ADA441366

Air Gun Launch Simulation Modeling and Finite
Element Model Sensitivity Analysis
by Mostafiz R. Chowdhury and Ala Tabiei
ARL-TR-3703 January 2006
Approved for public release; distribution is unlimited.

NOTICES
Disclaimers
The findings in this report are not to be construed as an official Department of the Army position
unless so designated by other authorized documents.
Citation of manufacturer’s or trade names does not constitute an official endorsement or

approval of the use thereof.
DESTRUCTION NOTICE⎯Destroy this report when it is no longer needed. Do not return it to

the originator.
Army Research Laboratory
Adelphi, MD 20783-1145
ARL-TR-3703 January 2006
Air Gun Launch Simulation Modeling and Finite

Element Model Sensitivity Analysis
Mostafiz R. Chowdhury
Weapons and Materials Research Directorate, ARL
Ala Tabiei
Private Contractor

Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data
needed, and completing and reviewing the collection information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden,
to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302.
Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid
OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
January 2006 Final May to December 2003

4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Air Gun Launch Simulation Modeling and Finite Element Model Sensitivity 5b. GRANT NUMBER
Analysis
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
622618H80
Mostafiz R. Chowdhury (ARL); Ala Tabiei (private contractor) 5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

REPORT NUMBER
U.S. Army Research Laboratory
Weapons & Materials Research Directorate ARL-TR-3703
Adelphi, MD 20783-1197
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT

NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT

13. SUPPLEMENTARY NOTES
14. ABSTRACT
This report presents in two parts analytical simulation results of a test item launched in an air gun test. The first part of the
report presents a discrete mass-spring model to predict the transient response of a generic artillery component subjected to
launch simulation in an air gun test environment. A 2-degree-of-freedom model is developed to simulate the air gun launch
environment in which a test object mounted on a projectile is launched through the air gun and decelerated by a crushing
aluminum honeycomb mitigator. The mitigator in turn impacts the momentum exchange mass (MEM) before being stopped at
the retrieving end. The present model achieved a good prediction of the period and peak acceleration of the on-board recorder.
The sensitivity of finite element (FE) model parameters affecting the dynamics of a test object launched in an air gun test is
discussed in the second part of the report. An LS-DYNA1 FE model previously developed to simulate the impact mitigation
environment is used to investigate the sensitivity of FE model parameters. The sensitivity study clearly indicates the necessity of
honeycomb material characterization under very high impact loading for effective modeling of the strain rate effects. The FE
analysis also suggests that intermittent eigenvalue analyses during an impact simulation could be useful in identifying the
dominant modes of vibration governing the dynamics of the test items and verifying the predictive capability of the FE model.
An efficient procedure using a decoupled model is proposed to simulate the response of a test item launched in an air gun test.
In this approach, the contact force generated from the discrete analytical model is applied on a detailed FE mesh of the test item
for a given impact velocity. The acceleration data obtained with this procedure compared well with those predicted from the full
FE analysis. The simplified model runs in less than 1 minute.
1
LS-DYNA, which is not an acronym, is a trademark of Livermore Software Technology Corp.
15. SUBJECT TERMS
air gun launch simulation; aluminum honeycomb; analytical simulation; discrete spring-mass model; sensitivity
17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
16. SECURITY CLASSIFICATION OF: OF ABSTRACT OF PAGES
Mostafiz Chowdhury
a. REPORT b. ABSTRACT c. THIS PAGE SAR 65 19b. TELEPHONE NUMBER (Include area code)
Unclassified Unclassified Unclassified 301-394-6308
Standard Form 298 (Rev. 8/98)
Prescribed by ANSI Std. Z39.18
ii
Contents
List of Figures iv
Acknowledgments vii
1. Part I: Air Gun Modeling 1

1.1 Introduction .....................................................................................................................1
1.2 Model Formulation..........................................................................................................2
1.2.1 Assumptions ........................................................................................................2
1.2.2 Rigid Body Dynamics System ............................................................................2
1.3 System Parameters Definition .........................................................................................5
1.3.1 OBR and MEM ...................................................................................................5
1.3.2 Mitigator..............................................................................................................5
1.3.3 Strain Rate Effects...............................................................................................8
1.4 Results and Discussion....................................................................................................9
1.4.1 Validation of Present Model................................................................................9
1.5 Parameter Variation Sensitivity.....................................................................................16
1.6 Conclusion.....................................................................................................................19
2. Part II: Air Gun Finite Element Simulation 20

2.1 Review of Previous Work .............................................................................................20
2.2 Finite Element Model Sensitivity Study .......................................................................21
2.2.1 Double Versus Single Precision Code...............................................................21
2.2.2 Time Step Size Effect........................................................................................23
2.2.3 Different Output Intervals .................................................................................24
2.2.4 All Damping Options in LS-DYNA Investigated .............................................25
2.2.5 The OBR Filled With Solid Elements to Represent Beads ...............................28
2.2.6 Intermittent Eigenvalues Analysis.....................................................................29
2.3 Simplified Simulation ...................................................................................................31
2.4 Conclusion.....................................................................................................................32
3. References 34
Appendix A. Program I/O Files and FORTRAN Source Code 35
Distribution List 42
iii
List of Figures
Figure 1. Schematic of the air gun. ................................................................................................ 1

Figure 2. Dynamic system representation...................................................................................... 3
Figure 3. Honeycomb stress-strain curve....................................................................................... 6
Figure 4. Plastic shock wave.......................................................................................................... 6
Figure 5. Mitigator deformation. ................................................................................................... 7
Figure 6. Mitigator wedge deformation zones............................................................................... 7
Figure 7. Top acceleration of the OBR unfiltered. ...................................................................... 10
Figure 8. Top acceleration of the OBR filtered at 7000 Hz......................................................... 11
Figure 9. Top acceleration of the OBR filtered at 2500 Hz......................................................... 11
Figure 10. Top acceleration of the OBR filtered at 1000 Hz....................................................... 12
Figure 11. Fast Fourier transform (FFT) of the top acceleration of the OBR. ............................ 12
Figure 12. Top velocity of the OBR. ........................................................................................... 13
Figure 13. Top displacement of the OBR. ................................................................................... 14
Figure 14. Mitigator contact forces.............................................................................................. 14
Figure 15. Mitigator contact force impulse.................................................................................. 15
Figure 16. Length of mitigator plastic region. ............................................................................. 15
Figure 17. m1:m2 mass ratio sensitivity of model. ....................................................................... 16
Figure 18. Model sensitivity to 10% variation of the value of k1. ............................................... 17
Figure 19. Model sensitivity to large variation of the value of k1. .............................................. 17
Figure 20. Model sensitivity to variation of the solution time step. ............................................ 18
Figure 21. Model sensitivity to 10% variation of the value of mitigator radius.......................... 19
Figure 22. Acceleration data for the same FE model with two different precisions. .................. 22
Figure 23. FFT of the acceleration data. ...................................................................................... 22
Figure 24. Acceleration data for different integration time steps. ............................................... 23
Figure 25. FFT of the acceleration data for different integration time steps. .............................. 23
Figure 26. Acceleration data sampled at different frequencies.................................................... 24
Figure 27. FFT of the acceleration data sampled at different frequencies................................... 24
Figure 28. Acceleration data for simulation with mass proportional damping............................ 25
Figure 29. FFT of the acceleration data for simulation with mass proportional damping........... 25
Figure 30. Acceleration data for simulation and test with mass proportional damping. ............. 26
Figure 32. FFT of the acceleration data for simulation with stiffness proportional damping. .... 27
Figure 33. Acceleration data for simulation and test with stiffness proportional damping......... 27
Figure 34. FFT of the acceleration data for simulation and test with stiffness proportional
damping................................................................................................................................. 28
Figure 35. Acceleration data for simulation with filled OBR...................................................... 28
Figure 36. FFT of the acceleration data for simulation ............................................................... 29
iv
Figure 37. Mode shapes for a fundamental frequency................................................................. 30
Figure 38. Contact force between the OBR and the mitigator..................................................... 31
Figure 39. Acceleration data of the simplified and full model. ................................................... 32
v
INTENTIONALLY LEFT BLANK
vi
Acknowledgments
The analytical work presented here was accomplished under a contract (Contract No. DAAD19-
02-D-001, TCN: 03-054) with the Scientific Services Program of U.S. Army Research Office,
Research Triangle Park, North Carolina. Air gun tests were conducted by Mr. Ara Abrahamian
of the Weapons and Materials Research Directorate, U.S. Army Research Laboratory (ARL).
The contributions of Mr. Ara Abrahamian, Mr. Edward Szymanski, and Mr. William McIntosh
of ARL in delivering the air gun test are greatly appreciated. Dr. Dave Hampton of the
U.S. Military Academy at West Point provided a helpful review of the report, which is greatly
appreciated.
vii
INTENTIONALLY LEFT BLANK
viii
1. Part I: Air Gun Modeling
1.1 Introduction
In an air gun test, it is critical to identify the characteristics of the test environment such as the
dynamic/physical properties of momentum exchange mass (MEM) and mitigator combinations in
order to obtain a target forcing function. Figure 1a depicts a schematic of the air gun test setup.
The test setup consists of an air gun, a pressure tube, a catch tube, and a MEM. The test item,
which is placed inside the pressure tube, is launched to impact an aluminum honeycomb mitigator.
The test item contains an on-board recorder (OBR) that records responses during launch simula-
tion. The OBR case, which is hereafter referred to as the “OBR,” is shown in figure 1b with
several accelerometers and strain gauges. The model developed in this study will be used to
determine the response behavior of the test item mounted on a given projectile during a launch
simulation air gun test. This methodology requires the development of a predictive model of
responses of the test article. An analytical model is herein developed to simulate an air gun launch
environment in which a test object mounted on a projectile is launched through the air gun and
decelerated by the crushing of an aluminum honeycomb mitigator that impacts the MEM before
being stopped at the retrieving end.
CATCH TUBE PRESSURE TUBE GUN EXIT
Energy
MEM MITIGATOR TEST Item Projectile
a. schematic of air gun test setup.
b. The OBR with test item.

Figure 1. Schematic of the air gun.
1
The objective of this effort is to develop an overall analytical model to predict the transient
response of a generic artillery component subjected to launch simulation. The simulation is
achieved via impact mitigation techniques whereby the component is propelled to a target inside
a 4-inch carrier equipped to measure component response during the simulation. A detailed
description of the developed formulation is presented in the next section.
1.2 Model Formulation
1.2.1 Assumptions
To model the response of the OBR-mitigator-MEM system during impact, an analytical model is
herein developed. The model provides a closed form solution to the dynamic system driven by
the central difference equations and is based on the following assumptions:
The OBR and MEM can be accurately represented by rigid bodies since they do not dissipate
energy through plastic deformation. Only the lowest natural frequency of the OBR is of interest
when we are predicting the acceleration output; therefore, the OBR is represented by a 2-degree-
of-freedom spring-mass system. It is assumed that this frequency is readily available (if not, it
can be easily determined through simple analysis).
The mass of the mitigator is neglected. The mitigator is entirely represented by the force it exerts
on the OBR and the MEM. This force is determined from the geometric and material properties
of the mitigator and from balance and conservation considerations during the analysis stage.
1.2.2 Rigid Body Dynamics System

The impact event is represented by a rigid body mass-spring system with 3 degrees of freedom
acting along the horizontal axis. The system consists of three masses and one spring, as
illustrated in figure 2. The OBR has initial velocity v0 and is represented by two of the masses,
m1 and m2, and a connecting spring, k1. The mitigator is represented by the force F, and the
MEM is represented by its mass m3. Note that the whole system is divided into two separate
systems coupled by the force F representing the mitigator.
This assembly is a very simple representation of a complex deforming system. If the first OBR
frequency mode is to be neglected, we can further simplify it by combining masses m1 and m2
into a single body of mass mOBR = m1 + m2 and dropping spring k1. This will result in a 2-degree-
of-freedom system. However, the essential part of this system formulation is the representation
of the mitigator with the force F. To accurately represent the behavior of the mitigator, an
elaborate and efficient formulation needs to be developed. Such a formulation is described in
detail as follows. This simple dynamic system representation is accurate, efficient, and provides
adequate results.
2
v0 v0
k2
k1
m1 m2 m3
C1
OBR Mitig. MEM
The system above is equivalent to the system below
v0 v0
k1
F F
m1 m2 m3
OBR Mitigator MEM
Figure 2. Dynamic system representation.
First, let us describe a time-stepping scheme to solve the dynamic system. The system’s motion
can be represented by the system of equations:
[M ]{u&&}+ [K ]{u} = {f } (1)
in which
[M ] = ⎡⎢
m1 0⎤
m2 ⎥⎦
is the mass matrix of the left-hand side system,
⎣0
− k1 ⎤
[K ] = ⎡⎢
k1
k1 ⎥⎦
is the stiffness matrix of the left-hand side system,
⎣ − k1
⎧ u1 ⎫
{u} = ⎨ ⎬ is the displacement vector with ui being the displacement of mass mi
⎩u 2 ⎭
where i = 1,2; {u& } and {u
&&} are the mass velocities and accelerations, respectively.
{f } = ⎧⎨
0 ⎫
⎬ is the external force vector.
⎩− F ⎭
To solve this system, we define a central difference time-stepping scheme. At the initial moment
(t = 0), we have
3
{u}t = 0 = ⎧⎨
0⎫
⎬,
⎩0⎭
⎧v0 ⎫
{u& }t =0 = ⎨ ⎬,
v
⎩ 0⎭
− F0 ⎧0⎫ ⎧0⎫
{&u&}t =0 = [M ]−1 ({f }t =0 − [K ]{u}t =0 ) = ⎨ ⎬=⎨ ⎬
m2 ⎩1⎭ ⎩0⎭
To define the central difference time-stepping scheme, we expand {u}n+1 and {u}n–1 about {u}n
using a Taylor series as follows:
{u}n +1 = {u}n + Δt{u& }n + Δt {&u&}n + O(Δt 3 )

2
2 (2)
( )
2
{u}n −1 = {u}n − Δt{u& }n + Δt {&u&}n − O Δt 3
2
in which Δt is the central difference time step and O(Δt3) represents additional terms of the order
of Δt3 considered negligible. Then, from equation 2 we have
{u& }n = 1
({u}n+1 − {u}n−1 )
2Δt
(3)
{u&&}n = 2 ({u}n+1 − 2{u}n + {u}n−1 )
1
Δt
Now, from the equation of motion, equation 1, it follows:
⎛ 1 ⎞
⎜ 2 [M ]⎟{u}n +1 = {f }n − [K ]{u}n + 2 [M ](2{u}n − {u}n −1 )
1
⎝ Δt ⎠ Δt
or {u}n +1 = 2{u}n − {u}n −1 + [M ]−1 ({f }n − [K ]{u}n ) (4)

in which
[M~] = 1
[M ]
Δt 2
Equations 4 and 3 are sufficient to define a forward time-stepping scheme, provided that a
sufficiently small time step, Δt, is used. To start the time-stepping scheme, {u}–1 needs to be
available, which can be determined from the second equation of equation 2:
2
{u}−1 = {u}0 − Δt{u& }0 + Δt {&u&}0
2
For the right-hand side system representing the MEM we have
4
m3 u&&3 = F
m3
((u3 )n+1 − 2(u 3 )n + (u 3 )n−1 ) = Fn
Δt 2
Δt 2 Fn
(u3 )n+1 = + 2(u 3 )n − (u 3 )n −1
m3
and the velocity and acceleration of the MEM can be calculated from equation 3.
1.3 System Parameters Definition
1.3.1 OBR and MEM

The simplest part of the system (figure 2) is the MEM, which is represented by its mass only.
The OBR is represented by two masses and a connecting spring. The sum of the two masses is
equal to the total mass of the actual OBR. In the determination of the way in which the total
mass is distributed between m1 and m2, different values were used for the investigated system.
It was determined that a ratio of m1:m2 anywhere from 1:2 to 2:1 gives reasonable results for the
second mode or lowest fundamental frequency of the OBR (the first mode is a rigid body mode).
The spring stiffness (k1) is determined so that the frequency of the second mode is matched.
ω 2 (m1 + m 2 )
k1 = (5)
4
Equation 5 is derived from K − ω 2 M = 0, for m1 = m2 = m, in which ω is the angular

frequency to be matched in the model. If the value of ω is not readily available, it can be easily
determined by simple testing.
1.3.2 Mitigator
The representation of the mitigator is vital to the accuracy of the simulation. In this development,
it is assumed that the mitigator is built of honeycomb and consists of one continuous piece of
honeycomb structure. It has one cylindrical section and one wedge-shaped part with a varying
cross section on the side (see figure 1a) that is being impacted by the OBR. A typical orthotropic
crush model behavior for the honeycomb material is composed of three phases as shown in figure
3 (1). These three phases include linear elastic loading, volumetric crush, and hardening to full
compaction. The initial linear elastic loading phase can be represented by an elastic modulus Ee,
yield stress and strain σy and εy, respectively. The volumetric crush behavior of the honeycomb
material model (up to compaction strain), εc, is assumed to be linear, as indicated in figure 3b.
The fluctuation in volumetric strain relations observed in tests (see figure 3a) is believed to have
resulted from unstable buckling of honeycomb cells. The hardening phase is assumed to have a
linear relation during compaction with modulus Ec. To model the mitigator, the relation between
the contact force and the mitigator front deformation and their progress in time is established.
5
Here, it is assumed that as the wedge section deforms, it takes the shape of a truncated wedge, as
was indicated by the finite element analysis (FEA) results. It is also assumed that within the
wedge part, the cross-sectional area varies linearly over the wedge height. In the development of
this mitigator deformation model, its mass is herein neglected; its mass and velocity throughout
the event are small compared to the other moving parts. The mitigator deformation is driven in
time by its elastic and plastic wave propagation speeds. Since the elastic propagation speed is
much higher than the velocities in the event, it is assumed that elasticity-driven events propagate
with infinite speed. Therefore, before the stress-strain state attains material yielding, it must be
homogeneous along the cylinder axis.
16 σ
14
H38_06
Average Crush Strength: 7.18 ksi
Ec
Crush Efficiency: 64.4 %
12
10 σy
F/Ao, ksi
8 εc
6
Ee
4
2 εy
0
ε
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
ΔL/Lo
a b
Figure 3. Honeycomb stress-strain curve.
Our aim is to determine the force resulting from the mitigator deformation within the central
difference time-stepping scheme. Assuming that we have found a solution at time equal to tn, we
need to express the force at time tn+1 = tn + Δt.
When the OBR hits the mitigator, a plastic region forms immediately at the wedge top. As
indicated in figure 4, the boundary between this plastic region and the rest of the mitigator,
which is still elastic, would travel down the length of the mitigator with variable velocity equal
to the current plastic wave speed, cpl (2, 3). Figure 4 also shows the respective material states
(density [ρ], stress [σ]) of the crushed and elastic parts of the mitigator. The stresses, velocities,
and densities on both sides of the plastic wave front are further explained in figure 5. With the
approach suggested in (2) and (3), the plastic wave speed can be determined as follows.
Plastic wave front
σb σy
m2 V = VOBR V=0 m3
ρ = ρy/(1-εc) ρy
cpl
Figure 4. Plastic shock wave.
6
m2 σb σy m3
VOBR V = VOBR V = VMEM VMEM
ρ = ρy/(1-εc) ρy
ΔL0 x Lel
L0
Figure 5. Mitigator deformation.
Following the notation of figure 5, let the initial length of the mitigator be L0. At the current
analysis time, t, the total deformation of the mitigator is ΔL0 and the lengths of the crushed and
the elastic parts are x and Lel, respectively. Since the yield strain of the mitigator is negligible
compared to its compaction strain, we can assume that all mitigator deformation, ΔL0, occurs in
the compacted region:
ΔL0
εc = (6)
x + ΔL0
or
1− εc
x= ΔL0
εc
Ld Lp
Lw x
Figure 6. Mitigator wedge deformation zones.
The mitigator deformation can be expressed through the displacements of both its ends, which
are equal to the displacements of the adjacent masses m2 and m3.
ΔL0 = uOBR − u MEM (7)
The plastic wave speed, cpl, is equal to the time rate of change in length of the crushed portion as
shown next.
7
d ( x + ΔL0 ) 1 d ( ΔL0 ) 1 d (uOBR − u MEM ) (VOBR − VMEM )
c pl = = ⋅ = ⋅ = (8)
dt εc dt εc dt εc
Knowing the velocity of the OBR at time tn, we can calculate this speed and revise the plastic
wave front location:
(x )
front n +1 = (x front )n + (c pl )n Δt (9)
The stresses immediately below the plastic wave front (assuming a vertical mitigator with the
wedge on top, figure 6) are equal to the yield stress. Thus, knowing the position of the wave
front, one can easily determine the area at the plastic wave front (Af) and the corresponding
cross-sectional resultant contact force, Fcont as shown next.
Fcont = σy⋅(Af) (10)
The front area, Af, is determined from the plastic wave front displacement. Within the wedge
portion of the mitigator, we assume that there are two regions: the plastic zone and the elastic
zone. As illustrated in figure 6, Lw is the wedge initial height, and Ld is its current deformed
height to be determined from loading. When the plastic zone length ( L p ) is less than the wedge
length ( Lw ), then the front area can be determined as follows:
Lp
Af= Ao (11)
Lw
Otherwise,
Af = Ao
in which A0 is the cross-sectional area of the cylindrical mitigator.

Equation 10 is assumed to represent the mitigator force, F, the contact force between the mitigator
and OBR, and the mitigator and MEM in figure 2. Computation of mitigator force using equation
10 is believed to be valid to a sufficient degree of accuracy as long as the plastic wave front
progresses forward (cpl > 0). When the plastic wave motion no longer propagates, then the miti-
gator force decreases after the unloading of the elastic stiffness of the mitigator. This scheme
can be easily programmed as a stand-alone routine, as shown in appendix A.
1.3.3 Strain Rate Effects

Any analysis model of the impact event would rely on the constitutive representation of the
mitigator honeycomb. The associated relationship would usually be strain rate dependent, with the
dependency being very important for the accuracy of representation. Significant stress enhance-
ment under impact loading is reported for most of the cellular honeycomb materials by Zhao (4).
Since in most cases it is impossible to develop a closed form model of the strain rate effect, some
empirical approaches have been developed and successfully applied in this investigation. One
such example is a power law relation used in (5), in which the stress is expressed as a function of
8
the quasi-static stress-strain relationship and of the current strain rate. Experimental data for
different rates can be used to obtain the parameters of the power law relationship, thus expressing
the stress enhancement as a function of variable strain rates. Since no such experimental data are
available for the present study, the available quasi-static relations are used, and a stress enhance-
ment coefficient of 1.5 times the static stress is assumed to model the crushing stress on the
honeycomb material. In a previous finite element (FE) simulation of the air-gun environment (6),
the authors have successfully demonstrated the effectiveness of such a constant stress enhancement
factor of 1.5 in modeling the strain rate effect on the honeycomb material.
1.4 Results and Discussion
1.4.1 Validation of Present Model

To validate the present model and assess its performance, the predicted responses with the
present model are compared with the available test and simulation results. Measured responses
from the air gun launch simulation of the test item shown in figure 1b, as well as the results from
a detailed FE model built to represent the system behavior of the air gun environment, are used
for validation and verification of the present model.
To model the honeycomb material, the static stress-strain curve from figure 3 was used, scaled to
1.5 times the static relationship to incorporate the strain rate effect. This approach was identical
to the one used in the FE analysis (6). The total mass of the OBR was distributed between the
two masses m1 and m2, with a ratio 1:1. To determine the k1 (spring stiffness), the period of
oscillation (T) of the OBR was estimated to be about 1.36×10–4 s, from the FE acceleration curve
as filtered at 2500 Hz. The corresponding frequency was determined to be 1/T, or 7353 Hz, and
the angular frequency ω equals 46,200 rad/s. Then from equation 5 we have
ω 2 (m1 + m 2 ) 46200 2 ⋅ 3.75

k1 = = = 2.0 × 10 9 N/m
4 4
The values of all parameters used in the present analysis are now provided:
m1 – mass of OBR, back: 1.25 kg
m2 – mass of OBR, front: 2.5 kg
m3 – mass of MEM: 31.2 kg
k1 – OBR spring stiffness: 2.0×109 N/m
v0 – OBR initial velocity: 83.566 m/s
ΔT – End-time of analysis: 0.002 s
Δt – Time-step for analysis: 1.0×10–6 s
δt – Output time step: 4.0×10–6 s
εc – compacting strain: 0.64
9
σy – yield stress: 55.16 MPa
σcrush –1.5⋅(static yield) = 1.5⋅55.16 = 82.74 MPa
E – elastic modulus: 4,060 MPa
ρ0 – density of honeycomb: 650 kg/m3
r – radius of mitigator: 0.05 m
Lw – length of mitigator wedge: 0.038 m
Lc– length of cylindrical section: 0.219 m
These values, converted for the convenience of comparing the results, were entered into the input
file provided in appendix A. The analysis was performed with the FORTRAN (Formula Trans-
lator) program, and the corresponding results are presented and compared to the FE analysis and
test results in figures 7 to 16. Figure 7 shows the unfiltered acceleration curves of the OBR top
(attachment point of mass m2 for present approach). Predicted accelerations filtered at 7000 Hz,
2500 Hz, and 1000 Hz are presented in figures 8, 9, and 10, respectively. Since the only frequency
present in the current model is low, the results from the present approach have been filtered only in
figure 10, which portrays the results of applying a 1000-Hz filter to all three curves. In figures 7
through 9, only the FE and test curves were filtered. As seen from the first two figures, the present
approach matches the average of both the test and the FEA curves. Figure 9 shows that the present
model gives excellent results for both the acceleration peak and the acceleration pulse duration.
Figure 10, data filtered at 1000 Hz, indicates excellent agreement between the present and the FE
approach. Figure 11 shows frequency domain plot comparison of the acceleration data. As seen in
figure 11, the first fundamental frequency of the present model matches that of the test item. This
comparison suggests that the modeling of OBR spring stiffness, k1, based on test item’s first
fundamental mode, reasonably captures the dynamics of the proposed mass-spring discrete system
of the air gun launch environment.
Figure 7. Top acceleration of the OBR unfiltered.
10
Figure 8. Top acceleration of the OBR filtered at 7000 Hz.
11
Figure 11. Fast Fourier transform (FFT) of the top acceleration of the OBR.
12
OBR velocities for the test, FE simulation, and present approach are compared in figure 12. As
shown in the figure, a very good agreement is observed between the FE and present approach.
Test velocity history is obtained from the integration of test acceleration data. A similar compari-
son of OBR displacement results for all three cases is presented in figure 13. Figure 13 shows
that the present approach matches the maximum displacement from the test very closely. The
discrepancies we observe to the right of the peak displacements (when the OBR starts to retract)
can be attributed to friction in the “catch tube” and to other mechanisms not accounted for within
the FE and the present analysis.
Figure 12. Top velocity of the OBR.
Figure 14 shows the contact forces between the mitigator and the OBR and between the
mitigator and the MEM. In the FE analysis, these forces differ because of the elastic wave
propagation speed, inertial effects in the mitigator, etc. These phenomena contribute
insignificantly to the dynamics of the system, especially if the OBR is to be the primary object of
interest; therefore, they were neglected within the present approach. This simplification makes
the contact forces at the two interfaces equal, but as we see, they are very close to the average of
both curves. Furthermore, when we compare impulses by integrating the force over time, we get
a very good agreement among the impulses at both interfaces, as shown in figure 15.
13
Figure 13. Top displacement of the OBR.
Figure 14. Mitigator contact forces.
14
Figure 15. Mitigator contact force impulse.
The present approach provides data for the length of the compacted zone within the mitigator.
The position of the crush front within the undeformed mitigator (“initial”) and the length of the
crushed part (“compacted”) are shown in figure 16.
Figure 16. Length of mitigator plastic region.
15
1.5 Parameter Variation Sensitivity
In this section, we illustrate the model’s response to varying the parameters of the system. As
explained earlier, the representation of the OBR by two separate masses and a spring aims to
capture one additional frequency of the OBR which is considered of interest. However, this
representation requires the total physical mass of the OBR to be distributed between the two
representative masses in the model: m1 and m2. Furthermore, the connecting spring stiffness
has to be determined. These processes introduce additional uncertainties into the model; this is
especially true of the mass distribution process since it is not uniquely defined but depends on
the user’s judgment. In general, varying the m1:m2 mass ratio will only change the oscillation
amplitude, not the oscillation frequency. Figure 17 shows the acceleration curves of masses m1
and m2 for two different distribution ratios. First, the total mass of the OBR (3.75 kg in this
example) was distributed to 1.25 kg for m1 and 2.5 kg for m2 (ratio 1:2). The corresponding
accelerations are shown with curves A and B in figure 17. Curves C and D show the accelerations
of the masses for a ratio of 2:1 (m1 = 2.5 kg and m2 = 1.25 kg). The frequency of oscillation for all
four curves is the same. Only the amplitude of oscillation is changed.
Figure 17. m1:m2 mass ratio sensitivity of model.
The stiffness of the spring connecting the two OBR masses, k1, is determined by equation 5
which requires the determination of the angular frequency of the OBR. Since this would
normally be determined experimentally, some error might be introduced into the spring stiffness
value. Therefore, it is interesting to see how small variations in the value of k1 would affect the
model output. Effects of the variation of k1 values are illustrated in figure 18, which compares
16
the initial acceleration response to that obtained for k1 values 10% smaller and larger than the
initial stiffness determined by equation 5. As seen, the system behavior, especially in the area of
interest (first millisecond), does not change significantly. Furthermore, when drastically
changing the spring stiffness value, as shown in figure 19, we still get the correct shape and peak
of the deceleration pulse.
Figure 18. Model sensitivity to 10% variation of the value of k1.
Figure 19. Model sensitivity to large variation of the value of k1.
17
The integration time step is a very important parameter of the central difference time-stepping
scheme. In order to get an accurate solution, the time step needs to be smaller than a certain
value called the critical time step. If this requirement is not satisfied, the solution quickly
diverges. Therefore, determining the critical time step for a given solution is very important to
ensure accuracy in the results. However, in most cases this is not an easy task, and for complex
systems, only threshold limits of the critical time step are determined. In the present model,
there is no need to determine the critical time step explicitly. Since the solution does not take a
significant amount of run time, we can easily re-run it with a smaller time step and compare the
two solutions to assess the accuracy of the results. As illustrated in figure 20, the initial solution
using a time step Δt = 1.0×10–6 s is identical to the solution generated with a 10-times-smaller
time step, Δt = 1.0×10–7 s. Even with a much coarser time resolution of Δt = 2.0×10–5 s, we get a
reasonably accurate solution. This shows that the chosen value of the time step is well below the
critical and illustrates the good overall stability and robustness of the solution scheme.
Figure 20. Model sensitivity to variation of the solution time step.
Finally, to illustrate how some of the geometric parameters of the system affect the solution
response, the radius of the mitigator honeycomb cylinder was varied. The corresponding OBR
acceleration responses are shown in figure 21. As would be expected, a smaller radius produced
a lower acceleration peak and a wider pulse, whereas a larger radius produced a higher
acceleration peak and a narrower pulse.
18
Figure 21. Model sensitivity to 10% variation of the value of mitigator radius.
1.6 Conclusion
The objective of this effort was to develop an overall analytical model to predict the transient
response of a generic artillery component subjected to launch simulation. The simulation is
achieved by impact mitigation techniques in which the component is propelled to a target inside
a carrier equipped to measure component response during the simulation. A 2-degree-of-
freedom model is developed to simulate the air gun launch environment in which a test object
mounted on a projectile is fired through the air gun and decelerated by the crushing of an
aluminum honeycomb mitigator, which in turn impacts the MEM before being stopped at the
retrieving end. The presented model achieved a good prediction of the period and peak
acceleration of the OBR.
Sensitivity of parametric study indicates that the distribution of masses in the spring-mass model
of the test item does change the oscillation amplitude without affecting the system’s frequency of
vibration. A slight variation of spring stiffness (±10%) does not change the shape and neither
does it affect the peak of the deceleration pulse of the test item. The sensitivity study of initial
time step illustrates the good overall stability and robustness of the proposed solution scheme.
19
2. Part II: Air Gun Finite Element Simulation
2.1 Review of Previous Work

Several LS-DYNA models were developed to simulate an air gun launch environment. In this test,
a test object is mounted on a projectile and fired through the air gun. The object is decelerated by
the crushing of an aluminum honeycomb mitigator MEM complex before being stopped at the
retrieving end (6).
The objective of the effort was to develop an overall simulation model to predict the three-
dimensional transient response of a generic artillery component subjected to launch simulation in
the U.S. Army Research Laboratory’s air gun facility at Adelphi, Maryland. The simulation is
achieved by impact mitigation techniques whereby the component is propelled to a target inside a
4-inch carrier equipped to measure component response during the simulation. These models
include all elements of the simulation apparatus that contribute to the total response of the
component under test.
Two formulations are employed in the simulations: Lagrangian and arbitrary Lagrangian/Eulerian
(ALE). The aluminum honeycomb mitigator undergoes a significant amount of deformation,
which could render a Lagrangian simulation severely unstable. For this reason, an ALE simulation
is also considered. The Eulerian method is more suitable for problems involved in severe mesh
distortion. The Lagrangian method, on the other hand, is limited in how much an element can
deform. The Lagrangian method is easy to set up and visualize since the material point moves
with the mesh. However, the Eulerian method is more difficult to set up and the mesh is stationary
so that material points are advected from one element to the next.
The aluminum honeycomb mitigator is modeled with *MAT_MDIFIED_HONEYCOMB in the
LS-DYNA simulation code. This is material model number 126 in the code. This material model
is available for the Lagrangian method only and it is not yet implemented in the Eulerian solver of
the LS-DYNA. Therefore, an alternate material model is considered that can simulate the
behavior of aluminum honeycomb and is available in the LS-DYNA Eulerian solver. The
material model considered for this case is *MAT_CRUSHABLE_FOAM which is material model
number 63 in LS-DYNA. This material model can simulate an isotropic crushable material. The
*MAT_MDIFIED_HONEYCOMB can simulate an orthotropic crushable material. Most of the
aluminum honeycomb material in the air gun simulation is crushed axially. Therefore, the
*MAT_CRUSHABLE_FOAM model is considered to be appropriate for this model. Several
LS-DYNA models are developed to simulate the air gun launch environment with Lagrangian
and ALE computational methods. The Lagrangian method is simpler to set up, simpler to post
process, and require less computational time. However, it requires significant expertise in the
explicit FE codes to make the simulation numerically stable. This is because of the significant
20
large deformation of the mitigator. On the other hand, the ALE method is more difficult to set
up, more difficult to post process, and requires much more CPU (central processing unit) time.
However, the ALE simulation is more suitable for very large deformation-like flow problems.
Both computational methods lead to the same prediction for the acceleration of the OBR. An
accurate prediction of the period and peak acceleration of the OBR is achieved with the presented
models and the methods employed.
2.2 Finite Element Model Sensitivity Study

It is very important that the FE simulation captures the contributions of high frequency oscillations
to the net acceleration of the model. This will allow failure mode simulation of the embedded
MEMS (micro-electrical-mechanical systems). Improving the correlation between test and FE
results allows the FE simulation to become a tool for detecting/predicting failure modes of
physical components subject to an air gun test. For this purpose, the developed FE model is
further examined to fully understand the differences between the model and the experimental data
and to acquire more confidence in the predictability of the simulation. Several model sensitivity
issues are considered in detail as follows:
• double versus single precision code,
• time step size effect,
• different output intervals,
• different damping options in LS-DYNA (all options are investigated),
• representation of the OBR’s glass beads with solid elements, and
• intermittent eigenvalue analysis.
Each one of these items is studied in detail, and the corresponding results are presented.
2.2.1 Double Versus Single Precision Code

The default LS-DYNA code is a single precision code. When the integration time step is small
enough to satisfy the Courant stability condition, many integration cycles are possible. Since the
LS-DYNA code is an explicit code based on second order central difference equations, numerical
error propagation is possible. The same model was run with single and double precision code.
The results are presented in figure 22. Fast Fourier transforms (FFTs) of the results are presented
in figure 23. One can observe that both codes give the same results.
21
Figure 22. Acceleration data for the same FE model with two different precisions.
Figure 23. FFT of the acceleration data.
22
2.2.2 Time Step Size Effect
Integration time step size variation is used to measure the stability of a model. Time step
variation is also used to detect any anti-aliasing in the collection of the data from the explicit time
integration simulation. The default integration time step is normally taken to be 90% of the
stability limit to avoid any filtering error such as beating in the data. This report considers two
additional integration time steps. One is taken to be 80% of the critical time step and the other is
taken to be 40% of the critical time step. The results of all three time steps are depicted in figure
24. FFTs of the data are depicted in figure 25. One can observe that all the important frequencies
are the same in all the considered integration time steps. Therefore, the default integration time
step is appropriate for this problem.
Figure 24. Acceleration data for different integration time steps.
Figure 25. FFT of the acceleration data for different integration time steps.
23
2.2.3 Different Output Intervals
The data collection from the FE simulation should be the same as the test setup; therefore, the
two data can be compared meaningfully. The sampling rate in the test is 4 x 10-6 seconds, which
corresponds to 500 data sampling points in the 2-millisecond simulation. To investigate if any
data are missing from the simulation or if any abnormality exists in the numerical calculations,
data sampling rates of 10 and 20 times the test sampling rate are considered (5,000 and 10,000
points, respectively). In addition, data are collected at every integration time step. Figure 26
depicts the acceleration data for the entire sampling rate considered. FFTs of the data are
depicted in figure 27. All major contributing and significant frequencies are present in the
lower rate, which is also the sampling rate of the experimental data.
Figure 26. Acceleration data sampled at different frequencies.
Figure 27. FFT of the acceleration data sampled at different frequencies.
24
2.2.4 All Damping Options in LS-DYNA Investigated
There are three options for damping in the LS-DYNA code: global, mass, and stiffness propor-
tional damping. Global damping is similar in formulation as the mass proportional damping, and
therefore, results for this case are not presented. Different damping constants are considered for
each damping option. For the mass proportional damping, constants of magnitudes 100, 500,
1000, and 1500 are considered. Results for these cases are depicted in figure 28. FFTs of the
acceleration data are depicted in figure 29. One can observe that frequencies are damped signifi-
cantly, especially the lower frequencies of the system. Figure 30 depicts the test data and the
damped acceleration data from the simulation.
Figure 28. Acceleration data for simulation with mass proportional damping.
Figure 29. FFT of the acceleration data for simulation with mass proportional damping.
25
Figure 30. Acceleration data for simulation and test with mass proportional damping.
Stiffness proportional damping is considered in the simulation as 1%, 10%, 25%, 25%, and 35%
of the critical damping. The acceleration data are depicted in figure 31. FFTs of the data are
presented in figure 32. One can observe that the low frequencies are not damped, while the
higher frequencies are damped. Figure 33 depicts accelerations from the test data, the simulation
with no damping, and the simulation with 10% of the critical damping. Figure 34 depicts the
FFT of these accelerations. One can observe, again, that higher order frequencies are affected
more significantly than lower order frequencies. In general, mass proportional damping should
be used for low frequency content and pure deformation with no rigid body motion since it
damps the rigid body motion. Stiffness proportional damping is orthogonal to rigid body motion
and would not affect this type of motion. Stiffness proportional damping is good for high
frequency content and should be used in such problems where attenuation of high frequency
vibration is desired.
Figure 31. Acceleration data for simulation with stiffness proportional damping.
26
Figure 32. FFT of the acceleration data for simulation with stiffness proportional damping.
Figure 33. Acceleration data for simulation and test with stiffness proportional damping.
27
Figure 34. FFT of the acceleration data for simulation and test with stiffness proportional damping.
2.2.5 The OBR Filled With Solid Elements to Represent Beads

The OBR is filled with beads in the experiment. However, the OBR is hollow in the simulation.
These beads exist in the OBR to protect the accelerometer during tests. There is no material
model in any FE code to represent the behavior of these beads. In this report, a model of the
OBR is considered that is filled with solid elements. Different material models such as soil,
foam, viscoelastic, and crushable foam are considered for the beads in order to simulate the
beads’ damping high frequency oscillations. Results of the simulation are depicted in figure 35.
One can observe that high frequency oscillations are more significant because of waves traveling
back and forth in the solid element.
Figure 35. Acceleration data for simulation with filled OBR.
28
2.2.6 Intermittent Eigenvalues Analysis
To understand the frequency content in this simulation, an intermittent eigenvalue analysis is
conducted. The LS-DYNA code allows for the extraction of frequency contents during an impact
simulation. This is performed to relate the frequency content to the event and deformation mode in
the OBR. The FFT of the unfiltered acceleration of the OBR is depicted in figure 36. One can see
several important peaks in the data. The largest peak corresponds to the first fundamental mode.
The next peak is at about 7400 Hz. The next ones are at about 8000, 14600, 21000, 23800, 27700,
and 29500 Hz. The intermittent eigenvalue analysis revealed the sources of these frequencies.
One hundred modes were extracted from the analysis. Only the modes corresponding to the peaks
in the data are presented here. Figure 37 depicts the undeformed (line mesh) and the deformed
mode (solid shaded mesh) corresponding to the peaks in the frequency spectrum graph (figure 36).
It is clear that these peaks correspond to the axial mode of the OBR. High frequency content was
not observed in the experimental data. However, high frequency content is observed in the
simulation data. Since the beads are not present in the simulation and the high frequencies are
attributable to the axial deformation modes of the OBR, one can conclude that these high
frequencies in the simulation results are not realistic. This conclusion is based on the assumption
that high frequency oscillations in the axial direction of the OBR are absorbed by the glass beads
surrounding the accelerometer.
Figure 36. FFT of the acceleration data for simulation
29
a. b.
c. d.
e. f.
Figure 37. Mode shapes for a fundamental frequency.
30
2.3 Simplified Simulation
The full FE model runs for few hours on a single processor CPU. For parametric studies, this run
time is considered to be long. A simplified methodology in which an FE model can run in few
minutes is developed. This methodology requires that a set of simulations be run for each OBR
configuration (see figure 1b, for example) with the discrete spring-mass model shown in figure 2,
and extract the force exerted between the test item and the mitigator. The contact forces (F in
figure 2) between the OBR and the mitigator could subsequently be used in a simplified FE
model of the test item only. For a given contact force corresponding to a given OBR weight and
configuration, an FE mesh of the OBR (test item) could be analyzed that requires a shorter run
time. Any modification in the OBR content that may be required because of the addition of a
new electronics package to be tested could easily be modeled in the simplified FE mesh of the
OBR for a given force. To simulate the response of the OBR, the contact force is applied as an
external force to the OBR to resist the momentum of the OBR. This methodology is applied to
the current model and presented next.
For the test item used in figure 1b, the contact force between the mitigator and the OBR is
presented in figure 38. This force is then applied as an external force to the OBR FE model (the
only mesh is the OBR) for a given initial impact velocity. The acceleration data are collected
from the simplified model and compared to full model in figure 39. One can observe that they
are practically the same. The simplified model runs in less than a minute.
Figure 38. Contact force between the OBR and the mitigator.
31
Figure 39. Acceleration data of the simplified and full model.
2.4 Conclusion
The sensitivity of FE model parameters affecting the dynamics of a test object launched in an air
gun test is discussed in this part of the report. An LS-DYNA FE model previously developed to
simulate the impact mitigation environment is used to investigate the sensitivity of FE model
parameters. Previous study indicates the necessity of honeycomb material characterization under
very high impact loading for effective modeling of the strain rate effects. The sensitivity study
clearly indicates that the model is numerically stable and can aid in failure prediction if damping
is added. A comparison of FE simulation responses with various damping options suggests that
the mass proportional damping attenuates the low frequency vibration in the simulated response
while the stiffness proportional damping attenuates the high frequency oscillations. The FE
analysis also suggests that intermittent eigenvalue analyses during an impact simulation could be
useful in identifying the dominant modes of vibration governing the dynamics of the test items
and verifying the predictive capability of the FE model. The reasons that the FE model predicted
high frequency content (above 14,000 Hz) that is not observed in the test could include (a) a lack
of information for material characterization of the honeycomb mitigator under very high strain
rate, or (b) lack of a proper material model for the beads which has significant damping at high
frequencies.
An efficient procedure using a simplified decoupled model is proposed to simulate the response
of a test item launched in an air gun test. In this approach, the contact force generated from the
discrete analytical model is applied on a detailed FE mesh of the test item for a given impact
velocity. The acceleration data obtained by this procedure compared well with those of the full
32
FE analysis. This model is programmed in a FORTRAN code that can produce results with only
a few minutes of run time which are comparable to the FE and test results. This procedure can
aid in the prediction of the failure of MEMS in the OBR with deformable FE model.
33
3. References
1. Lu, Wei-Yang; Hinnerichs, T. Crush of High Density Aluminum Honeycombs.

Proceedings of 2001 ASME International Mechanical Engineering Congress and Exposition,
AMD-Vol. 251, pp. 203–211, 2001.
2. Ashby, M.F.; Evans, A.G.; Hutchinson, J.W.; Fleck, N. A. Metal Foams: A Design Guide,
University of Cambridge: Cambridge, 1998.
3. Reid, S.R.; Peng, C. Dynamic Uniaxial Crushing of Wood, International Journal of Impact
Engineering, Vol. 19, Nos. 5–6, pp. 531–570, 1997.
4. Zhao, H.; Elnasri, I.; Abdennadher, S. An experimental study on the behavior under impact
loading of metallic cellular materials, International Journal of material Sciences, Vol. 47,
pp. 757–774, 2005.
5. Zhang, J.; Kikuchi, N.; Li, V.; Yee, A.; Nusholtz, G. Constitutive Modeling of Polymeric
Foam Material Subjected to Dynamic Crash Loading, International Journal of Impact
Engineering, Vol. 21, No. 5, pp. 369–386, 1998.
6. Chowdhury, M.; Tabiei, A. Transient Response of Projectile in Gun Launch Simulation
Using LS-DYNA Models; ARL-TR-125; U.S. Army Research Laboratory: Aberdeen Proving
Ground, MD, 2003.
34
Appendix A. Program I/O Files and FORTRAN Source Code
Input File
The input parameters necessary for the impact model are entered in an ASCII text file named
“input.dat,” which needs to be located in the same directory as the program executable. A
sample input file with brief description of the input parameters follows:
1.25 M1 - mass of OBR back
2.5 M2 - mass of OBR front
31.2 M3 - mass of MEM
2.0E6 K1 - OBR spring stiffness
83.566 V0 - OBR initial velocity
0.002 End time of analysis
1.0E-6 Time step for analysis
4.0E-6 Output time step
0.64 Compacting strain
82.74E3 Yield stress
4.060E6 Elastic modulus
650.0 Density of honeycomb
0.05 Radius of mitigator
0.038 Length of wedge part
0.219 Length of cylindrical part
Notes:
The input parameters must be in consistent units. The program does not check for unit
consistency nor does it convert the input values to ensure consistency. The units used in the
above input are kilogram, second, meter, Newton.
The program reads only the first input parameter from each line of the input file. Everything
after the first parameter is treated as comments. However, the first parameter should have the
appropriate value of the corresponding variable and be in the appropriate numeric format.
The input parameters should be entered in the exact order as shown in the example above. There
should be no empty lines at the beginning of the input file.
35
Output Files
The program produces several output files:
output.dat – an ASCII file with all the output information in different columns. There are 12
columns having the values of current analysis time, displacement, velocity, and acceleration of
masses M1, M2, and M3, current force value, and current plastic front location.
LS_POSTdis.dat – an ASCII file with the displacements of the three mass attachment points.
This file is in LSTC LS-POST format.
LS_POSTvel.dat – an ASCII file with the velocities of the three mass attachment points. This
file is in LSTC LS-POST format.
LS_POSTacc.dat – an ASCII file with the accelerations of the three mass attachment points. This
file is in LSTC LS-POST format.
LS_POSTfor.dat – an ASCII file with the contact force between the mitigator and OBR (MEM).
This file is in LSTC LS-POST format.
36
FORTRAN Source Code
The FORTRAN source code of the program implementing the formulation described in section
“Model Formulation” follows below.
program main
double precision um1(3),un(3),up1(3),udn(3),uddn(3),xk1u,
* buff(10),cpl,xfront,dxfront,Force
c Initialize
xpi=3.1415926536
open (10,file='input.dat')
read(10,*)xM1
read(10,*)xM2
read(10,*)xM3
read(10,*)xK1
read(10,*)xV0
curtime=0.
read(10,*)endtime
read(10,*)dt
dt2=dt**2
curout=0.
read(10,*)dtout
mcount=0
read(10,*)epsc
read(10,*)sigy
read(10,*)Emod
epsy=sigy/Emod
read(10,*)rho0
rhoy=rho0/(1.-epsy)
rhofrnt=rho0/(1.-epsc)
read(10,*)radius
read(10,*)xLwed
read(10,*)xLcyl
close(10)
xfront=0.
area0=xpi*radius**2
Force=0.
c Time = 0
37
c Displacement
un(1)=0.
un(2)=0.
un(3)=0.
c Velocity
udn(1)=xV0
udn(2)=xV0
udn(3)=0.
c Acceleration
uddn(1)=0.
uddn(2)=0.
uddn(3)=0.
c Displacement at -dt
do i=1,3
um1(i)=-dt*udn(i)
end do
c Define output
open (10,file='output.dat')
c End Initialization, Define Time Stepping Loop
c
c Stop when acceleration of MEM gets zero
c do while (uddn(3).gt.0.0)
c
c Or stop at endtime
c
do while (curtime.lt.endtime)
c Write values for displacement, velocity and acceleration
if (curtime.ge.(curout+dtout)) then
curout=curtime
write(10,10)curtime,(un(i),udn(i),uddn(i),i=1,3),Force,xfront
mcount=mcount+1
end if
c Calculate mitigator force
cpl=(udn(2)-udn(3))/epsc
if (cpl.gt.0.) then
c Plastic region grows
dxfront=cpl*dt
xfront=xfront+dxfront
38
if (xfront.lt.xLwed) then
areac=area0*xfront/xLwed
else
areac=area0
end if
Force=sigy*areac
else
c Elasricity only; unloading
Force=Force+
* area0*Emod*(udn(2)-udn(3))*dt/(xLwed+xLcyl+un(3)-un(2)-xfront)
end if
if (Force.lt.0.) Force=0.
c Get new displacement values
xk1u=xK1*(un(1)-un(2))
up1(1)=2.*un(1)-um1(1)-dt2*xk1u/xM1
up1(2)=2.*un(2)-um1(2)+dt2*(xk1u-Force)/xM2
up1(3)=2.*un(3)-um1(3)+dt2*Force/xM3
do i=1,3
c Velocity and acceleration
udn(i)=.5*(up1(i)-um1(i))/dt
uddn(i)=(up1(i)-2.*un(i)+um1(i))/dt2
c Shift displacements in matrix
um1(i)=un(i)
un(i)=up1(i)
end do
c Step forward in time
curtime=curtime+dt
end do
c Write LS-POST output files
open (11,file='LS_POSTdis.dat')
write(11,20)
do i=1,3
rewind(10)
write (11,30)i,mcount
do j=1,mcount
read(10,*)curtime,(buff(k),k=1,10)
write(11,40)curtime, buff(3*i-2)
end do
39
write(11,50)
end do
close(11)
open (11,file='LS_POSTvel.dat')
write(11,120)
do i=1,3
rewind(10)
do j=1,mcount
write(11,140)curtime, buff(3*i-1)
end do
write(11,50)
end do
close(11)
open (11,file='LS_POSTacc.dat')
write(11,220)
do i=1,3
rewind(10)
do j=1,mcount
write(11,240)curtime, buff(3*i)
end do
write(11,50)
end do
close(11)
open (11,file='LS_POSTfor.dat')
write(11,320)
rewind(10)
write (11,330)mcount
do j=1,mcount
write(11,340)curtime, buff(10)
end do
write(11,50)
close(11)
close(10)
40
10 format(11(E12.6,','1X),E12.6)
20 format ('Curveplot',/'ARL GUN SIMULATION',/'Time',
* /'Displacement',/'Mass')
30 format('Displ. of M',I1,' #pts=',I10)
40 format(E12.6,2X,E12.6)
50 format('endcurve')
* /'Velocity',/'Mass')
130 format('Veloc. of M',I1,' #pts=',I10)
140 format(E12.6,2X,E12.6)
* /'Acceleration',/'Mass')
230 format('Accel. of M',I1,' #pts=',I10)
240 format(E12.6,2X,E12.6)
* /'Force',/'Mass')
330 format('Contact Force',' #pts=',I10)
340 format(E12.6,2X,E12.6)
stop
end
41
NO. OF NO. OF
COPIES ORGANIZATION COPIES ORGANIZATION
1 DEFENSE TECHNICAL 1 LAWRENCE LIVERMORE NATL LABS

(PDF INFORMATION CTR ATTN MS9042 WEI-YANG LU
ONLY) DTIC OCA PO BOX 808 L 125
8725 JOHN J KINGMAN RD LIVERMORE CA 94551-0969
STE 0944
FORT BELVOIR VA 22060-6218 5 LAWRENCE LIVERMORE NATL LABS
ATTN R CHRISTENSEN S DETERESA
1 US ARMY RSRCH DEV & ENGRG CMD F MAGNESS M FINGER MS 313
SYSTEMS OF SYSTEMS M MURPHY L 099
INTEGRATION PO BOX 808
AMSRD SS T LIVERMORE CA 94550
6000 6TH ST STE 100
FORT BELVOIR VA 22060-5608 1 SANDIA NATL LABS
ATTN MS0847 T HINNERICHS
1 INST FOR ADVNCD TCHNLGY PO BOX 969
THE UNIV OF TEXAS AT AUSTIN ALBUQUERQUE NM 87123
3925 W BRAKER LN STE 400
AUSTIN TX 78759-5316 3 DIR SANDIA NATL LABS
APPLIED MECHANICS DEPT
1 DIRECTOR ATTN MS 9042 J HANDROCK
US ARMY RESEARCH LAB Y R KAN J LAUFFER
IMNE ALC IMS PO BOX 969
2800 POWDER MILL RD ALBUQUERQUE NM 87123
ADELPHI MD 20783-1197
5 UNIV OF CINCINNATI
1 DIRECTOR AEROSPACE ENGINEERING DEPT
US ARMY RESEARCH LAB ATTN PROF ALA TABIEI
AMSRD ARL CI OK TL CINCINNATI OH 45221
2800 POWDER MILL RD
ADELPHI MD 20783-1197 1 CDR US ARMY ARDEC
ATTN AMSTA AR CC COL JANKER
2 DIRECTOR PICATINNY ARSENAL NJ 07806-5000
US ARMY RESEARCH LAB
AMSRD ARL CS OK T 6 CDR US ARMY ARDEC
2800 POWDER MILL RD ATTN AMSTA AR CCH A F ALTAMURA
ADELPHI MD 20783-1197 M NICOLICH M PALATHINGUL
R HOWELL A VELLA
21 DIRECTOR M YOUNG
US ARMY RSCH LABORATORY PICATINNY ARSENAL NJ 07806-5000
ATTN AMSRD ARL WM MB A FRYDMAN
M CHOWDHURY 1 CDR US ARMY ARDEC
2800 POWDER MILL RD ATTN AMSRD AAR AEM D VO
ADELPHI MD 20783-1197 PICATINNY ARSENAL NJ 07806-5000
1 DIRECTOR 5 CDR US ARMY ARDEC

US ARMY RSCH LABORATORY ATTN AMSTA AR CCH A L MANOLE
ATTN AMSRD ARL SE DE R ATKINSON S MUSALLI M LUCIANO
2800 POWDER MILL RD E LOGSDEN T LOUZEIRO
ADELPHI MD 20783-1197 PICATINNY ARSENAL NJ 07806-5000
1 RD&E COMMAND 1 CDR US ARMY ARDEC

SYSTEMS OF SYSTEMS INTEGRATION ATTN AMSRD AAR AEM L R CARR
ATTN AMSRD SS T PICATINNY ARSENAL NJ 07806-5000
6000 6TH STREET STE 100
FT BELVOIR VA 22060-5688
42
NO. OF NO. OF
5 CDR US ARMY ARDEC 1 CDR US ARMY ARDEC

ATTN AMSTA AR CCH B P DONADIA ATTN AMSTA AR QAC T C J PAGE
F DONLON P VALENTI PICATINNY ARSENAL NJ 07806-5000
C KNUTSON G EUSTICE
PICATINNY ARSENAL NJ 07806-5000 1 CDR US ARMY ARDEC
ATTN AMSTA AR TD
4 CDR US ARMY ARDEC PICATINNY ARSENAL NJ 07806-5000
ATTN AMSTA AR CCH B K HENRY
J MCNABOC R SAYER 1 CDR US ARMY ARDEC
F CHANG ATTN AMSTA AR WEA J BRESCIA
PICATINNY ARSENAL NJ 07806-5000 PICATINNY ARSENAL NJ 07806-5000
2 CDR US ARMY ARDEC 1 CDR US ARMY ARDEC

ATTN AMSTA AR CCH C H CHANIN ATTN AMSTA AR WEL F M GUERRIERE
S CHICO PICATINNY ARSENAL NJ 07806-5000
PICATINNY ARSENAL NJ 07806-5000
1 CDR US ARMY ARDEC
1 CDR US ARMY ARDEC PRODUCTION BASE MODERN ACTY
ATTN AMSTA AR CCH P J LUTZ ATTN AMSMC PBM K
PICATINNY ARSENAL NJ 07806-5000 PICATINNY ARSENAL NJ 07806-5000
2 CDR US ARMY ARDEC 1 PM ARMS

ATTN AMSRD AAR AEM I ATTN SFAE GCSS ARMS
D CONWAY BLDG 171
ATTN AMSRD AAR AEW E(D) PICATINNY ARSENAL NJ 07806-5000
R SCHLENNER
PICATINNY ARSENAL NJ 07806-5000 1 PM MAS
ATTN SFAE AMO MAS
ATTN AMSTA AR FSA A WARNASH
B MACHAK M CHIEFA 1 PM MAS
PICATINNY ARSENAL NJ 07806-5000 ATTN SFAE AMO MAS
CHIEF ENGINEER
ATTN AMSTA AR FSE
ATTN SFAE AMO MAS PS
ATTN AMSTA AR FSF T C LIVECCHIA
ATTN SFAE AMO MAS LC
ATTN AMSTA AR FSP G M SCHIKSNIS
D CARLUCCI 1 PM MAS
PICATINNY ARSENAL NJ 07806-5000 ATTN SFAE AMO MAS MC
1 CDR US ARMY ARDEC
ATTN AMSTA AR M D DEMELLA 1 CDR US ARMY TACOM
PICATINNY ARSENAL NJ 07806-5000 PM ABRAMS
ATTN SFAE ASM AB
1 CDR US ARMY ARDEC 6501 ELEVEN MILE RD
ATTN AMSTA AR QAC T D RIGOGLIOSO WARREN MI 48397-5000
43
NO. OF NO. OF
1 CDR US ARMY TACOM 4 CDR US ARMY TACOM

ATTN AMSTA SF ATTN AMSTA TR R R MCCLELLAND
WARREN MI 48397-5000 D THOMAS J BENNETT
D HANSEN
1 CDR US ARMY TACOM WARREN MI 48397-5000
PM BFVS
ATTN SFAE GCSS W BV 1 CDR US ARMY TACOM
6501 ELEVEN MILE RD ATTN AMSTA JSK A SCHUMACHER
WARREN MI 48397-5000 WARREN MI 48397-5000
1 CDR US ARMY TACOM 2 CDR US ARMY TACOM

CHIEF ABRAMS TESTING ATTN AMSTA TR R S GOODMAN
ATTN SFAE GCSS W AB QT T KRASKIEWICZ D TEMPLETON MS-263
6501 ELEVEN MILE RD WARREN MI 48397-5000
WARREN MI 48397-5000
3 CDR US ARMY TACOM
1 DIR AIR FORCE RSCH LAB ATTN AMSTA TR D D OSTBERG
ATTN MLLMD D MIRACLE L HINOJOSA B RAJU
2230 TENTH ST WARREN MI 48397-5000
WRIGHT PATTERSON AFB OH 45433-7817
2 CDR US ARMY TACOM
1 OFC OF NAVAL RESEARCH ATTN AMSTA CS SF H HUTCHINSON
ATTN J CHRISTODOULOU F SCHWARZ
ONR CODE 332 WARREN MI 48397-5000
800 N QUINCY ST
ARLINGTON VA 22217-5600 10 BENET LABORATORIES
ATTN AMSTA AR CCB R FISCELLA
1 CDR WATERVLIET ARSENAL M SOJA E KATHE M SCAVULO
ATTN SMCWV QAE Q B VANINA G SPENCER P WHEELER
BLDG 44 S KRUPSKI J VASILAKIS
WATERVLIET NY 12189-4050 G FRIAR R HASENBEIN
WATERVLIET NY 12189-4050
1 TNG DOC & CBT DEV
ATTN ATZK TDD IRSA A POMEY 4 BENET LABORATORIES
FT KNOX KY 40121 ATTN AMSTA CCB R S SOPOK
E HYLAND D CRAYON
2 HQ IOC TANK R DILLON
AMMUNITION TEAM WATERVLIET NY 12189-4050
ATTN AMSIO SMT R CRAWFORD
W HARRIS 2 US ARMY CORPS OF ENGINEERS
ROCK ISLAND IL 61299-6000 ATTN CERD C T LIU
CEW ET T TAN
1 CDR US ARMY AMCOM 20 MASSACHUSETTS AVE NW
AVIATION APPLIED TECH DIR WASHINGTON DC 20314
ATTN J SCHUCK
FT EUSTIS VA 23604-5577 1 US ARMY COLD REGIONS
RSCH & ENGRNG LAB
1 NAVAL SURFACE WARFARE CTR ATTN P DUTTA
ATTN DAHLGREN DIV CODE G06 72 LYME RD
DAHLGREN VA 22448 HANOVER NH 03755
1 USA SBCCOM PM SOLDIER SPT

ATTN AMSSB PM RSS A J CONNORS
KANSAS ST
NATICK MA 01760-5057
44
NO. OF NO. OF
2 USA SBCCOM 3 DIR US ARMY NATL GROUND INTEL CTR

MATERIAL SCIENCE TEAM ATTN IANG TMT
ATTN AMSSB RSS J HERBERT 2055 BOULDERS RD
M SENNETT CHARLOTTESVILLE VA 22911-8318
KANSAS ST
NATICK MA 01760-5057 1 NAVAL SEA SYSTEMS CMD
ATTN D LIESE
2 OFC OF NAVAL RESEARCH 1333 ISAAC HULL AVE SE 1100
ATTN D SIEGEL CODE 351 WASHINGTON DC 20376-1100
J KELLY
800 N QUINCY ST 4 US ARMY SBCCOM
ARLINGTON VA 22217-5660 SOLDIER SYSTEMS CTR
BALLISTICS TEAM
1 NAVAL SURFACE WARFARE CTR ATTN J WARD W ZUKAS J SONG
TECH LIBRARY CODE 323 P CUNNIFF
17320 DAHLGREN RD KANSAS ST
DAHLGREN VA 22448 NATICK MA 01760-5019
1 NAVAL SURFACE WARFARE CTR 1 US ARMY SBCCOM

CRANE DIVISION SOLDIER SYSTEMS CTR
ATTN M JOHNSON CODE 20H4 MARINE CORPS TEAM
LOUISVILLE KY 40214-5245 ATTN J MACKIEWICZ
KANSAS ST
2 NAVAL SURFACE WARFARE CTR NATICK MA 01760-5019
ATTN U SORATHIA
C WILLIAMS CD 6551 2 US ARMY SBCCOM
9500 MACARTHUR BLVD SOLDIER SYSTEMS CTR
WEST BETHESDA MD 20817-5700 ATTN AMSSB RCP SS W NYKVIST
S BEAUDOIN
2 CDR NAVAL SURFACE WARFARE CTR KANSAS ST
CARDEROCK DIVISION NATICK MA 01760-5019
ATTN R PETERSON CODE 2020
M CRITCHFIELD CODE 1730 7 US ARMY RSCH OFC
BETHESDA MD 20084 ATTN A CROWSON H EVERITT
J PRATER G ANDERSON
1 NAVAL SURFACE WARFARE CTR D STEPP D KISEROW
CARDEROCK DIVISION J CHANG
ATTN R CRANE CODE 6553 PO BOX 12211
9500 MACARTHUR BLVD RSCH TRIANGLE PARK NC 27709-2211
WEST BETHESDA MD 20817-5700
1 AFRL MLBC
3 DIR US ARMY NATL GROUND INTEL CTR 2941 P ST RM 136
ATTN D LEITER MS 404 WRIGHT PATTERSON AFB OH 45433-7750
M HOLTUS MS 301
M WOLFE MS 307 1 AFRL/MLMP
2055 BOULDERS RD ATTN R THOMSON
CHARLOTTESVILLE VA 22911-8318 2977 HOBSON WAY BLDG 653 RM 215
2 DIR US ARMY NATL GROUND INTEL CTR
ATTN S MINGLEDORF MS 504 2 AFRL
J GASTON MS 301 ATTN F ABRAMS J BROWN
2055 BOULDERS RD BLDG 653
CHARLOTTESVILLE VA 22911-8318 2977 P ST STE 6
45
NO. OF NO. OF
1 AFRL MLS OL 1 HYDROGEOLOGIC INC

ATTN L COULTER SERDP ESTCP SPT OFC
5851 F AVE ATTN S WALSH
BLDG 849 RM AD1A 1155 HERNDON PKWY STE 900
HILL AFB UT 84056-5713 HERNDON VA 20170
4 NAVAL SURFACE WARFARE CTR 2 US ARMY RSRCH LAB NASA

ATTN J FRANCIS CODE T08 VEHICLE TECHNOLOGY CENTER
D WILSON CODE B56 ATTN AMSRL VS W ELBER MS 266
R D COOPER CODE G32 F BARTLETT JR MS 266
J FRAYSSE CODE G33 HAMPTON VA 23681-0001
DAHLGREN VA 22448
1 NASA LANGLEY RSCH CTR
2 NAVAL SURFACE WARFARE CTR ATTN G FARLEY MS 266
ATTN L DE SIMONE CODE G33 HAMPTON VA 23661-3400
R HUBBARD CODE G33
DAHLGREN VA 22448 1 FHWA
ATTN E MUNLEY
1 DIR LOS ALAMOS NATIONAL LAB 6300 GEORGETOWN PIKE
F L ADDESSIO T 3 MS 5000 MCLEAN VA 22101
PO BOX 1633
LOS ALAMOS NM 87545 1 USDOT FEDERAL RAILRD
ATTN M FATEH RDV 31
1 OSD JOINT CCD TEST FORCE WASHINGTON DC 20590
OSD JCCD
ATTN R WILLIAMS 3 CYTEC FIBERITE
3909 HALLS FERRY RD ATTN R DUNNE D KOHLI
VICKSBURG MS 29180-6199 R MAYHEW
1300 REVOLUTION ST
2 DARPA HAVRE DE GRACE MD 21078
ATTN S WAX
L CHRISTODOULOU 1 3TEX CORPORATION
3701 N FAIRFAX DR ATTN A BOGDANOVICH
ARLINGTON VA 22203-1714 109 MACKENAN DR
CARY NC 27511
2 SERDP PROGRAM OFC
PM P2 1 DIR DEFENSE INTEL AGNCY
ATTN B SMITH ATTN TA 5 K CRELLING
901 N STUART ST STE 303 WASHINGTON DC 20310
ARLINGTON VA 22203
1 COMPOSITE MATERIALS INC
3 OAK RIDGE NATL LABORATORY ATTN D SHORTT
ATTN R M DAVIS 19105 63 AVE NE
C EBERLE MS 8048 PO BOX 25
C D WARREN MS 8039 ARLINGTON WA 98223
PO BOX 2008
OAK RIDGE TN 37831-6195 1 JPS GLASS
ATTN L CARTER
4 NIST PO BOX 260
ATTN M VANLANDINGHAM MS 8621 SLATER RD
J CHIN MS 8621 SLATER SC 29683
J MARTIN MS 8621
D DUTHINH MS 8611
100 BUREAU DR
GAITHERSBURG MD 20899
46
NO. OF NO. OF
1 COMPOSITE MATERIALS INC 3 PACIFIC NORTHWEST LAB

ATTN R HOLLAND ATTN M SMITH G VAN ARSDALE
11 JEWEL CT R SHIPPELL
ORINDA CA 94563 PO BOX 999
RICHLAND WA 99352
1 SIMULA
ATTN R HUYETT 5 ALLIANT TECHSYSTEMS INC
10016 S 51ST ST ATTN C CANDLAND MN11 2830
PHOENIX AZ 85044 C AAKHUS MN11 2830
B SEE MN11 2439
2 PROTECTION MATERIALS INC R DOHRN MN11 2830
ATTN M MILLER F CRILLEY D KAMDAR MN11 2830
14000 NW 58 CT 5050 LINCOLN DR
MIAMI LAKES FL 33014 MINNEAPOLIS MN 55436-1097
3 FOSTER MILLER 3 ALLIANT TECHSYSTEMS INC

M ROYLANCE W ZUKAS ATTN N VLAHAKUS MN11 2145
195 BEAR HILL RD S HAGLUND MN11 2439
WALTHAM MA 02354-1196 M HISSONG MN11 2830
4700 NATHAN LANE N
1 ROM DEVELOPMENT CORP PLYMOUTH MN 55442-2512
ATTN R O MEARA
136 SWINEBURNE ROW 1 R FIELDS
BRICK MARKET PLACE 4680 OAKCREEK ST
NEWPORT RI 02840 APT 206
ORLANDO FL 32835
1 TEXTRON SYSTEMS
ATTN M TREASURE 1 APPLIED COMPOSITES
1449 MIDDLESEX ST ATTN W GRISCH
LOWELL MA 01851 333 NORTH SIXTH ST
ST CHARLES IL 60174
1 O GARA HESS & EISENHARDT
ATTN M GILLESPIE 1 CUSTOM ANALYTICAL ENG SYS INC
9113 LESAINT DR ATTN A ALEXANDER
FAIRFIELD OH 45014 13000 TENSOR LANE NE
FLINTSTONE MD 21530
1 MILLIKEN RSCH CORP
ATTN M MACLEOD 1 AAI CORPORATION
PO BOX 1926 ATTN DR N B MCNELLIS
SPARTANBURG SC 29303 PO BOX 126
HUNT VALLEY MD 21030-0126
1 CONNEAUGHT INDUSTRIES INC
ATTN J SANTOS 1 OFC DEPUTY UNDER SEC DEFNS
PO BOX 1425 ATTN J THOMPSON
COVENTRY RI 02816 1745 JEFFERSON DAVIS HWY
CRYSTAL SQ 4 STE 501
1 ARMTEC DEFENSE PRODUCTS ARLINGTON VA 22202
ATTN S DYER
85 901 AVE 53 3 ALLIANT TECHSYSTEMS INC
PO BOX 848 ATTN J CONDON E LYNAM
COACHELLA CA 92236 J GERHARD
WV01 16 STATE RT 956
PO BOX 210
ROCKET CENTER WV 26726-0210
47
NO. OF NO. OF
1 PROJECTILE TECHNOLOGY INC 2 UDLP

515 GILES ST ATTN G THOMAS M MACLEAN
HAVRE DE GRACE MD 21078 PO BOX 58123
SANTA CLARA CA 95052
5 AEROJET GEN CORP
ATTN D PILLASCH T COULTER 2 UDLP
C FLYNN D RUBAREZUL ATTN R BRYNSVOLD
M GREINER P JANKE MS 170
1100 WEST HOLLYVALE ST 4800 EAST RIVER RD
AZUSA CA 91702-0296 MINNEAPOLIS MN 55421-1498
1 BRIGS COMPANY 1 LOCKHEED MARTIN

ATTN J BACKOFEN SKUNK WORKS
2668 PETERBOROUGH ST ATTN D FORTNEY
HERNDON VA 22071-2443 1011 LOCKHEED WAY
PALMDALE CA 93599-2502
1 ZERNOW TECHNICAL SERVICES
ATTN L ZERNOW 1 NORTHRUP GRUMMAN CORP
425 W BONITA AVE STE 208 ELECTRONIC SENSORS & SYSTEMS DIV
SAN DIMAS CA 91773 ATTN E SCHOCH MS V 16
1745A W NURSERY RD
2 GENERAL DYNAMICS OTS LINTHICUM MD 21090
FLINCHBAUGH DIV
ATTN K LINDE T LYNCH 1 GDLS DIVISION
PO BOX 127 ATTN D BARTLE
RED LION PA 17356 PO BOX 1901
WARREN MI 48090
1 GKN WESTLAND AEROSPACE
ATTN D OLDS 1 GDLS
450 MURDOCK AVE ATTN M PASIK
MERIDEN CT 06450-8324 PO BOX 1800
STERLING HEIGHTS MI 48090-1800
5 SIKORSKY AIRCRAFT
ATTN G JACARUSO B KAY 1 GDLS
T CARSTENSAN MUSKEGON OPERATIONS
S GARBO MS S330A ATTN M SOIMAR
J ADELMANN 76 GETTY ST
6900 MAIN ST MUSKEGON MI 49442
PO BOX 9729
STRATFORD CT 06497-9729 1 GENERAL DYNAMICS
AMPHIBIOUS SYS
1 PRATT & WHITNEY SURVIVABILITY LEAD
ATTN C WATSON ATTN G WALKER
400 MAIN ST MS 114 37 991 ANNAPOLIS WAY
EAST HARTFORD CT 06108 WOODBRIDGE VA 22191
1 AEROSPACE CORP 5 INST FOR ADVANCED TECH

ATTN G HAWKINS M4 945 ATTN H FAIR I MCNAB
2350 E EL SEGUNDO BLVD P SULLIVAN S BLESS
EL SEGUNDO CA 90245 W REINECKE C PERSAD
4030-2 W BRAKER LN
2 CYTEC FIBERITE AUSTIN TX 78759-5329
ATTN M LIN W WEB
1440 N KRAEMER BLVD
ANAHEIM CA 92806
48
NO. OF NO. OF
1 ARROW TECH ASSO 1 UNIV OF MAINE

1233 SHELBURNE RD STE D8 ADV STR & COMP LAB
SOUTH BURLINGTON VT 05403-7700 ATTN R LOPEZ ANIDO
5793 AEWC BLDG
1 R EICHELBERGER ORONO ME 04469-5793
CONSULTANT
409 W CATHERINE ST 1 JOHNS HOPKINS UNIV
BEL AIR MD 21014-3613 APPLIED PHYSICS LAB
ATTN P WIENHOLD
1 SAIC 11100 JOHNS HOPKINS RD
ATTN G CHRYSSOMALLIS LAUREL MD 20723-6099
8500 NORMANDALE LAKE BLVD
SUITE 1610 1 UNIV OF DAYTON
BLOOMINGTON MN 55437-3828 ATTN J M WHITNEY
COLLEGE PARK AVE
1 UCLA MANE DEPT ENGR IV DAYTON OH 45469-0240
ATTN H T HAHN
LOS ANGELES CA 90024-1597 5 UNIV OF DELAWARE
CTR FOR COMPOSITE MTRLS
1 UMASS LOWELL ATTN J GILLESPIE M SANTARE
PLASTICS DEPT S YARLAGADDA S ADVANI
ATTN N SCHOTT D HEIDER
1 UNIVERSITY AVE 201 SPENCER LABORATORY
LOWELL MA 01854 NEWARK DE 19716
1 IIT RESEARCH CENTER 1 UNIV OF ILLINOIS AT URBANA CHAMPAIGN

ATTN D ROSE DEPT OF MATERIALS
201 MILL ST SCIENCE & ENGINEERING
ROME NY 13440-6916 ATTN J ECONOMY
1304 WEST GREEN ST 115B
1 MICHIGAN ST UNIV URBANA IL 61801
MSM DEPT
ATTN R AVERILL 1 MISSISSIPPI STATE UNIVERSITY
3515 EB DEPT OF AEROSPACE ENGNRNG
EAST LANSING MI 48824-1226 ATTN A J VIZZINI
MISSISSIPPI STATE MS 39762
1 PENN STATE UNIV
ATTN C BAKIS 1 DREXEL UNIV
212 EARTH ENGR ATTN A S D WANG
SCIENCES BLDG 32ND & CHESTNUT ST
UNIVERSITY PARK PA 16802 PHILADELPHIA PA 19104
1 PENN STATE UNIV 3 UNIV OF TEXAS AT AUSTIN

ATTN R S ENGEL CTR FOR ELECTROMECHANICS
245 HAMMOND BLDG ATTN J PRICE A WALLS
UNIVERSITY PARK PA 16801 J KITZMILLER
10100 BURNET RD
1 PURDUE UNIV AUSTIN TX 78758-4497
SCHOOL OF AERO & ASTRO
ATTN C T SUN 1 VA POLYTECHNICAL INST & STATE UNIV
W LAFAYETTE IN 47907-1282 DEPT OF ESM
ATTN M W HYER
BLACKSBURG VA 24061-0219
49
NO. OF NO. OF
1 SOUTHWEST RSCH INST 1 DIRECTOR

ENGR & MATL SCIENCES DIV US ARMY RSCH LABORATORY
ATTN J RIEGEL ATTN AMSRD ARL WM BA (CHIEF)
6220 CULEBRA RD BLDG 4600
PO DRAWER 28510
SAN ANTONIO TX 78228-0510 2 DIRECTOR
US ARMY RSCH LABORATORY
1 BATELLE NATICK OPERATIONS ATTN AMSRD ARL WM BC P PLOSTINS
ATTN B HALPIN J NEWILL
313 SPEEN ST BLDG 390
NATICK MA 01760
7 DIRECTOR
2 US MILITARY ACADEMY US ARMY RSCH LABORATORY
ATTN R D HAMPTON ATTN AMSRD ARL WM BD P CONROY
BLDG 752 MAHAN HALL B FORCH M LEADORE R LIEB
WEST POINT NY 10996 B RICE R PESCE RODRIGUEZ
A ZIELINSKI
ABERDEEN PROVING GROUND BLDG 4600
1 DIRECTOR 1 DIRECTOR
US ARMY RSCH LABORATORY US ARMY RSCH LABORATORY
ATTN AMSRD ARL CI OK TECH LIB ATTN AMSRD ARL WM BD C LEVERITT
BLDG 4600 BLDG 390
1 US AMSAA 1 DIRECTOR
ATTN AMXSY TD P DIETZ US ARMY RSCH LABORATORY
BLDG 392 ATTN AMSRD ARL WM BF S WILKERSON
BLDG 390
1 US ARMY ATC
ATTN CSTE DTC AT AC I W C FRAZER 2 DIRECTOR
BLDG 400 US ARMY RSCH LABORATORY
ATTN AMSRD ARL WM M S MCKNIGHT
1 DIRECTOR J MCCAULEY
US ARMY RSCH LABORATORY BLDG 4600
ATTN AMSRD ARL O AP EG
M ADAMSON 3 DIRECTOR
BLDG 245 US ARMY RSCH LABORATORY
ATTN AMSRD ARL WM MA (CHIEF)
1 DIRECTOR L GHIORSE E WETZEL
US ARMY RSCH LABORATORY BLDG 4600
ATTN AMSRD ARL SL BM D BELY
BLDG 328 19 DIRECTOR
3 DIRECTOR ATTN AMSRD ARL WM MC (CHIEF)
US ARMY RSCH LABORATORY R BOSSOLI E CHIN
ATTN AMSRD ARL WM J SMITH S CORNELISON D GRANVILLE
J MCCAULEY M ZOLTOSKI B HART F PIERCE
BLDG 4600 E RIGAS W SPURGEON
BLDG 4600
1 DIRECTOR
ATTN AMSRD ARL WM B (CHIEF)
BLDG 4600
50
NO. OF NO. OF
22 DIRECTOR 1 DIRECTOR
US ARMY RSCH LABORATORY US ARMY RSCH LABORATORY
ATTN AMSRD ARL WM MB J BENDER ATTN AMSRD ARL WM TC R COATES
T BOGETTI J BROWN L BURTON BLDG 309
R CARTER K CHO W DEROSSET
G DEWING R DOWDING 4 DIRECTOR
W DRYSDALE R EMERSON US ARMY RSCH LABORATORY
D GRAY D HOPKINS R KASTE ATTN AMSRD ARL WM TD D DANDEKAR
L KECSKES M MINNICINO M RAFTENBERG S SCHOENFELD
B POWERS D SNOHA J SOUTH T WEERASOORIYA
M STAKER J SWAB J TZENG BLDG 4600
BLDG 4600
1 DIRECTOR
12 DIRECTOR US ARMY RSCH LABORATORY
US ARMY RSCH LABORATORY ATTN AMSRD ARL WM TE (CHIEF)
ATTN AMSRD ARL WM MD P DEHMER BLDG 1116A
B CHEESEMAN R DOOLEY
G GAZONAS S GHIORSE
M KLUSEWITZ J LASALVIA
J MONTGOMERY W ROY
J SANDS D SPAGNUOLO
S WALSH S WOLF
BLDG 4600
2 DIRECTOR
ATTN AMSRD ARL WM RP C SHOEMAKER
J BORNSTEIN
BLDG 1121
1 DIRECTOR
ATTN AMSRD ARL WM T B BURNS
BLDG 309
3 DIRECTOR
ATTN AMSRD ARL WM TA W GILLICH
C HOPPEL M ZOLTOSKI
BLDG 4600
5 DIRECTOR
ATTN AMSRD ARL WM TA T HAVEL
J RUNYEON M BURKINS
E HORWATH B GOOCH
BLDG 393
1 DIRECTOR
ATTN AMSRD ARL WM TB P BAKER
BLDG 390
51
NO. OF NO. OF
1 DEF RES ESTABLISHMENT

FOREIGN ADDRESSES VALCARTIER
A DUPUIS
1 LTD 2459 BOULEVARD PIE XI NORTH
R MARTIN VALCARTIER QUEBEC
MERL CANADA
TAMWORTH RD PO BOX 8800 COURCELETTE
HERTFORD SG13 7DG GOA IRO QUEBEC
UK CANADA
1 CIVIL AVIATION 1 INSTITUT FRANCO ALLEMAND

ADMINSTRATION DE RECHERCHES DE SAINT LOUIS
T GOTTESMAN DE M GIRAUD
PO BOX 8 RUE DU GENERAL
BEN GURION INTERNL AIRPORT CASSAGNOU
LOD 70150 BOITE POSTALE 34
ISRAEL F 68301 SAINT LOUIS CEDEX
FRANCE
1 AEROSPATIALE
S ANDRE 1 LAB TECHNOLOGIES COMPOSTS
A BTE CC RTE MD132 & POLYMERS
316 ROUTE DE BAYONNE J MANSON
TOULOUSE 31060 DMX LTC
FRANCE CH 105 LAUSANNE EPFL-IMX
SWITZERLAND
1 DRA FORT HALSTEAD
P N JONES 1 TNO DEFENSE RESEARCH
SEVEN OAKS KENT TN 147BP R IJSSELSTEIN
UK ACCOUNT DIRECTOR
R&D ARMEE
1 SWISS FEDERAL ARMAMENTS PO BOX 6006
WKS 2600 JA DELFT
W LANZ THE NETHERLANDS
ALLMENDSTRASSE 86
3602 THUN 1 FOA NATL DEFENSE RESEARCH
SWITZERLAND ESTAB
DIR DEPT OF WEAPONS &
1 DYNAMEC RESEARCH AB PROTECTION
AKE PERSSON B JANZON
BOX 201 R HOLMLIN
SE 151 23 SODERTALJE S 172 90 STOCKHOLM
SWEDEN SWEDEN
1 ISRAEL INST OF TECHNOLOGY 1 DEFENSE TECH & PROC AGENCY

S BODNER GROUND
FACULTY OF MECHANICAL ENGR I CREWTHER
HAIFA 3200 GENERAL HERZOG HAUS
ISRAEL 3602 THUN
SWITZERLAND
1 DSTO
WEAPONS SYSTEMS DIVISION
N BURMAN RLLWS
SALISBURY
SOUTH AUSTRALIA 5108
AUSTRALIA
52
NO. OF
COPIES ORGANIZATION
1 MINISTRY OF DEFENCE
RAFAEL
ARMAMENT DEVELOPMENT
AUTH
M MAYSELESS
PO BOX 2250
HAIFA 31021
ISRAEL
B HIRSCH
TACHKEMONY ST 6
NETAMUA 42611
ISRAEL
1 DEUTSCHE AEROSPACE AG
DYNAMICS SYSTEMS
M HELD
PO BOX 1340
D 86523 SCHROBENHAUSEN
GERMANY
53

ADA441366

Uploaded by

Copyright:

Available Formats

ADA441366

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ADA441366

Uploaded by

Copyright:

Available Formats

Air Gun Launch Simulation Modeling and Finite

Element Model Sensitivity Analysis

by Mostafiz R. Chowdhury and Ala Tabiei

ARL-TR-3703 January 2006

Approved for public release; distribution is unlimited.

Citation of manufacturer’s or trade names does not constitute an official endorsement or

DESTRUCTION NOTICE⎯Destroy this report when it is no longer needed. Do not return it to

ARL-TR-3703 January 2006

Air Gun Launch Simulation Modeling and Finite

Approved for public release; distribution is unlimited.

January 2006 Final May to December 2003

6. AUTHOR(S) 5d. PROJECT NUMBER

5f. WORK UNIT NUMBER

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

11. SPONSOR/MONITOR’S REPORT

12. DISTRIBUTION/AVAILABILITY STATEMENT

Approved for public release; distribution is unlimited.

1. Part I: Air Gun Modeling 1

2. Part II: Air Gun Finite Element Simulation 20

Appendix A. Program I/O Files and FORTRAN Source Code 35

Figure 1. Schematic of the air gun. ................................................................................................ 1

a. schematic of air gun test setup.

b. The OBR with test item.

1.2 Model Formulation

1.2.2 Rigid Body Dynamics System

The system above is equivalent to the system below

OBR Mitigator MEM

Figure 2. Dynamic system representation.

{u}n +1 = {u}n + Δt{u& }n + Δt {&u&}n + O(Δt 3 )

Now, from the equation of motion, equation 1, it follows:

or {u}n +1 = 2{u}n − {u}n −1 + [M ]−1 ({f }n − [K ]{u}n ) (4)

For the right-hand side system representing the MEM we have

1.3 System Parameters Definition

1.3.1 OBR and MEM

Equation 5 is derived from K − ω 2 M = 0, for m1 = m2 = m, in which ω is the angular

Figure 6. Mitigator wedge deformation zones.

in which A0 is the cross-sectional area of the cylindrical mitigator.

1.3.3 Strain Rate Effects

1.4 Results and Discussion

1.4.1 Validation of Present Model

ω 2 (m1 + m 2 ) 46200 2 ⋅ 3.75

Figure 7. Top acceleration of the OBR unfiltered.

Figure 9. Top acceleration of the OBR filtered at 2500 Hz.

Figure 12. Top velocity of the OBR.

Figure 14. Mitigator contact forces.

Figure 16. Length of mitigator plastic region.

Figure 17. m1:m2 mass ratio sensitivity of model.

Figure 18. Model sensitivity to 10% variation of the value of k1.

Figure 19. Model sensitivity to large variation of the value of k1.

Figure 20. Model sensitivity to variation of the solution time step.

2.1 Review of Previous Work

2.2 Finite Element Model Sensitivity Study

2.2.1 Double Versus Single Precision Code

Figure 23. FFT of the acceleration data.

Figure 24. Acceleration data for different integration time steps.

Figure 26. Acceleration data sampled at different frequencies.

Figure 27. FFT of the acceleration data sampled at different frequencies.

2.2.5 The OBR Filled With Solid Elements to Represent Beads

Figure 35. Acceleration data for simulation with filled OBR.

Figure 36. FFT of the acceleration data for simulation