MATHEMATICAL

AND
COMPUTER
MODELLING
PERGAMON Mathematical and Computer Modelling 35 (2002) 821-834
www.elsevier.com/locate/mcm

Prediction of Swerving M o t i o n of a Dual-Spin

Projectile with Lateral Pulse Jets in
Atmospheric Flight
B. BURCHETT, A. PETERSON AND M. COSTELLO
D e p a r t m e n t of Mechanical Engineering, Oregon S t a t e University
Corvallis, O R 97331, U.S.A.

(Received December 2000; revised and accepted May 2001)

Abstract--Using the linear theory for a dual-spin projectile in atmospheric flight, closed form
expressions are obtained for swerving motion under the action of lateral pulse jets. Trajectory results
generated by the linear theory equations and a fully nonlinear seven degree-of-freedom dual spin
projectile model agree favorably. The analytic solution provides a relatively straightforward and
computationally efficient means of trajectory estimation which is useful within smart weapon flight
control systems. In order to accurately predict the impact point using the analytic solution, the
dual-spin projectile linear model must be updated periodically. Terminal impact point prediction
degrades rapidly as the linear model update interval is increased beyond a critical value. Control
authority, as defined by the change in impact location due to a pulse jet firing, steadily decreases as
a function of projectile down range position. (~) 2002 Elsevier Science Ltd. All rights reserved.

Keywords--Linear projectile theory, Impulse, Swerve.

SYMBOLS

C x FCxA0
0 zero yaw axial force aerodynamic coefficient for the forward and aft bodies
F A yaw angle squared axial force aerodynamic coefficient for the forward and aft bodies
CX2, CX2
F A normal force aerodynamic coefficient for the forward and aft bodies
CN A, CNA
F A magnus force aerodynamic coefficient for the forward and aft bodies
CNpA, CNPA
F A roll moment aerodynamic coefficient due to fin cant for the forward and aft bodies
CDD, CDD
c;p, c5~ roll damping moment aerodynamic coefficient for the forward and aft bodies
cf, Q,c~q pitch rate damping moment aerodynamic coefficient for the forward and aft bodies
Cv viscous damping coefficient for hydrodynamic bearing
D projectile characteristic length
FX , I yFy
IX roll and pitch inertia of the forward body projectile section
.4 A roll and pitch inertia of the aft body projectile section
IXX, I y y
m total projectile mass
rrtA aft body mass
mF forward body mass

0895-7177/02/$ - see front matter (~) 2002 Elsevier Science Ltd. All rights reserved. Typeset by .AA/eS-TEX
PII: S0895-7177(02)00053-5
 822 B. BURCHETT et al.

PF roll axis component of the angular velocity vector of the forward body expressed in the fixed
plane reference frame

PA roll axis component of the angular velocity vector of the aft body expressed in the fixed plane
reference frame

g>r components of the angular velocity vector of both the forward and aft bodies expressed in the
fixed plane reference frame

Tfx,rfyrTfz fixed plane components of vector from composite center of mass to forward body mass center

Tax, ray, raz fixed plane components of vector from composite center of mass to aft body mass center

Rfs 2Q,, Rfz fixed plane components of vector from forward body mass center to forward body center of
pressure
fixed plane components of vector from aft body mass center to aft body center of pressure

fixed plane components of vector from aft body msss center to aft body Magnus center of
pressure
fixed plane components of vector from forward body mass center to forward body Magnus
center of pressure
fixed plane components of vector from composite body mass center to the impulse application
point
position vector components of the composite center of mass expressed in the inertial reference
frame
translation velocity components of the composite center of mass resolved in the fixed plane
reference frame
magnitude of mass center velocity

Euler pitch, and yaw angles

Euler roll angle of the forward body

Euler roll angle of the aft body

pulse jet force impulse components in projectile body axes

1. INTRODUCTION
With clever integration of microelectromechanical sensors and actuators into new projectile con-
figurations, future weapon systems hope to achieve a notable leap in accuracy and lethality.
Because some sensors and control mechanisms require different levels of spin to operate properly,
new multicomponent projectile configurations are being investigated. The dual-spin projectile
configuration is one such concept that consists of forward and aft components, connected through
a bearing, which roll at different rates.
A potential control mechanism for smart weapons trajectory control is a cluster of lateral
pulse jets. Each pulse jet imparts a relatively short duration, yet large lateral body force on the
projectile. By mounting a set of pulse jets on either the forward or aft section of a dual-spin
projectile, a specific trajectory can be tracked. When designing a projectile pulse jet flight control
system to minimize terminal miss distance, pulse jet firing logic relies on real-time estimation
of the effect of firing a particular jet on the miss distance at an arbitrary time and system
state. In theory, this can be accomplished by numerically integrating the nonlinear equations
of motion from a given point to the target. However, from a practical point of view, real-time
numerical integration of the nonlinear equations of motion from a given state to the target at
each flight control system computation cycle represents too large a computational burden on the
on-board microprocessor. Hence, computationally efficient trajectory prediction of smart weapon
projectile configurations is needed for insertion into control laws. Along these lines, Guidos et
al. [l] developed closed form expressions for swerve of a rigid nonrolling projectile with a lateral
pulse jet applied. Their analysis allows for finite duration impulse loads.
Although the dynamics of dual-spin spacecraft have been extensively studied [2-121, work on
the atmospheric flight mechanics of dual-spin projectiles has been comparatively sparse. Topliffe
et al. [13] developed a dynamic model of a dual-spin projectile and employed this model to sim-
 Prediction of Swerving Motion 823

u&e a smart artillery shell. More recently, Costello et al. [14] reported a more general dual-spin
projectile dynamic model which permits unsymmetric mass properties and generalized bearing
friction. This model was subsequently used to develop a linear theory and stability boundaries
for this configuration. The work reported here modifies the dual-spin projectile linear equations
to predict swerving motion and also accounts for the effect of lateral pulse jets. The swerving
dynamics are solved in closed form, resulting in computationally simple algebraic expressions for
the projectile trajectory under the influence of lateral pulse jets.

2. DUAL-SPIN PROJECTILE DYNAMIC MODEL

The dynamic model for a dual-spin projectile admits a total of seven degrees of freedom in-
cluding the components of the position vector of the mass center and the Euler yaw, pitch, aft
body roll, and forward body roll angles. Figures 1 and 2 indicate coordinates used to specify
position and orientation for the dual-spin projectile dynamic model. The linear equations devel-
oped by [14] neglected gravity, since their primary interest was dynamic stability, and also did
not include lateral pulse jet forces in the formulation. Equations (l)-(10) remove these two lim-
itations and represent a modified set of dynamic equations describing the motion of a dual-spin
projectile in atmospheric flight.

(1)
(2)
(3)

(4)

-YI \
-A 0 0 -D m
V' 0 -A D 0 G, + $
’ , (5)
-~RIXZI
IYY
RIZYI
IYY /

0’ = ;q, (6)
l/J’ = ;r, (7)
2’ = D, (8)

y’= ;v+$D, (9)

D
z’=,w-eL3: (10)

where

A=
[2 1
(CNA), (11)

B= [e][g](;) ((R,.+r,Z)
+,,.+ (&,+r,:,) *PA), (12)
824 B. BUFLCHETTet al.

Figure 1. Projectile position coordinate definitions.

,. . . “...

_^x
..,_,._.,,..
x._, I_ .,

Figure 2. Projectile orientation definitions.

C= [$] [g]CMA, (13)

E= [fg] [E] + 04)

D (&PF + I&PA) (15)

F= ,
v ITYY

pSD%,F, DCv (16)

H= --,
4% WFX
DCv (17)
J=
Eg’
DCv
T= (18)
 Prediction of Swerving Motion 825

pSD3C&-- DCv
(19)
L =

4I;, VI&/

(20)
N = PSD~VC& (21)
21& ’

C&IA = ((& + rfs)C;A + (&z + ?&$A), (22)

I& = I& + my& + IyAy + mar&. (23)
The bearing coupling rolling resistance is assumed to be of the hydrodynamic type.

3. ANALYTIC SOLUTION
Pulse jet forces and moments that appear in equation (5) are assumed to occur over a relatively
short duration of time compared to the motion of the projectile, hence, their effect is impulsive
in nature. An impulsive force acting on a rigid body creates a jump in the translational and
rotational velocity of the body at the time of the impulse. The trajectory solution provided
below does not directly include the effect of pulse jet forces and moments. Instead, to simulate
motion of the projectile with pulse jets acting on the system, a piecewise solution is constructed
that consists of free flight and pulse jet discontinuity segments. Expressions for free flight solution
segments are given below while the pulse jet state discontinuity conditions are provided in the
next section.
As seen in equation (l), the total velocity is dynamically uncoupled from the system but
appears in a nonlinear fashion in the remaining state dynamic equations. As is customary in the
linear theory for projectiles, it is assumed that the total velocity changes slowly with respect to
the other states of the system so it can be treated as a constant in the other dynamic equations.
When viewed in this light, the equations of motion become linear and the equations of motion
for angle of attack and roll motion become uncoupled.
The total velocity of the projectile exhibits exponential decay as the round flies down range,
Equation (24) provides the total velocity solution.

v = Q,-PS~Cxo~l(2~)~
(24)

The solution to the roll rate equations is aided by first defining the decay rates as shown in
equation (25),
H+LfdH2-22HL+L2+4KJ
Cl,2 = (25)
2
With the roll rate decay rate given by equation (25), the roll rate solution can be written as

1
+ rzl -f $r22 $rz2
where,
PA = T
01 - ff2 N rll + -P12
Cl >
eals - -$
( >
elrzs -
1
, (27)

rll = TPF, + (L - Q)P&, (28)

r12 = TM + (L - 02)N, (29)
h = TPF, + (al - L)pAor (30)
rz2 = TM + co1 - L)N. (31)
 826 B. BURCHETT etal.

As given in equation (5), the equations of motion for epicyclic pitching and yawing behavior
consists of coupled motion of ZI,w, q, and r. The slow and fast modes of the epicyclic dynamic
equations are given by equations (32) and (33),

XF = ffF + i$‘F, (32)

xs = gs + i$s, (33)
where,

I
(2AF + 2B)
OF =
F(A-E) ’

I
-(A - E)
us =
2 ’

#s = f [F - @zE] , (37)
Corresponding to the modes given in equations (32) and (33), the right and left eigenvector
matrices are given by equation (38) and (39), respectively,

(38)

-ip -1
(-z/+/L) (-YYP) (-YfII) (-u-L)
* * -1 (39)
+*“” CL*) (v* “_ CL*) (V’ - #LL*) (v* ” CL*)
--i$ *
(V* _ II*) (+*-IIp*) (V’ : p*) (zJ* IZp*) -
The constants utilized in equations (38) and (39) are defined below. Note, that the * symbol
denotes complex conjugate.

K=(E-A)+2A+iF, (40)
Q = (E - A)2 + 4AE + 4C - F2 + 2i(F(E - A) + 2(AF + B)), (41)
R= (E-A)+2A-iF, (42)
S = (E - A)2 + 4AE + 4C - F2 - 2i(F(E - A) + 2(AF + B)), (43)

(44)

(45)
The vector x, which represents the initial conditions with the steady-state portion removed, is
used to define the homogeneous response of the epicyclic equations

--A 0 0 -D--l

(46)
 Prediction of Swerving Motion 827

Using the definitions provided by equations (38), (39), and (46), the homogeneous solution for
each epicyclic state variable is expressed as a combination of the fast and slow modes,

while the particular solution is given by equation (51),

--A 0 0 -D’-1
VP 0

1)_ _I1
0 -A D 0
WP Gg
ZZ-
B ECE-F
i5 0 .
(51)
4P
0
I-P
-5
’ 5B F E
Expressions for the modal amplitudes and phase angles used in equations (47)-(50) are provided
in the Appendix. The total epicyclic solution is the summation of the homogeneous and particular
solutions

(52)

The solutions for the Euler pitch and yaw angles are obtained by integrating the solutions for
the pitch rate and yaw rate, respectively. The resulting formulae are

uF8cos(#J,s + e,,) - cos e,F] + z[eup’sin(C$Fsf erF) - Sin erF] %F

>

+ F z[egS’ COS($~S + eTs) - cose,S] + z[eDSS sin(&s + ers) - sin&s] Q-s,

( >

e = e. + $(qps) (54)

COS($bpS f e@) - cOSeqF] i- z [capssin(C$&?s+ &F)

PF
- sin erlF] fiqF

-4s
+g
v (zIe oSscos(@~s + eqs) - cOse,s] + s[eOSs
PS
sin(4ss + eqs) - sineqs] fl,S,
)

where

PF =&+d& (55)

ps = Cl; + 0;. (56)

The trajectory or swerve solution is then obtained by substitutin, 0 equations (47)-(49) into equa-
tions (9) and (10) and subsequently integrating. The final results are given by equations (57)
and (58).
828 B. BURCHETT etal.

x
(
y(e
S
oz3 sin(&s + &S - 7r) - sin(&s - 7r) - q5sCOS(&~- n)s)
1
,

[
PSD
2m 1
(cxo -

vo
CNA)

*” sin(4Fs f &F - n) - sin(&F - 7r) - $F COS(6,F - 7~)s) (58)

+
[ 2m
PSD
1
(Cxo-CNA)
v,

x
( $(e u:3 sin(&s + tL.9 - r) - sin(8,s - 7~)- $s COS(O,~- T)S)
>
.

Equations (57) and (58) provide relatively simple expressions for the altitude and cross range of
a dual-spin projectile as a function of range. Both the cross range and altitude solutions contain
constant, linear, and oscillatory terms. The altitude formula has an additional parabolic term to
account for trajectory bending due to gravity.

4. PULSE JET FORCE AND MOMENT JUMP CONDITIONS

The net effect of a pulse jet force is a discontinuity in the translational and rotational velocity
of the body at the time of the impulse. The discontinuity conditions are found by transforming
equation (5) to the Laplace domain and noting the additional impulse terms adding with the
pre-impulse initial conditions. The results are given by equations (59)-(62).

+=Vg+--, YI
UO m
(59)

w$=w,+--g, ZI (60)
RIJI
go+= 4; - -> (61)
IYY

ro+=Ti+-.
&,YI
(62)
IYY

In equations (52)-(55), the superscripted t symbol denotes after impulse and the super-
scripted - symbol denotes before impulse. Since the epicyclic solution amplitude and phase
angles (~,F,~~SI~,F,~~S,~~F,~~S,~TF,~~TS,~~F~~~S,~~F,~ZUS,~~F~@,S,~,F,~~S) all depend
on ~0, wo, go, and TO through x, the dynamic model must be updated at t+ to accurately account
for the discontinuity in the epicyclic solution formulae.
 Prediction of Swerving Motion 829

Table 1. Projectile physical properties.

Lateral pulse jet stationline location

Lateral pulse jet waterline location

Table 2. Projectile state conditions.

5. RESULTS
In order to investigate the utility of the simple trajectory formulae derived above, the results
that follow compare the trajectory of a dual-spin projectile computed by numerically integrating
the fully nonlinear equations of motion [13] with the linear theory analytical solution. The
numerical solution was generated using a fixed step fourth-order Runge-Kutta algorithm with a
 830 B. BURCHETT et al.

Range (It)

Figure 3. Cross range vs. range.

800

s 700

2
5 600

500

400

300 r,

200
0 1500 2000 2500 3000 3500 4000
Range (tt)

Figure 4. Altitude vs. range.

time step of 0.00001 sec. Tables 1 and 2 list the physical parameters and state conditions for the
projectile considered in trajectory comparison. The linear model is updated for each change in
range of one caliber. In all cases, the numerical nonlinear simulation and the analytic closed form
solution are in good agreement. Figures 3 and 4 show the swerve of the projectile. A difference
of approximately 10 ft between the numerical nonlinear solution and the analytic solution at a
range of 3750 ft is apparent in the altitude trace. When viewed from the rear, the projectile drifts
slightly to the left prior to the pulse jet firing. After the pulse jet is fired, the pulse jet turns and
drifts to the right. Toward the end of the trajectory, the lateral pulse jet has changed the cross
 __
Prediction of Swerving Motion 831

0.25

0.24

0.23 - _-_-_-
-

‘I

-_

_
-i
-

-_
1

, .-_- _-_ _______L________._________~_____

______ _L_

_L_
I I
-
I 1

Non-linear
-m-e Linear

‘I
G

.__
s 0.22 __--_-_- ; __

1
al
T?

-
2 0.21 ______; __ _'
_
0 ?
E
,& 0.2 _______; _ _ _.
I;
0.19 -- _______: -
:_,
.
0.18 _______ c- - _._.
~
i- i
o.170L 500 1000 1500 2000 2500 3000 3500 4 OCIO
Range (fi)

Figure 5. Pitch angle vs. range.

0.08 I I I I I I I

- Non-linear
-*-a Linear
0.06 _-______.________L________._____-_1_-__-______*________~______~

0.04
B
s
g 0.02
2
P
T 0
4
L;
-0.02

-0.04

-0.06 -
0 500 IWO 1500 2000 2500 3000 3500 4
Range (f~)

Figure 6. Yaw angle vs. range.

range impact point by approximately 17ft. Figures 5 and 6 show that the action of the pulse
jet induces Euler angle changes exceeding five degrees. Because of the roll orientation when the
pulse jet was fired, the altitude response does not show an effect from the pulse jet. The fins
mounted on the aft body along with inherent roll damping combine to create the aft body roll
rate history shown in Figure 7. The forward body roll rate history is driven by the viscous friction
in the bearing and the roll rate of the aft body. Figure 8 displays the difference in the predicted
impact point of the numerical nonlinear simulation and the analytic solution as a function of the
linear model update interval. At the range of 3750 ft, the impact error prediction is below 10 ft.
 832 B. BURCHETT
et al.

10 I I I 1 I I I

0 _______:________,________:_________~________;________,________~_______-

-10 _______i________i________t________’_ i _______i________f________i_________

G -20 _______ ~~_-_-_-_._-_-____L________L-____-_-_~~~_._~___*________._______-

B
@.30 .______.________:_.______,________;________~__.___._;________~_______-

*b-
: : .a-
2*
g -40 ._-___-I I________’ _ r___.____:________;___.-____~__=-~~_~______._
s
tic :
I _______

= # ’
g _!jfJ ..,,,~~~‘r.._~__._....i________
~~.~_~~~_~_._~_
.------L----.-..L---.---- ; .--_-___A_ ;

$
m-60

-70
; i
\ ; ; /j)“_
i_____.‘________‘________.____

r,.....;.._._._.,._,~___~________~____.
’ ’

-80 .!b~~~~_i_/_
, yi
_ ..~...._._.~......_.,____
’
i
j
.**-*- AtI Linear
1 1
I I I I I I I
-90
500 lmo 154x 2000 2500 3000 3500 4000
Range (ft)

Figure 7. Roll rate vs. range.

140 I I I I n I I I

1x ________:_______:_______~____._-:_______~_______~_______~_____
_;______
2

0 2Mo 4000 6CQO 6000 10000 12000 14000 16000 IGO0

Model Update Interval (Calibers of Travel)

Figure 8. Target prediction error vs. model update interval.

Updating the linear model used for the analytical solution beyond 4000 calibers of travel results
in significant increase in impact point error. Figure 9 examines the potential change in impact
point location as a function of down range position. After the epicyclic dynamics have settled, the
swerve solution is dominated by the linear term. Firing a pulse jet changes the slope of the linear
term. As the projectile closes on the target, the potential change in position of the round at the
target due to firing a lateral pulse jet decreases, since a reduced distance to the target reduces
the linear swerve changes at the target. Defining control authority as the potential change in
swerve at the target, the control authority of a lateral pulse jet decreases as the projectile travels
down range.
 Prediction of SwervingMotion 833

1500 2m 2500 3om 3500

Rangeto Targetat LateralPulse Jet Firing (tl)

Figure 9. Induced pulse jet swerve vs. range to target.

6. CONCLUSIONS

A simple set of formulae has been developed for the swerve motion of a dual-spin projectile
under the action of lateral pulse jets. The solutions provide a computationally low cost and
reasonably accurate means of predicting impact points of a projectile that can be employed in
future smart weapon flight control systems. For the best impact point prediction performance,
the linear model that underlies the analytic solution must be periodically updated. For the
example configuration examined here, the linear model should be updated no less than once
per 4000 calibers of range. The ability of lateral pulse jets to change the impact point steadily
decreases with range to the target.

APPENDIX

GF = x)~+ (2Im@117T),~
(2h (EPIT),~ xJ2, (AlI
aUs = (2% (C2$>,r x)2 + (2Im (E277F),r xj2, (A%
&F = (b??)*,xl23
4(2& (&77;)*2x)?+ c21rn (A31

%S = J(2Re (t2$)*2 xl2 + (2Im (&G$)*~ ~12~ (A41

&F = (2Re (b& Xl2 + (2Im @117:),~ Xl27 (A51

2b (6713~~~)~ + (21m(t2v2T)*3x)27 (A@

f&F = (2Re (bf),_, Xl2 + (2Im (b& Xl21 (A71

G-S = (2Re(&2112T)*4 Xl23

x)~+ (2Im(&?2T)*4 (Ag)
834 B. BURCHETT et al.

2 Re (<I$‘)*~x
&,F = tan-l C-49)
-2 Im (&VT) *1 X

(
2 Re (<2$‘)*, x
8,~ = tan-l (AW
-2 Im (12772T)
*1X)

_1 2Re(C1$)32x
e wF = tan (All)
( -2 Im (F117T)
*2X1

2Re (&rl?)*, x
e ws = tan-l t-412)
(5277:) *2 X
( -2 Im )

2 Re (&$) *3 x
6@ = tan-’ (~413)
-2Im (EI$),~ X
( 1

2 Re (&‘> *3 x
Oqs = tan-’ (A14)
(
-2 Im (J2172T)*3 X 1

2 Re (CI$‘)*~ x
&.F = tan-’ (A151
-2 Im (EI$) +4 X

2 Re(52772T)
r4 x
8?s = tan-’ (A161
-2 Im (C&J *4 x ’

In equations (Al)-(A16), the notation r*j denotes the jth row of the matrix r.

REFERENCES

1. B. Guidos and G. Cooper, Closed form solution of finned projectile motion subjected to a simple inflight
lateral impulse, 38fh AIAA Aerospace Sciences Meeting and Exhibit, Rena, NV (2000).
2. P.W. Likins, Attitude stability criteria for dual-spin spacecraft, Journal of Spacecraft and Rockets 4 (12),
1638-1643 (1967).
3. G.J. Cloutier, Stable rotation states of dual-spin spacecraft, Journal of Spacecraft and Rockets 5 (4), 490-492
(1968).
4. D.L. Mingori, Effects of energy dissipation on the attitude stability of dual spin satellites, AIAA Journal 7
(lo), 20-27 (1969).
5. B.T. Fang, Energy considerations for attitude stability of dual-spin spacecraft, Journal of Spacecraft and
Rockets 5 (5), 1241-1243 (1968).
6. C.D. Hall and R.H. Rand, Spinup dynamics of axial dual-spin spacecraft, Journal of Guidance, Control,
and Dynamics 17 (l), 39-37 (1994).
7. A.C. Or, Resonances in the design dynamics of dual-spin spacecraft, Journal of Guidance, Control, and
Dynamics 14 (2), 321-329 (1991).
8. J.E. Cochran, P.H. Shu and S.D. Rew, Attitude motion of asymmetric dual-spin spacecraft, Journal of
Guidance, Control, and Dynamics 5 (l), 37-42 (1982).
9. K. Tsuchiya, Attitude behavior of a dual-spin spacecraft composed of asymmetric bodies, Journal of Guid-
ance, Control, and Dynamics 2 (4), 328-333 (1979).
10. H.X. Yang, Method for stability analysis of asymmetric dual-spin spacecraft, Jownal of Guidance, Control,
and Dynamics 12 (l), 123-125 (1989).
11. Z. Viderman, F.P.J. Rimrott and W.L. Cleghorn, Stability of an asymmetric dual-spin spacecraft with
flexible platform, Journal of Guidance, Control, and Dynamics 14 (4), 751-760 (1991).
12. M.C. Stabb and A.L. Schlach, Pointing accuracy of a dual-spin satellite due to torsional appendage vibra-
tions, Journal of Guidance, Control, and Dynamics 16 (4), 630-635 (1991).
13. J.A. Smith, K.A. Smith and R. Topliffe, Feasibility study for application of modular guidance and control
units to existing ICM projectiles, Final Technical Report, Contractor Report ARLCD-CR-79001, U.S. Army
Armanent Research and Development Command, Picatinny Arsenal, Dover, NJ (1978).
14. M. Costello and A. Peterson, Linear theory of a dual-spin projectile in atmospheric flight, Journal of
Guidance, Control, and Dynamics 23 (5) (2000).