UCRL-JC-127524

PREPRINT

Measurement of Projectile Trajectory in Dielectric Target

with Micropower-Impulse Radar

D. W. Baum
R. M Kddo
E. T. Rosenbury
S. C. Simonson

This paper was prepared for submittal to the

17th International Symposium on Ballistics
Midrand, South Africa
March 23-27,1998

November 20,1997
 DISCLAIMER

This document was prepared as an account of work sponsored by an agency of

the United States Government. Neither the United States Government nor the
University of California nor any of their employees, makes any warranty, express
or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by trade
name, trademark, manufacturer, or otherwise, does not necessarily constitute or
imply its endorsement, recommendation, or favoring by the United States
Government or the University of California. The views and opinions of authors
expressed herein do not necessarily state or reflect those of the United States
Government or the University of California, and shall not be used for advertising
or product endorsement purposes.
MEASUREMENT OF PROJECTILE TRAJECTORY IN DIELECTRIC TARGET
WITH MICROPOWER-IMPULSE RADAR

Dr. Dennis W. Baum, Mr. Robert M. Kuklo, Mr. E. Tom

Rosenbury, and Dr. S. Christian Simonson

Lawrence Livermore National Laboratory, P.O. Box 808, L-170, Liverrnore,CA 94550, USA

The micropower-irnpulse radar has been adapted for non-intrusive tracking of

projectiles in dielectric targets. The main application of this technique is in-
tended to be the validation of continuum mechanics simulation codes and mate-
rial models used in the study of the interaction between high-velocity penetra-
tors and concrete targets. Two experiments have been conducted in which a
gun-launched 90-mm-diameter projectile was fired at velocities of 160 and 23(I
m/s into a cubical box ffled with dry sand and tracked with the micropower-
impulse radar. The system was adjusted so that a 2-m range in sand was di-
vided into 511 timing intervals, which were swept every 0.1 rns. k the pro-
jectile took approximately 40 ms to come to resl this meant that the~ were 40C
measurements of its position. The CALE continuum mechanics simulation was
used to model the projectile motion in the targeg and close agreement was found
with the memured traiectmv

INTRODUCTION
For the purpose of studying the interaction between munitions and materials used to protect
hardened targets, such as concrete or rock, it is useful to employ a non-intrusive diagnostic
which will provide trajectory information. The recent development of the micropower-impulse
radar, or MIR [1], has made available a new diagnostic technique which is not only non-
intrusive but which also is capable of measuring a large number of data points along a trajec-
tory. In order to evaluate the utility of the MIR in this application, two experiments have been
conducted with gun-launched projectiles and the results have been analyzed. Since this diag-
nostic technique tracks the interface between media of different dielectric properties, for examp-
le, a conducting penetrator and a dielectric target, it can in principle be applied to shaped
charge jets and explosively formed projectiles as well as to gun-launched or free-falling projec-
tiles.
The main use of the data from MIR measurements in such experiments is expected to be the
validation of continuum mechanics simulation codes and material models. As an example of
such an application, the CALE code [2] was used, together with a sand model reported at a
previous International Ballistics Symposium [3], to model the projectile motion in the present
experiments.

EXPERIMENTS

Projectile and Target

The experimental setup is shown in Fig. 1. A 90-mm projectile is launched by a smooth-bore
powder gun, firing horizontally. The projectile impacts a target consisting of a 2-m cubical box
ffled with dry sand. The micropower-impulse radar antennas are positioned just behind the
targe~ and the radar electronic equipment is located to one side.
 . Antennas

b
160-230 mki

~=
MIR
90 mm
Gun projectile

Fig. 1,—Top view of the experimental setup (schematic), showing 90-mm gun, projectile,
sand target, and micropower-impulse radar antennas positioned at rear of target. -

The projectiles consisted of hollow cylinders of aluminum, 400 mm long and 90 mm in di-
ameter, with a beveled nose extending back 19 mm and a central cavity 60 mm in diameter and
350 mm deep. These were adopted for convenience, as they had been fabricated for another
experiment, in which the central cavity was intended to be filled with steel ballast.
The sides of the target box were constructed of plywood sheets to give a total thickness of 50
mm. External stiffeners were constructed of 50 mm x 150 mm to provide extra protection for
the radar antennas. The back also projected 600 mm downward into the firing table gravel bed
to provide additional reaction mass during projectile deceleration. An entrance hole of diameter
150 mm was cut in the front of the box and covered by a 6-mm-thick polycarbonate plate to
retain the sand. The box was fdled with No. 3 coarse silica sand which had been kiln dried
and bagged. It was kept dry during the experiments by use of a blower and heater, and a tent
was erected over the experiment for protection from the weather.
For purposes of calibrating the MIR, a projectile was buried in the sand while the box was be-
ing filled, at a distance of 1.5 m from the antennas. In addition, a PVC pipe was installed run-
ning through the box so that a projectile could be moved back and forth manually and imaged
by the MIR.
The projectiles were launched with a 90-mm smoothbore powder gun, firing horizontally. The
distance from the muzzle to the target was 1.052 m. The muzzle velocity was obtained from a
pair of shorting pins at the muzzle. A third shorting switch of parallel-conductor braided wire
was installed on the face of the target. This initiated the radar data acquisition and gave a free-
flight velocity measurement.
The projectile velocity for these experiments was chosen so that there would be little danger of
the projectile penetrating the rear wall and damaging the radar antennas. For the first experi-
ment, the velocity was 160 m/s and the penetration was 0.917 m. For the second, the velocity
was 230 m/s and the penetration was 1.298 m.

Micropower-Impulse Radar
The micropower-impulse radar that was used in these experiments operates by sending a
stream of transmitted pulses at a pulse repetition frequency of 2 MHz. Each puke contains
Fourier components covering a frequency range of 1.5-5 GHz. The system rise time is about
 60 ps, and the pLdscwidth is abou[ 1 ns. A gated mixer opens to admit the return wave form N
increasing intervals of around 10 ps. After a number of range cells have been recorded, in this
case 511, a certain intend of range will have been covered, and rhe process is then repeated.
It maybe noted that there are no range ambiguities witfr this technique.
For the experiments reported here, the system was adjusted so that the 2-m range in sand was
divided into 511 timing intervals, which were swept every 0,1 ms. In both experiments, the
projectile took approximately 40 ms to come to rest, which meant that there were 400 meas-
urements of its position from which its trajectory could be derived.
For each O.l-ins range sweep, the return signal has a form that represents the interaction of the
transmitted wave form with the shape and material of the projectile. This return signal has a
complicated form, but it stays relatively constant from one sweep to the next, except for tie
displacement which is due to the projectile motion,
For the purpose of analysis, the range sweeps, each consisting of 511 data points, may be
stacked side by side and viewed as an image. This process is illustrated in Fig. 2. The various
features of the return wave form appear as bands in the image.

[ [

/ /

TiMe (milliseconds to seconds)

Fig. 2.—Combination of hundreds of range sweeps into an image. Bands in the image indi-
cate the projectile moving into the range gate and coming to rest in the target.
 The data for ExpenmenL 2 are shown in such an image in Fig. 3. (For Experiment 1, the en-
tmnce pin failed to report a time, although it did trigger the radar data acquisition system.
Therefore the entrance velocity was less certain, and this experiment was not further analyzed.)
The background, consisting of tfre average signal before tie projectile entered the sand, has
been subtracted. This has left artifacts at ranges of approximately 230-320 mm and 1OOO-105O
mm into the target. If we pick the leading band, we may follow the leading edge of the projec-
tile as it enters the sand, slows down, and eventually stops. (Note that a faint earlier band ap-
pears, which is due to a low-energy precursor to the main radar puke.)

125

83

42

0
0 10 20 30 40 50 60
Time (ins)

Fig. 3.—Range sweep data displayed in the form of an image for the 230 M projectile travel-
ing into the sand target. The horizontal bands at ranges of 23-32 cm and 100-105 cm represent
background subtraction anomalies.

The location of the leading band was measured on each sweep, using a simple level detection
algorithm. The results are shown in Fig. 4 for Experiment 2. In order to obtain vafues for
velocity and acceleration, a polynomird (of degree 9) was fitted to the axiaf positions as a func-
tion of time. The projectile came to rest after approximately 40 ms. Thus at 0.1 ms per sweep,
there are approximately 400 positional data points. The r.m.s. deviation of the data points from
the polynomial is +6.5 mm. The velocity and acceleration obtained from the polynomial tit are
also shown in Fig. 4. The p+xakvelocity is 203.8 m/s, while the time difference measurement
 140

120
10)

0 10 20 30 40
250
l!/ 200.
~ 150
“g 100
; 50
>

0 10 20 30 40
8000

4000
\

2000t+F
o 10 20 30 40
Time (ins)
Fig. 4.—(a) Positional points for the range sweep data represented in Fig. 3. The position data
have been fit with a 9-term polynomial, which is also shown. (b) Velocity derived from the
polynomial fit. (c) Derived acceleration.

gave 230 m/s. However, as the range gates did not extend into the air, the incoming velocity
was not recorded by the MIR.
The rapid changes upon initial impact were not well resolved in the method of recording the
dat~ because the range cells were distributed uniformly along the entire thickness of tie target.
Th; peak deceleration was 6,500 G.

SIMULATION
To model the entrance and initial deceleration of the projectile into the sand targe~ the CALE
(Arbitrary Lagrangian-Edefim code in C) continuum mechanics simulation code [2] was
 used. The sand was modeled using the parameters reported by Simonson et al. [3]. The initial
layout and mesh are shown in Fig. 5a.

gsee TIme . 0. OOWOe+OO

Cycle = 0 gsee Time . 1.~023e~3
mo(o. o) . l.~~oe.o~
Cgcle . 3202
ntc(28.9) = 3.3343f&-oj

II II
JJ

I
1
El

-1
El

-2

-3

gsee El
Time = 2.0w29e~3 El
Cycle = 6633 gsee
Dtc(28.1) = 3.33354e-01 Time = 4.oo~ie+03
Cgcle = 12260 Dtc(51.6) . 3.73727=41
3
El 3
El

2
2

o
El
&!- 1

0
El

-1
El -1
El

-2
-2

-3
-3

El -.
cl

Fig. 5.—Projectile motion in sand target as modeled with CALE. (a) t = O ms; initial layout .
and mesh (in cm). (b) t = 1 ms: crater develops around front of projectile and sand column .
grows in central cavity. (c)t= 2 ms: sand column collides with bottom of cavity. (d) t = 4 ms:
sand fills front of projectile, effectively forming a blunt nose.
>
The various stages in the penetration are shown in the cross-sectional snapsho~ in Fig. 5b -5d.
hpact with the polycarbonate cover plate causes rapid initial deceleration at t = Oms. A cmter
then starts to grow around the hollow cylindrical nose, and a sand column forms in the central
cavity (t= 1 ins). The sand column collides with the bottom of the cavity at t =
2 ms. Eventu-
ally, compressed sand fills the front of the cavity and acts as a blunt nose for steady-sta~ de-
celeration (t= 4 ins).
 A comparison of the axial position of the leading edge of the projectile as simulated with CALE
and as measured with the MIR is shown in Fig. 6a. The figure covers the first 5 ms of the 40-
ms event. A segment of the polynomial fit for the full 40 ms is also shown. The r.m.s. devia-
tion of the CALE simulation for the polynomial fit is *3.4 mm. This is within the inherent
r.m.s. scatter in the data, which is *6.5 mm with respect to the polynomial fit.

600 1,11‘i,I1III1IattrI,I1,,,I,1,I11,I1 #1,,

I II111I * I 1 0 11

500 a

z 400
g

!3
c 300
a
z
.—
n 200
~
x
< 100

0 1
lll,,s,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 1 , I t #t , I I t I
o
, 1

1 2 3 4 5
Time (ins)
240 , aI , 1a, I 1I 1I 1, 1 , # , I I , , , I I 1 , , 1
1 , , I 1 I , , , 1 , t , , 1 1 1 I , I
1

%
200 i \o b

160
— Fit to Data
o CALE Calculation
120

80 J

40

0 1119, ,1,,,,,,,,,,,,,,,,,,,, t, I I 1 1 , , I I , I # I I , 1 * I#,

o I
I I

1 2 3 4 5
Time (ins)
Fig. 6.—(a) Comparison of displacement data along with polynomial fit and CALE simulation.
(b) Comparison of derived velocity and CALE simulation.
The velocities from the CALE simulation and from the polynomial fit to the data are shown in
Fig. 6b. Until t = 3 ms, the CALE velocities show various features as the successive impact
events occur. After t = 3 ms, CALE shows steady-state deceleration, like the measurements.
The CALE velocities are somewhat noisy, for three reasons, First, there are physical oscilla-
tions in the projectile owing to shock reverberations and impact pressure variations. Second,
there is numerical sampling noise, as the edited velocity for a test particle located near the nose
happens to refer to the nearest node, not to an interpolation between nodes. Third, the mesh
resolution may have been too coarse to record all the impact phenomena in a smooth fashion.
Despite the noisiness in the velocity record, the general behavior clearly reflects the initial
stages in the penetration event (irnpac~ cratering, central column collision, steady state dtxel-
eration). The deviation of the CALE velocities from the polynomial fit velocities (which are
also imprecise in the first few milliseconds), is B.6 m/s for t = O-3 ms and +06 m/s for t = 3-
5 ms (with a bias in the latter interval of +1.5 M/s}.

CONCLUSIONS
The following general conclusions may be drawn
The micropower-irnpulse radar gave a clear record of the motion of a metallic penetrator in a
dielectric target. The signal-to-noise ratio was adequate to allow O.l-ins time resolution with
511 range cells distributed across a 2-m target depth. The repetition rate was rapid enough to
yield numerous data points, allowing reasonably precise measurements of position, velocity
and acceleration from this single non-intrusive diagnostic.
The impact event was successfully modeled with CALE. A the purpose of the MIR exper-
imentwas to record the entire penetration event within the target material, localized details of the
impact were somewhat imprecise. In future experiments, the number of range cells could be
disposed more advantageously. For example, range cells should record the incoming projcxtile
in air. Also, depending on the application, more spatial resolution and less time resolution
might be useful, or the reverse. By the use of mukipIe elements, 3-D trajectory measurement
may also be possible.
The main disadvantage in using the MIR is its sensitivity to free water molecules and to em-
bedded conducting elements, such as rebw or instrumentation wiring. In the current applica-
tion, it worked well with dry target material. However, it is also expected to work under more
ordinary environmental conditions, as it has been employed to survey conc~te highways and
bridges for internal flaws and to detect metal pipes and similar targets buried at depths of up to
a meter underground.

ACKNOWLEDGMENTS
Work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore
National Laboratory under contract #W-7405 -Eng-48. This work was supported in part by the
Joint DoD/DOE Munitions Technology Development Program.

REFERENCES
1. Science and Technology Review, Lawrence Livermore National Laboratory, UCRL-52000-
96-1, January/February 1996, p. 17.
2. Tipton, R. E. 1997, “CALE User’s Manual”, private communication.
3. Simonson, S. C., Wirier, K. A., Reaugh, J. E., Breithaupt, R. D., Baum, D. W. 1995,
Proceedings 15th Int. Symp. on Ballistics, Jerusalem, Israel, Vol. 1, p. 51.