AD-A245 957

INAVAL rvUTGRADUATE SCHOOL

Monterey, California

DTIC
A ELECTE
FEB 1 41992 J

D M
THESIS
DYNAMIC MODELING AND MODAL ANALYSIS
OF AN AIR-TO-AIR MISSILE
by
Michael Allen Shutty
September, 1991

Thesis Advisor:

Edward M. Wu

Approved fur public release; distribution is unlimited

92-03668

z
s;z41,

Unclassified
SECURITY CLASSIFICATION OF THIiS PACL

REPORT DOCIUMENTATION PAGE

1a REPORT SECURI ri- CLASSIFICATION
Unoclassified

I0 RESTRICTIVE MARKINGS

2a SECURITY CLASSIFICATION AUTHORITY

3 DISTRIBUTIONIAVAILABILITY OF REPORT
ApprovedlIirpuiirlas;isrutoisiiniie

2b DECLASS:FICATION/DOW'NGRADING

SCh EDULE
5 MONITORING ORGANIZATION REPORT NUMPER(S)

4 PERFORMING ORGANIZATION REPORT NUMBER(S)

RFORMING ORGANiZTIN
6a NAME OF PRE
Naval Postgraduate School

1b

7a NAME OF MONITORING ORGANIZATION

Naval PosItgraduate School

OFFICE SYMBOL

(If applica ble)

6c ADDRESS (City, State, and ZIP Code)

7b ADDRESS (City, State, and ZIP Code)

Monterey, UA 93943 50001

Mtontere-y. CA 9394356000

180

8a NAME OF FUNDINGISPONSORiNG

ORGANIZATION

9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMFBER

OFFICE SYMBOL

(If aPplica ble)

6c ADDRE SS (City, State, and ZIP Code)

10 SOURCE OF FUNDING NUMBERS

W{

_______________________________I,-,______t

1`101-1 N'O,,

iiltAetf
I

,k

MtA-.0

11 TITLE (include Security Classification)

DYNAMIC MODE)1.1ING ANI) MOI)Al.ANAI.YS1I 01"AN AlI-.TO-AII{MISSILE;
12 PERSONAL AU1iT OR(S)

Nljehoel A.Sliutty
13b ThIME COVERED
114p4ii1e
Fromn

13a TYPE OF REPORT

Mi.sser','I'hviiso

4DT

OF
E REPORT (year. month, day)
JIII1

15 PAGE -COUNT

16 SUPPLEMENTARY NOTATION
IThe views expressed in this thesis tire those of the author aid do not reflect theColbc al policy or posi tionot t.if DeI)epartmflent of IDefeiist. or thle UiS.
Governiniei.
17 COSA~tCODES
18 SUEBJECT
TERMS (continue on reverse it necessary 'ind identity by block number)
FI1I D

GROUP

SUBGROUP

Vibration, modal analysis. finite element modeling, dylinaiie modeling, P 3, AIM 9

19 ABSTRACT (continue on reverse if necessary and identify by block number)

~iiiisiiii.

III
Uglist cliiiiixal~icrahutn IL' kiug raligci
uFIVIUL
air miissile sYslIUi 101jis
response tsi this need, the Navktl Air lest Cetnter was tasked in 1989 toj conduit a 1) /AIM.9 (Side%iidudr integration programl. In support il this
prtigrani, a vibration test stand was established at NIPS, and a ground vibratioii characterizatzioi, was cuiidLiCtid to determine ifo! potential flutter
m t of a two-degree (,I IreediiIorn iuped-mssndI 1ileCeruuui Idtrmiatilitli
problIi ii exstoted. This Lesi. resulted in the d eve ltIii
niissile's resoniance rudes iii pitch With the reuent terminationoifthe P-7A. the P-3 community is now looking to the P-3 Orion If Programt to
cairry it inuto the 21st century. Tl'l Orion 11 will must likely have a beefed tip wing structure. necessitating ain analysis Ofthibwinl, in conjunction
with the AIM-9tmissile. This iivestugation responids to that requiremeont by conicurrently developing a miathienatical model if the AIM- numissile
usi ng fitnite e lemen t techaniq ues to analytically determ inte its, niidalI pi arameters, a rid settinrg op a ruiodalI test system to qua ntify the paramnete rs of
thib model bye), perinientally deterrni niiug Lte niissulesb natural freqtuencies. moude shapes, and t ra nsient respoinse- This folIly uiistrumiented te.iL
system aud associapted methotdologies COUItldten be thiebasis foe cuuiductliug a ci~nililete mo~dal testtif thei AIM 9 nuissile systemn,as wll as to
qucantulfy the v ibration charac~teristi cs (it iothier cuaidid idte ni issi le synitems firtthe P-3 a nd its eventualI ucceswur.

oLr
Th~e P'-3 Orion liatruil aiirplaiii has u nuied [or ii:air

120
El

21 ABSTRACT SECURITY CLASSiI iCATION

[)IS1RIBUTION/AVAILABILI .Y OJ ABS TRACT

'INLLASS11IWAi,.litiM

i, Ii

O.

AM
Mi HtPOR]
ii

22a NAME Of RESPONSIBLE INDIVIDUAL

______
P'rofessor Edward M. Wu

DD FORM 1473,84 MAR

MR,

3t

icas
i2c OFFICE SYMBOL
22b TELEPHONE (!nclude Area code)
7
AA/Wu
408) 646 3459
__
SLCUFIIJYCLASSIFICATINON 1TISPG
83 APH edition riuuy be used Until exhiitstedl
All other editionb airc obsolete
U nclassified
i Is

Approved for public release; distribution is unlimited.

Dynamic Modeling and Modal Analysis
of an Air-to-Air Missile
by
Michael Allen Shutty
B.S., Pennsylvania State University, 1983
Submitted in partial fulfillment
of the requirements for the degree of
MASTER OF SCIENCE IN AERONAUTICAL ENGINEERING
from the
NAVAL POSTGRADUATE SCHOOL
September 1991

Author:

Approved by-

Michael A. Shutty

.___,
Edw

d M. Wu, The::is Advisor

./

Gerald -. Lindsey, Secon

Daniel J. Coiliiir', Chairman

Department of Aeronautics and Astronautics

iii

ABSTRACT
The P-3 Orion patrol airplane has a netd for an air-to-air missile system for defense
against enemy aircraft on its long-range missions. In response to this need, the Naval Air
Test Centcr was tasked in 1989 to condvct a P-3/AIM-9 (Sidewinder) integration program.
In support of this program, a vibration test stand was established at NPS, and a ground
vibration characterization was conducted to determine if a potential flutter problem existed.
This test resulted in the development of a twc degree-of-freedom lumped-mass model and
experimental determination of the missile's iesonance modes in pitch. With the recent
termination of the P-7A, the P-3 community is now looking at the Orion II program to
carry it into the 21st century- The Orion II will most likely have a beefed-up wing
structure, necessitating an analysis of this wing in conjunction with the AIM-9 missile.
This investigation responds to that requirement by concurrently developing a mathematical
W

model of the AIM-9 missile using finite element methods to analytically determine its
modal parameters, and setting up a modal test system to quantify the parameters of this
model by experimentally determining the missile's natural frequencies, mode shapes and
transient response.

This fully instrumented test system and associated methodologies

could then be the basis for condu ting a comprehensive modal test of the AIM-9 missile
system, as well as to quantify the vibration characteristics of other candidate missile
systems for the P-3 and its eventual successor.
.at

'ACo.e.10:,

NT;S

L.
Di~J,t
_

i:Or

C,:&

.'.2

TAPLE OF CONTENTS

I.

II.

INTRODUCTION

...................

..................

A.

BACKGROUND

B.

MISSILE MODELING TECHNIQUES .........

C.

IMPORTANT

..................

...............

DEFINITIONS ............

ANALYTICAL MODELING

..........

.............

...............

..............

A.

FINITE ELEMENT METHOD BACKGROUND

......

.......

B.

MSC/PAL2 FINITE

SOFTWARE

ELEMENT ANALYSIS

1. Introduction ...........
2.

................

Basic Program Operation .....

C.

11
11

..........

12

a.

File .............

b.

Edit ...............

..................

13

c.

Analysis ...........

................

14

d.

Graphics ...........

................

16

e.

Composite Capabilities ...

.........

17

f.

Advanced Capabilities

.........

17

..................

...

12

3. MSC/PAL2 Analysis Procedure ...

4.

MSC/PAL2 Limitations .......

MSC-PAL2

........

............

20

SIDEWINDER MISS T LE ANALYTICAL MODEL

1. Model Geometry ...........

...............

2.

Determination of Modal Parameters

3.

Modification of Modal Parameters

iv

18

22
22

.....
......

26
29

EXPERIMENTAL MODAL ANALYSIS

Ill.

............
.

3Z
.

A.

GEINERAI . .

B.

PREVIOUS WORK ACCOMPLISHED .........

1. Test Setup ..........
2.

C.

Previous Results ........

32
35

.................

35
36

MODAL TEST SETUP CONSIDERATIONS ..........

38
38

............

2.

Transducer Mounting .......

............

40

3.

Transducer Calibration ....

...........

41

4.

Excitation

...........

41

...........

41

Techniques ......

a.

Shaker Excitation .....

b.

Impulse

(Impact)

Excitation

......

43

Specifying the Number of Degrees-of-Freedom

SIDEWINDER MISSILE MODAL

2T.ALYSIS

TEST SETUP

44
45

1. Test Structure ..........

...............

45

2.

...............

46

Test Equipment ..........

MODAL-PC SOFTWARE BACKGROUND

1. Introduction ..........
2.

F.

..............

5..

E.

..........

1. Transducer Selection ..........

D.

.........

...

50

................

Major Features ..............

50
...............

50

EMODAL-PC MODAL TESTING FUNDAMENTALS ...

1. Test Procedures..............
2.

Modal

Testing vs.

Operating

Test ................

3.

.....

..............

51

Deflection

....................

a.

Modal Testing .......

b.

Operating Deflection Shape Test

Curvefitting ..........
v

Shape
52

.............

................

51

52
....

52
53

G.

(Co/Quad)

a.

Coincident/Quadrature

b.

Circle Fit ........

c.

Rational Fraction Polynomial Fit

53

Fit

54

...............

EMODAL-PC TEST PROCEDURE .....

..............

IV.

V.

63

6. Recall Project ........

...............

64

..............

..............

66

.............

67

ANALYTICAL MODELING ........

B.

EXPERIMENTAL

C.

COMBINED TESTING AND ANALYSIS ...

MODELING ......

.........

68

69

..................
..............

69

.............

70

A.

ANALYTICAL MODELING ........

B.

EXPERIMENTAL

C.

COMBINED TESTING AND ANALYSIS ...

MODELING ......

.........

......................

APPENDIX B ...........................................

vi

64

66

...................

A.

APPENDIX A .................

58

..............

RECOMMENDATIONS ............

55

61

5. Store Project .........

CONCLUSIONS ...............

..

.........

and Animate Shapes ...

7. Prepare Reports .........

55

57

..................

3. Data Acquisition ........

4. Extract

54

...........

1. Configuring the General Set-up .........

2. Geometry ............

..

70

72

82

APPENDIX C

APPE DIX 0L) .

..

90

93

LIST OF REFPERENCES............................................98

BIBLIOGRAPHiY

..............

.....................

INITIAL DISt mRIBUTION LIST .....

100

..............

101

-'i

a=v

S-

",1,

IP 'I'

'

"

l~

'a,'

a ,'

Vi I

' I

-"

'

_-

LIST OF FIGURES
Figure 1

Siduwindul

Figure 2

AIM-9 Missile

Figure 3

Modeling

Figure 4

I'AL2 Program Exccution

F igure

MSC/PAL2

Missile ............

.............

System.... ..........

............

Techniques ............

Suggested

.............

.....

Oider

...........
of

13

Analysis

2]

rigure 6

MSC/PAL2 AIM-9 Geometry Model ..........

23

Figure 7

Combined Testing and Analysis ..........

33

Figure

Missile Modes for

Figure 9

Oingle-Rail AIM-9

Spring - Mass Model ..........

.....

37

.............

36

Figure 10

Typical Shaker Setup ....

Figure 11

AIM-9 Missile Modal Analysis Block Diagram

47

Figure

AIM-9 Missile Modal Analysis Test. Setup

48

12

............

Figure 13

EMODAL-PC lest Procedure ........

Figure

14

AIM-9 Missile Model Mode

Figure

15

AIM-9 Missile Model

Figure

16

Figure

1 Response

..

..........

43

56

84

85

AIM-9 Missile Model Mode 3 Response

86

17

AIM-9 Missile Model Mode 4 Response

87

Figure

18

AIM-9 Missile Model Mode 5 Response

Figure

19

AIM-9 Missile Model Mode 6 Response

viii

Mode 2 Response

88
.

89

ACKNOWLEDGUMENT

thu

following

whose

people,

First,

I would
foi

advisoi,

fifteen

like to thank Professor

Ed Wu,

guidance,

his

outlook on life.

confidence

through these last

helped me make it

abilities

and

support

patience

to

and appreciation

1 would like to expiess my gratitude

and

in

my

months.
my thesis

always

optimistic

has been a privelege working with you,

It.

and an even greater pleasure knowing you.

Secondly,
and

support

would

advice

liko

to

from Mr.

received

MacNeal Schwendler Corporation,

and

thank

acknowledge

the

of

the

Ken

Ranger

without whose help I might not

have been able to get as far on the AIM-9 model as I did.

Next,

I would like to thank Mi..

Air Tcst Center,

place,

Tim Twigg

fLion the Naval

who not only got me the missiles in

but also

obtained

much-needed

information

the first
about

the

missiles on the spur of the moment.

My deepest gratitude goes to my wife Maureen,
children,
for all.

Michael

and Meagan.

good cause.
Jove,

can probably

never make up

those times when I coulan't go where you wanted to go,

or do what you wanted to do,

I

and our two

but please believe it

was for a

You may never realize how important your constant

understanding and encouragement

mention the help with the typing.

have been to me,

not to

Thanks Lor being there when

I needed you.

ix

I'

-sI?

m r q

P q'" T '

1M

W ''I:

" " . I ,,,

" , r

n-

-"-

INTRODUCTION

1I.
A.

BACKGROUND
Over the last

has performed
(ASW)

25-30

its

years,

primary

with virtually

the 1-3 Orion patrol

mission

airplane

of Anti-Submarine

Warfare

no deuensive weap>onry to protect

itself

on

its

solitary,

During

its

long-range
first

15-20

carried armament
as "Zuni"

in

years

of

on the world's

existence,

rockets for ginund suppression.

capability.

It

airplane with ninimal,

However,

the Harpoon

in

the

anti-ship

platform,

the

platform.

Throughout

vulnerable

to,

r-3

and

the

1970's

missile

ud'decilly
all

late

was

became

this

time,

defenseless

it

if

any,

as well

offensive

and early

1980's,
into

offensive

has

against,

only

had been viewed

integrated
an

oceans.

F-3

the form of mines and torpedoes,

primarily as a patrol

when

missions

the

weapon

remained highly

any

and

all

air

threats.
""'he
of

the

years,
is
that

issue then becomes one of increasing the survivability

F-3

against

airborne

threats.

Defensive Air Combat Maneuvering

currently

being taught,

to train

would lengthen the P-3's

engagement.
realized that,

This training
if

an air-to-air

P-3

survival

During

the

(DACM)

has been,

aircrews

time in

has shown promise,

in

last

few
and

tactics

an air-to-air
but it

was soon

defensive missile were to be

illtefIg]-ated

int~o

muich bot~tori
*

weapuox

the

chance

tht3 V-3

invenitoLy,

ot !-'.ii-vi vinq,

would

have

not possibly dfe-st royinci,

i t

the air- thicat.

This

reaIi.-ationi
progi-lam,

integtatioii

the Naval Aii

Riverqj,

thu

tu

pO55~bo

Iin this effo.rt,

Fig-ur.es

nine,
inl

and

1 and 2,
to

['-3 .

defin-e

Wing

anld currLently Onl-going

stat

as

installed

system
ion

Side-w-indei

One- of the maini

at

Patuxentt

Sidewilider.
oin r-3 outer

requirements

nine

is

of

for

critical

allows the carriage

losing a Harpoon wing station.

Motor

S I rLMF Warlirad

Fiigrure 1

AIM-9

t-hu

of

imipor~tance to the teot. pr.ogram becauise it

of anl AIM-9 missile without

proof-oi-concept

NATC was ta.sked to evaluate

Iha
-t;t
i CL;

shown in

Ilse

P-3/A1M-9

at. Naval Air Station

(NATC)

9Ce'Z-11at ionl

station

the

inl 1989,

begunl

Tvst Center

Ma.ryland.

muissile,
wing

led

Missile

concerns with carrying AIM-9 miasiles was

whether or not the natural

vibr-ation inodes

and frequoncies

of

the missile system would interfere with the two low frequmeu12y,

kg.ure 2

AIM-9 Missile System

outer wing vibration modes of the P-3:

a Ground

This could be determined by conducting

17-22 hz.
Vibration

and the outer wing torsion mode at

(hz)

mode at 4.7-8 hertz

the outer wing bending

Test

(GVT),

the

results

could

of which

then be

incorporated into a computer model of the P-3 wing, leading to

a determination of whether or not a potential flutter problem
exists with an AIM-9 installation on a P-3 wing.
In

support

on-going program,

of this

Hollyer

[Ref.

1]

designed and constructed a GVT stand at the Naval Postgraduate

School.

A ground vibration

of the AIM-9

resulting in the development of a two

missile was conducted,

degree-of-freedom

characterization

and

model,

lumped-mass

experimental

determination of the missile's resonance modes in

pitch.

With the recent termination of the P-7A and P-3H programs,

the P-3 community is
to carry

likely

it

into the

have

now looking at the P-3 Orion II

2 1

beefed-up

analysis of this wing in

This

investigation

concurrently

"t

century.

wing

structure,

II

will

most

necessitating

an

conjunction with the AIM-9 missile.

responds

developing

The Orion

program

to

that

a mathematical

requirement

model

by

of the AIM-9

missile using finite element methods to analytically determine

its

modal parameters,

and setting up a modal test system to

verify the mathematical

model by experimentally

determining
1

the missile's natural frequencies,

response.

r4

IB

The analytical modeling,

mode shapes and transient

combined with the fully

instrumented

experimental

modal

testing

pursued

in

this

investigation will lead tc the following:

"* Full characterization of the dynamic ch'"racteristics of

the AIM-9 missile.
When combined with an appropriate
analytical model of the wing, these methodologies will
allow the analytical assessment of different missile
mounting locations and configurations without the risk and
expense of flight
tests.

"*Modification of the parameters in

the analytical AIM-9

model to analytically assess the changes in the modal
characteristics for the upgrades and modifications of the
missile.
Upgrades
and
modifications
may
include
propellent change in the rocket motor for range extension,
and modified guidance and control system resulting in a
mass change.
Again, this
is possible without the expenses
of prototype construction and testing.

a Characterization of other candidate missile systems (e.g.,

AGM-65 Maverick and AGM-88 HARM) for the P-3 and its
eventual successor.

B.

MISSILE MODELING TECHNIQUES

To

obtain

structure's

vibration

different methods can be used.

modeling which utilizes
in

The first

the finite

characteristics,
method is

two

analytical

element method, i.mplemented

a computer software program.

This is

analysis, whereby the structure is

a "forward-type"

of

partitioned into components

whose mass,

stiffness

structure's

natural frequencies and mode shapes are predicted

and damping are parameterized,

from these input parameters.

modeling,

is

The second method,

and the

experimental

usually referred to as modal analysis.

This is

an "inverse-type" of analysis,

whereby the actual structure is

subjected to

function,

a known

forcing

and the

structure's

response to the excitation is measured at multiple locations.

at

responses

individual

the

From

locations,

different

the

overall structure's stiffness and damping can be determined,

along with the natural. frequencies and modes of vibration.
block

of both

diagram

Figure 3

[Ref.

:p.1-

of

Finite
Element Analysis

Frequencynalyica
Response
Measuremeos

E ignvalue

Estimation

EFove_

Figure 3

C.

Model

Parameter

Faramplpr_

in

].

Modal Testing

shown

is

methodologies

these

M1dlModel

Li

"""a""

Modal

t '

Modeling Techniques

IMPORTANT DEFINITIONS
A compilation of terminologies

the

discussions

techniques.

of

the

finite

Definitions

of

is

presented to facilitate
analysis

element

and

modal

terms

are

available

the

textbooks of vibration analysis,

and the following listing

adopted from two of these books

[Refs.

* Modal
shape

parameters

was

3 and 4].

frequency,

natural

from

damping

and

mode

Natural frequency - a frequency at which a system vibrates

and then released; that frequency
when disturbed from rest
where,
in
theory,
a minute force can produce large
movements of the structure
* Resonance - occurs when a force coincides with the natural
frequency;
movement of the structure increases either
until
it
is limited by damping or non-linear effects,
or
until
something breaks
a

Mode - resonance or peak in

a Frequency Response :'unction

* Modal model - set of modal parameters

dynamic behavior of a structure

representing

the

* Modal mass - mathematical concept that can be thought of

as being proportional to the amount of mass actually in
motion in a given mode shape; it is not the physical mass
of the structure and cannot be measured
* Modal stiffness
- mathematical concept that
can be tho~aght
of as being proportional to the amount of stored energy in
a structure in
a given mode shape at the instant it
reaches the point of extreme motion
Damping resonance;

quantity that controls the sharpness

usually expressed in percent of critical

of

* Frequency Response Function (FRF) - ratio

of the response
(either
displacement,
velocity
or
acceleration)
at
coordinate i to the excitation at coordinate
j
(for
example,

* Measurement

, = X!/F,)
-

FRF acquired from a spectrum analyzer

Degree-of-freedom (DOF)
defined on a structure

- measurement point and direction

Driving point EIRE - measurement made with the excitation

and response at the same point and direction
*
Reference
transducer
transducer
connected
to the
reference channel (channel A) of the analyzer; normally,
this
is the force transducer
e Response transducer - transducer connected to a response
channel (channel B, C, or D) of the analyzer; normally,
this
is an accelerometer

-1VP~rlF'Im

ql'~

'

r""-

P'"nMIi

,=,,

..

""Mode shape

- a set of shape coefficients at different

locations along a structure that defines the structure's
deformation at a specific frequency

"*Real (normal) shape - purely real (or imaginary) numbers

that describe the vibration pattern of the structure; all
parts of the structure are moving either in phase, or 1800
out of phase, with each other

"* Complex shape - when each deformation value requires two

numbers (magnitude and phase, or real and imaginary) to
describe the vibration at a location; can have any phase
relationship between different parts of the structure

"*Curvefitting - matching a mathematical equation to a set

of data points obtained from an FRF measurement

Br

II.

A.

ANALYTICAL MODELING

FINITE ELEMENT METHOD BACKGROUND

In

designing

structures,

the

finite

element

often used to determine the structure's

and mode shapes before
foresee any potential
The finite

the

can

element

analysis

In

initial

the

general
location

of

leading to

the

The finite

and coded

programs

design,

the

the

configuration

final

to

and

can

Generally

block

and material
or at

one of the finite

on

will

the market.

have
the

mechanical
of the

confirmation

at

least

structure,

the

interfaces

and

structure

can be

of

changes

the

design of the structure.

also

and mode

are

an effort

building

mass

into

of

any

currently

designer

and

iteration

element
It

frequencies

blocks are known,

Constituent properties
with

frequencies

advance

components,

estimated,

designers.

in

of

attachments.

in

damping,
building

software

idea

built,

be viewed as

stiffness,

for the individual

estimated,

structures

actually

is

implementation problems which may occur.

whereby

properties

built.

is

element method

approach

least

it

natural

method

method is
be

used

shapes

of

speaking,

in

represented

by

which are connected by structural

not
to

just

determine

structures
the
discrete

usefil

which
finite

tool
the
are

element
grid

elements.

or
In

for

natural
already
method,

node points
other words,

the

finite

simulate

of

representation
its

can

method

element

continuous

structural

forces or loadings.

behavior

be

as

viewed

system
and

made

in

response

discrete

order

to

to

expected

in the form of a

This representation is

mathematical model consisting of discrete elements connecting

The more node points that are chosen

discrete nodal points.

to

represent

will be.
chosen,

the

Conversely,
the stiffer
or

designer
structure,

analyst

the

fewer the number

the model will be.

is

detailed

of node points
the aim of the

If

stress

analysis

of

the

such as around attachment points

and around the location of applied forces.

aim is

solution

then the nodal density must be increased in regions

stress gradients,

of large

accurate the

the more

structure,

a deflection shape analysis only,

be used. [Ref.

5:Section l,pp.

However,

if

the

then fewer nodes may

2-2,2-3]

For either an existing structure or one that has yet to be

built,

following

typical
[Ref.

finite

element

5:Section l,p.

analysis

requires

the

2-1]:

"*Nodal point spatial locations

"* Structural

elements that connect the

nodal
representing the stiffness of the structure

* Mass properties of the structure

* Boundary conditions or structural constraints
* Static and dynamic load application

10

points,

The mathematical model that results from the analysis can then
be used to determine the structure's natural frequencies, mode
shapes,

stress levels and displacements.

An added benefit of

conducting a finite element analysis on a particular structure

is

Fo

and faster than by altering the actual hardware.

if

cheaper

that design iterations can easily be accommodated,

example,

the designer or analyst wished to determine the effect of

additional mass on the structure's natural frequencies,

that would be required is

material type,

damping,

easily

handled

The effect of changing the

as

shape or size of the structure can be

well.

However,

care must be taken in

model,

structure's geometry,
MSC/PAL2

B.

modifying a few lines of software

and rerunning the analysis.

code,

all

to

achieve

an

accurate

the assumptions made about the

material properties and so on.

FINITE ELEMENT ANALYSIS SOFTWARE

Introduction

1.

The finite element analysis software package used in

this

MSC/PAL2

Advanced

Version 4.0.
for a

Stress

and Vibration

MSC/PAL2 uses the finite

structure's

displacement,

different points on the body.

of both static
it

Corporation's

investigation was the MacNeal-Schwendler

of

element method to solve

forces

and/or

stresses

and dynamic analysis of a structure,

node

points,

Software,

at

The software has the capability

limits the number of degrees

number

Analysis

for

11

of

freedom,

certain

although

and hence

dynamic

the

analysis

In addition to tabular output of nodal locations,

techniques.

element connectivity and nodal deformations,

the software also

has the capability to produce graphical output for analyzing

model

geometry,

structural

deformations,

functions of time and frequency,

displacements

as

animated deformation plot:,

and stress contour plots.

MSC/PAL2

requires

computer with at least

an

80386

(or

a 20 megabyte

(Mb)

least 2 Mb of random access memory,

space

being

requirements
3.1

required
include

(or higher)

very

personal

hard disk and at

with additional hard disk

large

Other

problems.

at least one floppy disk drive,

operating system,

and either a CGA,

2.

for

80486)

EGA,

PC DOS

an 80387 math coprocessor

or VGA monitor.

Basic Program Operation

MSC/PAL2

Version

4.0

is

menu

driven

program,

beginning at the top level with six main menus.

Each of these

menus

of submenus,

then branches

out

into varying

depending on the application chosen.

these menus

is

shown

in

Figure

4,

numbers

The interrelationship of
and is

addressed

in

the

discussion that follows.

a.

File
The

selection,

file

menu has

"Configure",

only two

allows

choices.

the user to change

drive destination to which the database files

It

is

The

first

the disk

will be saved.

primarily used to store databases on alternate drives

12

[II
AVNE

BUILD MODEL

DYNAMICS

STATICS

PAL2 Program Execution

Figure 4

B,

D)

for very large problems.

"Quit",

is

the command used to exit

(for example A,
selection,

The second
the program.

Edit

b.

The Edit menu likewise has only two choices.

first

choice,

edit

an

editor.

text

existing
This is

file

the

utilizing

and the load file(s)

The

second

choice,

which are utilized

is

"MSC-MOD",

an

program,

for the MSC/PAL2

13

create or

software's

the option used to create and edit

file

preprocessor

allows the user to view,

"Text file",

in

The

text

the model

the analysis.

interactive
enabling

graphical
the user to

build and edit two-dimensional and three-dimensional

element

models representing

actual

finite

Instead of

structures.

wciting software code in a text file, rISC-MOD uses a menuing

system overlaid on a graphical display which has the advantage
of letting the user observe on the display the results of a
pcrticular command.

MSC-MOD can be used in place of MSC/PAL2

to create a PAL2 wodel,

or it

can be used to edit an existing

PAL2 model.
c.

Analysis
The Analysis menu serves the purpose of providing

commands

for

both

building

the

model

from

the

previously

defined model text file, and also performing the actual static
and/or

dynamic

aralyses

using the

file(s).

The "Build Model"

model

the

of

consists
equations

of
of

mathematical
system

structure
a

set

of

motion,

previously

defined

in

the

second-order
representing

when

solved,

model

ordinary
the

give

displacements and element stresses.

formed from the mass,

load

submenu generates a mathematical

mass

expressions for the structure.

equations,

defined

the

file,

which

differential
and

stiffness

These generated
structure's

nodal

Since these equations are

stiffness and damping of each element

making up the structure,

and are specified for each nodal DOF

in the system, matrices are employed to help keep track of all

the different equations.

The system stiffness matrix is built

up from the stiffnesses of each element,

which are dependent

14

F_

'l~

',

lm

-!''

r|l:lrl~

r~i:

on the size,

shape,

type and material make-up of each element.

The element stiffnesses

command,

are defined via the PAL2 "3EAM

from which

moments

of

cciiputed.

inertia,

beam

and

cross-sectional

toisioiial

area,

moment

of

The system mass and damping matrices

by the program in
originating
"MATERIAL

the

an analogous manner,

originating

from

ELEMENT".[Ref.

the

conunand,
PAL2

5:Section l,pp.

aie

are computed

with the mass matrix

and

commands

bending

inertia

from the mass density specification

PROPERTIES"

TYPE"

the

in

the PAL2

damping

"DAMPER"

and

matrix
"DAMPING

2-27 through 2-30]

The second submenu under the Analysis heading is

the "Static Analysis" menu option.

Static analysis is used to

determine the stiffness and element stresses of the structure

when

subjected

static

to

load file.

(built

using

some

static

load,

The static

the

"Build

Model"

applied to the same model,

In

fact,

if

submenu

interest

have

been

defined

the

run,

analysis

and

the

without building the

can be

After all
the

command)

"Data

run repeatedly

of the loading cases

Recovery"

invoked in order to obtain element stresses. [Ref.

p.

in

Additional loads can be

desired,

without rebuilding the model.

of

is

loads are applied to the model

program computes the displacements.

model again.

which

option

io

6:Section 2,

3-8]
The third submenu under the Analysis heading is the

"Dynamic Analysis" menu option.

a similar manner to static

Dynamic analysis proceeds in

analysis,
15

although multiple load

conditions
Analysis"
user

cannot

run

same

the

a structure' s

model.

natural

The

that

Analysis"

option) ,

and mode

frequencies

the

"Transient

dynamic

analysis

Analysis"

run

modes

d.

option)

to

the

assess

loads

Dynamic
the static

should be the

analysis

of the structure. [Ref.

characteristics

time-varying

should be run only after

and the normal

analyois,

to

responses

Response

of a structure

analysis

responses

(using the "Frequency Response

forces

and/or

"Dynamic

allows the

it

(using the "Normal Modes Analysis" option),

to frequency-dependent

(using

on

very flexible in

menu option is

to compute

shapes

be

overall

first

dynamic

5:Section l,p.

3-4]

Graphics
The Graphics menu of MSC/PAL2 has two submenus from

which to choose:

two ways.

bc used in

prior

geometry,
certain

"View"

that

elements.

to

all

First,

model

deformation

to

deformation.
option.

The

The "View" option can

be used to chtck xuud.l

can

the

analyzing
are

points

model.

properly

to

make

connected

with

accomplished using the "F2:

option from the first-level

after

it

actually

node

This is

and "XYPlot".

has

"View"
been

_.ome

built

menu.
to

"View"
view

including

load,

Get Dataset"

the

accomplished using the "Fl:

"View"

option

used

to

also used
structural

animation

This is

is

is

of

the

Get Model"

view the

graphical
b

results

of

both

static

deformed and undeformed

and

dynamic

cases.

16

analyses,

in

both

the

The second option under the Graphics heading is the

"XYPlot" option.

This option is primarily used to graphically

represent frequency and transient response results in X-Y plot

format,

and provides

function of time,

graphical

displays

of

response

as

frequency, element number and subcase.

the "View" option,

"XYPlot"

is

Like

menu-driven and uses the ten

function keys to select the different plot options.

e.

Composite Capabilities
The Composite menu is

the

Analysis

section,

used

to

a pre- and postprocessor to

cr tate

composite

laminate

descriptions for plate material properties and for printing or

plotting composite results.

Laminates with up to 200 layers

made from up to ten different materials can be made.

the

AIM-9

section

missile

of

the

investigation.
[Ref.

It

is

of

composite

software

was

construction,
rot

used

this

in

this

mentioned here simply for completeness.

1-1]

Advanced Capabilities
The

capabilities
following.

Advanced
of

the

menu

MSC/PAJL2

presents
software,

some
and

additional
includes

the

The "Status" menu option shows whether the DOFs of

current

"Equations"
stiffness

not

MSC/PAL2

6.Section 3,p.
f.

the

is

Since

model

are

active,

eliminated

menu option prints out

and

damping

matrices

for

created by the "Build Model" option.

17

or

the terms
the

most

zeroed.

The

of the mass,
recent

model

The "Expanded File" menu

option provides expanded MSC/PAL2 model data sets,

significantly
they

were

larger than the original data sets from which

created,

connectivity
CON-VERSION"

are

since

all

explicitly

nodal

points

specified.

and

The

element

"MSC/NASTRAN

menu option translates an MSC/PAL2 model dataset

to MSC/NASTRAN
which is

which are

format.

Finally,

the "REPLAY2"

a separate program from MSC/PAL2,

continuously

replay text,

slide-show format.

menu option,

allows the user to

picture and animation files

in

Since the MSC/PAL2 program was only used

in this in-estigation to build a model file

and obtain natural

frequencies and mode shapes, these advanced capabilities were

not utilized, and are mentioned here simply for completeness.
[Ref.

6:Section 4,pp.

3.

1-1 through 6-1)

MSC/PAL2 Analysis Procedure

The anplysis of a generic structure using the MSC/PAL2

finite element software program can be broken down into three

general

steps.

process

in

which

mathematical
elements.

The

first

the

model,

step,

actual

defined

system definition,

structure
by

node

is

origin
nodal

LOCATIONS"

locations
commands,

(Young's modulus,

system chosen.

are
the

defined

struzture's

with

relative to the

the

"NODAL

material

of the
POINT

properties

shear modulus, mass density, Poisson's ratio

18

A
I-

and the

After all

using

the
by

connected

The analyst chooses a coordinate system,

of the coordinate
point

simulated

points

locations of all node points are specified

is

and tensile

yield stress)

PROPERTIES"

command.

specified,
choices
elements,

types

several

quadrilateral

hexahedral,

the

and

element

are designated.

types

pentahedral

Following

of curved

triangular

or

plate

tetrahedral

type

are

joined

boundary

by

each

using

element

defined,

(for

the

"OFFSET CONNECTION"

example,

elements,

and

elements.

the

element

"DO CONNECT"

and/or

model.

Finally,

the

points)

are

attachment

"ATTACH

AT"

and

After the model has been fully

menu option is

engaged to assembla

at which time the model

is

ready

for

and/or dynamic analysis.

The second general step,

which

different

structure's

loading

static

Static

and

model

via

dynamic

loads

separate

AND MOMENTS APPLIED",

analysis,

conditions

and/or

dynamic
are

load

"DISELACEMENTS APPLIED",

are

the process

applied

behavior

applied
files,

is

to the
using

conducted

Analysis"

using

submenus,

is

among others,

the

"Static

the

computed.

the

commands
and "FORCES

and the resulting nodal

Analysis"

described previously.
19

and

in

alrezdy-built

"ACCELERATIONS APPLIED",

displacements and element stresses are computed.

is

beam

is

"ELIMINATE",

commands.

the "Build Model"

the

in

"ZERO",

the system equations,

static

straight

solid

specification

type

commands to specify the nodal points that

conditions

specified

Element

and

connectivity process, using the "CONNECT",

"GENERATE CONNECTS"

"MATERIAL

Once the material properties have been

the element

include

are specified using the

The analysis
and

"Dynamic

the

in

process

which

the

static

used

to

structural

show

the

structure's
and

deformation
option

"Graphics/XYPlot"

function

of time or

previously described in

be

can

used

consists

of

These options

mode shapes,

were

also

element analysis,

shown in

This breakdown,
If

process.

the

Figure
is

analyst

complete with frequencies,

stresses and deflections caused by a variety of

conditions,

different

loading

completed

in

analyst is

interested in

natural

nodal

steps can be further broken down into a

conducting a full-blown analysis,

rat

plot

to

detail.

ten-step

The

and also element stresses, as

frequency.

more specific plan of attack.

5,

geometry,

animation.

For the conduct of an actual finite

these three general

are

results

undeformed

deformed

displacements and accelerations,

a

dynamic

and

The "Graphics/View" option can

presented in graphical format.

be

is

general step, graphical postprocessing,

The third

its

if

entirety,

frequencies

then

the

possible.

process
If,

only certain parameters

and mode shapes),

should

however,

be
the

(for example,

then only those

steps

relevant to that particular analysis need be completed.

4.

MSC/PAL2 Limitations
The maximum problem size for this

dependent upon how many DOFs are in

type
model,

of DOF.

Initially,

where G is

there

version of PAL2 is

the model,

are

G global

as well as the
DOFs

six times the number of nodal points,

20

in

each
since

RUN NORMAL
MODES ANALYSIS

CREATE MODEL
FILE

VERIFY GEOMETRY
WITH VIEW OPTION

PLOT MODE SHAPES

WITH VIEW OPTION

RUN STATIC

VERIFY FREr ENCIEc

ANALYSIS

SIM ..JIED
ANALYSIS

IWITH

PLOT STATIC

APPLY TRANSIENT

RESPONSE WITH
VIEW

AND FREQUENCY
LOADS

RUN TRANSIENT

VERIFY RESPONSE

AND FRPEQ UEC

WITH SIMPLIFIEL)I

ANALYSES

ANALYSIS

Figure 5

nodal

each

MSC/PAL2

Suggested Order of Analysis

point's

motion

and three

translations

system

chosen.

zeroed

DOFs

The

DOFs

are

the

coordinate

into

partitioned

"Build Model"

three

by

command

and the

(created with the

active DOFs.
21

in

the

into

"Build Model"

With this

axis

"ZERO")

The free DOFs are further partitioned

eliminated DOFs

"Eliminate")

global

represented

be

about

rotations

(created with the

and the free DOFs.

the

can

command

mind,

the

problem

size

limitations

software are as follows

"*Build Model:
* Statics:

for this version

[Ref.

2000 nodes,

5:p.

A-2]:

12000 DOF

Normal modes
Subspace iteration:
QR method:

12000 active DOF

225 active DOF

Jacobi method:

a model

MSC/PAL2

12000 free DOF or 225 active DOF

"*Dynamics:

If

of the

225 active DOF

Transient response:

225 active DOF

Frequency response:

150 active DOF

has

more

active DOFs

than

shown

here,

and the

analyst wishes to conduct a detailed analysis, then the number

~f
static

t-i

~fl~F'
mst
e

analysis)

-~~~ued

iastatic

or Guyan reduction

MacNeal-Schwendler

[Ref.

some detail the procedure

5:Section

-ondensa-tion

(for

(for dynamic analysis).

l,p.

2-43]

addresses

in

for accomplishing this reduction.

Since the main thrust of developing the AIM-9 model was to

analytically determine the missile's natural frequencies and
mode shapes,
used,
C.

the normal modes subspace iteration method was

and no reduction was necessary.

MSC-PAL2 SIDEWINDER MISSILE ANALYTICAL

1.

MODEL

Model Geometry

The MSC/PAL2 software code written to define the AIM-9

geometry is

shown in Appendix A.

22

The actual missile geometry

resulting from this

code is

shown in

Figure 6,

both with the

hidden elements showing and with them not showing.

code,

it

body,

of 706 node points w'-re used

can be seen that a total

to define

this

structure:

36 for the front fins,

From the

610 node

points

for

the missile

and 60 for the rear wings.

This

large number of points was required to describe the missile

Figure G

MSC/PAL2 AIM-9 Geometry Model

23

for

several

segments

reasons.

1irst,

instead of using

curved beam

(which cannot be graphically displayed in

software) to make up the cylindrical missile body,

rings

of

16

nodes

joined

with

small

the PAL2
concentric

quadrilateral

plate

elements were spaced three inches apart to define the missile

body.

Secondly,

these rings were spaced this

of the recommendation by MSC

(length/width ratios)
to one.
node

Thirdly,

points

frequencies
built

is

in

7]

to avoid aspect ratios

the plate elements larger than four

a fairly
required

and mode

[Ref.

closely because

uniform
to

shapes.

spatial distribution

properly
Since

the

compute
missile

using the actual dimensions of the missile,

was 35 concentric

of

resonant
model

was

the result

rings of node points making up the uniform

cross-section of the missile body and three concentric rings

of node points making up the tapered nose cone,

all

into

the missile.

one

node point

representing

the

nose

of

tapering

Quadrilateral plate elements of 0.25 inch thickness were used

to define the missile body and the tapered nosecone,

while

triangular plate elements of 0.25 inch thickness were used for

the tip

of the nose cone and the back plate.

The three points

where the AIM-9 missile attaches to the LAU-7 launcher rail

were modeled by zeroing the displacements
node points.

Translation in

the x,

y,

at the applicable

and z directions and

rotations about the y and z axes were zeroed for node points
85,

213,

and 373.

Rotation about the x axis was not zeroed,

24

since motion in

the pitch direction is

the primary motionof.

interest.
After the missile body was constructed, the front fins
were added using a combination of quadrilateral and triangular
plates,

each of 0.10 inch thickness, with great care taken to

keep the individual plate aspect ratios at less than four to

one.

Connecting the front fins to the missile body proved to

be a problem because of the fact that the AIM-9's front fins

are

actually movable

rotation

which exhibit

canards

The PAL2

their attachment points to provide missile guidance.

has

software

arrangement,

no

for

provisions

this

modeling

about

type

of

so the fins were attached to the missile body via

thereby prohibiting the rotation of

a small 0.35 inch plate,

the fins in the x-y plane about their attachment points.

This

resulted in a stiffness problem, with the fins rotating about

the

z-axis.

Additional

work

is

required

in

area to

this

reflect a more representative attachment method.

The rear wings were the next building block added to
the inodel.

These wings were modeled using a combination of

0.5 inch thick quadrilateral

and triangular plates,

and were

rigidly attached to the missile body using previously defined

missile body node points.
All that remained to complete the missile model was
the

addition

warhead,
of the

of

mass

parameters

to

represent

the

motor,

target detector and guidance and control components

AIM-9 missile.

This

was
25

accomplished

using point

of each component's

masses located at the midpoint

of

exception

the

with

component,

motor

the

was

which

70 inch length.

represented by two point masses because of its

Finally,

length,

the material composition of each element in

the model was defined using generic steel values of Young's

modulus,

modulus,

shear

Poisson's

density,

mass

ratio

and

tensile yield strength for the missile body and front fins,
and

aluminum

generic

material

properites,

element

connectivity,

along

with

rear

the

nodal

are used to generate

These

wings.

the

for

values

and

locations

that,

equations

when solved, give the system displacements and stresses.

The completed AIM-9

missile model

consisted

of 706

a total of 4221 DOFs.

nodal point locations, which re3ulted in

The model was constructed of 717 elements of three different

types:

660 quadrilateral plate elements,

52 triangular plate
Quadrilateral plate

elements and five point mass elements.

elements were used as much as possible because they are less

the

than

stiff

triangular

better and more accurate results

2.

[Ref.

thereby

5:Section 2,p.

giving
2-3].

Determination of Modal Parameters

Upon

verification
using

elements,

plate

the

completion

of

of the nodal point

"Graphics-View"

the

model

locations

option,

the

geometry,

and connectivity

"Build Model"

option was invoked to form the system stiffness,

mass matrices.

and

menu

damping and

These matrices arise from Newton's Second Law

26

Sr

, , q,",

! ' !" .

..I

' I ' I' I

of Motion,
a

upon which all

second-order

vibration work is

differential

based.

in

equation

the

Written as

time

domain,

Newton's Second Law can be written as follows:

Mx(t)

+ Ci(t)

+ Kx(t)

= f(t)

(1)

where M = system mass maLrix

C = system damping matrix
K = system stiffness

matrix

x(t) = nodal displacement vector

x(t) = nodal velocity vector
(t)

= nodal acceleration vector

f(t)

= applied force vector

a rnormal modes analysis was

After the mode] was built,

conducted to determine the missile's

written in

include

the

system

stiffness

mass

and

matrices,

matrix notation as follows:

Mk + Kx = 0

The

right-hand

since
motion

there
is

is

side
no

of

this

forcing

assumed to be

is

equation
in

steady-state

the AIM-9 missile,

frequency,

f,

with each point moving either

acceleration

out

vector

of
may

phase.
be

equal

motion.

vibrates
in

in

zero
All
The

at some

phase with the

Therefore,

written

to

analysis.

harmonic

this

or

set

modes

normal

in

points

case

(2)

structure,

other

and

The system equations formed for a normal modes

mode shapes.
analysis

natural frequencies

the

terms

of

nodal
hhe

displacement vector as follows:

x= -(2tf)

2x

27

(3)

Im
Substituting this into Equation 2 gives:
[K

(2-7f)

M]x

In the case of the AIM-9 missile,

and stiffness matrices
4221 active DOFs.

is

(4)

the dimensions of the mass

4221 x 4221,

since the model has

One can immediately see the necessity of

using a computer program for this type of analysis,

as the

manual manipulation of this many equations would be virtually

impossible.

The nontrivial solution to Equation 4 yields the

system resonant frequencies,

it

each of which has associated with

a characteristic vibration, or mode shape.

no forcing in a normal modes analysis,

each mode shape are arbitrary,
important.[Ref.

Since there is

the scaling values for

and only the relative shape is

5:Section l,p. 2-40]

A normal modes analysis can be accomplished by three

ditterent

methods.

extraction,
method,

The

default

the QR method,

which

structures.

is

a more

Both

of

is

these

typically

commands

only

the

method

methods

for

require

This reduction is

"Eliminate"

lowest

eigenvalue

unrestrained
the

system

size to the point where there are

no more than 225 active DOFs.

with the PAL2

of

much faster than the Jacobi

accurate

equations to be reduced in

method

few

and

modes

accomplished

"Activate".
are

excited

Since
to

any

appreciable degree by dynamic forces,

these are the modes of

primary interest to the designer or

analyst.

iteration method is
frequencies

The

subspace

the method of choice when only the lowest

are needed,

because

it

accomodates

up to

12000

28
l

p' '

'r"'

, ' ,

Il lr tiT1r}S

.''

" . . . .

"

active

and

DOFs

This

required.

reduction

no

the

of

for

solves

method

4,

matrices in

these equations

obtaining

is
and

eigenvalues

the

eigenvectors with the mass and stiffness

of Equation

equations

system

the form

from the physical

properites of the structure coded into the model,

and was the

method used for conducting a normal modes analysis of the AIMof this

The results

9 missile model.

analysis

are given

in

Appendix B.
Modification of Modal Parameters

3.

The modal parameters of the AIM-9 model,

the

frequencies

natural

and

mode

may

shapes,

simply

by

the

altering

mass,

readily

or

stiffness

In the construction

damping properties input into the model.

of the missile model,

be

of the missile

modified to reflect alternative arrangements

configuration,

specifically

diftlerent configurations were

various

analyzed to determine the effect of changing these properties.

The

of the

results

five

It

Appendix C.

discussed in

cases

analyzed

are presented

and

can be seen from these results

that the addition or movement of the point masses can have a

effect on both the natural frequencies

substantial
mode

Adding

shapes.

resonance
stiffness,

to

lower

while

higher frequency,
helpful in

mass

to

the

frequency,

removing

mass

shifts

the

decreasing

the

structure

as

does

shifts

the

resonance

as does increasing the stiffness.

avoiding a known excitation.

and the

to

This is

Moving the location

29

Fl
r

'1!

,-,

l|I!!
-

ll!

"'

'

r"r

'

'

'

.
--

-,

of

point

mass

only

short

distance

completely changing the mode shapes.

p-Luperties

of

response.

Since these

parts

of

the

the

effect

of

Changing the material

structure

results

had

also

affected

the

were obtained from a normal

modus analysis in which only the relative shape of tile mode is

important
However,

and not the

damping

would

amplitude,
need

damping is

to be

not addressed.

increased

reduce

t.o

the

response at. a particular resonance.

Knowing the effects of changing the various properties
of a model
different

enables

the designer

design iterations

or

analyst

tv investigate

or feasibility studies prior to

making any hai dware change6.

For example,

if

ascertaining

the effect of adding more fuel to the AIM-9 missile to extend

its

range was desired,

model LU represent

mass would be added to the existing

the increoseci tuei required, and additional

node points would be added to represent the increased length

of the fuel tank component.
then be run,
obtairied.

A new normal modes analysis could

and the new natural frequencies

2Alother

possible

missile

incorporating a new type of seeker head.

potential

upgrade

includes

The effects of this

modification could also be observed quite easily.

For these studies to be of value,

baseline model accurately
just geometrically,
this reason,
and

and mode shapes

analysis

it

is

but in

it

is

represent the actual structure not

the modal responses as well.

important to utilize

approach

to

imperative that the

first
30

obtain

For

the combined testing

the

missile's

modal

parameters experimentally, then modify the analytical model ,to.match the experimental results.

It

is

at this point that the

design iterations or feasibility studies can be conducted on

the analytical model with a reasonable amount of confidence in

the results obtained.

II.A

31

EXPERIMENTAL MODAL ANALYSIS

II.
A.

GENERAL
Iii the same vein that the finite element method can be

considered a "forward-type"
constituent

properties

in that a structure's

analysis,
input

are

and

the

response

is

predicLed, modal analysis can be considered an "inverse-type"

analysis,

that a structure's response to an excitation is

in

measured,

and

from

resulting

the

transfer

the

function,

structure's modal parameters can be obtained using analytical

estimation

parameter

The

methods.

n.ajor

conducting a modal analysis of a structure is

mas-sprn-da.pcr

can

modclo

be

advantage

of

that simplified

co-nzstructed

g oo

degree of certainty from the actual measured response of the

structure.

The

drawback

of modal

testing

is

that

design

iterations or feasibility studies cannot be accomplished

unless the structure is
physically modified.
The best
approach is

to combine the two methodologies,

modal analysis

of the test

structure

to

by conducting a

obtain the

actual

response, and then u.e this information to make the PAL2 model
more

accurate

combined

for use

testing

diagram form in

r!

in

future

and analysis

Figure 7

[Ref.

32

design

approach
2

:p.

3-4].

iterations.
is

shown

in

This
block

MODAL
TESTING

rIrITE
ELEMENT

MFASUntMFNTSMOE

PARAMETERL
TIONESCRD~~MC

Fon~

MODALj

EN ALUE~

MODAL~lA

FigureER
7 CmbneATstnganIAaRSi
St~~~ucturaODE
Mesrmnt Sses
In.
[e.
:.
defines
modal analysis a
ALthe
TpONeso
hrctrzn
dynamic
properties of an el
LAstcstutueintrm
modes
~virto.A
~~ ~

~~~~SNHE~
of h
aoiyo
'MO3L

rsn

12
h
t

oa

analysis practice,

as well as

mentioned previously,

described

linear differential
Mx(t)

If

system

dimensional

has

n-DOFs,

by a

then

the

X(s)
is

in
It

is

the transfer

of

= H(s)

modal

are

n-

function form

resulting in

the relation:

F(s)

(6)

matrix,

frequency,

important to note that

which

can then be

damping

and

modal

each row and column of

function matrix contains the same modal vector;

to identify

provided

the

assumiption
described

is
by

assumption

is

all

the modal

following
that

the

linear
that

parameters

assumptions
motion

of

second-order

symmetry

Reciprocity

Therefore,

only one

matrix needs to be measured in

row or column of the transfer

Maxwell's

vectors

These system equations

multiplied by some component of itself.

order

as

(5)

equivalent transfer

cal~ed the transfer

terms

= f(t)

above

and the matrices are nxn.

where H(s)

analysis

set of simultaneous

+ Kx(t)

by taking the Laplace transform,

vectors.

element

equations of the form:

+ Cx (t)

are transformed into their

written

finite

the motion of a physical system can be

assumed to be adequately
second-order

in

met.

The

the

structure

equations.

which

structure,

are

of motion

Theorem,

of a

first
can

The

exists,
implies

second

described
that

be

the

by
FRF

measured between any two DOFs is independent of which of them

is

used

for

assumption is

excitation
that

or

response.

third

and

final

no more than one mode exists at each pole

location of the system transfer

matrix. [Ref.
34

rI1

The

8]

Modal
reasons.
help

analysis

is

conducted on a structure

Perhaps one of the most important is

improve

Experimental

the

accuracy

measurements

analyzed result in
mathematical model.

of

finite

obtained

a physical

on the

to

match

models.

structure

check of the accuracy

being
of the

If the model predicts the same modes and

considered to be accurate.

modified

to verify and

element

frequencies as those obtained experimentally,

is

for various

the

Otherwise,

same

then the model

the model must be

response

as

that

obtained

experimentally, assuming the experimental response is correct.

Other important reasons for doing modal testing and analysis
include troubleshooting noise and vibration problems,
evaluating fixes made to structures in
and vibration problems,

quickly

order to solve noise

and formulating dynamic models

for

parts of structures that are too difficult or time-consuming

to

model

analytically.

In

this

investigation,

modal

analysis test setup was completed to be used to analyze the

response

of

the

structure

of

interest

(AIM-9

conjunction with another structural member

B.

missile)

(P-3 outer wing).

PREVIOUS WORK ACCOMPLISHED

Test Setup

1.

Work was begun on this effort by Hollyer [Ref.

designed

and built

the

GVT

stand,

which was

then

1] who
used to

conduct a vibration characterization of the AIM-9 missile

pitch

only),

employing

forced

35

I;1

in

oscillatory

inputs

over

(in
a

frequency range of 0-50 hz.

instrumentation

In this test,

Hollyer used analog

(vice a digital spectrum analyzer) in which a

shaker assembly was utilized in

conjunction with a function

generator to excite the structure

sinusoidally.

To measure

the missile's response, two piezoelectric accelerometers were

used,

one mounted rigidly in

series with the shaker assembly

(the reference transducer) and one roving over the missile to

read magnitude
missile
2.

and phase

shift at

various

locations

on the

(the response transducer) using an oscilloscope.

Previous Results
Using

the

above

apparatus,

the

missile

system

was

excited in pitch, with the excitation force varying from 10 to

40 pounds peak-to-peak.
of

24.4

to

25.6

hz

Pitch modes were observed over a band

and

near

34

and a

heave

mode was

elastic mode of the missile,

The first

observed near 38 hz.

hz,

obtained by placing the shaker at the node of the first

mode,

appeared to occur at about 52.5 hz.

However,

rigid

this mode

was terribly polluted with the motion of the GVT stand itself.
Exact

frequencies

inaccuracies
results

of

could

the

not

data

from this test

primary thrust of this

be

determined

collection

are

shown

in

equipment
Figure

built

and stiffened

direction only.

Even so,

for excitation

in

the
The

Since the
the natural

pitch,

the GVT

the longitudinal

the system was excited in

36

to

used.

8.

effort was to determine

frequencies and mode shapes of the missile in

was

due

sway and

yaw,

with a sway mode observed at about 7.2 hz and a yaw mode,

at about 9.8 hz.

However,

these modes are very suspect due to

significant motion of the GVT during excitation.

more accurate missile response in

sway and yaw,

need to be stiffened laterally.[Ref.

To obtain

the GVT would

1]

Cl,
I'-

W
C-,
L-

N,

C) MISSILE AT 24 4 HZ
*D MISSUILE AT 38 HZ

<

MISSILE AT 52 5 HZ

Cr
20

'20

100

80

60

40

POSITION AFT OF THE NOSE OF THE MISILE

S/!NCHES)

Figure 8

Missile Modes for Single-Rail AIM-9

After

the

above

modes

were

[Ref.

obtained,

the

1]

missile

system was modeled as the simplified spring lumped mass model

shown in

Figure 9.

This model was represented as a two lumped

mass system possessing no internal damping,

to provide

the

two

degrees

of

freedom

resonance of the pitch and heave modes.

with two springs

to model
The theory,

thought process beiiind this

model can be found in

in

[Ref.

Hollyer's investigation

37

1].

the

lowest

logic and

more detail

VERTICAL
4 X4.
+X

mr=4.82

9
x=1.599

ft

It
It

x=1.599

slugs
m=4.82
OO[b/ft

k=274

SP

-r=2.43

ft

Spring - Mass Model

Figure 9

slugs

Ir=2.43

(Ref.

it

1]

MODAL TEST SETUP CONSIDERATIONS

C.

Transducer Selection

I.

Prior

to

conducting

modal

test,

co.,siderations must be taken into account.

most

important
A

transducer.

is

considerations
transducer

is

the

device

multitude

of

Perhaps one of the

selection
which

of

converts

a
the

vibratory motion of the structure into an analogous electrical

signal.

SI

.;J-I,

velocity

accelerometer is
velocity

come in

Transducers

three basic types:

transducers

and

accelerometers.

by far the most common and is

transducer

in

the majority of today's

38

displacement
The

replacing the
applications

because of its
the

most

superior phase and frequency response.

common

piezoelectric
transducer,

accelerometers

accelerometer,

in

which

use

is

today

a small,

is

the

lightweight

useable over a wide frequency range.

output of a piezoelectric accelerometer is

One of

Since the

a very low voltage,

a charge amplifier must be used to amplify the signal to a

high enough level to be displayed by the spectrum analyzer.
A

recent

development

in

circuit-piezoelectric

accelerometers

(ICP)

concept,

in

is

the

integrated-

which amplifiers

in

the form of microelectronic integrated circuits are built into

the

transducer.

amplifier,

This

precludes

the

need

for

but a constant current power supply is

charge

required to

operate an ICP transducer.

Considerations

in

selecting

an accelerometer

for a

frequency range,

size,

specific project include sensitivity,

and long-term stability
properties

is

[Ref.

important

for

tradeoffs

may

have

to

properties

may

have

to be

usually

selected

frequency

to

9].

be

have

If
a

made,

particular
and

sacrificed.
a

one or more of these

some

of

the

Accelerometers

sufficiently

so that the frequency

application,

high

other
are

resonance

response can be considered

flat throughout the operating frequency range.

Accordingly,

the accelerometer reference frequency should be at least five

times

the

maximum

frequency

of

interest

[Ref.

10].

By

following this rule, slight variations in resonance frequency

due to different mounting methods can be ignored.
39

Ira

2.

Transducar Mounting
selecting a transducer

in

The primary consideration

mounting method is that the mounting method does not introduce

any impedance into the system that would result in
the generated data.

errors in

Various mounting methods are available,

including stud mounting, wax mounting, cementing and hand-held

probes

and

magnets.

isolated

stud

attached

to

provides

the best

The

mount

preferred

whereby

the structure

the

via

frequency

method

is

to

transducer

is

physically

screw stud.

range,

accuracy

use

This
and

an

method

long-term

stability, and problems with dust, moisture or temperature are

normally

not

encountered.

However,

this

is

somewhat

permanent method which modifies the test structure.

test structure is
not

desired

alternative

one in

(for
is

and transducer movement,

the

which drilled and tapped holes are

example,

beeswax.

If

missile),

the

next

best

This method allows for quick setup

while still

frequency range and accuracy.

providing for a moderate

Manufacturer testing [Ref.

11]

has shown that frequency response equivalent to stud mounting

can be achieved
surfaces.
can

be

with a thin coating

If the surface is curved,

used,

which

contoured surfaces.

Fi+

easily

Cementing is

be

flat

a composite mounting base

sanded

for mounting

to

sometimes difficult because

of the setting time required,

and also because

problems to certain surfaces.

Magnets allow for quick setup

and movement,

can

of wax on clean,

but this method is

40

of adhesion

frequency-limited due to the

lack

a rigid

of

structures

may

simplest method,
3.

occur.

Hand-held

loading

probes,

of small

while being

the

are not recommended.

is

not necessary to calibrate the transducers used

essential

when

transducer calibration is

However,

developing

modal

(for example,
read

which is

lg at 1Khz)

directly

to

be

calibration that can be used.

of a single transducer

calibration

model

on

used

for

There are basically three

structural modification studies.

types of transducer

the acquiring of a

desired is

all that is

a modal test if

structure's mode shapes.

and

mass

Transducer Calibration

It
in

Also,

attachment.

requires

Absolute

a known source

fed through the transducer

spectrum

analyzer.

Referential

calibration requires a well-calibrated transducer to be used

as a standard to which the user's transducers are referenced.
Finally, the constitutive or ratio calibration, which combines
aspects of the absolute
constitutive

law

(for example,

the ratio (A*

01 ./Fo 1 t.
each other. [Ref.
4.

and referential

:p.

methods,

uses some

dividing all measurements by

m) to relate a pair of transducers to

10-2]

Excitation Techniques
a.

Shaker Excitation
The objective of an exciter in

a modal test is

to

input a force into the structure under test to excite all the
modes of interest.

Various excitation techniques have been

41

developed to help the analyst

problem.

One

of

the most

solve a particular vibration

common

hydraulic or electro-mechanical

techniques

is

to

shaker assembly.

shaker setup for a modal test

is

shown in

use a

A typical

Figure 10.

shaker should be attached to the test structure in

The

such a way

that excitation of the structure occurs at a single point and

in a single direction.

A thin flexible "stinger" with a force

transducer mounted at the end can be used to reduce bending

moments.

Using a shaker assembly requires a certain amount of

fixturing,

but

it

gives

good

control

of

the

excitation

bandwidth and level, and allows several choices of excitation

signals.

These

deterministic
periodic

chirp

excitation

signals

excitation
4,

(which

are

includes

pure

excitation techniques has its

characteristics

not

discussed in great detail by SMS,

9-44].

The

point

is

that

excitation

technique

constraints

of a particular

considerations

are

based

test

types:

sinusoidal
a

random,

sweep,
a

pseudo

random,

Each of these types of

advantages

will

primary

relaxation)

ep

periodic random and burst random).

These

two

(which includes
--an.A

Il

of

be
Inc.

and disadvantages.

listed

the

analyst

on

the

but

are

:pp. 9-2 through

must

select

considerations

application.

measurement

[Ref.

here,

an
and

Examples of these

time,

signal

leakage

control, bandwidth control, level of signal-to-noise ratio and

distortion removal.

42

CO

Force Iransducer

Notion
TransducerDgla
D;gltal
Signal

5tlnger
(Ea

DerAnalyzer

5ignal Generator
Figure 10
b.

Typical Shaker Setup

Impulse (Impact) Excitation

The latest

of an impulse,
to the test

tip

or tap,

structure,

The excitation
hammer

advancement in

the use

hanmmer to impart broadband excitation

thus exciting all

modes simultaneously.

force response can be altered by changing the

material,

range of interest.

modal analysis is

whicl,

is

determined

A soft rubber tip

I|
43

by

will result

the
in

frequency
a longer

impact

frequency

lower
result

in

frequency

ranges.

shorter

of

a hard steel tip

Conversely,

impact

is

range

into the

concentrate the input energy

time which will

times

for

cases

The

interest.

where

impact

low

convenient

cost,

and

and

portable,
However,

require any external signal source.

higher

technique

several advantagcs over a shaker assembly in that it

easy,

will

is

it

has

fast,

does

not

provides only

it

a low input energy with no randomization of the input,

a low

signal-to-noise ratio, and no control of the frequency content

if

a particular excitation bandwidth is

Brown
the

12]

[Ref.
worst

state that,

to

method

use

for

most

as the technical disadvantages far outweigh

advantages.

technical

Corelli and

impact excitation is

general,

excitation

possible

structural tests,
the

in

desired.

However,

the

convenience

so dominant that it

quickness of impact testing is

outweighs the technical disadvantages and is

and

generally

often chosen as

the excitation technique.

5.

Specifying the Nunmber of Degrees-of-Freedom

As

stated

on

direction

defined

structure

generally

three rotational.

test.

If

has

a measurement

Any

structure.
six

DOFs:

free

point and

point

translational

three

on

and

When deciding on the number of DOFs to use

for a particular test,

of the test,

a DOF is

earlier,

the analyst must consider the purpose

and the geometry of the actual structure under

the purpose of the modal test

44

is

simply to verify

modal frequencies previously preaicted anplytically, then only

a few DOFs would be required.
test

is

to cor.struct

However,

a mathematical

modal measurements at liscrete

are

mode shapes in

distinguishable

geometrical

and

complexity

interpolated

then sufficient

independent.

structure's

shapes of similar stractures tested

possible to

begin.

DOFs

determine

mode

It

is

shapes

which

Known mode

previously should be used

not using more DOFs than are

necessary to define the mode shapes for a structure,

in

the

how many DOFs with which to

Care should be taken in

results

to

the frequency range of interest

ultimately deteri, ines the nunber of DOFs required.

whenever

from

geometry into account,

linearly
of

the purpose of the

model

locations,

must be used, taking the structure's

ensure that all

if

longer computation and test

as this

time.

SIDEWINDER MISSILE MODAL ANALYSIS TEST SETJUP

1.

Test Structure
The GVT stand utilized in the missile modal analysis

setup is

located on the basement floor of Halligan Hall at the

Naval Postgraduate School in Monterey,

Californip.

The GVT is

the same structure constr-ucted and used by Hollyer [Ref.

except

for

the

addition

of

dozen

additional

1],

fasteners

joining the top and bottom plates to the 8 inch by 8 inch I

beams,

and two 14 inch bolts joining the top and bottom plates

together
the

for tl~e purpose

u-tructure.

This

of providing

structure

45

was

further
initially

stiffenina
designed

of
to

the frequency range of interest

provide minimal motion in

to

50

hz)

longitudinal

to

structure

no

has

of

frequency

resonant

initial

added

member excitation,

test

an
The

in

the

usually comprised

of a

for

stiffening

with
hz.

52.5

approximately

(0

excitation

direction.

lateral

Thst Equipment

2.

system is

A basic modal test

spectrum analyzer,

a computer with associated modal analysis

software,

a shaker or impact hammer to provide the excitation

(with an

external

if

generator,

driving

provide

to

transducer

function

point

needed),

measurements,

response

transducers to measure the response to the excitation,

test

Figure

17.,

setup
and

details

Specific

the

is

in

depicted
.

actual

of the

is

setup

individual

Scientific--Atlanta Model SD-380,

diagram
..

'block

this

test

of

simultaneous

r..n.

Figure
of this

setup is

12.
test

the

with the four channel option

sampling of four

This option allows the parallel

channels

in

shown

components

set'ap are discussed in Appendix D.

The spectrum analyzer used in

installed.

and the

The Sidewinder missile modal

actual structure to be tested.

analysis

force

information,

permitting

thus

force transducer to be connected to channel A,

the

and the three

Flexcel response transducers to be connected to channels B, C,

and D.

The SD-380 used in

the missile test

Synchronous Signal Generator (SSG)

46

-4i

setup also has a

capability, precluding tha

LU

i
w

C9,

Iq:
W

U.1
&0
Cj)j

ww
----

0
0

cc

cc)

L_

a..
:o

11__
C,,ur

(
_

cc
_

0<

M- 0C

Boc

9MsieMdlAayi

IUUJ(

ci47

Diaga

77

Ik'

Figure 12

AIM-9 Missile Modal Analysis Test Setup

48

need for a separate function generator to send the excitation

signal to the shaker.
include

broadband

limited noise,
with

the

white

computer,

the

program,

EMODAL

modal

interface,

board.

analysis

using a General

specifically

or

from

the

an actual modal test,

vibration

signal,

frequency content.
(full

the

This

National

interface

analyzer

to

in

the SD-380 is

this

case

an

allows

optional

When a structure is

the modes of the structure.

FRFs

from

structure,

the

are

center

modes

shown

as

frequencies,

these

FRFs that

which

c-reates

extracts

resonances

its

excited by white noise

the superposition

The analyzer

measurement

determines

locations

at

each

on

the

series

location on the structure.

of

identifying

are transferred
a

mathematical

peaks,
them as

with

to the EMODAL-PC

expression

for

calculate

and

display

of interest.
49

modal

The

characteristic

resonances.

the mode shapes and modal parameters.

then

for

and performs a Fast Fourier Transform to break out

the separate
FRFs

different

used to analyze

acceleration,

spectrum excitation) , the response is

of all

can

interface

band

pen plotter.
In

(GPIB),

noise,

of measurement data from the spectrum analyzer to

EMODAL

digital

Bus

random

The SD-380 communicates

the

via an IEEE-488

AT-GPIB

the transfer

pseudo

specifically

Interface

Instruments

noise,

sine and burst random.

software package,
Purpose

The SSG signal output types available

each

It

is

software,
FRY,

and

The software

animation

for

all

E.

MODAL-PC SOFTWARE BACKGROUND

1.

Introduction
With the

growth in

several good modal analysis

recent years,

have recently been released.

the

market

is

EMODAL-PC is

of modal

popularity

EMODAL-PC

analysis

in

software packages

One of the packages currently on

Entek

by

Corporation.

Scientific

a user-friendly personal computer-based software

specifically designed and written for modal testing,

package,

troubleshooting.
125

shapes,

shape

deflection

operating

It

analysis

vibration

and

allows 250 measurement locations for real

measurement

locations

for

complex

750

shapes,

total DOFs and 20 inode shapes.

EMODAL-PC requires an MS-DOS-based PC/XT/AT,
100% compatible computer,

with at least a 10 Mb hard disk and

(of which at least 420

at least 640 kb of random access memory

kb

must

be

available

80386, or

for

use

by

requirements include at least one floppy disk drive,

80287 or 80387 math coprocessor,

Other

EMODAL-PC).

a CGA,

an 8087,

EGA or VGA monitor,

and an IEEE-488 interface port on the computer on which it

is

to be used.
2.

Major Features
The version

of the

EMODAL-PC

Sidewinder missile modal analysis test

software
setup is

used in

the

version 2.75.

The major features of this version of the EMODAL-PC software

include the following

[Ref.

3:p.

50

2-2]:

* Modal test
outline to lead the user
necessary to conduct a modal test.
* Easy editing of measurement
geometry connectivity.
Automatic communication
IEEE-488 interface.

* Auto-configuration

through

location

the

steps

coordinates

and

with a spectrum analyzer via an

of spectrum analyzer.

a Transfer of measured FRFs

the program.

from the spectrum analyzer to

* Three
different
curvefitters:
rational fractional polynomial fit.
* Real-time animation of deflection

co/quad,

circle,

or

shapes.

a Ability to plot undeformed/deformed geometry.

* Ability
to
plot
measurements
in
real,
magnitude, phase, Bode or Nyquist format.
F.

imaginary,

EMODAL-PC MODAL TESTING FUNDAMENTALS

1.
m ^. Proc
Vta
du..A.
The

analyst

EMODAL-PC

program

or designer wishing

is

valuable

to analyze

and

for

the

solve noise

and

vibration problems of structures or mechanical

tool

systems.

The

most successful approaches to solving these problems usually

utilize
the

a combination of analysis and testing.

EMODAL-PC

program

can

be

used

following general sequence of steps

by
[Ref.

In this

the
3

analyst
:p.

vein,
in

the

6-2].

* Step 1:
Identification of problem components or systems,
usually done by operating vibration tests.
a Step 2:
Perform experimental modal analysis, in which
F-R.s are measured and the system's modal parameters are
identified.

51

Curvefitting and calibration of modal parameters

to extract the mode shapes from the frequency response

"*Step 3:
data.

" Step

4:
Predictions of dynamic
structural modification analysis.

2.

response,

prior

to

Modal Testing vs. Operating Deflection Shape Test

a. Modal Testing
Modal testing can be considered the experimental,

or measurement-taking approach,
or system's modal parameters.
test structure is
An FRF is

to determining a structure's
In conducting a modal test, the

subjected to some externally-applied force.

obtained (actually estimated) from the simultaneous

measurement of the exciting force and the response.

The modal

parameters are then extracted from a set of FRFs obtained at

different locations on the structure.
b.

Operating Deflection Shape Tist

An operating deflection shape test

is the process

of determining the vibration of some structure or system while

under

operating

It

conditions.

is

not a prediction

of a

system's dynamic behavior such as would be the result of a

modal test, but instead it
dynamic behavior.
test,

is

a direct measure of the system's

To conduct

the structure

an operating deflection shape

(usually running machinery)

to self-induced forces.

The forces generated by the structure

are not measured as in modal testing.

deflection

shape

is

vibrates due

determined

52

from

Instead, the vibration

a

set

of

vibration

amplitude and phase measurements obtained at many different

locations on the structure.
3.

Curvefitting
In

curvefitting

terms,

general

is

process

the

of

matching a set of data points obtained from FRF measurements

expression

to a mathematical

In the world of modal analysis,

linear regression analysis.

specifically

step

second

at

least

in

the

in

measurement
modal

global
damping,

frequency,

are

theory,

This

the structure.

on

position

natural

which,

stiffness,

few)

structure's

test

the

(or

parameter

estimation

the parameter

in

step

first

one

which

in

estimate

the

is

curvefitting
process

line

straight

recognized by most engineers as the result of a

is

curvefit,

the

of curvefit,

type

simplest

The

model.

based on a mass-spring-damper

is

then

estimation

used

is

to

parameters,

modal

and

mass

independent

of

followed by the
process,

shape

extraction, which uses all of the measurements obtained during

the test to estimate the structure's mode shapes.
coefficients

are known

Mode shape

as local parameters because

they do

depend on the position on the structure.

a.

currently

Coincident/Quadrature (Co/Quad) Fit

There

are

being

used

numerous
in

modal

-VJ

chioce

of

three

analysis,

each

techniques
having

its

The EMODAL-PC program gives the

advantages and limitations.

user

curvefitting

of

the

more

common

curvefitting

53

techniques.
available
method,

The
is

modal

the

estimate

measurement

of the

extracting
mode

for

shape

operating

shapes

curvefit

(co/quad)

method

fit.

quickly

the

line

with

the

band, and uses this amplitude as

coefficient,

deflection
to

spectrum

check

is

shapes,
the

used mainly

and

To use this method,

for

for extracting

integrity

of

the

Damping and other modal parameters are not estimated in

method.

This

also called the peak pick method because it

maximum magnitude in the fit

an

analysis

the coincident/quadrature

which is

searches

simplest

data.
this

all modes must be uncoupled and

the resonance peaks must be well-separated.

b.

Circle Fit
The second type of curvefitting technique available

in EMODAL-PC is

the circle fit.

This curvefit method is

used

primarily for extracting modal parameters when the modes are

well-separated and when accurate estimates of damping are not
required.
gives the

It

performs a least squared error algorithm which

shape

coefficient,

natural

frequency

and a

rough

estimate of the damping.

c. Rational Fraction Polynomial Fit

The third type of curvefitting technique available
in

EMODAL-PC is

the rational

curvefit method is

a good,

fraction polynomial fit.

This

general purpose curvefitter that

gives good results in most measurement

include single DOF curvefitting

54

cases.

Its main uses

on coupled modes

(situation

when two or more modes

to

difficult

distinguish

curvefitting.

The

curvefit method is
DOF fit,

but if

to attempt to

and
to

use

is

the

with

modes using the

fit

DOF
this

single

curvefitting
are desired.

results

EMODAL-PC TEST PROCEDURE

Conducting a modal analysis of a complex structure is
an

necessarily
software

task,

easy

decreases

the

the EMODAL-PC

from

through

entire

the

measurements,

outline

defined in
1.

of

step in

the

directory
type of test

project

and

to store the

It

the

user

test,

taking

and storing all

the data

up

the

as presented in
shown in

the modal

Figure 13,

and is

Set-up
using the

conducting a modal test

the software for the test

to conficT

about to be conducted.

leads

follow.

Configuring the General

EMODAL-PC software is

which

setting

is

2-4],

the sections that

The first

names

obtained by pressing

procedure,

1 3 :p.

analysis

outline window,

process

[Ref.

modal

The EMODAL-PC

main monitor,

The test

of

use

not

of difficulty.

extracting mode shapes,

for future use.

test

the

but

level

software has a modal test

<?>

is

move on to the

recommended

the most accurate

it.

multiple

obtained are poor,

This

technique to use if
G.

procedure

recommended

fit.

peaks)

between

the results

DOF

multiple

frequency that

are so close in

is

tells

in

the

project

to be conducted

this

program

step that the user

on what

and measurement

(either

55

first

modal test

drive

files.

and
The

or operating

START

1.

CONFIGURE THE GENERAL SETUP

GEOMETRY

2.

3.

4.

DATA ACQUISITION

EXTRACT AND ANIMATE SHlAPES

5.

STORE PROJECT

L
6.

RECALL PROJECT

7.

PREPARE REPORTS

END

Figure

13

EMODAL-PC

56

Test Procedure

shape

deflection

test)

is

output will be sent to

plotter.

The

analyzer

OPIB

most

chosen,

the

well

computer screen,

important

address,

as

input

in

which

without

as

the

a printer or a

this
the

whether

step

is

program

the

cannot

transfer measurements trom the analyzer.

2.

Geometry
The second step of the modal test. outline is the test

structure

in which the user enters the

geometry definition,

coordinates

of

measurement

the

structure, as well. as the links

measurement locations.

locations

(or elemen-Lts)

on

the

test

connecting the

There are three separate and necessary

processes that take place in defining the structure's geometry

within the modal analysis software.

The first is the process

of entering the x, y and z coordinates of all the measurement

locations

into

the

software.

EMODA.L-PC

These

measurement

locations should be completely thought-out and defined by the

analyst prior to entering themr into the computer.
function

can

be

used

to

inpuit

additional

The add

measurement

locat'ions, should this be necessary, and the edit functicn can

be used to mod-ify or correct locations that have already been
entered.

Once all of the measurement locations are entered,

the next

step

is to connect

these locations together.

The

links, or elements, defined in this step are the instructions

to the program to accurately draw the test structure.
the coordinates

Like

commands, the add or insert function can be

57

used

to

enter

new

links

and

the

edit

function

is

used

After

modify or correct links that have already been entered.

all

locaticis

geometry

and
function

links

have

should

be

been

selected

structure has been drawni correctly.

dispJayed

structure

to be

entered,
to

verify

parts of the structure.

that

the

translated

or

rotated

about

the

hidden or complex

Once the displayed geometry has been

the analyst is

verified,

display

This function allows the

as well as magnified to verify small,

axes,

the

to

ready to begin the actual conduct of

the modal test.

3.

Data Acquisition
data

The

acquisition

results.

a minimum,
spectrum
software

package,

the

most

analyst
in

understand

conjunction

that the

with

the operation
the

modal

use

in

acquiring

stationary

data

transducer(s)

The

correct

is

the

reference

to

the

to various

analysis

careful and

technique.

throughout

of the

the frequency

range of interest, and that the analyst use good,

measurement

as

transducers being used have been

calibrated correctly and can measure response in

consistent

important

Obtaining valid measurements requires,

that the
analyzer

is

as only valid measurements will give

part. of a modal analysis,

valid test

process

keep
test

and

locations

cn

move
the

procedure

the

to

transducer
response

structure.

The

spectrum analyzer will then measure FRFs between the reference

and response

coordinates.

58

tft

A:

--

The

first

and

acquisition process is
spectrum

analyzer

most

important

step

in

the

data

transferring the measurements from the

to

the

program.

To

accomplish

this

successfully, the data acquisition parameters of the EMODAL-PC

software

program

parameters

measurement
has

drive

and

auto-store
been

configured.
has

functions

feature.

Upon

properly.

reference

These

coordinate,

calibration

setup,

or modes,

setting

up

the user

completion

After the data acquisition

the

analyzer

can

be

properly

each

de 'icated to a specific

arid discussed
Manual

in

detail

[Ref.

14].

in

group of

the SD-380
However,

analyzer

front

panel"

with some help in

of analyzer

setup,

selection,

this
the

the program prompting the user

location

entered,

and

direction

(for

which

setup procedure.

"Read measurement"

coordinate for which a transfer function is

the

the

the modal test outline contains

command can be chosen from the data acquisition menu,

results in

and

This can be a rather complex task as the SD-380

data acquisition menu in

provides

name,

directory,

Signal Analyzer Operator's

"Help

configured

completed,

setup pages,

11

be

include the project

measurement

setup

must

example,

which

for the response

desired.
lx)

have

After
been

the program will begin the data transfer process.

When completed, the response transducer(s) may be moved to new

lonations,
have

and the process continued until transfer functions

;en obtained for all measurement locations of interest.

59

The second step in the data acquisition process is the

storing and recalling of measurements.
easiest

procedure

for storing measurements

program automatically
from the

The recommended and

spectrum

is

store them when they

analyzer.

This

is

to have the

are transferred

accomplished

via the

"Auto-store measurements=ON" parameter in the data acquisition

setup

menu.

manually

via

Otherwise,
the

the

"Store

acquisition menu.

measurements

measurement"

command

program

to

the

same

particular
as

measurement

number

select

"Measurement

acquisition menu,
enter "a"

the

location

number

assigned

to

directory"

and
on

is

not

the

test

To determine the

particular
command

measurement,
from the

data

enter the project name at the prompt,

then

to obtain a directory listing of all measurements.

When the measurement number of interest is

that

data

the number the

measurement,

structure at which the measurement was made.

the

the

and the program will prompt the user for a

The measurement number is

necessarily

in

stored

command from the data

measurement number,
assigns

be

To recall a measurement previously stored,

simply select the "Recall measurement"

acquisition menu,

can

nunt..er

at

the

"Recall

measurement"

determined,
prompt,

enter

and

the

program will then recall that measurement.

The
viewing

third step

- measurement,

measurement.

ilm

from

the

data

in

thq data

accquisition

usually per:Tormed

after

process

is

recalling a

After selecting the "View measurement" command

acquisition

menu,
60

the

program

displays

the

recalled measurement,
to magnify

or

along with commands which allow the user

expand the

view,

set

up

frequency

bands

where fit

bands are

for

shape extraction and execute curvefits.

4.

Extract and Animate Shapes

This is

set

up

and

the step in

curvefits

are

extracted and animated.

is

to

identify

measurement
procedure
is

the

for

a modal test
performed,

for setting up fit

for the desired input.

set up fit
menu,

of

curvefit

bands" option in

interest
will

on a
be

are

bands

displayed

conducted.

The

with the program prompting the

From the "View measurement

and

the "Extract and Animate Shapes"

the <SETUP FIT> option should be selected.

will prompt the user for a shape number

increasing frequency),

shapes

bands and conducting a curvefit

simple and straight-forward,

uscr

mode

The purpose of setting up fit

frequencies

which

and

curvefit type

The program

(assigned in

order of

(either co/quad,

circle

or polynomial)

and the upper and lower limits of the desired

curvefit band.

The user then selects the <CURVEFIT> option to

execute the curvefit,

for which the program displays a graph

of the curvefit results in

Nyquist formats.

real,

imaginary,

After the curvefit,

log magnitude and

the user has the option

of rejecting the curvefit or accepting it,

if

satisfactory,

in

and

saving

the

parameter table.

61

results

the results are

the

project's

To
curvefit

extract

using

the

initiate

the

the

shapes"

above

shape

mode

extraction

then select tne "Shape

(real

or

as

For

process,

"Extract

on the disk

each

fit,

After

are

these

the

To

"Extract

Shapes"

Setup" option

menu,

and set the

include extract
and

parameters

select the "Extract shapes"

shape

generated.

and Animate

coordinate,

is

shape file,

select

The parameters

reference

drive/directory.

shapes

Extraction

desired.

complex),

properly,

process.

structure's

option from the

param.eters

each measurement

computed and saved it,the project's

coefficient is
from which

shapes,

measurement

have

option,

type

been

set

and enter the

measurement numbers for which extracted shape coefficients are

desired.
and

The program will recall these measurements from disk

curvefit

structure's

them,

resulting

mode shapes,

in

the

generation

different

either in

the

which can then be animated.

The purpose of animating the shapes is

the

of

measurement

locations

on

the

to observe how
structure

move,

translation or rotation or both, as a result of the

excitation force.

After the shapes have been extracted,

the

user selects tne "Animate shapes" option from the "Extract and
Animate
The

Shapes"

program

structure.
including
structure,

menu,

then

and enters

displays

tne

the

desired

animation

Numerous options are given in

the

ability

to zoom into

of

shape
the

deformed

the animation menu,

different

rotate or translate the geometry,

parts

of the

and display both

the deformed and undeformed geometry on dual plots.

62

number.

5.

Store Project
All of the files

can be
menu

stored very

associated with a particular

easily by

from the modal test

store

the project

files

files

will be lost

if

power

outage

or

selecting

outline.

It

immediately

the
is

"Store

very

upon

project
Project"

important

completion

the program stops for any reason

user

exiting

the

program

without

as

when saved,

is

option is

selected.

the

(e.g.,

saving)

There are actually six different types of project files

are stored when this

to

that

Each of these files,

given the same name as the project name chosen

when the general setup is

filename extension,

configured, but each has a different

as summarized below

[Ref.

13 :p.

2-61].

* Coordinates file
- contains the x, y, and z coordinates of
each measurement location; has extension .CRD
* Links file
- contains the connectivity that
structure; has extension .LNK

draws the test

* Parameter table file

- contains the fit
band definitions
and the modal parameters extracted by curvefitting;
extension .PAR
* Shapes file
- contains the deflection shapes,
columns by shape number; has extension .Sxx

SSetup

menus
file
parameter settings;

* Title
.TTL

Once

all

computer's

contains
program
has extension .SET

file

contains

of

the

project

hard disk,

they

-che project

files
can

disks for future reference.

63

have
easily

title;

be

arranged in

configuration

has

been

has

extension

stored

copied

on

on

the

floppy

Recall Project

6.

into computer

back

or

results

may need to be recalled

stored project

A previously

memory

preparing

recall

To

reports.

outline and

The program will recall

enter the project name at the prompt.

and upon completion,

of the project files

the modal test

outline.

final

This

allows

step

the

user

to

print

Prior to receiving any of these

format.

the program, however,

to

plotter

be

of the correct configuration,

given

four

files

And tables",

which

the

or

file

second option,

options.

The

displays

the print

files

to

"Plot shapes",

be

"The third

option,

After

option,
selection
are

the

outline,

all

"Print
menu,

from

selected.

The
the

sending these plots either

or to an external printer

"Recall a measurement from disk",

64
IM`

printed

step.

will produce plots of either

undeformed or deformed structure,

to the computer screen,

first

is

This

and upon selection

of the "Prepare Reports" option of the modal test

is

outputs from

used.

accomplished during the "General Program Setup"

verification

report-

in

the program must be correctly configured

and/or

printer

the

all

measurements,

of specific

plots

and produce

user

project

produce plots of the undeformed and deformed structure,

files,

for

will redisplay

Prepare Reports

7.

ready

simply

project,

select the "Recall Project" step of the modal test

all

the

of reviewing

for the purpose

or plotter.
is

used to

recall a specific measure.lent from the measurement directory

for the

of plotting

purpose

it.

The

fourth

option,

"Plot

measurement",

can then be selected to produce a plot of that

measurement.

However,

must

first

configured

In this

interest.
including
type,

be

title,

type

the "Measurement plotting setup" menu

menu,

of plot

for

plotting

the

measurement

of

the user selects plotting parameters

imaginary,

(real,

Bode,

etc.),

x-axis and y-axis labels and divisions,

and so on.

65

line

scaling,

CONCLUSIONS

IV.

A.

ANALYTICAL MODELING

Within
finite

the

scope

element

analytical

its

of

of

dynamic

behavior

mode

the

for

system,

and

and

mode

analysis

structure

yields
under

MSC/PAL2

obtaining

an

analyticafly
The

shapes.

insight

into the

a wide variety

of

Knowing the lowest few resonant frequencies

which

redesign

components

satisfactory

frequencies

of the

shapes

information

investigation,

missile

a normal modes

dynamic loading.

requiring

is

natural

conduct

and

this

software

model

determining

of

provides

the

may

in

or

aid

designer
identifying

possible

interface

or

analyst

with

potential
problems

areas

between

of a system.

In the case of the AIM-9 missile, the material properties

and

mass

software

distributions
can

be

in

modified,

the model

obtained via the

on

based

expecimental comprehensive modal test,

obtained, feasibility

studies can easily be performed

the AIM-9M-8/9 program will retrofit

counter-countermeasures

incorporate

an

Once an accurate

for the AIM-9 upgrades currenuly being planned.

infrared

of

to accurately represent

the dynamic response of the actual missile.

model is

outcome

the

PAL2

Specifically,

existing missiles with an

feature,

the

AIM-9R

will

an imaging focal plane array to increase target

66

acquisition range while improving the missile' s countermeasure

and clutter
will

rejection capabilities,

feature

[Ref.

15].

an

improved seeker,

All

of these

parameters

and

modified in

the finite

The

mass

AIM-9

satisfactorily

airframe,

changes may

distributions,

and warhead/fuze

affect

the material's

which

can

be

readily

element model.

model

developed

represents

material's

and the yeai 2000 AIM-9X

parameters

an

in

actual

and mass

this

investigation

missile;

distribution

matched to the natural frequencies

however,

have

yet

to be

and mode shapes determined

by the dynamic modal measureiment of an actual missile.

tests

the

Modal

are planned for a follow-.on phase of the investigation,

at which point the model can easily be modified to reflect

accurate
element
B.

dynamic
stiffness

by

altering

the

mass

locations,

or material properties.

EXPERIMENTAL MODELING
While

the

excitation

GVT

stand

of a test

article,

excitation of a test
be

response

an

structurally

is

satisfactory
it

article

in

stiffened

is

for

presently unsuitable for

the lateral
prior

longitudinal

to

direction and must

the

conduct

of

comprehensive missile modal test.

Within the scope of this
test

investigation,

setup and the associated

the modal analysis

methodologies

established

are

satisfactory for conducting a comprehensive modal analysis of

th

AIM-9

missile

system

for

the

67

purpose

of

experimentally

determining
parameters.

the

missile's

transient

response

and

modal

The Fast Fourier Transform-based analyzer and the

EMODAL-PC modal analysis software have been incorporated into

the

test

setup,

measurement

and

and

are

ready

subsequent

for

transfer

determination

of

function
a

structure's

modal parameters.
C.

COMBINED TESTING AND ANALYSIS

The use of the analytical and experimental

techniques

in

a complementary manner allows the designer or analyst to draw

on

the

strengths

of

each

technique

to

thoroughly

conpletely describe the dynamic response

of a structure.

first

dynamic

determining

experimentally

and

structure's

then

modifying

the

analytical model to match that response,

this

called

"what

if?"

studies,

approach

problems

can be

modifications

on

the

actual

thereby

avoiding

the

approach to solving
hardware.

are implemented.

68

an

design options,

The

iterated on the model before

or upgrades

in

the analyst can use

costly and time-consuming trial-and-error

vibration

By

response

parameters

analytical model to investigate different

sometimes

and

"what

if?"

any hardware

LJ1m

V.
A.

RECOMMENDATIONS

ANALYTICAL MODELING
After

completion

system,
same

of

modal

mode

missile's

shapes

and

This
material

natural

can

be

points
final

front

fin.3

across

an

and refined

the

the

AIM-9

actual

entirely

and

missile

by

observed

changing

by re-distributing

by

missile

body,

defining

element

vice

version

the
at

of the

mcdifi.cations
on the

frequencies

accomplished

properties,

masses more evenly along the

the

on

the analytical model should be modified to match the

experimentally.

to

test

made

the

missile's

single

the

lumped

by adding stiffness
attachment

node

analytical
to

the

point.

model,
model,

mode shapes and frc.quencies

The

and hence

will

depend

obtained via

modal

analysis.

Upon

completion

model,

design

AIM-9

upgrades

frequencies
ascertain

iterations
to

whether a
(or the

This

can

be

distribution

Orion

AIM-9 missi1a

should be conducted
analytically

flutter
II

and

accomplished primarily
for

upgrades,

problem will

wing)

the

for

determine

and mode shapes of these

wing

mass

of an accurate

the

exist

upgraded

and

the

planned

the
in

natural

an effort

between the
AIM-9

by modifying

AIM-9M-8/9

analytical

to
P-3

missile.

the

model's

AIM-9R,

and

by

69

tA

lq]-'

1!'i

Nrt.~~.-1

'p

"i

modifying

the

model's

mass

distribution,

fin

surfaces

and

possibly the shape of the body for the AIM-9X version.

B.

EXPERIMENTAL MODELING
The

GVT stand needs

excitation

measurements

structure

could

replacing

the upper and

plates,

to be

be

in

stiffened

the

stiffened
lower

laterally

lateral

in

to allow

direction.

several

ways,

3/16" plates

The

including

by thicker

1/2"

adding sandbags between the plates to provide damping,

and adding transverse bracing between the three internal

"I"

beams.
The AIM-9 missile system should undergo
modal test

to completely determine its

frequencies
lateral

experimentally

directions,

setup completed in

using
this

and

flutter

Orion

II

problems

both

the

missile

longitudinal

modal

analysis

and
test

The results of this

should then be compared to the P-3

wing
at

the

mode shapes and natural

investigation.

comprehensive modal test

Orion

in

a comprehensive

models

any

of

to

the

examine

proposed

any
wing

potential
carriage

stations.
C.

COMBINED TESTING AND ANALYSIS

The

this

facilities

investigation

should be utilized
analysis

on other

and methodologies
of the AIM-9
to

conduct

candidate

missile's

similar

missiLe

70

discussed

and used

dynamic

in

response

combined testing and

systems,

including

the

AGM-65 Maverick and AGM-M88 BALM, considered for incorporation

into the P-3's weapon inventory.

71

APPENDIX A
The MSC/PAL2 software code for

the

AIM-9

missile

statements were

is

given

the

finite

below.

element model of

Explanatory

included where appropriate.

commands which could not be fully explained in

statement,
[Ref.

6]

command

comment

Foc those
a brief

PAL2

comment

the appro-riate page of the MSC/PAL,2 User's Manual

is

given,

on

which

detailed

discussion

of the

can be found.

TITLE - AIM9 GEOMETRY FILE

C
C DEFINE NODES FOR MISSILE BODY IN
C
NODE -3
C
-

CYLINDRICAL COORDINATES

C NODE NUMBER, RADIUS, THETA, Z

C
(Ref. e:p. 3-92)
1 2.50, 0, 0 THROUGH 16 2.50, 33-7.5, 0
545 2.50, 0,
102 THROUGH 560 2.50, 337.5,
51 0,
0, 0
C
C GENERATE NODES OF MISSILE
C
(Ref. 6 :p.
9-99)
NODE 22
1, 16, 56'1,
1, 16
NODE 22
1, 1-6, 560:

15,

102

BODY AUTOMATICALLY

16

C
C GENERATE 1ODES OF TAPERED NOSE'ONE

AUTOMATICALLY

C
NODE 3
562 2.50,
594 1.25,

0,
0,

NODE
L,,,,

bC9,

22
577,

103 TiROUGH
.. 12 TIROUGH

1,

577 2.50,
609 1.25,

16

72

337.5,
3-;7.5,

103
112

NODE 22
562, 577,

609,

NODE 3
610 0, 0,

113

15,

16

C
C CHANGE POSITION OF CERTAIN NODES TO ALLOW ATTACHMENT OF FINS
C
C MOVE SECOND & TENTH CONCENTRIC RINGS 1 INCH IN -Z DIRECTION
C
NODE 1
17 0, 0, -1 THROUGH 32 0, 0, -1
NODE 1
145 0, 0,

-1,

THROUGH 160 0

0,

-1

C MOVE 34TH CONCENTRIC RING 2 INCHES

NODE
529 0,

1
0,

-2,

THROUGH

544 0,

0,

-2

C
C DEFINE NODE POINTS FOR FRONT FINS
C
NODE 3
C

C FIRST FRONT FIN

C
611 2.75, 45, 93
612 2.75, 45, 96
613 2.75, 45, 97
614 2.75, 45, 101.5
615 7.00, 45, 97.25
616 7.00, 45, 95
617 7.00, 45, 94.5
618 7 .00,
45, 93
619 11.25, 45, 93
C
C SECOND FRONT FIN
C
620 2.75, 135, 93
621 2. 75, 135, 96
622 2.75, 135, 97
623 2.75, 135, 101.5
624 7.00, 135, 97.25
625 7.00, 135, 5
626 7.00, 135, 94.5
627 7.00, 135, 92
628 11.25, 135, 93

73

IN THE -Z

DIRECTION

C
C THIRD FRONT FIN
C
629 2.75, 225, 93
630 2.75, 225, 96
631 2.7Z, 225, 97
632 2.75, 225, 101.5
633 7.00, 225, 97.25
634 7.00, 225, 95
635 7.00, 225, 94.5
636 7.00, 225, 93
637 11.25, 225, 93
c
C FOURTH FRONT FIN
C

638
639
640
641
642
643
644
645
646

2"75,
2.75,
2.75,
2.75,
7.00,
7.00,
7.00,
7.00,
11.25,

315,
315,
315,
315,
315,
315,
3-5,
315,
315,

93
96
97
101.5
97.25
95
94.5
93
93

C DEFINE NODE POINTS FOR -EAR WIN4GS

C
NODE 3
C
C FIRST R.AR WING

647 4.75,

648 4.75,

45,

45,

649 4.75, 45, 9

650 4.75, 45, 12
65. 1.75' 45, 15
652 4.75, 45, 18
653 4.757 45, 21
654 4.75, 45, 24
655 7.88, 45, 21
656 11.50, 45, 18
657 Ii.50, 45. 15
658 11.50, 45, 12
659 11.50, 45r 9
660 11.50, 45, 6
661 11.50, 45, 2
c
C SECOND REAR WING
C
662 4.75, 135, 2

r14

663 4.75,
664 4.75,
666 4.75,

135,
135,
135,

6
9
12

667 4.75,

135,

18

666 4.75, 135, 15

669 4.75, 135, 21
669 4.75, 135, 24
670 7.88, 135, 21
671 11.50, 135, 18
672 11.50, 135, 15
673 11.50, 135, 12
674 11.50, 135, 9
675 11.50, 135, 6
676 11.50, 135, 2
C
C THIRD REAR WING
C
677 4.75, 225, 2
678 4.75, 225, 6
679 4.75, 225, 9
680 4.75, 225, 12
681 4.75, 225, 15
682 4.75, 225, 18
683 4.75, 225, 21
684 4.75, 225, 24
685 7.88, 225, 21
687 11.50, 225, 15
687 11.50, 225, 15
688 13.50, 225, 12
689 11.50, 225, 9
690 11.50, 225, 6
691 11.50, 225, 2
C

C FOURTH REAR WINGC

692
693
694
695
696
697
698

"699
700
701
702
703
704
705
706

4.75,
*.75,
4.75,
4.75,
4.75,
4.75,
4.75,
4.75,
7.88,
11.50,
11.50,
11.50,
11.50,
11.50,
11.50,

315,
315,
315,
315,
315,
315,
315,
315,
315,
315,
315,
315,
315,
315,
315,

2
6
9
12
15
18
21
24
21
18
15
12
9
6
2

75

I2:
I"I

Il

C
C ENTER MATERIAL PROPERTIES
C
C YOUNG'S MODULUS,
SHEAR MODULUS, MASS DENSITY, POISSON'S
C RATIO, TENSILE YIELD STRESS
C
(Ref. 6:p. 3-83)
MAT 30.OE6,
12.0E6, 7.76E-4, 0.25, 30.0E3
C
C DEFINE PLATE TYPE FOR MISSILE BODY
C
C QUADRILATERAL PLATE ELEMENT, Ni, N2, THICKNESS
C
(Ref. 6:p. 3-113)
QUAD 1, 1, 0.25
C

C GENERATE QUAD PLATE ELEMENTS ON MISSILE BODY AUTOMATICALLY

C
(Ref. 6:p. 3-56)
GENERATE CONNECTS 1, 16, 560,
1, 16
GENERATE CONNECTS 1, 16, 560, 15,
16
GENERATE CONNECTS 562, 577, 609, 1, 16
GENERATE CONNECTS 562, 577, 609, 15, 16
C
C DEFINE REAR MISSILE PLATE AND CONNECTIVITY
C
C TRIANGULAR PLATE ELEMENT, Ni, N2, THICKNESS
C
(Ref. 6:p. 3-121)
TRI 1, 2,
0.25
CW

C CONNECT NODE A TO NODE B TO NODE C

C
(Ref. 6:p. 3-27)
CON 1 TO 2 TO 561
CON 2 TO 3 TO 561
CON 3 TO 4 TO 561
CON 4 TO 5 TO 561
CON 5 TO 6 TO 561
CON 6 TO 7 TO 561
CON 7 TO 8 TO 561
CON 8 TO 9 TO 561
CON 9 TO 10 TO 561
CON 10 TO 11 TO 561
CON II TO 12 TO 561
CON 12 TO 13 TO 561
CON 23 TO 14 TO 561
CON 14 TO 3.5 TO 561
CON 15 TO 16 TO 561
CON 16 TO 1 TO 561
C
C DEFINE PLATE CAP CONNECTIVITY OF NOSECONE
C
CON 594 TO 595 TO 610
CON 595 TO 596 TO 610
CON 596 TO 597 TO 610

76

CON 597

TO 59

TO 610

CON
CON
CON
CON
CON
CON
CON
CON
CON
CON
CON
CON
CON
C

TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO

TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO

597
598
599
600
602
602
603
604
605
606
607
608
609

598
599
60C
601
602
603
604
605
606
607
608
609
594

610
610
610
610
610
610
610
610
610
610
610
610
610

C CONNECT BODY TO NOSECONE

C

QUAD 1, 1, 0.25
C

C CONNECT NODE A TO NODE B TO NODE C TO NODE D

C (CONNECT COUNTERCLOCKWISE)
C
(Ref. 6:p. 3-27)
CON 545 TO 562 TO 563 TO 546
CON 546 TO 563 TO 564 TO 547
CON 547 TO 564 TO 565 TO 548
CON 548 TO 565 TO 566 TO 549
CON 549 TO 566 TO 567 TO 550
CON 550 TO 567 TO 568 TO 551
CON 551 TO 568 TO 569 TO 552
CON 552 TO 569 TO 570 TO 553
CON 553 TO 570 TO 571 TO 554
CON 554 TO 571 TO 572 7) 555
CON 555 TO 572 TO 573 T) 556
CON 556 TO 573 TO 574 TO 557
CON 557 TO 574 TO 575 TO 558
CON 558 TO 575 TO 576 TO 559
CON 559 TO 576 TO 577 TO 560
CON 560 TO 577 TO 562 TO 545
C
C DEFINE FRONT FIN PLATE TYPE AND CONNECTIVITY
C
C FIRST FIN
C
C QUAD PLATE TO ATTACH FIN TO MISSILE BODY
C
QUAD 1, 1, 0.35
CON 515 TO 531 TO 613 TO 612
C
C FIN CONSTRUCTION
C
QUAD 1,
1,
0.10
CON 611 TO 612 TO 617 TO 618

77

CON 612 TO 613 TO 616 TO 617

CON 613 TO 614 TO 615 TO 616

TRI 1, 2,

0.10

CON 619 TO 618 TO 617

CON 619 TO 617 TO 616
CON 619 TO 616 TO 615

C
C
C 9ECOND FRONT FIN
QUAD 1, 1, 0.35
CON 519 TO 535 TO 622 TO 621

QUAD 1,

1,

0.1

CON 620 TO 621 TO

CON 621 TO 622 TO
CON 622 TO 623 TO
TRI 1, 2, 0.1
CON 628 TO 627 TO
CON 628 TO 626 TO
CON 628 TO 625 TO
C
C THIRD FRONT FIN
C
QUAD 1, 1, 0.35
CON 523 TO 539 TO
QUAD 1, 1, 0.1
CON 629 TO 630 TO
CON 630 TO 631 TO
CON 631 TO 632 TO
TRI 1, 2, 0.1
CON 637 TO 636 TO
CON 637 TO 635 TO
CON 637 TO 634 TO
C

626 TO 627
625 TO 626
624 TO 625
626
625
624

631 TO 630
635 TO 636
634 TO 635
633 TO 634
635
634
633

C FOURTH FRONT FIN

C
QUAD 1, 1, 0.35
CON 527 TO 543 TO 640 TO 639
QUAD 1, 1, 0.1
CON 638 TO 639 TO 644 TO 645
CON 639 TO 640 TO 643 TO 644
CON 640 TO 641 TO 642 TO 643
TRI 1, 2, 0.1
CON 646 TO 645 TO 644
CON 646 TO 644 TO 643
CON 646 TO 643 TO 642
C
C DEFINE MATERIAL PROPERTIES FOR REAR WINGS
C
MAT 10EG, 4E6, 2.59E-4, 0.25, 10E3

78

C
C DEFINE REAR WING PLATE TYPE AND CONNECTIVITY
C
C FIRST REAF WING
C
QUAD 1, 1, 0.5
CON 19 TO 35 TO 648 TO 647
CON 647 TO 648 TO 660 TO 661
CON 35 TO 51 TO 649 TO 648
CON 648 TO 649 TO 659 TO 660
CON 51 TO 67 TO 650 TO 649
CON 649 TO 650 TO 658 TO 659
CON 67 TO 83 TO 651 TO 650
CON 650 TO 651 TO 657 TO 658
CON 83 TO 99 TO 652 TO 651
CON 651 TO 652 TO 656 TO 657
CON 99 TO 115 TO 653 TO 652
CON 652 TO 653 TO 655 TO 656
CON 115 TO 131 TO 654 TO 653
2, 0.5
TRI 1,
CON 654 TO 131 TO 147
CON 655 TO 653 TO 654
C
C SECOND REAR WING
C
QUAD 1, 1, 0.5
CON 23 TO 39 TO 663 TO 662
CON

662

TO

663

TO

675

TO

676

CON 39 TO 55 TO 664 TO 663

CON 663 TO 664 TO 674 TO 675
CON 55 TO 71 TO 665 TO 664
CON 664 TO 665 TO 673 TO 674
CON 71 TO 87 TO 666 TO 665
CON 665 TO 666 TO 672 TO 673
CON 87 TO 103 TO 667 TO 666
CON 666 TO 667 TO 671 TO 67W
CON 103 TO 119 TO 668 TO 66;
CON 667 TO 668 TO 670 TO 671
CON 119 TO 135 TO 669 TO 668
TRI 1, 2,
0.5
CON 669 TO 135 TO 151
CON 670 TO 668 TO 669
C
C THIRD REAR WING
C
QUAD 1, 1, 0.5
CON 27 TO 43 TO 678 TO 677
CON 677 TO 678 TO 690 TO 691
CON 43 TO 59 TO 679 TO 678
CON 678 TO 679 TO 689 TO 690
CON 59 TO 75 TO 680 TO 679

79

CON 679 TO 680 TO 688 TO 689

CON 75 TO 91 TO 68" TO 680
CON 680 TO 681 TO
87 TO 688
CON 91 TO 107 TO 682 TO 681
CON 681 TO 682 TO 686 TO 687
CON 107 TO 123 TO 683 TO 682
CON 682 TO 683 TO 685 TO 686
CON 123 TO 139 TO 684 TO 683
TRI 1, 2, 0.5
CON 684 TO 139 TO 155
CON 685 TO 683 TO 684
C
C FOURTH REAR WING
C

moCON

QUAD 1, 1, 0.5
CON 31 TO 47 TO 693 TO 692
CON 692 TO 693 TO 705 TO 706
CON 47 TO 63 TO 694 TO 693
CON 693 TO 694 TO 704 TO 705
CON 63 TO 79 TO 695 TO 694
CON 694 TO 695 TO 703 TO 704
CON 79 TO 95 TO 696 TO 695
CON 695 TO 696 TO 702 TO 703A
CON 95 TO 111 TO 697 TO 696
CON 696 TO 697 TO 701 TO 702
CON Il1 TO 12"7 TO 698 TO 697
CON 697 TO 698 TO 700 TO '701

127 TO 143 TO 699 TO 698

TRI 1, 2, 0.5
CON 699 TO 143 TO 159
CON 700 TO 698 TO 699
C
C DEFINE ATTACHMENT

POINTS

C
C AT ATTACHMENT POINTS, ALLOW ROTkTION ABOUT X AXIS ONLY
C
(Ref. 6:p. 3-124)
ZERO

TA 85, 213, 373

RY 85, 213, 373
RZ 85, 213, 373
C
C DEFINE POINT MASSES FOR SEPARATE MISSILE COMPONENTS
C
C MASS COMMI4AND,
C

cC

NODE NUMBER,

MASS OF MOTOR COMPONENT

MASS

141,

0.128

MASS 269,

0.128

MASS OF COMPONENT

(TWO POINT MASSES SINCE 70" LONG)

80
,

ci

C MASS
C
M.iSS

OF WARHEAD

413,

COMPONNT

0.0538

C MASS OF TARGET DETECTOR COMIPONENT

MA
MASS 477, 0.0233

C
MASS

557,

0.0673

81

APPENDIX B
Figures 14 through 19 depict the first
the

AIM-9

They

missile

model

represent

natural

the

frequencies

58.80

hz.

It

of 29.65,

50.79,

57.36,

be

the

missile

and

individual

optimum planar
still

strictly

the

results

the

model

for

third

While not totally

and

of
obtained

geometry,

mass

representing

the

presented

in

reponses,
the

first

the
show
view

second showing the

showing the

obvious

57.89,

representative

shapes:
the

six

and

viewing the

mode

first

were

masses

of deformation,
and the

its

frequencies,

The figures,

animated

upper limit

these

These

lumped

components.

undeformed structure,
deformation.

that

from

additional

of

52.97,

are not necessarily

representation

frames

showing the

noted

response.

missile

investigation.
at

mode shapes,

density,

this

deformations

should

mathematically,

in

missile's

hence these
actual

constructed

six mode shapes for

from the

lower limit

of

figures,

the

missile model exhibited the following modal responses.

"*Mode

1 (29.65 hz) - whole body z-axis rotation

attachment points with some bending in x-z plane

"* Mode 2 (50.79 hz) - bending mode in

front fin z-axis

"*Mode 4

(57.36
body motion

hz)

y-z plane with slight

rotation

"* Mode 3 (52.97 hz) - bending mode in

front fin z-axis

about

x-z plane with slight

rotation
-

front

" Mode 5

fin

z-axis

(!' 89 hz) - front fin z-axis

body torsion in the x-y plane
82

rotation only;

rotation with

no

szme

Mode

(58.80 hz)

hbody bending in
Th,

the y-z plane and torsion in

fins,

the x-y plane

are actually gross exaggerations

from two possible phenomena.

fins are

for these fins.

shape in

external
important.

load

The first

is

The second is

resultin~g in

applied;

The deformationis

only by a

the
thin

a lack of stiffness

the fact thaZ scaling for each

a normal modes analysis

is

arising

the fact that

attacned to the missile body

quadrilateral plate element,

mode

7-axis rotation with some

figures, showing what appear to be strange contortions

of the front

fronit

fin

front

only

the

is

arbi-crary
relative

of the front fins in

was actually less than 0.3 inches.

I3

83

sinc-e no
shape

is

tl.ese modes

t7
-t

S....

Figure 14

11

AIM-9 Missile Model Mode

84

I Response

S..

Ii~I ]
a-

7-q7

-j1

Rsos
5AI-9-i -ieMode.M
Iiiu -43
U8

-3-

..
....
.j

4j

._--

:.-'..,

~Figure

6Ab-Misl

olNde3esoe

86

"-.

K ;",

27

'C

_-

:'

yQ

Figure 18

AIM-9 Missile Model Mode 5 Response

88

Figure 19

AIM-9 Missile Model Mode 6 Response

89

APPENDIX C
TRADE-OFF STUDIES
Several studies were conducted on the basic AIM-9 model in
an attempt to determine the effects of changing the location
of

the

point

structure.
in

masses

and

the

material

properties

were conducted,

the first

corresponding natural
in

which they

with the current AIM-9 model being Case V.

For each of the cases,

II,

the

The results of these trade-off studies are given

the five cases discussed below in the order in

Case

of

which

frequencies

only

the

first

six mode shapes and their

are described,
four

determined.
In studying these cases,
g
theg mass location or mL aterLAi a I-

mode
one

except for
shapes

can

both the mode shapes and the resonant frequencies;

see

were
that

sometimes

quite drastically.

Case I - missile body only with point masses located on top of

missile body.

"*Mode 1 (50.02 hz)

- bending in

x-z plane

"*Mode 2 (53.36 hz) - bending in y-z plane

"*Mode 3 (71.39 hz)
" Mode 4 (209.60 hz)
"*Mode 5 (262.70 hz)
torsion in

rotation about z-axis

complex bending in

y-z plane

complex bending in y-z plane and body

x-y plane

90

y-z plane and bQdy

* Mode 6 (291.80 hz) - complex bending in

torsioni in x-y plane
Case II

- missile body cnly with masses located on bottom of

samc z--axis location.

missile body but in

"*Mod, 1 (33.42 hz)

rotation about z-axis,

with bending in

x-z plane

"*Mode 2 (53.02 hz)

bending in

y-z plane

(59.75 hz)

bending in

x-z plane

- Mode 3

"*Mode 4 (140.60 hz)

Case

III

full

AIM-9

model

with

fins

and wings,

and

five

located on bottom of missile body.

point masses

"o Mode

1 (24.56 hz)
attachment points)

"*Mode 2 (50 80 hz)

""Mode 3

y-z plane

complex bending in

(50

torsion in

hz

rotation

bending in

hendiinn

in

about

z-axis

the

(about

y-z plane
-x-z plane with some body

x-y plane

*, Mode 4 (57.36 hz)

body motion

fin rotation about

front

z-axis;

" Mode 5

front fin xotation about z-axis

(57.72 hz)
some body torsion in x-y plane

no

with

"*Mode 6 (58.70 hz) - front fin rotation about z-axis with

x-y plane and bending in

some body torsion in

Case IV

full

AIM-9 model with ten point masses

x-z plane
located on

top and bottom of missile body at the same z-axis location.

"" Mode

1 (26.68 hz)
attachment points)

rotation

about

z-axis

" Mode 2 (50.58 hz) - bending in

x-z plane

" Mode 3 (51.01 hz) - bending in

y-z plane

91

(about

the

"*Mode 4

(57.36 hz)
body motion

" Mode 5 (57.57 hz)

some body torsion in

front fin rotation about z-axis;

no

front fin rotation about z-axis with

x-y plane

"*Mode 6 (58.71 hz) - front fin rotation about z-axis with

some body torsion in

x-y plane and bending in

x-z plane

Case V - full AIM-9 model with five point masses located on

bottom of missile body and the rear wings changed from steel
to aluminum.
* Modes given previously in Appendix B.

92

APPENDIX D
MODAL TEST SYSTEM TEST EQUIPMENT

Spectrum Analyzer Model SD-380

Scientific-Atlanta

signal

channel

four

and in

measures

the case

it

signal analysis capabilities,

feature,

(or cascade)

interface,

IEEE

signal(s)

of the two and four channel

and

the

equipped with a built-in

is
a

an

56K word input memory,

to

alility

perform

Adaitional

ditterentiation/integration.

options,

In addition to standard

the relationship between two signals.

waterfall

the

or

two or four BNC connectors located on the rear

present at one,
panel,

It

analyzer.

two

one,

a microprocessor-based

is

SD-380

T-ne Model

single/double

specifications

and

information on the SD-380 are supplied by Scientific-Atlanta

[Ref.

14].

COMPAQ Deskpro 386/25e Model 60 Computer with EMODAL Software

The Deskpro 386/25e includes a 25 Mhz 386 microprocessor,
four (Mb)

of random access memory,

high-density 1.2 Mb,

density

r:
4.

'-

1.44

Mb,

a 60 Mb hard disk drive,

5 1/4 inch floppy disk drive,

1/2

inch

floppy

equipped

with an enhanced keyboard,

monitoL,

and a mouse.

93

and a high-

disk drive.
a COMPAQ Video

It

comes

Gr-aphics

The software used in

this

test

setup is

Entek Scientific

Corporation's EMODAL modal analysis software package.

be used as a stand-alone modal testing
collect and process data.

It

without peripheral devices;

be connected when

tool,

It

carn

or as a tool to

has been designed to run with or

however,

a spectrum analyzer must

acquiring measurement

data.

To allow the

EMODAL software to communicate with the spectrum analyzer, the

Deskpro 386/25e has had a National
installed,
More

which
specific

supplies

necessary

thc

information

Instruments AT-GPIB board

on

the

IEEE-488

EMODAL

software

obtained from the EMODAL-PC Instruction Manual

Bruel & Kjaar

(B&K)

of

several

different

complete exciter
up to a
portable
pounds,

but

it

with

optimum flexibility

uo 0.5 inches.
with

its

be

heads,

assembly,

rather

It

the base

capable of generating
It

at

in

is

not considered

any

weight

angle

to

of

180

provide

With the general purpose head,

capable of providing displacements up

has an internal natural frequency of 7200

having

More detailed information

natural

frequency

can be found in

Instructions and Applications booklet

Exciter

3].

resulting

substantial

positioned

and use.

the Type 4801 exciter is

hz,

exciter

(or shaker)

can

be

to he combined with any

100 pound excitation force.

shaker

[Ref.

can

Exciter Tyte 4801

The Tvne 4801 exciter was dsianed

one

interface.

System V Instruction Manual

94

[Ref.
[Ref.

of

10-14

hz.

tne manufacture-'s
16] and Vibration
17].

B&K General Purpose Exciter Head Type 4812

The Type 4812 general purpose head was designed to provide
a low distortion

operating band from DC to 10, 000 liz when used

with the Type 4801 exciter and the Type 2707 amplifier.
'g'

loads available for testing

vary with the test

and run from a low of zero in

1hz.

This

general

purpose

The

frequency

the DC range to 100g at 150-160

impedance

head has both

fotce

transducer and a response transducer incorporated into it,

is

used to obtain driving point measurements.

informati.on can be found

in

and Applications booklet

(Ref.

and

More detailed

the manufacturer's

Instruct ions

18].

B&K Power Amplifier Type 2707

The
proper

Type

2707

power

and

Protective
include

power

protection

circuits

power

amplifier
to

incorporated
phase

overtemperature,

designed

the

Type

into

protection,

displacement,

was

the

to

4801
power

signal

exciter.
amplifier

ground

overcurrent

and

provide

fault,
waveform

clip[ping indications.

More detailed information can be found

in

Instruction Manual

the manufacturer's

Type 2707

[Ref.

for Power Amplifier

19]

PCB Piezoelectronics 3'lexcel Modal Accelerometer Model 336A

I

The Model 336A modal accelerometer

vibratory
convenient

motion
display

into
and

measurable
analysis.
95

functions to transfer
voltage

signal

It

is

for

low-mass

acculcromctez

which serv'es to minimize mass-loading effects.

Characteristics

of

this

accelerometei_

important

analysis applications include high output,

phase

response,

importantly
response,

low transient

for the

missil'_

thermal

modal

modal
matched

and most

good

low frequency

provided by a composite construction which isolates

contains

amplifier,

built-in
a

it.

a resonant

constant

currenlt

Specifics

concerning

frequency

can

operating instructions

power

be

[Ref.

of

accelerometer

charge-to-voltage

unit

is

the Model

a sensitivity

(mounted)

information

PCB Piezoelectronics

Since this

microelectronic

frequency range of 1-2000 liz,

detailed

low noise,

sansitivity,

test,

from grouna and thermal transients.

operate

in

required

336A include

of 1000 mV/g,

greater than 9 kliz.

found

to

in

the

and
More

manufacturer's

20].

Dual--Mode Charge Amplifier Model 464A

The Model 464A is

a dual-mode device which can be operated

either as a charge amplifier

(for transducers without builc-in

microelectroniics)

Integrated-Circuit-Piezoelectric

(ICP)

constant

built-in

or as

current

an

power

source

microelectronics) .

(for

Operation

is

transducers
simple

switching to the mode of operation (charge or TCP),

the transducer sensitivity,
is

all

response

that

is

required.

from near

in

with
that

dialing in

and switching to the desired range

The Model

DC to 100

khz.

plug-in 180 khz low-pass filter

464A has

For this

test

was installed, as this

95

frequency
setup,

was all

that

was

availablie

at the time of. the test

180

khiz.

information

Moi~e detailed

instr~uct ions

niaiiutrcturer' s cperating

1Lowevto

varying .9izes, irauging from 1 khiz

thu [ilteLs a.Le availab~le ii.

to

seLup.

can be

LRef .

found

in

Lhe

211

PCB Piezoolectronic-n Mechanical Impedance Sensor Model 28BAll

4

~The

Model

is

288A11

mass,

low

com~binatio-i

transducer and- accelerometer which functions to transier

the

excitation

electrical

(f oxce)

andi

The2 built-in

signals.

conxverts

the. high impedancu

element

force

impedance
built-in

amplifier,

to operate it.
mass

gauge

vo.itagje

and

Since

this

to

by the. quartz

of

this

low

relatively

sensor

contains

a constant. current power unit is

Characteristics

(only 45 grams)

generated

int~o

amplifier

microelectronic

accelt!.eroweter

Loth

(miot ion1)

response

signals

signals.

sensor

required

include low

nominal sensitivity of 100 mV/g (miotioni)

and 1000 niiV/g (force)

detailed

information

operating

instructions

I9

the

foLC~e

and f requency range of 5-5000 hz.

can

be

[Rcf.

found
22].

in

the

more

manufacturer' s

LIST OF REFERENCES9

1.
Hollyer, J. D3., Ground Vibration C-haracte~r,.zation of a
M~issile Syrsteni for-Flutter Energy,Definition, Master's The3is,
Nz~val Postgraduate School, Monterey, Califorxnia, June 1990.
n.

asurt.iient

Sucua

Systemns,

Testing and

Basic Modal

Analysi~s Techniques, 19Q1.

3.

Entek

Scientific

EMODAL -PC

Cor-poration,

Instruction

4. bramei-, T.P .C., and cthers, Bas.c Vibration Control, Sound

Research Laboratories Linlited, pp. 117-121, 7977.
MacNeal-Schwendler Corp. , MS3C/PAL2 Advanced Stress and
5.
Vlibrat ion Anialysis Verusion 4. 0 Reference Manual, 1990.
Advanced Stress and
Corp.,M/AL
6.
MacN(--al-SchwendeLe
Vi brzat ion Analy~sis Versioni 4. 0 User's Manual, 1990.
'7.
Private corniunicat~ion
author-, 7 August 19191.

I9.

boetween

K~en

Richardson, Marx, Modial Analysis

8.
Systems, Hewlett: -Packard Com~pany.

Spectral

Institute,

Ranger,
Uising

MSC

and

Digit-il

the
Test

Modal Analysis Training Course and

Application No".:CS, p. 2-(, 199'..

10 .
American Society of Mecbanical Engineers, vibcation
Testing - lnistrurnentation and Data Analy~sis, prp. 37--38, 3.975.
11.
1C13 Pie z -electronics,
111st~ructi ons.

Inc.,

Petro

Wax

Model

080A--4

Testing
"Impact
D.L.,
and
Brown,
D.
12.
Corelli,
Cons iderat ions", E:xz::.rpt from ENTEK Scientific Cor~poration
Modal Analysis Training Course Notes, undated.
Entek Seient-ific Corporation, EMODAL-PC Learning Guide,
13.
1988.
Scientific
14.
Manual.

Atlanta,

SD-380

96

Signal Analyzer

Operator's

15.
Scott, W.B.,
Missile Upgrades",
53, 22 July 1991.
16.

Eruel

"Experience From Desert Storm May Speed

Aviation Week & Space Technology, pp. 52-

& Kjaer

(B&K),

Instructions

and Applications -

Exciter Body Type 4801.

17.

B&K,

Instruction

Manual

- Vibration Exciter

System V.

18. B&K, Instructions and Applications - General Purpose Bead

Type 4812.
19.

B&K,

Instruction

Manual - Power Amplifier Type 2707.

20.

PCB Piezoelectronics,

Inc.,

Transducer Instrumentation

Operating Instructions - FLEXCEL Accelerometer 336A Series.

21.

PCB Piezoelectronics,

Operating
464A.
22.

Instructions -

PCB Piezoelectronics,

Inc.,

Electronic

Instrumentation

Dual Mode Charge Amplifier Model

Inc.,

Transducer Instrumentation

Operating Instructions - Impedance Head Model 288A11.

99

BIBLIOGRAPHY

1.

American Society of Mechanical Engineers

Testing

Instrumentation

Engineering Center,

and

Data

(ASME) , Vibraticn

Analysis,

United

1975.

2.
Cook, R.D., Malkus, D.S., and Plesha,
Applications
of Finite
Elemrent Analysis,
& Sons, Inc., 1989.

M.E., Concepts and

John Wiley
3 r` ed.,

3.
Elishakoff, I.
and Lyon, R.H., Random Vibration and Recent Developments, Elsevier Science Publishers,
4.
Harris,
C.M.,
and Crede,
C.E.,
Handbook, MCGraw-Hill Book Co.,
1976.
5.

Shock

and

Status
1986.

Vibration

Naval Air Systems Command Technical Manual NAVAIRO1-AIM9-

2, Sidewinder Guided Missile AIM-9HiL/M Organizational and

Intermediate Maintenance With Illustrated
Parts Breakdown.
6.

Naval Air Systems Command Technical Manual NAVAIROI-75PA-

75, Airborne Weapons/Stores Loading Manual,

41 9.
Aircraft, 15 April
7.

Thomson,

Prentice-Hall,
8.
Tse,
Vibrations
1978.

W.T.,
Inc.,

Vibration

Thecry

and

Navy Model P-3

Applications,

1965.

F.S.,
Morse,
I.E.,
and
Theory and Applications,

100

Hinkle,
R.T.,
Mechanical
Allyn and Bacon, I-ac.,

INITIAL DISTRIBUTION LIST

1.

Defense Technical Information Center

Cameron Station
Alexandria, VA 22304-6145

2.

Library, Code 52
Naval Postgraduate School
Monterey, CA 93943-5002

3.

Commander (AIR-5114B)
Attn: CDR Boyington
Naval Air Systems Command
Washington, D.C. 23061-0001

4.

Department Chairman
Department of Aeronautics
Naval Postgraduate School
Monterey, CA 93943-5000

2
and Astronautics

5.

Commander, Naval Air Test Center

Attn: Mr. Tim Twigg
FWATD, Code FW52B

6.

Commander, Patrol Squadron 47

Attn: LCDR R. Bobbitt

FP0 San Francisco,

CA 96601-5920

7.

Mr. Michael Shutty

Naval Air Test Center
FWATD, Code FW522MS
NAS Patuxent River, MD 20670

8.

Professor Edward Wu
Department of Aeronautics
Naval Postgraduate School
Monterey, CA 93943-5000

3
and Astronautics

101

AI