TECHNICAL REPORT ARCCB-TR-95011

RIFLING TWIST DESIGN

n~fL~

S~~i1~r

F

ROYCE W. SOANES

FEBRUARY 1995

US ARMY ARMAMENT RESEARCH.

o EVE LOP MEN TAN 0 ENG I NEE R I N GeE N tE R CLOSE COMBAT ARMAMENTS CENTER

,

BENET LABORATORIES

WATERVLIET, N.Y. 12189-4060

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED

19950608 074

,

DESTRUCTION NOTICE

For classified documents, follow the procedures in 000 S200.22-M, Industrial Security Manual, Section 11-19 or DoD S200.1-R, Information Security Program Regulation, Chapter IX.

For unclassified, limited documents, destroy by a~y method that will prevent disclosure of contents or reconstruction of the document.

For unclassified, unlimited documents, destroy when the report is no longer needed. Do not return it to the oriiinator.

DISCLAIMER

The findings in this report are not to be construed as an official Department of the Army position unless so designated by other authorl%ed documents.

The use of trade name(s) and/or manufacturer(s) does not constitute an official indorsement or approval.

REPORT DOCUMENTATION PAGE Form Approved •
OMS No. 0704-0188
Public reoonmg bu"oen for thl) couect.on 01 In'formatlon I~ E!'Strmat@ld to alo'@rage 1 hour per res.ponse. including th@ tim@for reoviewlng instructions,_ searching e'ltlstlng data sources.
gat~er,e.g and m.,nta,nlng t~e data npeded. and completing and reviewing tne collection of Information. Send comments :r.arding this burden estimate or any other aSPeCt of thi>
couect.on ot mtorrnanon, ,ncludlng ,ugg",tlon, tor reduCIng ttus burden. to Wa,nlOgton HeadQuarters ServICes. DirectOrate or Information Operatlom. and Reports. 1215 Jeffenon
Dav" Hlgt-"'ay. Su,'< 1204. Arlington, VA 22202-4302, and to tne Office of Management and Budget. Paperwork Reduction PrOject (07~188), Washington. DC 20503.
1. AGENCY L':' E ONLY (Leave blank) 12. REPORT DATE 13. REPORT TYPE AND DATES COVERED
February 1995 FmaI
4, TITLE AN ;UBTITLE 5. FUNDING NUMBERS
RIFLING 1WIST DESIGN
AMCMS: 6111.02.H611.100
6. AUTHOR(S)
Royce W. Soanes
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) •• PERfORMING ORGANIZATION
U.S. Army ARDEC REPORT NUMBER
Benet Laboratories, AMSTA-AR-CCB-O ARCCB-1R-9S011
Watervliet, NY 12189-4050
9. SPONSORING I MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING
AGENCY REPORT NUMBER
U.S. Army ARDEC
Case Combat Armaments Center
Picatinny Arsenal, NJ 07806-5000
11. SUPPLEMENTARY NOTES
12a, DISTRIBUTION / AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Approved for public release; distribution unlimited
13. ABSTRACT (MaXImum 200 words)
This report considers the efficient solution of the following problem: given pressure travel shape data for a particular round.
compute the rifling cUIVe which will produce a projectile torque cUIVe of virtually any desired shape:
14RlHBJECT TERMS 15. NUMBER OF PAGES
. ing, Twist, Progressive, Gain, Torque 32
16. PRICE CODE
17. SECURITY CLASSIFICATION 1 S, SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF REPORT OF THIS PAGE OF ABSTRACT
UN ClASSIFIED UNCIASSIFIED UNCLASSIFIED _ill
- . . , I NS~ 7540-0,-2BO-550u

Sta nnarc r or rr <98 ,~e .. 2-89) P~,?,~:rlO€"O tl\ AI\j)L St: i~~·'t

2gB·' "2

TABLE OF CONTENTS

ACKNOWLEDGEMENT................................................... ii

INfRODUCTION 1

RIFLING SYNTHESIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

FIGURES OF MERIT 8

PRESSURE TRAVEL FUNCTIONS , 10

RIFLING ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 19

REFERENCES , 21

APPENDIX 22

ii

ACKNOWLEDGEMENT

The author would like to thank David Finlayson of Benet Laboratories for suggesting that this rifling analysis focus on projectile velocity, since (for a given round) it is roughly invariant with respect to any form of rifling.

INTRODUCTION

There are many aspects to the design of rifling in guns. Here, we are concerned with only one--specification of the rifling angle or slope to produce a desired projectile torque. A rigid gun and a pre-engraved projectile are assumed. No frictional, thermal, or plastic effects are considered.

Constant twist rifling has been in use ever since the earliest rifled cannon, and in early cannon which used pre-engraved projectiles, constant twist rifling was mandatory. With the advent of the soft metal rotating band, however, it became possible to vary the rifling angle for various reasons.

Progressive twist rifling, for instance, can be used to reduce projectile torque near the origin of rifling where the base pressure is highest, while constant twist rifling produces a torque which is proportional to the base pressure. Progressive twist can be particularly important after wear has taken place near the origin of rifling. Progressive twist can still engrave the rotating band reliably, while the velocity gained by the projectile over the worn section of a constant twist gun can ruin the rotating band as it hits the unworn area.

Unfortunately, most efforts at variable twist rifling design have used analytical forms for the rifling curve and then computed the resulting torque in retrospect. This trial and error method may have been convenient before the advent of computers, but it is now an unnecessarily restrictive design technique. For instance, although traditional progressive twist rifling does decrease torque near the origin of rifling, it raises the torque considerably near the muzzle, whereas constant twist rifling has a considerably lower torque near the muzzle due to the lower base pressure there.

It would seem that we cannot get everything that we want, namely lower torques near the origin of rifling and no torque at all as the projectile exits the muzzle. Fortunately, what seems to be the case is not the case at all. In fact, if we are prepared to spend a few seconds of computer time doing some simple numerical integrations, we can have virtually any shape torque curve we want. Another comment is perhaps in order for the uninitiated: although the purpose of rifling and the torque produced thereby is to spin stabilize the projectile in flight for far greater accuracy, the shape of the torque curve has absolutely nothing whatsoever to do with the amount of spin imparted to the projectile in flight. The only determinant of this spin is the slope of the rifling curve at the muzzle. As long as this slope at this single point is held constant, the exit spin will remain constant and we are free to pick virtually any shape torque curve we want.

1

Dist

Avail ai·"r~/cr Special

Accesion For \ J

~----------\---

NTIS CRA&I ~ I

OTIC T.~B [_J I

Unannounced [J j

Justitica Lon

By _ ._=~~:.~:~~.~~~_~~.=J

Distribution f I

1------.--.-------- . ..;

Availabi~t~_~~~~.:~_ ... J

A-}

RIFLING SYNTHESIS

Let P(x) be the pressure travel curve, A be the bore area, r be the bore radius, and m be the projectile mass. The movement of the projectile through the rifled bore is described by the following equations:

PA -nSin4J=m dv dt

where n is the collective normal force on the lands, <I> is the rifling angle, 8 is the amount of rotation of the projectile, and i is the projectile moment of inertia. L is the torque applied to the projectile and v is its velocity.

Let f(x) be the rifling curve which is related to the amount of rotation by f(x);;;:r8(x)

First,

de de v /, -;;;:v-;;;:-f(x)

dt dx r

d2e v d /,

- =- -(vf(x»

dt2 r dx

and

2

dv dvldz

-;;;:v-;;;:--v

dt dx 2dx

Hence, the equations of motion become

AP(x) - n Sin t/J _1 m!!._v2 z dx

L;;;:nrCost/J;;;:iv !!'_(vfl(X» rdx

Solving for n in the second equation gives us

iv d (vfl(X»

r2Cos¢dx

n

3

Inserting this in the first equation yields

. d d

AP(x) - IV Tan ~_(vf'(x» = !.m-v2

,2 dx 2 dx

but

Tan~=t'(x)

Hence, we have

or

2A d(' 1

-P(x) = - v2 +-'-(vf'(x)l

m dx mr2

Integrating this, we have

where R is the projectile radius of gyration. Hence, projectile velocity is given by

v(x) =

Let I be the length of travel (of the rotating band). P(x) is the projectile base pressure when the rotating band is at x. A good typical value for

is 0.1. We therefore see that the muzzle velocity of a rifled gun is about one-half of one percent less than the smoothbore muzzle velocity. Indeed we can see as Finlayson points out that the projectile velocity is roughly invariant with respect to virtually all forms of rifling (ref 1).

4

Now, define the torque function to be the product of some constant and some torque shape function:

L(x)= el(x)

Recalling that

We have

and integrating this equation, we get

fl(x) = e.L rJ: l(u) du i vex) 10 v(u)

Now let

lex) = ~ ;:"P(u)du

and

Hence, v is given by

v(x) = ~ 2A lex) m U(x)

and therefore

fl(x) = emr U(x) (;r; l(u)U(u) du

2A i lex) 10 leu)

Letting

J(x) = /? A(u)U(u) du

10 I(u)

We have

fl(x) = cr U(x)J(x) 2AR2 J(x)

And hence, when f (I) has been specified, c is defined as 2A R'1I(l)J(l)

c ----"-~-'-'-

rU(l)J(l)

Therefore

jl(X) = fl(l)J(I)U(x)J(x) U(I)J(I)I(x)

Rewriting this last equation in the form

where

J(l)l(x) g(x) = J(1)I(x)

we have

which we can solve for f (x), getting

5

6

Now, for the sake of numerical accuracy in computing the integral J(x) (especially for small values of x), it is necessary to perform an integration by parts on J. Recall that J is given by

J(x) =;:% A(u)lJ(u)dU

o l(u)

We must have

A(O)=O

for this integral to exist, since I is proportional to x for small values of x. Since

we have

Hence,

J(x) =;:% 2 A(U)lJ(U) J1(u)dU :;: PK(u)dI(u)

o p(u) Jo

where

K(x) = 21(x)lJ(x) p(x)

Note that

by I'Hospital's rule.

Using integration by parts on J, we have

J(x) = K(x)J(x) - K(O)I(O) - S(x) = K(x)/(x) - Sex)

where

The Stieltjes integral S should be computed directly by trapezoidal rule because K may not be differentiable. Indeed, K need not even be continuous.

Now, in order to compute g(O), it is sufficient to note the asymptotic behavior of ,t, P, I, and J for small values of x:

A(X) - x AI(O) P(x) - Xpl(O)

l(x)2: (XP(u)du. - !..X2pl(0)

10 2

J(x)- X11(0)U(0)~ 2 p'(O)

We therefore have

J(x) _ 211(0)U(0) :K(O)

I(x) pl(O)

and

(0): K(O)I(I)

g J(I)

We now summarize in one place the equations applicable to rifling synthesis.

P(O) =0 1(0)=0 I(x) = ~ J;xp(u)du.

U(x) = 1 { ~f'(X) r

K(x) = 2.l(x)U(x) P(x)

K(O): 211(0)U(0) pl(O)

Six)» foxJ(u)dK(u) J(x) = K(x)J(x) - Sex)

7

g(x) = i:~~=~

(0) = K(O)J(l)

g J(/)

r(x) = JI(l) U(X) g(X) U(l)

f1x) = f1Qg(x)

1 {~f'(l)r(l-g(xY~)

fix) = fojl(u)du

The equations defining f (x) do so implicitly because f (x) enters the right-hand sides through U. Functional iteration is used to obtain f (x). Convergence is extremely rapid, taking only a few iterations. In fact, if we take f (x)=f (I) initially, the very first iteration gives a result which is in error by only about one quarter of a percent for most reasonable A.'s. The reason for this is that U is only very weakly dependent on f (x), since the square of the latter is small relative to unity. Hence, U is roughly a constant near unity for all practical rifling configurations.

Also note from the equations that for the first iteration, U is a constant which cancels out. Hence the rifling slope function is determinable within about a quarter of a percent error without knowledge of either the projectile radius or radius of gyration. The absolute values of pressure are also unnecessary. Only the shapes of the pressure travel function and the desired torque function are necessary to determine a good approximation to the rifling configuration for any size gun with any muzzle twist and any projectile regardless of its mass or moment of inertia!

FIGURES OF MERIT

8

Let us take

Jt(x) = kP(x)

For the first (and last) iteration, set

fl(x)=r(l)

Hence,

U(x) == const ==U K(x)=2kU=const

S(x) == ;;"ldK =0 J(x) == 2 k U l(x) (x)== 1(1)2kUJ(x) =1

g 2kUl(I)J(x)

Therefore

JI(X) == JI(l)

(immediate convergence to constant twist rifling).

Now, recall that

where

2A R2JI(1)/(1)

c ----"--'-'---"-!..

rU(l)J(l)

Denote the max norm by I g. The maximum torque force will then be

and the maximum pressure force is /PUA·

We define JJ to be the ratio of maximum torque force to maximum pressure force:

9

10

For the constant twist case (AocP), we have

_ 2klP ~rf'(I)J(I) _ (~rf'(l) Pc - jPIU(I)2kU(l)J(I) - U(ll

We can also define another figure of merit (a) as the ratio of maximum torque to maximum constant twist torque:

q:_E_ 2 I). IJ(l)U(l)

/Jc IPIJ(l)

Note that a=l in the constant twist case and that a is virtually independent of R, r, andf(l). We can now compare different torque shape functions by comparing their /J's and O"s.

We would presumably like these figures of merit to be acceptably small (up to a point).

Now, the smallest values for /J and 0' one can typically get (depending on the pressure travel curve shape) are about 0.03 and 0.45, respectively. We get these rather small values by letting A be constant except in a small neighborhood of x=O where A' must be quite large. Unfortunately, the byproduct of this A is an f having f (0) considerably larger than f (l), an obviously impractical situation. However, if we let A be proportional to P up to maximum pressure and constant everywhere else except for the last five percent of travel where we drop A linearly down to zero, we get a JJ of about 0.04 and a 0' of about 0.6 with an f (0) of about 0.6 f (l). The rifling is therefore progressive, but not radically so, and it substantially reduces maximum torque on the projectile while giving it a torqueless exit.

PRESSURE TRAVEL FUNCTIONS

Ideally, one might want to work with the best available pressure travel data for a given round; however, the need will generally exist to fire more than one type of round or charge in

the gun. If each round type displayed the same shape pressure travel function, the rifling configuration could be determined uniquely. In general, however, the pressure travel function will vary in shape for rounds of different types and it will be necessary to use some representative or average pressure travel shape function. This being the case, it makes sense to approximate

the pressure travel data with some smooth closed form function.

If the reader has some favorite numerical method for approximating pressure travel data, this section can be completely ignored. In any case, no claim of superiority is made for the simple approximation offered here; it at least suffices for testing the software associated with the problem at hand.

A simple approximation one might use is

P(x)=axe iir

To determine the three parameters a, b, and c, one might enforce the three conditions P(X1)=Pl

pl(X1) = 0

P(X2)=P2

This would allow us to put the peak pressure (PI) where we wanted it (Xl) and specify the pressure (P2) at the muzzle (x2)·

This representation seems insufficiently flexible, however, so the following piecewise exponential representation is suggested:

where

Ir~2)

P (x) = a -b e \ c2

2 2 2

PlX1)=Pl

I Plx1)=O

PlX2)=P2 PlX2)=P2 I I

Plx2)=Plx2)

PlxJ=P3

11

Determining the parameters in p}(x):

-

1

and

1 e PIe 1

P/X1)=PI - a]=-x]

We therefore have

in which we must now determine C}'

We have the order relationships

12

Therefore

13

or

Letting

We have the equation which must be solved for C1:

Define function / as

Clearly,

and since a > 1,/(00)=00.

Differentiating, we have

j'(x) = rrlna+p jll(X} = rr(lna)2 > 0

Hence,

Therefore, since/(O)=O,f(O) < O,f(x) > 0 and/(oo)=oo, a positive zero of/is guaranteed. Also, there must be an m such thatf(m)=O. Hence,

-2f(m) fll(m)

Expandingfin a Taylor series around m, we have

fix) t::: f(m) +¥II(m)(x -ml

Hence,

/(z)=o -

This is our first estimate of z (or cj) which we will subsequently use in Newton iteration to obtain z exactly.

Now,

k= -/J _ Ina

Ink

m=-

Ina

/JInk

f(m) = a'" -Bm -1 =emma +-- -1 =k-klnk-I

Ina

fl/(m) = a'"(1n a)2 =emlna(ln a)2 =k(1n a)2 -2f(m) 2(lnk+i-1)

111(m) (In al

Therefore the first estimate of z reduces to

zz_1_(Ink+J2(Ink~ -1»

Ina j;

And the Newton iteration for z is

/(z) rr+/Jz-1 rr(zIn a-I) + 1

z .... z--=z-

fl(z) filln a+/J rrIn a +/J

14

15

Now, determining the parameters of P2(x):

P (x)=a -b e f:)

222

So,

and

Therefore, we have

Now,

which must be solved for c2•

Define/by

Clearly,

-- --------------------------------- ...

16

and for large values of x,

~

-_ 0

e x -1-x

Therefore,

Now,

and since,

~ lim ~=O x-o x

we have

Also, for large values of x,

s

-_ 8

t :;c -1-x

Furthermore,

Therefore, f is monotone increasing and can never be positive. We can therefore finally conclude that f is monotone decreasing and is guaranteed to have exactly one positive zero provided

The Newton iteration for [(z) =0 is

Taking z=O initially, we have

as our first guess at z,

Computing

4 z/I(1.)-f(Z);;;;P3-P2-8p;e %

we have the Newton iteration for the correct z:

Summarizing the results of this section:

Pix) if 0 s: x s: x2 P(x) = Pix) if x2 < X s X3

X '1

g(x):;;; 1-(-)

Xl

17

-------------------------------------------

/3=ljX1P2) '1XV'1

k=i

Ina

aC1(c1lna-l)-rJ aC1lna+/3

18

C2 ...

pi [1-(1 + :'r -~ 1

19

RIFLING ANALYSIS

For the sake of completeness, we briefly derive the equations to compute torque and pressure in terms of rifling twist. Recalling that

L = iv ~(v!,(x»= i 2vj'(x)~(vj'(x»

r dx 2rj'(x) dx

= i d (vj'(X»2

2r!'(x) dx

and further recalling that

>(x): ~ 2A I(x) m U(x)

we have

L(x)= i d (2Af'(X)2](X)2)

2rj'(x) dx m U(X)2

AR2 d /,(xi Jo1Cp(U)du.

=

r/,(x) dx (R/'(X»)2

1+--

r

For constant twist rifling, where f(x)=constant=k, we have L(x)= nkrR2 P(x)

l+(k:r

If one wished to write P in terms of L, one could begin with P :!!!'_'!!'_(U2v2) 2Adx

and since

d 'f'< rL

-(v x»:-.

dx rv

20

we have

2r u.x)f/(X)

2vf'(x) ~ (vf'(x»;;:; i

;;:; '!!__(Vfl(X»2

dx

which gives us

and hence that

REFERENCES

1. David Finlayson, Private Communication, U.S. Army ARDEC, Benet Laboratories, Watervliet, NY, 1993.

21

APPENDIX

22

-

m ~ ~ ~ ~ ~ .~ N ~ 0

. . . . . . . . .

9JnSS9Jd

23

-

~ m ~ ~ ~ ~ M N ~ 0

. . . . . . . . .

anbJo.l

24

~--------------------~~

~ 0

.

aA.lnO 5u!ltJl::I !O adolS

-

-

CD

.

It)

.

N ~

. .

o

anbJo..L

26

-

,.. 0

•

al\Jn~ 5u!I!!CJ !O adolS

27

-

~ ~ ~ ~ ~ M N ~ 0

. . . . . . . .

anbJol.

28

-

,.... 0

.

aAJn8 5u!l!!l:J !O adOIS

29

-

....

•

o

TECHNICAL REPORT INTERNAL DISTRmUTION LIST

NO. OF COPIES

CHIEF, DEVELOPMENT ENGINEERING DIVISION

ATIN: AMSTA-AR-CCB-DA -DB

-DC

-DD

-DE

1 1 1 1 1

CHIEF, ENGINEERING DIVISION

ATIN: AMSTA-AR-CCB-E -EA

-EB

-EC

1 1 1

CHIEF, TECHNOLOGY DIVISION

AnN: AMSTA-AR-CCB-T -TA

-TB

-TC

2 1 1 1

TECHNICAL LIBRARY

AnN: AMSTA-AR-CCB-O

5

TECHNICAL PUBLICATIONS & EDITING SECTION ATIN: AMSTA-AR-CCB-O

3

OPERATIONS DIRECTORATE AnN: SMCWV-ODP-P

1

DIRECTOR, PROCUREMENT & CONTRACTING DIRECTORATE AnN: SMCWV-PP

1

DIRECTOR, PRODUCT ASSURANCE & TEST DIRECTORATE AnN: SMCWV-QA

1

N01E: PLEASE NOTIFY DIRECfOR, BENET lABORATORIES, ATTN: AMSTA·AR-CCB-O OF ADDRESS CHANGES.

TECHNICAL REPORT EXTERNAL DISTRmUTION LIST

NO. OF COPIES

ASST SEC OF TIlE ARMY RESEARCH AND DEVELOPMENT

ATTN: DEPT FOR SCI AND TECH TIlE PENTAGON

WASHINGTON, D.C. 20310-0103

ADMINISTRATOR

DEFENSE 1ECHNICAL INFO CENTER 2

ATTN: DTIC-OCP (ACQUISmON GROUP) BLDG. 5, CAMERON STATION ALEXANDRIA, VA 22304-6145

COMMANDER

U.S. ARMY ARDEC

ATTN: SMCAR-AEE 1

SMCAR-AES, BLDG. 321 1

SMCAR-AET-O, BLDG. 351N 1

SMCAR-FSA 1

SMCAR-FSM-E 1

SMCAR-FSS-D, BLDG. 94 1

SMCAR-IMI-I, (STINFO) BLDG. 59 2

PICATINNY ARSENAL, NJ 07806-5000

DIRECTOR

U.S. ARMY RESEARCH LABORATORY

ATTN: AMSRL-DD-T, BLDG. 305 1

ABERDEEN PROVING GROUND, MD

21005-5066

DIRECTOR

U.S. ARMY RESEARCH LABORATORY

ATTN: AMSRL-WT-PD (DR. B. BURNS) 1

ABERDEEN PROVING GROUND, MD

21005-5066

DIRECTOR

U.S. MATERIEL SYSTEMS ANALYSIS ACTV ATTN: AMXSY-MP 1 ABERDEEN PROVING GROUND, MD

21005-5071

NO. OF COPIES

1

COMMANDER

ROCK ISLAND ARSENAL ATTN: SMCRI-ENM

ROCK ISLAND, IL 61299-5000

1

MIAC/CINDAS PURDUE UNIVERSITY P.O. BOX 2634

WEST LAP A YETTE, IN 47906

1

COMMANDER

U.S. ARMY TANK-AUTMV R&D COMMAND ATTN: AMSTA-DDL (TECH LmRARy) 1 WARREN, MI 48397-5000

COMMANDER

U.S. MILITARY ACADEMY

ATTN: DEPAR1MENT OF MECHANICS 1

WEST POINT, NY 10966-1792

U.S. ARMY MISSILE COMMAND

REDSTONE SCIENTIFIC INFO CENTER 2

ATTN: DOCUMENTS SECTION, BLDG. 4484 REDSTONE ARSENAL, AL 35898-5241

COMMANDER

U.S. ARMY FOREIGN SCI & TECH CENTER ATTN: DRXST-SD 1 220 7TH STREET, N.E.

CHARLOTTESVILLE, VA 22901

COMMANDER

U.S. ARMY LABCOM

MATERIALS TECHNOLOGY LABORATORY ATTN: SLCMT-IML (TECH LmRARy) 2 WATERTOWN, MA 02172-0001

COMMANDER

U.S. ARMY LABCOM, ISA ATTN: SLCIS-IM-TL

2800 POWER MILL ROAD

ADELPHI, MD 20783-1145

1

NOTE: PLEASE NOTIFY COMMANDER, ARMAMENT RESEARCH, DEVELOPMENT, AND ENGINEERING CENTER, BENET LABORATORIES, CCAC, U.S. ARMY TANK-AUTOMOTIVE AND ARMAMENTS COMMAND, AMSTA-AR-CCB-O, WATERVLIET, NY 12189-4050 OF ADDRESS CHANGES.

TECHNICAL REPORT EXTERNAL DISTRIDUTION LIST (CONT'D) NO. OF

COPIES

COMMANDER

U.S. ARMY RESEARCH OFFICE

ATfN: CHIEF, IPO 1

P.O. BOX 12211

RESEARCH TRIANGLE PARK, NC 27709-2211

DIRECTOR

U.S. NAVAL RESEARCH LABORATORY

ATfN: MATERIALS SCI & TECH DIV 1

CODE 26-27 (DOC LIDRARY) 1

WASHINGTON, D.C. 20375

NO. OF COPIES

WRIGHT LABORATORY ARMAMENT DIRECTORATE ATfN: WL/MNM

EGLIN AFB, FL 32542-6810

WRIGHT LABORATORY ARMAMENT DIRECTORATE ATfN: WLJMNMF

EGLIN AFB, FL 32542-6810

1

1

N01E: PLEASE NOTIFY COMMANDER, ARMAMENT RESEARCH, DEVELOPMENT, AND ENGINEERING CEN1ER, BENET LABORATORIES, CCAC, U.S. ARMY TANK-AUTOMOTIVE AND ARMAMENTS COMMAND, AMSTA-AR-CCB-O, WATERVUET, NY 12189-4050 OF ADDRESS CHANGES.