NASA Technical Memorandum 104744

Space Shuttle Entry Terminal Area

Energy Management

Thomas E. Moore

November 1991

Nq2-_9930
(f_._A_.A-TM-I047_4) SPACE SHUTTL£ ENTP, Y
Tr:R.u,[_.;AL AP.EA _:NERGY MANAGEMENT (NASA)
C_CL 01_
50 D uncl ds
G3/OI 005711_

rg/ A
NASA Technical Memorandum 104744

Space Shuttle Entry Terminal Area

Energy Management

Thomas E. Moore
Lyndon B. Johnson Space Center
Houston, Texas

National Aeronautics and Space Administration

Lyndon B. Johnson Space Center
Houston, Texas

November 1991
 CONTENTS

Section Page

1.0 Introduction ...................................................................................................... 1

1.1 Guidance Modifications ................................................................................... 1
1.2 Entry to Landing Guidance Functions ............................................................. 1

2
2.0 Development History ....................................................................................
4
2.1 Energy to Attitude Control ..........................................................................
2.2 Optional TAEM Targeting (OTT) ..................................................................... 6
2.3 Bailout Modes .................................................................................................. 9
2.4 Smart Speedbrake and Bank Limit for OI22 .................................................... 10
2.5 Theta Limits .................................................................................................. 11

3.0 Detailed Description of Algorithms .................................................................. 12

3.1 Technique of Constraints Limits ...................................................................... 12
3.2 Overview ........................................................................................ _................ 13
3.3 TAEM and GRTLS Executives - TGEXEC & GRE','EC .................................. 14
3.4 Initialization - TO/NIT & GRINIT .................................................................... 15
3.5 Heading Alignment- TGXHAC ........................................................................ 15
3.6 Navigation User PQ_meter Processor - HAC Control .................................... 16
3.7 Ground Track Predictor - GTP ........................................................................ 17
3.8 Computations - TGCOMP ............................................................................... 20
TAEM Transitions - TGTRAN .......................................................................... 23
3.10 TAEM Body Vertical Acceleration - TGNZC .................................................... 28
3.11 TAEM Speedbral_8 - TGSBC .......................................................................... 36
3.12 TAEM Bank - TGI°--IIC .................................................................................... 36
3.13 Overview of GRTLS Open Loop Guidance Phases 6 to 4 .............................. 41
3.14 GRTLS Transitions - GRTRN .......................................................................... 42
3_5 GRTLS Body Vertical Acceleration - GRNZC ................................................. 44
3.,6 GRTLS Alpha Recovery - GRALPC ................................................................ 44
3.17 GRTLS Speedbrake - GRSBC ........................................................................ 45
46
3.18 GRTLS Bank - GRPHIC ..................................................................................

ORIGINAL PAGE IS
OF POOR QUALITY

.o,

111

PRECEDING PAGE BLANK NOT FILMED

1.0 _rTRODUCTION

An historical account of the development for Shuttle's Terminal Area Energy Management
(TAEM), is presented. A derivation and explanation of logic and equations is provided as a
supplement to the well documented guidance computation requirements contained within the
official Functional Subsystem Software Requirements (FSSR) published by Rockwell for
NASA. This FSSR contains the full set of equations and logic, whereas this document will
address Just certain areas for amplification.

1.1 GUIDANCE MODIFICATIONS

The abort Glide Return to Launch Site (GRTLS) is a high altitude extension of TAEM.

TAEM was initially developed to do its primary function of energy management with an energy
controller, but has since been modified to an altitude controller for better energy management
in the presence of unknown winds.

Optional TAEM Targeting (OTr) provides greater flexibility for pilot control of the ground
track by providing the option for an overhead approach to the runway in addition to a straight-
in approach. This modification has been in use since STS-5.

A bailout guidance mode was added to implement one of the recommendations of the Rogers
Commission Report for crew capability to escape from an Orbiter in controlled subsonic
gliding flight. The software changes to both TAEM and GRTLS, along with the associated
escape pole, were the means of implementing this recommendation starting with STS-26.

A pending modification, Change Request 89979, involves a smarter speedbrake for energy
control and a smarter vehicle vertical acceleration (NZ) limiter that limits total NZ by
11miting bank attitude.

A potential modification, previously considered and still a potential change, is Theta Limits
which would compute pitch attitude limits to constrain airspeed in the presence of degraded
airdata.

1.2 ENTRY TO LANDING GUIDANCE FUNCTIONS

The Entry guidance flies the low lift-to-drag (L/D) and high angle-of-attack (a) lifting body
vehicle from atmospheric entry to the higher L/D and low a aeroplane flight region of TAEM.
The F rdry vertical component command of the aero lift vector commands the bank angle
magn,_le to control the downrange component relative to the runway landing site. The
laterCl(_-,mponent command of the aero lift vector, commands the bank angle direction (roll
reve1_s) to control the crossrange component relative to the runway landing site, whenever
the la.eral deviation from the target exceeds a deadband level. As the TAEM target altitude is
approached, entry guidance transitions from the high a of about 40 ° to the TAEM level of about
10 °. Speedbrake during entry is to a preprogrammed profile to assist attitude control and is not
used for guidance.

TAEM starts at a mach level of 2.5 at about 85,000 ii altitude. The lateral component command
of the aero lift vector commands bank angle magnitude and sign to control crossrange relative
to a prescribed or calculated (direct route) ground track toward the runway. The vertical
channel basically controls energy and altitude with a. Speedbrake control during TAEM
continues from entry to a preprogrammed profile until subsonic and then controls airspeed
indirectly through dynamic prez ;ure.

ORIGINAL PAGE IS
OF POOR QUALITY
TAEM delivers the vehicle into the final approach plane of the runway and then delivers
control to the approach and landing guidance at from 5000 to 10000 feet altitude. The
approach guidance continues to fly the reference glide slope with ct, the lateral track with bank,
and the airspeed with speedbrakes. The landing guidance then flares from the steep reference
to a shallow glide slope and then flares to the runway for touchdown.

Entry to Landing Ground Track Sketch

The GRTLS guidance is identical to TAEM, except that it starts it is TAEM phase at mach 3.2
and has three guidance open loop phases with closed loop flight control of either angle-of-
attack or acceleration in the region between external tank separation (about mach 6) and mach
3.2.

GRTLS Vertical Plane Sketch

2.0 DEVELOPMENT HISTORY

The three flight control variables given for the development of TAEM are speedbrake for drag
modulation below maeh .95, normal acceleration along the body axis Z (NZ) for longitudinal
state control, commanded from the guidance to the flight control which controls ct to get NZ,
and vehicle bank angle or roll about the velocity vector for lateral state control or direction of
flight.

FlaredRudder
Speedbrake

Flight Control Variables V

2
 Energy is the longitudinal variable to be controlled.

E = wh + mV2/2

E/W = h + V2/(2g) = h + qbar/(pg)

During the early 1970's, many organizations were involved in the development of thL TAEM
guidance. This author, in JSC's Avionics System Division, produced a Variable Entry Point
guidance. Candidates from others were the Racetrack from JSC's MPAD Division, the VO1_I'AC
from Charles Stark Draper Labs, the Cylinder from McDonnell Douglas, and the Spiral from
Rockwell International. A study contract was given to MDAC/St. Louis to pick, or blend into
another form, the final _uidance. The result was a "Hybrid" that incorporated good features
from all the candidates.

[22223

ariableOEntryPoint

VOR
_'T==

vo.T,c

fixed

. jo i"'"

Baseline

Candidate Guidance Schemes

3
Simplicity was a major emphasis in the formulation of the Hybrid guidance. Instead of
computing a ground track to satisfy the energy dissipation required from any initial condition
to a runway landing, as was done with the VEP and Racetrack, the Hybrid would use a
prescribed direct path to a cylinder fixed on the same side of the runway as the vehicle, nstead
of using an offset target to provide an acceptable ground track in case of a high energy
overshooting approach, as was done with the VORTAC and Cylinder, the Hybrid would Just turn
away (S-Turn) from the target cylinder (Heading Alignment Cylinder-HAC) for extreme high
energy cases, until the energy entered an acceptable level to head for the HAC. Instead of a
complex coordination of both vertical and lateral channels, as was done with the Spiral, the
Hybrid would separate the channels: a for vertical, and bank for lateral control.

2.1 ENERGY TO ATTITUDE CONTROL

This baseline guidance, established in 1975, was then subjected to additional testing that found
some poor performance and response characteristics. This baseline performed the major task
for TAEM of energy control, by controlling the energy with speedbrakes, and by controlling the
dynamic pressure with normal acceleration NZ. The first problem was that a firm control of
energy is not desirable for a situation where the vehicle flies toward the HAC in a taflwind, and
then turns 180" into a headwind. Energy control will bias altitude {h) below nominal and
dynamic pressure (qbar) above nominal so that energy remains nominal. At the end of the
tum, the low h and high qbar are the reverse of what is required to fly into a headwind.

EIW

Al_kudo Cor'_rol i . _i_"

)
i
Range

gO' of T_n

Energy Loss Turning into a Headwind

The second problem was the oscillatory response characteristics of controlling qbar with Just
NZ caused by qbar being a function also of drag in addition to _ (NZ).

4
 RI

D/M 'P o_

NZ Control of Altitude or Qbar

The vertical acceleration component shows close correlation with NZ,

g.NZ=L/M Cos((x)+D/M Sin((x)-g Cos0 Cos¢

hdbldot = I_/M Cos(C) Cos(y) - DIM Sin(y) - g

i.e., exactly equal when ot =-1' and ¢ = 0, and therefore good predictable quadratic response
can be expected from:

NZC = kl (hc - h) - k2 hdot

but, looking at qbar,

qbar = p V2/2

where,
P = P0 e'Kh
pdot/p = -K hdot
Vdot = -g Sin(y) -D/M

and the derivative of qbar,

qbardot = 2 qbar Vdot t V + qbar pdot/p

= -2 qbar ( g Sin(x) + D/M) / V - qbar K hdot

shows a gooC correlation with hdot, which by itself, would give a good correlation to NZ by the
above argument. But it also has a predominant drag term, where the representative numeric
example;

g=-12, g=32, D/M=2o32/4=16, K=5E-5,

qbar=200, hdot=-150

qbardot = +3 -8 +1.5

shows the -8 drag term to be larger and opposite in direction, and therefore poor unpredictable
response can be expect from:

NZC = -k3 (qbarc - qbar) + k4 qbardot

5
The change to controning Just the altitude component of energy with NZ solved both of these
problems. Maintaining reference h eliminates erroneous altitude biasing of energy control,
and the close relationship of NZ to altitude acceleration produces predictable and smooth
response. For simplicity sake, the speedbrake was changed to the other component of energy
which is velocity or qbar, and for the sake of approximately matching the controller of the
Approach & Landing guidance, which is based on airspeed.

This altitude controller also maintains control over both energy and dynamic pressure by the
utilization of midvalue limits that are calculated. This will be discussed in more detail later.

2.2 OPTIONAL TAEM TARG_[NG (OTT)

The Shuttle flight crew in 1976 expressed an interest in an overhead (OH) type approach as an
option to the straight-in (SI) approach to the HAC. The objective is to conserve energy for
subsonic dissipation in the vicinity of the runway. Another reason which became the real
selling point for making the change was thunderstorm avoidance at the Cape. The altitude
difference between SI and OH increases the probability for successfully flying either under or
over a weather disturbance along the landing site approach.

Note: OH/SI toggling should be done eady In

__K Entry guidance to rnlnlmlze trajectory maneuver.

Range-to-go OH is
Weather .. - 14n.mi.> than SI,
Disturbance i '_-_:_ & altitude is about
25000 ft higher.

OTI" Option

The development of this guidance modification started in 1976 as the Reverse Targeting
Scheme and later became known as Optional TAEM Targeting (OT13. The initial developers
and testers were the author of this paper, Vance Brand/Astronaut, Bob McNenny/MDAC, Gil
Carman/JSC-MPAD, Chuck Bowser/JSC-Crew Systems, and Ellis Henry/JSC-MPAD. The
initial development was on an off-line Univac batch computer, and then evolved to more
detailed testing and development on the Crew Procedures Simulator-Bldg 5, Shuttle Procedures
Simulator-BIdg 35, Shuttle Engineering Simulator-Bldg 16, and the Shuttle Avionics
Integration Lab-Bldg 16. The first flight of OTT was on STS-5 in 1982.

In addition to the option of SI or OH, OFT added or modified the following features: 1. The
Heading Alignment Circle-HAC was changed to a spiral. In three dimensions the circle is a
cylinder, and the spiral a cone, and therefore the HAC is now a Heading Alignment Cone.

6
 i 14O00
5000

Spiral HAC

RTURN = RF + RI_+ R2_2

Range = (RF + R1_/2 + R2_2/3)_/57.
RF =5K to 14K, R1 =0., R2=.093

2. The HAC radius is now adjustable after the _-,C turn phase starts. A low energy condition
of sufficient magnitude will start shrinking th_ radius.

AI!itude
HREF

HREFOH
H

Range

Spiral Adjustment

RF = RF -0.8 ( HREFOH - H) / ( aH/aR • aR/_RF)

where,
al-I/aR = Tan(15 °)
aR/aRF = _° / 57.3

7
 3. Provisions for HAC turns greater than 360 ° were made because Entry guidance roll reversals
can wrap an OH which starts less than 360 ° to greater than 360 ° .

>360" TAEM

4. An energy dump phase on approach to the HAC was added to alleviate the ground track overshoot that
results from large turn, supersonic HAC starts. Supersonic bank is limited to keep sonic boom overpressure
low. This procedure biases the target energy lower then nominal for large turn angles, which produces a dive
to dump energy, and then just prior to transonic, a pull-up maneuver targets for subsonic qbarmin. This
converts kinetic to potential energy, allowing the HAC to be followed more closely; because the subsonic
region starts sooner, the higher subsonic bank is used, and the velocity to be tumed is smaller.

/| ....iffo,,ow
if energy
dump & pull-up ,5 \
ma neuve/._f,_p---..._

Mach .9

ql:)ar140

/, Lr3OO
Start Pull-up

Dive to Energy
Bias

Large Turn, High Energy HAC Overshoot

8
 FJW
Nominal
ENV
P uli-uP _-_F.JW

f! ,._m_p =_

19 n.ml
'
26 n.ml
Range
270" 360"

Energy Dump for High Speed Turn

5. Also this low supersonic bank limit could, with default airdata, produce a serious HAC
overshoot and loss of vehicle because of delayed transonic indication from AD. The solution
implemented is that the low bank limit is increased if airdata is default.

2.3 BAILOUT MODES

The bailout mode was developed to provide an automatic flight system to maintain the Orbiter
n controlled flight to allow time for the crew to escape using the newly developed parachute/
, ole escape system. The design requirements:

_,irloads minimized by minim!zing airspeed.

Two fault tolerance on mode engagement resulting from either crew action or hardware failure.
Nominal manual flight operations maintained.
-_ngle-of-attack constrained to airdata maximum of 20".
=Vlinimlzesoftware to fit within the AP101B flight computer.

The level B changes were to set a bailout flag if mach <.95 and the pilot had both moved the
abort selection switch to the Abort to Orbit position and pushed the abort PBI. The bailout flag
would then force GRTLS to re-lnitialize to the acquisition phase.

The TAEM/GRTLS guidances monitor the bailout flag and the manual/auto status of both pitch
and roll. so that command values are obtained at the point that the pilot has selected those
values by going from manual to auto, individually in each axis. The pitch command selected is
dynamic pressure, which is controlled with the standard logic, by making the minimum and
maximum values the same as the command value. Also a new angle-of-attack pitch axis
constraint is added for both TAEM and GRTLS. The roll command is simply the bank angle
selected at the above snapshot. The speedbrake command is simply zero when the bailout flag
is on.

9
2.4 SMART SPEEDBRAKE AND BANK LIMIT FOR 0122

The baseline speedbrake controller has gone from energy to qbar and now eventually back to
energy although with a di_erent formulation from the original. The change request CR89979
is scheduled for O122, Jan 1993.

11=EN°
h
| ,

I/--'-
O.u-z------ H

Equk, alent signal : -EDTERI - (dEc I dR - dE / dR) dR / dt

Smart Spee_rake

The present qbar algorithm generally moves the brakes in the proper direction, except in
transient conditions such as a vehicle (energy & altitude) high, qbar low case. Energy will put
them out, whereas 0bar will tuck them erroneously in, until the altitude channel drives the
speed up.

The NZ and bank limit coordination are also included in CR89979. This will produce a
cooperative coupling, such as for a vehicle high condition during a turn maneuver in a
crosswind.

I0
 Equilibrium component of normal acceleration,
NZC Cos _ = Gravity (1.0) Cos 0

Total normal acceleration,

D=,_ li_n_
_ 2.2 [ NZ total ] = Cos e / Cos $ + [ t_c ] t_ _.s

To constrain NZ total to NZmax

then solve for bank limit,
NZmax = Cos e / Cos$ lim + NZC
Cos$ lira = Cos O/(NZmax - NZC)
$lim = Cos-l(Cos $ lira)

Smart NZ - Bank Limiter

The new computed bar]_ ]j'mit will provide more lateral acceleration to counter the crosswind
as the vehicle pitches down toward the reference altitude. The NZ is l:_ _ited for structural
reasons, and now if the vertical channel does not require much lift, then the lateral channel
can get more force up to the NZ limit. The reverse situation can also be found where the vertical
channel can get the maximum from NZ if the lateral does nnt require much force.

2,5 TI_TA

Erroneous airdata to TAEM could result in the NZ(qbar) control laws either overspeed
overloading the vehicle, or underspeed stalling, or losing energy on the backside of the L/D
curve. This can be prevented with additional constraints on NZC of upper and lower pitch
limits (0).

Omax = f(V, qbarmin, Speedbrake, Bank)

Omin = f(V, qbarmax, Speedbrake, Bank)

These limits can be determined empirically by flying an open loop entry simulation to the min
or max qbar for various setting on speedbrake and bank. This has been done and the results
were applied to the pilot displays to allow a manual override of the guidance. This could be
applied to the guidance utilizing another midval limiter added to the present NZC.

11 ORIGINAL PAGE 15
OF POOR QUALITY
 M
I S NZC
D E to FC
NZC V L
fromTAEM f

A E
L C
U T
f
E

Potential Theta (0) Limits for TAEM

An open issue though is where to place this logic. It is a guidance function, but it may require
the higher computation frequency of the flight control loops.

3.0 DETAILED DESCI_rION OF AI_ORITHMS

3.1 TECHNIQUE OF CONSTRAINTS LIMITS

A utility routine MIDVAL is used throughout the guidance that selects the middle value of three
inputs that can be mixtures of variables and constants or all variables. A cascade of these
software devices has the effect of prioritizing constraints. For example

KI
h

M M
I S I S
D E D E
V L V L f

A E A E
L C L
13

F
uT E

Cascaded Midval Functions

The lowest level constraints say that 12 will never exceed 11 , nor go below 13 , assuming 11 > 13.
The highest priority constraints say that no matter what 1 1,2 or 3 do, _ will never go above KI,
nor below K2, assuming KI > K2.

The most detailed application of this technique is seen in the TGNZC routine of TAEM where
three midvals are cascaded.

12
 -- NZ(qmin)
-- NZL
NZ(EUL) M M M
I I I
D D D NZC
NZ(h) V V V
A A A
L L L
NZ(ELL) U NZ(qmax) U U
NZLL
E __ E E

TGNZC Constraint Priorities

with the result:

1. NZ(h) is the nominal driver, so that usually NZC = NZ(H)

2. But if energy goes too far beyond energy nominal then the dynamic commands
NZ(E/Wnom+8000) Or NZ(E/Wnom:4000) take over. For example, a simple straight-in approach
to the runway in a headwind, could possibly not make it to the runway if the vehicle starts pitching
down to a less efficient L/D as soon as h is satisfied.
At V = 2000, where

E/W = h + V * V / (2g) = h + qbar / (p* g )

t_E/W = _Lh+ Aqbar / ( p ° g) = _h + V*V * Z_qbar/ (2g*qbar)

tLE/W = Ah + 500 Aqbar

Aqbar can be about 80 iow(140) from the nominal(220) for a L/Dmax case, which then at mach 2
would be equivalent to 40,000 ft of altitude. That is E/Wnom is not achieved until h goes 40000
higher than its nominal.
3. But regardless of the h or E/W situation, a higher priority is NZ(qbar), i.e., 140 min for max L/D or
300 max for max dive. If energy were low, it would only go lower if qbar were allowed to go below
qbarmin.
4. And the highest priority is vertical acceleration which prevents the wings from being pulled off
trying to get to an optimum qbar.

The remaining discussior_ _rfll follow the format of the FSSR's for both TAEM and GRTLS.
There are unique routine _or GRTLS which are given names that start with GR The
basically common routine reed by both TAEM and CRTLS are given names that start with
____.. The basic functions of TG routines between TAEM and GRTLS are the same, although
they may not be exactly identical. For example, TGCOMP has more linear reference profile
segments for GRTLS than TAEM.

Any I-Load constants values should be considered approximate where exact values should be
obtai_ed from the I-Load database.

13
The logic flow:

CONDITION

means IF (CONDITION) is true, then do Block 1,

else, if false, do Block 2. Return and continue past
CONDITION when finished either.

3.3 TAEM AND GRTLS _ - TGEXEC & GREXEC

The flow chart of all TAEM subroutines also shows the interface with the pilot for OTT, the
various phases of TAEM, and a brief description of changes to SB & Bank with CR89979.

Runway TGINIT _ i In= PILOT toggle HAC

Rodeslgnate P==_

I i =-'°
J TGXHAC ck_wB_
Iocam hedng
rekd_dlgnmm¢
t= Ihemrm=y
_ne i 1NAV UPP HAC
L='tmmm_t¢=_ _ C°nlr°l reld_
k>ca==hacrr'g
to turret/ ]

J GTP 4:ak,_ ,._ b tm ,_.,,y i_ha

re_t
I

I
J _
I o
I I[ S-T.m
HACA_pim'lim I

j ............

_ Is=_=_,=
I_'_ded prmo_n= _--E,_=r_,_,,
to qba re.orI=r ]

[ TGSBC =.c_l l , _l ¢ewl_ _°l_°ach &LW¢IGddanel ' J

_
[-v,=,_
Sce_IbrmkeWoq_rlor',al
_ inle_ral
i
h_4LoD_iml _ Bal_ Un_ maxknumIsa fun_lonol

TAEM Executive Row Logic - TGEXEC

The flow chart of all GRTLS subroutines also shows the interface with the pilot for OTI', and
the various phases of GRTLS.

14
 .I

Runway GRINIT
pauaii=Jial_._
II _ I I PILOT toggle HAC 1
Redeslgnate
l
t I
V
tog_ac

I TGXHAC do,
mranoer_lve
Ioca_ he_ _ Io iherurmay
corn J I NAV UPPom'_crosua_
_lnm,lml HAC Controlrslalk_
lec=u.s
_ Mad_
Rmway j

I
I G'n= _ r_p b I_ _

PHASEz4 _ IPHASE=4

1
_,d It._ , I & S-Turn GRTRN toM=.,
61o4
transibe= I
I_ 0ulmm_ i S NZHdd

I 0 S-T_m_
L o_wi_31 ! = P..oo_
I_ a -t I w¢_
! I._C Tom

TGNZC i '

.,om_li,oollWd_
_m_nd I ........
I
I
i

I
GRPHIC only,in _tl_mto,,,4_lHAC
TGPHIC bmkm'm_m_ ] t _k =le_n¢
crlelly II_n IbI l_,=o 4
S.Turn

GRTLS Executive Flow Logic - GREXEC

3.4 INITIALIZATION- TGINIT & GRINIT

All variables t_:c require inl_izatlon are done here. for either the first pass through TAEM
or GRTLS, or whenever during these guidances, the opposite end of the runway or a new runway
is selected.

3.5 HEADING ALIGNMENT- TGXHAC

The inplane X coordinate points of the Heading AJignment Cone are computed from the
corresponding pre-flight trajectory design input (I-Load) altitude points.

15
 l=Rat Caatrd af JtHAC p

0 I
- it l DR3

• • _ p •

• NominnH_e_mu

""" ' "' ! ,Ji:3

CROSS VL_/ i I ' '

I i IEp I 11011 1 _*1_ |

Ioos =T,_ T)

Geometry calculations of the Heading Alignment Cone

- TGXHAC

The vehicle nominally arrives in the final approach plane on the steep glide slope at HFTC
prior to the Approach & Landing Guidance at HALI, but if an energy low condition were
encountered and the pilot assesses that HALl will not be achieved, then the pilot can move the
HAC to XMEP to shorten the range to the runway.

The adjustment of the glideslope intersect point XA (IGI) is an autoland adjustment for
headwind. The adjustment should be made in TAEM, to allow time to settle on the new
trajectory, but it is not intended to be a TAEM energy control variable.

3.6 NAVIGATION USER PARAMETER PROCESSOR - HAC CONTROL

The pilot can toggle the HAC between overhead (OH) and stralght-ln (ST) with a keyboard item
entry. The software is outside the guidance and documented in the Navigation FSSR 6/30/85,
pages 4-207 to 210.

L
I I 1
I PlLOTtogglel-tkC
X

TOGHAC. OFF
PSHA = PSHARS-300

_OVHD = OFF I
YSGN - SIGNY I

OVHD = ON
YSGN = -SIGNY I

=+1
i'

NAV UPP HAC Control

16
The HAC information sent to the guidance specifies the side of the runway for the HA and the
targeting for either OH or SI. The HAC turn angle PSHA is reset to PSHARS at a toggle to
restart properly from a potential wrap around condition where PSHA > 3C_ _'.

3.7 GROUND TRACK PREDICTOR - GTP

The ground track, or trajectory component in the ground plane, is predicted for three major
phases by calculating the turn geometry of the acquisition phase by rotating the present
velocity vector toward the final turn spiral HAC, and then wings level flight to the HAC. The
HAC phase distance is calculated from the acquisition point around the HAC and into the final
approach plane, and then to the runway for the pre-final phase.

Pref_proach

GTP Phases

The GTP is not concerned with energy state at the present or end conditions. Other elements, to
be discussed later, will guide the vehicle energy to a reference at the predicted range.

The navigation state of the vehicle and the state of the HAC are specified relative to the runway
from which the GTP computes the state of the vehicle relative to the HAC and the vehicle range
from the runway.

Working backwards along a trajectory, the last part of GTP is simply the direct range vector to
the runway.

X " -XHAC

_ v J I
D_

Prefinal Approach Range Prediction

17
This direct range algorithm used to start with the last guidance phase (3), but as part of the OTT
modifications, this start was delayed until XCIR < DR4(2000) to provide a smooth range
calculation transition. A smooth bank calculation transition occurs with the phase 3 point at
DR3(5300 ft).

The next part of GTP is the analytical range around the HAC. For the spiral HAC,

RTURN = RF + R 1 PSHA(deg) + R2 PSHA 2

(14K) (0) (.093)

The range around the spiral,

RPRED2 = J Tan Vel • dt = j" (RTURN,PSHAdot/57.3) odt

PSHA
= j'RTURN/57.3 ,dPSHA
0

= (RF-PSHA + RlopSHA2/2 + R2*PSHA3/3)/57.3

and to the runway,

RPRED2= RPRED2+ XHAC

L------i I.... _RE_

" o)
RPRED = RPRED2 + RTAN _cA v

HAC Turn Range Prediction

The acquisition phase establishes RTAN the same as the HAC phase. The velocity vector is
then projected through a circular turn until parallel to RTAN. From that point the RC vector is
proJected to the HAC intersect.

18
 i

ARCAC = RTAC- I DPSAC I

RPRED = ARCAC + RC + RPRED2

Acquisition Range Prediction

The circular turn radius is calculated from the present velocity and an average bank of mach.

average

Mach

VH

I./M Sin(@ _

19
For 7dot =0,
L/M Cos(_) = g Cos(l)

RTAC = VH2/(L/M Sin(_)) = VH2/(g Cos(l) Tan(_))

= VHoV/(g Tan(_))

Acquisition Turn

An additional term is added for GRTLS because of the early S-Turn at high velocity with a low
bank rate of only 5 deg/sec.

DELRNG= V * t
ARCAC= ARCAC+ DELRNG

Acquisition Range Prediction

Additional Term for GRTLS

COMPUTATIONS - TGCOMP

General calculations that will be used elsewhere are basically done here, such as energy,
reference values for energy, altitude and dynamic pressure, and boundary values for energy. A
few exceptions, though, are that the action of filtering qbar and the control of the HAC final
turn radius are done here.

There are three variable indexing parameters for TAEM.: 1. IGI, which was discussed in
TGHAC, has two settings for glide slope intersect for final approach; 2. IGS has two settings to
allow reshaping of profiles for an optional heavy payload return from orbit. GRTLS does not
have this index, because the weight is known before lift-off; 3. ]EL allows two segments for
TAEM and four segments for GRTLS to generate reference profiles.

E/W = h + V / (2g) _ EN

IEL- 1

ENC2(_,S,tEL)

IEL. 2 ---"'_'t _
E NC 1(IGS,2)
15,000 ft
"" t Range to Nom H - 10018 [i
I
ENCI(IGS,1)
EOWSPT(IGS) DRPRED
18 n.ml

Energy Reference (Nominal) Profile

2O
GRTLS also does additional indexing with IES, IEST and IEM to generate S-Turn and
minimum energy profiles. These indexing differences are the main reason that the TGCOMP
software, used by both TAEM and GRTLS, are not identical, although the functions are the
sal-i'le.

The objective of the Energy Dump Maneuver, in cooperation with the Pull-up Maneuver m the
TGNZC, is to target for subsonic conditions at HAC initiation, to facilitate tracking the HAC.

I I

P_MAX

ENV
EN

E.o y
Dump,
___:_:_:_.i.-max ESHFIVlX,
.RPRED2 20O00
-R2MAX
DRPRED

Energy Dump

The empirically determined R2MAX is the maximum value of RPRED2 to start the turn phase
at the nominal energy and be at a low enough speed to track the HAC. The empirically
determined ESHFMX is set to minimize energy loss for large turns > 400".

The energy profiles are computed here and used for decisions in TGTRAN, or constraints in
TGNZC. One exception is that EMOH for TAEM, but not GRTLS, is computed in TGTRAN.
ES

EST
ENV
EMAX EN

_/" _MIN EMOH

__ EMEP

DRPRED

Energy Profiles

21
The energy boundary lines of ES and EMEP represent the limits of wings level dive or range
stretch respectively. The max stretch for an overhead approach, EMOH, is less than EMEP
because higher energy dissipation during the HAC turn. The EMAX & EMIN are used to
constrain the altitude controller in TGNZC. Energy is a better control parameter for a straight
in approach with either a wind condition where energy will bias altitude to compensate
velocity, or a navigation altitude error. But altitude is a better control parameter for non-
straight in wind conditions. The energy constraint on altitude tends to capture the advantages
of each. The bias values from EN are empirically determined from 180" turns into either a
head or taft wind on final approach.

The altitude reference is computed from linear and cubic function segments.

_1
HREF

HALl , cuucf°rGRTLS
_ HERROR

J l ln_g
DRPRED
33-45rl.rrl[.

Altitude Reference Profile

The equations for the Tan of the flight path angle reference, DHDRRF, is obtained by
differentiating the href equations.

The dynamic pressure reference is generated from two limited linear segments.

qbar
275

QBREF t QBERR
_-- QBARF j

DRPP_D
18 _.mi.
_R(NI

Dynamic Pressure Reference Profile

The radius of the HAC spiral at the final approach, RF, is adjusted ff the altitude goes below
HREFOH during the turn phase if the turn angle > 90'.

Altitude
HREF

HREFOH

Range

Spiral Adjustment

22
The adjustment utilizes a filter,

RF Filter

Assuming the HAC flight path angle is about 15 °, then

h = R Tan15, hc = Rc Tan15
R = RF-_, Rc= RFc-

Solving for command values,

RF c - RF = (h-hc) / (_ Tan15)

and substituting HREFOH for command altitude for the rate of change of RF,

DRF=(-.8/Tan15)(HREFOH-h)I(PSHM57.3)

The input dynamic pres _ ire to the qbar filter,

Qbar Filter

is selected from either measured airdata, if it is good and V < 1500, or from nax_ation derived
qbar.

3.9 TAEM TRANSITIONS - TGTRAN

This routine, used by both TAEM and GRTLS, performs transition of guidance phases and
issuance of alert conditi¢_, _ There are additional GRTLS phase transitions in GRTRN.

23
 IPHASE Description

S-Turn
HAC Acquisition
HAC Tum TGTRAN
Pre-final Approach

4
5 NZ Hold GRTRN
Alpha Transition ]
6 Alpha Recovery

End Start Approach &

TAEM Landing Guidance

In addition, the bailout mode is managed. When the bailout flag has been set then individually
the pitch and bank flags are set for bailout control of each axis whenever auto mode of that axis
has been selected after having been in manual mode. The manual mode drives to the
conditions that will be held by the auto mode.

The transition out of TAEM and into AL guidance for normal non-bailout operation is a
function of four error functions on h, Y, 7 and qbar, or it is forced at h of 5000 ft. The
transition criteria was provided for TAEM implementation by the AL development group.

The transition to phase 3 usually occurs at a range of RPRED3, but will be forced to phase 3 at h
< HMIN3(7000) to enable the AL transition for a low energy approach which can occur only
from phase 3.

The transition to phase 2 occurs when the vehicle gets to within 10% of the tum radius.

RCIR < 1.1 RTURN

Transition to Phase 2

This empirical transition point provides good bank transition by avoiding the HAC overshoot
of a later transition and a bank command sign reversal of an earlier transition. This earlier
start problem results from the radial rate damping term for phase 2 bank command. The ideal
RCIR start produces and initial bank command equal to the steady state value required for the
turn.

24
 Initial
phase 2
1.1 RTURN,_
_c RCIR
eady
rate

Initial HAC Bank Command

The transition to an S-Turn is done if E/W exceeds the profile line ES. The transition from the
S-Turn to phase 1 occurs tr_ TAEM when the energy returns below ES by a bias amount ENBIAS,
and in GRTLS when en_.',gy goes below the profile EST(see TGCOMP). The objective of the S-
Turn is to do an open ioop maneuver turning away from the straight-in solution until energy
gets to an acceptable value(EST) for straight-in. Attention is also given to the direction of turn
because a turn which intersects the HAC in the wrong direction is undesirable.

Wrong Direction S-Turn

The direction of turn logic that unwraps the HAC during an S-Turn prevents this problem,

S (Sign of Bank, + right) = - YSGN

25
 !

_ _IP "_

Right Direction S-Turns

Additional turn direction considerations are given for turns < 90".

<g0"

Right Turn (SPSI > 0, S=1) PS_D_, _'

LeftTum (SPSI<O,S=-I) _._"K,_

Additional Turn Direction Logic

To avoid an S-Turn geometry problem for low range condition where an acquisition solution is
not achieved after the S-Turn,

Low Range S-Turn

26
S-Turns are inhibited if DRPRED < RMINST (20n.mi.). At this minimum range S-Tum,

Minimum Range S-Turn

there is Sufficient room to turn and acquire the HAC.

The S-Turn is also inhibited in TAEM for tv: .-n angles > PSSTRN(200"),

.Ac
l

No TAEM S-Turn ">200 °

because the high energy HAC overshoot wig gain enough range and tra_,sitlon to a satisfactory
energy-range state without encountering a geometry problem.

B1,t for GRTLS that nominaUy requ='fe_ S-Turns in phase 4, PSSTRN is set at i000" to enable S-
T L_7nS.

Low energy alerts are issued to t_ _llot for his action. An energy lower than EMOH during
overhead approach is suggesting _¢pflot that he consider downmoding to straight in.

ORIGINAL PAGE
OF POOR QUALll

27
 Downmode
to
SI

to

Energy Alerts

An energy lower than EMEP is advising that the HAC be moved closer to the runway.

3.10 TAEM BODY VERTICAL ACCELERATION - TGN'ZC

The normal acceleration command from guidance to the flight control is an incremental
command of acceleration along the body negative Z axis,

NZ

/Y

Z
NZ Direction

and the flight control system adds to this a feed forward, steady state gravity compensation
term computed by matching the vertical component of NZ with gravity along that direction.

28
 Az Measured

Az
m Cos $ _$1

_m
-_Cos
,_'
Nzss = Azm _ ./4,

g _/_ gc_e

Pitch
Rate
Nzss = Cos 0 / Cos ¢ Command

from
NZc +_
Guidance

Az _ Azm

Flight Control Nz Loop

The primary guidance vertical controller on altitude,

HREF .1 +1 _c DNZC
.__ "_ERROR IHOREO G GI:)H

14K 24K

Altitude Controller

29
 is simply a second order controller, with the assumption that DNZC produces hdbldot without
time delay or error. The stability and response characteristics are specified by design to be at

a natural frequency con, and a damping ratio _, determined by the products of gains and the

ratios of gains.

COn= GDH_/(HDREQG ,.322)

= .054 to .179

The gain ratio is held constant as frequency varies with altitude and therefore the damping
ratio is slightly under critical damping.

4_2 = .322 GDH / (HDREQG*GDH) = 3.22

t;--.9

A more rigorous evaluation of the altitude controller transfer function involves NZc to hdbldot
transfer.

NZc = hdbldotc I g

Assuming that Az converges to the command state, i.e., flight control gains >> guidance gains
then

Az= NZc + Cos 0I Cos _

The measured acceleration is given by,

Az = L (Cos(z+ Sincd(l_/D))/(Mg)

and altitude acceleration is given by,

hdbldot = L (Cos_ Cosy- SinT/(L/D))/M - g

Solving from the above,

hdbldot= Q ( hdbldotc + g Cos 0 / Cos _ ) - g

where,
O = ( Cos $ Cos 3'- Sin _'/(L/D)) / ( Cos c¢+ Sin a/(L/D))

Substituting hdbldotc = (03n/(2_) ( hc - h ) - hdot ) 2_0_n, and transforming to LaPlace, with the
linearizing assumption that Q is constant,

h(S)=(Q COn2 hc(S)+L{g(QCos 0/Cos _ -1 )})

/ ( S2 + Q2_(.0nS + Q COn2 )

Numerical evaluation with liD=4, _=10, 7=-10, 0=0, and cl_=0gives,

Q=I
h converges to hc without error

3O
With $ =45, and COn=.179,

Q = .72
h converges to hc +hE
he = g ( Q Cos0/Cos¢ - 1)/(Q (On2) = 26ft
and the actual damping decreases

actual = "_ -E,design (.9)= .76

A modification that divided hdbldotc by Cos $ would change this damping decrease to a small increase,
.92.

The guidance has been formulated without this Cos_ division and, so far, no justification has been found
for a change.

Stability analyses of Gain and Pt_n_e Margin, Root Locus, and mainly integrated simulation
were conducted for the total G&C tern where flight control gain is not >> guidance gain. The
next lowest gain(qc/Aa) in the fhc control NZ loop is .47 at qbar of 200 which is not much
higher than the highest guidance gain of .322. The stability analyses, though, have all verified
satisfactory response characteristics and control margin.

The secondary vertical controller, _re for the dynamic constraints of dynamic pressure and
energy with NZ.

The maximum dynamic pressure constraint QBMXNZ is determined from hinge moment
constraints. The minimum QBMNNZ is the results from a flight at maximum L/D and is a
function of weight and ban1_ The profiles are generated as a function of mach.

340

OBMXNZ-Z_ 275 ( 300 for GRTLS)

184
QBMNNZ
138
5905.4 slugs

112
I I I I I I I

1. 1.5 2. 2.5 3. _¢
Mach
Dynam¢ Pressure Constraints

To facilitate trajectory tracking of _le HAC, a pull-up maneuver that lowers QBMXNZ acts in
cooperation with the energy dump of TGCOMP to target for subsonic initial conditions for a
large turn HAC. The maximum turn angle at which the HAC initiation is subsonic, with a
nominal energy approach, is approximately 270" at a range-to-go RPRED2 of R2MAX. At
higher turn angles the supersonic low L/D and low sonic boom bank limit can cause the range
to increase at HAC initiation as the vehicle flies past the HAC.

31
 V
RPRED2 RP_

I fl
/

R2MAX

EN

DRPRED
PO.MAX RPRElY2

High Speed Turn

The energy dump shifts the reference energy profile to target the same energy state of R2MAX at
RPRED2. The pull-up starts at an energy state EQLOWU 85,000ft, by lowering QBMXNZ by an
amount depending on the range error term and any flight deviation from the shifted energy
reference.

EtVV

EOLOWU
S:J
,, Du&mp .A
En_ ,
_ Energy Dump

Pull-up Dump ,
= Only i
! !

R21_0( RPI_ED2 ORPRED

Management of HAC Initiation Energy State

This energy dump & pull-up maneuver lowers the energy deficit that must be made up
subsonically.

The secondary guidance vertical controllers for minimum and maximum dynamic pressure
constraints each use the same flow diagram.

32
 QBMNNZ QBNZUL
&
& .oTs .o_s QBNZLL

Dynamic Pressure Controller

An added feature m GRTLS is that QBG2 is a function of roach.

Analytical stabfli analysis of qbar is not as clean as with altitude because there is no simple
relationship with NZ, but is a function of many variables and derivatives. Also, it is non-
linear.

qbardbldot=pVVdbldot+pVdot 2
+2pdotVVdot+pdbldotV2/2

The alternative approach used was to determine gains and stability empirically. The response
at the flight boundaries of either qbar min or max, is satisfactory. The unpredictable and
oscillatory respon.ce, though when used as a primary controller, was a part of the reason for
selecting altitude as previously discussed under "Energy to Altitude Control."

The secondary guidance vertical controllers for upper and lower energy constraints each use
basically the same controller as altitude, except that energy error replaces altitude error, and
altitude rate da.- -_ng remains the same. The gains were given new names, but to date, the
values have beer. _ _ same.

EMAX
or .1
DNZC
_ _. °_
GELL

GEHDUL
or

GEHDLL
= .322 / g= .01
Energy Controller

This controller assigns all of the energy change to the potential component. The final energy
will converge to the appropriate potential (initially unknown because the integral of drag*time
is the only thing that changes energy) and kinetic components. The stability is approximately
that of the altitude controller, but is mainly verified by simulation to be satisfactory.

Constraints on angle-of-attack are imposed for GRTLS only.

33
 2O

15

(X 10

5
/ ALPLL

0
I I I I I I I
.5 1. 1.5 2. 2.5 3. 3.5
Mach

Alpha Constraints

The controller for a constraints,

ALPUL 25. .005 ALPNZUL

ALPLL ALPNZLL
OR _____]-- OR

Alpha Controller

The convergence to a steady state will occur when ydot = 0.

1¢1ot = (L/M Cos(I)-g Cosy)/V = 0.

NZ=(Cos(z+Sinod(IJD))lJ(Mg)=DNZC+Cose/Cos¢

Solving for DNZC and a_ = ac - ct,

DNZC=(Ca+Scd(L/D))Cy/C¢- Ce/C$ = ou_25...005

Numerical evaluation with

I_/D=4, (x=10, 7=-10, e=o, and (1)=0, and
15 20 0 20 0, gives,

o_ = .1 to 1.8"

and therefore ct converges closely to (xc.

The baseline altitude and the six other dynamic constraint signals that have been generated
above are now combined using the "Technique of Constraints Limits" that was previously
discussed.

34
 Vertical Controller Constraints and Filter

where the alpha constraints are for GRTLS only.

The bailout mode takes a snapshot of dynamic pressure QBARS whenever bailout_pitch goes
from false to true. The upper and lower qbar constraints are then set to QBARS so that
NZ(QBARS) will be midvalue selected, regardless of NZ(h, E/W or alpha), to go to the CQG filter.
An additional angle-of-attack constraint is added for both TAEM and GRTLS (which replaces
ALPNZUL previously computed in GRTLS).

.005
ALPCMD___. ALPNZUL MI
18"
ql Z(QBARS) V

L
_ -5. A

NZ(QBARS)

DNZC IDI DNZCLID Dt

V" V

NZ(C_ARS)

Vertical Controller Constraints for Bailout

In the pre-final approach phase 3(and not bailout), when it is too late to do anything about
energy and maximum control tightness is desired, the filter is removed and Just qbar
constraints are used.

DNZC NZC ID tN_C

Iv t-_-
DNZUL ]A I FCS

DNZLL, _ ,

Vertical Controller Constraints for pre-final

35
3.11 TAEM SPEEDBRAKE - TGSBC

g8.8

65

Speedbrake DSBCU_
Limits

DSBCLL _//

Mach ._

80.8

DSBCE2_ ,015I/DSBCUL
_
2.5 3.2
a_d GRTLS
EOW ,w__; DSBCUL
DS_Mach

o D
>._

DSBLIM V
DSBCAT
98.6' _ S-Turn A
DSBNOM L

Speedbrake Controller

The speedbrake controller is basically a proportional plus integral on qbar error. An energy
term has been added to compensate for situations where qbar and energy may not be in sync,
such as, a 180" turn from taft to head wind. The energy can drop sooner than qbar. The energy
term will start to modify the brake upper limit at -427 ft., and will have it reduced to fun in by -
7000ft of energy error. A replacement of this controUer with a direct function of energy, at
least for the early part of TAEM, was discussed previously under "Smart Speedbrake for OI22."

The bailout mode sets the speedbrake command DSBC to zero so that the output DSBCAT will
be the lower limit DSBCLL function of mach.

3.12 TAEM BANK - TGPHIC

The upper and lower llmlt imposed on all the bank commands at the end of this routine arc
computed at the beginning as functions of mach and phase.

PHILIMIT
PHILIM
Ph=o
TJ_-UGRTtS
0 50 45
PHILMSUP
1 50 45
30"
2 6O 55
3 3O 3O
(wA£rdata
good)
4" 80 60
• Part of Phase 3
.85 Mach ._

Bank Limits

36
The S-Turn bank command is simply at the bank limit, in the direction of "S", as determined
in TGTRAN.
PHIC = S " PHILIMrr

Y j
YSGN! 1

S-Turn Pha

The acquisition bank command is proportional to heading error.

PHIC = GPHI (2.5) • DPSAC

- Acquisition Phase 1 Bank Command

or Phase 2 with RERRC • RERRLM (7000)

The re]ationship of lateral acce]eration and bank with steady state vertical conditions,

L/M Cos (I)= g Cos 7

L/M Sin _ = a
Tan ¢ = a / (g Cos 7)
(I)"~ 57.3 Tan (I)

The rotation rate of the velocity in degrees per second,

co= a/V = g Cos _,Tan (1)57.3 / V

(I)"~ (o V 57.3 / ( g Cos 7 57.3)

Substituting command values,

¢'c - ocV/(g CosT)

= 2.5 DPSAC

37
Solving for the control gain,

0)c I DPSAC = 2.5 g Cos Tl V

This lateral controller gain is small compared to the next inner loop gain(.5 to 1.2) in the flight
control,

PST DPSAC I (l]c _ _c I "

_°1 -.o8 *1 ' ' '_ t _l

FXght Control

Lateral Acquisition Controller

and therefore the response is exponential at about 1/.08 = 12 second time constant.

The HAC turn bank command is proportional to position and rate errors relative to the HAC,
and includes a feed forward centrifugal force acceleration term.

PHIC = -YSGN(Rdbldotref + GR.AR+ GRDOT ,ARdot)

=-1 _VH

VII

%
HAC Tum Phase 2 Bank Command

The relationship of lateral acceleration and bank with steady state vertical conditions,

Tan _ = Rdbldotc / (g Cos 7)

and the partial derivative,

a_'/aRdbldot = 57.3 Cos2_/(g CosT)

are used to derive the lateral controller,

38
 Rdotn' Rdbldatff . _c

.cl. t .oo_ ",_l "_ k

' .025 _ .22
0_12_ 2 C,(On

Lateral HAC Tum Controller

which has response characteristics for a natural frequency of .074, and is overdamped with
damping ratio _ of 1.5.

The reference radial rate is derived by differentiating the spiral radial command.

RTURN = RF + R1.PSHA + R2-PSHA 2

RDOT = (R1 + 2 R2.PSHA)-PSHAdot
PSHAdot = -VH 57.3 / RTURN
RDOTRF = -VH'RI+ 2 R2.PSHA) 57.3 / RTURN

The reference radial acceleration counter balances centrifugal force,

Rdbldotf = -Vt2 / RTURN

The implementation of this term involves the conversion to ¢c,

PHIP2C = Rdbldoff 57.3 / g

and then a quadratic, _rve fit from TanCc to ¢c.

TanCc = PHIP2C
PHIP2C= PHI2CloTan0c - PHI2C2.Tan2@_
1.13 - .0055

OC PHIP2C

30 31.4
45 46.7
60 58.0

The pre-final bank command is proportional to lateral position and rate errors relative to the
runway.

PHIC = - GYoY - GYDOT-Ydot

Pre-final Phase 3 Bank Command

39
The relationship of lateral acceleration and bank with steady state vertical conditions,

Tan ¢c = Ydbldotc / (g Cos 7)

_c ~ 57.3 Ydbldotc / (g CosT)

The lateral controller,

. Ill Yd=c _c $c

i .1 L_..J I .22 1.86

(tin 12 r, 2 r, ruOn

Lateral Pre-final Controller

has response characteristics for a natural frequency of. 15, and a damping ratio _ of. 73.

A fader which starts at the first pre-final pass with ISR=5 allows the bank command to change
by only 1/5 the difference between old and new command. On the second pass 1/4 is dumped in,
and so on, until at ISR= 1 the system fuUy uses the phase 3 PHIC.

Simulation testing revealed two oscillation problems. A case of high energy with HAC
overshoot and spiral shrink can produce both situations.

A Previous Bank Oscillation Case

For the HAC phase the oscillation was ellminated by reverting back to the acquisition bank
command and its lower limits when RERRC > RERRLM(7000ft). For the pre-final phase the
oscillation was eliminated by allowing the bank limit to increase from 30" to 60' as the
command increases to 100".

PHILIMIT = MIDVAL(30,60, .43(ABS(PHIC)) + 17.)

The bailout mode takes a snapshot of bank, PHIC_ATS, whenever bailout_bank goes from
false to true, and outputs PHIC_ATS as the bank command.

4O
3.13 OVERVIEW OF GRTLS OPEN LOOP _ANCE PHASES 6 TO 4

GRTLS phases 6 and 4 use the angle-of-attack command mode of the flight control system(FCS)
that is not otherwise used in TAEM. The regular normal acceleration command
NZC/TAEM/FCS mode is used in phase 5.

The open loop feature of these phases is that guidance does not close the loop with longitudinal
state commands to drive ac or NZc, although a and NZ are of course closed loop within the FCS.
An exception is that phase 4 lateral control is a guidance closed loop direction command that
is either toward the HAC, or away from it for a high energy S-Tum.

The objective of this part of GRTLS is to aerodynamically reduce altitude rate after falling into
the atmosphere, and to transition to front side L/D flight at mach 3.2 for normal TAEM closed
loop guidance energy management. Typical GRTLS atmosphere re-entry profiles show the
angle-of-attack going from about 10" at ET/SEP to, and holding, 50' for the phase 6 Alpha
Recovery.

50

(%

0 I l I I I

7 6 5 4 3 2

Mach

Alpha Profile

After the rate-of-descent peaks, a calculation is made for the NZ level to transition to phase 5
NZ Hold at which point a constant NZ near 2 is held to reduce the rate-of-descent.

NZ

0 t I I I I

7 6 5 4 3 2

Mach

NZ Profile

41
At a rate-of-descent of-HDTRN,

1500

1000

-hdot phase

5OO
phase
4 to ISTP4

I .J. l _ t

8 5 4 E3

Mach

Rate of Descent Profile

the transition to phase 4 Alpha Transition, then commands alpha to a profile GRALPR, until
TAEM phase 1 (or 0) starts at mach 3.2.

3.14 GRTLS TRANSITIONS - GRTRN

The transition to phase 5 occurs when NZ builds to a computed value, NZ + DGRNZ.

_ - phase

:_ NZSW1 + DGRNZ
NZ

1.85 + lime GRALPC)

M=m.x
(m GRALPC)

0 I l I i I

-1700
-1600 -I._0 -144_ -1300

hdot

Transition Phase 6 to 5

The alpha profile to be flown in phase 4 is computed in GRTRN, SO that a test with it can also be
used in the transition to phase 4. The transition is mainly on hdot > HDTRN, but also alpha
must be > GRALP_

42
 18

(X

13 I I I I

7 5 4 3 2

Mach

Phase 4 Alpha Transition Profile

This a is increased for phase _ -Turns, up to the lmUt of AXL .

GRALPR = Min(GRALPR / ]Cos ¢1, AMAXLD)

A correction to NZ,

CORN7 NZ - Cos0 / Cos_) (TAS/(TAS+VCO)

involves subtracting the C0/C¢ t: _rom the total NZ to get an initial NZC reset of the filter
used in TGNZC. This replaces the el of NZC last used in phase 5 which would produce a big
transient at phase I initiate if left

"he logic for transition to phase IS:_ o_ (1 or 0) involves energy, NZ, and mach, but the I-Load
_,sage of EOWL1--0,and MSW2=MSWi=3.2 produces a transition based only on mach of 3.2.

Low energy alerts are issued during phase 4 to the pilot for his action. An energy lower than
EMOH during overhead approach is suggesting to the pilot that he consider downmoding to
straight in.

Downmode
to
SI

Energy Alerts

An energy lower than EMEP is advising that the HAC be moved closer to the runway.

43
ORIGINAL PAGE IS
OF POOR QUALITY
3.15 GRTLS BODY VERTICAL ACCELERATION - GRNZC

The NZC for phase 5 is generated as a function of a linear and an exponential term to go from
the initial to the final constant command value.

+2

DGRNZ
(+.2,-.3) -.3

GRNZC1 NZC
1.2 _ =-
1. - SMNZC3 .125,:

.___ .25e E SMNZCl(_

SMNZ2L
MIN -.05

1.2

1.1
NZC

1.0

.9

.8 I I I J

0 10 20 30 4O
Time, sec

Phase 5 NZC

3.16 GRTLS ALPHA RECOVERY - GRALPC

The angle-of-attack command for phase 6 is ALPREC 50".

The maximum descent rate is captured here, and the DGRNZ used in GRTRN & GRNZC are
calculated.

44
 .2 m

I_RNZ hdot
-I
,8 .!sss ,4 8
HDMAX ___
-.3

DGRNZ Calculation
DGRNZ=MIDVAL((HDNOM-HDMAX) DHDNZ, DHDLL,DHDUL)
-1558 .002 2. -.3

The angle-of-attack command for phase 4 is initiallzed at the a at the start of phase 4. The
command is then ramped down toward GRALPR at a limit change of GRAL, + 1°. When a falls
below GRALPR, then ac is clamped to GRALPI_

28 ¢x =t sta/t
04phase 4

20

I I !
12
5 4 3 2

Mach

Phase 4 Alpha

3,17 GRTI_ SPEEDBRAKE - GRSBC

The speedbrakes stay tucked in at zero unU1 pitch Jets are shut off at qbar of 20.

QBA_

> IC
SBQ O. GRSBL1

DEL1SB
20. _ DSBCAT

t__ =..
2.5 3.2
Math

GRTLS Speedbrake _

The command then ramps up to 80.6 and, then at roach 4, ramps back down to 65. for TAEM
interface.

45
3.18 GRTLS BANK - GRP/tIC

The bank command is zero until phase 4 at which point the command is the same as phase 1 in
TGPHIC to acquire the HAC, and the same as phase 0 in TGTRAN for high energy S-Turns. The
flag ISTP4 tracks the S-Turn status so that the transition in GRTRN to TAEM will be either to
S-Turn phase 0, or acquisition phase 1.

S-Turn

V -DPSAC

Acquisition

([_c= 2.5 • DPSAC

Right Turn ( SP81 > O, S,,1) PSD

Additional S-Turn Direction Logic

Bank Directions

ES

E/VV S-Tu_ EST

EMEP

DRPRED

S-Turn Energy Profiles

GRTLS Phase 4 BankCommands

The only exceptions from normal TAEM operation are that mach must be less than MSW3(7.0),
and the I-Load PSSTRN is set high at 1000. to enable S-Turns for all overhead approaches.

46
 REPORT DOCUMENTATION PAGE Form Approved
OMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering 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 collection of information,
including suggestions for 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 Reduction Project (0704-01BB), Washington, 0C 20503.

1. AGENCY USE ONLY _eave blank) 2. REPORT DATE 3, REPORT TYPE AND DATES COVERED
November 1991 Technical Memorandum
4. TITLEANDSUBTITLE 5. FUNDING NUMBERS
Space Shuttle Entry Terminal Area Energy Management 551-15-1216

6. AUTHOR(S)
Thomas E. Moore (NASA)

7, PERFORMINGORGANiZATiONNAME(S)ANDADDRESS(ES) 8. PERFORMING ORGANIZATION

REPORT NUMBER
Guidance and Control Branch/EG2
S-661
Ascent/Entry Section
Navigation, Control, and Aeronautics Division
Johnson Space Center
Houston, Texas 77058
SPONSORING/MONITORINGAGENCYNAME(S)ANDADDRESS(ES) 10. SPONSORING / MONITORING
AGENCY REPORT NUMBER
National Aeronautic and Space Administration
TM 104744
Washington, D.C. 20546-001

11. SUPPLEMENTARY NOTES

12a. DISTRIBUTION / AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Unclassified/Unlimited
Subject Category 01

13. ABSTRACT (Maximum 200 words)

A historical account of the development for Shuttle's Terminal Area Energy Management
(TAEM) is presented. A derivation and explanation of logic and equations are provided
as a supplement to the well documented guidance computation requirements contained
within the official Functional Subsystem Software Requirements (FSSR) published by
Rockwell for NASA. The FSSR contains the full set of equations and logic, whereas this
document will address just certain areas for amplification.

14. SUBJECT TERMS 15. NUMBER OF PAGES

Shuttle Guidance
TAEM 16. PRICE CODE

17. SECURITY CLASSIFICATION 18. SECURITYCLASS'IFICATION 1'9. SECURITY CLASSIFICATION 20.LIMITATION

OF ABSTRACT
OF REPORT OF THIS PAGE OF ABSTRACT Unlimited
Unclassified Unclassified Unclassified

Standard Form 2gB (Bey. 2-89)

Prescribed by ANSi Std 23g-18 NASA-JSC
2gB-102