Professional Documents
Culture Documents
Ada204865 PDF
Ada204865 PDF
j-4
Wq
~-
'~f'-~~'----'
-~ti~t-o-,
-k
IV
-- Ui
'i~3
AIAA
MONTEREY,~
CAIORI
Lr-f E
NAYBLSI
MISSILE
SCECSCNERENCEOG
MONTES
REYTE CALEIFONIA
CTEm-!
LD
00
r
"
AIAA
0
NMISSILE
SCIENCES CONFERENCE
IMONTEREY,
CALIFORNIA
VOLUME 6
NAVY BALLISTIC MISSILE TECHNOLOGY
Chairman:
Vice Chairman:
QD, COt
ot-''.
(1)
-AVIONICS
Robert E. Brainard
Lockheed Missiles & Space Co., Inc.
Advanced Concepts, 0/81-71, Bldg. 157
Sunnyvale, CA
94088-3504
Phone:
(408) 756-3122
(2)
of
a solid propellant missile ladened with
aluminum oxiie particulate. The analyses focused
on absolute ultraviolet U)
radiometrics of the
booster oljme and its afterburn as a function of
altitude.
(Afterburn is the reignition of fuel
rich
missile
exhaust
with
the
atmospheric
oxidizers.) The stability of the booster plume as
a pointing and tracking source for hardbody
in azimuth.
_ACKGROUND
the aircraft.
There is a limited UV data base on plumes for
acquisition, oointing,
tracking, and hardbody
hand-off on liquid and solid missile launches. In
Seotember
1987,
an
ultraviolet
and
visible
observation of the scheduled test for November
1997 was organized to view a solid orooellent
missile.
Through short-term Herculean efforts of
the team from Jot Prooulsion Laboratory, LORAL
Electro-Ootical
oivision,
and
Aeromet
(the
ooerator of the Learjet optical observatory), an
5.5"
ultraviolet
installed
in
the
telescope
Learjet.
and
camera
.... I
W.D
'
,_
STATIONRY
MIROR
____
was
swc,o
successful
TELSCO.
,,.,,,oAUV
CAAEA
40dM 01*.P0OC*TOI[
00f[S~
,,0USOTOAurOO
~, Coo ,v,,
-A QRAi
[SOLUTION
~fl.E
SYSTEM DESCRIP'rrON
AfCAJTr0,
HHTE,,,OA
S,,
SMember of AIAA
Dati
recorded
M SSE
TRAJECTORY
5.5- ASPECT
A
ALT ,
In Chamber
Exhausted
Pressure
400 psia
1 psia
Temperature
5046OF
1844 0 F
H2
1.407
1.512
A120 3 (solid)
0.283
0.297
CO
1.116
1.059
N2
0.293
0.294
WC
0.526
0.579
0.032
0.088
C02
H2 0
0.297
0.021
Cl
0.021
trace
AiC1 2
0.005
trace
HO
0.006
trace
AICI
0.016
trace
0.084
trace
174
230
ELEVATION
1101.
68T
---
32096
=:GLE.EV*_TN
AL
A2
230;5
'ALT
INAETERS
OISTANCE
EGRES
ANGLESN
r"
Figure 2.
ELEANGLE
455"
1200 noon,
engagement
the night
that of
nearly matched
geometry
The geometry is shown in Figure 3.
exoeriment.
At launch time, the sun was about 48 degrees from
the local Zenith. The ultraviolet camera operator
acquired and tracked the first stage motor nozzle
A
through second stage ignition and burnout.
lower gain setting was initially was used to limit
and not saturate the ultraviolet camera. As with
the night mission, automatic gain control limited
the visible camera input, but lid not record its
settings
Taole 1.
o"
22,70
,/
A)C
OPERATIONS
A2Lo
(mole/100 grams)
3770
"
so
that
the
camera
did
not
yield
radiometric data.
Night Mission
On 10 Nove-mber, the HALM was able to observe the
This was
missile launched at 17:49 local time.
anoroximately 45 minutes after sundown, making the
Figure 2
solar tangent ray at 150 kilometers.
shows the geometry. The missile stage burned from
34.5 to 42.8 kilometers altitude causing the
Separation
camera system to look uo 45 to 55*.
an-i thrust neutralization (three plugs on the too
of the stage releasing to allow venting through
them) folloed by a destruct sequence which blew
venting)
occurred
the
case
(called
case
iimeiteiy lfter burnout,
Aooroach
Since the data were of varied quality, several
aooroaches were taken to maximize the technical
gain.
The manual tracking resulted in residual
line of sight wander and aircraft disturbances
caused line of sight jitter. Plume jitter values
using centroid and edge tracking algorithms were
estimated.
The ultraviolet camera was radiometrically
calibrated,
but
the
visible
data
were
ASPECTANGLE
,
32246
AC
ALT 12894
4 AL
,
,
ALT
'
11523
AFT LOOKING TO AC
43011.4
-8315.6
'69.1.
11464
1000
lOG
---------------.
25365
A LT 1284 5
78.7
... .
/ 101.3'
SE
C T ANGLE
5r
12882
10
3,'4696
ALT
0.1
,
Wa
0.001
a4
I E-5
GROUND',,TRACE
'
'
ELIE 44.5
"
-"&
ANGLE
1000
2000
3000
4000
6000
6000
WAVELENGTH
7'.3*
Sensor
Ultraviolet
Extraction Discussion
Calibration
and
Data
ATMOSPHERE
out fro
between
@3300 A
HI0.82
[HIT.S
0.70
QUARTZ
STEERING
a few meters.
H1 0.88
FERRANT
.JPL
UV
TELESCOPE
HOO.95
UVDIAGONAL
LO090
F
r-
MOST PROABLE
___________K
__33_A
ATIOSPHERC
A3
0.79
AITZwoow
o0
70
FERRNTEMI
Vmost
MOSTq
THRU
PUTI
SYSTEM
I
I
0./
0.5
V) 0.4
;ANGLE
0.5/93.0-
OAY
48 DEG/
ANGLE/
0.3 -
difficulties
02
---
2400
i/quantitative
I0.2
S
2600 2600
""
3200
3400 3600
3600 4000
WAVEI.FNOTH
7igure 6.
Unfortunately, tracking
and
saturated
data
make
the
reduction of this set most difficult.
,!
3000
degrees
Our best
11
This means
data
which
are not
that the analysis
of the
night
saturated
needs only
minor corrections for solar
/ZENITH
about
.
IGHT
0.
0.0
....
0.7
Z 0.6
ZENITH
ANGLE
that
T.02523
0.9
is
so
090
STErEAGM=O
J.uvr EsCT
E
source at 40 km
altitude, with a
line in the top
flux, and the
of
-o
:NOU
7054
.................
SPECTRA
V.,
L
-5,
."
711
RW
AND FILTER
5.0
IlSTCkbody
Fig-
0I2
7.01
F
S2.34
3360
WAVELENGTH
0460
15600
.00
6000
5120
4240
(A)
I---- I
*I
36
3360
2480
0Night
4240
CkboAy
,I,,
00
1600
5t20
6000
WAVELENGTH (Al
rRAOIATION
0.6
SEEN
,N
I-22)
,.,,,,SRS*,-0,
4
T
06
IT
Figure 9 developed
observed light.
0.2
1600
2460
3360
4240
WAV' ENGTH (A)
5120
Atfts.*
L 0. 3
U 0.07
"..
2400
is
radiation
at
the
the
lower panel
with aircraft
the quantun
of the photocathode, which is the
incoming
The
convolvel
in
0600
260
3000
CMC"*'uW"
L .13
X
3200
3400
3600
WAEENO H (A)
Fl.
3800
4000
.200
,3
.0
2400
the m issile .
60
2600
0.1 2
L
U
2800
3000
3,0
--3 0 ---360
3
3200
34
3600
3600 -'-000" 200
WAVELENGTT.
O t.
0. 16
ii.ii
NPUT
SPECTRA
AN FIL.TER
I----I
K , 1,4 ,,f
............
0,
Chu2.in
of
0.2
efficiency
spectrum
effective
6000
-1gure 7.
the
.,.
0.06
0
1600
2033
2467
2900
3333
3767
4200
3333
3767
4200
WAVELENGTH
(A)
RADIATION SEEN
in
the
plume
intensity
or
3260
0.6 o,
shape,
tracker
that
might have
F,9/i.S
o.
C
s
,.,
"
0.2
0
1600
its centroid
2033
2467
2900
WAVELENGTH(A)
RESULTS
atmospheric
transparency
between
the
40
Im
altitude and the aircraft altitude of 11 km is the
dashed function. The resultant relative intensity
and the
response.
signature of
the particle
size. A simple single number is not sufficient to
orooerly address this issue.
Similarily, the
optical
thickness
of the
plume at(280several
-avelengths
in our region
of interest
- 400
nm) is not a constant. The observed radiation is
-ompared on the solar scatter from the spatially
varying particle size distribution, an] by the
emission of the hot plume particles as they leave
the nozzle.
There is also a possibility that a
hotter source region inside the nozzle can have a
single scatter interaction with particles just
outs-ie the nozzle to cause observed radiation at
an apparent blackbody temperature higher than that
of nolten alumina particles (about 2420 K).
Figure 11.
---- --ULTRAVIOLET
--
,400300
14.
,
0
WAIST WIDTH
BLOOMING
VISIBLE
VIOTH AT--------
90
---
32
304
ULTRAVIOLET
34
36
30
3B
42
40
TIME (SEC)
44
30
40
37
36
40
41
image
error
---....
20 :PLUME HEAD
42
ALTITUDE (XFT)
12.
the
radiance
of
those
pixels
that
Figure 14.
iso0
are
1400
Z200
1000
1@DO
0
400
-1-
200
0
HALO A
POE
00
Visible
uv
2 40
101.2
20
033
3435
3 18 30
3
4'0 4'14'2 -4
4
TIME (SEC)
PWR>0%34.5
1 :
'445463'
35
36S
37
368
39
40
41
42
ALT'TUDE (KFT)
I0.8
Figure 15.
0.4
The
0
MISSILE
ALTITUDE (m
1000)
along
T hir
ig~ure
3,
a visible plume
for
A centroid tracker
of 1500 meters.
handoff to the hardbody of about 750 meters wouli
require error limits of less than 0.1% in listance
,upwards
the
plume just
to
intersect
the hardboiy.
HALO 8
0.0ao
a
0.065
A binary
POWER>90%
0.050
POWER >50
Thresholding
TOTAL AVERAGE POWER
-igure 16.
36
that
eliminate
0.0_
34
is a technique
36
40
42
MISSILE ALTITUDE (m X 1000)
44
.-
..
.-
P.--.--
hintensity.
"atmospheric,
ZfJ
..
.. ..
-and
_
-
"-'
.-
-of
.
._-__
_______......___
.-
cannot
be
_r
motion.
:igure 17.
that
,_...___
error
the
low
frequency
incoherent
plume
oThrelhold
IS
A lgori the
%|
901
So
Y
Leading
ginary
21.0
11.9
21.0
11.1
10.2
14.4
2.6
12.5
L,
e.din
21.3
12.7
21.3
11.4
12.1
14.4
7.0
14.9
Errors (Pixels)
-Track
.. . .....
350
3000
2500
2000
1500
000
500
00
DATA
INDEX
:igure 18.
____
CL
_indicates
00
_____
1000
O-
500
500
500
2000
00
000
SO02000
25000F
3000
7
7.54
3505
1600
ot
A11.-ort:
2460
5120
600
3360
.240
WAVELENGTH (A1
13. 7
21.2
14.6
12 .4
16.4
7.8
Cont.ro LO
9
y41
5.9
8.1
5.4
8.2
5.9
10.1
6.1
9.1
lmt0-.LtY
0,1
7.6
0.7
7.5
0.
1.2
6.1
FILTER
INPUT SPECTRA&0
o.0
2~dn
1.2
Edge
0005
,,i/.A
1400
90
50
y
0.
5120
XU
234
5
Y
4.67
360
2400
7.01
Threshold
.'
,.oe'v!
DATA INDEX
igure l9.
INPUTSPECTROAAND FILTER
Vml.........................................
17.00.4
fl
'
00.0,0..
10~0
r
04
30
04
,
10
io
60
WAV tr ELEN
7.2
POOI0W
0.6
a0le 2.
_AA.60
SEEN
00.n6/-.
n,?I
-!r of the intrinsic plume stability is the
relative motion of the plTns leading edge to the
,)line centroid.
To study this measure, the
Jifference between the leading edge an.dcentroiW
Qrr')r' as
9 nouted (ee
Trable 3).
a
MOO
2.40
Figure 20.
'A
120
Sim
6000
o-inel)
also convolves
this
input
function
CONWSIONS
(solid
line)
with
the transmission characteristics
expected from the new filter.
The third panel
convolves these two functions in the solid line,
and clearly shows the proposed improvement.
The
Final
convolution
with
the
present
photocathoie is given in the lower panel.
The
third panel shows the filter convolution (solid
line) and the photocathode response (dashcd line),
The resultant performance is in the lower panel
v-I illustrates the superb quality of the proposed
The total transnission of this
filter addition.
filter is greater than 70%, so the losses are not
i major issue.
the automatic
in the visible.
The consequences of this tyoe of plume are that a
centroid tracker that works well with a round
target would be hard pressed to find the centroid
position along the olume unless it is tracking the
UV plume ootside the atmosphere.
Thus, other
Other Improveoents
Other imorovenent
include
removing
IRIG
10
AVIONICS
FOR A FRACTIONATED
SLBM
Robert E. Brainard
Lockheed Missiles and Space Company, Inc
Sunnyvale, California
Abstract
Ballistic missile vulnerability to
exoatmospheric encounter by space deployed
weapons can be significiently reduced by early
deployment of reentry bodies. For a Submarine
Launched Ballistic Missile (SLBM) the bus can be
eliminated and the third stage replaced by third
stages attached to each Maneuvering Reentry
Body (MARB). This missile configuration
requires a different approach to avionic design.
An integrated avionic system is described
which provides guidance and control for the
booster stages, fractionated third stage and
MARB. Avionics include missile electronics
located with the booster second stage and
guidance and control sensors and electronics
located in each MARB.
MARB avionics requires high accuracy and
high reliability with severe constraints from
available space and the rugged environment,
Lack of operational accessibility for repair or
replacement demands high reliability.
The guidane and control sensor is a Stellar
Inertial Measurement Unit (SIMU). It is a
strapdown, configuration which includes a
stellar sensor, ring laser gyros and solid state
accelerometers.
Technical challanges for avionic contractors
are the design and development of a strapdown
stellar sensor and solid state accelerometers
which can meet requirements for high accuracy
and ran operate in the rugged environments,
Eleronic miniaturization also requires
innovative design.
Introduction
Current SLBM configurations have three
boost stages and a bus. The bus is used to
provide precision impact accuracy, footprint
capability and precision deployment of multiple
independently targeted reentry bodies. The bus
operation requires several minutes of time to
accomplish its objectives. The bus operating
time plus the three stage booster operating
time results in reentry body deployment above
the Earth's atmosphere where the missile with
undeployed reentry bodies becomes vulnerable
to exoatmospheric encounter by space deployed
weapons. 1,2
A proposed solution to this problem is to
eliminate the bus and replace the third stage
with a third stage attached to each reentry
body. If the reentry body is a MARB it can
correct for third stage propulsion errors and for
prelaunch and boost guidance errors which
become observable during coast after third
stage burnout.
This paper describes a reference missile
configuration, addresses the avionic
requirement issues and describes an avionic
configuration which can meet those
requirements.
Design Objectives
Design objectives are to decrease the time
after launch for deployment of all reentry
bodies while maintaining missile accuracy,
range, throw weight, foot print capability and
reliability.
Missile Description
THIROTAGE
MAR
gas
MARS AVIONICS
Bus
FRACTIONATED
THIRoSTAGE
SECONDSTAGE
SECONDSTAGE
FIRSTSTAGE
FIRSTSTAGE
AVIONICS
Ecs-TRONICS
Avionic Description
The missile electronics contains
* Missile computer
* Interfaces
* Missile electrical power
The missile computer performs guidance and
control computations during standby, prelaunch,
first stage flight and second stage flight until
MARB deployment occurs. Guidance and control
navigation data is generated by each MARB
avionic system. That data is sent to the
missile computer for missile guidance and
control computation. The missile computer also
cntrols and responds to the transfer of data
between submarine fire control via the missile
umbilical and each MARB computer. Missile
electronics also contains missile electrical
power which includes the missile battery and a
power distribution system. The battery
provides missile internal electrical power. The
power distribution system controls the
availability of missile internal and external
power and MAr 3 external power.
MARB avionics include
" Stellar Inertial Measurement Unit (SIMU)
" Electronics
" MARB electrical power
Figure 1
FRACTIONATED SLBM
SLBM Configurations
BOOST
COASTJOI4WdG
PLOY FRACTINATEO
MAR'STAGS
ROfiECET OR
BOOST
FISTSTAGE
TIDSA
*T"A"O
STAGE
SECOND STAGE
ACCELCA.
TARGETATA
NAVYC
Low power
UPDATE NAWIC
L
ACCEL BIAS CAL
ACCELCAL
Figure 2
Flight Sequence
Flight Sequence
The flight sequence is illustrated in figure 2.
During prelaunch, target data from fire control
can be updated, navigation initial conditions are
provided, and the component of combined
longitudinal accelerometer bias and scale factor
is estimated for each MARB SIMU. The
navigation initial condition data is obtained
from SINS. This includes longitude, latitude,
height, azimuth, and velocity,
During first and second stage flight, the
missile computer will process the MARB
navigation data from each MARB to determine a
best estimate of navigation data for boost
guidance and control. By processing all
available MARB navigation data, mission
accuracy and reliability is enhanced.
Reliability Issues
As a design objective, missile reliability is
to be maintained. The Fractionated SLBM
configuration has major avionic components
installed in each MARB, These components are
not physically accessible for repair or
replacement while the MARB is installed in the
missile while in the submarine launch tube.
Consequently, the standby, or dormant,
reliability of the avionic components must be
very high.
Periodic testing with built-in-self-test
(BITE) will be performed to determine the
existence of disabling failures in the MARB
avionics. Because MARB avionics provide
redundent guidance and control navigation data
for the first two stages of the missile, it can
still be launched when MARB avionic failures
have been detected. The missile guidance will
be designed to operate in the presence of failed
RB avionics and still meet safety requirements.
It is of course prudent to repair that missile at
the earliest opportunity.
To achieve high reliability, design simplicity,
minimum parts, an effective parts reliability
program and component burn-in are required.
System design must achieve low power, with
efficient heat dissipation to minimize operating
temperature. The design must be constrained for
power off during standby to minimize the initial
starting temperature and contributions from the
higher failure rates associated with energized
conditions. Of course, these reliability design
constraints and objectives can make attainment
of accuracy objectives more difficult. The
selection of the strapdown, solid state, and
miniature electronic design approach is intended
to meet the reliability design constraints and
objectives as well as the functional and
accuracy requirements.
Available Technology
Avionic tecinologj which is available for
implementation of the Fractionated SLBM design
are
Flight computers
. Miniature, low power electronics
* Hardened digital electronics
* X ray shielding
. Ring laser gyro
* High
reliability with restricted
maintenance constraints
Lockheed Missiles and Space Co., Inc (LMSC) is
developing a radiation hardened computer which
can perform the required computations for the
missile computer and the MARB computer. This
computer can meet reentry radiation hardness
levels. It has a throughput capability of 5
million instructions per second (mips). It is
compatible with high level languages, including
ADA, for cost effective firmware development
and maintenance. This computer has such a high
capability that the specific design of
operational algorithms become less critical in
the sense that creative design compromises to
minimize throughput and memory storage are not
major design drivers.
Radiation hardened, low power electronics
can be miniaturized by using VHSIC, CMOS/SOS,
hybrid chip modules and innovative packaging
design. X ray shielding designs and radiation
hardened digital parts are available.
Strapdown ring laser gyros are being used in
many applications. Accuracy, low power, high
reliability and radiation hardness are available
features. Designs with smaller size and shape
with higher performance are becoming more
readily available.
Currently available reliability capability can
satisfy the Fractionated SLBM design
constraints and meet reliability requirements.
*Selected reliability features include strapdown,
solid state sensors, solid state digital
electronics, effective part reliability programs,
component burn-in and understanding of part
failure mechanisms with effective design
corrective action.
Advanced Technology
The Fractionated SLBM design requires
advanced technology which includes
*
*
*
*
Conclusion
To reduce SLBM vulnerability to
exoatmospheric encounter by space deployed
weapons a missile configuration has been
defined which significantly reduces reentry body
deployment time. The configuration includes a
fractionated third stage, an integrated missile
avionic configuration and a MARB for subsequent
correction of midcourse determined navigation
errors. Avionics include a strapdown, solid
state SIMU for each MARB. The SIMUs are the
source for navigation data used for guidance and
control of the boost stages, fractionated third
stage and MARB. The required technology is
available. However, the strapdown stellar
sensor, solid state accelerometers, and
electronics require advanced development.
References
1. Dietz, R.C. "Fast Burn Missile Requirements
Analysis" LMSC 8171/157, Ballistic Missile
Future Systems and Technology Workshop,
AFBMO, April 1986
2. Dietz, R.C. "Post Boost Vehicle Design in the
Future" LMSC 8171/157, AIAA Future
Systems Workshop, AFBMO, MARCH 1988
3. Albert, W.C., Weber, R.E. "Vibrating Beam
Accelerometer for Strapdown Applications"
IEEE CH1820, 1982
4. Kancler, H.C. "VLSI Rad Hard Computer" LMSC
8160/1 57
co
cl
-n
0U
zc
ClL
C/))
(fl
00
LL
C'?
0
1
F.-.~
ZU
0
W()
cc z
<
WU
.. ~w 0
<
2r
.1
0(Lool=-
IILL
W
U=C
C0
aW -L
a- -LZUO
LL
20z
zW)
clWW
W''
IinW LU0
L -
Cl)M<
Cl)
xz
LU
3l
W U
C.) C
-J
UJ
0
0
LU
oc
C.
S4
CC1
L-
LU
wc
Cl
L0
U4
22uM
CC
0
LU
1
cc
LLL
I-
I(zZW
< )
C0
00
CC z
z
4:)
Z.Jf
(ml
(U)L
(/)
L LIC
ClOEE
lL
C.)
I-)
z
0
COC'?
zz
M~
C/)
2ZwLL
mn
<Zccw
<
0=01
o<
m~
JW
z
O
F-
0001-
wo<wom
-0<<T
M0CO -0
<
IC,)
z
:z
xo-12
~cnw
C..)
imLL
I-
LAI
0
W
.
JC>
LU
zo
4
U.",
U-A
COlW
LU
0u0
7cLjo
U.COc
cq
w
z
ClCO
C,)l
W
Cf)
44 LL
I-
C).j-
w"ZCuQ
>w
-
W
<
0
zml
-L
<Z-
n Ls
(I)<
>OZLL
z c
-j
mmmlM
>...0M
> -
Zw
-1
-+
ZW
;L
LZ.
0
>.
m~
cy
CI)Ex
1Wgo0m
-0 a
0
w0U
LLwjjj-
m m
a1-mE-
0
F-~
ccZ.-0
c)a.a
U
ol
0AC/)
0
--
0~
0
.))C/
.)
rz
w
Wf
00
LLU
00
z
0
OD
Cl')-
F-0
0>
ww
WD,
0)j0
0
Wl
LLF'
JC/)
z 2w
00
swa
CL')
00ZY
Z
C
W
4
>oO
rL
ic~C/)
ow 0
Mo=
wUgs2
0
M-
F=
wL
>
z
C)
C.)
Cf)
aI-
01
I0.j
Cu)
Wj
20C'
0ow
LLWC
0L
0
Cl) 2
Cl)
LIIgr,
L)
oL
w
a:-2
Oa
WU
a-WUJ im
1-D
0
M~
C/)
wL
Cl)
0L
Cl)
<
0~
I-
0 wI a .
)
-cl
o5
Cl-
= = .
L)
C/_
Cu
LJwFIwL
<
(0__
2o
Co ~LL
Cl)0"
< 0Z .
_0
</ .j<
-li
~z
L
0u
W
w
"
~C/)
WLI~
LU >IJ
0:
This paper presents the concept for sevcral stcllar aidcd Ring La,wr C'\rjo .LC) leitt,.I Measuremlnt Unit IlMU) sy ,WnS suitable
for ballistic missile guidance applications. Top level design tradeoff analysis results for configurations ranging from pure strapdown
to two-axis gimbaled strapdown mechanizations are presented. System mechanization features overcoming the constraints of a strapdown configuration for observability of online determination of key
sensor calibration and alignment parameters are described. Stellar
aiding relaxes the accuracy required of preflight azimuth alignment.
The unparalleled scale factor performance of the RLG offers new
flexibility in star sighting strategy. In addition, the RLG offers cost
and reliability advantages over conventional gyros.
1. Introduction
The key performance parameters for aballistic missile guidance
system are preflight azimuth alignment and thrust acceleration
measurement during the boost flight. These translate into gyro drift
and accelerometer scale factor as the critical IMU performance parametcrs. Current systems require elaborate inertial sensor and platform system designs, and complex preflight flight calibration and
alignment operations are performed to accomplish and maintain the
accurate azimuth alignment and accelerometer scale factor calibration. The stringent CEP accuracy required of the strategic weapon
system has driven up the complexity and cost of booster inertial
guidance systems mechanized using conventional spinning mass
gyro and stabilized platform technologies. Cost escalations also resuit from the reliability and maintainability problems that may arise
through the life time of such systems where continuous active operalion is necessary to maintain the critical thermal environment required for achieving the CEP accuracy.
Stellar aiding can provide the required total system accuracy with
greatly relieved IMU requirements. Even with a conventional technology IMU, this relaxation is translated directly into significant
cost savings. The accuracy of stellar aiding has been verified in
Navy systems and more recently in the flight tests of the GE AINS
(Altemate Inertial Navigation System) as part of the USAF SICBM
program. For these systems the star sensor is placed in the inner clement of a gimbaled platform. The inner element is inertially stabilized during the boost flight with the star sensor LOS pointing in the
direction anticipated for a single optimal star sighting.
The development of RLG systems for strategic weapon systems
application by the Air Force Ballistic Missile Office has been underway for twelve years. The Honeywell Dormant Inertial Navigation
System (DINS) program demonstrated strapdown RLG navigation
for maneuvering reentry vehicles.The Altemate Inertial Navigation
System (AINS) provided RLG performance improvements and nuclear hardening. The second Generation RLG strapdown IMU is
been developed on the current Honeywell Advanced IMU (AIMU)
program,
The development of Ring Laser Gyro systems and the stellar sensor offers new flexibility in the design of a stellar aided IMU with
attending accuracy, cost and reliability benefits. Due to the absence
of moving parts, the RLG strapdown IMU mechanization is inherently lower cost and more reliable than astabilized platform system
using conventional spinning mass gyros. The unparalleled scale factor performance of the RLG provides accurate measurement of angular rotations. The star sensor LOS can be maneuvered to actively
seek out any pair of bright and isolated stars with adequate angular
separation to acquire complete three axes attitude information for
stellar aiding without being constrained by the target position. In addition, a 180-deg roll about the LOS to the star can be performed as
part of the sighting maneuver to enable the inflight calibration of the
star sensorLOS with respect to the IMU sensitive axis.The excellent
gyro scale factor performance ensures the preservation of the star
sighting accuracy throughout the large angle maneuvers.
The pendulus accelerometers are typically of small size, low cost
and high reliability. The bias and scale factor stability is moderate
over the long term. However, over the short term on a continuous
operating basis, they are excellent making them suitable for applications where online calibration is available. As aminimum, they arc
suitable formeasurement of applied acceleration on the missile piich
and yaw axes where the performance sensitivities are low. High performance accelerometers, e.g. SFIR, offering excellent long term
bias and scale factor stability and modelability, are suitable for
measurement of applied acceleration on missile roll axis, the nominal thrust axis. This becomes necessary when online calibration is
not available. Under development arc vibrating beam accelerometers with potentials of high performance, low cost and high reliability. They are candidates for thrust axis application when the
technology matures.
This paper presents the concept for several stellar aided RLG
IMU systems suitable for ballistic missile guidance applications.
Top level design trade-off analysis results for configurations ranging from pure strapdown to two-axis gimbaled strapdown mechanizations are presented. System mechanization features overcoming
the constraints of astrapdown configuration for observability of online determination of key sensor calibration and alignment parameters are described. This assures the high accuracy required for
strategic booster guidance applications.
Variations ofthe aiding mechanizationoptions are presented suiting the capability of the missile vehicle to which the system is applied. This includes providing the environment for online
accelerometer calibration, performing the maneuver for star acquisition, and correcting the missile trajectory for the navigation errors
determined from online calibration and stellar aiding. Modern missile systems, e.g. Minuteman Ill, Peacekeeper and Trident, are
equipped with a post boost propulsion system capable of fully supporting these operations. Older existing systems, e.g. Minuteman II,
do not include apost boost stage. Simplified operations conforming
to the constraints on vehicle capability can be performed. System
accuracy reduced from the full potential but satisfying the original
weapon system requirements can be achieved.
2. System Confiourations
Four system configurations are considered in the concept cvaluation of stellar aided RLG IMUs as depicted in the geometry presented in Figure 2-1. All four configurations assume the Advanced
Inertial Measurement Unit (AIMU) presently in development under
Ballistic Missile Office (BMO) contract. The AIMU program has
the objectiveof developing an RLG strapdown IMU with the Inertial
Sensor Assembly (ISA) ready for strategic system applications. In
increasing complexity, the systems are:
(1) (SD/3A): A pure strapdown mechanization in which the
AIMU and the star sensor are mounted fixed to the vehicle
SS
G1
SSA4(HA)
AAIDING
AA-
GA3
STAR
G3
A3
NOMINAL
MISSILE
AIMU
AXES
EN3X1
G3
HIGH
AACCURACY(H
CE
CH
G1
Al (HA)
AI
GaA
AS
G2
A2
A3
G3
GIMBAL
AXIS
The accuracy of a stellar aided system is derived from supplemeriting the preflight alignment with highly accurate attitude updates available from inflight star sightings. The RLGs with excellent
scale factoraccuracy enable the use oflargc angle maneuvers forstar
acquisition and LOS to IMU calibration while still preserving the
basic accuracy associated with the star sightings. However, the stellar aiding does nothing to improve the scale factor performance of
the thrust axis accelerometer, the other critical performance parameter for the ICBM boost guidance IMU. This is accomplished through
observing accelerometer measurements in the preflight one-g environment and during the post boost zero-g environment. If the post
boost zcro-g environment is not available, the ability to point the IA
up and down in the preflight one-g environment is essential to provide the calibration obscrvability forseparation of thc accelerometer
bias and scale factor errors. System operations during the preflight,
boost and post boost mission phases are depicted in Figure 3-1, and
discussed in the following.
(3) (1X/3A): A strapdown system with a high accuracy accelerometer in the thrust axis and a single 180 deg turn axis. The
turn axis is nominally aligned with the missile pitch axis to
provide the capability to position the IA of the thrust axis accclcrometer up and down during the preflight operation. The
turn axis will be controlled during the boost flight with the
IA of the thrust axis accelerometer aligned to the nominal
missile thrust axis. The re-orientation capability enables the
calibration of the thrust axis accelerometer bias and scale factor
during the preflight operation and reduces vehicle attitude maneuvering required for star sightings. Full attitude maneuvering is still
required for star sensor LOS calibration,
(I)
COAST
POST BOOST
UPDATE
AN#ZERO-G
#STELLAR UPDATE
SP
#180-DEG ROLL
ABOUT STAR LOS
#CORRECTION
OF NAV ERROR
RETYS
BOOST
#NAVIGATION
NA'T REF
/
T
F.
SA &
j
PRELAUN(QH
STAR NEAR
O era T
T (FOR
I-STARTARGE
UPDATE)
TARG Er
#ALIGNMENT
#CALIBRATION
Fig. 3-1. Stellar Aided tMU System Operations
For the single axis system the stellar sighting is accomplished using vehicle roll maneuvers and the system axis of
turn about vehicle pitch. For the gimbaled system, the star
sighting maneuvers are accomplished using the two-axis
gimbal. If sightings to a case fixed simulated star is performed preflight, the 180-degree roll about the star LOS
MANEUVER
TO ACQUIRE
STAR-1
180 DEG
ROLL
ABOUT
LOS TO S1
MANEUVER
TO ACQUIRE
STAR -2
S2
MANEUVER
TO ATTITUDE
FOR TRAJ
CORRECTION
THRUSTING
Fig. 3-2. Star Acquisition Maneuver and
180-Degree Roll about LOS to Star
only one star is sighted for update.
Figure 4-1 compares the navigation performance in terms of normalized CEPof the four system configurations for various mechanization options. It is shown that zero-g update is necessary for the
pendulus accelerometer based strapdown system, i.e. SD/3A. Notice
that the CEPs for the cases of no zero-g update are not plotted due to
the excessive error magnitude. This is because the post boost zero-g
update complements preflight one-g measurement for calibration
observability of the roll axis accelerometer. A high performance accelerometer, e.g. SFIR, with good bias and scale factor stability aiding the strapdown IMU triad, i.e. the SD/4A configuration, allows
medium accuracy performance for applications where zero-g update can not be performed. The single turn axis in the 1X/3A configuration enables the calibration of the bias and scale factor of the
thrust axis accelerometer without the zero-g update. With the degrees of freedom provided by a two-axis gimbal, i.e. the 2X/3A configuration, complete IMU calibration can be accomplished preflight.
Figure 4-1 also shows the benefits of the 180-degree roll about the
LOS to the star. The difference between calibrating the star sensor
LOS to IMU alignment using case fixed simulated star preflight and
using the actual star post boost is minimum in terms of the CEP performance as indicated in comparing options A and B for the twoaxis gimbaled configuration. 'he benefit of two-star update is not
apparent in the Figure 4-1 comparison of options 2 and 3. This is
because a star at target zenith is used in the case of one-star update
offering near optimal performance. The real advantage of the twostar update lies in the lifting of constraints on star location. Any pair
of bright and isolated star with adequate angular separation can be
used to provide complete three axes attitude information. This relaxes the star sensor design requirement in that better signal to noise
ratio can be assured,
Additional performance sensitivities to system parameter variations are presented in Figure 4-3. The performance for 50 % increases in zcro-g noise and gyro angular random walk are
designated as ZRG and ARW respectively. The effect of uncertainty
in initial position, designated as PIC, represents 100 fold increases in
north and east components of launch point uncertainty. The zero-g
update noise determines the accelerometer calibration accuracy
which translates directly to down range target miss. The gyro angular random walk degrades the correlation between the navigation
and attitude errors and thus reduces the effectiveness of the stellar
update which is basically an accurate attitude measurement. The
horizontal components of launch point uncertainty produce preflight
level alignment errors similar to horizontal components of accclerometer bias and IA misalignment. Zcro-g update provides direct observation on accelerometer bias. The preflight level alignment error
is observed by the star sightings leaving the effect of accelerometer
IA misalignment and launch point uncertainty indistinguishable.
This is not a problem for a silo based system since launch point is
well surveyed. Nor is it a problem for a two axes gimbaled system
because accelerometer bias and IA misalignment arc fully calibrated
Benefits of the single turn axis and the 180-degree roll about star
LOS are not fully revealed in Figure 4-1 because of the good (and
realistic) accelerometer bias and scale factor performance, and the
alignment stability considered. Figure 4-2 compares the perfcrmance sensitivities to parameter variations for the four system configurations. Without zcro-g update, the performance of the pure
strapdown system, i.e. SD/4A, will be severely degraded in the pres4
SD14A
1X(2A
ZM1A
4U 3
G
2
3
1
2
MECHANIZATION
OPTIONS
NOTES:
0 PLOTS FOR SD/2A-NO 0-0 NOT SHOWN
OPTION
I
NO OF STAR
I
2
3
1
2
2
B2
-LOS-ROILL
NO
YES
YES
YES (PREFLIGHT)
YES
2.STAR &LOS-ROLL
NOM
HAXIO
LOSXlO
NOMI
HAXIG
PARAMETER VARIATIONS
LOSXIO
NOTES:
'PR: PREFLIGHT ROLL ABOUT LOS TO CASE FIXED SIMULATED STAR
H0OPR: NO PREFLIGHT ROLL
'NOM: NOMINAL ERROR BUDGET
*HAXIO: 10 TIMES BUDGET HIGH ACCURACY ACCEL BIAS. SF AND IA MISALIGNMENT
ILOSXI0: 10 TIMES BUDGET STAR SENSOR LOS MISALIGNMENT TO IMU
NOMI
NOES
wu
ARW
PIC
HAX10
PARAMETERVARIATIONS
LOSXio
zero-g update. The same system aided with sighting to a single star
is analyzed with the ranking of the top fifteen errors presented in
Figure 4-7. The ordering of the error sources is quite similar to that
presented in Figure 4-5.
pre flight. The two-axis gimbalcd system may be required for a mobile missile system because of the IA stability capability of the pendulus accelerometer.
The top fifteen error sources for the SD/4A configuration are presented in Figure 4-4 for the case of a full complement of inflight
update operations. The CEP is used as a scalar figure of merit in the
ranking of the significance of error sources. As expected the top error source is the gyro angular random walk because of its degrading
correlation effects. This is followed by the zero-g update noise and
the velocity noise during preflight C & A because of their influence
on the accelerometer calibration. Almost equally significant are the
aiding accelerometer scale factor and bias errors. The accelerometer
IA misalignments are contributing to the target miss but at lesser signi ficance.
The top fifteen error sources for the 2X/3A configuration are presented in Figure 4-8 for the case of a full complement of inflight
updates. This represents the best performing system. The gyro angular random walk and velocity noise during preflight calibration and
alignment are the top error sources.
5. Technology Status
The key elements of a stellar aided RLG IMU include the RLGs,
the accelerometers and the star sensor. Recent developments in these
key technology areas include the development by BMO of a RLG
based strapdown Advanced IMU (AIMU) for application to strategic weapon systems and the demonstration of a stellar aided stable
platform IMU as part of the Alternate Inertial Navigation System
(AINS) on Air Force's SICBM program. High performance accclcromelres such as SFIR are already in production for existing missile
systems with proven performance and radiation hardness. The developmcnt of a high performance Vibrating Beam Accelerometer
(VBA), also under BMO sponsorship is in progress aiming toward
lower cost, higher reliability and improved size, power and weight
characteristics.
The top fifteen error sources for the I X/3A configuration are presented in Figure 4-6 for the case of a full complement of inflight
updates. The gyro angular random walk again is the top error source.
The ability to calibrate the aiding accelerometer scale factor preflight reduces the significance of zero-g update noise only slightly.
This is so because resolutioi of pitch and yaw axis accelerometer
bias and accelerometer IA misalignment are still dependent on
-2.
U
10
11
12
13
14
1s
16
17
ERRORSOURCES
INFLIGHT UPDATES:
2-STAR STELLAR AIDING, 180-DEG ROLL ABOUT LOS TO STAR, ZERO-G UPDATE
TOP IS ERROR SOURCES:
I GYRO ANGULAR RANDOM WALK
3 PREFLIGHT VELOCITY NOISE
S GYRO IA MISALIGN tG2 ABOUT 1)
7 ACCEL BIAS (AIDING A4-HA)
9 STAR SENSOR NOISE
11 ACCEL IA MISALIGN (Al ABOUT 2)
12 ACCEL IA MISALIGN (A2 ABOUT 1)
1I GYRO IA MISALIGN tGI ABOUT 2)
17 ALL ERROR SOURCES
2 ZERO-G NOISE
4 ACCEL SF (AIDING AA.HA)
6 ACCEL IA MISALIGN (A3 ABOUT 2)
4 GYRO BIAS (GI)
10 ACCEL IA MISALIGN (A2 ABOUT 3)
13 GYRO SF (G3)
:4GYRO BIAS (02)
16 ALL OTHER ERRORS
Fig. 4-4. Top Fifteen Error Sources for SDI4A System Configuration
With Full Complement of Post Boost Aiding Operations
Small Missile program was proven to meet all requirements as demonstratcd in component, sled, and MM III flight tests. The hcart of
the GE/AINS stellar subsystem is a telescope with a Charge Injection Device (CID). The CID is a MOS imaging array of adjacent
photo sensitive sites called pixels that are sensitive to photons in the
the photon-gencrated charge in each site. The CID's behavior in radiation allows the sighting to occur despite nuclear events.
Overall performance of this system was successfully demonstrated in sled tests and two MM III piggyback flight tests. In addition, the CID was proven to meet all specified hostile environments
by component testing.
.2 Stellar Technoloey Demonstrated Apolication. Stellar technology is a mature concept having been reliably deployed in boost
guidance applications on both Navy Trident I and 11programs. The
4
0.
0
1
INFLIGHT UPDATES:
1-STAR STELLAR AIDING,
6
9
10 11 12
ERROR SOURCES
13
14
15
16
17
Fig 4-5. Top Fifteen Error Sources for SDI4A System Configuration
Aiding With Sighting to a Single Star
0
1
10
11
12
13
14
15
16
17
ERROR SOURCES
INFLIGHT UPDATES:
2-STAR STELLAR
ZERO-G UPDATE
Fig. 4-6. Top Fifteen Error Sources for IXI3A System Configuration
With Full Complement of Post Boost Aiding Operations
p7,
02-
10 i1
12
13
14
1S
16
17
ERROR SOURCES
INFLIGHT UPDATES:
I-STAR STELLAR AIDING, NO ROLL ABOUT LOS TO STAR, NO ZERO-G UPDATE
TOP IS ERROR SOURCES:
I ACCEL BIAS (A3)
2 ACCEL
3)
ABOUT 2)
Fig 4-7. Top Fifteen Error Sources for IX13A System Configuration
Aiding With Sighting to a Single Star
2-
0I
1O 11
12
13
14 1
16
17
ERROR SOURCES
INFLIGHT UPDATES:
2-STAR STELLAR AIDING, 160-DEG ROLL ABOUT LOS TO STAR. ZERO-O UPDATE
TOP 1S ERROR SOURCES:
I GYRO ANGULAR RANDOM WALK
3 STAR SENSOR NOISE
S GYRO IA MISALIGN (G1 ABOUT 2)
7 ACCEL SF (AI-HA)
9 GYRO IA MISALIGN (01 ABOUT 3)
11 GYRO IA MISALIGN (G2 ABOUT 3)
13 ACCEL IA MISALIGN (A3 ABOUT 1)
15 GYRO BIAS (G2)
17 ALL ERROR SOURCES
Fig. 4-8. Top Fifteen Error Sources for 2X13A System Configuration
With Full Complement of Post Boost Aiding Operations
6. ConclusIons
Stellar aided RLG IMUs offer high accuracy missile guidance
RLG. Use of single turn axis offers a cost effective solution to pre-
azimuth alignment eliminating the need of cxpcnsive high performance gyros. The use of a semi-dormant system with a stellar sensor
and the laser gyros is the primary driver for lower cost and high reliability. This is possible because of the cost and reliability benefits
directly associated with RLG and the simplification in star .sensor