AD-A242 988

I~
IIIIII~
IIIil~ IItl II ______________ _______________
-®

TECHNICAL REPORT BRL-TR-3289

DTIC
B3 R1 L s 99
urr

DESIGN AND ANALYSIS OF

KINETIC ENERGY PROJECTILES
USING FINITE ELEMENT OPTIMIZATION

BRETT R. SORENSEN

NOVEMBER 1991

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION IS UNLIMITED. "

U.S. ARMY LABORATORY COMMAND

BALLISTIC RESEARCH LABORATORY

ABERDEEN PROVING GROUND, MARYLAND
 NOTICES

Destroy this report when it is no longer needed. DO NOT return it to the originator.

Additional copies of this report may be obtained from the National Technical Information Service,
U.S. Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161.

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

The use of trade names or manufacturers' names in this report does not constitute indorsement
of any commercial product.
 Ui~v
Mh ~)________"__

ohc
REPORT DOCUMENTATION PAGE
tfi.OolIng bujen fof this collection oft•"tOrnatiOr' is stimated to averaq•
gamfer4g an maintainig the data needed, &n completing andreviewing the collectmon
hour d•r
Of
resoorse. including time
informtiOn en
[ Form Approved
No.n070-0o18
for reviewing instructiOns. searhcning
existing cata sources.
Ominents regarding this b•rden estimate or anv Other aspideOf this
including suggietions for reducing this ourcden to Washington 'eadauartersServices. Directorate for information Operations and Reveorts. 12 1 JeSlerson
cO~llctin Of *informatlon,
Davis Highwy. Suite 1204. Arlington. VA 22202-4302. and to the Office of Management and Budget. Parwwork Reduction Prolect(0704-0l11).WlWashngtOn, DC 20S03

1. AGENCY USE ONLY (Leave bWank) 2. REPORT DATE 13. RE PORT TYPE AND DATES COVERED
I November 1991 I Final, 1 Jan 90 - 1 Jul 90
4. TITLE AND SUBTITLE S. FUNDING NUMBERS

Design and Analysis of Kinetic Energy Projectiles Using Finite Element PR: IL162618AH80
Optimization
6. AUTHOR(S)

Brett R. Sorensen

7. PERFORMING ORGANIZATION NAME(S) AND AODRESS(ES) 8. PERFORMING ORGANIZATION

REPORT NUMBER
U.S. Army Ballistic Research Laboratory
ATTN: SLCBR-TB-P
Aberdeen Proving Ground, MD 21005-5066

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING

AGENCY REPORT NUMBER

U.S. Army Ballistic Research Laboratory

A'TN: SLCBR-DD-T BRL-TR-3289
Aberdeen Proving Ground, MD 21005-5066

11. SUPPLEMENTARY NOTES

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release; distribution is unlimited.

13. ABSTRACT (Maximum 200 words)

This report discusses the minimization of the parasitic mass of a 120-mm cannon launched kinetic
energy projectile. The purpose of the study was to design a minimum mass aluminum sabot to launch both
depleted uranium and tungsten heavy alloy penetrator materials for the M829 penetrator geometry. The
minimization was conducted by implementing finite element techniques. A parametric model of a kinetic
energy projectile using a double ramp traction sabot was constructed and an input stream for the ANSYS
engineering analysis software was created. This input stream created the mesh for the projectile, solved
for the stresses for each penetrator material, and implemented the optimization capabilities within ANSYS
to minimize the mass of the sabot. Twenty-five iterations were required to reach a local minimum of the
objective function (sabot mass) and resulted in a 15% decrease in sabot mass. The entire process (initial
design and evaluation, optimization, and final analysis) was completed in one day (9 hours wall clock and
2 man-hours) on a personal workstation.

14. SUBJECT TERMS 15. NUMBER OF PAGES

28
kinetic energy projectiles; sabots; sabot design; finite element analysis; 16. PRICE CODE
optimization
17. SECURITY CLASSIFICATION
OF REPORT I 18. SECURITY CLASSIFICATION
OF THIS PAGE
19. SECURITY CLASSIFICATION
OF ABSTRACT
20. LIMITATION OF ABSTRACT

UNCLASSIFIED UNCLASSIFIED I UNCLASSIFIED SAR

NSN 7540-01-280-5500 Standard Form 298 (Rev 2-89)
UN LASSIFIEDPrescribed ft"ANSI$to Z1.
uUNCLASSIFIED 961
INTENTIONALLY LEFT BLANK.
 TABLE OF CONTENTS

Page

LIST OF FIGURES .......................................... v

LIST OF TABLES ........................................... v

1. BACKGROUND ........................................... 1

2. INTRODUCTION .......................................... 3

3. MODELING ASSUMPTIONS .................................. 4

4. PARAMETRIC MODEL ...................................... 6

5. DESIGN PROCESS ........................................ 9

5.1 Initial Design ........................................... 9

5.2 O ptim ization ........................................... 9

6. RESULTS ............................................... 12

7. CONCLUSIONS ........................................... 16

8. REFERENCES ............................................ 17
APPENDIX: PARTIAL LISTING OF THE ANSYS INPUT DECK ........ 19

DISTRIBUTION LIST ....................................... 25

•Ace..s3 Wg'

Dl st rtj lw 1til./

IAWMUJ~ii iyolfsr
//0 ' l .t l~ iIV

III V \
INTENTIONALLY LEFT BLANK.

iv
 LIST .OF FIGURES

Figure Page

1. Isometric View of a KE Projectile Fired From a Tank Cannon ............. 1

2. Pressure, Displacement, Velocity and Acceleration Curves for a

KE Projectile ............................................... 2

3. Axisymmetric Profile of a Penetrator and Sabot Designed by KEPDEP

and the Finite Element Mesh Generated by ANSYS via the KEPDEP
Interface ..................................................

4. Parametric Model of a KE Projectile and the Resulting Finite

Elem ent Mesh .............................................. 8

5. Flow Chart of the Design Logic Beginning With the Initial Design
and Progressing Through the Optimization and Final Analysis ........... 11

6. Results of the Objective and State Variables From the Optimization

Process ................................................... 14

7. Results for the Final Analysis for Each Penetrator Material Using
the "Best" Sabot ............................................ 15

LIST OF TABLES

Table Page

1. Material Properties ............................................ 3

2. KE Projectile Parameters ....................................... 7

v
INTENTIONALLY LEFT BLANK.

vi
 1. BACKGROUND

As emerging technologies and materials advance tomorrow's heavy armor threat, the need
to develop improved cannon-launched kinetic energy (KE) projectiles continues to be an
Important issue within the Army. Historically, as the armor protection levels have increased,
penetrator aspect ratio (length-to-diameter), mass, and velocity have also increased. The
current trends in modern antiarmor KE ammunition have been in high length-to-diameter ratio
(L/D), fin-stabilized penetrators utilizing discarding sabots. An example of a fielded projectile,
the M829, is depicted in Figure 1. Figure 1 presents an isometric view of the penetrator, three
of four sabot petals, and associated hardware (windshield, fins, and obturator).

Figure 1. Isometric View of a KE Proiectile Fired From a Tank Cannon.

gFin
S

Penetrator

Windshield

A long rod penetrator is essentially a right circular cylinder with an aspect ratio of 10 or
larger and is made of a high density material. The geometry requirement of a cylinder is not
absolute. Concessions are made to attach the windshield and fin, and more importantly, to
provide an interface between the penetrator and the sabot. This interface transfers the
pressure incident on the sabot to the penetrator. It consists of annular buttress grooves in the
forward section of the penetrator/sabot interface and a friction drive (e.g., fine threads) in the
aft section.
 The sabot is made of aluminum (or some other low density material) and consists of three
or four sections (petals). When the sabot is assembled around the penetrator, a one-piece
plastic obturator is pressed onto the sabot. This provides a seal between the sabot and the
cannon to prevent combustion products from leaking and also provides radial compression to
hold the sabot on the penetrator prior to placing the projectile into the cannon.

The sequence of events for a launch is as follows. The propelling charge is ignited and
begins to bum. Pressure inside the cannon builds rapidly as the projectile starts to move,
accelerating the projectile along the length of the cannon toward the muzzle. Typical
pressure, displacement, velocity, and acceleration histories experienced by a KE projectile are
plotted in Figure 2. As the projectile exits the muzzle, the radial constraint of the cannon is
removed and the high-pressure gases exiting the cannon cause the obturator to fracture (hoop
failure). As the projectile enters the ambient atmosphere, aerodynamic forces on the sabot,
along with affects from the gun gases, cause the sabot petals to separate and disengage from
the penetrator. After the sabot discards, the lethal mechanism (penetrator, windshield, and
fin) travels down range to the target.

500 -o70 1 1750

400- -/ aO~ ]4-1500. o

n *# -125
%300 40I I,12
X V/
2 30 2/ 750

o00 o >.
0 1 2 3 4 5 6 0 2 3 4 5 6
Time (msec) Time (msec)

Figure 2. Pressure, Displacement, Velocity and Acceleration Curves for a KE Prooectile.

An automated design technique has been developed to maintain structural integrity during
the interior ballistic phase of launch. The Kinetic Energy Projectile Design Program, KEPDEP,
is an interactive program implemented on a computer-aided design (CAD) network to generate
projectile geometry and to interface with several finite element programs, including ANSYS

2
(Sorensen 1991; DeSalvo and Gorman 1987; PRISM/DDM User Manuals 1990; Hallquist
1983). Using 4inite element techniques, simplified models of the projectile can be
implemented to insure that the in-bore stresses do not exceed the allowable material limits.
KEPDEP is used to generate a complete mesh with boundary conditions, or a parameter list
to be used in a parameterized optimization as will be discussed in this report.

2. INTRODUCTION

An experiment aimed at evaluating the terminal ballistic performance of two ballistic

materials, depleted uranium (DU) and a tungsten heavy alloy (WHA), for a particular
penetrator geometry posed a new design problem. It was decided that one sabot design
would be utilized to launch both penetrator types; furthermore, due to the velocity
requirements of the experiment, the sabot mass must be minimized. This created a problem
because DU and WHA have different mechanical properties (Table 1). DU is alloyed with
0.75% of titanium by weight and is aged with a heat treatment; it has very good elongation
properties and a nonlinear stress-strain behavior. The WHA used is 93% tungsten by weight
and is produced by a liquid phase sintering process. The tungsten grains are in a tungsten-
iron-nickel-cobalt matrix. The material is thermomechanically processed by swaging and then
heat treated. The material is almost perfectly elastic-plastic, and due to low elongation and its
microstructure, is prone to fracture. Considering the facts that DU is a nonlinear material
while WHA is not, and the large difference in elastic moduli, obtaining a mass efficient sabot
design would be difficult by manual means. Therefore, a parametric model of a KE projectile
was devised to be used in an optimization study.

Table 1. Material Properties

Elastic Poisson's Compressive Tensile

Material Modulus Density Ratio Yield Yield
(GPa) (kg/mi) (MPa) (MPa)
DU 165 18,600 .22 931 862
WHA 338 17,600 .29 1,207 1,303
Al 69 2,800 .33 496 496

3
 3. MODELING ASSUMPTIONS

Almost any finite element analysis makes some simplifying assumptions and several are
made in this analysis. The primary assumptions are that the model is axisymmetric and
quasi-static in nature (Drysdale 1981). The axisymmetry assumption can be used since the
sabot is under radial compression, thus preventing the petals from sliding relative to each
other. The quasi-static solution requires that the model be constrained in the axial direction
and that force be conserved by applying an acceleration in the direction opposite that of the
applied pressure. This assumption ignores any transverse loads, which is acceptable since
the maximum axial load is at least an order of magnitude larger. Furthermore, although the
rise time for the pressure is very short, dynamic analyses show that wave propagation is not
significant. Therefore, the maximum dynamic loads can be replaced by quasi-static loads.
Additional assumptions are made to simplify the geometry and can be seen in Figure 3.
These geometry assumptions simplify the bulkhead, the bell arid the penetrator/sabot
interface. The interface is simplified by smearing the details into a homogenous material with
shared nodes. The homogenous material properties are the same as the sabot material
properties with the exception of using an average density. Also note the addition of lumped
masses to represent the windshield and fin. Figure 3 also names the significant features of
the penetrator and sabot to clarify future discussions.

Saddle Fore taper

addl rIt unsupported
F~ore
Aft unsupported Aft tapr
penetrator length penetrator length

SWindshiield
Fin lumped moss Wnped mass

Figure 3. Axisymmetric Profile of a Penetrator and Sabot Designed by KEPDEP

and the Finite Element Mesh Generated by ANSYS via the KEPDEP
Interface.

4
 The axisymmetry option in ANSYS predetermines the coordinate system. The four-node,
quadrilateral element utilized requires that the y-axis be the axis of symmetry and suggests
that all elements be in the first quadrant (+x, +y). Therefore, the +x direction is radial and the
+y direction is axial. The origin is placed at the front of the windshield's lumped mass and the
penetrator extends along the positive +y axis.

The maximum pressure due to the combustion of propellant occurs at the breech of the
gun. Since the projecti'e and some of the propellant is accelerating, the pressure seen by the
projectile, the base pressure, is less than the breech pressure. A force balance about the
breech of the cannon at maximum pressure is seen in Equation 1 and is the result of the
Lagrange correction (Department of the Army 1965),

PAB= +-21C}M' (1)

where
Pc = Breech pressure (Pa)
AB = Area of the cannon bore (M2)
M = Projectile mass (kg)
C = Propellant mass (kg)
z = Acceleration (m/s2).

Summing the forces about the projectile leads to

P8 AB = Mg-, (2)
where
P8 = Base pressure (Pa).

Reorganizing Equations 1 and 2, the base pressure is determined as

Pe Pc
= P(3)
C
1
2M
 To obtain the quasi-static solution, the node at the origin is assumed to be stress free and
is constrained in the y direction. The base pressure, computed by Equation 3, is applied on
the tree edges of the model which are behind the bulkhead. To balance the force introduced
by the base pressure, the acceleration computed by Equation 2 must be applied in the -y
direction. The model is very sensitive to the balance of the base pressure and axial
acceleration, and due to numerical inaccuracies, the computed acceleration may induce a
stress at the axially constrained node. This can be checked in the post-processor by
examining the axial stress component at this node. If the magnitude of the stress is greater
than 0.5% of the maximum stress in the model, the acceleration is modified accordingly to
bring this stress level to acceptable limits, thus balancing the forces at the constrained node.

An additional static pressure exists within the model. As the projectile is forced into the
cannon, the obturator is radially compressed to provide a seal for the propellant gasses. This
exerts a pressure on the bulkhead since the obturator material is relatively incompressible.
Therefore, a 140 MPa pressure is placed on the aft portion of the bulkhead. The last
boundary conditions applied are to constrain the remaining nodes along the sabot/cannon
interface. In reality, this is a sliding contact where the projectile can move radially inward, but
since the projectile is experiencing radial expansion in these areas, using a radial constraint is
accurate. The boundary conditions are shown on the nodal presented in Figure 4.

4. PARAMETRIC MODEL

The projectile profile displayed in Figure 3 will be adapted to a parametric model. To

create the model, 25 geometric parameters, 10 material parameters, and 3 cannon
parameters are required. A list of these parameters with a short description is presented in
Table 2 and a sketch of the parameterized model is provided in Figure 4. A brief discussion
on generating the model will be presented

The mesh generation capability of the ANSYS preprocessor will be implemented to create
the nodes and elements; therefore, the model is divided into quadrilateral areas to define the
geometry. All necessary keypoints and line segments are defined first. (The keypoints and
line segments are required to define the quadrilateral areas.) The areas are created and
meshed using a "different element type for each material type. The constraints are set and the

6
 Table 2. KE Projectile Parameters

GEOMETRIC PARAMETERS

XR Penetrator radius. YR Penetrator length.

YF Fore unsupported penetrator length. XFI Forward flat radius of the sabot.
YFF1 Length of the forward flat. MF1 Slope of the forward taper.
XF2 Starting radius of the saddle. YFF2 Starting location of the saddle.
MF2 Slope of the saddle. YB Beginning of the bell.
DB Initial thickness of the bell. MB1 Slope of the front of the bell.
DMB Difference in slope between bell surfaces. YO Beginning of the bulkhead.
DO Thickness of the bulkhead. YA Aft unsupported penetrator length.
XA Aft flat radius of the sabot. YAF Length of the aft flat.
MA Slope of the aft taper. R1 Forward blend radius of the bell.
R2 Aft blend radius of the bell. R3 Forward blend radius of the bulkhead.
R4 Aft blend radius of the bulkhead. WWS Windshield mass.
WFN Fin mass.

MATERIAL PROPERTIES

PE Young's modulus of the penetrator. PRO Penetrator density.

PNU Poisson's ratio of the penetrator. PCOM Compressive yield of the penetrator.
PTEN Tensile yield of the penetrator. SRO Young's modulus of the sabot.
SRO Sabot density. SNU Poisson's ratio of the sabot.
SCOM Compressive yield of the sabot. STEN Tensile yield of the sabot.

CANNON PARAMETERS

XB Gun bore radius PC Chamber pressure.

CM Charge mass.

Prefix definitions:
X - Radial variable Y - Axial variable
D - Delta distance M- Slope
S - Sabot material property P - Penetrator material property
R - Fillet (blend) radius

7
 X*-.

C4C

a-,

wl
a.

. .... 01

1 0~

* CI

* 0L

Do --

C')4-
pressure on the bulkhead from the obturator being radially compressed is applied. The
remaining boundary conditions, base pressure and acceleration, cannot be set until the
projectile mass is known because they are functions of the mass. A solution input deck is
written and executed. By entering the ANSYS post-processor, the volumes for each element
type can be obtained. Multiplying each resulting volume by the appropriate density and
summing, the projectile mass can be determined. The objective function, sabot mass, is also
defined at this point. Reentering the preprocessor, the remaining boundary conditions are
calculated and set, thus completing the model.

5. DESIGN PROCESS

5.1 Initial Design. Before the optimization process could begin, an initial design was
required. For the purpose of this report, penetrator geometry from an existing projectile, the
M829, was used. The goal was to design a minimum weight sabot to launch both DU and
WHA penetrators. To assist the optimization process, the initial projectile design should be
feasible as specified by the state variables. This is not a requirement for the process to work,
but if the design parameters are too far from the acceptable design space, a feasible solution
may not be obtained.

The process for obtaining the initial design was as follows. The penetrator design was
known and the initial sabot design is to be provided by KEPDEP. Based on prior design
experience with both materials, WHA was chosen as the initial material. After the sabot
design was obtained, a finite element analysis was completed for each penetrator material. If
post-processing revealed that the sabot design was feasible for both penetrator materials, an
initial configuration had been obtained. If not, the sabot was modified and the analysis
process repeated until a feasible design was reached. At this point, a parameter file defining
the geometry (Table 2) was generated by KEPDEP.

5.2 Optimization. The ANSYS optimization module uses three different types of variables:
design, state, and objective. The design variables are the parameters which are allowed to
change from iteration to iteration. These variables are drawn from the pool of parameters
which specify the finite element model which are listed in Table 2. For each iteration, a
unique set of design variables will exist to provide a new finite element model each time. The

9
objective variable is the parameter which is being minimized, in this case, sabot mass. The
state variables define the optimization function over which the objective variable is minimized.
State variables can be almost any retrievable data from the finite element analysis, but in this
example, they are stresses from specified regions of the finite element mesh. Both design
and state variables are provided operating ranges. For the design variables, this range
specifies the acceptable values which can be used during the optimization. The state variable
ranges specify whether each state variable is acceptable for any given iteration and provide
the rules governing the objective function.

At this point, the user must decide which of the parameters that define the sabot are to be
used as design variables in the optimization process. Acceptable ranges for the design
variables must also be determined. Furthermore, if any additional state variables are desired,
they must also be defined. After all these decisions are made, the input deck in the Appendix
must be modified appropriately and the analysis can begin. A flow chart of this process is
presented in Figure 5.

In this analysis, the sabot details which remained constant were the locations of the sabot
ends, and the geometry of the bell and bulkhead. The features which were allowed to change
and their associated parameters are: the fore taper (XF1, YFF1, MF1), the saddle (XF2,
YFF2, MF2); the aft taper (XA, YAF, MA), and the locations of the bell (YB) and bulkhead
(YO). These 11 parameters were assigned design variable status. In addition to the 11
design variables, R1 and R2 were also allowed to change to prevent an error in geometry
creation from occuring. That is to say, if the design variables for a particular iteration had
values such that a particular blend radius could not be placed between the specified lines, R1
and R2 would be changed to accommodate the design variable set to prevent premature
termination. This is not an advisable solution, but is deemed better than program termination.
If either of these radii are modified, a flag is set to inform the designer to inspect the printed
output and optimization parameters for any unacceptable affects. The best solution is to
insure beforehand that the ranges on the design variables will not allow the creation of faulty
geometry. This was the case for the analysis presented in this report.

10
 U-

i E
lo .0 0
00

c.77

4-c 0 z , 1

0LO= L.o >

c~ -o M

~~CL

0))

- -- - a-

a-

*0 S.0 .0

V U

L)L))

0 cu

IL )

U-1

11c
 Six state variables were used in the optimization and are all yield limits of the various
materials in the analysis. The state variables are the maximum axial compressive stress in
the penetrator, the maximum effective tensile stress in the penetrator, and the maximum
effective stress in the sabot (excluding stress concentrations) for each penetrator material.
The limits are determined by the yield criteria for each material and the maximum stress level
acceptable.

6. RESULTS

Execution of the input deck (see Appendix) resulted in a local minimum for the objective
variable in 25 iterations. Of these 25 iterations, 14 were feasible and 5 resulted in values for
the objective variable which are within 2% of the minimum. Iterations were judged infeasible if
any state variable exceeded its limits. The mass of the sabot was reduced by 15% from
3.06 kg for initial design to 2.60 kg for the "best" solution. Numerical simulations of the
combustion process were performed using a standard propellant and a maximum breech
pressure of 655 MPa (Anderson and Ficke 1987). This operating pressure is the same as
used in the analysis and is 93% of the 120-mm cannon's maximum rating. This analysis
provided an accurate estimate for maximum loading conditions and the muzzle velocity.
Using the initial and the "best" designs, the calculations showed a 2.9% increase in muzzle
velocity, from 1,769 m/s to 1,820 m/s. For the "best" design, the projectile experienced a
maximum base pressure of 403 MPa and an acceleration of 65,700 g's.

The finite element mesh used within the optimization process consisted of approximately
450 nodes and 350 elements with 900 active degrees of freedom. The material properties of
the lump masses and WHA penetrator were linear (elastic), where as the remaining materials
were nonlinear (elastic-plastic). The initial solution passes, used to determine the projectile
mass and boundary conditions, were performed with linear material properties to minimize the
number of iterations required. The final solution pass for each penetrator material was
performed with the nonlinear materials active to obtain an accurate stress state. The
optimization analysis was executed on an APOLLO DN4500 as a background process in
7.25 hours.

12
 In Figure 6, the objective variable and state variables are plotted against iteration number
for the optimization process. In each of the plots, feasible solutions are defined by circles and
the state variable limits are denoted by broken lines. Additionally, state variables for DU and
WHA are labeled. The objective variable (sabot mass) is presented in Figure 6a, and the
state variables of sabot stress, compressive penetrator stress, and effective tensile penetrator
stress are in Figures 6b, 6c, and 6d, respectively. The first impression of these curves is the
oscillatory nature with a large magnitude early in the optimization process. Somewhere
between iterations 10 and 15, the curves dampen considerable, with the exception of the
compressive penetrator stresses for WHA. The large oscillations suddenly dampening is
explained by the number of design variables. In order to sufficiently describe the design
surface (objective function), at least one iteration per variable is required. To minimize the
complexity of the input stream, these initial iterations are randomly generated by the
optimizat*.*,n roitine. Once the surface is constructed, the optimization routine can select the
design variables for subsequent iterations much better and the state variables fluxuate less.
Examination of the state variables show that all three DU state variables are at the design
limits whereas the WHA state variables for the feasible designs are generally five percent
below their limits. Therefore, the three DU state variables were the dominating factors in the
analysis.

The design variables for the "best" sabot design were utilized to construct the final
projectile design. After completing this design, a final analysis was conducted using this sabot
with both penetrator materials. The resulting penetrator stresses are presented in Figure 7
with the WHA results in Figure 7a and the DU results in 7b. The stress profiles presented
represent the axial and the effective stresses along the penetrator centerline (solid lines) and
the minor diameter (dashed lines). The state variable limits for the penetrator materials are
presented as the horizontal phantom lines and are also labeled. Several points can be made
about this figure. Examination of the stresses at the penetrator's diameter reveals two sharp
discontinuities. These discontinuities occur at each end of the sabot and can be used to
define each of the unsupported penetrator lengths and the length of the sabot. Furthermore,
these features can be used to define the penetrator stress at each unsupported length.
Figure 7b shows that the stress limits for the DU penetrator are slightly exceeded, this is due
to the tolerance placed on the state variables. In the aft section of the penetrator, the
elevated stress level will have minimal effect since the overstressed state does not exist

13
 c
0

C Cl)
C44 I 4 C
L- 0n
CD) U) (
0
e4 O
01 EL
>0I

0 - a

Cu)C

CIDdV4) SS381S (Odfl) SS381S

UI)

cuC0
Cl

II = Ii 0U"
> 0)

00
0 0
U)i U,

(B~~~~o~ DdoS3NS LLr

CJ', -

14u

..........
 0
C

C4l
aca)
N -
w~*o W

a)
(DdY4
SSOIS 0
ca

pn 0)
o- 0)

0
0 0)
-n a:
75

0 0)00

0 0 00 000000

(DdY4) SS;DJIS

15
throughout the crosssection. Furthermore, since DU has excellent ductility and an ultimate
tensile stress exceeding 1,275 MPa, this slight overstressed condition is acceptable.
However, the stress at the forward, unsupported length should be examined more carefully
because the stress state exceeds the design limit through the entire cross section and
buckling could be a concern. Buckling, or yielding of the WHA penetrator is not a concern
because the maximum stresses in the penetrator and sabot are 5% below their respective
yield limits.

The decision to use the M829 penetrator to document this design procedure had a
significant drawback. Due to the aspect ratio and length of the penetrator, the differences
between the penetrator materials did not have a significant affect on the design process. In
this case, the sabot could have been designed for the DU penetrator alone and resulted in
nearly the same design. However, if a penetrator with increased length or aspect ratio were
used in the analysis, the result would most likely be different. As penetrator length increases,
the modulus mismatch between the WHA penetrator and the sabot will result in the
compressive stresses in the portion of the penetrator under the sabot saddle to become the
determining factor. But in the aft portion of the penetrator, and throughout the sabot, the DU
projectile will still be the driving influence. When this happens, the analysis is much more
complex and design optimization will be the most beneficial.

7. CONCLUSIONS

The implementation of parameterization and optimization techniques within ANSYS were

extremely helpful in solving the complex and difficult design problem presented in this report.
The parametric analysis and design language proved to be quite adept in creating a finite
element mesh for complex geometry. By using KEPDEP to provide an initial design and
ANSYS to minimize the sabot mass, an attractive result was produced with one day's effort.
Attempting a problem of this type without the methods discussed here could take days or
even weeks; therefore, a substantial decrease in design time was realized, leading to
increased productivity. Furthermore, since the optimization process was run in a background
environment, the actual commitment in time was less than two hours. One hour or less was
required to generate the initial design and set the design and state variables and another hour
to perform the final detailed analysis and review all results.

16
 8. REFERENCES

Anderson, R. D., and K. D. Ficke. "IBHVG2 -- A User's Guide." BRL TR-2829, U.S. Army
Ballistic Research Laboratory, Aberdeen Proving Ground, Aberdeen, MD, July 1987.

Department of the Army. Engineering Design Handbook: Interior Ballistics of Guns. AMC
Pamphlet 706-150, Washington, DC, February 1965.

DeSalvo, G. J., and R. W. Gorman. ANSYS Engineering Analysis Systems User's Manual,
Vol. I and II (Rev. 4.3), Swanson Analysis Systems Inc., Houston, PA, June 1987.

Drysdale, W. H. "Design of Kinetic Energy Projectiles for Structural Integrity." BRL-TR-02365,

U.S. Army Ballistic Research Laboratory, Aberdeen Proving Ground, MD, September
1981.

Hallquist, J. 0. "MAZE - An Input Generator for DYNA2D and NIKE2D." LLNL Report UCID
19029, Rev. 2, Lawrence Livermore National Laboratory, Livermore, CA, June 1983.

Hallquist, J. 0. "NIKE2D - A Vectorized Implicit, Finite Deformation Finite Element Code for
Analyzing the Static and Dynamic Response of 2-D Solids with Interactive Rezoning and
Graphics." LLNL Report UCID 19677m, Rev. 1, Lawrence Livermore National Laboratory,
Livermore, CA, December 1986.

PRISM/DDM User's Manuals. CALMA, A Division of Prime Computer, Inc., San Diego, CA,
(-1990).

Sorensen, B. R. "Implementation of a CAD/CAM System to Design, Analyze, and

Manufacture Prototype Kinetic Energy Projectiles." BRL-TR-3272, U.S. Army Ballistic
Research Laboratory, Aberdeen Proving Ground, MD, September 1991.

17
INTENTIONALLY LEFT BLANK.

18
 APPENDIX:

PARTIAL LISTING OF THE ANSYS INPUT DECK

19
INTENTIONALLY LEFT BLANK.

20
 A partial listing of the ANSYS (Revision 4.3A) input deck used in the preceding analysis is
presented. Most of the file has been omitted to maintain confidentiality. The following
variables are defined: vI, ..., v5, are the volumes of each element type; dl, ..., d5 are the
densities for each element type; yacc, yacl, yac2 are axial acceleration values; sy0l, sy02
are axial stresses at the constrained node; mt and ms are the projectile and sabot masses;
the remaining variables are self-explanatory.

/com, Read initial parameter file /com, Use post-processing to determine projectile
*use,parm.opt /com, and sabot masses.
Icom, Set initial values and constants /com, calculate pressure and acceleration.
/prep7 /postl
Ititle,Sabot optimization stress,volu
set,1
/com, Set materials for tungsten projectile ersel,type,1
/com, Linear material for tungsten nelem
/com, Non-linear for grooves and sabot ssum
*get,vl ,ssum,volu
nall$eall
ersel,type,2
/com, Construct geometry and mesh nelem
ssum
*getv2,ssum,volu
nall$eall
/com, Set switches for one iteration solution ersel,type,3
knl,O nelem
ter,1 ssum
afwr *get,v3,ssum,volu
fini nall$eall
ersel,type,4
/com, Execute solution nelem
/exe ssum
/input,27 *get,v4,ssum,volu
fini nall$eall

21
ersel,type,5 /com, Execute solution for two load steps
nelemn lexe
ssum linput,27
*get,v5,ssum,volu f in
nall$eall
mt-(((vl *dl)+(v2*d2))+((v3*d3) +(v4*d4))) +(v5*d5) loom, Post-process and interpolate the <hrt>
ms=(v3*d3) lcom, acceleration to minimize the
pb~pc/l( +(cnV(2*mt))) lcom, axial stress and the windshield
zaci =(pb'ab)lmtd lpostl
zac2=zacl +100 set,l
fini nrsel,y,
/prep7 nrsel,x
resume nsort,sy
*use,parm.ans *get,sy~l ,max
/com, Add boundary conditions at R=XB nail
nrsel,x,xb set,2
nrsel,y,,((yol +yo2)/2)+1 nrsel,y,
d,all,ux,0 nrsel ,x
nail nsort,sy
nrsei,x,xb *get,syO2,max
nrsel,y,((yol +yo2)12)+1 ,yo2+1 nail
psI ,all...20000 xt=(zacl -zac2)l(sy~l -syO2)
nail zacc=(-xtrsyO2)+zac2
/com, Add base pressure and two accelerations fin!
lsrsel,,13,21
nline,1 /com, Enter PREP7 and apply accurate <hrt>
psI ,all...
pb /com, acceleration. Use three load steps to <hrt>
nail /com, minimize the plasticity ratio.
acel,,-zacc /com, Set switches for multiple iterations
Iwri /prep7
acel,,-zacl resume
lwri use,parm.ans
afwr knl,1
fini iter,-20

22
cnvr,.1 Icom, Compute new projectile mass and B.C.s
Isrsel,,13.21 mtO=mt
nline,11 mt=(((vl 'dl )+(v2*d2))+((v3*d3)+(v4*d4)))+(v5*d5)
ps ,aii, ,,pb*.5 pb=pC/(1 +(Crnl(2*mt)))
nail zacc=zacc~mtO/mt
acel,,-zacc*.5 nline,1
wNd psf,all...pb*.5
nline,1 nail
psf,all...pb*.8 acel,,-zacc*.5
nail lwri
acel,,-zacc'.8 MOOe,
lwri psf,a~ll...pb*.8
nline,1 nail
psi ,all...
pb acel,,-zacc.8
nail 1wni
acel,,-zacc nline,11
lwri psf,all,,,pb
afwr nail
Ieini acel,,-zacc
lwri
/com, Solve for the tungsten projectile stresses afwr
/exe fini
/input,27
fmin

/com, Save the tungsten solution /com, Solve for the uranium projectile stresses
/copy, 12,31 /exe
/input,27
/prep7 fmin
resume
*use,parm.ans /com, Save the uranium solution
Icom, Change to uranium material properties /copy, 12,32

/com, Post-process to obtain state variables

/postl

23
nfile,31 opvar,ste2,sv,O,tend
set,3 opvar,sse2,sv,O,ssbd
Icom, Tungsten state variables
/com, Design variables
opvar,yff 1,dv, .

opvar,xfl1,dv,
nfile,32 opvar,mfl,dv,
set,3 opvar,yff 2, ..
Icom, Uranium state variables opvar,xf2,dv,
opvar,mf2,dv,
opvar,ma,dv,
opvar,yaf,dv,
finish opvar,xa,dv,
opvar~yol,dv,
Icom, Optimization routine opvar,ybl ,dv,
/opt
/com, Objective variable OPCOPY
opvar,ms,obj,,,.OO1 oprun,10 .... ,,30
oplist,all,,O
/com, State variables for tungsten finish
opvar,scyl ,sv,cmpw,O
opvar,stel ,sv,O,tenw
opvar,ssel ,sv,O,ssbw
/com, State variables for uranium
opvar,scy2,sv,cmpd,O

24
 No. of No. of
Copies Organization Copies Ornanization
2 Administrator 1 Commander
Defense Technical Info Center U.S. Army Missile Command
ATTN: DTIC-DDA ATTN: AMSMI-RD-CS-R (DOC)
Cameron Station Redstone Arsenal, AL 35898-5010
Alexandria, VA 22304-6145
1 Commander
Commander U.S. Army Tank-Automotive Command
U.S. Army Materiel Command ATTN: ASONC-TAC-DIT (Technical
ATTN: AMCAM Information Center)
5001 Eisenhower Avenue Warren, MI 48397-5000
Alexandria, VA 22333-0001
1 Director
Commander U.S. Army TRADOC Analysis Command
U.S. Army Laboratory Command ATTN: ATRC-WSR
ATTN: AMSLC-DL White Sands Missile Range, NM 88002-5502
2800 Powder Mill Road
Adelphi, MD 20783-1145 1 Commandant
U.S. Army Field Artillery School
2 Commander ATTN: ATSF-CSI
U.S. Army Armament Research, Ft. Sill, OK 73503-5000
Development, and Engineering Center
ATTN: SMCAR-IMI-I (Ck,. only)1 Commandant
Picatinny Arsenal, NJ 07806-5000 U.S. Army Infantry School
ATTN: ATSH-CD (Security Mgr.)
2 Commander Fort Benning, GA 31905-5660
U.S. Army Armament Research,
Development, and Engineering Center (unw,. ony)l Commandant
ATTN: SMCAR-TDC U.S. Army Infantry School
Picatinny Arsenal, NJ 07806-5000 ATTN: ATSH-CD-CSO-OR
Fort Benning, GA 31905-5660
Director
Benet Weapons Laboratory 1 Air Force Armament Laboratory
U.S. Army Armament Research, ATTN: WL/MNOI
Development, and Engineering Center Eglin AFB, FL 32542-5000
ATTN: SMCAR-CCB-TL
Watervliet, NY 12189-4050 Aberdeen Proving Ground
(u.ncm. only)1 Commander 2 Dir, USAMSAA
U.S. Army Armament, Munitions ATTN: AMXSY-D
and Chemical Command AMXSY-MP, H. Cohen
ATTN: AMSMC-IMF-L
Rock Island, IL 61299-5000 1 Cdr, USATECOM
ATTN: AMSTE-TC
Director
U.S. Army Aviation Research 3 Cdr, CRDEC, AMCCOM
and Technology Activity ATTN: SMCCR-RSP-A
ATTN: SAVRT-R (Library) SMCCR-MU
M/S 219-3 SMCCR-MSI
Ames Research Center
Moffett Field, CA 94035-1000 1 Dir, VLAMO
ATTN: AMSLC-VL-D
10 Dir, BRL
ATTN: SLCBR-DD-T

25
No. of No. of
Copies Oroanization Copies Oranization
5 Director 2 Director
Benet Weapons Laboratory Lawrence Livermore National Laboratory
U.S. Army Armament Research, ATTN: R. Christensen
Development, and Engineering Center S. deTeresa
ATTN: SMCAR-CCB, P.O. Box 808
J. Keane Livermore, CA 94550
T. Allen
J. Vasilankis 2 Pacific Northwest Lab
J. Zweig A Div of Battelle Memorial Inst.
L. Johnson Technical Information Section
Watervliet, NY 12189 ATTN: M. Smith
M. Gamich
7 Commander P.O. Box 999
U.S. Army Armament Research, Richland, WA 99352
Development, and Engineering Center
ATTN: SMCAR-CCH-T, 4 Director
S. Musalli Sandia National Laboratories
P. Christian Applied Mechanics Department,
K. Fehsal Division-8241
SMCAR-CCH-V, E. Fennel! ATTN: C. W. Robinson
SMCAR-CCH, J. DeLorenzo G. A. Benedetti
SMCAR-CC, K. Perano
R. Price W. Kawahara
J. Hedderich P.O. Box 969
Picatinny Arsenal, NJ 07806-5000 Livermore, CA 94550-0096
2 Commander I Director
U.S. Army Armament Research, Los Alamos National Laboratory
Development, and Engineering Center ATTN: D. Rabem
ATTN: SMCAR-TD, WX-4 Division, Mail Stop G-787
M. V. Lindner P.O. Box 1633
T. Davidson Los Alamos, NM 87545
Picatinny Arsenal, NJ 07806-5000
1 Olin Corporation
Commander Flinchbaugh Division
Production Base Modernization Agency ATTN: E. Steiner
U.S. Army A.,ament Research, P.O. Box 127
Development, and Engineering Center Red Lion, PA 17356
ATTN: AMSMC-PBM-K
Picatinny Arsenal, NJ 07806-5000 1 Olin Corporation
ATTN: L. Whitmore
3 PEO-Armaments 10101 9th Street North
Project Manager St. Petersburg, FL 33702
Tank Main Armament Systems
ATTN: SFAE-AR-TLtA, COL Hartline 2 Alliant Techsystems, Inc.
SFAE-AR-TMA-MD, C. Kimker ATTN: C. Candland
SFAE-AR-TMA-ME, K. Russell K.Ward
Picatinny Arsenal, NJ 07806-5000 5640 Smetana Drive
Minnetonka, MN 55343

26
No. ot
Copies Organization
2 Chamberlain Manufacturing Corp.
Waterloo Facility
ATTN: T. Lynch
550 Ester Street
P.O. Box 2335
Waterloo, IA 50704

27
INTENTIONALLY LEFT BLANK.

28
 USER EVALUATION SHEET/CHANGE OF ADDRESS

This laboratory undertakes a continuing effort to improve the quality of the reports it
publishes. Your comments/answers below will aid us in our efforts.

1. Does this report satisfy a need? (Comment on purpose, related project, or other area of
interest for which the report will be used.)

2. How, specifically, is the report being used? (Information source, design data, procedure,
source of ideas, etc.)

3. Has the information in this report led to any quantitative savings as far as man-hours or
dollars saved, operating costs avoided, or efficiencies achieved, etc? If so, please
elaborate.

4. General Comments. What do you think should be changed to improve future reports?
(Indicate changes to organization, technical content, format, etc.)

BRL Report Number BRL-TR-3289 Division Symbol

Check here if desire to be removed from distribution list.
Check here for address change.
Current address: Organization
Address

DEPARTMENT OF THE ARMY

Director
111 NO POSTAGE
U.S. Army Ballistic Research Laboratory NECESSARY
ATTN: SLCBR-DD-T IF MAILED
Aberdeen Proving Ground, MD 21005-5066 IN THE

OFFICIA BUSINES BUSINESS REPLY MAIL UNITED STATES

PRR• C11S FEWTNo 0001, AP, W

Postage will be paid by addressee.

Director
U.S. Army Ballistic Research Laboratory
ATTN: SLCBR-DD-T
Aberdeen Proving Ground, MD 21005-5066