italian Space Agency (Agenzia Spaziale Italiana, ASI), carried a huge nuclear-powered probe
programmed to land on the surface of Titan, the largest of moon of Saturn. The return of
Titan-Centaur, the last and most expensive of NASAs large spacecraft, underlined the appro-
priateness of expendable launch vehicles for ambitious space science missions that required the
power and precision of Centaur.
A Changing Tolerance for Risk
President Nixon had closed his 1972 announcement of the Shuttle program with a
prophetic quotation from Oliver Wendell Holmes: We must sail sometimes with the wind
and sometimes against it, but we must sail and not drift nor lie at anchor.110 Few other state-
ments could have so accurately captured Centaurs relationship with the Space Shuttle.
Implicit within Holmess words was a notion of risk. Sailing is a dangerous endeavor. When
107
David Raspet to L. J. Ross, 4 November 1986, Box 2142, Division STED Project Management Office, Folder Air Force
Correspondence, NASA GRC Records.
Nathan J. Lindsay to NASA Lewis Research Center, 9 December 1986, Box 2142, Division STED Project Management
108
Office, Folder Air Force Correspondence, NASA GRC Records.
Larry Ross to Major General Ralph H. Jacobson and Brigadier General Nathan J. Lindsay, 30 October 1986, Box 2142,
109
Division STED Project Management Office, Folder Air Force Correspondence, NASA GRC Records.
110
White House Press Secretary, The White House: Statement by the President, 5 January 1972, Richard M. Nixon
Presidential Files, NASA Historical Reference Collection, NASA Headquarters, Washington, DC, as found in Nixon Approves
the Space Shuttle, in NASA: A History of the U.S. Civil Space Program, ed. Roger D. Launius (Malabar, Florida: Krieger
Publishing Company, 1994), 235.
chapter 7 3/9/05 10:52 AM Page 216
the wind blows at our back, the chance of safely completing the voyage is increased. However,
when we set sail with the gusts directly in our face, churning the waters against the ships bulk-
head, the conditions for ultimate success are reduced. Holmes said that inherent in the nature
of humanity is the desire to raise anchors and explore the unknown. We endure the dangers.
We accept the risks of what we do because we believe that the rewards ultimately justify the
sacrifice.
Holmess eloquent argument for the adventurous spirit suggests that risk is not a constant
factor, equally endured by everyone. The tolerance for risk changes like the wind, deviating
from one person, organization, or era to the next. What one person might consider an accept-
able risk assumed to achieve some larger goal, a person with a more conservative outlook might
reject for one that yields less reward, but also less potential for failure.
Safety concerns are now given as the official reason for termination of the Shuttle/Centaur
program. One author reported in 1999 that this Shuttle configuration was eventually scuttled
because of safety concerns stemming from the Centaurs common-bulkhead tank design and
from the difficulties of dumping propellants in an abort situation.111 However, blaming the
entire failure upon technical issues related to safety obscures the engineering triumphs that
made Shuttle/Centaur flight-ready. These safety concerns were nothing new and did not
emerge suddenly after the tragic Challenger loss. Instead, they were evident from the very start
of the program and were embedded in the Centaur design. Many engineers considered that the
tank design and propellant dumping concerns were part of the acceptable risks of flight. They
worked many years to reduce the risks even further to meet NASA standards. In this they were
successful. Before the Challenger tragedy, General Dynamics confidently proclaimed, The
proven Centaur System can now provide maximum benefits to Shuttle users.112 Vernon
Weyers concluded, despite the technical challenges and problems encountered, At the time of
the Challenger accident . . . there was little doubt that the Centaur would be ready on time to
support the 1986 dual planetary launches.113 Marty Winkler of General Dynamics concurred
with this assessment: Im convinced to this day we would have made the launch window in
May of 86, but it was a sprint to the finish. It was like the racehorse that overtakes you at the
end. Had not the Shuttle accident occurred, we would have been ready for launch.114 So, if
111
Ivan Bekey, Exploring Future Space Transportation Possibilities, in Exploring the Unknown, Volume IV, ed. John M. Logsdon
(NASA SP-4407, 1999), 509.
General Dynamics, Box 2146, Division STED Project Management Office, Folder S/C Marketing Vugraphs from GDC,
112
NASA GRC Records.
Vernon Weyers, Development of the Shuttle/Centaur Upper Stage, paper delivered at the IAF Congress, Innsbruck, Austria,
113
October 1986, Box 2143, Division STED Project Management Office, Folder IAF Paper, NASA GRC Records.
114
Interview with Martin Winkler by Virginia Dawson, 21 March 2001.
chapter 7 3/9/05 10:52 AM Page 217
technically the Centaur would have been ready, something beyond safety was at play in the
decision to terminate.
The first reason was a monetary issue. NASA management knew that the effort required
to return the Space Shuttle to flight-ready status would be enormously expensive. As a result,
there would not be enough left in the budget to work out any remaining technical issues with
Shuttle/Centaur. Marty Winkler said, it wasnt canceled because of technical problems,
although there were technical problems, because everyone agreed that the technical problems
were solvable. It was canceled because NASA had concluded internally . . . that they didnt
have enough manpower, energy, and money to fix both the Shuttles return to flight and the
Centaur. Thats a little-known fact, but its absolutely true.115 Also, a repaired Shuttle would
never have had the lift capability to launch a safer and heavier Centaur.
Another component was even more central to the cancellation of the programa
changing tolerance of risk. Liquid hydrogen was such an alluring fuel because of the capabili-
ties that it gave to explore space. The tradeoff for greater power and control was heightened
safety concerns. Liquid hydrogen was dangerous, and there was never any question that it was
more volatile than solid propellants. However, during the 1970s and early 1980s, these risks
were considered acceptable. Ironically, the Challenger explosion was caused by the failure of
one of the solid-propellant booster rocketspreviously considered very reliable. The tragedy
brought with it a new era of conservatism and a diminished acceptance of risk. As Larry Ross
argued, After Challenger the equation changed.116 The design specifications that had been
acceptable before Challenger were based upon theoretical and technical judgments. In the wake
of the explosion, new emotional judgments were added to the mix. Because of these strong
emotional factors, Ross concluded, There was no crew that would fly it because they were
staggered by Challenger. It was not rational; it didnt have to be rational. It was human life.117
The astronauts agreed that the level of acceptable risk changed, and that NASA had been
willing to assume too many risks in the pre-Challenger days. Rick Hauck said that the changing
levels of risk tolerance were a part of NASAs history. He said, It is all part of the ebb and flow
of the desire to succeed and the desire to meet schedules. Probably there was a certain amount
of confidence born of never having an in-flight failure before.118 There was also a risk toler-
ance difference in 1986 between Johnson Space Center and Lewis Research Center that he
called cultural. He said that the Lewis engineers were highly motivated professionals who
believed they were exercising as much caution as needed to be exercised. But generally these
115
Ibid.
116
117
Ibid.
118
chapter 7 3/9/05 10:52 AM Page 218
engineers believed Centaur could fly in the Shuttle. As a whole, Lewis always believed that the
human spaceflight guys at Johnson were a bit too cautious. Hauck concluded, There was a
natural cultural difference between the groups. The issues about risk are tied up in the cultural
issues.119 While Hauck was on the side that supported grounding Shuttle/Centaur forever, he
conceded that there was an irresolvable difference between the two sides. Eventually, the less
risky solution won out.
Richard Kohrs was right in the middle of the debate between the engineers and the astro-
nauts. Kohrs was the manager of the Technical Integration Office for the Space Shuttle from
1980 to 1985 at Johnson Space Center. He was also the main point of contact between Lewis
Research Center and Johnson, so he understood both cultures. He conceded that
Shuttle/Centaur seemed like a good idea before they got to the now lets go make it work
phase. But when they began to go through all of the battles concerning weight issues, single
point failures, interfaces, redundancy, and so on, he and his colleagues at Johnson came to
realize that the plan was less desirable. Although initially the risks seemed worth taking, the
risk tolerance changed as the program progressed. Kohrs recalled, I think a lot of these risks
were accepted before . . . the [technical] guys had to go make it happen. Then of course you
got into the flight crew concern of this whole thing. Johnson protected the astronauts, and
because the astronaut crews united against the Shuttle/Centaur plan, Kohrs and his colleagues
also withdrew support. Although Kohrs ultimately adopted the more conservative attitude of
the astronauts, he still had faith that from a technical perspective, I think it [Shuttle/Centaur]
would have flown.120
Those engineers who devoted years of their lives to readying Centaur for the Shuttle were
part of a culture with a tolerance for risk. They knew their rocket and they believed the mission
would succeed. To the engineers involved in Shuttle/Centaur, the rocket represented hope for
a new era of space exploration. Their enthusiasm did not blind them to safety problems, but
they believed that they had reduced the risk to manageable levels for mission success and pilot
safety. These engineers also correctly pointed out that the human risk factor should never have
entered the equation in the first place. Joe Nieberding concluded, These missions should have
flown on an expendable rocket. They needed no human crew. Again, the decision to fly every-
thing on STS was a disaster.121 Historian William Burrows agreed with this assessment of the
tragic loss of life in the deployment of a satellite that could easily have been launched with an
expendable rocket.122
119
Ibid.
120
Interview with Richard Kohrs by Mark Bowles, 13 July 2001.
121
Joe Nieberding correspondence with authors, 7 November 2001.
122
William E. Burrows, This New Ocean: The Story of the First Space Age (New York: Random House, 1998), 556.
chapter 7 3/9/05 10:52 AM Page 219
In 1990, a NASA advisory committee produced a report on the future of the United States
space program and discussed the significance of risk for the Agency. The committee concluded
that risk has always been a central feature of all of the greatest human adventures. When
Magellan first circumnavigated Earth in 1519, he started his voyage with five ships and a crew
of 280. After the three-year voyage, only one ship and thirty-four crewmen returned. Early test
pilots in the 1950s faced a similar risk factor while pushing the limits of supersonic flight. The
committee wrote, Risk and sacrifice are seen to be constant features of the American experi-
ence. There is a national heritage of risk taking handed down from early explorers . . . it is this
element of our national character that is the wellspring of the U.S. space program.123
Despite this bold statement of our adventurous heritage, NASA worried that the spark of
adventure is flickering. As a nation, we are becoming risk averse.124 NASA knew this from
experience. Risk aversion had kept them from launching Shuttle/Centaur. Richard H. Truly,
Associate Administrator for space flight and charged with the responsibility to return the Space
Shuttle safely to flight, wrote, I know that the business of space flight can never be made to
be totally risk-free, but this conservative return to operations will continue . . . .125 But risk is
a subjective, ever-present force that waxes and wanes in the minds of all who step into a NASA
program. NASA and America would recover from the Challenger tragedy. They would never
forget their fallen colleagues, but they would move on and again accept the risk inherent in the
exploration of space. Left behind was Shuttle/Centaur. It would forever remain an untested
dream, eclipsed by a tragedy not of its own making. Marty Winkler expressed the unease that
settled over the Centaur program in the wake of the Shuttle/Centaur cancellation. He said,
We were absolutely consumed through the early to mid-80s with Shuttle/Centaur. It sucked
up all our energy. And when the program was canceled, we said, What do we do now?126
123
Report of the Advisory Committee on the Future of the U.S. Space Program (NASA publication, December 1990), 1617.
124
Ibid.
125
Richard H. Truly to NASA distribution, Strategy for Safely Returning the Space Shuttle to Flight Status, 24 March 1986,
in Exploring the Unknown, Volume IV, ed. John M. Logsdon (NASA SP 4407, 1999), 378.
126
chapter 7 3/9/05 10:52 AM Page 220
31276-chapter 8 3/9/05 11:24 AM Page 221
Like a Phoenix 221
Chapter 8
Like a Phoenix
The sale of one commercial launch by a United States company is equiv-
alent to the import of ten thousand Toyotas.
Congressman Bill Nelson
Like a phoenix, Centaur was reborn in the 1980s and would remain a dependable and
important upper stage launch vehicle into the new millennium. The source of this rebirth
came from the commercialization of expendable launch vehicles. Just as it had survived the end
of the Surveyor program in the 1960s and at the cancellation of the expendable launch vehicle
program at the end of the 1970s, Atlas-Centaur emerged once again after the loss of
Challenger. The impetus came this time from the communications satellite industry and the
military.
In hindsight, it appears that NASAs decision in the 1970s to phase out all expendable
launch vehicles in favor of the Shuttle was seriously flawed. This decision made it easier for
Arianespace, a corporation backed by a consortium of European governments, to break the
monopoly over space transportation that NASA had enjoyed since the 1960s. In the early
1980s, Arianespace became NASAs competitor for commercial satellite business, and other
countries like China, Russia, and Japan followed somewhat later. The Europeans correctly
predicted that this market would grow in the 1980s and 1990s and that the Shuttle would not
be able to keep up with the demands of this market.1 Ariane had a competitive edge. Because
the new Arianespace launch facilities in Kourou, French Guyana, were closer to the equator,
Ariane required a shorter trajectory to reach orbit than rockets launched from Cape Canaveral.
Moreover, unlike launch facilities at Cape Canaveral, which had been allowed to deteriorate in
anticipation of closing down expendable launch vehicle production, they were new and
designed to serve the customer.
The reusable Shuttle was originally pitched to Congress as more economical than expend-
able launch vehicles. By flying the Shuttle often, it was thought that the cost per launch would
1
See Andrew J. Butrica, ed., Beyond the Ionosphere: Fifty Years of Satellite Communication (NASA SP-4217, 1997), especially
Lorenza Sebesta, U.S.-European Relations and the Decision to Build Ariane, the European Launch Vehicle, 137156. See also
Bruce D. Berkowitz, Energizing the Space Launch Industry, Issues in Science and Technology 6 (Winter 19891990): 7783.
31276-chapter 8 3/9/05 11:24 AM Page 222
be significantly lower than that of an expendable launch vehicle that is used only once. The
Shuttle, however, proved to be vastly more expensive than predicted and capable of far fewer
flights per year. Moreover, a two-stage expendable rocket like Atlas-Centaur (or Ariane) is
vastly more efficient than the Shuttle because it can carry a satellite directly to a geosynchro-
nous transfer orbit. After separation, the satellite motor pushes it a relatively short distance to
the desired geosynchronous orbit, where it moves at the same speed as the rotating Earth. By
contrast, because the Shuttle is limited to low-Earth orbit (about 200 miles from Earth), a
satellite launched from the Shuttle bay needs to use more of its own fuel to reach orbit, thus
shortening its life as a communications relay.
Despite these significant drawbacks, NASA had pressed forward with its plans to make the
Shuttle the countrys only space transportation system. All three expendable launch vehicle
manufacturers were in the process of closing down production. Termination of the Atlas-
Centaur program was forestalled in 1980 when Intelsat, concerned over delays in Shuttle
development, insisted on being able to use Atlas-Centaur for at least four more launches.2
Advocates of Atlas-Centaur urged NASA to articulate a formal policy stating that it was in the
national interest to allow expendable rockets to coexist with the Space Shuttle.3 In the early
1980s, the Reagan administration took the first tentative steps toward the commercialization
of expendable launch vehicles by setting up an Office of Commercial Space Transportation.
However, expendable launch vehicle manufacturers found it impossible to compete against the
highly subsidized Shuttle and the European Ariane rocket until the loss of Challenger precipi-
tated a change in national space policy.
Global Communications Satellites: Intelsat, Comstar, FLTSATCOM
Although science fiction writer and visionary Arthur C. Clarke had imagined an impor-
tant role for communications satellites as early as 1945, he had tempered this vision by calling
attention to a technical hurdle that many thought might scuttle the dream of an artificial satel-
lite communications systemthe lack of a rocket powerful enough to launch a satellite into
geosynchronous orbit.4 Despite his grand hopes for satellite communications, he thought it
would be many decades before the world would have a rocket with sufficient power to break
the bonds of gravity. Twelve years later, the Soviet Union launched Sputnik, pushing the
United States to develop a new satellite communications industry that would dominate into
the next century.
2
Report of the Atlas/Centaur Review Board, February 1981, NASA Glenn Records.
3
Ibid.
4
Arthur C. Clarke, Extra-Terrestrial Relays: Can Rocket Stations Give World-Wide Radio Coverage? Wireless World (October
1945): 305.
31276-chapter 8 3/9/05 11:24 AM Page 223
Like a Phoenix 223
The commercial satellite industry came into being when Congress passed the
Communications Satellite Act in 1962 and created the quasi-commercial American
Communications Satellite Corporation, or Comsat.5 Several large American telecommunica-
tions companies such as AT&T equally owned Comsat, the corporate entity that managed the
emerging global satellite system for the International Telecommunications Satellite
Organization (Intelsat), organized in 1964. The system was designed to carry voice, telegraph,
data, facsimile, and television transmissions. Although Intelsat was an international organiza-
tion, it was heavily weighted to favor American interests.
Date Mission Vehicle Payload Weight to Result

Synchronous Orbit*
25 Jan. 1971 Intelsat IV AC-25 1,600 lbs Success
19 Dec. 1971 Intelsat IV AC-26 1,600 lbs Success
22 Jan. 1972 Intelsat IV AC-28 1,600 lbs Success
13 June 1972 Intelsat IV AC-29 1,600 lbs Success
23 Aug. 1973 Intelsat IV AC-31 1,600 lbs Success
21 Nov. 1974 Intelsat IV AC-32 1,600 lbs Success
20 Feb. 1975 Intelsat IV AC-33 1,600 lbs FailedAtlas
22 May 1975 Intelsat IV AC-35 1,600 lbs Success
25 Sept. 1975 Intelsat IVA AC-36 1,820 lbs Success
29 Jan. 1976 Intelsat IVA AC-37 1,820 lbs Success
13 May 1976 Comstar I AC-38 1,787 lbs Success
22 July 1976 Comstar I AC-40 1,787 lbs Success
26 May 1977 Intelsat IVA AC-39 1,820 lbs Success
29 Sept. 1977 Intelsat IVA AC-43 1,820 lbs FailedAtlas
6 Jan. 1978 Intelsat IVA AC-46 1,820 lbs Success
9 Feb. 1978 FLTSATCOM F1 AC-44 2,250 lbs Success
31 Mar. 1978 Intelsat IVA AC-48 1,820 lbs Success
29 June 1978 Comstar I AC-41 1,787 lbs Success
4 May 1979 FLTSATCOM F2 AC-47 2,250 lbs Success
17 Jan. 1980 FLTSATCOM F3 AC-49 2,250 lbs Success
30 Oct. 1980 FLTSATCOM F4 AC-57 2,250 lbs Success
6 Dec. 1980 Intelsat V AC-54 2,280 lbs Success
21 Feb. 1981 Comstar I AC-42 1,787 lbs Success
15 Dec. 1981 Intelsat V AC-55 2,280 lbs Success
23 May 1981 Intelsat V AC-56 2,280 lbs Success
6 Aug. 1981 FLTSATCOM F5 AC-59 2,250 lbs Failed
4 Mar. 1982 Intelsat V AC-58 2,280 lbs Success
28 Sept. 1982 Intelsat V AC-60 2,280 lbs Success
19 May 1983 Intelsat V AC-61 2,280 lbs Success
9 June 1984 Intelsat V AC-62 2,280 lbs FailedCentaur
5
Public Law 87-624, 31 August 1962, as found in Exploring the Unknown, Volume III, ed. John Logsdon (NASA SP-4407,
1998), 77. See also Hugh R. Slotten, Satellite Communications, Globalization, and the Cold War, Technology and Culture 43
(2002): 315350.
31276-chapter 8 3/9/05 11:24 AM Page 224
Date Mission Vehicle Payload Weight to Result

Synchronous Orbit*
22 Mar. 1985 Intelsat VA AC-63 2,420 lbs Success
29 June 1985 Intelsat VA AC-64 2,420 lbs Success
28 Sept. 1985 Intelsat VA AC-65 2,420 lbs Success
4 Dec. 1986 FLTSATCOM F7 AC-66 2,700 lbs Success
26 Mar. 1987 FLTSATCOM F6 AC-67 2,250 lbs Failedlightning
25 Sept. 1989 FLTSATCOM F8 AC-68 2,700 lbs Success
*The actual weight carried by Centaur to transfer orbit was approximately double these numbers.
Because of the rapid development of satellite technology, communications satellites needed

to be redesigned every four to five years.6 The Delta rocket, manufactured by the Douglas
Aircraft Corporation, launched the Intelsat satellites of the 1960s. However, as satellites increased
in weight from 85 pounds to 330 pounds, they reached the limits of Deltas ability to lift them.
The jump from Intelsat III to IV was a dramatic fivefold increase in weight from 330 pounds to
1,600 pounds sent to geosynchronous orbit. Centaur was the only upper stage available to lift
this fourth generation of Intelsat satellites in 1971. By 1985, Atlas-Centaur had launched
twenty-four Intelsat satellites with only three failures.
Driven by the popularity of telecommunications, the demand for satellites continued to
grow. The weight of the next generation of satellites, Intelsat IVA, increased to 1,820 pounds,
with a communications capability twice that of its predecessor. This new Intelsat was launched
by an upgraded Atlas-Centaur D-1AR. The new features of Centaur, the result of knowledge
gained from the Titan-Centaur missions, included a computer-controlled vent and pressurization
system to control tank pressures, new six-pound thrust attitude control engines, and a redundant
hydrogen peroxide supply. The payload fairing was redesigned to accommodate the lengthened
Intelsat IVA.7 Both the upgraded Centaur and the new Intelsat worked flawlessly. The next series,
Intelsat V, developed by the Ford Aerospace and Communications Corporation, promised an
even greater capacity for global communications because of its improved relay capacity.8
The commercial success of Intelsat spurred the development of a series of four commercial
satellites called Comstar, also launched between 1976 and 1981 on Atlas-Centaur. Roughly the
same size and design as the Intelsat IV, Comstar satellites had a longer life because of more effi-
cient solar cells.9 Comsat General, a subsidiary of the Comsat Corporation, owned the Comstar
6
Martin, Intelsat V, Communication Satellites, 62.
7
Launch Vehicles Directorate, Lewis Research Center, AC-36 Flight Data Report, March 1976, Glenn Research Center, Box
Flight Reports and Historical Data AC-33 to AC-41, Division Atlas/Centaur Project Office, NASA GRC Records.
8
For more information on the Intelsat V launches, see Flight Reports from Launch Vehicles Division, NASA GRC Records.
9
See Flight Reports and Historical Data AC-33 to AC-41 and Flight Reports AC-42 to AC-49, Division Atlas/Centaur
31276-chapter 8 3/9/05 11:24 AM Page 225
Like a Phoenix 225
satellites. Comsat leased the satellites to AT&T; this meant that GTE Satellite Corporation
(GSAT), a subsidiary of General Telephone and Electronics, could also use them as a relay
between their networks of ground stations. AT&T had ground stations in New York, Chicago,
San Francisco, and Atlanta, while GSAT served Tampa, Los Angeles, and Hawaii. Comstar
provided increased long-distance service in the United States without increasing the number of
telephone poles, overhead wires, or microwave towers.10
The military also depended on Atlas-Centaur to deliver heavy satellites to orbit. A series
of eight satellites, the Fleet Satellite Communications (FLTSATCOM) system, designed and
built by TRW Systems, served the space communications network of the Navy, the Air Force,
and the Department of Defense.11 It enabled ships, aircraft, submarines, and ground stations
to communicate effectively with each other.
Cost benefits to NASA were significant. Because national policy precluded satellite
companies from contracting directly with the launch vehicle manufacturers for launch services,
NASA, which owned the vehicles, could charge the makers of commercial satellites for the use
of its vehicles, launch facilities, and services. In 1970, the cost of an Atlas-Centaur launch was
fourteen million dollars.12 By 1980, the cost of launching one Comstar satellite had climbed
to over twenty-three million dollars. Only about half of this amount came from the launch
vehicle hardware. These hardware costs included $4,368,502 for the Centaur stage, $777,058
for the Centaur engines, $2,114,065 for Atlas-Centaur hardware support, $2,563,133 for the
Atlas airframe, $1,156,481 for the Atlas engines, $460,701 for guidance hardware, $675,172
for computer hardware, and $305,597 for government-furnished property and services. The
other half came from various government support services, including $2,802,724 for contract
launch services, $3,852,491 for contract support services, $556,816 for miscellaneous
program support, $151,671 for DOD contract administration, $1,519,000 for Air Force range
support, $785,322 for NASA project management and engineering, $973,028 for Agency
overhead, and $615,000 for use charges.13 Often it would take NASA up to three years to bill
companies for these additional costs. NASA anticipated that it could use revenues from satel-
lite launches to help offset the enormous costs of a Shuttle launch, once all satellite launches
were transferred from expendable launch vehicles to the Shuttle.
Launch Mission Summary Comstar D-3 AC-41, 29 June 1978, Box FLTSATCOM 19781983, Division STED, NASA
10
GRC Records.
11
For more information on the FLTSATCOM launches, see Flight Reports and Historical Data, NASA GRC Records.
12
Joan Lisa Bromberg, NASA and the Space Industry (Baltimore: The Johns Hopkins University Press, 1999), 111.
13
Charles J. Tiede to Chief, Launch Vehicles Division, Final Government Costs for the Third Comstar I Launch, 9 April 1980,
Box FLTSATCOM 19781983, Division STED, NASA GRC Records.
31276-chapter 8 3/9/05 11:24 AM Page 226
Old Centaur, New Ariane
The early 1980s proved to be an extraordinarily demanding time for Atlas-Centaur

managers as they juggled the competing demands of industry, military, and commercial
launches. For example, Lewis engineer James Patterson wrote to NASA Headquarters, You
will note that the Comstar vehicle could be launched 5 February 1981 from Pad A only if
Intelsat would be willing to slip their AC-55 launch from 26 February to 5 March 1980, and
their AC-58 launch from 8 May to 4 June 1981. The new proposed schedule also takes into
account the Navys request to move up the AC-57 FLTSATCOM launch from 13 November
to 28 October 1980.14
Despite this heavy use of Atlas-Centaur, Lewis Research Center began to prepare for the
termination of all engineering activities associated with its management in anticipation of the
Shuttles taking over its payloads. Termination of the program was forestalled when Intelsat,
concerned over delays in Shuttle development, insisted on being able to contract with General
Dynamics for four additional Centaurs to be launched prior to 1984. At the same time,
NASA, recognizing the limitations of the Shuttle, began to consider using Centaur as an upper
stage for the Shuttle (as discussed previously).
This small glimmer of hope in the otherwise gloomy outlook for the Launch Vehicles
Division at Lewis permitted Centaur supporters to mount a campaign to keep Atlas-Centaur
alive. NASA permitted an Atlas-Centaur Review Board to be convened at Lewis Research
Center in October 1980. The blue ribbon committee was composed of representatives from
Goddard, Langley, JPL, and the Air Force Space Division. Two significant individuals in the
history of Centaur, retired Lewis Center Director Bruce Lundin and former Deputy Associate
Administrator for Space Science Vincent Johnson, were called out of retirement to serve as co-
chairs. The final report of the committee, submitted February 1981, strongly recommended
that Atlas-Centaur be continued as a program into the foreseeable future.15
The board emphasized the reliability of Centaur and the soundness of Lewiss manage-
ment of the program over a period of nearly two decades. The technical penetration of Lewis
staff included supervision of both General Dynamics, the vehicles integrating contractor, and
associate contractors Honeywell, Teledyne, and Pratt & Whitney. The report recommended
against allowing the associate contractors to report to General Dynamics directly because
involvement of the Lewis project staff was invaluable: This supplies both a corporate
memory and person-to-person familiarity with their counterparts among contractors.16
Despite the decade-long threat that the program would be phased out, morale at Lewis and
14
James E. Patterson to NASA Headquarters, AC-42 Launch Schedule, 5 March 1980, Box FLTSATCOM 19781983,
Division STED, NASA GRC Records.
15
Report of the Atlas/Centaur Review Board, February 1981, 1, Box 10, NASA GRC Records.
16
Ibid., 5, Box 10, NASA GRC Records.
31276-chapter 8 3/9/05 11:24 AM Page 227
Like a Phoenix 227
among the industry contractors remained high. They were a true government-industry team
that supports, stimulates, and challenges the other to produce a whole that [was] greater than
the sum of the parts.17 However, the board warned that Centaur reliability could be affected
by a decrease in personnel assigned to Centaur and other cost-cutting measures. It was espe-
cially important to keep Lewis Research Center in the picture because of its expertise in
software design and verificationan area that General Dynamics had been unable to staff
adequately because of imminent phase-out. Another reason was the significant Lewis Research
Center role in postflight analysis. Because of their long-term involvement, Lewis engineers had
become particularly skilled at uncovering subtle anomalies in test data.
However, there were also significant drawbacks to keeping Atlas-Centaur flying. The
board criticized Lewis management for casualness born of familiarity that was evident in
preflight checkout. Another shortcoming of the Launch Vehicles Division was its adversarial
relationship with Headquarters, a result of the protracted struggle to keep the program from
being terminated. The report noted the loss of skilled people both at Lewis and among the
contractors. At General Dynamics, the welding of pressure-stabilized tanks for both Atlas and
Centaur was a highly specialized skill that few technicians possessed. The same was true of the
special expertise needed to manufacture nose fairings and insulation panels. Investigation
revealed that an increasing number of workmanship discrepancies had occurred since 1978,
possibly due to the perception of the pending discontinuation of the program. At Honeywell,
technical know-how had declined, along with the reliability of worn-out test equipment. In
response to this criticism, the head of the Lewis Launch Vehicles Division wrote, This situa-
tion demonstrates the classical dilemma of a program limping along on small size buys. At
this rate our 1960 vintage test equipment will be seeing action cum bailing wire until it dies
of exhaustion!18
Even more worrisome was the loss of Pratt & Whitneys expertise in hydrogen/oxygen
technology. Staffing at the company had dropped from a level of about 150 engineers in 1967
to just 9. By 1982, the company expected to have only three engineers assigned to the RL10
engine. The technical know-how involved in fabricating thrust chambers, understanding the
design and performance margins of the engine, predicting the behavior of cryogenic propel-
lants for cooling thrust chambers, and manufacturing turbopump bearings and injectors could
never be replaced.19 Another area of concern was the anticipated retirement of long-time super-
visors, technicians, and inspectors at the Eastern Space and Missiles Center at Cape Canaveral.
The ground-support infrastructure had been allowed to deteriorate, with worn-out equipment
17
Ibid., 1, Box 10, NASA GRC Records.
18
Memo from Larry Ross, Director of Space, to Center Director John F. McCarthy, Jr., 4 June 1981, Box 10, NASA GRC
Records.
19
Report of the Atlas/Centaur Review Board, February 1981, 14, NASA GRC Records.
31276-chapter 8 3/9/05 11:24 AM Page 228
being repaired instead of replaced. The board recommended that NASA adopt a more flexible
system of personnel management that would allow staff to be shifted between the Shuttle and
expendable launch vehicle operations. NASA was urged to articulate a formal policy stating
that it was in the national interest to allow expendable rockets to coexist with the Space
Shuttle, with specific reference to the Atlas-Centaur launch vehicle as a necessary element of
the U.S. Space Transportation System into the indefinite future.20
It must have come as a relief to the Launch Vehicles Division that a proposed reduction
of the number of engineers from 123 to 65 might be averted. In his formal response, Larry
Ross, Director of the Space Division, outlined steps for implementing the report recommen-
dations. He believed that most of the divisions shortcomings noted in the report reflected a
going-out-of-business mindset.21 Predictions of the last launch throughout the previous
decade had always proved to be mistaken. Although the future of Centaur remained in limbo,
he recommended that the Launch Vehicles Division go resolutely forward with planning for
future launches.
During the same decade that Atlas-Centaur was limping along on outdated rocket tech-
nology, a brand-new rocket was being developed by a consortium of European countries.
Motivation for financing the building of Ariane differed from country to country; for the
French, it was a political decision driven by French nationalism. France contributed about 60
percent of the development costs and wanted to be able to launch its own military satellites
independently of the United States. The fact that Ariane proved to be an outstanding commer-
cial success came unexpectedly as a result of NASAs policy blunders.22
Ariane, comparable to Atlas-Centaur in power and cost per launch, also had a cryogenic
upper stage. The development of this stage appears to have proved both difficult and contro-
versial, since the United States regarded cryogenic know-how as an issue of national security.23
Despite NASAs skepticism that the Europeans were capable of developing a launch tech-
nology, the European Space Agency pressed forward with development of the new launch
vehicle. When Intelsat purchased a ride for an Intelsat V satellite on Ariane in 1978, it was
hailed by the Europeans as a signal that Europe, notwithstanding the threat of the Shuttle,
20
Ibid., 2.
21
Memo from Larry Ross, Director of Space, to Center Director John F. McCarthy, Jr., 4 June 1981, Box 10, NASA GRC
Records.
John Krige, The Decision Taken in the Early 1970s To Develop an Expendable European Heavy Satellite Launcher, in A
22
History of the European Space Agency, 19581987, vol. 2 (ESA SP-1235, April 2000), 389421.
23
See references in Krige, 413, 421, and L. Sebesta, The Availability of American Launchers and Europes Decision To Go It
Alone, History of ESA, vol. 2, 437.
31276-chapter 8 3/9/05 11:24 AM Page 229
Like a Phoenix 229
was now perceived as having a viable, alternative launch service which was going to be compet-
itive on the international launcher market.24
The Contradictions of Shuttle Policy
Although President Reagan announced in 1982 that the government would no longer
order Titan, Atlas, or Delta launch vehicles, his administration began to work behind the
scenes to prepare legislation to commercialize the American expendable launch vehicle
industry. An advocate of laissez-faire economics, Reagan championed competition and may
have believed that, once liberated from its dependence on government contracting, the
expendable launch vehicle industry would be able to win commercial satellite business. The
Presidents Commercial Launch Directive (NSDD 94) set up a new Office of Commercial
Space Transportation within the Department of Transportation. Secretary of Transportation
Elizabeth Dole organized the Commercial Space Transportation Advisory Committee
(COMSTAC) to facilitate communication between the industry and her new office.25
COMSTAC included representatives of the big three launch vehicle manufacturersMartin
Marietta, General Dynamics, and McDonnell Douglasand about a dozen startup companies
seeking to get into the business of small expendable launch vehicles, the one market not served
by the Shuttle. The charge to COMSTAC was to assist the government in drawing up the
Commercial Space Launch Act of 1984.26
The Space Systems Division at General Dynamics now began to take the first tentative
steps toward commercialization. In July 1981, Alan Lovelace, who had served as acting
Administrator for NASA during the Carter administration, had become Corporate Vice-pres-
ident and General Manager of the Space Systems Division. He had great credibility within the
aerospace community because of the many years he had spent as head of the Materials
Laboratory at Wright-Patterson Air Force Base, Deputy Chief of Staff for R&D for the Air
Force Systems Command at Andrews Air Force Base, and then Deputy Assistant Secretary for
R&D at the Pentagon. Lovelace recalled that as acting Administrator of NASA, he had
strongly believed the Shuttle capable of handling all of NASAs missions, and in this role, he
was responsible for informing Lewis Research Center of the imminent cancellation of NASAs
Expendable Launch Vehicle Programs.
After he joined General Dynamics, however, Lovelace began to see new opportunities for
commercial launch vehicles. In his view, NASAs monopoly of launch services had created a
business of arrogance. A commercial launch service offered customers more control over
24
J. Krige, The Move from Ariane Development of Production and the Establishment of Arianespace, History of ESA, vol. 2,
472.
25
John M. Logsdon and Craig Reed, Commercializing Space Transportation, Exploring the Unknown, Volume IV, Accessing
Space (NASA SP-4407, 1999), 411.
26
For an insightful account of the entrepreneurial activity stimulated by the new legislation, see Bromberg, 114131.
31276-chapter 8 3/9/05 11:24 AM Page 230
their payloads. He strongly believed that if General Dynamics were allowed to sell a launch
service directly to the customer, it could manage indirect overhead and become competitive
with Ariane.27 Standing in the way of commercialization was a government contracting envi-
ronment that stifled initiative. In his view, the cost-plus mentality had to go.28
A 1982 General Dynamics study determined that the company could anticipate an
expanding market for commercial rockets. Liberated from dependency on government
contracts, the company looked forward to exerting greater control over marketing, design, and
production. The study speculated that Atlas-Centaur had the potential to tap 90 percent of
this market by upgrading Centaur to reduce costs and increase performance. Systems proposed
for modernization included the Centaur avionics, nose fairing, and hydraulics system. The
Atlas vernier propulsion system was to be replaced with a simple hydrazine roll control module
mounted in the Interstage Adapter between Atlas and Centaur. This change would reduce cost
and improve reliability.29
In anticipation of commercialization, General Dynamics formed a new commercial entity
called CommSpace in March 1983 to market the projected $10-billion international commer-
cial satellite industry. The company promoted the new commercial launch service as a backup
launch capability for the Space Shuttlea capability for which the government would incur
no additional costs. General Dynamics offered to assume the risk of a failed launch. It would
maintain existing launch facilities in danger of becoming obsolete. The company anticipated
that the new commercial initiative would contribute significantly to the U.S. economy.
One of the prime considerations in the move toward commercialization was the possibility
of realizing the economies of scale inherent in continuous production. NASAs procurement of
Centaur rockets one or two at a time since the end of the Surveyor program had made it diffi-
cult to maintain manufacturing expertise.30 For example, because welding the Atlas and
Centaur tanks was highly skilled labor, it took at least three months to get welders retrained
after a layoff. Training new workers in the field of electronics was an even greater problem.31
Quality suffered when well-trained engineers and production workers were spirited away by
companies with better long-term prospects.
27
Interview with Alan Lovelace by Virginia Dawson, 19 July 2002.
28
Ibid.
29
D. E. Charhut and J. E. Niesley, The Commercial Centaur Family, paper to 34th Congress of the International Astronautical
Federation, IAA-83-233, 1015 October 1983. See also W. F. Rector III and D. E. Charhut, The Commercial Centaur Family,
Convair Division, General Dynamics Corp., 1983; M. C. Simon and O. Steinbronn, The Economics of Space Development,
General Dynamics, 1984, Box 57, NASA GRC Records.
30
In the decade of the 1970s, there were only two large contracts: one for nine Centaur stages, dated 24 September 1973, and
one for eight stages, dated 8 September 1976. See NASA Historical Data Book, vol. III (NASA SP-4012, 1988), 2526.
31
Interview with Stanley Chamberlain by Virginia Dawson, 7 June 2000.
31276-chapter 8 3/9/05 11:24 AM Page 231
Like a Phoenix 231
Although the Reagan administration encouraged commercialization in principle, manu-

facturers of expendable launch vehicles were effectively shut out of the market by the heavily
subsidized price of a Shuttle launch once the Shuttle became operational in 1982. As histo-
rians John M. Logsdon and Craig Reed have noted, The reality was that the United States
during the 198385 period was pursuing two policy goals that were clearly inconsistent.32
Commercialization was not possible as long as artificially low Shuttle pricing made it impos-
sible for the expendable launch vehicle manufacturers to compete for satellite business.
NASA demonstrated formal compliance with the 1984 Commercial Launch Directive by
issuing a request for proposals for commercializing Delta and Atlas-Centaur. Initial agreements
with Transpace Carriers (which marketed Thor-Delta for McDonnell Douglas) and General
Dynamics were reached in December 1983. They were to go into effect for Delta in 1986 and
for Atlas-Centaur in 1987. Also, in September 1984, NASA set up an Office of Commercial
Programs focused largely on selling Delta launch services. The budget of this office for its first
year was a mere $17.1 million, about 0.2 percent of NASAs total budget.33
In the policy debates over the implementation of the 1984 Space Commercialization Act,
members of Congress and representatives of the expendable launch vehicle industry argued
that Shuttle pricing should at least approximate actual costs. NASA resisted with the rationale
that a low Shuttle price was needed to remain competitive with Ariane, which was subsidized
by the European governments. The price question pitted the Department of Transportation,
which represented the manufacturers of expendable launch vehicles, against both NASA and
the satellite makers, who wanted space transportation offered at the lowest possible cost.
Clevelands local paper, the Plain Dealer, chided NASA for paying lip-service to space commer-
cialization while working behind the scenes to maintain its monopoly. Since the government
contributed between half and three-quarters of the actual cost of a commercial Shuttle launch,
expendable launch vehicles could not compete with the Shuttle.34 As the Wall Street Journal
reported in March 1985, The shuttle price question has turned into a noisy bureaucratic trial
of strength between NASA and the Transportation Department, and will, in the end, be
decided by President Reagan himself.35 Rudolph Penner, Director of the Congressional
Budget Office, predicted that pricing would determine what gets done in space, how it is
32
Logsdon and Reed, Commercializing Space Transportation, Exploring the Unknown, Volume IV, 412.
33
Joan Lisa Bromberg, NASA and the Space Industry, 122.
34
Patrick Cox, What Price the Shuttle? Plain Dealer 3 (February 1985): Section B, 1.
35
Arlen J. Large, Price Policy on Space Shuttles Commercial Use Could Launchor GroundNASAs Rockets, Wall Street
Journal (21 March 1985): 54.
31276-chapter 8 3/9/05 11:24 AM Page 232
done and the fate of competing means of space transportation.36 Ariane had a growing market
share, and Roland Deschamps, Secretary General of Arianespace, could gloat: The big error
of NASA was to abandon the expendable launch vehicles to private companies and to compete
with them through Shuttle.37
At the same time, the Department of Defense questioned NASAs insistence that all
payloads, both civil and military, be sent aloft on the Shuttle. Concerned that a Shuttle failure
might delay sensitive military payloads, the Air Force insisted on keeping at least one expend-
able launch vehicle manufacturer in business.38 During congressional hearings, Administrator
James Beggs resisted the Air Force initiative, arguing that expendable launch vehicles were
obsolete and that support of the expendable launch vehicle industry would undermine the
cost-effectiveness of the Shuttle. In the spring of 1984, an Air Force contract for ten upgraded
Titan IV rockets was the first indication that at least the Martin Marietta rocket business
might survive.39
For General Dynamics, the future was more problematic. Prior to 1984, Centaur had enjoyed
its most successful record of launchesan impressive string of twenty-eight in a row without a
failure. The magic was broken when the first flight of an improved Atlas G (AC-62, 81 inches
longer than the previous Atlas), mated to an upgraded Centaur vehicle, failed.40 The failure inves-
tigation kept the next three Atlas-Centaur launches on hold. Investigations completed in August
1985 concluded that the failure was caused by a liquid-oxygen leak through a 4-inch crack in the
intermediate bulkhead, which occurred during Centaurs separation from the booster.41 The
company could ill afford this interruption in service because, by this time, Ariane had gained an
impressive 30 percent of the world market for satellite launchers.42
36
Ibid.
37
NASA Does Not Admit Its Error in Putting All Eggs on Shuttle, Interspace: The European Satellite & Space News (11 January 1985).
38
The role of the Air Force is discussed by Craig Reed in Factors Affecting U.S. Commercial Space Launch Industry
Competitiveness, Business and Economic History 27 (1998): 229; Ragsdale, 150.
39
See discussion by Joan Lisa Bromberg, 131132. See also previous note. (The Air Force ordered another twenty-five Titan IV
vehicles at a total cost of more than $4 billion after the Challenger accident.)
40
Mary Ann Peto and Roland Raab, Stretched Atlas-Centaur Will Boost Payload Capacity, 3 May 1984, NASA Historical
Reference Collection, Washington, DC. Improvements to the new silver-throated Centaur included a pressure-fed system to
replace the boost pumps, a new hydrazine attitude-control rocket system, and silver plating on the interior of the rocket nozzle
throat to increase thrust. The result was an extremely powerful unpiloted vehicle with the capability of placing an extra 500
pounds into geosynchronous orbit.
41
General Dynamics, Atlas-Centaur AC-62 Failure Investigation Interim Report, Report No. GDC-SP-84-045, November
1984, Box AC-62 Failure Investigation, Division Atlas-Centaur Project Office, NASA GRC Records; General Dynamics, Atlas-
Centaur AC-62 Failure Investigation Final Report, Report No. GDC-SP-85-018, August 1985, Box AC-62 Failure
Investigation, Division Atlas-Centaur Project Office, NASA GRC Records.
42
J. Krige, The Move from Ariane Development of Production and the Establishment of Arianespace, History of ESA,
vol. 2, 481.
31276-chapter 8 3/9/05 11:24 AM Page 233
Like a Phoenix 233
A suit by Transpace Carriers, Inc. (which marketed Delta for McDonnell Douglas),
claimed that Arianespace was unfairly and illegally engaging in predatory pricing of launch
services. This attack on Ariane pricing was indirectly an attack on Shuttle pricing, since both
were heavily government-subsidized. Even after raising the price of a Shuttle launch from $38
million to $71 million, this price was still far below the actual cost of a launch. U.S. pricing
policy on launchers in 1984/5, historian John Krige wrote, was thus torn between the funda-
mental contradiction of trying to protect the Shuttle from Ariane while encouraging private
industry to develop a viable commercial ELV service in parallel.43
After Challenger
The Challenger disaster in January 1986 precipitated a complete reversal of Shuttle policy
and opened the way for the development of a viable expendable launch vehicle industry in the
United States. In December 1986, President Reagan issued National Security Decision
Directive 254, stipulating that NASA would no longer launch satellites for private companies
or foreign governments, except those that required a human presence in space or involved
national security. This decision took the Shuttle out of competition with expendable launch
vehicles for revenue-generating business.
Analysis of the Shuttles manifest after the loss of Challenger showed that of a total of 186
payloads scheduled for launch between 1986 and 1989, 103 were appropriate to be flown on
expendable launch vehicles. Lee R. Scherer, director of the commercial space systems division
for General Dynamics, pointed out that the Shuttle pricing policy had precipitated the closing
down of the production lines for Delta, Atlas-Centaur, and Titan 34D rockets. Now the hiatus
in Shuttle flights left commercial satellite customers with only one optionthe Ariane rocket.
He concluded that the Challenger tragedy had created many uncertainties, but one thing was
clear: Our country must never again allow itself to get into the current posture where we have
been completely dependent on a single form of space transportation.44
NASAs Mixed Fleet Study, completed in May 1987, complemented the new policy. It
emphasized that in addition to the Shuttle, the nation needed expendable launch vehicles of
different sizes and capabilities. Richard Truly, Associate Administrator for Space Flight,
parceled out the management of NASAs expendable launch vehicles among NASA Centers.
Lewis Research Center received the management of intermediate- and large-class expendable
launch vehicles, namely Atlas-Centaur and Titan-Centaur. Goddard took over the manage-
ment of small- and medium-class expendables, such as the Delta vehicle, as well as payload
43
J. Krige, 483.
44
L. R. Scherer, Expendable Launch Vehicles as a Complement to the Space Shuttle, paper presented to the AIAA
Communications Satellite Systems Conference, March 1986. See also Keith Mordoff, ELV Makers Gear for Production
Restart, Commercial Space (Spring 1986): 4447
31276-chapter 8 3/9/05 11:24 AM Page 234
processing at Wallops Flight Facility. Kennedy Space Center was assigned the assessment of
expendable launch vehicle processing and launch countdown for launches involving NASA.45
The Challenger disaster not only changed NASA policy, but also raised questions over
NASAs managerial competence. While the problems related to the Challenger tragedy received
wide media coverage, there were Agency concerns for NASAs management of expendable
launch vehicles. An Atlas-Centaur carrying a FLTSATCOM was struck by lightning during a
driving rainstorm on 26 March 1987. Forty-eight seconds into launch, the Centaur computer
ordered an incorrect yaw steering command that caused the rocket and its payload to explode
50 seconds later. This failure in the wake of Challenger precipitated a congressional investiga-
tion. NASA revealed that the lightning strike had caused the computer to issue an erroneous
command.46 But the failure involved more than just the weather. The Centaur team at Lewis
Research Center was reprimanded for work practices that exhibited inadequate emphasis on
procedural rigor and discipline.47 The Centaur team admitted that the obvious indication of
a lightning hazard escaped the launch management team because of imprecise communica-
tions, lack of awareness, or both.48 Just six months before this accident, another NASA review
board had questioned procedures for making decisions that related to weather at Cape
Canaveral.49 Had the recommendations from this report been followed, NASA would have
waited for better weather to launch AC-67. Political cartoons satirized NASA weather fore-
casting, and one caption stated that even Ben Franklin would not have flown his kite during
such lightning conditions. An editorial in Aviation Week & Space Technology described the AC-
67 accident in the context of the previous years Challenger tragedy. The editor wondered,
What other time bombs are ticking away in the agencys hidebound management structure
that need to be rooted out to prevent further loss of life, limb, and property?50
The answer came four months later. During preparation for the last scheduled NASA
launch of an Atlas-Centaur, a workstand bounced against the empty but pressurized hydrogen
tank. It sliced a gash in the side of the tank, causing the vessel to rip violently apart. Several
Field Center Expendable Launch Vehicle (ELV) Roles, Responsibilities, and Organizational Structure, 17 December 1987,
45
Box 57, NASA GRC Records.
46
General Dynamics, Atlas-Centaur Postflight Analysis AC-67, Report No. GDSS-SP-87-011, July 1987, and
Determination, Findings, Observations, and Recommendations, report given before Congress, 30 June 1987, Box AC-67
Failure Investigation, Division 6500 John Gibb, NASA GRC Records.
Atlas-Centaur 68 Investigation Board Report on the 13 July 1987 Centaur Liquid Hydrogen Tank Mishap, Box AC-68,
47
Division Atlas-Centaur Project Office, NASA GRC Records.
48
Ibid.
49
Craig Covault, Atlas Accident Inquiry Board Finds Violation of Launch Commit Criteria, Aviation Week & Space Technology
(16 May 1987): 25.
50
Atlas Launch Follies, Aviation Week & Space Technology (18 May 1987): 11.
31276-chapter 8 3/9/05 11:24 AM Page 235
Like a Phoenix 235
workers sustained minor injuries. Investigation revealed a lack of discipline on the part of the
workers at Cape Canaveral, again attributed to poor NASA management.51
NASA was also criticized for the inefficiency of payload processing at the Cape and for
allowing the launch infrastructure to deteriorate. Investigation revealed that about half of the
seventy-seven expendable launch vehicle payloads launched between January 1977 and
February 1988 had experienced delays of an average of two weeks. Sixty-five percent of the
delays were caused by payload or launch vehicle problems, with nearly 30 percent of the
payloads requiring a storage period.52 It took longer to process vehicles for launch because the
Air Force controlled the launch pads, while the ground-support equipment, including mission
control and the tracking stations, were NASAs responsibility. The expendable launch vehicle
manufacturers and their satellite customers had to contend with NASA rules, practices, and
priorities, as well as those of the Air Force. Arianespace had the advantage of a streamlined
launch facility without any kind of government oversight.
To address these and other problems, the Air Force drew up a Model Range Use Agreement
and sought ways to keep the expendable launch vehicle manufacturers in business. In 1984, to
ensure that the military had access to space, the Air Force had contracted with Martin Marietta
for the ten Titan IV boosters. After Challenger, the Air Force increased this order to twenty-three
vehicles at a cost of $5.1 billion. Secretary of the Air Force Edward (Pete) Aldridge wanted to end
Air Force dependency on NASA by creating a robust commercial industry. The new contract
permitted Martin Marietta to market a commercial version of Titan III to ten satellite customers.
In the spring of 1987, the Air Force contracted with McDonnell Douglas for eleven Delta II
medium launch vehicles with an option for twenty additional vehicles. At this point, of the three
established manufacturers in the expendable launch vehicle business, only General Dynamics
lacked a significant new contract with the Air Force.53
Despite the loss of the medium launch vehicle competition to McDonnell Douglas, Alan
Lovelace forged ahead with a major financial commitment to commercialization. The cancel-
lation of Shuttle/Centaur after Challenger left General Dynamics with a cadre of experienced
Centaur people to initiate the commercialization effort.54 Lovelace convinced the companys
board of directors to finance the manufacture of eighteen Atlas-Centaur vehicles before they
actually had orders for them. The company granted the Space Systems Division $125 million
to start up the assembly line.
Atlas-Centaur 68 Investigation Board Report on the 13 July 1987 Centaur Liquid Hydrogen Tank Mishap, Box AC-68,
51
Division Atlas-Centaur Project Office, NASA GRC Records.
Creighton A. Terhune and Shirley P. Green, Payload Processing Study for ELV Launches, paper presented at the 26th Space
52
Congress, 1989, AEDC Tech Library, Arnold AFB, TN.
53
Ramon L. Lopez and Greg Waskul, New Life for Expendable Launchers, Space Markets (Spring 1987): 4.
54
31276-chapter 8 3/9/05 11:24 AM Page 236
In the fall of 1986, an Air Force Request for Proposal for twelve medium launch vehicles
once more pitted the three major expendable launch vehicle manufacturers, McDonnell
Douglas, Martin Marietta, and General Dynamics, against each other. The competition for
this Air Force contract created a make-or-break situation for General Dynamics. The second
Medium Launch Vehicle (MLV II) contract would give an edge to the winner over its competi-
tors because the military order would allow the company to leverage the Air Force contract
against the startup costs of commercial production.55 At this time, at the urging of the Air
Force, General Dynamics dropped the name Centaur when referring to its commercial vehicle.
In May 1988, Lovelace, who presided over the American Institute of Aeronautics and
Astronautics (AIAA) that year, found a note from Pete Aldridge on his plate when he returned
to his seat on the dais during the annual banquet. It simply said, Congratulations, you have
won the MLV II competition. This welcome news vindicated Lovelaces audacious decision to
start up production at company expense. The $340-million Air Force contract for ten launch
vehicles represented an important step toward underwriting commercialization of the entire
American expendable launch vehicle fleet. With commercialization, the contractor, not the
government, was responsible for the integration of the launch process. To make sure everyone
in the company focused on manufacturing a quality product, General Dynamics planned to
self-insure the vehicles, offering customers a free ride should the booster fail. This commit-
ment to a program with a value of more than $1 billion will create over 6,000 jobs for our
domestic work force, Lovelace stated in congressional testimony on 5 October 1987.56
Because of his NASA background, Lovelace understood the cultural changes NASA
needed to make in order to transition from purchasing hardware to purchasing launch services,
but he was convinced that commercialization would result not only in greater reliability, but
also in significant cost savings for NASA. The company planned to offer launch services to
customers, including NASA, at a fixed price. Lovelace insisted that it was in the best interest
of the government for NASA to become a customer of American companies in the launch
vehicle business. He stated to Congress, Advantages to the United States Government in
procuring launch services from a robust and competitive stable of domestic commercial launch
vehicles are numerous. Effectively the United States Government can enjoy the advantages of
a multiyear procurement and quantity buys with only limited financial obligation and risk.
Attendant lower cost, flexibility in scheduling, and resource management can all be realized.57
55
See Theresa Foley, U.S. Manufacturers Begin the Job of Rebuilding the U.S. Space Program: ELVs, Commercial Space (Fall
1986): 1621.
56
U.S. Senate Committee on Commerce, Science, and Transportation, Subcommittee on Science, Technology, and Space, 5
October 1987, 60.
57
Ibid., 62.
31276-chapter 8 3/9/05 11:24 AM Page 237
Like a Phoenix 237
NASA, however, resisted the idea of contracting for a launch service. The Commerce
Department, which operated the Geosynchronous Operational Environmental Satellite
(GOES) weather satellites, forced the issue. It insisted that NASA use a commercial vendor to
launch the three GOES weather satellites and, when NASA resisted, simply threatened to find
another procurement agent. GOES became NASAs first commercial contract and represented
a milestone in the effort to commercialize the launch vehicle industry.58
The Barter Agreement
One of the most important elements of the GOES contract was the successful negotiation
of a barter agreement to turn over government property and equipment in exchange for launch
services.59 Government property involved in this agreement included all NASA-owned equip-
mentgantries, storage tanks, bunkers, etcat Launch Pads 36A and B and test facilities, as
well as equipment owned by NASA and located on contractor property. Crafting the agree-
ment took enormous contracting creativity on the part of procurement officers at Lewis
because there was no precedent for turning over government property to the private sector at
fair market value rather than at cost. The launch pads were not part of the barter agreement
because they were still owned by the Air Force, which leased them to General Dynamics. But
complicating the determination of value of NASA-owned facilities and equipment at the Cape
was the fact that they had been allowed to deteriorate in anticipation of cancellation of the
program. They were worth far less than the governments initial investment. Another compli-
cating factor was the unusual method of Lewis contracting. When Silverstein had taken over
the Centaur program in 1962, he had insisted that General Dynamics be treated as an asso-
ciate contractor. Contractors for subsystems like Pratt & Whitney and Honeywell had all
contracted directly with Lewis. By keeping Lewis at the center of all aspects of vehicle manage-
ment, the government was able to control costs and leverage its considerable technical
expertise. However, this model for contracting made crafting the barter agreement more
complicated. In 1987, instead of managing a single contract with General Dynamics as a
prime contractor, Lewis was managing about fifty-nine major contracts, valued in excess of
$1.5 billion.
The termination of the Shuttle/Centaur program in 1986 and the commercialization of
Atlas-Centaur called for the disposition of about $360 million in government-owned property
that remained in contractor hands.60 This included vehicle components, special test equip-
58
Joan Lisa Bromberg, NASA and the Space Industry, 158159.
59
Interview with Ronald Everett by Virginia Dawson, 17 December 2001.
Memo, Report on Audit of the Disposition of Atlas and Centaur Property, A-LE-88-002, 18 November 1988, part of the
60
Audit Report, Office of Inspector General, files of Ronald Everett, NASA GRC.
31276-chapter 8 3/9/05 11:24 AM Page 238
ment, tooling, ground-support equipment, raw materials, and partially finished work at the
time of cancellation. NASA had to find a way to place a value on the government property
involved in these contracts. Once General Dynamics owned this property, the company could
legally offer the entire Atlas-Centaur package as a launch service to commercial satellite
customers and NASA.
General Dynamics agreed to rehabilitate and provide launch services for AC-68, damaged
extensively in July 1987 by the careless workstand accident. The company also agreed to
supply the entire vehicle and launch services for the Combined Release and Radiation Effects
Satellite (CRRES) mission (AC-69), originally scheduled for flight on the Shuttle. In
exchange, General Dynamics received the government-owned property from Honeywell, Pratt
& Whitney, Teledyne, and Rocketdyne, as well as the legal rights to NASA-owned property at
company headquarters and at the Cape. It also received rights to the vehicle design. Thanks to
the barter agreement, General Dynamics could proceed with the upgrade of the launch pads
and changes to the vehicle design to lower costs and increase reliability. This was a major finan-
cial undertaking, but one that General Dynamics was willing to finance. One of the companys
immediate priorities was to reduce the inefficiencies that had crept into launch practices over
a period of thirty years.
A significant problem remained. The loss of AC-68 was not only costly, but it also came
at a point when many NASA engineers with critical experience had been reassigned to other
projects. John Gibb, manager of the Centaur project office at Lewis, was assigned the daunting
task of supervising the rebuilding of the last NASA-owned Centaur. For General Dynamics, it
was an equally challenging undertaking that involved restarting the Centaur assembly line after
it had been shut down for several years. Because of the production hiatus and the loss of key
personnel, NASA noted that the company had difficulty in recapturing the formula for
manufacturing Centaur.61 After two years, Gibb was able to accept delivery of a new rocket.
On 25 September 1989, the FLTSATCOM satellite was placed into orbit by AC-68.62 James
Womack, Director of Expendable Vehicles at the Kennedy Space Center, congratulated Gibb
with appreciation mixed with relief when a week plagued with uncertainties ended with a
flawless launch.63
61
Jean E. Klick to John Klineberg, 29 September 1989, Box AC-68 Flight Reports and Historical Data, Division Atlas-Centaur
Atlas-Centaur Postflight Analysis AC-68, Box AC-68 Flight Reports and Historical Data, Division Atlas-Centaur Project
62
63
John W. Gibb to James L. Womack, 5 October 1989, Box AC-68 Flight Reports and Historical Data, Division Atlas-Centaur
31276-chapter 8 3/9/05 11:24 AM Page 239
Like a Phoenix 239
Headlines in the Washington Times noted, Nostalgic NASA launches last unmanned
payload.64 After 448 launches, dating back to the Thor-Able on 11 October 1958, the Agency
ended its management of expendable launch vehicle launches. From this point on, NASA
would have to purchase launch services from its former contractors. William B. Lenoir,
NASAs acting Associate Administrator for spaceflight, officially commemorated the end of the
old way of doing business in a letter to Lewis Director John M. Klineberg. Lenoir, a former
astronaut, was responsible for developing, operating, and implementing all policy for the
Space Shuttle and all U.S. government civil launch activities. Lenoir congratulated John Gibb,
Vernon Weyers, and the Lewis team for exemplary management of the mission.65 The success
of the last Centaur mission, he said, was indicative of the outstanding cooperation between
government and industry that had been the hallmark of the Centaur program.
Congress, which saw commercialization as a means to reduce the enormous federal deficit
and improve the balance of trade, fully supported the barter agreement. Amendments in 1988
to the 1984 space commercialization act added guidelines for establishing insurance limits for
loss or damage to government property, as well as U.S. government indemnification for third-
party liability above $500 million, or what was reasonably available. The purpose of these
amendments was to provide stability and support for a fledgling commercial space industry.
Because one launch cost between $50 and $100 million, Congress focused on the fact that a
viable industry had the potential to affect the United States balance of trade significantly. Bill
Nelson, Chairman of the Subcommittee on Space Science and Applications, stressed that
nurturing this industry was in the national interest. The sale of one commercial launch by a
United States company is equivalent to the import of ten thousand Toyotas,66 he said on more
than one occasion.
Nelson was not satisfied with NASAs apparent lack of effort to simplify its procurement
guidelines. In a letter to Dale D. Myers, acting Administrator, on 3 May 1989, Nelson asked
why NASA seemed unwilling to work toward the modification of the Federal Acquisitions
Regulations in order to expedite NASA procurement. Most of these guidelines were not
compatible with a fixed-price contract. It is time that NASA recognized that it is not the only
participant in our space program, he said acerbically. This nation has a mature commercial
launch industry that is fully capable of meeting the governments launch requirement.67
64
Nostalgic NASA Launches Last Unmanned Payload, Washington Times (26 September 1989).
William B. Lenoir to John M. Klineberg, 10 October 1989, Box AC-68 Flight Reports and Historical Data, Division Atlas-
65
Centaur Project Office, NASA GRC Records.
66
Summary of November 9, 1989, House Subcommittee Hearing on Commercial Space Launch Activities, 27 November
l989, Box 57, DEB vault. See also discussion by Craig Russell Reed, U.S. Commercial Space Launch Policy Implementation,
19861992 (Ph.D. dissertation, George Washington University, 1998), 2124.
67
Bill Nelson to Dale Myers, 3 May 1989, Box 57, NASA GRC Records.
31276-chapter 8 3/9/05 11:24 AM Page 240
Asked to respond to this critique, Centaur staff at Lewis Research Center took issue with
the premise that the nation had a mature commercial launch vehicle industry, noting that
although there were at least thirteen firms that wanted to get into the business, there were only
three firms (General Dynamics, McDonnell Douglas, and Martin Marietta) with a demon-
strated technical capability to provide launch services. Although NASA was attempting to
widen as far as possible the gate of opportunity for all domestic launch service firms,68 the
Agency needed to balance its different obligations to itself, to industry, and to the taxpayer.
The issue was the degree to which NASA should risk expensive payloads on commercial launch
vehicles. Simplistically, the easiest approach from NASAs perspective, and the most purely
commercial would be to limit awards to only those pre-qualified firms with a proven track
record of successful performance and not risk NASA payloads on unproven designs.69 The
Lewis staff emphasized that NASA payloads were unique, one-of-a-kind, high cost, virtually
irreplaceable spacecraft that required unique launch services. In contrast, commercial
launches were more repetitive, less unique, and far less expensive to replace. They cited the
successful negotiation of the first NASA contract for General Dynamics launch services in
1988. During negotiations with General Dynamics for GOES for the National Oceanic and
Atmospheric Administration, they were able to waive the certified cost and pricing data
requirement. They believed that extreme caution and insistence on NASA oversight were justi-
fied. While we have dramatically scaled back our oversight requirements wherever possible,
prudence dictates that we cannot blindly accept assertions that all is being done properly.70
Members of COMSTAC pushed NASA to accept the new commercial mandate. For
example, Dennis Dunbar, Vice President of Programs and Technical Operations for General
Dynamics Commercial Launch Services, Inc., pointed out in testimony before COMSTACs
Subcommittee on Procurement that commercialization entailed a complete change of mind
set on the part of those doing the procuring for the government.71 To procure only one vehicle
at a time could never be cost-effective. There were distinct advantages to the government in
procuring launch vehicles in blocks with common specifications. Over the previous decade,
the government had contracted for an average of one and a half vehicles per year. With the
approach advocated by General Dynamics of eight vehicles per year, costs were projected to be
reduced by 64 percent.
68
Talking Points on ELV Commercialization, in draft of response to Bill Nelson, 22 May 1989, Box 57, NASA GRC Records.
69
Ibid.
70
Ibid.
71
Testimony of COMSTAC Subcommittee on Procurement to Subcommittee on Space Science and Applications on
Implementation of the Commercial Space Launch Act and its Amendments and Space Transportations Services Purchase Act of
1989, 9 November 1989, ELV Program Planning, 19861989, NASA GRC Records.
31276-chapter 8 3/9/05 11:24 AM Page 241
Like a Phoenix 241
Dunbar compared the General Dynamics first commercial contract with Eutelsat, a
French satellite company, with the GOES contract. The Eutelsat contract was 91 pages,
compared to 4,250 pages for the NASA GOES proposal. Nevertheless, Dunbar thought that
the NASA contract represented significant progress. After spirited negotiations, GD was able
to convince the government that many of their usual requirements just did not fit in a
commercial buy. There were many argumentswe won some and lost somebut there was
healthy movement in what we consider the right direction.72 Subsequent requests for
proposals were answered in an average of 500 pages, a great step forward for the industry and
NASA. The principal contracting officer on the NASA side of these negotiations, Thomas D.
Tokmenko, incorporated numerous unique provisions that became a model not only for
buying commercial launch services, but also for the purchase of commercial satellites by the
government.
Officers of General Dynamics, however, continued to press the Office of Federal
Procurement Policy to simplify the Federal Acquisition Regulations. To facilitate this process,
the COMSTAC subcommittee presented the government with a model request for proposals
(RFP) and model contract to assist in the procurement process. Some of the recommendations
included a fixed payment schedule rather than progress payments. This change was needed
because the manufacturer now produced launch vehicles on speculation, at contractor risk,
and not allocated to specific missions until just before completion.73 Second, since the
contract was for launch services, not hardware, the statement of work was limited to perform-
ance requirements such as a peculiar orbit requirement, interfaces, and environments. The
government no longer had the right to dictate how a contractor achieved that performance.
Third, for the first time, the contractor could insist that the government adhere to a strict
schedule with penalties for unreasonable delays. The contractor was liable for costs associated
with termination. To allow the contractor to make a profit, the government permitted pricing
to be considered independently of the cost of a particular launch.
As a condition for bidding the MLV II contract, Lovelace succeeded in persuading the
directors of General Dynamics to increase planned production from eighteen to sixty-two
vehicles. This increase was again internally financed. In fact, since the restart of the companys
factory, Al Lovelace had asked the companys board of directors for funding in increments of
$100 million so often that $100 million became known as the Lovelace unit. Fortunately,
this significant investment began to show results. The next year, in addition to contracts for
Intelsat VII and Navy ultra-high-frequency communications satellites, General Dynamics won
72
Ibid.
73
Ibid.
31276-chapter 8 3/9/05 11:24 AM Page 242
business from the Italian Space Agency for its Satellite per Astronomia X (SAX) and from
NASA to launch the Solar and Heliospheric Observatory (SOHO) mission in August 1989. It
was fortuitous for American companies that Arianespace experienced a failure of an Ariane II,
carrying an Intelsat V satellite, the same year as the Challenger disaster. This worldwide hiatus
in commercial launch services allowed General Dynamics and Transpace Carriers, Inc., an
entrepreneurial company set up to market the McDonnell Douglas Delta rocket, to break into
a market dominated by Ariane.74
An Evolving Family of Launch Vehicles
In the days when NASA had procured launch vehicles from General Dynamics one or two
at a time, NASA had supplied the engines and the Centaur guidance system and computer as
government-furnished equipment. In the new world of commercialization, the company took
responsibility for procurement. Roy Roberts commented that for the first time people really
understood now what NASA was bringing to the table.75 Faced with Arianes daunting head
start in commercialization, General Dynamics engineers dusted off NASA studies for Centaur
improvements never implemented because of the going out of business mentality. The
company could no longer expect the government to finance development costs. However,
contrary to the expectation that innovations would cease once NASA no longer financed
improvements to Centaur, commercialization actually promoted innovations.
To make Atlas-Centaur more appealing to the satellite market, General Dynamics began
to think in terms of a family of rockets that could be swiftly adapted to different customer
needs. Within six years of the decision to restart Atlas-Centaur production, the company had
developed four new designs: Atlas I, Atlas II, Atlas IIA, and Atlas IIAS. (These were all Atlas-
Centaur vehicles, but the company dropped the hyphenated name in favor of Atlas.) Each
represented an advance in performance, reliability, and payload envelopes, and reduced the
time it took to manufacture and launch the vehicle. All four models were included in the sixty-
two vehicles built at company expense. Atlas-Centaur engineers emphasized that each
successive model had evolved from the previous one.
Building more than one or two vehicles at a time allowed the company to formulate a
strategic plan that included continuous improvements to the vehicle. This strategy required an
enormous investment on the part of the company. The philosophy of continuous improve-
ment encouraged greater standardization, which in turn reduced costs and increased reliability.
The Atlas I, actually an improved Atlas G-Centaur D-1A, became General Dynamicss first
commercial launch vehicle. A pressure-fed system replaced the Centaur boost pumps, an
74
Craig Russell Reed, U.S. Commercial Space Launch Policy Implementation, 19861992 (Ph.D. dissertation, George
Washington University, 1998), 306.
75
31276-chapter 8 3/9/05 11:24 AM Page 243
Like a Phoenix 243
improvement whose feasibility had been demonstrated by NASA years before. This change
reduced complexity and cost. The avionics system was upgraded. Because it was less expensive
to manufacture than a composite nose fairing, a new 14-foot-diameter metal nose fairing was
developed to accommodate larger payloads originally intended for the Shuttle or Ariane. In the
Atlas booster, a computer-controlled pressurization system with redundant sensors replaced
the old mechanical regulators. This innovation made the system not only more reliable, but
also more versatile.
Atlas II represented an important evolutionary step. One key change in Atlas II and all
later versions was a fixed foam insulation system glued to the outside of Centaurs hydrogen
tank. This eliminated the mechanical complexity and weight of jettisonable insulation panels.
Although the company had been working on the concept of foam insulation for almost a
decade using NASA contracted research and development funds, there was no incentive to
implement these changes under the old system. Up to this point, conventional wisdom in the
company dictated that change involved compromise. It was thought that increased reliability
would affect cost and performance. Edward Bock recalled that during the years when Atlas-
Centaur vehicles were being produced one or two at a time, the message was dont make
changes.76 The switch to foam insulation, however, not only increased performance and reli-
ability, but also reduced cost. As a result of successful implementation, the new philosophy of
continuous improvement permeated the company.
A completely new solid-state avionics suite replaced the obsolete guidance and navigation
system. The new avionics system, also the fruit of NASA contracted research and development
funds sponsored by Lewis Research Center in the 1970s, increased computing capability, reli-
ability, and redundancy while it simultaneously reduced cost from approximately $8 million
per vehicle to $2 million.77 At the same time, the company contracted with Rocketdyne to
upgrade the Atlas engine system to provide greater thrust. Atlas and Centaur were stretched 9
feet and 3 feet, respectively. The first Atlas II launched Eutelsat II into geosynchronous transfer
orbit in December 1991.
Quality control and on-time delivery improved dramatically after General Dynamics
established a mentoring program to encourage strong relationships with engineering peers in
supplier companies. Marty Winkler said, The production line wound up being a real produc-
tion line rather than a haphazard collection of things to do.78 Suppliers learned in far greater
detail how their particular part was going to be used in the vehicle, and they began to take
responsibility that went beyond the specified form, fit, and function.
76
Interview with Edward Bock by Virginia Dawson, 22 March 2001.
77
Interview with Martin Winkler by Virginia Dawson, 21 March 2001.
78
Ibid.
31276-chapter 8 3/9/05 11:24 AM Page 244
Atlas-Centaur 69, the first commercial launch of a NASA mission. The CRRES satellite, originally scheduled for launch from the
Shuttle, was part of the barter agreement that paved the way for commercialization of Atlas-Centaur. (NASA C-91-02792)
31276-chapter 8 3/9/05 11:24 AM Page 245
Like a Phoenix 245
Greater standardization meant that the vehicle no longer had to be virtually taken apart,
tested, and rebuilt at the Cape. This improvement had an enormous impact on how long it
took to prepare for launch. Management of the launch manifest changed significantly with
commercialization. In the NASA era, launches were scheduled years in advance and General
Dynamics planned no more than three or four launches a year. After commercialization, the
number of launches averaged to 6.36 per year, with a peak number of launches occurring in
1995 when 11 Atlas-Centaur vehicles were launched. Even more important than the number
of launches was the new interaction between manufacturer and customer that commercializa-
tion allowed for the first time. The manifest became dynamicdriven by customers. The
company was able to reduce the time between the date of signing a contract and actual launch
from thirty months to about six. The need to be responsive to many different customers
demanded greater flexibility. Roy Roberts pointed out that because a commercial customer
regards an Atlas-Centaur simply as transportation, they expect you to be ready, to go when
they want to go and be successful.79
For the next iteration, the Atlas IIA, Pratt & Whitney redesigned the RL10 to increase
thrust and added an extendable nozzle. The added length allowed the rocket to be more effi-
cient at higher altitudes by holding in the exhaust plume, rather than allowing it to expand
rapidly. The Atlas IIAS featured four solid rocket boosters to increase overall booster capability.
This was a major design change that the company had resisted for many years because of the
belief that a pressure-stabilized tank was too delicate to support solid rocket boosters. AC-69,
the first commercial launch of a NASA mission, carried the Combined Release and Radiation
Effects Satellite (CRRES) in July 1990. For the first time, no NASA official sat in the block-
house at the Cape during the countdown. Even more startling, General Dynamics was now
emblazoned down the side of the rocket where United States had always been proudly
displayed before. It was indeed the end of an era.
Continuity and Change
Despite an enormous General Dynamics investment in commercialization, it was not clear

by the early 1990s whether the companys space division could survive. Most of the startup
companies in the launch vehicle business had already failed, but even the future of the three
established manufacturers looked doubtful. NASA engineers Nieberding and Spurlock argued
that expendable launch vehicles should be considered a national resource. In an article for
Space Commerce in 1990, they emphasized that although the government now purchased
launch services, NASA missions continued to be extremely expensive because NASA payloads
were unique and required large-scale government involvement. They pointed out that when
Atlas-Centaur was developed in the 1960s, reliability, not low cost, was the criterion. From
79
31276-chapter 8 3/9/05 11:24 AM Page 246
our experience with Atlas/Centaur and Titan/Centaur, these systems are unforgiving and require
scrupulous, expensive attention to detail to achieve a high probability of success.80 They asserted
the value of continued government oversight of unique and costly NASA missions. They ques-
tioned whether the industry could afford the large investment required to develop a reliable,
low-cost system, particularly when they were faced with government-subsidized foreign competi-
tion. Fierce international competition might drive American companies out of the expendable
launch vehicle business, cutting off United States government access to space.
Continued government involvement, however, made it more difficult for U.S. companies
to compete with Ariane. Bruce Berkowitz, a policy analyst, pointed out that the expense and
rigidity of specification systems used by NASA and the Air Force to ensure quality tended to
discourage innovation. Government specifications controlled the design of vehicles built for
the commercial market because it was not economical to manufacture different models for
different customers. The use of the same launch facilities for both government and commer-
cial payloads also complicated operations and caused delays. NASA and Air Force missions
took precedence over commercial payloads, while Arianes manifest put the commercial
customer first. Berkowitz found it ironic that a country that extolled the benefits of a free
market system allowed the U.S. government to call the shots. Unless we adopt policies that
unleash the potential of the private sector, he warned, we are using only one leg to catch up
in the race.81
Of the three big manufacturers, the future of General Dynamics was the most problem-
atic because of its large investment in commercialization. In 1990, the Atlas program lost $300
million. During the next two years, the company lost more revenues when three Atlas I
launches of commercial payloads failed (AC-70, -71, and -74). These failures made it
extremely difficult for the company to sell its services to satellite customers. Two of these fail-
ures were attributed to the icing of a valve in the Pratt & Whitney RL10 engine. The valve was
required when boost pumps were eliminated in 1984 and worked well for several flights. Then
a slight reduction in valve clearance caused the valve to stick. The third loss of an Atlas-
Centaur was caused by the flawed design of a small screw in the Atlas Rocketdyne engine. The
faulty screw had been used without incident for thirty years prior to this failure, whose timing
was especially unfortunate.82
After the redesign of the screw, the company launched Telstar-401 into geosynchronous
transfer orbit in December 1993. The success of this launch did little to dispel the gloom that
80
J. J. Nieberding and O. F. Spurlock, U.S. ELVs: A Perspective on Their Past and Future, Space Commerce 1 (1990): 32.
81
Bruce D. Berkowitz, Energizing the Space Launch Industry, Issues in Science and Technology (Winter 19891990): 7783.
82
Interview with Marty Winkler by Virginia Dawson, 21 March 2001. See also Builder Suspects Engine Failure in Failed
Liftoff, New York Times (20 April 1991).
31276-chapter 8 3/9/05 11:24 AM Page 247
Like a Phoenix 247
had settled over the company. By this time, the company had launched eighty-one Atlas-
Centaur vehicles with a success rate of over 93 percent, but three failures in twenty-three
months sealed the fate of the company.83 To the General Dynamics corporate headquarters, it
appeared that the enormous investment in upgrading the launch pads, incorporating new tech-
nology into the launch vehicle, and starting up the assembly line could never be recovered.
Former Shuttle commander turned insurance company executive Frederick Hauck recalled
that at this time, Atlas was virtually uninsurable.84 A three-year hiatus in sales resulted in large
losses for the company and prompted the decision to sell the Space Systems Division of the
company. A new chief executive officer, Michael Wynne, was brought in to prepare the
company for sale.
In March 1994, Martin Marietta, manufacturer of Titan, bought the Space Systems
Division of General Dynamics for $208.5 million. This was considered a bargain price for a
major company division with a backlog of twenty-nine Atlas-Centaur contracts and fifteen
upper-stage Titan-Centaur contracts through the end of the decade. The chairman and chief
executive of Martin Marietta, Norman Augustine, commented, There just isnt enough work
to be done to justify all the companies in the business.85 The deal included a large tax write-
off of debt for General Dynamics.
Initially, many General Dynamics employees believed that Centaur as an upper stage for Titan
IV was the quarry that prompted the Martin Marietta purchase. After several unsuccessful
missions with the Titan II-Transtage, the company had withdrawn from the commercial market.86
However, the company soon recognized that the new Atlas family of vehicles had greater commer-
cial potential than the powerful, but far less technically advanced, Titan. Martin Marietta would
later decide to halt production of Titan II and III and declare them obsolete.87
The sale of the General Dynamics Space Systems Division to Martin Marietta
Astronautics included all of the Atlas-Centaur and Centaur for the Titan IV intellectual prop-
erty, tooling, inventory, launch-site assets, and work in progress. Martin Marietta swiftly
relocated the Space Systems Division to its Waterton Plant south of Denver in the winter of
1995. Manufacture of adapters and fairings remained at the General Dynamics plant in
Harlingen, Texas. Tank manufacture, because of the advanced technical skills required for
83
Mike Patzer, Robert C. White, and Terry A. Bohlen, Status and Review of the Commercial Atlas Program, 15th AIAA
International Communications Satellite Systems Conference, San Diego, February 1994.
84
Interview with Richard Hauck by Mark Bowles, 23 August 2001.
Marietta Warns on Acquisition, New York Times (19 February 1994); Martin Marietta To Buy General Dynamics,
85
Washington Post (23 December 1993).
86
Comments on manuscript by Ed Bock, 11 March 2002.
87
The information in this paragraph is from comments on the manuscript by Ed Bock, 11 March 2002.
31276-chapter 8 3/9/05 11:24 AM Page 248
welding, remained at Air Force Plant 19 in San Diego. Martin Marietta assumed the General
Dynamics lease. The entire Kearny Mesa plant in San Diego, including all manufacturing, test,
and engineering facilities, was razed.
Most of the engineers supporting these two programs were given job offers, and about half
of them chose to make the move. Approximately eight hundred former General Dynamics
employees transferred to Denver in 1994. Surprisingly, Martin Marietta offered them most of
the top jobs. Mike Wynne was named executive vice-president of all launch vehicles, with the
exception of Titan, and Edward Bock became the vice-president of the Atlas program. Edward
Squires became the vice-president of all Martin Marietta Astronautics operations, and Robert
DiNal became vice-president/division chief engineer. Only Tom Knapp, who became director
of the Atlas launch vehicle, came from the ranks of the Martin Marietta Titan program. Knapp
played a key role in providing a smooth transition and helping former Atlas program people
deal with the Titan culture in Denver. Gene Fourney, a longtime member of NASAs resident
office in San Diego, worried that precious engineering expertise would be lost in the move.
You know you had this terrific organization back there, and it was really putting a lot of
successful flights together. And then youd move them, uproot them, and wonder what youre
going to do to that record.88 He was soon reassured.
Martin Marietta did not try to impose control over the program but gave the Atlas
managers authority to hire additional people from the Denver area to fill vacancies. In May
1995, production began at the new Atlas and Centaur Final Assembly Building at the
Waterton facility in Colorado. In September, the companys International Launch Services
opened a new headquarters in San Diego, California. That same month, the last Atlas booster
to be built in San Diego came off the assembly line, commemorated by a mass signing of the
aft end by former General Dynamics workers.
The takeover of General Dynamics by Martin Marietta, a long-time competitor in the
launch vehicle business, proved more positive than many former General Dynamics and Lewis
Research Center employees could have imagined. Air Force oversight of Titan had produced a
distinctive Titan culture at Martin Marietta. Because Air Force personnel changed frequently,
they had less time to become involved in the nuts and bolts of Titan-Centaur. In contrast,
NASA had extensive in-house technical expertise and had always worked closely with General
Dynamics to solve engineering problems. Although commercialization would break this close
tie between government and industry, General Dynamics reaped the many benefits that this
unique association had produced over the course of more than thirty years.
The forbearance of Martin Marietta management in allowing former General Dynamics
employees to play an important role in the transition paid off for the company. Because the
first launches were successful, the company did not try to impose Titan processes and proce-
88
Interview with Gene Fourney by Virginia Dawson, 21 March 2001.
31276-chapter 8 3/9/05 11:24 AM Page 249
Like a Phoenix 249
No longer called Centaur, the upper stage for the Atlas II is raised to a vertical position on Pad 36A for a launch of a GOES satellite
for the National Weather Service system, March 2000.
31276-chapter 8 3/9/05 11:24 AM Page 250
dures on Atlas. Mike Benik, at that time head of the NASA resident office, observed, What
happened was the Atlas culture survived even though they were now in Denver at Martin
Marietta, and in effect, [NASA] ended up dealing with two different companies, all under one
roof.89 Ed Bock commented, I think if wed had a failure shortly after we moved here, we
would have been Titanized at the drop of a hat.90
Used to Air Force oversight, the Martin Marietta General Manager was at first visibly
uncomfortable when he was informed that it was not the customers responsibility to make the
decision to launch. Roy Roberts said, We were making a decision on our own to push the
button, and all we had to know was that the spacecraft was ready.91 By the end of 1995, Atlas-
Centaur had accomplished a record-breaking eleven successful launches. Far from being
discontinued, Atlas-Centaur became the divisions cash cow.92
A group of young engineers at General Dynamics, mentored by the small cadre of Atlas-
Centaur veterans, continued the tradition of innovation at General Dynamics. They designed
the series of evolved commercial vehicles and persevered after the Atlas I engine failures and
the move to Denver with Martin Marietta. Since the three engine failures, there has been a
100-percent success record in all fifty-nine flights completed through February 2002, despite
significant design modifications in each of the II, IIA, IIAS, IIIA, and IIIB versions of Atlas.
In 1995, further consolidation of the aerospace industry occurred with the merger
between Martin Marietta and Lockheed. This set the stage for establishing the Lockheed
Martin International Launch Services (ILS), a joint venture owned by Lockheed Martin
Corporation Commercial Launch Services and Lockheed Khrunichev Energia International,
with headquarters near Washington, District of Columbia. ILS markets both the Atlas family
of vehicles and the Russian Proton, built by Khrunichev State Research and Production Space
Center in Moscow.
This joint venture in marketing proved less difficult than the next step in collaboration
with the Russians. Atlas III, completed in Denver in 1998, was the culmination of General
Dynamicss ambitious ten-year development program. The company chose the Russian-built
RD-180 engine to power the Atlas III booster because of its superiority in terms of perform-
ance, technology, and cost to its nearest competitor. General Dynamics was so enthusiastic
over its choice of the Russian engine that it was willing to put considerable time and energy
into overcoming various objections from the U.S. State Department. (Atlas III had a perform-
89
Interview with Mike Benik by Virginia Dawson, 9 November 1999.
90
Interview with Ed Bock by Virginia Dawson, 22 March 2001.
91
92
Interview with Matthew Smith by Virginia Dawson, 21 March 2001.
31276-chapter 8 3/9/05 11:24 AM Page 251
Like a Phoenix 251
ance capability of carrying 9,920 pounds to geosynchronous transfer orbit, or an increase of

1,720 pounds over the Atlas IIAS). It was so powerful that the Atlas booster could again be
stretched. The engine was throttleable, a feature that enabled it to provide a range of efficient
performance for different payloads.
The Russian booster engines paved the way for utilizing the single-engine Centaur for
many missions. Pratt & Whitney upgraded the RL10 engine in many steps to provide 20,800
pounds of thrust for both the dual- and single-engine versions of Atlas. Years earlier, Lewis had
laid the foundation for this upgrade when Centaur had demonstrated a capability for long
coasts and three, four, or five burns. Particularly significant were tests of Centaurs multiple-
restart capability in a zero-gravity environment in NASAs Space Power Facility at Plum Brook
prior to the Helios (TC-5) mission in 1976.93
The single-engine Centaur increased the inherent reliability of the upper stage dramati-
cally because rocket failures are often related to the engine. Half of the propulsion hardware,
including the thrust vector control actuators, propellant feed ducting, pneumatic supply, purge
tubing, and controls, could be eliminated. The electromechanical engine control system
contributed to a leap in performance. Previously, Centaur engines had been chilled down
during the coast phase, after separation from Atlas. Now the Centaur single engine could be
cooled down during the Atlas booster phase. This saved precious seconds and conserved fuel,
which in turn allowed the vehicle to reach orbit more efficiently.
To achieve a greater propellant load, the length of Centaurs tank was stretched to fit on
either an Atlas III or a redesigned 12-foot-diameter Atlas V first stage. Carrying the satellite
Echostar, the new stretched Centaur was successfully flown on the first Atlas IIIB flight on 21
February 2002.
Atlas V, the culmination of the development program that began with commercialization,
is also the companys response to the Air Force demand in 1996 for a more efficient, less
expensive launch vehiclean Evolved Expendable Launch Vehicle, or EELV. Atlas V reflects
the global marketplace in which it operates. The first stage has Russian RD-180 engines. It has
a thick, structurally stabilized aluminum tank, reinforced with rings and stringer for the first
time. Its interstage adapter was built by Construcciones Aeronauticas S.A. (CASA) of Spain
and its composite fairings by Contraves of Switzerland. Its standardized features make it
extremely versatile. Depending on the weight of the payload and power requirements, up to
five large solid rocket motors can be attached to its common core or three boosters can be
strapped together. The boosters are identical to facilitate ease of production.
93
Interview with Bill Johns by Virginia Dawson, 22 March 2001. It should be noted that a 35,000-pound-thrust RL10 deriva-
tive was developed for the new hydrogen/oxygen upper stage manufactured by McDonnell Douglas/Boeing for Delta III.
31276-chapter 8 3/9/05 11:24 AM Page 252
Atlas-Centaur, now simply called Atlas, remains one of the most efficient rocket systems
ever built. Systems engineering and technical management processes honed over three decades
of NASA-industry collaboration are still used in the commercial Atlas program, although the
role of NASA in the new world of commercialization is now greatly diminished. If Atlas-
Centaur is still flying, both the space systems division at General Dynamics and its counterpart
at NASA Glenn have vanished. Bought by Martin Marietta and then merged with Lockheed,
all that remains of the famed missile makers Kearney Mesa complex in San Diego is a lone
administration building. To complete the process of commercialization in 1998, NASA phased
out the launch vehicles division in Cleveland. A few of the men and women in the division at
Glenn transferred to Kennedy Space Center; others retired or moved into other areas of
research and technology. Nevertheless, the unique rocket test facilities conceived and built by
NASA Lewis engineers continue to provide the nation with a unique capability.
Taming liquid hydrogen is a striking example of how government research contributed to
the development of a new technology. Fifty years ago, when engineers at the Lewis Flight
Propulsion Laboratory began experimenting with exotic rocket fuels, they could never have
imagined that their work would make a significant contribution to a race with the Soviet
Union to land human beings on the Moon. Chance also played a role in the development of
liquid-hydrogen technology. Without the aborted development by the Air Force of a high-alti-
tude spy plane in the 1950s, the country would have lacked the technology infrastructure to
produce liquid hydrogen in the enormous quantities needed for the space program.
Research and testing of liquid-hydrogen rockets at Lewis Research Center directly influ-
enced the 1959 decision to use liquid hydrogen in the upper stages of the Saturn V. Once the
commitment to liquid hydrogen had been made, the development of Centaur became of
central importance to proving its feasibility. Because of their expertise in high-energy liquid
fuels, engineers at Lewis Research Center served as a clearinghouse between 1959 and 1962
for technical know-how vitally important to the design of liquid-hydrogen-fueled rockets.
General Dynamics, Pratt & Whitney, Rocketdyne, and Douglas all benefited from their inter-
action with Lewis rocket researchers, but whether a liquid-hydrogen upper stage could actually
perform with the flawless precision required of spacecraft remained unclear. This was still in
doubt as late as October 1962, when management of Centaur was transferred to from Marshall
to Lewis.
With the benefit of hindsight, Marshall scientist Ernst Stuhlinger has called liquid
hydrogen one of the most momentous innovations in the history of rockets during the second
half of the twentieth century. He credits Abe Silverstein with the courageous decision to use
it in Saturns upper stages. He writes:
Hydrogen rocket technology played an absolutely decisive role not only in the Saturn-
Apollo Moon Project, but also in the Shuttle Project that came to life during the
1970s. A new, high performance engine, the Space Shuttle Main Engine (SSME), was
31276-chapter 8 3/9/05 11:24 AM Page 253
Like a Phoenix 253
developed in a joint effort by the G. C. Marshall Space Flight Center and the
Rocketdyne Division of North American Aviation. By that time, hydrogen technology
for rocket engines had reached its full maturity.94
The Centaur saga illustrates how success or failure of a technology ultimately rests on
peoplethe decisions they make and their determination to influence the outcome. To adopt
a new technology involved risk. That Centaur ultimately won the backing of President
Kennedy himself demonstrates the national importance of this technology. A determined band
of government engineers at Lewis persevered in the face of rancorous congressional criticism,
bad press, and active opposition by Marshall Space Flight Center. A collaborative relationship
built on mutual respect between NASA civil servants and government contractors kept the
program alive.
However, if liquid-hydrogen technology moved into NASAs mainstream, neither Centaur
nor Lewis Research Center ever did. By the 1970s, Centaur was regarded as yesterdays tech-
nology, a dependable workhorse for launching planetary probes and commercial satellites
while NASA pushed forward with the development of a reusable Space Shuttle. When the
Shuttles technical limitations for launching planetary missions became clear, NASA turned
once again to Centaur. Ironically, while Marshall Space Flight Center actively campaigned for
the termination of the Centaur program in the 1960s, it fought unsuccessfully for manage-
ment of the Shuttle/Centaur program.
Nowhere is the theme of NASAs changing tolerance for risk more forcefully demonstrated
than in the tortuous history of Shuttle/Centaur. Lewis engineers had confidence in the systems
they designed to protect the astronauts should a mission be aborted, but they had great diffi-
culty in negotiating the politics of the manned spaceflight establishment. After the loss of
Challenger in 1986, it was clear that the program was doomed.
In the 1970s, when NASA decided to phase out its expendable launch vehicles, the
successful commercialization of Centaur was inconceivable. Yet global competition for
commercial satellite business and the loss of Challenger precipitated a change in national prior-
ities. Today, Atlas-Centaur remains one of the most efficient rocket systems ever built.
Although its identity is now submerged within the Atlas, Centaur has proven itself as an upper
stage beyond the wildest dreams of one of the great visionaries of the space age. Someday a
probe powered by Centaur may yet reach Alpha Centauri, the star that tantalized the rockets
gifted designer Krafft Ehricke.
94
Ernst Stuhlinger, Enabling Technology for Space Transportation, The Century of Space Science, vol. 1 (Dordrecht: Kluwer
Academic Publishers, 2001), 7374.
31276-chapter 8 3/9/05 11:24 AM Page 254
31276-back matter 3/9/05 11:41 AM Page 255
List of Interviews 255
List of Interviews
Eighty-nine members of the Lewis Research Center Centaur team celebrate the fiftieth Atlas-Centaur Flight, June 1979: Joe
Kubancik, Gil Widra, Jack Herman, Steve Szabo, Norm Weisberg, Bill Groesbeck, Maurie Baddour, Dick Woelfle, Henry Synor,
Lillian Manning, Ed Ziemba, Larry Ross, Jim Patterson, Maynard Weston, Dick Heath, Tom Hill, Joe Nieberding, Sharon Huber,
Bob Edwards, Floyd Smith, Barbara Troyan, Konnie Semenchuk, Carl Wentworth, Ted Gerus, Jim Magrini, Don Garman, Paul
Kuebeler, Dean Bitler, Bill Kress, Sandy Tingley, Carl Monnin, George Schaefer, Harlan Simon, Peggy Schuler, Rita Hart, Irene
Blanchard, Steve Kiacz, Bob Firestone, Bill Misichko, Bill Hulzman, Harry Groth, Tom Seeholzer, Bob Reinberger, Jim Stoll, Baxter
Beaton, Merle Jones, Dick Kalo, John Kramer, Bill Middendorf, Gene Cieslewicz, Ed Procasky, Jack Farley, Pam Foerster, John
Feagan, Sandy Hanes, Pat Miller, Jack Estes, John Gibb, Ed Muckley, Bob Miller, Dick Flage, Dick Orzechowski, John Roberts,
Charley Eastwood, Ralph Kuivinen, Frank Gue, Janos Borsody, Ken Baud, Gene Fourney, Ray Salmi, John Nechvatal , Bob Jabo,
Ed Jeris, Al Hahn, Cecil ODear, Andy Dobos, Gus Delaney, Bob Lubick, Steve Szpatura, Ted Porada, Art Zimmerman, John Riehl,
Ray Lacovic, Bob Rohal, John Bulloch, Jim McAleese, Chuck Tiede, John Catone, and Earl Hanes. (Glenn Research Center
unprocessed photo)
Glenn Research Center

Brun, Rinaldo J., by Virginia Dawson, 17 June 1999
Everett, Ronald, by V. Dawson, 17 December 2001
Fourney, Eugene, by V. Dawson, 21 March 2001 (at Lockheed Martin, Denver)
Geye, Richard, by V. Dawson, 27 May 1999
Himmel, Seymour, by V. Dawson, 1 March 2000
Kozar, Robert, by V. Dawson and Mark Bowles (at Plum Brook)

Lundin, Bruce, by V. Dawson, 7 March 2000
Muckley, Edwin, by V. Dawson and M. Bowles, 29 April 1999
Nettles, Cary, by V. Dawson, 4 March 1999
Nieberding, Joe, by V. Dawson and M. Bowles, 15 April 1999, 9 May
2000, 7 November 2001
Reshotko, Meyer, by V. Dawson, 17 December 2001 (Cleveland Heights)
Robbins, Red, by M. Bowles, 21 March 2000
Ross, Larry, by M. Bowles, 1 March 2000
Simon, Harlan, by V. Dawson, 20 March 1985
Spurlock, Frank, by V. Dawson, 6 April 1999
Stofan, Andrew, by M. Bowles, 13 April 2000 (by phone)
Swavely, Jim, by V. Dawson, 16 November 1999
Weyers, Vern, by M. Bowles, 8 April 2000
Ziemba, Edmund R., by V. Dawson, 19 May 1999
Zimmerman, Art, by V. Dawson and M. Bowles, 5 August 1999
Johnson Space Center

Hauck, Rick, by M. Bowles, 23 August 2001 (by phone)
Kohrs, Richard, by M. Bowles, 13 July 2001 (by phone)
Kennedy Space Center

Benik, Mike, by V. Dawson, 9 November 1999
Carpenter, Mary Sue, by V. Dawson, 11 November 1999
Curington, Floyd, by V. Dawson, 9 November 1999
Franois, Steve, by V. Dawson, 9 November 1999
Gossett, John, by V. Dawson, 2 July 2002
Gray, Robert, by V. Dawson, 9 November 1999
Mackey, Arthur (Skip), by V. Dawson, 9 November 1999
Neilon, John, by V. Dawson, 1 July 2002
Womack, James, by V. Dawson, 11 November 1999
Former General Dynamics employees, San Diego

Anthony, Frank, by V. Dawson, 6 June 2000
Benzwi, Robert, by V. Dawson, 7 June 2000
Bonesteel, Howard, by V. Dawson, 7 June 2000
Bradley, Robert, by V. Dawson, 7 June 2000
Chamberlain, Stanley, by V. Dawson, 7 June 2000
Chandik, Thomas, by V. Dawson, 6 June 2000
Hansen, Grant, by V. Dawson, 6 June 2000
Kachigan, Karl, by V. Dawson, 7 June 2000
Kaukonen, Everett, by V. Dawson, 6 June 2000
Laird, Hugh, by V. Dawson, 6 June 2000
Lesney, Donald, by V. Dawson, 5 June 2000
Lovelace, Alan, by V. Dawson, 19 July 2002 (by phone)
Lynch, Roger, by V. Dawson, 5 June 2000
Martin, Richard, by V. Dawson, 5 June 2000
Merino, Frederick, by V. Dawson, 5 June 2000
List of Interviews 257
Perry, L. Charles, by V. Dawson, 6 June 2000

Wilmot, Allan, by V. Dawson, 6 June 2000
Wilson, Charles (Chuck), by 7 June 2000
Winkler, Martin, by V. Dawson, 21 March 2001
Zeenkov, Seymour, by V. Dawson, 5 June 2000
Lockheed Martin, Denver, Colorado

Bock, Edward, by V. Dawson, 22 March 2001
Jensen, Michael, by V. Dawson, 22 March 2001
Johns, William, by V. Dawson, 22 March 2001
Owara, Vern, by V. Dawson, 22 March 2001
Reynolds, Robert, by V. Dawson, 21 March 2001
Roberts, Roy, by V. Dawson, 21 March 2001
Smith, Matthew, by V. Dawson, 21 March 2001
JPL
Haynes, Norm, by V. Dawson, 9 June 2000
ONeil, William, by V. Dawson, 9 June 2000
Barnett, Phillip, by V. Dawson, 9 June 2000
Kohlhase, Charles, by V. Dawson, 9 June 2000
Shaw, Lutha (Tom), by V. Dawson, 10 November 1999 (at Kennedy Space Center)
About the Authors 259
About the Authors

Virginia P. Dawson is the author of Natures Enigma, Engines and Innovation: Lewis Laboratory
and American Propulsion Technology and Lincoln Electric, a History and has contributed chapters in
several collections on aerospace history. She is founder and president of History Enterprises, Inc.,
http://www.HistoryEnterprises.com, and an adjunct professor of history at Case Western Reserve
University in Cleveland, Ohio.
Mark D. Bowles received his B.A. in psychology and M.A. in history from the University of
Akron. He earned his Ph.D. from Case Western Reserve University in history in 1999. He has also
been a Tomash Fellow from the Charles Babbage Institute at the University of Minnesota. Dr.
Bowles is the author of Our Healing Mission, a history of Saint Francis Hospital and Medical
Center in Hartford, Connecticut. He is currently writing Reactor in the Garden, a history of
NASAs nuclear research reactor at Plum Brook Station.
The NASA History Series 261
The NASA History Series

Reference Works, NASA SP-4000
Grimwood, James M. Project Mercury: A Chronology. NASA SP-4001, 1963.
Grimwood, James M., and C. Barton Hacker, with Peter J. Vorzimmer. Project Gemini Technology and
Operations: A Chronology. NASA SP-4002, 1969.
Link, Mae Mills. Space Medicine in Project Mercury. NASA SP-4003, 1965.
Astronautics and Aeronautics, 1963: Chronology of Science, Technology, and Policy. NASA SP-4004, 1964.
Ertel, Ivan D., and Mary Louise Morse. The Apollo Spacecraft: A Chronology, Volume I, Through November 7,
1962. NASA SP-4009, 1969.
Morse, Mary Louise, and Jean Kernahan Bays. The Apollo Spacecraft: A Chronology, Volume II, November 8,
1962September 30, 1964. NASA SP-4009, 1973.
Brooks, Courtney G., and Ivan D. Ertel. The Apollo Spacecraft: A Chronology, Volume III, October 1, 1964-
January 20, 1966. NASA SP-4009, 1973.
Ertel, Ivan D., and Roland W. Newkirk, with Courtney G. Brooks. The Apollo Spacecraft: A Chronology, Volume
IV, January 21, 1966July 13, 1974. NASA SP-4009, 1978.
Newkirk, Roland W., and Ivan D. Ertel, with Courtney G. Brooks. Skylab: A Chronology. NASA SP-4011, 1977.
Van Nimmen, Jane, and Leonard C. Bruno, with Robert L. Rosholt. NASA Historical Data Book, Volume I:
NASA Resources, 19581968. NASA SP-4012, 1976, rep. ed. 1988.
Ezell, Linda Neuman. NASA Historical Data Book, Volume II: Programs and Projects, 19581968. NASA SP-
4012, 1988.
Ezell, Linda Neuman. NASA Historical Data Book, Volume III: Programs and Projects, 19691978. NASA SP-
4012, 1988.
Gawdiak, Ihor Y., with Helen Fedor, compilers. NASA Historical Data Book, Volume IV: NASA Resources,
19691978. NASA SP-4012, 1994.
Rumerman, Judy A., compiler. NASA Historical Data Book, 1979-1988: Volume V, NASA Launch Systems, Space
Transportation, Human Spaceflight, and Space Science. NASA SP-4012, 1999.
Rumerman, Judy A., compiler. NASA Historical Data Book, Volume VI: NASA Space Applications, Aeronautics
and Space Research and Technology, Tracking and Data Acquisition/Space Operations, Commercial Programs, and
Resources, 19791988. NASA SP-2000-4012, 2000.
Noordung, Hermann. The Problem of Space Travel: The Rocket Motor. Edited by Ernst Stuhlinger and J. D.
Hunley, with Jennifer Garland. NASA SP-4026, 1995.
Astronautics and Aeronautics, 19861990: A Chronology. NASA SP-4027, 1997.
Astronautics and Aeronautics, 19901995: A Chronology. NASA SP-2000-4028, 2000.
Management Histories, NASA SP-4100

Rosholt, Robert L. An Administrative History of NASA, 19581963. NASA SP-4101, 1966.
Levine, Arnold S. Managing NASA in the Apollo Era. NASA SP-4102, 1982.
Roland, Alex. Model Research: The National Advisory Committee for Aeronautics, 19151958. NASA SP-4103, 1985.
Fries, Sylvia D. NASA Engineers and the Age of Apollo. NASA SP-4104, 1992.
Glennan, T. Keith. The Birth of NASA: The Diary of T. Keith Glennan. J. D. Hunley, ed. NASA SP-4105, 1993.
Seamans, Robert C., Jr. Aiming at Targets: The Autobiography of Robert C. Seamans, Jr. NASA SP-4106, 1996.
Garber, Stephen J., Looking Backward, Looking Forward: Forty Years of U.S. Human Spaceflight Symposium.
NASA SP-4107, 2002.
Project Histories, NASA SP-4200

Swenson, Loyd S., Jr., James M. Grimwood, and Charles C. Alexander. This New Ocean: A History of Project
Mercury. NASA SP-4201, 1966; rep. ed. 1998.
Green, Constance McLaughlin, and Milton Lomask. Vanguard: A History. NASA SP-4202, 1970; rep. ed.
Smithsonian Institution Press, 1971.
Hacker, Barton C., and James M. Grimwood. On Shoulders of Titans: A History of Project Gemini. NASA SP-
4203, 1977.
Benson, Charles D., and William Barnaby Faherty. Moonport: A History of Apollo Launch Facilities and
Operations. NASA SP-4204, 1978.
Brooks, Courtney G., James M. Grimwood, and Loyd S. Swenson, Jr. Chariots for Apollo: A History of Manned
Lunar Spacecraft. NASA SP-4205, 1979.
Bilstein, Roger E. Stages to Saturn: A Technological History of the Apollo/Saturn Launch Vehicles. NASA SP-4206,
1980, rep. ed. 1997.
SP-4207 not published.
Compton, W. David, and Charles D. Benson. Living and Working in Space: A History of Skylab. NASA SP-4208,
1983.
Ezell, Edward Clinton, and Linda Neuman Ezell. The Partnership: A History of the Apollo-Soyuz Test Project.
NASA SP-4209, 1978.
Hall, R. Cargill. Lunar Impact: A History of Project Ranger. NASA SP-4210, 1977.
Newell, Homer E. Beyond the Atmosphere: Early Years of Space Science. NASA SP-4211, 1980.
Ezell, Edward Clinton, and Linda Neuman Ezell. On Mars: Exploration of the Red Planet, 19581978. NASA
SP-4212, 1984.
Pitts, John A. The Human Factor: Biomedicine in the Manned Space Program to 1980. NASA SP-4213, 1985.
Compton, W. David. Where No Man Has Gone Before: A History of Apollo Lunar Exploration Missions. NASA SP-
4214, 1989.
Naugle, John E. First Among Equals: The Selection of NASA Space Science Experiments. NASA SP-4215, 1991.
Wallace, Lane E. Airborne Trailblazer: Two Decades with NASA Langleys Boeing 737 Flying Laboratory. NASA SP-
4216, 1994.
Butrica, Andrew J., ed. Beyond the Ionosphere: Fifty Years of Satellite Communication. NASA SP-4217, 1997.
Butrica, Andrew J. To See the Unseen: A History of Planetary Radar Astronomy. NASA SP-4218, 1996.
Mack, Pamela E., ed. From Engineering Science to Big Science: The NACA and NASA Collier Trophy Research
Project Winners. NASA SP-4219, 1998.
Reed, R. Dale, with Darlene Lister. Wingless Flight: The Lifting Body Story. NASA SP-4220, 1997.
Heppenheimer, T. A. The Space Shuttle Decision: NASAs Search for a Reusable Space Vehicle. NASA SP-4221,
1999.
Hunley, J. D., ed. Toward Mach 2: The Douglas D-558 Program. NASA SP-4222, 1999.
Swanson, Glen E., ed. Before this Decade is Out . . .: Personal Reflections on the Apollo Program. NASA SP-4223,
1999.
Tomayko, James E. Computers Take Flight: A History of NASAs Pioneering Digital Fly-by-Wire Project. NASA SP-
2000-4224, 2000.
Morgan, Clay. Shuttle-Mir: The U.S. and Russia Share Historys Highest Stage. NASA SP-2001-4225, 2001.
Mudgway, Douglas J. Uplink-Downlink: A History of the Deep Space Network, 19571997. NASA SP-2001-
4227, 2002.
Center Histories, NASA SP-4300

Rosenthal, Alfred. Venture into Space: Early Years of Goddard Space Flight Center. NASA SP-4301, 1985.
Hartman, Edwin P. Adventures in Research: A History of Ames Research Center, 19401965. NASA SP-4302,
1970.
Hallion, Richard P. On the Frontier: Flight Research at Dryden, 1946-1981. NASA SP-4303, 1984.
Muenger, Elizabeth A. Searching the Horizon: A History of Ames Research Center, 19401976. NASA SP-4304,
1985.
Hansen, James R. Engineer in Charge: A History of the Langley Aeronautical Laboratory, 19171958. NASA SP-
4305, 1987.
Dawson, Virginia P. Engines and Innovation: Lewis Laboratory and American Propulsion Technology. NASA SP-
4306, 1991.
Dethloff, Henry C. Suddenly Tomorrow Came . . .: A History of the Johnson Space Center. NASA SP-4307,
1993.
Hansen, James R. Spaceflight Revolution: NASA Langley Research Center from Sputnik to Apollo. NASA SP-4308,
1995.
Wallace, Lane E. Flights of Discovery: 50 Years at the NASA Dryden Flight Research Center. NASA SP-4309, 1996.
Herring, Mack R. Way Station to Space: A History of the John C. Stennis Space Center. NASA SP-4310, 1997.
Wallace, Harold D., Jr. Wallops Station and the Creation of the American Space Program. NASA SP-4311, 1997.
Wallace, Lane E. Dreams, Hopes, Realities: NASAs Goddard Space Flight Center, The First Forty Years. NASA SP-
4312, 1999.
Dunar, Andrew J., and Stephen P. Waring. Power to Explore: A History of the Marshall Space Flight Center. NASA
SP-4313, 1999.
Bugos, Glenn E. Atmosphere of Freedom: Sixty Years at the NASA Ames Research Center. NASA SP-2000-4314,
2000.
General Histories, NASA SP-4400

Corliss, William R. NASA Sounding Rockets, 19581968: A Historical Summary. NASA SP-4401, 1971.
Wells, Helen T., Susan H. Whiteley, and Carrie Karegeannes. Origins of NASA Names. NASA SP-4402, 1976.
Anderson, Frank W., Jr. Orders of Magnitude: A History of NACA and NASA, 19151980. NASA SP-4403,
1981.
Sloop, John L. Liquid Hydrogen as a Propulsion Fuel, 19451959. NASA SP-4404, 1978.
Roland, Alex. A Spacefaring People: Perspectives on Early Spaceflight. NASA SP-4405, 1985.
Bilstein, Roger E. Orders of Magnitude: A History of the NACA and NASA, 19151990. NASA SP-4406, 1989.
Logsdon, John M., ed., with Linda J. Lear, Jannelle Warren-Findley, Ray A. Williamson, and Dwayne A. Day.
Exploring the Unknown: Selected Documents in the History of the U.S. Civil Space Program, Volume I, Organizing
for Exploration. NASA SP-4407, 1995.
Logsdon, John M., ed., with Dwayne A. Day and Roger D. Launius. Exploring the Unknown: Selected Documents
in the History of the U.S. Civil Space Program, Volume II, Relations with Other Organizations. NASA SP-4407,
1996.
Logsdon, John M., ed., with Roger D. Launius, David H. Onkst, and Stephen J. Garber. Exploring the
Unknown: Selected Documents in the History of the U.S. Civil Space Program, Volume III, Using Space. NASA SP-
4407, 1998.
Logsdon, John M., gen. ed., with Ray A. Williamson, Roger D. Launius, Russell J. Acker, Stephen J. Garber,
and Jonathan L. Friedman. Exploring the Unknown: Selected Documents in the History of the U.S. Civil Space
Program, Volume IV, Accessing Space. NASA SP-4407, 1999.
Logsdon, John M., gen. ed., with Amy Paige Snyder, Roger D. Launius, Stephen J. Garber, and Regan Anne
Newport. Exploring the Unknown: Selected Documents in the History of the U.S. Civil Space Program, Volume V,
Exploring the Cosmos. NASA SP-2001-4407, 2001.
Siddiqi, Asif A. Challenge to Apollo: The Soviet Union and the Space Race, 19451974. NASA SP-2000-4408, 2000.
Monographs in Aerospace History, NASA SP-4500

Launius, Roger D., and Aaron K. Gillette, comps. Toward a History of the Space Shuttle: An Annotated
Bibliography. Monograph in Aerospace History, No. 1, 1992. Out of print.
Launius, Roger D., and J. D. Hunley, comps. An Annotated Bibliography of the Apollo Program. Monograph in
Aerospace History, No. 2, 1994.
Launius, Roger D. Apollo: A Retrospective Analysis. Monograph in Aerospace History, No. 3, 1994.
Hansen, James R. Enchanted Rendezvous: John C. Houbolt and the Genesis of the Lunar-Orbit Rendezvous Concept.
Monograph in Aerospace History, No. 4, 1995.
Gorn, Michael H. Hugh L. Drydens Career in Aviation and Space. Monograph in Aerospace History, No. 5,
1996.
Powers, Sheryll Goecke. Women in Flight Research at NASA Dryden Flight Research Center from 1946 to 1995.
Monograph in Aerospace History, No. 6, 1997.
Portree, David S. F., and Robert C. Trevino. Walking to Olympus: An EVA Chronology. Monograph in Aerospace
History, No. 7, 1997.
Logsdon, John M., moderator. Legislative Origins of the National Aeronautics and Space Act of 1958: Proceedings
of an Oral History Workshop. Monograph in Aerospace History, No. 8, 1998.
Rumerman, Judy A., comp. U.S. Human Spaceflight, A Record of Achievement 19611998. Monograph in
Portree, David S. F. NASAs Origins and the Dawn of the Space Age. Monograph in Aerospace History, No. 10, 1998.
Logsdon, John M. Together in Orbit: The Origins of International Cooperation in the Space Station. Monograph in
Phillips, W. Hewitt. Journey in Aeronautical Research: A Career at NASA Langley Research Center. Monograph in
Braslow, Albert L. A History of Suction-Type Laminar-Flow Control with Emphasis on Flight Research. Monograph
in Aerospace History, No. 13, 1999.
Logsdon, John M., moderator. Managing the Moon Program: Lessons Learned From Apollo. Monograph in
Perminov, V. G. The Difficult Road to Mars: A Brief History of Mars Exploration in the Soviet Union. Monograph
in Aerospace History, No. 15, 1999.
Tucker, Tom. Touchdown: The Development of Propulsion Controlled Aircraft at NASA Dryden. Monograph in
Maisel, Martin D., Demo J. Giulianetti, and Daniel C. Dugan. The History of the XV-15 Tilt Rotor Research
Aircraft: From Concept to Flight. NASA SP-2000-4517, 2000.
Jenkins, Dennis R. Hypersonics Before the Shuttle: A Concise History of the X-15 Research Airplane. NASA SP-
2000-4518, 2000.
Chambers, Joseph R. Partners in Freedom: Contributions of the Langley Research Center to U.S. Military Aircraft in
the 1990s. NASA SP-2000-4519, 2000.
Waltman, Gene L. Black Magic and Gremlins: Analog Flight Simulations at NASAs Flight Research Center. NASA
SP-2000-4520, 2000.
Portree, David S. F. Humans to Mars: Fifty Years of Mission Planning, 19502000. NASA SP-2001-4521, 2001.
Thompson, Milton O., with J. D. Hunley. Flight Research: Problems Encountered and What They Should Teach
Us. NASA SP-2000-4522, 2000.
Tucker, Tom. The Eclipse Project. NASA SP-2000-4523, 2000.
Siddiqi, Asif A. Deep Space Chronicle: A Chrononology of Deep Space and Planetary Probes 19582000. NASA SP-
2002-4524, 2002.
Merlin, Peter W. Mach 3+: NASA/USAF YF-12 Flight Research, 19691979. NASA SP-2001-4525, 2001.
Anderson, Seth B. Memoirs of an Aeronautical Engineer: Flight Test at Ames Research Center: 19401970. NASA
SP-2002-4526, 2002.
Renstrom, Arthur G. Wilbur and Orville Wright: A Bibliography Commemorating the One-Hundredth Anniversary
of the First Powered Flight on December 17, 1903. NASA SP-2002-4527, 2002.
There is no monograph 28.
Chambers, Joseph R. Concept to Reality: Contributions of the NASA Langley Research Center to U.S. Civil Aircraft
of the 1990s. NASA SP-2003-4529, 2003.
Peebles, Curtis, ed. The Spoken Word: Recollections of Dryden History, The Early Years. NASA SP-2003-4530, 2003.
Jenkins, Dennis R., Tony Landis, and Jay Miller. American X-Vehicles: An InventoryX-1 to X-50. NASA SP-
2003-4531, 2003.
Renstrom, Arthur G. Wilbur and Orville Wright: A Reissue of a Chronology Commemorating the Hundredth
Anniversary of the Birth of Orville Wright, August 19, 1871. NASA SP-2003-4532, 2003.
Index 267
INDEX
A
ADDJUST steering program, 73, 128, 130, 144
Advanced Research Projects Agency (ARPA), iii, 13, 18, 20, 22
Advent Communications Satellite Program, 22, 27, 29, 32, 51, 67
Aerobee rocket, 27
Aerojet Engineering Corporation, 14, 15, 142
Aerospace Management, 66
Agena upper stage rocket, x, 22, 25, 26, 51, 52, 54, 55, 56, 51, 93. See also Atlas-Agena, Thor-
Agena
Air Force, ix, 13, 14, 18, 22, 26, 33, 40, 43, 44, 47, 48, 50, 62, 64, 71, 86, 130, 141, 144,
151, 168, 170, 172, 174, 176, 180, 184, 193, 194, 196, 205, 209, 214, 215; and commer-
cialization, 226, 232, 235, 236, 246
Aldridge, Edward (Pete), 235, 236
Aldrin, Edwin (Buzz), 93
American Rocket Society, 2
Ames Research Center, 62, 107
Analex, Inc., 209, 213
Andrews, Jan, 68
Angara engine, v
Anthony, Frank, 67, 68, 128, 256
Apollo 4, 96
Apollo 11, 100, 101
Apollo program, xi, 33, 50, 56, 64, 82, 83, 88, 89, 94, 100, 101, 163, 168, 174, 205
Applications Technology Satellite (ATS), 108, 110, 111, 113, 116, 123
Appold, Norman, 18
Ariane (launch vehicle), iv, v, vii, ix, 221, 228, 231, 232, 233, 236, 242, 243, 246
Arianespace, 221, 233, 235, 242
Army Ballistic Missile Agency, 3
Arnold Engineering Development Center, 27
Asteroid 29 Amphitrite, 190, 191
Asteroid Gaspra, 211
Astrotech International, 192
AT&T, 223, 225
Atlantis, 202
Atlas (launch vehicle), iv, v, ix, 1, 18, 20, 22, 3839, 4951, 6061, 74, 7678, 85, 124, 126,
178, 189
Atlas (missile), 1, 10, 39, 712, 39
Atlas G, 232, 242
Atlas I, 242, 246
Atlas II, 242, 243, 249, 250
Atlas IIA, 242, 245, 250

Atlas IIAS, 242, 245, 250, 251
Atlas III, 250, 251
Atlas IIIA, 250
Atlas IIIB, 250, 251
Atlas V, 251
Atlas-Agena, 20, 32, 44, 50, 61, 102, 104, 108, 113. See also Agena rocket
Atlas-Centaur 1 (F-1), 4647, 61, 72
Atlas-Centaur 2 (AC-2), 80, 81, 83, 90
Atlas-Centaur 3 (AC-3), 84, 90
Atlas-Centaur 5 (AC-5), 85, 87, 90
Atlas-Centaur 11 (AC-11), 90
Atlas-Centaur 19 (AC-19), 112, 116, 118
Atlas-Centaur 20 (AC-20), 112, 116, 118
Atlas-Centaur 22 (AC-22) 116, 117
Atlas-Centaur 30 (AC-30), 116, 125, 130
Atlas-Centaur 34 (AC-34), 116, 131, 132
Index 269

Atlas-Centaur 53 (AC-53), 116, 135, 138
Atlas-Centaur D-1AR, 224
Atlas-Centaur D-IA, 242
Atlas-Centaur diagram, 3
Atlas-Centaur, commercialization of, 237245

Augustine, Norman, 247
Aviation Week and Space Technology, 144, 150, 234
B
B-2 facility. See Plum Brook Spacecraft Propulsion Research Facility
Baddour, Maurie, 255
Baehr, Ed, 61
Barnett, Phillip, 257
Baud, Ken, 255
Baughman, Eugene, 54
Beaton, Baxter, 255
Beggs, James, 232
Bell Aerospace Co., 26
Bell Aircraft Corp., 7
Benik, Mike, 250, 256
Benzwi, Robert, 256
Berkowitz, Bruce, 246
Beverlin, Charles, 118
Bilstein, Roger E., x, 96
Bitler, Dean, 255
Bock, Edward, xii, 197, 199, 243, 248, 250, 257
Boeing, 172, 173, 175, 192, 209, 211
Boland, Edward P., 210
Bonesteel, Howard, 256
Borsody, Janos, 255
Bossart, Karel (Charlie), 3, 4, 7, 8, 9, 10, 34, 38
Bostwick, Leonard, 43
Bradley, Robert, xii, 256
Brahms, William, 65
Bridges, Roy D., 202
Brun, Rinaldo (Jack), 68, 255
Bulloch, John, 255
Burrows, William, 189, 219
C
Campbell, Richard, 65
Canright, Richard, 6
Cape Canaveral, vi, 10, 11, 27, 44, 51, 71, 75, 82, 106, 118, 120, 124, 150, 159, 164, 205,
207, 221, 228, 235, 236, 245
Cape Kennedy, 132, 144
Carpenter, M. Scott, 11
Carpenter, Mary Sue, 256
Index 271
Carter, President Jimmy, administration of, 229

Case Institute of Technology, 99
Cassini Program, 105, 163, 215
Catone, John, 255
Centaur attitude control, 108
Centaur autopilot, 112
Centaur coast period, 108, 109, 112114, 131, 143, 149152, 155, 156, 159, 160
Centaur D-1, 119
Centaur D-1A (improved), 125, 128, 131, 136
Centaur D-1T, 141143
Centaur Engineering Data Center, 213, 214
Centaur failures, iii, 4653, 72, 85, 87, 90, 113, 116, 117, 120, 123, 138, 145, 146
Centaur G, 184, 193, 194, 209, 215
Centaur G-prime, 184, 196, 215
Centaur insulation, 2829
Centaur Integrated Support Structure (CISS), 171, 185, 186, 194, 199, 205, 206, 214
Centaur intermediate bulkhead, 19, 34, 51, 74, 144, 174, 184, 216
Centaur liquid-hydrogen venting, 28, 137, 143, 150, 151, 186, 200
Centaur pressure-stabilized (balloon) structure, iii, 19, 3839, 4849, 227
Centaur Project Office, 109
Centaur Safety Waiver, 200
Centaur Standard Shroud (CSS), 143, 146, 149, 159
Centaur technical modifications after Surveyor, 109112
Centaur tenth anniversary, 99
Center for Space Research, 181
Cernan, Eugene, 93
Challenge to Apollo, 98
Challenger, vii, 167, 168, 189, 202, 207209, 211213, 216, 217, 219, 221, 222, 233, 234,
235, 242, 253
Chamberlain, Stanley, 256
Chance Vought Aircraft, 17
Chandik, Thomas, 256
Chicago Tribune, 94
Cieslewicz, Gene, 255
Clark, Victor, 66
Clarke, Arthur, C., 222
Clayton, Marvin, 65
Coar, Richard, 17
Cold War, iv, viii, 97, 101
Coleman, Kevin, xii
Collins, Michael, 101
Collins, Stuart, 118
Commerce, Department of, 237

Commercial Launch Directive 94 (NSDD 94), 229, 231
Commercial Space Launch Act, 229, 239
Commercial Space Transportation Advisory Committee (COMSTAC), 229, 240, 241
Committee on Science and Astronautics, U.S. House of Representatives, vii, 31, 33, 50, 52,
123
CommSpace, Inc., 230
Communications Satellite Act, 223
Communications satellites, 222225
Comsat (Communications Satellite Corp.), 125, 223, 225
Comstar I, 223
Comstar satellite, 222, 225, 226
Conrad, Charles, Jr., 95
Construcciones Aeronauticas, S.A. (CASA), 251
Contraves Space, A.G., 251
Convair Corp. (Consolidated Vultee Corp.), 7
Cooper, L. Gordon, 11
Copernicus, 117,
Cortright, Edgar, 33, 93
Courington, T. C., 65
Covault, Craig, 144, 150
Crawford, Fred, 99
Crooks, James, 4
CRRES (Combined Release and Radiation Effects Satellite), 238, 244, 245
Cryogenics Engineering Laboratory (U.S. National Bureau of Standards), 14, 29
Curington, Floyd, 256
Cutting, Elliot, 69
D
Dankhoff, Walter, 54
Dante, 192,
Davis, Deane, ix, 20, 31, 34, 35, 37, 38, 67, 71, 74, 91
Day, Dwayne, xii
Death Star, 189, 207
Debus, Kurt, 44, 77
Deep Space Network (DSN), 147, 148
Defense, Department of (DOD), 50, 61, 86, 108, 172, 176, 182, 184, 186, 193, 194, 215,
225, 232
Delaney, Gus, 255
Delta (launch vehicle), iv, v, 44, 53, 75, 124, 125, 134, 141, 149, 150, 224, 231, 233, 235,
242
Dempsey, James, 2, 4, 1012, 35, 36, 37, 65
Dempster, Harry, 108
Index 273
Deschamps, Roland, 232

Dewar, James, 45
DiNal, Robert, 248
Dobos, Andy, 255
Dole, Elizabeth, 229
Dore, Frank, 4
Dornberger, Walter, 7
Douglas Aircraft Corp., 25, 35, 75, 98, 252
Douglas Thor missile, 25, 35
Dryden, Hugh, 53
DUKSUP, 108
Dunbar, Dennis, 240, 241
Dunbar, William (Russ), 64, 125
Dunholter, Howard, 4
DX priority, 50, 54, 64, 86
E
Eastern Test Range, 130, 206
Eastman, Harry O., III, 183
Eastwood, Charles, 255
Eaton, D., 211, 212
Echo satellite, 61
Edwards Air Force Base, 10, 23
Edwards, Robert, 255
EELV (Evolved Expendable Launch Vehicle), 251
Ehricke, Krafft, ix, xii, 1, 2, 4, 67, 1113, 1820, 23, 25, 27, 29, 34, 35, 36, 48, 49, 50, 80,
100, 253
Eisenhower, President Dwight D., 3, 11
Element (Shuttle), 198200
Estes, Jack, 255
European Space Agency (ESA), v, x, 182, 191, 194, 211, 215, 228
Eutelsat II (satellite), 243
Eutelsat, 241
Everett, Ronald, xii, 255
Exceptional Bravery Medal, 118
Ezell, Edward and Linda, ix
F
F-1 Rocketdyne engine, 54
F-1(launch vehicle). See Atlas-Centaur 1
Faget, Max, 33
Failure Review Board, 146
Far Travelers, 33
Faraday, Michael, 4
Farley, Jack, 255
Feagan, John, 255
Federal Ministry for Research and Technology, 147
Files, Colonel William, 194
Firestone, Robert, 255
Flage, Richard, 68, 255
Fleet Satellite Communications. See FLTSATCOM
Fleming, William, 61, 82
Fletcher, James C., 101, 210
Flippo, Ronnie G., 174, 176
FLTSATCOM (Fleet Satellite Communications), 106, 108
FLTSATCOM satellite, 222
FLTSATCOM F1, 223
FLTSATCOM F2, 223
FLTSATCOM F3, 223
FLTSATCOM F4, 223, 226
FLTSATCOM F5, 223
FLTSATCOM F7, 224
Foerster, Pam, 255
Ford Aerospace and Communications Corp., 224
Forsythe, D. L., 25
Fourney, Eugene, 248, 255
Foushee, Robert, 68
Franois, Steve, 256
Frau im Mond, 1
Friedrich, Hans, 4
Frosch, Robert A., 136, 175
G
Gabriel, David, 6264, 66, 93
Galileo mission, vii, 163, 172, 177, 180, 190192, 194, 196, 198, 202, 204205, 207212
Galilei, Galileo, 154
Gardner, Trevor, 8
Garman, Don, 126, 255
Garside, Joseph, 68
Gemini mission, 141
General Accounting Office (GAO), 51, 6162
Index 275
General Dynamics, Astronautics Division (after 1961), viii, ix, x, 6069, 7175, 78, 85, 92,
100, 101, 105, 107, 108, 118, 126, 130, 141, 142, 144, 170, 172174, 180, 183, 193, 194,
196, 197, 199, 203205, 209, 210, 212, 213, 216, 217
General Dynamics, Convair/Astronautics Division (195761), 2, 4, 10, 18, 11, 27, 36, 2223,
3134, 43, 4647, 49, 51, 52, 55, 86
General Dynamics, German Society for Space Research, 5
General Dynamics, Space Systems Division, 226, 229, 232233, 236238, 240, 241242, 246,
250, 252
Gerus, Ted, 73, 255
Geye, Richard, 159, 255
Geyer, David, 126
Giauque, William, 13
Gibb, John, 238, 239, 255
Glenn L. Martin Co. See Martin Marietta Corp.
Glenn, John, 11, 49, 53
Glennan, T. Keith, iii, vii, 11, 22, 31, 99
Goddard Space Flight Center, 44, 75, 77, 107, 113, 147, 226
Goddard, Robert, 5
GOES (Geosynchronous Operational Environmental Satellite), 237, 240, 241, 249
Goette, William, 89
Gold, Thomas, 56
Gordon, W. L., 100
Gossett, John, 75, 144, 256
Grabe, Ronald J., 202
Graham, William R., 208
Grand Tour, 160, 162
Gray, Robert, 75, 7779, 105, 256
Gray, Vern, 2829
Green, Joseph, x
Green, William, 210
Groesbeck, William, 255
Groth, Harry, 255
GTE Satellite Corp. (GSAT), 225
Gue, Frank, 255
guidance systems, 91, 120. See also Minneapolis-Honeywell Co., Teledyne Industries, Inc.
H
Hahn, Al, 255
Hall, Eldon, 16, 24
Hanes, Earl, 255
Hanes, Sandy, 255
Hannum, Ned, 41
Hansen, Grant, 35, 36, 48, 49, 62, 66, 71, 80, 99, 256
Hart, Rita, 255

HATV (High Altitude Test Vehicle), 14
Hauck, Frederick (Rick) H., 189, 202, 204, 207, 210, 217, 218, 247, 256
Haynes, Norm, 105, 257
Heald, Daniel, 27
Heath, Richard, 255
Heatherly, Joe, 63
Heaton, Donald, 27, 33
Heineman, Ed, 80
Heisenberg, Wernher, 5
Helios 1, 140, 153
Helios 2, 140
Helios mission, 108, 139, 143, 147152, 154, 176, 251
Henneberry, Hugh, 67
Hennings, Glen, 126
Henry, Tom, 91
Heritage Foundation, 177
Herman, Jack, 255
Hicks, Edward, 65
High-Energy Astronomy Observatory (HEAO), 108, 136
High-Energy Astronomy Observatory-1, 116, 137
High-Energy Astronomy Observatory-2, 116
High-Energy Astronomy Observatory-3, 116, 135, 138
Hill, Robert, 65
Hill, Tom, 255
Hilmers, David C., 202
Himmel, Seymour, 61, 143, 146, 155, 256
Hindenberg dirigible, 5
Hines, William, 54
Hoagland, Richard C., 157
Hoften, James (Ox), 202
Holmes, Brainerd, 53
Holmes, Oliver Wendell, 215, 216
Honeywell. See Minneapolis-Honeywell Co.
Hubble telescope, 117
Huber, Denny, 27, 46, 66
Huber, Sharon, 255
Hueter, Hans, 32, 36, 37, 38, 47
Hughes Aircraft Co., 33, 69, 83, 92
Hughes, Thomas, vi
Hulzman, William, 255
Index 277
Hunley, John, xii

Huygens, Christian, 154
hydrogen liquifier (cryostat), 14, 17
I
Inertial Upper Stage (IUS), 171176, 178, 190, 192, 211
Integrate, Transfer and Launch (ITL) facility, 141
integrating Centaur with Atlas, 3940, 76, 77, 230
integrating Centaur with Shuttle, 178, 180, 185, 195, 198
integrating Centaur with Titan, 143, 144, 163
Intelsat. See International Telecommunications Satellite Organization
Intelsat IIIIV, 224
Intelsat IV, 223
Intelsat IVA, 223
Intelsat V, 223, 228, 242
Intelsat VA, 224
Intelsat VII, 241
Interim Orbital Transfer Vehicle (IOTV), 171
International Solar Polar Mission. See Ulysses spacecraft
International Telecommunications Satellite Organization (Intelsat), 106, 108, 125, 131, 138,
149, 155, 222, 223, 226, 228
Irene Blanchard, 255
Italian Space Agency (Agenzia Spaziale Italiana), 215, 242
J
J-2 Rocketdyne engine, 54, 85, 96, 98
Jabo, Robert, 255
Jakobowski, Walter, 151
James, J. N., 104
Jenkins, Dennis R., xi
Jensen, Michael, 257
Jeris, Ed, 255
Jet Propulsion Laboratory (JPL), 14, 15, 26, 29, 33, 52, 53, 55, 57, 59, 60, 6769, 83, 91, 92,
95, 104, 105, 107, 120, 157, 159, 165, 190, 191, 194, 197, 211, 212, 226
Johns, William, 257
Johnson Space Center (JSC), vii, viii, 168, 174, 177180, 182, 184, 185, 191, 194, 196202,
207209, 218
Johnson, Clarence (Kelly), 17, 18
Johnson, President Lyndon, 100
Johnson, Roy, 11
Johnson, Vincent, 53, 54, 73, 125, 226
Johnston, Herrick L., 13, 14
Jonash, Edmund, 62, 63, 64, 93
Jones, Fuller C., x

Jones, Merle, 255
Journal of the American Rocket Society, 6
K
Kachigan, Karl, 77, 78, 144, 145, 256
Kalo, Richard, 255
Karth, Congressman Joseph, 33, 47, 83, 85, 86, 88, 123
Katalinsky, Drew, 63
Kaukonen, Everett, 256
Kearny Mesa complex, 10, 71, 142, 248, 252
Kennedy Space Center (KSC), iv, ix, xii, 44, 75, 92, 105, 120, 129, 141, 146, 148, 151, 158,
161, 177180, 182, 194, 204206, 208, 209, 211, 214, 234, 252
Kennedy, President John F., iii, iv, viii, 50, 56, 80, 100, 101, 253
Khrunichev State Research and Production Space Center, 250
Kiacz, Steve, 255
Klineberg, John M., 239
Kloman, Erasmus, 55
Knapp, Tom, 248
Koester, F. W., 68
Kohlhase, Charles, 105, 257
Kohrs, Richard, 218, 256
Koppes, Clayton R., ix
Korolev, Sergei, 98
Kozar, Robert, 256
Kraemer, Robert S., 103, 120, 154, 157
Kraft, Chris C., Jr., 178
Kramer, John, 255
Kranzberg, Melvin, xi
Kress, William, 255
Krige, John, x, 233
Kubancik, Joe, 255
Kuebeler, Paul, 255
Kuivinen, Ralph, 255
L
Lacovic, Ray, 84, 255
Laird, Hugh, 256
Lang, Fritz, 1
Lange, Oswald, 24
Langley Laboratory, 13, 59
Langley Research Center, 169, 226
Launius, Roger, xi, 163, 165
Index 279
Lee, Jack, 68
Lenoir, William B., 239
Lesney, Don, 3940, 128, 256
Lewis (Glenn) Research Center, vi, ix, 23, 27, 43, 53, 55, 60, 64, 66, 68, 69, 71, 73, 80, 84,
89, 92, 98
Lewis Flight Propulsion Laboratory, v, 13, 15, 16, 17
Lewis Research Center Altitude Wind Tunnel, 117
Lewis Research Center Development Engineering Building (DEB), 106
Lewis Research Center Launch Vehicles Division, 106, 226, 227
Lewis Research Center Propellant Systems Section, 141, 181
Lewis Research Center Reduction in Force (RIF), 106, 177
Lewis Research Center Space Transportation Engineering Division, 182, 183
Lewis Research Center Trajectory Group, 107
Lewis Research Center, role in commercialization, 226, 227, 233, 240, 243, 248, 252
Lewis, Roger, 80
Lightbown, Red, 36, 77
Lindbergh Field, 10
Linde Air Products Company, 29
Lindsay, Brigadier General Nathan J., 215
liquid hydrogen, iii, v, 4, 5, 1314, 1719, 4950, 252
liquid oxygen, 45
Lockheed Aircraft Corp., 16, 17
Lockheed Khrunichev Energia International, 250
Lockheed Martin Corp., iv, vii, viii, 250
Lockheed Martin International Launch Services (ILS), 250
Lockheed Missiles and Space Co., 26, 101, 143, 209
Logsdon, John M., 231
Lounge, J. Mike, 202
Lovelace, Alan M., xii, 178, 180, 205, 229, 235, 236, 241
Low, George, 101, 120, 123, 169
Lowell, Percival, 154
Lowery, William, 174
Lubick, Robert, 255
Lucas, William R., 178, 180
Luna IX, 92
Lunar Excursion Module, 33, 88, 92, 93, 95
Lunar Orbiter Program, 55, 82
Lundin, Bruce, 53, 54, 65, 66, 73, 77, 99, 106, 128, 226, 256
Lunney, Glynn, 184, 199201
Lynch, Roger, xii, 44, 45, 46, 66, 79, 80, 91, 256
M
Mackey, Arthur (Skip), 79, 256
Magellan spacecraft, 163, 192
Magellan, Ferdinand, 219
Magrini, James, 255
Manganiello, Eugene, 53
Manning, Lillian, 255
Mansour, Ali, 41
Mariner 1, 105
Mariner 2, 105
Mariner 3, 105
Mariner 4, 105, 118
Mariner 6, 116, 120
Mariner 7, 116, 118, 120
Mariner 8, 116, 118, 120, 122, 123
Mariner 9, 116, 120, 122, 139, 154, 156
Mariner 10, 116, 131, 133
Mariner Program, 26, 29, 31, 34, 51, 54, 104, 108, 117, 129, 138
Mars Global Surveyor, 157
Marsch, Eckart, 152, 154
Marshall Space Flight Center (MSFC), vi, viii, 18, 25, 26, 29, 3132, 34, 3637, 43, 4448,
51, 52, 5455, 5960, 62, 66, 69, 7479, 85, 89, 107, 136, 174, 177180, 197, 206,
208209, 252253
Martin Marietta Corp., vii, 14, 86, 141, 144, 209, 229, 232, 235, 236, 240, 247, 248, 250
Martin, Richard, x, xii, 73
McAleese, James, 255
McCarthy, John F., Jr., 179, 181
McClure, Billy, 118
McDonnell Douglas Corp., 101, 229, 235, 236, 240, 242
Medium Launch Vehicle II (MLV II) contract, 236, 241
Mercury Program, 50
Merino, Fred, 2728, 84, 257
Meyers, Dale D., 239
Middendorf, William, 255
Miller, Pat, 255
Miller, Riley, 15
Miller, Robert, 73, 255
Milstar satellite, 193, 194
Milwitzky, Benjamin, 67, 88, 89
Minneapolis-Honeywell Co., 22, 46, 60, 63, 67, 79, 180, 194, 213, 226, 227, 237, 238
Minuteman missile, 12
Miramar Naval Air Station, 142
Index 281
Misichko, William, 255

Mission Integration Office (Shuttle/Centaur), 183
Mitchell, William, 20
Monnin, Carl, 255
Monzel, William, 65
Moore, Jesse, 167, 204, 206208
Mrazek, Willie, 38, 39
Muckley, Edwin, 107, 183, 255
Mulready, Dick, x, 42, 43
Multiprobe, 134, 136
Murray, Bruce, 159, 160, 174
MX-774, 8. See also Atlas (missile)
Mylar radiation shield, 150
N
NASA barter agreement, 237238
NASA Historical Reference Collection, xii
NASA Mixed Fleet Study, 233
National Advisory Committee for Aeronautics (NACA), 13, 15
National Oceanic and Atmospheric Administration, 240
National Security Directive 254, 233
Naugle, John, 123
Naval Research Laboratory (NRL), 3
Nechvatal, John, 255
Neilon, John, xii, 75, 141142, 256
Nelson, William, 231, 239
Nettles, J. Cary, xi, xii, 59, 62, 63, 65, 91, 93, 256
New York Times, 140, 187
Newell, Homer, iii, 32, 34, 53, 55, 82, 86, 88, 93, 99, 103, 104
Newton, Kenneth, 18, 44
Nicks, Oran, 3233, 44, 67, 95
Nieberding, Joe, xi, xii, 106, 128, 139, 144, 145, 163, 182, 183, 198, 208, 212, 218, 245, 255,
256
Nimbus satellite, 61
Nixon, President Richard, 101, 215
Norris, Chet, 120
North American Aviation, Inc., 14
North American Rockwell Corp., 101
nose fairing, 110, 131
Nova rocket, 51, 54
nuclear rocket, 5. See also Rover
O
ODear, Cecil, 255
ONeil, William, 69, 120, 257
Oberth, Hermann, 1, 5, 8
Office of Commercial Space Transportation, 222
Office of Management and Budget (OMB), 173, 174, 176
Office of Space Flight Programs, 18, 23
Ohio State University, 13, 14
Opp, Al, 152
Orbital Sciences Corporation, 192
Orbiting Astronomical Observatory (OAO), 108, 113, 121, 123, 136, 138
Orbiting Astronomical Observatory-2, 113, 116, 137
Orbiting Astronomical Observatory-3, 113, 116
Orbiting Astronomical Observatory-4, 116, 117
Ordin, Paul, 15
Orzechowski, Richard, 255
Ostrander, Don, 26, 31, 33, 42
Owara, Vern, 257
P
Palley, I. Nelvin, 86
parking orbit. See Centaur coast period
Patterson, James, 126, 226, 255
Patterson, William, 4, 12
payload (Shuttle), 198200
Payload Integration Plan, 193
Peery, David, 73
Penner, Rudolph, 231
Performance Trajectory Group (PTG), 6768
Perkins, Clay, 84
Perry, L. (Len) Charles, xii, 65, 257
Philadelphia Evening Bulletin, 97
Pickering, William, 82
Pioneer 4, 45
Pioneer 10, 116, 123125, 128, 130
Pioneer 11, 116, 124, 125, 128, 130, 131
Pioneer Venus, 108, 116, 133, 134, 136
Plain Dealer, 88, 231
planetary assists, 162, 175, 192, 208, 211, 212
Plum Brook Spacecraft Propulsion Research Facility (B-2), 73, 125127, 143, 146
Plum Brook Station, Sandusky, Ohio, 65, 73, 251, 104, 106, 121, 138, 146
Point Loma test facility, 20, 22, 65, 106
Index 283
Polaris missile, 37
Popular Science, 139
Porada, Ted, 255
Powell, Joel W., x
Prati, William, 117
Pratt & Whitney Aircraft Engine Co., 17, 18, 19, 26, 31, 32, 40, 42, 43, 47, 79, 89, 94, 155,
180, 194, 213, 214, 226, 227, 237, 238, 252. See also RL10 engine
Pratt, Perry, 40
Procasky, Ed, 255
Proctor, Richard, 65
Program Evaluation and Review Technique (PERT), 37, 66
Project Bee, 16, 74
Proof Flight (Viking mission), 140, 145, 147, 149
Propellant settling, 77, 108, 137, 159
Propulsion Systems Laboratory, 27
Proton (launch vehicle), 250
Q
Quarles, Donald, 18
R
Radcliffe, William, 4, 37
Ramo-Wooldridge Corp., 10, 60
Ranger Program, 33, 44, 47, 61, 82
Rasper, Colonel David, 215
RD-180 engine, 250, 251
Reaction Control System (RCS), 84
Reagan, President Ronald, 181, 233; administration of, iv, 222, 229, 231
Reed, Craig, 231
Reinberger, Robert, 255
Reshotko, Meyer, xi, 193, 256
Retargeting software, 136
Reusable Centaur Study, 170
Reynolds, Robert, 257
Richards, G. R., x
Riehl, John, 255
Rigi-Mesh, 42, 98
risk, relation to spaceflight, viiviii, 168, 170, 173, 176, 186, 189, 190, 198, 201, 204, 206,
207, 209, 210, 213, 215, 216, 218, 219
RL10 engine, x, 33, 54, 63, 72, 82, 85, 89, 96, 126, 180, 214, 227, 245, 246, 251; and 304
turbojet engine, 40; and Saturn, 2426, 32, 43, 6061, 98; diagram, 39; expander or boot-
strap cycle, 1820; failures, 246; ignition problem, 4243; regenerative cooling, 14; restart
capability, 93, 113, 126, 131, 133, 149, 150, 152. See also Pratt & Whitney Aircraft Engine
Company
RL10A-3-3, 89, 100

Robbins, William H. (Red), 182, 194, 196201, 207, 212, 256
Roberts, John, 255
Roberts, Roy, 68, 242, 245, 250
Rocketdyne Div., North American Aviation, Inc., 13, 75, 96, 98, 170, 238, 246, 252, 253
Rocketdyne engine (Atlas), 85, 246
Rocky, 196
Rohal, Robert, 255
Rosen, Milton, 3, 27
Rosenbaum, Mort, 35
Rosin, A., 68
Ross, Lawrence (Larry), xii, 64, 136, 155, 176, 179, 180, 194, 197, 201, 205, 206, 209, 210,
215, 217, 228, 255, 256
Rovenger, Ronald, 62, 63
Rover nuclear rocket, 50, 63
S
Salmi, Ray, 255
Sarokon, Dan, 79
Saturn (launch vehicle), iv, 23, 24, 25, 40, 43, 44, 48, 51, 54, 75, 80
Saturn C-1, 24, 25, 29, 31, 50, 52, 54, 55, 56, 61, 85
Saturn S-IV stage, 25, 26, 43, 56, 61, 82, 85, 98
Saturn V rocket, 24, 53, 85, 96, 139, 154, 169, 174, 175, 212, 252
SAX (Satellite per Astronomia X), 242
Schaefer, George, 255
Scherer, Lee R., 233
Schiaparelli, Giovanni, 154
Schirra, Walter, 11
Schneiderman, Dan, 120
Schriever, Bernard, 8
Schuler, Margaret, 255
Schwenn, Rainer, 152
Science, 167
SCORE (Signal Communication by Orbiting Relay Equipment) Project, 11
Scott, Ronald, 88, 89
Seaberg, John, 17, 18
Seamans, Robert, 52, 82, 88, 92
Seeholzer, Tom, 255
Semenchuk, Konnie, 255
Shaw, Lutha (Tom), 107, 165, 197, 212, 257
Shramo, Dan, 106
Shuttle Payload Integration Facility, 205
Index 285
Shuttle/Centaur: advantages, 176; cancellation, 168, 235, 237; fuel dump, 175, 185, 200, 206,
207, 216; mission patch, 169; rollout, 203205; safety, 170, 173177, 186, 189, 196, 198,
200202, 206, 207, 209, 215, 217, 253; weight issues, 208, 209
Shuttle/Centaur Project Office, 182, 191, 194
Shuttle/Centaur-1, 204206
Shuttle/Centaur-2, 205, 206
Shuttle-Centaur man-rating, 169, 178, 201
Siddiqi, Asif, 98
Silverstein, Abe, vi, ix, 12, 13, 18, 32, 33, 34, 99, 100, 237, 252; and direct ascent, 6667; and
Lewis management of Centaur, 5354, 57, 59, 62, 65, 7175, 77, 9293; and Saturn,
2325
Simon, Harlan, 59, 64, 65, 255, 256
Slayton, Deke, 93
Slone, Henry, 106
Sloop, John, ix, 5, 6, 15, 34, 35
Smith, Clyde Curry, 122
Smith, Dick G., 178
Smith, Floyd, 214, 255
Smith, Matthew, 257
SOHO (Solar and Heliospheric Observatory), 242
Solid Motor Assembly Building (SMAB), 141, 142
Solid Rocket Motors, 124, 142
Solomon, George, 66
Soviet Union, v, 3, 4, 11, 47, 98, 222; and space race, 104, 105, 134, 168
Space Commerce, 245
Space Plasma High Voltage Interaction Experiment (SPHINX), 145
Space Power Chamber, 71
space science, 103, 147, 152, 156, 162, 163, 165, 191, 192
Space Shuttle limitations, 169
Space Shuttle Main Engine (SSME), 255
Space Shuttle, iv, v, vi, 101, 102, 221, 222, 226, 228, 229, 231, 232, 233, 239, 243, 253
Space Taskforce Group, 169
Space Technology Laboratories (STL), 60, 66
Space Transportation System (STS). See Space Shuttle
space tug, 170, 171
Space, 5
Spacecraft Assembly and Encapsulation Facility, 161
Sparks, Brian, 52, 54, 55
Spurlock, Lewis Omer (Frank), xii, 107, 108, 183, 245, 256
Sputnik, 3, 11, 103, 222
Squires, Edward, 248
Stafford, Thomas, 93
Stages to Saturn, 96
Star Trek, 157
Star Wars, 205
Stofan, Andrew, 64, 84, 106, 107, 141, 163, 181, 182, 198, 213, 256
Stoll, James, 255
Strughold, Hubertus, 3
STS-6, 172
STS-7, 204
STS-51A, 204
Stuhlinger, Ernst, 252
Suntan 304 turbojet engine, 17, 18, 40
Suntan Project, 17, 40
Surveyor I, 91, 92
Surveyor II, 93
Surveyor III, 94, 95, 97
Surveyor IV, 94, 95
Surveyor V, 94
Surveyor Program, iv, 26, 29, 3134, 47, 51, 52, 55, 60, 61, 6668, 82, 83, 86, 8890, 96,
108, 128, 131, 179, 221
Surveyor, Sustaining Engineering and Maintenance (SE&M) contract, 65
Swavely, Jim, 256
Sycamore Canyon test facility, 10, 22, 34, 38, 44, 65, 106, 196, 209
Synor, Henry, 255
System Requirements Review, 183
Szabo, Steven V., xii, 84, 182, 183, 255
Szpatura, Steve, 255
T
Teledyne: avionics, 125; computer, 126, 130, 137, 142, 149, 184; guidance and control
systems, 104, 128
Teledyne Industries, Inc., 180, 194, 213, 226, 238
telemetry systems, 78, 79, 111, 113
Telstar-401, 246
Teren, Fred, 107
Terhune, Terry, 151
Thagard, Norman E., 202
thermal radiation shield, 143
Thiel, Walter, 56, 24
Thiokol Chemical Co., 83, 124
Thomas, Russ, 141
Thompson, Craig, 205
Thor-Able (launch vehicle), 239
Index 287
Thor-Agena (launch vehicle), 44, 61

Thor-Delta (launch vehicle), 231
Tiede, Chuck, 255
Tingley, Sandy, 255
Tischler, Adelbert O., xii, 26
Titan (launch vehicle), v, 141, 142, 189, 194, 224, 233, 247, 248, 250
Titan 34D booster, 172, 233
Titan I, 141
Titan II, 141, 153
Titan II-Transtage, 247
Titan III, 140
Titan IIIA, 141
Titan IIIC (launch vehicle), 86, 113, 136, 141
Titan IIIE, 139
Titan IV (launch vehicle), 143, 215, 232, 235, 247
Titan-Centaur 1, 140, 145
Titan-Centaur 2, 140, 153, 159
Titan-Centaur 5, 140, 159, 251
Titan-Centaur 6, 140
Titan-Centaur 7, 140, 159, 161, 163, 164
Tokmenko, Thomas D., 241
Torel, Bruce, 43
Tracking and Data Relay Satellite System (TDRSS), 172, 186, 189
trajectory analysis, 107, 108, 159, 183, 199
Transpace Carriers, Inc., 231, 233, 242
Transportation, Department of, 229, 231
Transtage, 86, 141, 171
Troyan, Barbara, 255
Truly, Richard H., 210, 219, 233
Truman, President Harry, 14
TRW, Inc., 124, 136, 209, 225
Tsiolkovskiy, Konstantin, 5
Tucker, Joel, x
tumble-recovery software, 137
Two-planet mission, 131, 132
U
U-2, 16
Ulysses spacecraft, 177, 180, 181, 191, 192, 194, 196, 202, 204, 207, 208, 211, 212
United Technologies, Inc., 141, 142, 214
upper stage rockets, iv

Urey, Harold, 13
V
V-2 rocket, 4, 5, 7, 8
Valentine, Jack, 68
Vandenberg Air Force Base, 61
Vanguard Program vi, 27, 44, 75, 79
Vega rocket, 27
Venera spacecraft, 134
Venus Radar Mapper (VRM). See Magellan spacecraft
Vertical Integration Building (VIB), 141, 142, 158
Vietnam War, 140
Viking 1, 140, 145, 155, 156, 159
Viking 2, 140, 156
Viking Dynamic Simulator (VDS), 145
Viking Mars lander, 96
Viking spacecraft, 108, 139, 151, 154, 157, 179, 204
Vivian, Weston E., 86
von Braun, Wernher, iii, viii, 3, 5, 7, 8, 29, 43, 52, 139, 174; and C1-Agena, 5456; and
Centaur management, 3139, 4749; and management at the Cape, 4546; and RL10
management, 43, 60; and Saturn upper stages, 2425
von Neumann, John, 10
Voyager 1, 140, 159, 160, 162
Voyager 2, 140, 159, 160, 162164
Voyager spacecraft, 108, 139, 157, 158, 161, 179, 204
W
Waldrop, M. Mitchell, 167
Walker, David M., 202, 204, 209
Wall Street Journal, 231
Wallops Flight Research Station, 27, 59, 234
Webb, James, iv, vii, viii, 33, 53, 56, 57, 98, 99, 100
Weisberg, Norm, 255
Wentink, Richard, 68
Wentworth, Carl, 68, 255
Weston, Maynard, 255
Weyers, Vernon, 194, 205, 206, 210, 216, 239, 256
Widra, Gil, 255
Wiesner, Jerome, 56
Wilcox, Fred, 54
Wilcox, Ward, 54
Wilford, John Noble, 140
Index 289
Wilmot, Allan, 257

Wilson, Chuck, 257
winds, upper altitude, 124, 128, 130. See also ADDJUST steering program
Winkler, Martin, xii, 189, 193, 196, 216, 217, 219, 243, 257
Winslow, Paul, 106
Woelfle, Richard, 255
Womack, James, 44, 238, 256
Wright-Patterson Air Force Base, 17, 18
Wynne, Michael, 247, 248
Y
Yardley, John, 178, 179
York, Herbert, iii, 13, 22
Young, John, 207
Z
Zeenkov, Seymour, 257
Ziemba, Edmund R., 120, 130, 255, 256
Zimmerman, Art, xii, 66, 68, 69, 255, 256
Zucrow, Maurice, 6

Centaur SP 4230

Uploaded by

Copyright:

Available Formats

Centaur SP 4230

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Centaur SP 4230

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What challenges did Centaur face early on?

What challenges did Centaur face early on?

31276-*front matter 3/7/05 3:18 PM Page 1

Virginia P. Dawson and Mark D. Bowles

The NASA History Series

National Aeronautics and Space Administration

Library of Congress Cataloging-in-Publication Data

Chapter 1: Centaurs Origins in Atlas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

Chapter 2: Marshalls Unruly Wards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31

Chapter 3: Abes Baby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59

Chapter 4: Heavy Lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103

Chapter 5: The Giant Titan-Centaur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139

Chapter 6: Centaur Reborn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167

Chapter 7: Eclipsed by Tragedy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189

Chapter 8: Like a Phoenix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221

List of Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255

About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259

The NASA History Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261

R i c h a rd Ma rtin, Larry Ross, Ed w a rd Bock, and John Neilon. History colleagues Jo h n

Centaurs Origins in Atlas 1

Centaurs Origins in Atlas

2 Taming Liquid Hydrogen

Centaurs Origins in Atlas 3

Diagram: Atlas-Centaur. (NASA E2283)

4 Taming Liquid Hydrogen

Centaurs Origins in Atlas 5

6 Taming Liquid Hydrogen

Centaurs Origins in Atlas 7

The Thin-Skinned Atlas

8 Taming Liquid Hydrogen

Centaurs Origins in Atlas 9

10 Taming Liquid Hydrogen

Centaurs Origins in Atlas 11

12 Taming Liquid Hydrogen

Centaurs Origins in Atlas 13

Early Liquid-Hydrogen Initiatives

14 Taming Liquid Hydrogen

Centaurs Origins in Atlas 15

16 Taming Liquid Hydrogen

Centaurs Origins in Atlas 17

Centaur and Project Suntan

18 Taming Liquid Hydrogen

Centaurs Origins in Atlas 19

Upper-Stage Rocket Engine, 126.

20 Taming Liquid Hydrogen

Centaur Program, 87th Congress, 15 and 18 May 1962, 67.

Centaurs Origins in Atlas 21

22 Taming Liquid Hydrogen

Centaurs Origins in Atlas 23

Centaur and Saturn

24 Taming Liquid Hydrogen

Centaurs Origins in Atlas 25

Caught Between DOD and NASA

26 Taming Liquid Hydrogen

Centaurs Origins in Atlas 27

28 Taming Liquid Hydrogen