AIAA-99-2823
NASA/TM--1999-209278
Development
Solid Rocket
Bruce
M. Steinetz
Glenn
Research
Patrick
Center,
H. Dunlap,
Modern
of Thermal
Barriers
Motor Nozzle Joints
Cleveland,
Jr.
Technologies
Prepared
for the
35th Joint Propulsion
Corporation,
Conference
cosponsored
by the AIAA,
Los Angeles,
California,
National
Space
Aeronautics
Administration
Glenn
Research
June
1999
Ohio
Center
and
ASME,
June
and
Middleburg
20-24,
Heights,
Exhibit
SAE,
1999
and
ASEE
Ohio
for
Acknowledgments
The authors
Bruce Bond's
greatly
acknowledge
Doug
Frost's
and Steve
Hicken's
(Thiokol)
thermal
(Albany-Techniweave)
assistance
in fabricating
the thermal barriers;
and Lauren Yee's (NASA) assistance
in test support.
barrier
test results;
and Tom Doeberling's
Trade names or manufacturers'
names are used in this report for
identification
only. This usage does not constitute
an official
endorsement,
either
Aeronautics
expressed
or implied,
by the National
and Space Administration.
Available
NASA Center
7121 Standard
for Aerospace
Drive
Hanover,
MD 21076
Price Code: A03
Information
from
National
Technical
Information
Service
5285 Port Royal Road
Springfield,
VA 22100
Price Code: A03
AIAA-99-2823
DEVELOPMENT
OF THERMAL
BARRIERS
FOR SOLID
MOTOR NOZZLE JOINTS
Bruce
National
M. Steinetz*
Aeronaulics
and Space
Glenn
Research
Cleveland,
Administration
Center
Ohio
44135
H. Dunlap,
Patrick
Modern
Technologies
Middleburg
Jr.;
Corporation
Heights,
Abstract
Ohio
600
44130
°F. Carbon-6
rocket
The
joints
Shuttle
Space
sealed
are
5500+°F
solid
using
combustion
from the seals
motor
casc
()-ring
conventional
gases
by thick
fill compounds
rocket
seals.
layers
of insulation.
The
arc used to fill the joints
Special
assembly
in the joint-fill
design
joints
hot gas penetration
The current
incorporates
most)
O-rings
current
motor
I out
to
()-ring.
safety,
Though
evidence
in extensive
reviews
solid
motor
rocket
improve
reliable
heft)re
resuming
manufacturer
braided
carbon-fiber
tormed
to determine
thermal
barriers
acetylene
thermal
(5500
°F),
types
and
to
representative
seals
research
drop,
of NASA
of the
of an oxy-
of the 5500
burn
time.
braid
architectures,
measure
Tests
were
performed
denoted
the temperature
in a compressed
state
oxyacetylene
torch.
barriers
excellent
drops
were
through
temperatures
within
their
drop across
when
and
insulators
diameter
Viton
from
°F
motor
thermal
barrier
and Carbon-6,
Carbon-6
causing
seals
do not
jetted
tothe
arc found
°F.
barrier
Gas
several
thereby
materials
transient
seal
designs
that
seal
as compliant
and pressure
require-
3 showed
that these
a few seconds
Shuttle
Solid
a thermal
de-
Previous
when
rocket
to evolve
barrier
motor.
the braided
for
sub-
gases
use
that
Thus,
rope
at extreme
temperatures.
The Space
Shuttle
joints
5500+
reusable
are scaled
the seals
solid rocket
motor
usingconventional
°F combustion
from
insulation.
than
led to
of newly
wires.
and Dunlap
were considered
into
out
rope
°F) com-
designs
hot 55(X)+ °Fcornbustion
in the Space
design
braided
temperature
last for more
seal
engine
braided
Steinetz
gases
by thick
Special.joint-fill
are kept
layers
of phenolic
compounds
seals.
two stages
of protection
a direct
llowpath
reaching
the seals.
charring
due to parasitic
leakage
seals
paths
to ['ill the
l]ow-patb
of the 900-psi
Occasionally,
or rubber
are used
a direct
these
seals.
a safe distance
to prevent
Normally,
(RS RM)
O-ring
.joints in the insulation
to the
are enough
hot gases
have
from
experienced
thai open
up in the
were
limit of
Branch,
Copyright © 1999 by the American Institute of Aeronautics and
Astronautics, Inc. No copyright is asserted in the United States under
Title 17, U.S. Code. The U.S. Government has a royahy-free license to
exercise all rights under the copyright claimed herein for Governmental
Purposes, All other rights arc reserved by the copyright owner.
1
American
by
spread
slag,
to both seal and serve
extremely
other
to prevent
thermal
temperature
*Senior Research Engineer. Mechanical Components
Member AIAA.
Resean:h Emgineer.
of an
temperature
25(_) to 2800
hot
( 1500 to 2000
seals
under aggressive
away
to
the barrier
to the flame
of the thermal
()-ring
rocket
and along
subjected
Carbon-3
1/4" downstream
the downstream
on two
Carbon-3
thermal
temperature
O-rings.
and supcralloy
the ability
1,2 However,
The
than the solid
blocked
gas turbine
fibers
yielded
over 6 rain,
longer
and
The
at
the
gap defect,
to the downstream
of rope
demonstrated
assembly
times
oversized
in advanced
ceramic
solid rocket
motor
combustion
temperatures.
Thermal
barriers
braided
out of carbon fibers endured
the flame tbr
three
gas
hot
gas
simulating
fill time.
cavity
The need tor high temperature
pliant
ments.
each
flame
firing,
in subscale
to hot
Introduction
mounts
tests were per-
through
to the
rocket
incoming
flow,
protection
veloped
barrier.
Burn
to burn
exposed
NASA
temperature
barriers.
the time
results
are working
for several
,jet
incoming
offering
subjected
joint
the
the development
by implementing
burn-resistance,
when
torch
flight.
and a thermal
test results
to
not threaten
()-ring
Thiokol
joint design
design
presents
and resiliency
does
the
In the
hot gas
of hot gas to the wiper
J-leg
This paper
flow,
the condition
the nozzle-to-case
a more
joint
(inner-
experience
downstream
reduced
well
when
defects
compound.
of 7 motors
l 1-see.
22(X) °F in an intentionally
and wiper
secondary
maxinmm
an
joint-
nozzle-to-case
and polysulfidejoint-fill
design,
the wiper
primary,
through
fl)r
extremely
tests
barrier
in the insulation
compound.
°F
perlormed
motor
away
prevent
a direct flowpath
to the seals. On a number
of
occasions,
NASA
has observed
in several
of the rocket
nozzle
"char"
3200
assembly
are kept a safe distance
ROCKET
Institute
of Aeronautics
and
Astronautics
joint-fill compounds
during rocket operation. Inspection
during disassembly
of Space Shuttle solid rocket motor
nozzle joints from RSRM-44
and RSRM-45
revealed
()-ring erosion of Joint 3 primary O-ring seals 4 (Fig. 1).
Subsequent
improvements
in joint-fill
compound
application-techniques
have apparently overcome the Joint
3 charring problem. However, a number of nozzle joints
including the nozzle-to-case
joint and Joint 2 continue to
show hot gas penetration through the joint fill compound.
The current nozzle-to-case
joint design incorporates
primary, secondary and wiper (innermost)
()-rings and
polysulfidc
joint-fill compound.
In the current design,
I out of 7 motors experience hot gas to the wiper ()-ring.
Though the condition does not threaten
motor safety,
evidence of hot gas to the wiper O-ring results in extensive
reviews belore resuming flight. NASA and solid rocket
motor manufacturer
Thiokol are working to improve the
nozzle-to-case joint design by implementing a more reliable
J-leg design and a thermal barrier, and eliminate the jointfill compound
(Fig. I). The J-leg is molded into the
insulation and contacts the mating surface of the adjoining
element. Rocket pressurization
acts to further preload the
J-leg increasing its effectiveness.
The basic J-leg design
has been applied successfully
to fixing the field joints in
the redesign el'fl+rt lollowing the Challenger
accident. s
The thermal barrier, compressed
between the J-leg and
adjoining clement, is intended to resist any hot gases the
J-leg does not block and prevent them from reaching the
wiper ()-ring. The braided carbon thermal barrier being
developed at NASA Glenn is the leading candidate based
on the results presented herein.
The thermal barrier for the Shuttle solid rocket motor
has unique requirements,
others:
including the following,
amongst
I. Sustain extreme temperatures
(2500 to 55(X) °F)
during solid rocket motor burn (2 min and 4 see.)
without loss of integrity.
2. Drop incoming gas temperatures (up to 3200 °F) in
the joint to levels acceptable
to Viton O-rings
(<600 °F, short-term) to prevent ()-ring damage+
including char and erosion.
3. Exhibit some permeability
to permit the joint
cavity (between thermal barrier and O-ring) to reach
chamber pressure (900-psi) in acceptable time.
4. Exhibit adequate resiliency/springback
to accommcx:late limited joint movement and manufacturing
tolerances in these large (8.5 It. diam.) nozzle
segments.
5. Diffuse/spread
incoming narrow (0.08 in. diam.)
hot gas jets to reduce their damaging effects on the
downstream ()-rings.
6. Block hot slag (i.e., molten alumina, etc.) entrained
in gas stream from reaching O-rings.
American
Institute
Steinetz and Dunlap 3 pertormed a number of tests on
0.125- and 0.2(X)-in. diameter braided carbon-fiber thermal
barriers demonstrating
that they met the burn-resistance,
permeability,
and resiliency criteria.
The main objective of the current study is to fully
characterize
two braided carbon fiber thermal barrier
designs (denoted Carbon-3 and Carbon-6)
by assessing
their transient thermal response when subjected to a high
temperature torch and by characterizing
their permeability,
resiliency, and burn-resistance.
Thc Carbon-6 design is
currently being tested by both NASA and Thiokol for the
nozzle-to-case
joints of the Shuttle solid rocket motor.
Subscale rocket "char" motor tests were performed to
assess the thermal barrier's (Carbon-6) thermal response
and heat resistance under actual rocket conditions.
Test Apparatus
Thermal
Barrier
and Procedures
Specimens
Carbon-3 and Carbon-6 were subjected to burn, temperature drop, flow, and compression
tests. Carbon-6
was also tested in a subscale char motor. Limited testing
was performed on the Carbon-4 design. Table I summarizes
the relevant architecture parameters for the thermal barrier
designs that were tested.
All thermal barriers wcrc composed ofa uniaxial corc
of fibers overbraided with various numbers of sheath layers.
The Carbon-6 design had ten sheath layers and a 0.26-in.
diameter. Carbon-6 had good flexibility and compliance
properties
because it was braided with a more open
architecture. The Carbon-3 design had a 0.20-in. diameter
and was made with a large degree of uniaxial core fibers
overbraidcd
with five sheath layers. Carbon-3 was a tight
braid that was not as flexible as Carbon-6. Carbon-4 had
4.4×10 -4 in. (I I Jam) pitch-based
Amoco
P25 fibers in its
core to evaluate core fiber diameter effects on performance,
while the core fibers of all the other carbon thcrmal barriers
were 2.76×10 -4 in. (6.9 Jam) PAN-based Grafil type 34700 fibers. PAN-based Thornel T-300 carbon fibers with
a 2.8× 10-4 in. (7 _m) diameter
all the thermal
Thermal
Barrier
were used in the sheaths of
barrier designs.
Porosity
Measurements
To assess thermal barrier porosity while under compression, samples of the Carbon-3 and Carbon-6 designs
were examined in a compressed state using a photographic
stereomicroscope.
Four !/2-in. long specimens of I'x)th types
of thermal barriers were prepared and weighed using a
precision electronic balance. The exact length of each specimen was measured using vernier calipers. Each specimen
was then clamped between two steel plates and subjected to
a 2()_: compression. While the specimens were compressed,
a light layer of cyanoacrylic glue was applied to the surface
2
of Aeronautics
and Astronautics
of eachspecimen
so thattheywouldmaintaintheir
compressed
shape
uponremoval
fromthefixture.
Fourspecimens
wereexamined
forbothCarbon-3
andCarbon-6.
Bothends
ofeach
specimen
were
examined
andphotographed
atI0Xinthemicroscope
sothateight
cross section photos were examined for both thermal
barrier designs. Each cross section assumed an ellipsoidal
shape in its compressed
state. The dimensions
of each
ellipse were measured
using vernier calipers.
These
dimensions were then used to calculate the cross sectional
. area of both ends of each specimen. An average cross
sectional area was calculated
tor each specimen
and
multiplied
by the specimen
length to determine
the
specimen volume. Specimen density was then calculated
by dividing the weight of the specimen by its volume. An
average density at 20% compression
was found for both
Carbon-3 and Carbon-6 by averaging the densities of the
four specimens
of each design. The porosity of each
thermal barrier design at 20% compression was calculated
using the following
Porosity
relationship:
= I- (gThe,',,,al Barrier/PCarbon
Fiber)
In this relationship, the density of each thermal barrier
design was divided by the density of an individual carbon
fiber (0.064 Ib/cu.in.). Thus, a thermal barrier design
would have a porosity of zero if it had no gaps and assumed
the density of an individual fiber.
Burn Tests
A screening test was developed to evaluate thermal
barrier burn resistance under sire ulated rocket motor combustion temperatures
(5500 °F) by aiming a "neutral"
flame 6 of an oxyacetylene
welding torch at the center
section of a 4-in. thermal barrier specimen. In these tests,
the amount of time required to completely cut through the
specimen
measured
was measured.
Time for cut-through
was
from the instant the flame touched the specimen
until the specimen was completely cut into two separate
pieces. A detailed description
and an illustration
of the
fixture used to perform these tests can be found in the
paper by Steinetz and Dunlap. 3
Temperature
Drop Tests
A test fixture was developed
to measure
the temper-
ature drop across and along the thermal barriers in a compressed state when subjected to the neutral flame of an
oxyacetylene
torch simulating
rocket
temperatures
(Fig. 2). Flow was drawn through the thermal barrier using
a vacuum roughing
pump to lower pressure
on the
downstream
side of the thermal barrier while leaving the
upstream side at ambient conditions. Flow through the
thermal barrier was measured using a flow meter positioned
between the fixture and the roughing pump. The volume
American
Institute
downstream
of the thermal barrier was an enclosed
plenum
chamber sealed by an O-ri ng between the bottom plate and
a top plate. The thermal barrier was compressed
at 20%
linear compression.
Other compressions
are possible by
placing shims under the thermal barrier. The fixture was
made out of phenolic
insulation
having low thermal
conductivity that simulates the solid rocket motor insulating
material and minimizes parasitic heat loss.
The torch flame was applied to the thermal barrier to
simulate a leak path of hot gases through the nozzle joint.
The flame with temperatures up to 32(XJ°F was positioned
on a small area of the thermal barrier. An "'iris plate" with
a 0.084-in.
diameter
hole concentrated
a "laser-like"
column of flame onto the thermal barrier, simulating a hot
gas jet flowing through the rocket nozzle joint. The iris
plate was positioned about I/4-in. away from the specimen.
The jet was directed at the center of the specimen both
span- and height-wise.
To measure the surface temperature distribution along
the thermal barrier span, thermocouples
were placed on
both the upstream (hot)and downstream (cold) sides. The
thermal barrier specimen sat between these two rows of
thcrmocouples
in a 0.040-in.-deep
groove. The thermocouples measured how the flame spread along the thermal
barrier, how much temperature drop occurred across the
thernnal barrier, and how heat was conducted along its
length. The fixture was instrumented
with seven thermocouples upstream of the thermal barrier and eight downstream Ihermocouples.
On the upstream side, the center
Type B thermocouple
was placed directly in line with the
center of the hole in the iris plate so that it measured the
hottest flame temperature
at the surface of the thermal
barrier. Type B thermocouplcs
were then positioned
I/4-in. on either side of the center thcrmocouple
(Fig. 2).
The remaining four thermocouplcs
on the hot side were
Type K thermocouples,
and they wcrc placed 1/2 and I in.
on either side of the center thermocouple.
Seven of the
eight Type K thermocouples
downstream
of the thermal
barrier were spaced so that they were directly in line with
those upstream of the thermal barrier. The remaining
Type K thermocouple
was positioned
I/4 in. (approxinmtely one thermal barrier diameter) downstream
of the
thermal barrier in line with the center thermocouple
and
measured the bulk air temperature.
Thermocouple
selection. Fine gage wire open-bead
thermocouples were used to quickly and accurately measure
changes in the surface temperature distribution along the
thermal barrier. The time constant and response rate of a
thermocouple
is controlled by the size of its wires and the
diameter of the junction ball that is lbrmed between the
wires. The wire diameters used lot the Type B and Type K
thermocouples
were 0.010 and 0.0125 in., respectively. A
typical thcrmocouple
junction ball has a diameter about
50% larger than the wires in the thermocouple.
Calculations
3
of Aeronautics
and Astronautics
ofthetimeconstants
forjunctionballswithadiameter
of
0.015 to 0.019 in. showed that these thermocouples
have a time constant of about 1/2 sec.
would
Pressure/Flow
Transducers.
An absolute pressure
transducer measured the pressure upstream of the thermal
barrier while a di fferential transducer measured the pressure
drop across a specimen. Flow through the thermal barrier
was measured using a 0 to 100 SLPM flowmeter. Data was
acquired from all of this instrumentation
at a sampling rate
of 10 Hz using Keithley data acquisition hardware and
Labtech Notebook software.
For each test, a 5-in. thermal barrier specimen was
prepared and installed into the groove in the fixture. The
14 thermocouples
that measured the surface temperature
along the specimen were slipped into the outer sheath
layer of the thermal barrier and adjusted so that they were
spaced properly. To prevent parasitic leakage, the plenum
chamber ()-ring was then positioned so that it was snug
against the ends of the thermal barrier. The vacuum pump
was turned on for several minutes, to cause the pressure
drop and to achieve a steady flow rate through the specimen
before applying the torch. The oxyacetylene
torch was
adjusted until a neutral flame was formed. The torch was
slid along a machined
groove until it was properly
positioned in front of the hole in the iris plate. The torch
was left on the specimen for 30 or 60 sec. and then pulled
away from the fixture and shut off. Sometimes repeat tests
were performed
on the same specimen to examine the
effects of repealed |'lame exposures. Torch nozzle spacing
to the iris plate proved to be important in controlling
the maximum
hot side temperature
without melting the
center Type B thermocouple
(platinum-rhodium,
Tmclt =
3308 °F). Torch spacings for Carbon-3 and Carbon-6
were 0.265 and 0.160 in. respectively.
Flow Tests
Flow tests were perR_rmed on the thermal barriers in
a high temperature
flow and durability
test rig shown
schematically
in Fig. 3. The test rig is capable of operating
at temperatures
from room temperature to 1500 °F, pressures between 0 and I(X) psig, and flows of 0 to 3.5 SCFM
(standard cubic feet per minute, conversion
I SCFM =
28.3 SLPM). Spccimcn length was 7.50-!-_0.05 in., and thc
thermal barriers were mounted into a groove in the piston.
The free ends of the specimens were joined together in the
piston groove using a I/4 in. lap joint. Preload was applied
to the specimens through a known interference fit between
the thermal barrier and the cylinder inner diameter. To vary
the amount ofpreload, the interference fit was modified by
mounting different thicknesses
of stainless steel shims
behind the specimen in the piston groove. During flow
testing, hot pressurized
air entered at the base of the
cylinder and flowed to the test specimen that sealed the
annulus created by the cylinder and piston walls (0.007 in.
American
Institute
radial gap). The durability of the thermal barriers at high
temperatures
was examined by subjecting them to scrub
cycles in which the piston and thermal barrier were
reciprocated
in the cylinder.
Flow data was recorded before scrubbing at temperatures of 70 and 500 °F and after scrubbing at 70, 500,
and 900 °F. Specimens were subjected to ten scrub cycles
at 500 °F. At each temperature, flow data was recorded at
pressures of 2, 5, 10, 30, 60, 90, and 100 psid (or as high
as could be recorded within the limits of the flowmeters)
with the downstream pressure at ambient pressure. Primary
and repeat flow tests were perlormed on the Carbon-3 and
Carbon-4 designs for a diametral or linear compression
of 0.040 and 0.050 in. (20 and 25% linear compression)
and on the Carbon-6 design at linear compressions
of
0.052 and 0.065 in. (20 and 25% linear compressions).
A
detailed description of the hardware and procedure used to
perform these tests can be lbund in the papers by Steinetz
et al. I and Steinetz and Adams. 2
Compression
Tests
Compression
tests were performed
to determine
thermal barrier preload and resiliency behavior at room
tcmpcrature
using a precision linear slide compression
test fixture shown schematically
in Fig. 4. A 1 1/2-in. long
specimen was loaded into a stationary grooved specimen
holder, and an opposing plate was compressed against the
specimen. Stainless steel shims were placed in the groove
behind the specimens
to vary the amount of linear
compression.
The amount of compressive
load on the
specimen was measured versus the amount of compression.
Multiple load cycles were applied to the specimen belore
the preload data point was recorded to remove effects of
the hysteresis and permanent set that accumulate with load
cycling of the specimens. Most permanent
set occurred
within the first tbur load cycles. A pressure sensitive film
mounted on the opposing plate was used to determine the
contact width of the specimen as it was compressively
loaded. The footprint length (nominal I in.) and width at
the end of the fourth load cycle were used along with the
measured load versus compression
data to calculate the
estimated prcload and residual interference corresponding
to a given linear crush value. I Residual interference
is
defined as the distance the specimen will spring back while
maintaining a load of at least I Ib/in. of specimen.
Compression
tests were per|ormed on the Carbon-3
and Carbon-6 designs to determine the specimen preloads
corresponding
to the linear crushes
used in the flow
experiments. Tests were performed at compressions of 20,
25, and 30% of each specimen's overall diameter. Primary
and repeat compression tests were performed. The hardware
and procedure used to perform these tests are described in
detail by Steinetz et al.I
4
of Aeronautics
and Astronautics
Subscale
Rocket
"Char"
Motor Tests
As part of the development
process of the thermal
barrier, Thiokol Corporation performed tests using a subscale (701 bm) rockct "char" motor. In these tests, the NASA
Carbon-6
0.260-in.
cross-sectional
diameter
thermal
barrier impeded hot gas flow through an intentional circam ferential defect between rocket-case insulation blocks.
The thermal barrier compression was 20%. The insulation
blocks were modi fled to accommodate a 5 1/8-in. diameter
thermal barrier. The 0.060-in. defect was much larger than
any defects that would normally lorm through the gap-fill
material in the actual rocket nozzle joint, but this size was
chosen to force gas flow through the thermal barrier under
very extreme conditions. Burning solid rocket propellant,
the rocket fired for I I sec. and generated 900 psi pressures
and 5000 °F (estimated) chamber temperatures.
Hot gas
l'lowed to the thermal barrier while upstream and downstream temperatures
and pressures were recorded. The
char motor incorporated an outboard plenum chamber, or
reservoir, to simulate the volume (80 in. J ) between the
thermal barrier and the Viton ()-ring seals. This reservoir
ensured that flow would pass through the thermal harrier.
The reservoir started at ambient pressure and then quickly
reached chamber pressure, simulating the actual RSRM
,joint fill-time. After the volume between the thermal barrier
and Viton ()-ring pressurizes
in the rocket nozzle joint,
charring risk to the Viton ()-ring is virtually eliminated.
Results
Burn Tcsl Results
The amount of time to burn through each type of
thermal barrier is shown in Fig. 5. In this figure, the
number of specimens that were tested is given next to the
name of each thermal barrier type, and the average burnthrough time is found above each bar. As shown previously
by Stcinctz and Dunlap, 3 carbon fiber thermal barriers
were the most burn-resistant.
Figure 5 summarizes
the
earlier tests done on I/8-in. diameter stainless steel rods.
Viton ()-rings. and all-ceramic
braided rope seals. It also
shows the burn times of the I/8-in. diameter (Carbon-1,
Carbon-2.
and Carbon-2A)
and 0.200-in.
diameter
(Carbon-3 and Carbon-4) carbon thermal barriers as well
as new data on the burn time of the 0.260-in. diameter
Carbon-6 design. The I/g-in. diameter designs all endured
the 55(X) °F oxyacetylene
torch tor about 2 rain, Even
more impressive burn times were seen h)r the 0.200-in.
diameter designs at about 6-1/2 rain. This is more than
three times the Shuttle solid rocket motor burn time of
2 min. 4 sec, However. an increase in diameter to 0.260 in.
did not produce an increase in burn time. Carbon-6 at
0.260 in. in diameter
had a similar burn time to the
0.2(X)-in. diameter designs at about 6-1/2 rain. Like the
other carbon thermal barriers. Carbon-6
was soft and
flexible after removal from the flame, even in the area
and Discussion
Thermal Barrier Porosity Measurements
Measured
values for thermal barrier density and
porosity at 20_ compression
are presented in Table II lbr
the Carbon-3 and Carbon-6 thermal barrier designs. A
20% compression
level was chosen, as this is the compression level selected lbr the nozzle-to-case
,joint thermal
barrier. The densities/porosities
of braided structures arc
important
for understanding
their thermal
and flow
response characteristics.
Carbon-3 had a higher density (0.041 Ib/cu.in.) and a
lower porosity (0.37) than did Carbon-6 (0.032 Ib/cu.in.
and 0.50, respectively).
This can be attributed to the differences in braid architecture
between these two designs
as shown in Table I. Carbon-3 had a core composed of ten
uniaxial 12K yarns of Grafi134-700
carbon fibers-a large
fraction of its cross-section,
while Carbon-6 only had one
12K yarn in its core. Carbon-6 had ten sheath layers of
braided carbon fibers, while Carbon-3 only had five layers.
Carbon-6 also had a lower sheath braid angle and fewer
carriers per sheath layer to produce a softer, more flexible
thermal barrier. Because the uniaxial fibers in the core pack
together much better than the braided fibers that cross over
each other in the sheath, the Carbon-3 design with a
American
greater percentage of core fibers is naturally more dense
and less porous. Steinetz and Dunlap 3 showed previously
that the density of a braided carbon thermal barrier was
inversely related to the number of sheath layers.
affected by the flame, with no evidence of charring or
melting. All of the non-carbon specimens showed signs of
charring or melting after removal from the flame, and
many became very brittle in the area that was burned.
The similarity in burn time between Carbon-6 and thc
smaller-diameter
Carbon-3 and Carbon-4 thermal barriers
is believed to be related to the difference
in porosity
between these designs. As shown in Table II, Carbon-6 is
more porous than Carbon-3 even in a compressed
state.
Steinctz
and Dunlap 3 theorized
that the mass-loss
mechanism during the oxyacetylene torch tests was carbon
oxidation.
Depending
on material type, carbon fibers
begin to oxidize at temperatures
in the range of 6(X) to
900 °F. 7"9 The oxyacetylene
torch burning at 5500 °F is
hot enough to cause oxidation to occur, but too cool for
carbon sublimation that occurs at 6900 °F. 10 It is believed
that the looser, more porous braid of Carbon-6 allowed
more of the hot, oxidizing torch flame to pass through it.
This allowed oxidation to occur more rapidly in the
innermost
fibers of Carbon-6 than in the less porous
Carbon-3 design. Even though there were more carbon
fibers in the larger Carbon-6 design, they were cut through
more quickly because they were exposed sooner to hot,
oxidizing gases. These results indicate that burn/oxidation
5
Institute of Aeronautics
and Astronautics
resistance
isdependent
onboththermal
barrierdiameter
andporosity.
Products
of combustion
in thesolidrocketmotor
include
liquidalumina
(A1203
jandgaseous
CO,CIO2,CI,
HCI.andH.,,noneof whichareoxidative.
Hence,
it is
believed
thattheneutralflameinambient
air(oxidizing)
isaconservative
(i.e.,moreaggressive)
environment
for
performing
material
screening
burntests.It isexpected
thatoxidation
rateswithinthercx:ket
environment
willbe
slowerthanthose
exhibited
herein.
Temperature
Drop Test Results
Temperature
drop tests were performed
on the
Carbon-3 and Carbon-6 thermal barrier designs using the
test fixture described that measured the temperature drop
across and along the thermal barrier in a compressed state
when subjected to the flame ol'an oxyacetylene
torch. Figure 6 shows temperature
versus time traces for a test
performed on a Carbon-3 specimen. Data recorded from
the center thermocouple
and the three thermocouples
to
the right of center on both the hot and cold sides of the
specimen are presented. Data from the thermocouples
to
the left of the center thermocouple
is not shown in this
figure for clarity. In general, the left and right sides
produced symmetric data. Also shown in the figure is the
temperature
trace from the "cold bulk" (Tbulk) thermocouple that measures the air temperature 1/4 in. downstream
of the specimen. For sensitivity purposes, we moved the
Tbulk thermocouple
spatially to see if we were missing
any local "'hot-streaks,'" and we did not find any. Figure 7
shows temperature
traces for a test performed
on a
Carbon-6 specimen.
Examining Figs. 6 and 7, it can be seen that the center
thermocouple
on the hot side (Tho I) and the center
thermocouplc
on the cold side (Tcold) of the thermal
barrier each recorded the hottest temperatures
on their
respective sides. This is expected as these thermocouples
are directly in line with the hottest part of the torch flame
as it passes through the hole in the iris plate. These figures
also show that the temperature got progressively
cooler
from the center thermocouple
to the R l, R2, and R3
thermocouples
on the hot and cold sides of the specimen.
This was also expected as the temperature decayed with
movement further away from the center heat source.
Figures 6 and 7 show that there was a lag between
increases in temperature on the hot and cold sides of the
specimen.
When the torch was applied to the thermal
harrier, the hot side thermocouples
instantly registered the
increase in temperature.
The insulating properties of the
thermal barrier delayed heat conduction to the cold side,
so the cold side thermocouples
did not register an increase
in temperature
until several seconds after the torch was
applied. The cold side temperatures measured were signiticantly lower than the hot side temperatures,
as will be
American
Institute
discussed
below. Alter the torch was pulled away from the
specimen, the hot side thermocouples
instantly showed a
decrease in temperature.
The cold side thermocouples,
though, continued to increase lor 3 to 5 sec before beginning
to decrease in temperature. Comparing the hot side temperatures in Figs. 6 and 7, one notes fluctuations in temperature
tor Carbon-6 but not Carbon-3. The origin of this fluctuation
is unclear at this point, but we could find no system source
of the variation (e.g. thermocouple
integrity, etc.).
Figure 8 shows the temperature drop across specimens
of Carbon-3 and Carbon-6 for flame applications of-30 sec.
The temperature
drop was calculated as the difference
between the temperature
recorded by the hot side center
thermocouple
and the cold side bulk temperature (Tbulk).
Over the 30-see. torch applicatiom the temperature drop
across the Carbon-3 specimen dropped from a high of
2870 to 2680 °F by the end of the test. This drop was
caused by a steady rise in the cold side bulk temperature
while the hot side temperature remained nearly constant.
Carbon-6 exhibited
a temperature
drop in the range of
2980 to 2600 °F. The uneven nature of the Carbon-6 trace
is duc to fluctuations
in the hot side temperature, as noted
above. As shown by these figures, both Carbon-3 and
Carbon-6 thermal barrier designs caused a comparable
temperature drop across the thermal barrier over a 30-see.
torch flame application.
Figure 9 illustrates the symmetry of the temperature
drop data [br Carbon-3 and Carbon-6. Figure 9(a) shows
the temperatures
recorded by the seven hot and cold side
thermocouples
that were in contact with the surface of a
Carbon-3 specimen 15 sec. into the test. Though the downstream volume in the nozzle-to-case
joint of the Shuttle
solid rocket motors is expected to fill in <10 see., 15 scc.
was chosen to include a safety factor of 5 sec. Figure 9(b)
shows similar data for a test performed on Carbon-6. Both
figures show the temperature distribution from left to right
across the hot and cold sides of the thermal barriers. The
center thermocouples
on the hot and cold sides correspond
to a position of zero. Thermocouples
to the left of center
have a negative position value, while those to the right have
a positive value. Both figures show a temperature distribution that is close to symmetric around the center thermocouples. Figure 9(a) shows that the data lor this Carbon-3
test is shifted slightly to the right. Both figures show a
temperature drop of about 2300 °F between the hot (Tho t)
and cold (Tcokl) center thermocouples
in contact with the
surface of the specimens.
Jet Spreading.
The jet spreading
capability
of
Carbon-3 and Carbon-6 is also shown in Fig. 9. Although
the hot (3000+ °F) torch was focused into a narrow
(0.084-in. diam.) column, the thermal barrier spread the
heat at least I in. on either side of the center thermocouples.
Figure 9(a) shows that for Carbon-3, temperatures
I/4 in.
away from the center hot side thermocouple
were about
6
of Aeronautics
and Astronautics
1200 °F on the left side and over 2000 °F on the right side.
Hot side data for Carbon-6 in Fig. 9(b) show a similar
trend with temperatures I/4 in. away from center over
2200 °F. Cold side data from both Figs. 9(a) and (b) show
that the hot gas ,jet was reduced in temperature and
diffused. Reducing the unit thermal energy per area is
beneficial in preventing hot gas effects on the downstream
()-rings.
Focused Jet Endurance Tests. Table Ili and Fig. 10
summarize the results of repeated temperature drop tests
performed on single specimens of Carbon-3 and Carbon-6
to examine their endurance alter multiple applications of
the oxyacetylene torch. For both thermal barrier designs,
a single specimen was subjected to the torch flame for two
30-sec. periods followed by two 60-see, pericvds. The
exposure times of 30 and 60 sec. are longer than the,joint
cavity fill time of 10 sec. but were selected to examine
the thermal
barrier's
insulation
and flame resistance
properties.
After each exposure,
the specimen
was
photographed
(with fixture cover plate removed) to record
any specimen damage before the next test was performed.
For reference, the Carbon-6 specimen was also exposed to
a 20-sec. flame application
before these endurance tests,
and no damage was observed.
Table III shows
several important
temperature
measurements
for each test after 15 sec. as well as the flow
through the specimen at fifteen seconds, the maximum
bulk temperature reached during a test, and the amount of
recession on the hot side of the specimen after the final
flame exposure. The data for Carbon-3 shows that tests 30,
3 I, and 32 were almost identical. Each showed a maximum
hot side
temperature
slightly
above
3000
°F and
a
temperature drop (Thot - Tbulk) of ovcr 2800 °F. The only
difference between these tests was the higher maximum
bulk temperature of 500 °F in tcst 32. This was due to thc
longer flame exposure
time that allowed
the bulk
temperature to keep increasing tor 60 sec as compared to
the 30-see exposures in tests 30 and 31. The maximum hot
side temperature in test 33 only reached 2590 °F compared
to 3(900+ "Fin the other tests. This caused lower temperaturc
differences across the specimen and lower bulk temperatures. For all four tests, the highest bulk temperature after
15 sec. was 230 °F. This is well below Viton's short term
maximum
operating
temperature
limit of 600 °F. l I Even
the maximum bulk temperature of 500 °F recorded after
60 sec. of flame exposure was within the limit. Figure IO(a)
shows the hot side of the Carbon-3 specimen after all four
flame exposures. No damage can be seen after the first
three tests with little if any damage evident alter the final
test. As shown in Table lit, there was a recession of
0.029 in. ( 13% of the compressed cross-section)
measured
alter 180 sec. of exposure. The thermal harrier should never
experiencc such a prolonged exposurc to.jets of hot gas in
the actual rocket application.
American
Institute
The endurance
tests performed
on Carbon-6
revealed
results slightly different than tor Carbon-3. After 15 see.,
the maximum temperature ranged from 2520 to 2730 °F
with temperature drops (Thot-Tbulk) that ranged from 2240
to 2560 °F. The maximum bulk temperature after 15 see.
was 280 °F, slightly higher than that for Carbon-3 but still
well below the Viton ()-ring temperature
limit. The
Carbon-6 series revealed a slightly higher maximum
overall bulk temperature
of 620 °F that occurred in the
final test after a 60-see. flame exposure. This temperature
is about the maximum that the ()-rings can withstand for
a short period of time, but as mentioned previously,
the
thermal barrier should not experience such a long flame
exposure in the rocket.
Figure 10(b) shows
the hot side of the Carbon-6
specimen after all lour flame exposures. Very little damage
can be seen after the first test. but the amount of damage
to the specimen increased to a maximum
recession of
0.092 in. (30% of the compressed
cross-section)
after the
final test. This recession likely contributed to the increased
maximum bulk temperature in the final test. These temperaturc drop tests were all performed in a more aggressive
oxidizing environment
than the thermal barrier would
experience in the rocket. The amount of damage observed
on thc Carbon-6 specimen after 2(X) sec. of flame exposure
would not bc expected to _v,:cur in a less oxidizing environment with much shorter hot gas exposures.
For both series of tests, the flow through the specimen
was almost identical from test to test. Flow rates through
Carbon-6 were higher than those through Carbon-3 as is
expected since Carbon-6 is more porous than Carbon-3
(Table I1).
Flow Test Results
Flow rates (measured
for Carbon-3, Carbon-4,
using the piston flow rig, Fig. 3)
and Carbon-6 at 20 and 25c)_
linear compression
are summarizcd in Fig. I I at 60 psid
and 70, 500, and 900 °F after scrubbing and 70 °F belorc
scrubbing. Application of the thermal barrier in the Shuttle
solid rocket motor nozzle-to-case
joint involves predominantly static (e.g. no scrubbing) loads. As shown by
the flow results, flow resistance
increased with higher
compression
levels. Figure 11 shows that the flow rates
for Carbon-6 were higher than those for Carbon-3 and
Carbon-4 at 60 psid at each temperature and compression
level. Carbon-6 flow rates were 2. I to 2.9 times higher
than Carbon-3 flow rates and 1.7 to 2.3 times higher than
Carbon-4
flow rates at comparable
temperatures
and
compression
levels. This difference is due to differences
in braid architecture between these thermal barrier designs.
The difference
in flow rates between Carbon-3
and
Carbon-4 was attributed to Carbon-4 incorporating
larger
core fibers resulting
in higher
seal porosity
than
Carbon-3. 3 Carbon-6 incorporating multiple sheath layers
7
of Aeronautics
and Astronautics
hasa higherporositythanCarbon-3
(Table1I)andis
therefore
more
permeable.
Discussions
between
theauthors
androcketmanufacturer
Thiokolhaveindicated
thatthe
thermal
barriers
have
highenough
permeability
topermit
thejoint-cavities
tofill inacceptable
times.
Effect of Temperature.
Figure II shows that flow
rates dropped for each thermal barrier as the temperature
was increased. This phenomenon
is explained by the
relationship that gas viscosity increases with temperature,
,IJo_ T 2/3. Thus, as the viscosity of the gas flowing through
the thermal barriers increased, the flow rate decreased. 2
Effect of Hot Scrubbing. Thermal barrier flow rates
typically rose after hot scrubbing during flow tests. Alter
500 °F testing Carbon-6 flow rates rose as much as 20_
as compared to the flow rates belore scrubbing. Post-scrub
room temperature flows lot all thermal barriers were done
after time spent at 500 °F (2 hr) and 900 °F ( 1.5 hr). Postscrub r_om temperature flow rates for Carbon-3 as much
as doubled
as compared
to their pre-scrub
values.
Carbon-6 exhibited similar flow growth after scrubbing
but tlows for pressure differentials
of 60 psid were not
within the range of the flow meter used. It is believed that
much of the flow rate increase is due to oxidation that
occurred
while the specimen
soaked
at these high
temperatures.
No major visible damage due to scrubbing
was observed on any of the thermal barrier designs at the
conclusion
of the flow tests. Only minor fraying was
observed at the specimen ends in the lap joint. Temperature
exposure tests performed on carbon fiber thermal barriers 3
showed that short lengths of carbon thermal barrier lost
weight when heated in a furnace at different temperatures
tbr two-hour exposures. This supported the theory that the
carbon
thermal
barriers
oxidized
when exposed
to
temperatures of 9(X) °F for extended periods of time. and
the associated
weight-loss
contributed
to the increased
flow rates after scrubbing.
Compression
Test Results
Table IV summarizes the results of the compression
tests performed on Carbon-3 and Carbon-6 and includes the
measured contact width, preload, and residual interference
Ior each amount of linear compression,
or crush, tested.
Contact Width. The contact width increased for the
Carbon-3
and -6 designs
as the amount
of linear crush was
increased. The thermal barriers continued to spread and
flatten out as they experienced
larger amounts
of
compression.
In each test, the footprint pattern left on the
pressure sensitive film after a compression cycle was solid
and continuous.
This indicates that during a flow test
continuous contact is made between the walls of the flow
fixture and the thermal
the specimen.
barrier,
minimizing
American
leakage past
Institute
The contact width at each compression
level for
Carbon-6 was over twice as large as it was for Carbon-3
even though the diameter of Carbon-6 was only 1.3 times
larger than tor Carbon-3. This shows that Carbon-6 had
a softer, more compressible
braid architecture
than
Carbon-3 allowing Carbon-6 to spread out more as it was
compressed.
Preload. The amount of preload or footprint contact
pressure increased with the amount of linear crush. However, Carbon-6 had preloads that were 1/6th to 1/9th those
ft_r Carbon-3
at each compression
level. As a result,
Carbon-6 will cause lighter loads on the adjoining rubber
J-leg element. The reason for this difference in preload is
believed to be related to the architectures
of these thermal
barrier designs (Table I). In Carbon-3 having a tightly
packed core of uniaxial fibers, there is little room lor
individual fibers to move with respect to one another when
they are compressed.
In contrast, in Carbon-6 the sheath
fibers are oriented at an angle with each other and arc
better able to slide past each other when the thermal barrier
is compressed.
Residual Interference. As with the contact width and
preload, thermal barrier residual interference
or spring
back also increased as percent linear crush increased.
Although contact width and preload were quite different
for Carbon-6 and Carbon-3, residual interference
scaled
with diameter lor these two designs. Increasing thermal
barrier diameter by a factor of 1.3 from 0.200 to 0.260 in.
resulted in an increase in residual interference by that ratio
for each level of compression.
Residual interference
for
Carbon-6 was 0.025 in. even for the lowest compression
(20%.) and meets the design requirement to Iollow nozzlc
joint movement during Shuttle solid rocket motor operati on,
as discussed with rocket manufacturer
Thiokol.
Comparison
of Carbon-3 and
Carbon-3 and Carbon-6
temperature
drop comparison
somewhat greater insulating
showed less recession
than
Carbon-6: Other Factors
both performed well in thc
tests. Carbon-3 did offer
effects than Carbon-6 and
Carbon-6.
We believe the
higher density of Carbon-3 is an important reason for
these results. However, there are many other factors to
consider
when deciding
between
these two braid
architectures.
Carbon-6 is braided using larger tows or
yarns that permits faster and therefore most cost-effective
production.
Carbon-6 is a more flexible braid that makes
it easier to spool for shipment and more accommodating
during installation. The current tests combined with other
planned rocket motor and joint-simulation
tests will enable
Thiokol
and NASA to decide on the optimal
braid
architecture
tot the thermal barrier.
8
of Aeronautics
and Astronautics
Results of Thiokol Char Motor Tests on Carbon
Thermal Barrier
Thiokol tested a 0.260-in.
diameter
Carbon-6
thermal
barrier for NASA in a subscale rocket motor to verify that
it would withstand the Shuttle solid rocket motor environment. The subscale motor, or "char" motor, simulates the
effective slag barrier. The inset photo in the figure shows
a close-up of an area where slag was trapped by the
thermal barrier, preventing it from reaching the downstream
()-rings. Minor fraying occurred in the area immediately
around the lap.joint during disassembly,
but the specimen
is otherwise in good condition.
thermal conditions of the full-scale motor by burning solid
rocket propellant at corresponding
chamber pressure and
temperature
conditions. The thermal barrier was placed
into an intentional gap defect between the phenolic insulation blocks, as shown in Fig. 12(a). The combination of an
outboard plenum chamber and the 0.060-in. circumferential
Comparison
gap extending
both upstream and downstream
of the
thermal barrier ensured thal hot gas flow would pass
through the thermal barrier.
Throughout the test duration of- 1 I see., a significant
drop in temperature
was measured across the thermal
phenolic material to simulate the material and boundary
conditions that the thermal barrier would be exposed to in
these other configurations.
The thermal barrier specimens
were subjected to 209b compression
as they were in the
char motor test and as planned for the rocket. The flame of
barrier. Figure 12(b) shows that the maximum temperature
seen on the hot side of the thermal barrier was over
the oxyacetylene
torch
that was used for the temperature
drop tests was directed through a 0.084-in. diameter hole
in an iris plate to simulate a hot gas jet that the barrier could
be exposed to in the rocket. Flame exposure times were
32(X) °F, while the cold side temperature
reached about
950 °F. Thus, a temperature drop of about 2200 °F occurred
across the 0.260-in. diameter thermal barrier. Pressure
readings upstream and downstream of the thermal barrier
and in the reservoir confirmed that there was gas flow
across the thermal barrier. The thermal barrier diffused the
focused nature of the hot gas jet, further reducing the jet's
potentially
damaging
effects on downstream
Viton
()-rings in the actual Shuttle solid rocket unotor.
Although the 950 °F temperature
recorded downstream of the thermal barrier is still higher than the temperature limits of the Viton nozzle ()-rings, the char motor
subjected the thermal barrier to more aggressive conditions
than would ever occur in the actual Shuttle solid rocket
motor, for the following reasons. First the gap defect was
purposely oversized at 0.060 in. to force flow through the
thermal barrier. In the actual nozzle ,joint, the gap between
adjoining blocks of insulation would be narrower as the
pieces of insulation are basically in contact with each
other. The narrow gaps between the phenolic insulation
would significantly cool the incoming gas temperature
impinging on the thermal barrier and would therefore
lower the temperature
of the gas that reaches the Viton
()-rings. Furthermore,
the downstream temperature in the
char motor test was recorded immediately downstream of
the thermal barrier. The ()-rings in the rocket nozzle .joint
are located several inches further downstream
of the
thermal barrier, allowing additional
heat to
from the gas before reaching the ()-rings.
Figure 13 shows the thermal barrier
removed from the char motor. There was
burning or charring of the thermal barrier.
Fig. 13 shows that the thermal barrier also
American
be removed
after it was
no apparent
In addition,
acted as an
Institute
of Thiokol
Char Motor
Test Results to
NASA Temperature
Drop Test Results
The fixture used to perform the temperature drop tests
on the Carbon-3
and Carbon-6
thermal barriers was
modelled after the char motor and the shuttle nozzle-tocase joint thermal conditions.
The fixture was made out of
intentionally
longer than they would be in the rocket
application to simulate extreme heating conditions.
Considering the results of Fig. I0 (NASA temperature
drop fixture), tests were performed with hot side temperatures ranging from 25(R) to nearly 32(X) °F. Carbon-6
temperature
drops ranged from 2240 to 2560 °F-I 5 sec.
into the test. These were somewhat
greater than the
2200 °F temperature drop exhibiled by Carbon-6 in the
char motor. The main reason for this difference
is that
9(X) psi pressures were generated by the char motor, while
only 10 psid pressures were applied across the thermal
barrier in the temperature drop tests. The higher-pressure
char motor test caused more hot gas to tlow through the
thermal thereby raising the downstream temperature causing a smaller temperature drop. Though there are some
differences
in the absolute results, the authors believe the
laboratory temperature-drop
test fixture simulates many
of the key factors at work in the rocket. The laboratory setup permits quick and easy comparisons between competing
architectures
and can be used to generate thermal data to
anchor thermal correlations
under development.
Summary
and Conclusions
The 55(X)+ °F combustion gases in the Space Shuttle
solid rocket nnotor are kept a safe distance away from the
assembly .joint seals by thick layers of insulation and by
special compounds
that fill the joint split-lines
in the
insulation. The current nozzle-to-case
joint design incorporates primary, secondary and wiper(innermost)
()-rings
9
of Aeronautics
and Astronautics
and polysulfidejoint-fill
compound. In the current design,
I out of 7 motors experience hot gas to the wiper O-ring.
Though the condition does not threaten motor safety,
evidence of hot gas to the wiper O-ring results in extensive
reviews before resuming flight. NASA and solid rocket
motor manufacturer
Thiokol are working to improve the
nozzle-to-case joint design by implementing a more reliable
J-leg design (successfully
used in the field and igniter
joints) and the thermal barrier Carbon-6 described herein.
The thermal resistance of two NASA thermal barriers,
denoted Carbon-3
and Carbon-6.
was assessed
Ik_rmed to measure the temperature drop across and along
the thermal barriers in a compressed
state when subjected
to the flame ofan oxyacetylene
torch. Flow and durability
tests were conducted on the thermal barriers to examine
their leakage characteristics
and durability at ambient and
high temperatures.
Room temperature compression
tests
were pertormed to determine load versus linear compression, preload, contact area. and residual interference/
resiliency characteristics.
Subscale rocket "'char" motor
tests were performed in which hot combustion gases were
directed at the Carbon-6 thermal barrier to assess its thermal
resistance
in a rocket environment.
The current tests
with other planned rocket motor and joint
tests will enable Thiokol and NASA to decide
on the optimal
braid architecture
for the thermal barrier.
Based on the results of the current
conclusions
are made:
tests, the following
I. The Carbon-6
(0.260-in.
diam.) and Carbon-3
(0.20-in. diam.) thermal barrier resisted the 5500 °F flame
of an oxyacetylene
torch for over 6 min before burn
through, greater than three times the Shuttle solid rocket
motor burn time.
2. Carbon-3
and
Carbon-6
thermal
barriers
were
excellent insulators causing temperature
drops through
their diameter from 25(X) to 2800 °F, depending on test
parameters.
Gas temperature
I/4" downstream
of the
thermal barrier were within the downstream Viton ()-ring
temperature
limit of <600 °F.
3. The Carbon-6 thermal barrier design performed
extremely well in subscale rocket "char" motor tests that
subjected it to hot gas at 3200 °F for an I I-see. rocket firing,
simulating
the maximum downstream
joint-cavity
filltime. The thermal barrier reduced the incoming hot gas
temperature by 2200 °F in an intentionally
oversized gap
American
References
by exposing
them to an oxyacetylene
torch at 5500 °F and measuring
time for burn through. Temperature
drop tests were per-
combined
simulation
defect, spread the incoming jet flow, and blocked hot slag,
thereby offering protection to the downstream O-rings.
4. Laboratory burn, temperature drop, flow, and compression tests and subscale rocket "char" motor tests
demonstrate
the thermal barrier's
feasibility
for use in
rocket applications and qualify it tbr comprehensive
motor
evaluation.
Institute
ISteinetz,
B.M.,Adams,M.L.,
Bartolotta,
P.A., Darolia,
R.,
and ()lsen, A., "High Temperature
Braided Rope Seals
tbr Static Sealing Applications,"
NASA TM- 107233,
rev., July 1996.
2Steinetz,
B.M.. and Adams, M.L., "Effects of Compression, Staging, and Braid Angle on Braided Rope
Seal Performance,"
NASA TM-107504,
July 1997.
3Steinetz, B.M.. and Dunlap, P.H.."Feasibility
Assessment
of Thermal
Barrier
Seals for Extreme
Transient
Temperatures,"
NASA TM-208484,
July 1998.
4Thiokol report TWR-7319 I. "RSRM-45A
Nozzle Joint
No. 3 ()-ring Erosion Investigation Team-Final Report,"
October 28, 1996.
5Rogers, W.P.. "Report of the Presidential Commission
on the Space Shuttle Challenger
1986.
6Ballis. W., ASM Handbook,
Accident,"Junc
Volume 6: Welding,
6,
Brazing,
and Soldering, ASM International,
1993, pp. 281-290.
7Bahl, O.P. and Dhami, T.L., "Oxidation
Resistance of
Carbon Fibers," High Temperatures
- High Pressures.
Vol. 19, pp. 211-214,
1987.
8Eckstein, B.H. and Barr, J.B., "An Accelerated Oxidation
Test for Oxidation Resistant Carbon Fibers,"MaterialsProcesses:
Twentieth
The Intercept
International
Point; Proceedings
of the
SAMPE
Technical
Conference,
Minneapolis,
MN, Sept. 27-29, 1988.
Covina, CA, Society for the Advancement
of Materials
and Process Engineering,
1988, pp. 379-391.
9Eckstein, B.H., "'The Weight Loss of Carbon Fibers in
Circulating Air," 18th International SAMPE Technical
Conference,
October 7-9, 1986, pp. 149-160.
l°Lide and Kehiaian CRC Handbook of Thermophysical
and Thermochemical
Data, CRC Press,
1994.
pp. 25-31.
I Iparker O-t4ng Handbook,
10
of Aeronautics
and Astronautics
Cleveland,
OH, 1992.
Barrier
type
Carbon-
TABLE
Core
Size
Diameter,
in.
I
(I. 125
Carbon-2A
0.125
Carbon-3
0.200
Carbon-4
0.194
Carbon-6
0.260
Grafil h
34-700 12K
Grafil
34-7011 12 K
34-711tl 3K
Grafil
34-71111 12K
34-700 3K
Grafil
34-7011 12K
Amoco"
P25 2K
Grafil
34-71)tl 12K
2.76xl0
72110
1800
2.76xl0"*
_
72110
2.76x
I1) "_
4.4x 10 "_
2.76xl0
Thermal
_Porosity
2.8x1(1 _
9
8
I
45
l0
Thornel
T-3111) I K
Thornel
T-30() I K
Thornel
T-30() 1K
T-300 3K
All-Ceramic
61tl)
2.8x10
5
I
61111
2.8x I 0 _
5
600
181111
2.gxl0
10
12 in I-2
24 in 3-5
12 in I-2
24 in 3-5
8 in I-5
12 in6-7
16 in 8-I1)
65 in I _'
61) in 5'"
65 in I _'
61) in 5 _"
17 in I "_
45 in 2-111
of
Diameter.
in.
0.2011
0.260
I
Exposure
31
32
33
Tesl
number
35
36
37
38
30
[
311
60
60
[
Exposure
Per test,
SCC
31)
311
60
60
5
8
I
45
600
2.8xl0
II)
8
I
45
"_
"_
a
I 700 I 3.2xl0_ [
Thermal
POROSITY
barrier
Carbon
density,. Ib/cu.in.
0.1141
0.032
Ill.--TEMPERATURE
time
T,,,,al
15see.
IP.,e,,.I
A.'o,,,uUed.
°E
SCC
sec
31)
2.8x111 a
I
2
8
I
I
I
'
AT
fiber
Porosity"
densit?', Ib/cu.in
0.06,4
t).1164
0.37
0.50
= I - Pu_/Pcl .
(a) Carbon-3
Test
number
6(XI
THERMAL
BARRIER
COMPRESSION
layers
TABLE
DROP
Temperature
RESULTS
Drop Test Results
J Th,,,_
at
T,.,, - Th,,it
15see,
J at 15see,
T,,,,_- T...... I
at 15 sec,
°F I
I
TEST
°F
Fiow at
15 sec.
SCFM/m.
I
Th,,ik
lnaxinlunl,
I
I
Recession
lest
in.
°F
31)
3070
I
2111
I
286t)
23311 111.14
310
611
121)
8(
3050
3020
2590
[
2311
21X)
50
[
I
28211
28211
24411
23/_1 I
22511
.14
.H
3411 l- -I
500
......
.14
340
time
Accuruulated.
sec
50
80
1411
200
921
Test Results
(b) Carbon-6
Temperature
1)1
, - T,, u
Flov, at
TI,,,tat
Ti,utk at
Tt>, - Tt,_,tl, at'k,,,
15 sec.
15 sec,
al 15 sec,
at 15 sec.
15 sec.
_'F
SCFM/in.
°F
°F
°F
2730
2690
25211
2701/
American
angle.
Braid
degrees
6011
I
5
10
I
Thornel
T-300 IK
II.--MEASURED
Number
sheath
yarns per
Numberof
bund c
I
I
20%
barrier t vpc
Carbon-3
Carbon-6
carriers
NumberofJ
per ayer
I
I
21
4
layers
Carbon
Thornel'
T-300 I K
Thornel
T-300 I K
NTW.a C-2 J
11.1211
J NX 551F
I 7°° I 3.2×10" I lilt, [ NX550
-'lxl0
in.=25gm.
hGrafil type 34-700 carbon libers, Gralil Inc. product.
12K-12.0()0
tiber ends.
_Thornel T-300 carbon fibers, Amoco Perlbrmance
Products,
Inc. producl.
"Amoco P25 pitch fibers. Amoco Peffurmance
Products,
Inc. product.
_NX 5511 = Nextel 550 fiber, 3M product, 73r;bAlzO, 27c/f SiO,
TABLE
MATRIX
diameter,
in?
4
720()
2.76x111 _
181111
2900
CONSTRUCTION
of yarns
7200
72011
BARRIER
Sheath
tliameter,
in/
0.114
Carbon-2
I.--THERMAL
171)
191)
2811
280
Institute
25611
25/)11
2240
2420
II
of Aeronautics
2050
1960
17611
1700
11.24
0.24
0,25
11.24
and Astronautics
Tt,,,,_
lllaxinlulll,
°F
320
350
481)
621)
I--
01129
after
Percent
I--
--3
Recessiun
after
test
in.
Percent
......
......
....
I).1)92
31)
I
s6
TABLE
IV.--THERMAL
BARRIER
CONTACTWIDTH, PRELOAD, AND RESIDUAL
INTERFERENCE FOR SEVERAL LINEAR CRUSH CONDITIONS
Diameter, Nominal percent Linear Number of Contact Preload,
Residual
in.
linearcrush,
crush,
sheath
width,
psi
interference,_'
percent
in.
lancers
in.
in.
0.2
20
0.040
5
0.063
310
0.019
25
.050
.082
490
.027
30
.060
.099
930
.I)33
Carbon-6
0.26
20
0.052
10
O.157
56
0.025
25
.(K_5
.192
81)
./)36
30
.078
.196
97
.041
"Residualinterferenceis defined as the distance that the thennal barrier will spring back while
maintaininga load of al leasl 1 Ib/m of specimen.
Themlal
barrier
type
Carbon-3
(a)
r Thermal barrier
I
_Vent
port
/
/-- Leak
check port
/
Secondary
O-ring
L Primary O-ring
L.Wiper O-ring
A
Rocket centerline
T-Throat
(b)
_assemblyr(_)
.- Beadng
assembly
Exhaust flow
_
,- Forward exit
,/
cone assembly
'//--_
/
L_Nozzle inlet
assembly
/ L Cowl
assembly
Section A - A
/
/
/- Aft exit cone
assembly
_ Nozzle-tocase joint
Figure 1.--Potential Shuttle solid rocket motor joint locations for thermal barrier.
(a) Enlarged view of nozzle-to-case joint showing J-leg, wiper, primary, and
secondary O-rings, leak-check port, and proposed thermal barrier location.
(b) Overall nozzle cross-section (half view).
American
Institute
12
of Aeronautics
and Astronautics
Top cover plate (phenolic)
removed for clarity _
r- Plenum chamber
/
_
F- Test
_
_- Iris plate
specimen
/
/
_
/
\
Vacuum
pump
roughing
_•
draws flow through
flow meter/fixture
Air
_
_
"<'_
---
_,/_--_
_
//"
-
_-__
/_
,_"_-_
Downstream
T/C's
Radiant
stroke
heating
ITI
0.25 in.
--"
,,,
_///
--8 each,
i
.-/
Piston
_T'/_ll
--.
"
K
Upstream ............. ambient
Differential ........ 10-11 psid
type K -_
of temperature
drop
Carriage
_/
test fixture.
nperature
Force
I
liFT7]_
/
Digital
Specimen
f"
in piston
(
groove
_,
"_
Y/A
J
_
g/) d_ ,
Square
i/
1
I_
_- Pressure
sensitive
radii -_ _
\_ _ 4-C_ _ _.
r----m',_-r-m-
-
Specimen
, Lap joint-7
holders_'l"
3. - ..(
Load ceil (2) _|,
.'r t
L
Ii.
_,/,
, _,/.
_
/| "--"
i''
"_
-
-- Test
specimen
'
_Stationary
plate
, I
T
0-100
plate)
film
I
Insulation
stationary
grooves
with corner
.
%
indicator
I( c°ntacts
\
-_-
plate
7Moving
__-
dia.lf-_
_mlZSin.
torch
_
:reesa22::pe
*II-:-
,
oxyacetylene
[_
IT I
":-Cylinder I
die hole) S.S.
Iris
Figure 2.--Schematic
:i
,
---
"-._
-_
i
C
0 nng--'
_ Miniature
/
Y-_/
\\_
surface
_
(0.084
%
z_Pressure
/
psi Hot air supply
Figure 3._Schematic
"/
of flow
,
/j
,
Figure 4._Schematic
fixture.
13
American
Institute
,
of Aeronautics
and Astronautics
/
/
/
of compression
/
fixture.
400
387
399
350
¢_
Diameters:
.¢E ¢_
_
300
C-3, C-4 = 0.2"; C-6 = 0.26"
O
¢_
_ 250
E >_
Reference:
o 200
O
1/8" nominal except
2 min 4 sec Shuttle solid
_
e-
rocket motor burn time -7
150
_: o_
.....
_'_
f>_"__ 100
||,,1
........
0
7_
__'0
,oo
I
I
i
i
e"
_
¢n ¢n
Figure
11 7
j"__133
<
•
_,o
5._Oxyacetylene
3500
y
thermal
Carbon
torch burn test results
¢?,
(n = number
I
barriers
of tests performed).
Distance
from center
(in.)
1 - Hot Center, Tho t
2
Hot R1
0.0
3
Hot R2
0.50
4
Hot R3
1.00
I1
30OO
5
Cold Center,
6 - Cold R1
2000
_ 1500
E
_,
Channel
I
2500
_
0.25
0.0
Tcold
0.25
7 - Cold R2
0.50
8 - Cold Bulk, Tbulk
9 - Cold R3
0.0
1.0
1000
500
0
0
10
20
30
Time
40
50
60
Figure 6._Temperature
rise vs. time for simulated hot gas exposure
upstream (hot) and downstream
(cold) temperatures
for Carbon-3.
temperatures
removed
for clarity.
American
Inslitute
70
(sec)
14
of Aeronautics
and
Astronautics
showing
Left hand
Channel
35O0
I
I
2500
1
2
3
4
2000
5
6
7
1500
8
9
3000
E
Distance
from center
(in.)
0.0
- Hot Center, Thot
Hot R1
0.25
Hot R2
0.50
Hot R3
1.00
0.0
.....Cold Center, Tcold
- Cold R1
0.25
- Cold R2
0.50
0.0
- Cold Bulk, Tbulk
• Cold R3
1.0
1000
500 __
•
:f
,_
°J-_7
___J.
8_
0
i
5
_--,_----__
...........
_
I
I
!
01
............
t
10
15
20
25
I
30
.....
_
I
35
L
40
I
45
_
50
55
Time (sec)
Figure 7.--Temperature rise vs. time for simulated hot gas exposure showing
upstream (hot) and downstream (cold) temperatures for Carbon-6. Left hand
temperatures removed for clarity.
3OO0
U.
o
F- Carbon-3
-1
2000
2500
6_
o
t-
/
//
Carbon-
h_
o
/
1500
"O
.9,=1000
Q.
E
t--
500
00
5I
_
1-
I
15
I
20
I
25
30
-
35
40
45
Time (see)
Figure 8.--Temperature drop vs. time
Carbon-3 and Carbon-6.
American
Institute
(Thot-Tbulk)
15
of Aeronautics
at flame location for
and Astronautics
5O
3500
3500
Hot side
3000
_"
-_
Cold side
(a)
/_
2500-
2500
.=
"¢ 2000
2000
-_
Cold_
0_
500 0.084"
_-
(b)
Hot side
3000
¢z 1500
E
1000
).084"
/
1000
et diamet_...._.
5OO
500
0
-1.00-0.75-0.5-0.25
0
Position
Figure 9.--Hot
diameter)
0.25
0.5
0.75
I
-0.75
b
for thermal
'-_1
0
Position
barriers
versus
axial position
(a) Carbon-3;
(b) Carbon-6.
at 15 seconds
16
American
I
-0.5 -0.25
(in.)
side and cold side temperatures
spreading
0
-1.00
1.00
Institute
of Aeronautics
and Astronautics
0.25
0.5
0.75
(in.)
showing
jet (0.082
in.
1.00
Test
Test
. rThermal
!J,
_
_
_
i
barder_
Cold
. _Thermal
barrier
31
_
36
Hot
Cold
. _Thermal
barrier _
32
37
Hot
,_ _ "tp_-,_'tll_ll_.j_
• _:_ :
....
_
.
.....
. _- Thermal
33
__
:
38
.ot
(a) Carbon-3
ITest#
Exposure
Temperature
Time
Per test Accumulated
30
(sec)
30
(sec)
30
Drop Test
Tbulk
Tho t
@15sec
@15sec
(°F)
210
(°F)
3070
(b) Carbon-6
Results
Thot-Tbulk
Test#
@ 15sec
Exposure
Temperature
Time
Per test Accumulated
(°F)
2860
35
Drop Test Results
Thot
Tbulk
!Thot-Tbulk
@ 15sec
@15sec
(sec)
(sec)
(°F)
(°F)
@ 15 sec
(°F)
30
50
2730
170
2560
2500
31
30
60
3050
230
2820
36
30
80
2690
190
32
60
120
3020
200
2820
37
60
140
2520
280
2240
33
60
180
2590
150
2440
38
60
200
2700
280
2420
versus
accumulated
Figure lO._Thermal
barrier
condition
and key temperatures
17
American
lnstilule
of Aeronautics
and Astronaulics
time. (a) Carbon-3,
(b) Carbon-6.
0.70
0.66
0.60
[]
70 °F - Before scrub
•
70 °F - After scrub
[]
500 °F - After scrub
[]
900 °F - After scrub
0.50
0.45
0.44
.c
0.40
o_
,T
0.30
0.39
off
0.27
0.26
0.
0.20
.21
;//ll
0.20
i
0.10
7"/ll
__
•
0.12
0.11
--
--
--
° °01JI
---
0.21
_--
0.00
Percent
compression[
20%
25%
Carbon-3
Figure 11 .--The
&P = 60 psid.
effect
(0.20" dia.)
of temperature,
"Char" moto_
nozzle
20%
test article
/
/
Intentional
/
/
Rocket
Carbon-4
thermal
barrier type,
/-- "/'cold
"
/-- Tho t
Carbon-6
scrubbing
(0.26" dia.)
and compression
on flow,
-
4000
/
og-
//
g 3000
7"hot
/__-
_- 2000
flow
°F
="-1
/-........
06_,_
[
barrier
-
(a)
Figure
25%
J
(0.20" dia.)
1000
Thermal
20%
J [
_',
(5000
25%
i [
• .....
Tcol d
/
I
20
J
tNI
11 -sec __J
burn time-'
I
30
Time, sec
(b)
12.--Subscale
configuration:
Temperature
(70 Ibm) "char"
Carbon-6
data:
thermal
Upstream
motor
tests examining
barrier impedes
thermal
hot gas flow
(Thot) and downstream
barrier (Carbon-6)
through
(Tcold) sides
intentional
of thermal
barrier.
18
American
Institute
of Aeronautics
and Astronautics
effectiveness.
joint defect
(Courtesy
(0.06
(a) Test
in. gap).
of Thiokol
(b)'
Corp.)
RTV
remaining
from test
--_.
\
\
5 1/8 in.
Diam.
Hot slag blocked
by barrier
Figure
13.--Photograph
effectively
Thiokol
blocks
of char motor
3200
thermal
barrier
(Carbon-6)
°F gas for 11 sec. (joint fill time)
after test. Thermal
and blocks
Corp.)
19
American
Institute
of Aeronautics
barrier
hot slag. (Courtesy
and Astronautics
of
REPORT
DOCUMENTATION
PAGE
FormApproved
OMB
No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, includingthe time for reviewing instructions, searching existing data sources,
galhering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this
colleclion of information, includingsuggestionsfor reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington. VA 22202-4302, and to the Office of Management and Budget, Paperwork ReductionProject (0704-0188), Washington, DC 20503
1. AGENCY
USE ONLY
(Leave
12. REPORT
blank)
DATE
4. TITLE
3.
REPORT
TYPE AND DATES
Technical
1999
June
AND SUBTITLE
Development
!5.
of Thermal
Barriers
lor
Solid
Rocket
Motor
Nozzle
COVERED
Memorandum
FUNDING
NUMBERS
Joints
WU-523-53-13-00
6. AUTHOR(S)
Bruce
M.
Steinctz
7. PERFORMING
Patrick
ORGANIZATION
National
John
and
Acn_nautics
H.
Glenn
Cleveland,
and
Space
Aeronautics
Washington,
Jr.
AND ADDRESS(ES)
8. PERFORMING
ORGANIZATION
REPORT NUMBER
Administration
Center
at Lewis
Field
AGENCY
NAME(S)
E-11738
44135-3191
9. SPONSORING/MONITORING
National
Dunlap,
NAME(S)
Research
Ohio
H.
and
DC
Space
AND ADDRESS(ES)
10. SPONSORING/MONITORING
AGENCY REPORTNUMBER
Administration
NASA
20546-(X)01
TM--1999-209278
AIAA-99-2823
11.
12a.
SUPPLEMENTARY
Prepared
for
Angeles,
California,
the
351h
Joint
June
Subject
This
Propulsion
20-24,
DISTRIBUTION/AVAILABILITY
Unclassified
13.
NOTES
Conference
&
Exhibit
cosponsored
by AIAA
ASME.
SAE,
and
ASEE,
Los
1999.
STATEMENT
12b.
DISTRIBUTION
CODE
- Unlimited
Category:
publication
37
is available
(Maximum
ABSTRACT
Tbe Space
Distribution:
Shuttle
from
200
solid
the NASA
Center
for AeroSpace
Nonstandard
Information.
(301)
621-0390
words)
rocket
motor case assembly
joints
arc sealed
using conventional
()-ring
seals. The 5500+°F
combustion
gases
arc kept a salt: dis-
lance away, from the ,,.cals by thick layers of insulation.
Special joint-fill compounds
are used to fill the joints in the insulaion to prevent a direcl llowpath
Io the seals, ()n a number of occasions,
NASA has observed in several of the rocket nozzle assembly joints hot gas penelmtion
through defecfs in the jointfill compound.
The current nozzle-to-case
.joint design incorporates
primary, secondary
and wiper (inner-rnost)
O-rings and polysulfidc
joinl-lill
cornlumnd,
In the current
design,
Iou!
to the wiper ()-ring results
nozzle-to-case
,joint design
of 7 motors
experience
hot gas to the wiper
in extensive
reviews bcfi)re resunfing
flight.
by irnplenventing
a more reliable J-leg design
()-ring.
Though
the condition
does
not threaten
motor
sali_ty, evidence
of hot gas
NASA and solid rocket motor nmnufacturer
Thiokol are working
to improve
the
and a thennal barrier. This paper presents burn-resistance,
temperature
drop. flow,
and resiliency test resuhs for several types of NASA braided c_on-fiber
thermal harriers. Burn tests were perforn_cd to determine
the lime to burn through
each of the thermal harriers when exposed to the llarne of an oxy-acetylene
torch (55(X) °F), representative
of the 55(X) °F solid rocket motor combustion
temperatures.
Thermal
barriers
braided
out of carbon
fibers endured
the flame
for over 6 rain, three times
longer
than solid rocket motor
burn time. Tests
were perlc,17ned on two thermal barrier braid architectures,
denoted Carbon-3
and Carbon-6,
to measure the temperature
drop across and along Ihe barrier
in a compressed
stale when subjected
to the flame of an oxyacetylene
torch. Carbon-3
and Carbon-6
thermal barriers were excellent
insulators
causing
temperature
drops through their dian,,eter of up to a 28OI) and 2560'F. respectively.
Gas lenlpemlure
1/4" downstream
of the thermal barrier were within
the dov, nsb'eam
Viton ()-ring
temperature
limit of 6(X) °F. Carbon-6
pertorrned
extremely
well in subscale
rocket
"char"
motor tests when
subjected
to hot
gas at 32(X) °F fi,',ran I I -see. rocket firing, simulating
Ihe maxinmrn
downstream
joint cavity fill time, The thermal barrier reduced the incoming
hot gas
temperature
by 2200 °F in an intentionally
oversized
gap delitct, spread the incoming jet flow. and blocked hot slag, thereby offering protection to the downstream ()-rings.
14.
SUBJECT
Seals:
TERMS
Space
Carbon:
Shuttle:
15. NUMBER
Solid
rocket
motor;
Fluid
tlow,
Design
thermal
barrier:
OF PAGES
25
Test:
Braid
16.
PRICE CODE
20.
LIMITATION
A03
17. SECURITY
CLASSIFICATION
OF REPORT
18. SECURITY
CLASSIFICATION
OF THIS PAGE
19. SECURITY CLASSIFICATION
OF ABSTRACT
Unclassified
Unclassified
Unclassified
NSN
7540-01-280-5500
Standard
Prescribed
298-102
OF ABSTRACT
Form 298 (Rev. 2-89)
by ANSI sir. Z39-18