This is a reproduction of a library book that was digitized

by Google as part of an ongoing effort to preserve the

information in books and make it universally accessible.

https://books.google.com
 SC - M -67-2933
February 1968
AEC
AERO - THERMODYNAMICS

A DETAILED DEVELOPMENT OF THE

TRICYCLIC THEORY

Harold R. Vaughn , 9321

SANDIA LABORATORIES पर
OPERATED FOR THE UNITED STATES ATOMIC ENERGY COMMISSION BY SANDIA CORPORATION | ALBUQUERQUE, NEW MEXICO: LIVERMORE, CALIFORNIA
UNIVERSITY OF MICHIGAN

3 9015 09522 4807

 Issued by Sandia Corporation ,
a prime contractor to the
United States Atomic Energy Commission
LEGAL NOTICE
This report was prepared as an account of Government sponsored work.
Neither the United States , nor the Commission, nor any person acting on behalf
of the Commission :

A. Makes any warranty or representation, expressed or implied, with re

spect to the accuracy, completeness , or usefulness of the information contained
in this report, or that the use of any information , apparatus , method, or process
disclosed in this report may not infringe privately owned rights ; or
B. Assumes any liabilities with respect to the use of, or for damages re
sulting from the use of any information, apparatus, method, or process disclosed
in this report ,

As used in the above, "person acting on behalf of the Commission includes

any employee or contractor of the Commission, or employee of such contractor,
to the extent that such employee or contractor of the Commission , or employee of
such contractor prepares , disseminates , or provides access to any information
pursuant to his employment or contract with the Commission, or his employment
with such contractor .
 SC - M - 67-2933

A DETAILED DEVELOPMENT OF THE

TRICYCLIC THEORY

Harold R. Vaughn, 9321

Sandia Laboratory, Albuquerque

February 1968

ABSTRACT

A detailed , step - by - step development of the Tricyclic Flight Dynamics Theory originally
derived by Dr. John Nicolaides is presented . Examples of application of the theory are included .

1-2
 CONTENTS

Page
5
SUMMARY

6
SYMBOLS

Introduction 9

9
Aeroballistic Equations of Motion
19
Tricyclic Solution to Aeroballistic Equations
36
Application
36
Simplified Equations
37
Example Problem
39
Tricyclic Case
40
Epicyclic Case

Discussion 41

Constant Coefficient Restriction 41

44
Small Angle Restriction
44
Physical Interpretation
Characteristic Motion 49
51
Magnus Moment
53
Conclusions

55
REFERENCES
APPENDIX A Sample Calculations 57

ILLUSTRATIONS

Figure
1. Inertial Axis System 11

2. Axis System 14

3. Aeroballistic System 18

4. Complex Plane 22

5. Comparison of Tricyclic Theory with 6 DOF Calculation 42

43
6. Comparison of Epicyclic Theory with 6 DOF Calculation
7. Effect of Large Angle on ā and ß 45

8. Physical Interpretation of Tricyclic Theory 46

9. ã - Ř Plot
- 48

10 . Characteristic Motion with Zero Roll Rate 50

11 . Characteristic Motion of Rolling Body 52

3-4
 SUMMARY

1,2
The Tricyclic Theory was first derived in 1953 by Dr. John D. Nicolaides , and has been
3
republished in different forms and extended by others . The Tricyclic Theory is a powerful
flight dynamics tool ; however , it is by necessity complicated and difficult to understand in detail .
The use of complex variables in the theory makes it particularly difficult to interpret on a physi
cal basis . In attempting to apply the theory the author has rederived it for his own use and added
some detail that is normally not presented in a formal document .. This approach may help other
ballisticians and flight dynamicists reach a more complete understanding of the theory .

5
 SYMBOLS

A Imaginary ( ā) component of K , rad

B Real (B) component of K , rad
CnB Yaw moment coefficient due to ß, per rad
Cma Pitch moment coefficient due to a , per rad
ac
n
С - Yaw moment coefficient due to rate of change of B , per rad
Cns a (ва
2V

cmo Pitch moment coefficient due to rate of change of a,

ac
m per rad

с
nr
-
- Yaw moment coefficient due to r ,
ac
n
rd per rad
ca
a
låd
2V

a
2V
ac
m
Cma Pitch moment coefficient due to q ,
qd per rad
a
2V
ac
n 2
с Magnus moment coefficient due to a , per ( rad )
ela

npa pd
laal
2V
ac
m 2
с
Cmp3 Magnus moment coefficient due to B,
pd
per (rad) 2
әр 2V

d Body diameter , feet

H Angular momentum
i, j , k Unit vectors

Ι
Ix - Roll moment of inertia about body x axis (also I ), slug-ft? XX
2
I
у
Moment of inertia about body y axis ( also I Tyy ), slug -ft?
2
Ι
I
IZ - Moment of inertia about body z axis (also I ), slug -ft? ZZ
2
I Lateral moment of inertia when I
y
= I
IZ slug -ft?
I
ху ?' Exz'” 'yx
Ixy. 2
I I Products of inertia , slug-ft?
yz ' Izx ' zy

Nutation arm , rad

K,
Precession arm , rad
K2
B? - Trim arm , rad

L - Roll moment, ft - lb

M - Pitch moment, ft - lb

N Yaw moment , ft - lb

6
 SYMBOLS ( cont)

M M M -

Total moments about x , y, and z axes, ft - lb

x ' y' z

- Moment due to control deflection , ft - lb / rad

M.
ft - lb
M Magnus moment due to B , (Equation 46) ,
pp 2
radº / sec

ft - lb
N
pa
- Magnus moment due to a (Equation 47 ) , 2
rad / sec

ft - lb
M α - Pitch moment due to a (Equation 48 ) ,
rad

ft - lb
NB - Yaw moment due to ß ( Equation 49 ) , rad

ft - lb
M Pitch moment due to q (Equation 50 ) , rad / sec
q

ft -lb
N Yaw moment due to r ( Equation 51 ) , rad / sec
r

ft - lb
M. Pitch moment due to 3 ( Equation 52 ) , rad / sec
α

ft - lb
N: Yaw moment due to B (Equation 53 ), rad / sec
Ni

N
N1 , N2 N3 Constants in differential equation

m1 , 2 - Two roots of differential equation

-

р Roll angular velocity, rad / sec

a Pitch angular velocity , rad/ sec

r
Yaw angular velocity, rad / sec

q' - Dynamic pressure , psf

164

- Vector distance from origin of axis to elemental mass , ft

2
s - Body cross sectional area,, ft?
S - Gyroscopic stability factor

t - Time, sec

- Vector velocity of elementary particle m, ft/ sec

V Total free stream velocity, ft / sec

7
 SYMBOLS (cont )

X - Body roll axis

X Aeroballistic body roll axis
у - Body pitch axis
у -

Aeroballistic body pitch axis

- Body yaw axis
-
N

Aeroballistic body yaw axis

NI

X, Y, Z - Inertial reference axis system

GREEK SYMBOLS

- Body fixed axis angle of attack

छा

α
छ
।
२२।

- Aeroballistic angle of attack

- Aeroballistic angle of attack in 6 DOF axis system
- Body fixed axis angle of side slip
B Aeroballistic angle of side slip
- Aeroballistic angle of sideslip in 6 DOF axis system
6 - Control deflection angle
-

11,2 - Damping exponent; real part of mm1,2

-

६ - Complex total angle of attack

T - Defined by Equation 103
-

- Angle of oscillatory plane in the ā -

-

plane
V -

- Angle of the K arm in complex plane with respect to ā axis

W1,2 - Nutation and precession frequencies; imaginary parts of mı1 ,. 2

‫ܝܐܠ‬ - Total angular velocity

1
 A DETAILED DEVELOPMENT OF THE TRICYCLIC THEORY

Introduction

The Tricyclic Theory is a solution to the aeroballistic equations of motion which makes use
of complex variables . It is the only theory that describes the free - flight motion of a rolling
vehicle in a relatively complete manner . It has the unique ability to predict the motion of a vehicle
quantitatively as planar , elliptical, epicyclical or tricyclical, depending on the physical and aero
dynamic characteristics of the vehicle . Consequently, it is important that flight dynamicists be
able to apply the theory in predicting and analyzing the motion of free flight vehicles. It is
equally important that flight dynamicists realize the restrictions inherent in the theory, most of
which stem from the fact that the theory is based on a solution to a linear differential equation,
which inherently has the assumptions of constant coefficients and small angles . However,
properly applied , the theory is a very useful tool in analyzing flight dynamics problems .

The first step in deriving the Tricyclic Theory is to derive the aeroballistic equations of
motion .

Aeroballistic Equations of Motion

The Tricyclic Theory is based on the aeroballistic axis system and the resulting equations
of motion. Since there may be some confusion involved in the differences between some of the
various axis systems ( i.e. , body fixed , aeroballistic and precessing aeroballistic ) the aeroballis
tic set of moment equations is derived and discussed in detail .

The equations of motion are based on Newton's statement that the time rate of change of
linear momentum must equal the sum of applied forces
8

Σι- Σα((mmgv) .
i= 1 i= 1
d
dt
1

(1)

where m is the elemental mass and F is the elemental force . The moment can be obtained by
simply multiplying by the distance r to the elemental mass and summing elemental moments

d
M =
Σ exiaΣα((mr xv) ,
Eixa dt
(2)

9
 but the vector velocity ✓ of the elementary particle m is also the angular velocity times ř .
Consequently

（
✓= ū xr . ( 3)

Substituting

Ed
M -Σ ε[[mPxfox7)]
= mr ( 4)

which from vector analysis can be written

d ( 5)
Σ [- 67
M -m
M =
- 717 .-os]
( • FH) TF dt
m

)
6
(
Ñ=
* * {2m-2)=
at
(6. y.
Σ -Empf.o mr mr

At this point it is tempting to take a giant step forward by simply noting that mrº is the moment
of inertia and that mrſr.W ) is equivalent to the product of inertia times the angular velocity.
However , the derivation will be done in detail . Similar developments are found in References 4 ,
5 , 66 and many others . To accomplish this we will resort to the Cartesian coordinate system
shown in Figure 1. Since r is a vector it can be written as

(
7
)
r = xi + yj + zk (7

and

2 2 2 2
r = X + y + Z ( 8)

where i, j, and k are unit vectors and the angular velocity vector w , which is not necessarily
coincident with r, is
(
)
9

‫ز به = د‬
w Xi + WI + w
w
Z
k

If we substitute ( 7 ) , ( 8 ) and ( 9 ) into ( 6 ) we have

d 2
M =

dt
ΣE m (x + y2 + 2?/2(w)‫په‬i+‫ ت‬w,7+
+
j )
у
w we
+ z
X у°
k) ( 10 )

m (xi + yj+ zk)(xwx ++ ywy + zwy

m
20,1}
W
10
 X

m
m

x
स
x

N
.
oy
y
N

Figure 1. Inertial Axis System

11
 and

2 2

Ñ - *{m [{(x2 + y²? + 2 ?)

M =
dt
Σπω, х
x у +
W ? + wу,xy + w_xz)
- ( x?
x ] Z
]?
i

2 2,

+ , + y2 + 2%)- (w xy+ x2 + w_vz)]]5

Σ m[[wy(z? у
+ у
X
tw ,
y
+
( 11 )

m [u_(x2+ y2 + 2?) . (w.*

Em [.
+ x (W xz + w
X wyz + W Z
Z =?]5}:
The scalar moments about the three inertial axes X , Y , and Z , which are fixed in space are
obtained by taking the i, j, k components , which are
>

d 2
M
x
[ ΣEm[ s +
dt X
z w у xy - wzXZ
osta] ( 12 )

d 2 2

у m[uycx +++ z - 28-09]

M, - * Σω dt
wexy 3 у
( 2 ) ху ( 13 )

d 2

x
M
z
=
dt
Em[ux[urs( x*+ y x W
х
XZ

]]
“уу? ( 14 )

2 2
) Em(x²
Now it can be seen that the terms involving Im(y² ++ 2z3), m (x² + y2) are
(x2 + 2%),) Σm(
moments of inertia I about the X , the Y , and the Z axis respectively and that the remaining terms
involve product of inertia . For instance mxy is the product of inertia about the Z axis and is
denoted as Iхуv Therefore

M
X
=

de(Ixx«x -- xyy - Exz

1xz®z
WZ? ( 15 )

d
M I
dt ( I уу
M, - & lygwy - 1xy "x - Wyzwy!?
у у ху х ᎩᏃ ( 16 )

z Izzu« zz
M - di (zz
M
dt
I
Exz^ I
x - fyzwy) .
( 17 )

This set of equations , which are the moments about the X , Y , and Z axes when they are fixed in
space , are often written as simply

d
N= dt
(I ) , ( 18 )

where I includes all the moments of inertia and products of inertia and w includes all the angular

velocities of the body (w,

X Wy, w.). Also Iw is the angular momentum #; consequently
12
 155
d
dt
= M
M = (15)) , dt
( 19 )

where

À= 1 .
( 20 )

Equation 19 is important since it says that the applied moments are equal to the time rate of
change of moment of momentum , which is the angular momentum . Unfortunately, Equation 19
refers to an axis system fixed in space where an axis system free to move in a specified manner
with respect to an axis system fixed in space is needed to describe the motion of a body. This
extension was done by Euler . Referring to Figure 2 we have a moving axis system x, y, z
referred to an inertially fixed axis system X Y Z, and û and û denote the total moment of momen
tum vector and ū the total angular velocity vector of the body . Therefore
w

( 21 )
= ‫ تيہد‬+ ‫ توه‬+ ‫تونس‬ у
H

H
= H 1+ H J++ h_k
х , yº z ( 22 )

where i j k are unit vectors, which are regarded as position vectors indicating the position of the
xyz axes . Then from Equation 19 the moment for the moving axis system

dH dH dH
х
y + Z
+ H
dì + H
dſ dk
( 23 )
M =
dt +j dt dt X dt y dt
+ H
z dt

since i, j and k change in the moving axis system . From vector analysis

j k
i jk
dí -üxía
w x1 wy) - w
-

( 24)
11

ху у

1 0 0

i j kk
- ô xy = X
- ‫ په‬+ ‫طه‬ ( 25 )
II

dt “ z
x y
0 1 0

i ិj ķ
dk
= xk =
w =
1

‫ه‬y‫ ت‬- ‫ نهد‬.

dt

. x wy wz
0 0 1
( 26 )

Substituting these derivatives in Equation 23 yields

13
 X

13

N
i
*1
0.

N
wy

Figure 2. Axis System

14
 dH
dH X

=
dt dt
- H,, + H_w)i
H
yz
+
y
i

dH
+ у + Hw H ( 27 )
dt
+H - ) X Z z x

dH

1k
+ z
- HW
dt Hw
хуy + Hy
ух

Since

ан ( 28 )
dt
Ñ = M X i + MJ + M
=
MK

the moments about the x, y and z axes are

dH
X
M + H W -
Hw ( 29 )
X dt z y y z

dH
у
M + Hw - Hw ( 30 )
у dt X Z Z X

dH
Z
M + HW - H
z dt ух ху , ( 31 )

where H X , H , and H are obtained from ( 15 ) , ( 16 ) , and ( 17 )

y' Z

( 32 )
Hx -= IXX
xxxxyy-
X ху у Exzz
H
у
w
Hy - Iİyyy
=
Iхуxyхox - Ifyzuz
уу у - Xyºx
-

( 33 )

H_ =- Izzwz - IW I -

İxzwx - Syzwy · ( 34 )

15
 Equations 29 through 34 are the complete equations for a body in a moving axis system . However ,
they are in an inconvenient form which is difficult to use . A better form can be obtained by noting
from (23 ), (24) , ( 25) and ( 26 ) that the total moment vector can be expressed as

195
M = + xÑ ,
X > ( 35 )

where is the total angular velocity of the moving axes x , y, z . And recalling ( 19 ) and ( 20 ) ,

d ( 36 )
À
M =
dt
(IW) + 2 x Iw .

Completing the differentiation

Iü + iW + à x Iū ,
Ñ = Iū ( 37 )

which is the complete general equation. Expressed in matrix form this equation becomes

M I - I - I ‫ن‬
x XX
İxx - İxy - İZ
ху XZ X XX ху XZ X

M - I i + ( 38 )
y -Тух yy
-Iyx уу - fyzy
wy İyx ivy - Z I
ух yy yz Фу

M
z
- Ixx - I1 , 1zz | |-ix -iy i
Ι
Zx zy
I
Z yx zy Zz z!

I - I - I
ху XZ x
3

X XX
By

+32 I I · I
Фу
Х

ух yy yz

- I
-

- I I
ON

Zx zy Zz
3
N

Now the symbols are replaced with symbols normally used in aerodynamics

L I - I i
Ixx -xyху XZ
р
İxx - Ixy
XX - IZ
ху XZ
р

I - I ( 39 )
M
-Iyx 4 ++ - iyx fyy - 12
N I
- 12x -Izy
р
ZX
yy yz

I
I
ZZ z
- I
RE
- I
-
ух

- izx - 1 y 22
ZX

р
уу

zy
yz

ZZ
r

XX ху XZ

Х I
+ q
- Ilyx
ух уу Tyy - lyz q

r I r
Izx - Izy 22 ZZ

16
where L, M , and N are roll, pitch, and yaw moments , and p , q, and r are roll, pitch, and yaw
angular velocities ( see Figure 3 ) .

Equation 39 is the complete set of moment equations for three degrees of angular freedom
( 4 , e, and o) in matrix form ; when the matrix algebra is completed , this equation can provide the
three moment equations for a body fixed axis system . Equation 39 can be used to derive the
moment equations for any type of axis system depending on the angular rates substituted for p, q ,
and r in the matrix marked with an asterisk since this matrix controls the rate at which the axis
system moves . For the aeroballistic system , p is made zero in this matrix since it is a non
rolling axis system . Furthermore , the tricyclic theory assumes that products of inertia and
rates of change of inertia are zero . This results in the following version of ( 39)

L I
XX о о р I
XX о о р

EEEE M

which is manipulated in this manner :

от
yy

ΟΙ
zz
0

r
+ q

r
X 0 I
yy

ΟΙ
0

ZZ
q

r
( 40 )

L Ixx'p 0 0 0 I
XX
p 0 0

M 0 Iyyģ 0 + Х 0 I 9 0 ( 41 )

500 N 0 0 Ir
zz

i
q

k
0
yy

0 I
ZZ
r
H.

L I р j
XX

BE M I
yy
g + a
q r
N

I r
ZZ IwxPlyyy, Iz
XX ZZ
r

I p + qri
LL = Ixx®
XX
· qr
zz yy

( 42 )
M = I yyq + rp XX

N = I1ZZ,r - qpIXX -

and substituting by virtue of symmetry

I = I = I
уу ZZ

T = I
XX X

17
 X
X

1. B
18

IN
18

N
8

ia
ia
이

18
9

Y
있
Figure 3. Aeroballistic System

18
The equations for the sum of inertial moments about the x , y, and z axes result :

L == IP
Ip ( 43 )
X

M = Iq + rpixX
M ( 44)

N = Ir qpI ( 45 )
X

which are the aeroballistic equations . It should be emphasized that, while the body is rolling at the
rate of p , the axis system is not rolling.

Tricyclic Solution to Aeroballistic Equations

To obtain an analytic solution to the three aeroballistic equations it is necessary to decouple

the roll equation from the pitch and yaw equations, which is done by assuming the roll rate p to be
constant. The effects of this assumption, as well as others made during the course of this theo
retical development, on the accuracy of application of the theory is discussed in the section entitled
Application . This assumption then leaves Equation 44 and 45 to be solved. The first step is to
define the applied aerodynamic moments M and N and equate them to the inertial moments on the
right side of the equations. Since this problem is complicated enough in its simplest form we will
consider only those moments due to q, r , å , ß , a, and ß, plus Magnus moment resulting from
combined roll rate and angle of attack ( which is only one of the several complex moments that can
be considered) .

2
M
Sa ? ( 46 )
q
pВ Cmpen
împp9' 2v
Magnus moments
2
Sd
N = C ( 47 )
pa 'npaa' 2v

M
α
=
С q'sd ( 48 )
ma

Static stability moments

N
NBВ Cmpq's
С ( 49 )

mq'sa ?
Cmg
M
IT

2V
( 50 )
q

Damping moments
2
Cog'sa ?
nr
-

N ( 51 )
r 2V

19
 Choq's
2
ma
M. = ( 52 )
α 2V

Aerodynamic lag moments

Cnhasq'sd
2

( 53 )
II

NB 2V

In addition to these moments, which are normally associated with a symmetrical vehicle, a
moment due to control deflection or an aerodynamic asymmetry must be added . Since a moment
associated with an aerodynamic asymmetry is fixed in the rolling body axis , we must break it up
into components around the ý and z aeroballistic axes. This can be accomplished by realizing
that the roll orientation of the body is simply a function of pt since p is treated as a constant .
Consequently, we can represent the two components of the moments due to asymmetries (or con
trol deflection ) as

and M sin pt
Mg 6 cos pt Mg 8& .
O

When all of these applied moments are substituted in Equations 44 and 45 , the complete pair of
equations to be solved is

M , ā- M q 9 - Ma ä - Mg
Iq + rpi - MaX
M.
6 cos pt - M PPRB PpßB = 0 ( 54 )

qpx - Ng B - N, r - NN B
Ir - qpi -

- Mg ô sin pt - N ,pa pā = 0 , ( 55)

NBX r

where Equation 54 is the sum of the moments about the y axis and Equation 55 is the sum of the
moments about the Z axis . The aeroballistic axis system is shown in Figure 3 , where the various
angles and their time derivatives are pictorially demonstrated. Equations 54 and 55 are scalar
equations , and the quantities in Figure 3 are scalars ( although complex ) even though their direc
tion is demonstrated by arrows . The complex angle of attack & is defined as a complex quantity
composed of the two angles ā and

& = iā + B , ( 56 )

where & is identical to ä used in Nicolaides' original treatment. The symbol is changed in an
effort to avoid confusion . The i simply denotes that ā is perpendicular to B. It should be noted
that ā and are positive when measured from the body x axis to the flight path , which is opposite
in sense to the usual aeroballistic angle of attack ã and ß used later in the report and obtained
from a 6 DOF calculation ( see Figure 8) . Notice that the position of the body x axis is defined by

20
the two angles ā and B and that angular rates of the y and ž axes is defined by q and r . The axis
system does not rotate about the x axis , but the body does roll about the x axis with the constant
rate p . The angular rate q is represented by an arrow perpendicular to the ā plane , and r is
represented by an arrow perpendicular to B. From Figure 3 it is possible to see that

ä
a qģ = ä
q = à,
( 57 )
r = -B, r = -3 .

All of the angular velocities in the figure are vectors in the sense that they can be represented by
arrows with a direction perpendicular to the plane of the angle . It should also be noted that
sq in Nicolaides' terminology) can be expressed in complex fashion as

12
S2 = q + ir . ( 58 )

It is now becoming clear that the complex plane has been chosen perpendicular to the x axis and
in fact is coincident with ā, , š , å, 3, į , q, r , s, and the y and ž axes . This is a tremendous
advantage in reducing the equations to a solvable form . A vertical view of this complex plane is
shown in Figure 4. Now we can easily see the rest of the relations :

2 = -ip ( 59 )

६ = iā + B ( 60 )

į=i + ( 61 )

12Ω = - i ( ià + B) = å - iß = - iſ .
=
( 62 )

With these relations we are ready to manipulate Equations 54 and 55 so that they can be solved
simultaneously . By multiplying one of the moment equations by i the problem is shifted into the
complex plane . Then the two equations are added to get the total moment equation in the complex
plane which can be solved in terms of a single complex angle ś. Multiplying ( 55) by i

α q
M.
Iq + rpix - Mgā - MM 9 - MMjå - Mg o cos pt -- MM PB PB == 0
a ( 63 )

ili - iqp1,X - iNg B - in, r - iN ; B - img 8 sin pt - in ( 64 )

-

r pa
pā = 0 ,

and adding the two equations and collecting terms yields

21
 ir

12

IN
r
o•

it

igliaå

Figure 4. Complex Plane

22
 Ilý + ir) - I pſiq - r ) - Maā - inB B α
-

M q - inr r
-

α
My 8o (cos pt + i sin pt)- MPB PB -
- Mää - iN ; B - Mg ß M M -

in
pa
=
pā = 0 . ( 65 )

Now if the vehicle is rotationally symmetrical , the static and dynamic stability derivatives about
the y axis are equal to the stability derivatives about the Z axis . Consequently

=
= N , M = -N : M N ( 66 )
Ma = -Ne, M a r ' a - NBg pВ pa

Using the normal symbols for moments about the axis (i.e. , MQ, M q ', Ma) , Equation 65 becomes
α ' Mo

Ilſ + ir) - I pliq r) - M α la - iß) - M ( q + ir )

q

- Mală - iß) - Mg 8 (cos pt + i sin pt) M plia + )

B) = 0 . ( 67 )
PB

Since the equation must be in terms of $ , the total angle of attack, Equations 56 through 62 can
>

be used to substitute for ā, B, å , ß, q, r , q, and r in Equation 67 as follows:

>
Š

d
ġ + ir ‫ܪ‬ =

( - ić) = -iſ (from ( 58 ) and ( 59 ) ]

dt

iq - r = ia + ß =3 (from ( 57 ) and ( 61 )]

2
=
2

a- ( ā- iß) (ia - i_B) = -i(ia +B ) = -ig

a- 18 2 2 (from ( 60 ) ]

q + ir = NΩ = -ig
-iξ (from ( 59 ) and ( 59 ) ]

ä- iß = -ig ( from (62)] .

Equation 67 becomes

- - =
M
I( - ië) - IPCĖ) - Mal- ig) - M q (-ig) - Mal- iš) - MMg, 8 (cos pt + i sin pt) pg pĚ = 0 . ( 69 )

Multiplying by

I M M M iM M
X a a 8 pß 10 ..
3-ipy Mouse
Š- 8( i sin pt + cos pt ) - ip . =
Š = 0 ( 70 )
I

I I I I

23
 Collecting terms

I M + M. M
х q a pß a

3 +(-ip_
+
I Mä):+(-ipMuse Mous
કે I I

( 71 )
M
8
I 8 ( - sin pt + i cos pt) = 0 .

Now

ipt -ipt ipt -ipt

sin pt + i cos pt = -
-1- ( 2
e

1 ) - (Me )
+ i
te
2

ipt ie -ipt + ie
ipt + ie -ipt
ie = ie ipt
=

Consequently Equation 71 becomes

I M + M M M iM
a pß 8
X g a ipt
8 + ((-ip
+
I I a) & + ( -ip I I
E
I
8 е ( 72 )

We can replace the coefficients with

N2 and N3
,,, N,
NZ N ' which are constants :

I M + M
α
N1 (-ip
X

I re)
q
I
( 73 )

M M
pß
N2 + (-ip 8 My a ) I 1
( 74 )

M
8
= i ( 75 )
Ng3 I

This yields

Z + N , E + N. & - N,3 eeipt ( 76 )

N2

which is a linear , second order differential equation with constant coefficients which can be solved.
The homogenous part of the equation is

N ++ N.
į+ , N2} = 0 . ( 77 )

24
The auxiliary equation of ( 77 ) is

2 =
m = 0 .
+ N,m
1 ++ N2

The two roots of the equation are obtained from the following algebraic manipulation :

2
m
+ Ni, m -N2

2 2
N N

* + N,*+(m * ) ( )
m
2
N.m +
2
-

2
1
-

N2

2 2
N N.

(m = ) (61)
m +
2
1
2
1
N2

2
N N
1 =
m +
2 2 N2

2
이지

N
1
m =
N2
+1

but since m is actually two separate quantities , m is replaced by m11,, 22 :

2
N N
1 1
m1,2 2 2 N2 ( 78 )

where mmi involves the positive radical, m, involves the negative radical, and they are both
complex values. The symbol m is identical to Qin Nicolaides ' notation. It was changed to avoid
confusion with the symbol for roll angle . The general solution to the homogeneous part of the
equation is known to be

m.t
mat
e ( 79 )
SHH K, e = K
+ K2 ee

To complete the solution the particular integral is evaluated :

mat ( 1, m )t
ei pt
$ .cm'y - e
c[ cm -m3") . *,
e
* S e
3
]at a
dt dt ( 80 )

25
 Integrating

m. t
m2 (m , -m2 )t N3 (-m ++ ip )t
5. " YSpa " (49 ) ----]"
&
р
= e e
m
1
+ ip
e dt

mat
e
N3 ( -m, + ipt
& T -my +
*s ip )
e dt

mot
(-m, + ipt
६ Nge
3 e

р (-m 1, + ip ) - m2 + ip

N e
ipt
3
( 81 )
р
(ip - m . )(ip - mm2,

pt
= RK ei ( 82 )
3

where

N.
3
( 83 )
11

K3 (ip - m .)(ip - m2 )

The complete solution is simply the sum of ( 79 ) and ( 82 ) , which is

Keipt
m. t m.t
1
& = Ке
Ke Ke
+ Ке + K >
( 84 )

where K , and K2
K are arbitrary constants that must be determined from the boundary conditions
(KZ3 is defined by (83 )) . The boundary conditions on this equation are simply that $ = $o and that
$ = 6 , at t = 0. Substituting into Equation 79 and its first derivative yields

5o == K. + K. + K. ( 85 )
Šo 1 3 K2 Kz

and

mt
mt 2 pt
=K m
, e
1
m
+ K m2e m e K , ipei
+ K3 ( 86 )

26
 ( 87 )
Šo = K, m, + K,m2 ++ K3
=
K , 2
( ip) .

Equations ( 85 ) and ( 87 ) are easily solved for K, and K.:

K

K
6. - m2% - (ip - m
m 2)Kg -
( 88 )
mi - m2
omi

m , $ o - (ip - m 1. )) K3
-

( 89 )
K2 m2
m
1

Unfortunately, these deceptively simple equations for K, and K , are highly complicated,
complex variable expressions . To use this solution we have to go back to Equation 73 , 74,
and m
and 78 and obtain the complete expressions for mi
m m2 in aerodynamic terms . They are

ipIX M + M.

m1 , 2 (** ) 21
+
a
21
α
+

2
iplX M + M. M M

( 21
q
21
α
+ ip
PB
I
+

. I
a

( 90 )

Completing the square under the radical,

2 2
iptx M + M.
a a
iplX ipiX Ma + Ma M +M
α
M
pß
M
a
m =

[***]:/ **) 2(*)(*_*).63 )

+ + + + ip ( 91 )
11

+
1, 2 21 21 21 21 21 I I

2
At this point an assumption is made that the term (M M
q
+ Ma) / (21)] 2 is small compared to the other
terms and may be neglected. This is often true and for many applications it is a legitimate
assumption. If this term is deleted and the other terms regrouped the following equation results .

2
ipi M + M.] PIxX M /M + M . M

m2 ( 9 :V |(95** )--(" X***;)

21
X
+
q
21
a
i
2
21 I
a
+ 2
21
X a
21
a
+ ip
pß
I
( 92 )

If we let

2
PIx M
E

2 2 α
= i
C ) 21 I
( 93 )

M + M M
72 X å pg
= 2(**)(*_* M4) .
2
21
q
21
+ ip
I
( 94 )

27
 then

ipI M + M:
α 2 2

* * V2
X
m
-
1, 2 21
+ a
21
E + F ( 95)

2
where EP + F2 can be approximated by the binomial expansion

4
2
+ F2 = E + F2 F
E

2E 3
+ ( 96 )
8E

2
If F << Eº the series converges rapidly and only the first two terms are required to approxi
mate the radical. Or ,

2
2 F
E + F2
F z E +
2E ( 97 )

For most cases this should be an excellent approximation . Substituting ( 93 ) and ( 94 ) in ( 97 )

yields

ipi M + M M
2 X a å "PB
E F
2
+
PIX M
a
2
21 21 )
+ ip I
( 98 )

12+ ) , 21 I

2i
PI
21
X
12

.
-
M

I
a

Now from ( 95 )

2
ipI M + M M
X q å PIX
) a
+

m1 , 2
( »)
- (» 21
+
21 21 I

pl M + M
X q M.á PB
21 21 + p 21 ( 99 )
2
PI M
X a

21 I

We can now separate real and imaginary parts of m and we let the real part equal a and
'1 , 2 1,2
the imaginary part equal w1
W1 , 2 :

28
 M
PIX /M a + Mà pß
M + M. 21 + p
q a 21 I
m
1,2
=
1 , 2 + 101,2
λ.
11,2 )
21
PIX
2
M
α
+

21 I

2
pI M
X x α

[( ) = M
i
21
+ 21
) I
( 100 )

Separating this into real and imaginary parts and labeling the real parti (which is often used
1, 2
in connection with damping ) and labeling the imaginary part w1 , 2 ( a symbol normally used for
frequency) was not accidental; it required considerable insight. A little rearrangement and
algebraic manipulation of ( 100 ) provides

PIx
M + M. .

a a 21
=

m
1,2 == λ^' 1, 2 + iw1.2 1 +

1+
21 2
pi M
x α

21 I

MO
B 2
PI M

( ( 21 pI
X -
α

PIx
21
2
X

I
α
Group
+ i
21
x
1+
21 I
( 101 )

Now we can let

2
pl
X
21
S =

() M

I
a
( 102 )

which is called the stability factor . It is a useful expression in determining whether or not a
spin stabilized shell will be stable . It must be greater than +1 for gyroscopic stability .
If we define t as

PIX
21
T =
11

1 ( 103)
PI M
S x a
21 I

then Equation 101 becomes

29
 M + M. M
q
au a
pß
m
m21 , 2 = 11
^1,, 2 + 191,2 iw

( 21
+
(1 + r) + I
X

PI
X ( 104)
i
([ (1++))
21
1

It should be recognized at this point that t is really an indicator of the magnitude of the contribu
tion of gyroscopic stability.

Equations 88 , 89 , 100 and 104 can be used to obtain usable equations for the two constants
and o
of integration , K, and Ky2 ,' in terms of 11 ,, 2 1, 2
which are defined by the aerodynamic
coefficients and moments of inertia . Substituting

m
1,2 == λ111 ,, 2 + iw 1,2

into Equations 88 and 89 yields

K
šo - ( 12 + iw, ) & - ( ip - 12 - iw ,) K3
o
( 106 )
1 w - ( ^2 + iw )
-

( 1, i iw
+ ive)

К. -
& - ( ^1, + io ) $ - (ip - 12 - iw ) K3 ( 107 )
Il

(^2 + iw ) - ( 1,
^ + iw )

Substituting

Š = iã + B from Equation 60

। = iã + ß from Equation 61

K = iA , + B , from Equation 118

K3 3 B3

and rearranging the denominators of ( 106 ) and ( 107 ) provides

Bo - (1, + iw ,) (iã + B ) - (ip - 12 - 1w) ( iAz

iä + B +
0
+ B
) ( 108 )
Il

K1 2 + ( ^, -12
i ( w - wy)

iw
К.
ia + (1 + iw ,) (iã + B.) - (ip
-, ( - 111 -1 ) ( iAZ3 + B3
B ) ( 109 )
Il

1 i (w2 - ^, ) + ( 12 - ^,)

Multiplying and collecting imaginary and real parts

30
 B
K
ilä -Ana
-1zão
λ
apa + λ ,Az
--WB 23 - (p -w ,)B3]
1
i (w, - wa ) + (^' _ - 2
1

tw B
A +»1 ,B3
[(7. - 12B ++ wnão ++ (p - wW )A3 +
( 110 )
^ -12 )
i (w , -w2 ) + ( ^,

w ) A -

K iſän - azā. -W ,B.+ 1 ,A3 - (p - «, B3] 1

II

λ
i (w2 -071 ) + ( 12-11 )

λ ω
+ (3. - 4,6.B. +®,ā,+(p-w
į (w2
. W , /A,+1,B3]
A3 1
( 111 )
i ) -1,7)
- w 1,) + (12-11

These equations are of the form

ia + b
K =
ic + d

and they can be separated into real and imaginary parts by

( ia + b ) ( -ic + d ) i ( ad - bc ) + ( ac + bd )
-

K
( ic + d) ( -ic + d ) 2.
(c² + d²

Consequently ,

. B
K
- 12ão -wzB. + 1 ,Az - (p - w,)B3)
-

( 4, -12) 2
1 2 2
(@_- w . ) . +
(1, -1,2) λ

(B. -12B +wzão + (p - w, )Az + 1 B3)

+
] (0,
W -w. ) -

2 1
-

2
2 2

( -W 2) + (1, -1,)?
1
w

lä.- dzāo- wzB. + 12Az

A3 - (p -0 )B3] ( , - ,) λ.A
23
ω
2
2
(w - w ))2 + (^( 4,, - λ1,2)
+

+ ( - 12B
.

++ ωwzão + (p - wy )A, + 11, B3 )(9, -12 )])

2 2
( 112 )
(wy - w.)? + a( ^ 1.)
1
- -
λ

31
 läo - 1 ,ão -W,B. ++2λ ,,A3
Α
- (p - @ )B3] ^2 -14)
.] ‫ܒܝܝܝ‬
;‫ܪ‬
K. = i
2
(W, - w ,)) + ((12₂ - 1 )
2
w +
1
2

B - 1,5+ wzão + (p -w,)A3+ 1 B3)(0 , - ) + λ

6.12.2016 0,32 + (12 - ‫ܝܢ‬ (wy -W , ) + (12 - 1

2

+ llão -- Ação -W,B. ++1,4

λ Α
1A33 -(p-w,)B, (( wą - wy )
2 2
+
(w, -w, (12 - 1 1

3( - sā . ++ wā. + (pp -w, )A, +1, B3 (^212 - xy ? -

(w , -w , )2 + (1, -2,1
) 1
A, ( 113 )

A similar procedure is required to separate the real and imaginary parts of' Kz
K
. From Equation 83
N
3 Ng3
K B3 ( 114 )
(ip - m 1, ) ( ip m ) (ip - ^ , -iw ) (ip - 12 - iw ! -

Completing the multiplication and separating the denominator into real and imaginary parts

N3
( 115 )
11

K; λ
ii [f12 (@(02; - p) + ^,1 (0, - p)]+[wy(p - wy) + pp (@ 7 - p)+1,1
1 2

This equation is of the form

N
3
K3 – if + g

and can be separated into real and imaginary parts by

N3 (-if + g )
N N, (-if + g)) iMgd ( -if + g
3
K3 ( if + g) ( -if + g)
(f2
(f + g )
2
3 y ;2 +
I
f + g
2

=
MS ſig + f
K
K3 I 2
f + g
2

Consequently,
32
 M.8
K. = i
8 wą (p -wy) + p (w - p) +1,12
3 I 2
λ
[12(wy
13 - p) +2,( , - p)] + (WZ(p
[ , =- 0,wy) + p («w, - p) +-22 ]
λ.
+
M8 12 (w 1 - p) +1 1, (w, - p)
( 116 )
I 2

(12 (@; - p) +1 , (0, - p)] + [w, (p - «1 ) + p (wy - p) +

1 2
+
2
ਆ 2

Since we now have equations for the three K's in terms of , and w , we are ready to break up the
exponentials part of the solution into real and imaginary parts . Then it will be possible to obtain
real expressions for ā and B. Recalling Equation 84

m . t mt
K , em 2 ° K ipt
-
Š = K, e
Ke"
1 :*1 + Ke * + Kye
and since

$ = iā + B
-

m1,2 - 11,2 + iw1,2

iw. )t
(^2 + iw , )t
ia + B = K,1 e ( ^2 + i , + Ke
2
+ K. e ipt ( 117 )

Referring to ( 112) , (113 ) and ( 116) , it can be seen that the K's are complex and are of the form

K = iA + B ( 118 )

and that the absolute value of K is

|K -VA
2
B?2
+ B

Hence

( 9, + iw , t) (^2 + iv , )t
jā + B = (iA , + B, je
=
B + (iA , + B e ipt
+ (iAz + BBzle
) ( 119 )

Now

iwt
e = coswt + i sinwt .

Consequently Equation 119 can be separated into real and imaginary parts by

33
 t

ia + B = ee =

i (A 1, cos wyt
1 + B, sin wyt) + ( B, cos wyt
A.
1 - A1
sin
w1 t )

te
12
[ i ( A2
A , coswyt + B, sin wyt) + ( B2 2 2t - A, sin wz
COSwt

+ li ,
+1 A cos pt + B, sin pt) + (B, cos pt - A, sin pt)1
(Az +
3
( 120 )

Equating iā to the imaginary part and ß to the real part of ( 120 ) yields the complete expressions
for ā and ß , which are

at
det
RI

1 +
ā
= (A , cos +
wyt + B,1 sin tht)ee
61 (A, cos watt + B, sin wyt )e

+ (Az3 cos pt + B sin pt) ( 121 )

it
1 12t
B/ cos w, t - A , sin w, t)e
B = ( B, +
+ (B, cos wyt - AA2 sin wyt)e

+ -
( 122 )
( B3 cos pt - A, sin pt)

where the A's and B's are obtained from (112 ) , ( 113) , ( 116 ) , and ( 118 ) . They are

+ - -

- 12ão - wzB 12A3 - ( p - wy ) B3 )( ‫^ܐ‬,n1 --12 )

1, A

-
λ
(wy - Wwy)2 + (17 -1,12
^,)?
2
2

(8 - 12Ã +wzão + (p
p -wy)Ag + 12B3) (wy -wy) ( 123 )
)‫ ) س‬- ‫)ي‬2 + )-۸
a
‫ه‬12‫ر‬,‫د‬2 )

λα
Azā - w ,B. ++ ,Az - (p - w ) B3) (w -wy)
2 ?
..& .. ݂‫ܒܒܠܕܫܝܸܟ‬
B

(wy - wy)2 + (1,

(417 - )?
2
+

+ λ ) + λ В.
+ (8. - 12B +wzã + (p -wy)Ag + 1, B3 (4, -12)
o
2
3
2
( 124 )
(wy - wy)? + (^4, -1)
12

34
 lä Az - (p - wy ) В.B3 ( ^2 - ^ ,)
B + 1213
^ ão -w7B.
-

- ***
A2
(wy - wy )2 + (12 - 1,79) -
2
134,-4
λ ΒB3
[ - A + +(p - B. - , B. + ,w ,a + (p - wy ) A3 + AB
1 o 1 1, o((wy,-- 1 2

*** (wy - ,w )2 + (12

2 (ang- ., ) +
2
( 125 )

B2
Ilklå - 1zā - -wzB +AA
1o + 1/Az-(p-
- (p - w ) B3) ((wy-- wy))
1o 1 3 1
2
В.

(wy - wy)2 + (1, --1,79

2 4)
1

w.ā
+ 1 B ( 2 -4
le - 17 ] +wão + (pp - wy) Ag ++ 1,334, ,) 1

2
( 126 )
(wy - wy)2+ (^, - 1,) -

Mgo wy (p - W 1) + p (w 1 - p) + ^_ ^2
Ag A
3 I
λ
2
+ -
)
( 127 )

Ez (wy - p) + 1, ((w, - p)] + [uz (p – wg) + p (w, - p) + 12 132

2 2 2

M 8
8 12 (w;1 - p) + ^, (w - p ) ( 128 )
II

B3
.

I 2 2

Az (wy, -p) + 1, (wz, - po]° + ["z (p - wy)+ p (w, - p)+ 1,1 ]?}
[12 + λ
2 2 1
- )

where ā, å, B, and B are the initial angle and angular rate conditions , and a1 , 2
and w
‫ܣ‬1i , 22
(which have been defined in Equations 101 and 104.) are

PIX
M + M. M M + M.
q a pВ q a 21
à
2,2-(M_ )(117)
T B+ 21 ) ) I
х
T =
21
pi
2
M
+ ( 129 )

Ver
21
X
I
a

pl
х
/ MpB
21 21

2
M

VE 21
X -

I
a

2
pI PI X M
PIX X å
w
1,2 () a 4, 21 (1 ++ ) 21
+
21 I
( 130 )

35
 where the subscript 1 refers to the + signs and the subscript 2 to the - signs . The equations for
a and B are obviously complicated and require the computation of ten constants before ã and B can
be calculated as a function of time . Fortunately they can be simplified to some extent for many
applications, which are considered in the next section .

Application

Assuming that the problem meets the linearity and small angle restrictions , that the physi
cal and aerodynamic characteristics are known, and that the initial angle conditions are known,
the motion of a vehicle can be calculated directly from the complete Equations 121 through 130 .
These equations are complicated and would normally require machine computation. Fortunately
it is possible to greatly simplify the equations for most applications .

Simplified Equations

The basic Equations 121 and 122 for ā and Bcannot be simplified ; however , most of the
complication is in the expressions for the A's and B's , which can be simplified . In many flight
systems the static stability is much greater than the dynamic stability, so that 1.1 , 2 « W1,2
This is usually, but not always , true for fin stabilized rockets and bombs , and is often true for
λ
re - entry vehicles. This condition provides the basis for eliminating all terms involving 11 , 2
in the equations for A and B :

B. ++ wzā .
B ão + (p - w .) A 3
A ,1
..[ w - w2 ng ( 131 )

o - wzB .
-B -

w2
( 132 )

+ ω, α
+ (p - W ) A3
‫܂ܫ‬..‫ܝܝܓܝܝ‬
A
A2
w2
( 133 )
Rii

ão - w, B. ( 134 )
II

B2 w2 - ♡ 1

M 8

An ,(pg
4, 6-41- W7 ) + p (wy - p)}
I
2
р (p † wa ( 135 )

( 136 )
B, = 0, (p #w ).
3

36
The percentage error involved in this simplification is approximately 12 (42,2/w7,2)
For two
^ w11 , 100%.
- arm or epicyclic motion ( zero trim) , the equations for the A's and B's reduce further to

+
B. + wc
=
-

A, WE -
( 137 )
2
RI

- BO
:

***
В.
1
1 2
( 138 )

B + w.ā
B. 1 .
A2 ( 139 )
w
®22 1
:RI

-W.B
-w
o l'o
( 140 )
11

B2 ~2 - 1

Az3 == B.
B
3
= 0 . ( 150 )

Example Problem

To demonstrate the accuracy of the simplified approach, two sample problems were calcu
lated using the Tricyclic Theory and a Six - Degree - of - Freedom computer program , 7 and the

results were compared. The example was deliberately chosen to be as linear as possible , yet it
is representative of a supersonic fin stabilized rocket. The rocket was launched at an angle of
1 ° above the horizontal and zero drag was assumed to maintain a constant dynamic pressure .
The vehicle characteristics are listed in Table I. The static moment derivative is calculated from
Equation 48 :

М. C
α = Cma q ' Sd

Maα = -30 ( 400.35 ) ( 0.442 ) ( .75) = -3982 .

The frequencies are obtained from Equation 130 :

2
PI IpIX M

“1,2 21
x
+
Here21 I
a

1/2
2
37.7 (3.37 -3982

0,2°(2.10:32)=[(0
“1
7:300597)*
37.3 (3.37)
( )
+
2 ( 101 ) 101

91,2 = 0.629 + 6.307 = + 6.936, - 5.678 rad / sec .

37
 TABLE I

Vehicle Characteristics and Flight Conditions

( Sample Problem)
q' 400.35 psf

II
2
Ι
IXX = 3.37 slug - ft?
2
I = 101 slug --ft

S = 0.442 ft2
d 0.75 ft

=
р 37.7 rad per sec

с = -30 per radian

Cma
с = -1400 per rad / sec
ma + C
ma
cmo = +30 per radian

8 = 2 degrees

The damping moment derivative is calculated from Equations 50 and 52 :

)
q' sa 2?
M
q
+ M.
a ( Cmq + Cm 2V

M + M. = (-1400 ) (400.35 ) (0.442) (0.75)2

(0.75 )
q a 2 ( 5000 )

M.q + Ma
Mà = -13.93 ft- pounds / rad per sec

The l's are obtained from Equation 129 :

M + M.
q a
(1 + T)
II

^1 , 2 21

-13.93
11,2 2 ( 101 )
=
( 1 + 0.0997 ) = -0.0758 , -0.0621 ,

where t is

pI
X

21 0.629
=

T = = 0.0997
=
.
6.307
Ma

2
PIX α

21 I

Both example problems are identical , except that the first calculation has a trim angle ( o = +2 °)
and is tricyclic (all three arms are present) and in the second example the motion is epicyclic .

38
The initial conditions are the same for both examples, that is ,

ở = 10 °

-
0
형
: 0
B.

§ =0 .
The A's and B's are computed from Equations 131 through 136 .

Tricyclic Case

M 8
8 1
Α.
3
o[ I
+
°2 (p - ~ .1 )) + pp (w - pp)

3982 ( 2 ) 1

h, 12
A
101
- ,a -
( -5.678 ) ( 37.7 - 6.936 ) + (37.7 ) ( 6.936 - 37.7 )

A
A3 -0.0591 °

В = 0
3

Α.
1 LOŽ. +wzão
W1-02
+ (p - wy ) Az 3

0 + (-5.678) (10) + (37.7 - ( 5.678 )] ( -0.0591 )

A
1 6.396 - ( -5.678 ) |(-0.0591)
A 1, = + 4.705 °

äo -w2B. 0 - ( -5.678 ) ( 0 )
RA w
1-2 ) - [៖ 6.936 - ( -5.678 )
-
0

B +wjā.
1 + (p – « _) Az3
[, 10 --, ]
11

A2 W 2 -- W 11

39
 -
0 + 6.936 ( 10 ) + ( 37.7 - 6.936 ) ( -0.0591 )
= +5.354

11
A2 -5.678 - 6.936 -

a
- wB. 0 - 6.936 ( 0 )
B2 C 02-01 LE
.

-5.678 -6.936
-
= 0 .

Since

2 2
K

A + B
|K| -
|K, = 4.705°

|KZ| = 5.354 °
= 0.0591 °
|K3|

Epicyclic Case

+ wzão 0 + ( -5.678 ) 10
AL - w
1 2
-- 6.936 - ( -5.678 ) -
+4.501 °

äo - B. --

24 1.
В.
«1-2
0

+
B. +wão 0 + (6.936 ) 10
= + 5.499 °
A

44
---( w
2 1 -5.678 - 6.936
-

0
Az == B3
B
11

| K, ) = 4.501 °
5.499 °
|K21
=
.
||K3| = 0

Note that rather small K,3 arm (0.059°) has a large effect on the size of the K , and KK,, arms ,
which causes a much larger effect on the motion than might be expected , and usually should
not be ignored . Also notice that initially both the ,
K , and K 2, .arms are pointed upward along the

40
+ @ axis and the K3 arm is pointed downward along the a axis . * The calculations for both the
tricyclic and epicyclic examples are shown in Appendix A and are plotted on Figures 5 and 6 in
comparison with the 6 DOF computer calculation . The agreement is excellent , as can be seen
on the plots and by comparing columns 21 and 22 with columns 23 and 24 in Appendix A for the
tricyclic example and comparing columns 31 and 32 with columns 33 and 34 for the epicyclic
case .

The error is approximately 1.5%, which is sufficiently accurate for most flight dynamics
purposes .

Discussion

The Tricyclic Theory is the only existing theory that is capable of predicting the motion of
a flight vehicle in considerable detail by means of a closed form solution. Other flight dynamics
techniques involve the numerical integration of the equations of motion . The most complete form
of numerical integration , the 6 DOF computer codes , provide complete nonlinear and time variant
solutions ; however , it is often difficult to obtain a thorough understanding of the motion required
to diagnose flight dynamics problems . Conversely, the Tricyclic Theory provides the basic
underlying principles of the motion but is restricted in application to linear or quasi - linear and
time invariant or quasi - time invariant problems . For these reasons it is important that the
Tricyclic Theory be applied to flight dynamics problems and that its restrictions be understood .

Constant Coefficient Restriction

Constant coefficients ( N 1 , N, and N,) are assumed in solving Equation 76. Unfortunately ,
this occurs in few flight dynamics problems , because Mgα and( M q + M.)α
almost always vary with
time and often change with angle of attack ( nonlinear) . Aeroelasticity is an often overlooked but
serious contributor to nonlinearity . To further complicate the situation , there are no simple
established specifications on the error involved in applying the theory to problems where the
coefficients vary with time or are nonlinear . However , two comments can be made in this
regard . First , a nonconstant coefficient problem can often be reduced to a quasi - constant
coefficient problem by dividing the problem into small time increments , so that the coefficients
do not change significantly during the time increments. Second, w1 , 2 appears to have the
largest effect on the magnitude of K 1, and K ,; consequently, variations in M , probably cause the
α

largest nonlinear effects . While indiscriminate application of this theory is inadvisable , appli
cation with caution can often solve a flight dynamics problem that otherwise would not be possible
to solve .

In the 6 DOF Aeroballistic System . See section entitled Physical Interpretation .

41
 ‫مل‬
‫ܐܐ‬
10
O TRICYCLIC THEORY
8
6 DEGREE - OF - FREEDOM
COMPUTER SOLUTION

2
10

Ã
8 6 4 2. -2 -4 -6 -E

-2

4
-

-6

-10

Figure 5. Comparison of Tricyclic Theory with 6 DOF Calculation

42
 oper
10
O EPICYCLIC THEORY
-8
6 DEGREE - OF- FREEDOM

oa
COMPUTER SOLUTION

-6

-4

- 2

L
8 6 4 2 -2 -4 -6 -8

-2

--4

--6

-8

-10

Figure 6. Comparison of Epicyclic Theory with 6 DOF Calculation

43
 Small Angle Restriction

Inherent in the definition of the applied aerodynamic moments in the aeroballistic equations
is a small angle of attack restriction that is not immediately apparent . This results because the
total velocity vector ( V) cannot remain perpendicular to both the ý and 2 axes when both ā and ß
are present . The distortion in the angles is demonstrated in Figure 7 , where both y and ž have
been rotated so that I $ i 90 and ã =B. It can be shown from trigonometry that
-

a = az = B,3 1 = B2 = 45°

a = B = 60° .
=

From this geometry it is possible to see that the true angle of attack and the sideslip (@c, Bed
that would normally be associated with the aerodynamic coefficients are
RI

= ā cos 3
R

, = B cos ā

The error caused by this assumption can be approximated as follows for circular coning motion .
The frequency w, is approximately

M
a
20

so that the error in w, is approximately

M
a α
Δω ,1 I osb) ,
ūcos ß

which for a 10° total angle of attack is

1 w1
= W
Δω, '1V cos 100 .99237

Or , the maximum error in frequency would be 0.76 % and the average error would be roughly
0.48 %. Therefore this restriction is not too significant as long as the angles are small .

Physical Interpretation

A physical interpretation of the Tricyclic solution is somewhat complicated by the complex

algebra that is involved. An attempt is made in Figure 8 to provide a graphical description of
the solution. The first thing to be recognized is that the equations of motion that are solved

44
 Х

>
‫ܐ‬

RI
ā
x

lan
!

Š α
2

ā
109

1
В

N
B.

IN J - Z PLANE AND PERPENDICULAR TO V

Figure 7. Effect of Large Angle on ā and ß

45
46
६

B
X

1x
К
W2 2
CENTER
OF α

20
GRAVITY

lun
Х
FLIGHT
PATH

10
Y
iā

:Ý
PNOTE
AND
AXIS
BARALLEL
AND
Z
AXIS
P
Ā ARALLEL
ãIN
ARE
iã
AND
PLANE
SAME
IN ND
ABŘRE
PLANE
SAME

18
ŘND
ãRE
A6DNORMAL
OF
ANGLES
AEROBALLISTIC

N
Figure
Physical
8.
Interpret
Tricyclic
of
Theory ation
constitute the sums of the moments about the body y and ž axes . Consequently , w and
2
must
1
be about the body axes . The solution , repeated below from Equation 84 , indicates that W1 W2 '
and p must all be about the same axis .

( ^ + iw ,)t 12 + ,
(^2 iw )t + K , eipt
& = K,1 e + K2 e
3

and since p is about the body x axis , w 1 and w must also be about the x axis . Consequently , the
2
three K's rotate about the x axis at their respective frequencies and describe the motion of the
flight path with respect to the body x axis . This is opposite to the normal convention used in the
aeroballistic option of 6 DOF programs. However , the whole thing can be inverted without any
sign reversals resulting (iā and ß are interchangeable with ã and ß ). This is shown in an ã -ß
plot in Figure 9, where ã and B are the aeroballistic angles measured from the flight path to the
body x axis . It should be noted from the solution that the order of addition of the K's is unimpor
tant and consequently the order of the K arms in Figure 9 is unimportant. Equation 119 shows
the complete solution where the K's are shown in complex form :

( ^4 + iw ,)t (^ 2 + iw ,)t
+
pt
Bg) eei
=
ia + B = (iA , + B ) e B2 e
+ (iA2 + ) e
+ (iAz ++ Bz

If we let t = o in this equation,

ia + B + B
B. = (iA, + BB ) + (iA , + B2 ) + ( iAz + B)

and

A , + Az
ā. == A1 + A2

B. == B,1 + B2 + Вз
B3
B B

Therefore, the A's and B's are simply the ā and B components under initial conditions. It also
follows that the 4 angles shown in Figure 9 can be calculated from

B
-1 1
= tan twt
41 Α.
A1 1

and similarly for 42 and

3.

47
 20
MISSILE NOSE

以3

K3

D 42

K2

W2

中,
K

FLIGHT PATH
ω

Figure 9. ã - B Plot

48
 The three K's or arms and the three associated frequencies have been previously defined
as follows:

K1, - nutation arm w1 nutation frequency ( high frequency)

K2 - precession arm 2 precession frequency ( low frequency)

K; - trim arm p - roll frequency

While these particular definitions may be in conflict with gyroscopic terminology it is important
that they be maintained ; otherwise confusion may result. The fact that the nutation arm (K )
always rotates in the same direction as the roll rate (p) and that the precession arm (K ) always
rotates opposite to p, for an aerodynamically statically stable body is important. This is easily
seen from Equation 130 , which is repeated below .

2
I I M

~ 1,2
х

21
NA
21
X

M
I
a

a
For a statically stable body I is always negative; consequently the radical is always greater
than (p 11)X / ( 21), and w1 will be positive and w2 negative for a positive p . Also if M a / I is small
( corresponding to exoatmospheric flight ), w become one frequency
1, 2

I
X

(4-12
PI /( 21) , which
which is the gyroscopic frequency . Conversely, if Ma/ I is large compared to (pIX)
corresponds to normal aerodynamically stable flight at high dynamic pressures , then

-M
a

wif I

which is a well known approximation of the nutation frequency. The gyroscopic frequency is
usually 5 to 10 % of the aerodynamic frequency on a fin stabilized rocket vehicle flying at high
dynamic pressure .

Characteristic Motion

While an infinite variety of motion patterns can be generated by the Tricyclic Theory, a
few types tend to predominate . Some of these have been generated by a mechanical device to
illustrate epicyclical motion .

The motion shown in Figure 10 is for the case of zero roll rate (p = 0 ) when the nutation
and precession frequencies are equal (w = w ) . Figure 10 shows the characteristic types of
motion with different initial conditions . In Figure 10 ( a) planar motion is generated when K and

K, are equal, which is the case when B, is zero. As a progressively larger negative value of

49
 el

18
Kz2 K2

N
K.
K,
B B

(o) B. : 00 (b)) o > B, <> w.zão <

P : 0 0 : 0
: K
K = K2 к > к.
RI

RI
a

K2
K
,

K,
B

B4

(c )) 0 >> Bo < wijzão

, (d) Bo = wizão
ps:
D 0 р : 0

K > 2 K2 : 0
K₂
K, *
Figure 10. Characteristic Motion with Zero Roll Rate

50
ė, is introduced, the K,2 arm will reduce ( see Equations 137 through 140) and K, is maintained
the same length, resulting in an elliptical motion ( Figures 10 (b ) and 10 (c ) ) . Finally as

९०
ā ,' the K2, arm becomes zero and circular motion results with only the
becomes equal to w1, 22ão
K, arm remaining. If the negative value for B is increased further so that it is greater than
o

w1 , zão'
ā elliptical motion results with the major axis being along the B axis. In this set of
sketches K 1, is kept constant , and a K , This
ā , varies and is simply equal to the sum of K , and K,
o

condition is inherent in the mechanical device used to generate the sketches .

Figure 11 is a similar set of motion patterns for a rolling vehicle. In this case the nutation
frequency is greater than the precessional frequency (w
. = -1.083 wy) , which is typical of fin
stabilized vehicles . The unequal frequencies cause the oscillatory plane to rotate in the direc
tion of and p , since The angular rate of this precessing motion ( j ) is
ယ၊ ယ 1 is larger than w2

WqAt ++ wzAt
‫ = ܩܢܐܬ‬wint
‫بله‬ PI
TA

- W
11

+ ‫به‬
11
o

At 1 2 I

Consequently the magnitude of ý, depends only on the roll rate and moment of inertia ratio of the
vehicle . Also , it can be shown that y depends only on
on wylwy
w :

2-( +) 2
1
3/30
O

Figure 11 (a) demonstrates a pointed cusp pattern that results from B, being zero. As B.
increases in a negative direction, the loops are formed which change into ellipses in Figures 11
(c ) and (d) . As before, K, is constant and K,2 gradually decreases as Ő is increased in a nega
.

tive direction. This reduction of , K2 with respect to K, as - B is increased can be seen in

Equations 137 and 139 .

Aerodynamic damping was ignored in generating the motion patterns . Normally, damping
would gradually decrease the magnitude of K, and K, if the vehicle is dynamically stable.

Magnus Moment

The magnus moment, which results from combined angle of attack and roll rate , was
included in the theory, since it can be important when the body is rolling at a high frequency.
Magnus moment is usually important on a spin stabilized shell and is normally considered
unimportant on a fin stabilized vehicle . This can be qualitatively demonstrated by examining the
^11 ,, 2 equation for a 6 - inch - diameter spin stabilized shell and a high fineness ratio fin stabilized
rocket.

51
 RI

RI
α

la
B

‫ (ܘ‬a ) Bo =0
: 0
(0)b B‫܂‬oro
PO D
ω
w , : - 1.182w2 W , ::-1.182w2
K , > K2 K
, > K
K2
RI

aa

RI
DOI

B B

(c)) B.YO ) Botho

(d)
DO DO

W , : -1,182w w , : -1.182
Κ
K , >K₂2 K >> K2

Figure 11. Characteristic Motion of Rolling Body

52
 For a typical 6 - inch spin stabilized shell , is approximately 2 ; therefore from
Equation 129 :

M + M M
pe
1,2 ME (1 + )on
q
+
21
α
I
X
(7 )

M + M. M
q pl
11,2 (MM) )1+2 + ‫ܩ‬ α
( 1 + 2) + I
X
(2) :

Now I is about 0.1 and I about 1.5 for the shell .

X

1 + 2 2
11 ,2 (M
q
+ M :)
(*2) + mpole ) (M.2 + Mu?(1+ 3);
α' 3
M
10.1
= (M
M M.)
:)
a
+
3
+ M
PB
( 20 ) .
.

From this it can be seen that, on a spin stabilized shell , M. can overwhelm ( M + M.
α
) even if
PB q

it is numerically much smaller ; in fact, M PB can provide a major part of the damping on a spin
stabilized shell . Now if we substitute the numbers for a high fineness ratio fin stabilized rocket
with an I of 2 and an I of 300 we have
X

M + M M
a α pB
11,2 2 ( 300 ) ( 1 + .01 ) + 2
( .01 )

1, 2 - (M.q + M.) 0.0017) + M pg ( .005 ) .

Converting the moments to coefficient form ,

11,2
λ
[Cma + C :)
må
.)
( .0017) + СCmpß (.005) qsd .
Wind tunnel tests have shown a value of 2500 / rad for (C
8
ma
+ C.)
ma
and 50 / rad for C
C mp
ß on the
Nike Tomahawk , a high - finenessratio fin - stabilized rocket. While these data indicate that
M has a relatively small effect on the motion of this type of vehicle, one should not make a
pВ
conclusive judgment on the relative significance of M in all cases on the basis of a single wind
pВ
tunnel test .

Conclusions

A highly detailed development of the Tricyclic Theory has been presented and its restric
tions discussed. A simplified version of the theory was obtained and two example calculations
compared with 6 DOF calculations. This work leads to the following conclusions .

53
 1.
Properly applied, the linear Tricyclic Theory provides an excellent means

.
2 of predicting the motion of rigid bodies .
The simplified method of evaluating the K arms will be adequate for many
if not most flight dynamics situations .

3. The contribution of the K , arm to the magnitude of the K, and K, arms, which
has often been neglected, is significant.
4.
The Tricyclic Theory provides a means for understanding flight dynamics prob
lems that are often otherwise incomprehensible .

54
 REFERENCES

1. Nicolaides , J. D. , On the Free Flight Motion of Missiles Having Slight Configurational

A symmetries , Ballistic Research Laboratories Report No. 858 , June 1953 .
2. Nicolaides , J. D. , Free Flight Dynamics , Text , Aero - Space Department, University of
Notre Dame , Notre Dame , Indiana .

3 . Murphy , C. H. , Free Flight Motion of Symmetric Missiles , Ballistic Research Laboratories

Report No. 1216 , July 1963 .

4. Scarborough , J. B. , The Gyroscope , University Press , Cambridge , 1958 .

5. Etkin , Bernard , Dynamics of Flight, John Wiley and Sons , 1962 .
6. Halfman , R. L. , Dynamics , Addison-Wesley , 1962 .
7. Six - Degree -of- Freedom Flight Path Study Generalized Computer Program , FDL - TDR - 64-1 ,
Air Force Flight Dynamics Laboratory , Wright - Patterson Air Force Base , Ohio ,
October 1964 .

8. Curry , W. H. , and Reed , James F. , Measurement of Magnus Effects on a Sounding Rocket

Model in a Supersonic Wind Tunnel, AIAA Paper , September 1966 .

55-56
 APPENDIX A
Sample Calculations

57-58
 Epicyclic Example

( 1) ( 26 ) ( 27 ) ( 28 ) ( 29 ) ( 30 ) ( 31 ) ( 32 ) ( 33 ) ( 34 )
7msoo

Tricyclic 6 DOF
Theory
λ.t
Azt
zy
‫ܬ‬at

‫ܬ‬st
20
e

?.ə
ət

RI
B
ə?

a
I

?R
A
A
A

0 1 5.499 +4,501 +5.499 0 0 +10.0000 +10.0000

0.37 5.488 +4.222 +5.268 -1.530 +1.542 + 9.490 ] +0.012 + 9.497 +0.010
0.35 5.466 +3,434 +4.542 -2.858 +2.957 + 7.976 +0.099 + 8.075 +0.080
0,12 5.455 +2,239 +3.573 -3.838 +4.118 +5.812 +0.280 + 5.886 +0.254

0.33 5.433 +0.794 +2,260 -4.362 +4.944 + 3.054 +0.582 + 3.162 +0.553

0.20 5.411 -0.729 +0.817 -4.358 +5.335 + 0.088 +0.977 + 0.179 +0.972

0.97 5.395 -2.172 -0.734 -3.821 +5.346 2.906 +1.525 - 2.760 +1.480
0.34 5.373 -3,323 -2,176 -2.858 +4.911 5.499 +2.053 5.369 +2,021
0.36 5.362 -4,082 -3.453 -1.550 +4.102 -
7.535 +2.552 - 7.397 +2.516
0.18 5.350 -4.348 -4.457 -0.070 +2.959 8.805 +2.889 8.657 +2.869
0.39 5.329 -4.105 -5.089 +1.415 +1.583 9.194 +2.998 · 9.047 +2.988
0.12 5,312 -3.355 -5.312 +2.708 +0.101 - 8.667 +2.809 · 8.554 +2.803
0.18 5.296 -2.226 -5.111 +3.675 -1.388 7.337 + 2 , 287 - 7.254 +2.288
0.0 5.279 -0.827 -4.508 +4.204 -2.750 5.335 +1.454 - 5.302 +1,460
0.7 5.263 +0.640 -3.542 +4.220 -3.889 2.902 ] +0.331 -
2.912 +0.376
0.13 5.240 +2,029 -2.298 +3.738 -4.716 0.269 -0.978 0.336 -0.878
0.31 5.2301 3.165 -0.889 +2.809 -5.152 + 2,276 -2.343 2.165 -2.194

0.17 5.21 +3,918 +0.589 +1.560 -5.182 + 4.507-3.622 4.349-3.441

0.19 5.1971 +4,199 +2.009 +0.130 -4.792 + 6,208 -4.662-

6.107 -4.475

0.36 5.180 +3.981 +3.258 -1.302 -4.025 + 7.239-5.327 7.030 -5.150

1.72 5.164 +3,288 +4 , 245 -2.570 -2.938 + 7.533 -5.508 - 7.324 -5.351
1.14 5.147 +2.206 +4.885 -3.518 -1.626
-
+ 7.091 -5.144-
6.910 ) -5.125

1.11 5.131 +0.874 +5.126 -4.046 -0.195 + 6.000-4 , 241 5.866 -4.139

59-60
DISTRIBUTION :
Standard Aero - Thermo Dynamics Distribution List ( 195 )
STA Membership and Military ( 100)
W. J. Howard , 1000
C. F. Bild , 1100
L. A. Hopkins , 1300
R. E. Fox , 1325
J. H. Findlay, 1400
W. A. Gardner , 1500
T. B. Lane , 1540
R. T. Othmer , 1541
F. J. Perdreauville , 1541
T. B. Cook , 5000
R. S. Claassen , 5100
J. W. Weihe , 5250
M. J. Norris , 5260
L. D. Smith , 5500
R. L. Peurifoy , 5510
A. A. Lieber , 5540
D. I. Shuster , 5600
G. 1. Rodgers , 5620
R. V. Clem , 5630
L. E. Hollingsworth , 7200
T. L. Pace , 7210
L. Gutierrez , 8100
A. N. Blackwell, 8110
G. W. Anderson , 8140
A. R. Willis , 8147
G. A. Fowler , 9000
J. H. Scott , 9200
J. C. Eckhart , 9220
A. Y. Pope , 9300
V. E. Blake , 9310
S. McAlees , 9314
G. W. Stone , 9314 ( 25 )
R. C. Maydew , 9320
H. R. Vaughn , 9321 ( 125 )
W. H. Curry , 9322
E. C. Rightley, 9323
W. R. Barton , 9324
J. K. Cole , 9325
K. J. Touryan , 9326
W. N. Caudle , 9327
D. F. McVey , 9328
A. J. Clark , 9330
W. C. Scrivner , 9400
J. A. Allen sworth , 9422
B. R. Allen , 3421
L. C. Baldwin , 3412
B. F. Hefley , 8232
R. S. Gillespie , 3413
For : DTIE (3)
For : H. F. Carroll ( 1 )
W. H. Swiss , 3413
C. H. Sproul , 3428-2 ( 10 )

61