Vous êtes sur la page 1sur 447



21-22 Nov 2008

Armament Research & Development Establishment

Pashan, Pune - 411 021
Paper Author Title

S1 Aerodynamics & Structures and Inbore, External & Terminal Ballistics

S1.1 Peyada NK, Sen A, Dutta GG, Singhal A, Effect of Mach Number on Supersonic
Rajan KM, Joshi DK, Ghosh AK Wrap Around Fin Aerodynamics

S1.2 Sonawane PD, Dr. Singh Amarjit Slender Aerodynamic Bodies for High
Speed Armaments

S1.3 Misra A, Ghosh AK, Ghosh K Stability & Control of Aircraft Bombs at
High Angle of Attack using Cascade Fins

S1.4 Chand KK Understanding Trajectory Modeling via

Error & Uncertainty Analysis

S1.5 Giri SN, Chand KK Role of Applied Physics & Ballistics in

Armaments : A Review

S1.6 Raj Venkatnarayan, Varadrajan S Weapon Locating Radar

S1.7 Gite LK, Padmanabham M, Joshi DK Parameter Estimation Techniques :

Application to Prediction Algorithm for
Weapon Locating Radar

S1.8 Padmanabham M, Joshi DK, Gite LK Ballistic Needs for Weapon Locating

S2 Warheads and Non Lethal Weapons

S2.1 Rajeev Prakash Non Lethal Weapons : Naval Perspective

S2.2 Anguswamy V, Sahu AC Dual Type Fragmentation Warhead

for Hard/Soft Targets

S2.3 Brig. HSN Sastry (Retd) Non Lethal Weapons : Concepts,

Technologies & developments

S2.4 Dhote KD, KPS Murthy Velocities of Pre-made Fragments in


S2.5 Harikrishnan S, KPS Muthy, Design & Evaluation of a Light Weight

Surendra Kumar Blast Barrier for Tandem Warhead of a
Third Generation Anti Tank Missile

S2.6 Mohammadali Partanian, Analysis of Jet Penetration of Hollow

Pralhad RN Charge from Concept of Hydrodynamic
Theory & Fracture Mechanics

S2.7 Padhan PK, Biswas PP, Dayanad, An Overview on End Game for Surface to
Bhattacherjee RN Air Missile (SAM)

S2.8 Kamlesh Kumar, Satwinder Singh, Thermobaric Warhead for Under-Ground

KPS Murthy Targets

NCAAT 2008 I
Paper Author Title

S3 Weapons, Launch Technology, SAM, Fuze & Armament Electronics

S3.1 Sen A, Subrahmanyam S, Wahi P, Five-Degree of Freedom Dynamic Wind

Ghosh AK Tunnel Test Rig for Estimation of Dynamic
Derivatives of Scale Down Models

S3.2 Rao MV, Mahalingam VS Pinaka Weapon System Integration With

ACCCS : Challenges & Solutions

S3.3 Salutagikar CS, Mukane SB, Pravker RC Enhancing Accuracy of Free Flight Rocket
Pawale SO, Rojatkat PT, Rajan KM By Minimizing Dynamic Unbalance :
A Case Study

S3.4 Mahmoud Zarrini & Pralhad RN Estimation of Shock Velocity & Pressure
for the Bullets Impacting on Human Body

S3.5 Pandey RP, Verma Gaurav, Dixit Aishwarya, Lethality Assessment of Small Arms
Gade SV, Datar AM Ammunition

S3.6 Pandey RP, Verma Gaurav, Dixit Aishwarya, Effect of Small Arms Bullet Construction
Gade SV, Datar AM on Penetration of Hard Targets

S3.7 Khilare TB Enhancement of Lethality of Rocket

Assisted Bomb Over Standard
HE Mortar Bomb

S3.8 Saini Vaibhav, Kumar Vinay, Saini DS, Integrated G-hardened Telemetry System
Rana SC, Kapil Deo for Cannon Launched Munition
Flight Testing

S3.9 Saha PK, Sahu BK, Dhande P, Application of Wireless Sensor Network in
Ms Mahajan CP the Selective & Intelligent Actuation of
Scattered Munition

S3.10 Dhande P, PK Saha, BK Sahu Remote Time Data Setting for Munitions
Vaishnavi Sharma, Ms Mahajan CP

S3.11 Ms Savita, Ms Patel Prachi, Dhande Pramod, Study of Frequency Selection for
Saha PK, Ms Mahajan CP Communication with Underground

S3.12 Deodhar KD, Kharat DK Hybrid Composite Technology For Gun


S3.13 Sachin Sood, Narasimha Rao G, Development Testing, System Integration

Chandramouli G, Panyam RR & Evaluation of Indigenous Surface to
Air Weapon System

S4 Test & Evaluation and Guidance & Control

S4.1 Kolhe JP, Shaik JB, Bokil GA, Talole SE Robust Roll Autopilot Design

S4.2 Bommu Mallikarjuna, Venkata Rama Missile Autopilot Design Using Dynamic
Subbaiah B, B Srinivas Kalyan, Inversion and Fuzzy Logic Control
C Arjun Rao, Talole SE

Paper Author Title

S4.3 Singh VK, Sharma IS, Dayal Arun and Inferencing from Limited Trials in
Rajiv Gupta Weapons Evaluation

S4.4 Nayak Navin, Bisval TK High Speed Imaging & Radar Based
Measurements for Performance Evaluation
of Armaments

S4.5 Bisval TK, Nayak Navin Dynamic Test & Evaluation Methodologies
for Measurement of Critical Test
Parameters of Armaments : Case Studies

S4.6 Saha Subhas, Kolhatkar DR Next Generation of Environmental Testing

S4.7 Bhange Niteen P, Ghosh AK Analytical Method for Predicting Sabot

Discard Phenomena

S4.8 Lakra S, PK Das Gupta, V Anguswamy Study of Gun Dynamics Using Data Mining

S4.9 Dr. Fleming R, Dr Gorge McConnell Advanced Miniature Precision Effectors

S4.10 Jha PK, Kumar Krishna Guidance for Perpendicular Hit for
Shaped Charge Warhead

S4.11 Shaikh AD, Salkar YB, Parate DA, Testing & Revalidation of Imported
Dr. Virendra Kumar, Dr. Kharat DK Cartridge PDO for Su-30 Fighter Aircraft

S4.12 Katare RV, Ravidas Rakesh Testing & Evaluation of Canopy Severance
System for Tejas Trainer Aircraft

S4.13 Mathew T, Chandel P, Rajeev Kumar, Impact Studies on Spaced Armour

Athwal M, Sewak KB

S5 Modeling & Simulation and Safety, Quality & Reliability

S5.1 Sharma IS, Singh VK, Dayal Arun, Modeling Weapon Effectiveness in
Rajiv Gupta Integrated Combat Scenario

S5.2 Singh SK, Kumar Manoj, Sharma VK, Mathematical Modeling for Earth
Srinivasa Rao HV Penetrating Bomb

S5.3 Kamble AD, Guru Prasad, Gandhi Vandana Design of MIL - STD - 1553B
Communication Interface Protocol &
Computer Models for a RTOS Based
Distributed Mission Critical
Electro - Hydraulic Control System

S5.4 KKK Sanyasi Raju, D Chakraborti Swarm Intelligence Based Algorithms for
Design Optimization of Stiffened
Cylindrical Shells for Underwater Weapons

S5.5 Deodhar RS, Anupam Anand Evaluation of Probability Grid Using

Simulation of Fragmentation
Pattern on Ground

S5.6 Philppe Ecolivet Development of Fuzes, Safe/Arm System

& Seekers


Paper Author Title

S5.8 Sunil Kumar K. Surya Kiran CH, Jaya Modeling & Simulation on Explosive
Ramiah M, Vara Prasad Rao S, Sekharan VG Dynamics Phenomena during Stage
Separation Mechanism

S5.9 Dr. Rajagopal Chitra Ammunition Storage and Safety Regime

in India. Comparison with Other Countries.

S5.11 Mishra GK Disposal of Life Expired and Regiime

in India : Comparison with other Countries

S5.12 Wadhawa KC Fire Safety Aspects of

Armaments & Ammunition

S5.13 Aguru NK Enhancement of Explosive safety in

Indian Army & Reduce Land Requirement
by Converting Ammunition of un Hazard
Division 1.1 To 1.2

S5.14 Bhupinder Sewak, Sharma AC, Numerical Simulation on Explosively

Saroha DR, Singh M Formed Projectile

S6 Explosives, Propellants, Pyrotechnics and Propulsion Technology

S6.1 Prof. Barpande Girish, Dr. Singh Amarjit CFD Simulation and Analysis of Flow
Through a Supersonic Combustor

S6.2 Sanghavi RR, Shelar SD, Tope BG, Studies on Effect of Picrite on
Chakraborthy TK, Singh Amarjit, Insensitive Nitramine Gun Propellants
A Subhananda Rao

S6.3 A Sudheer Babu Prediction of Impact Sensitivity of

CaHbNcOdExplosives Using Artificial
Neural Networks

S6.4 Saji J, Vishwakarma AK, Roy S, Apparao A Development of Aluminised Explosive

Composition for Thermobaric Warhead

S6.5 Daniel R, Danali SM, Kohadkar MJ, Development of Laser Initiated

Raha KC, A Subhananda Rao Pyrodevices for Futuristic Defence

S6.6 Danali SM, Swarge NG, Gaikwad VD, Experimental Studies on Accuracy
Kohadkar MJ, Raha KC Improvement of Pyrotechnic Delays

S6.7 Palaiah RS, Soman RR, Thakur BR, High Intensity Illuminating Flares
Kohadkar MJ, Raha KC, Based on Energetic Binder
A Subhananda Rao

S6.8 Debnath S, Rohit, VP Ambekar, Studies on Thermal and Structural

PM Deshmukh, Danali SM, Analysis of Pyrotechnic Thermite System
V Ramaswamy, Raha KC Designed for Fast Melting of Metals

S6.9 Swami Umesh, Pralhad RN Estimation of Flow Parameters of

Explosion at Shock Front for Different

Paper Author Title

S7 Advanced Materials and Manufacturing Techniques

S7.1 Ramakrishna B, Mishra B, Madhu VT, Perforated Armour : Improving

Balakrishna Bhat The Ballistic Efficiency of HH Armour Steel

S7.2 I Nithiya Priya, HH Kumar, DK Kharat Rheological study of Aqueous PZT slurry
for Tape Casting

S7.3 Premkumar S, B Praveen Kumar, Characteristics of 3-3 porous PZT

HH Kumar, DK Kharat Composites Evaluated by Finite Element

S7.4 Kolhe SM, HH Kumar, Koparkar PV, Piezoelectric Polymer 0-3 Composite
Lonkar CM, DK Kharat for Power Harvesting Application

S7.5 Chakradeo NM, Dighe NA Use of CAD/CAM Technology for

Manufacture of Critical Armament

S7.6 Chakradeo NM, Dighe NA Manufacturing of Barrel for "Under

Barrel Grenade Launcher" (UBGL)

S7.7 Dharia CJ Advances in Spring Materials for

Armament Application

S7.8 B Murlidhar, SM Nirgude, Effect of Fracture Toughness & Charpy

M Ashok Kumar Impact Value on Inbore Structural
Integrity of KE Projectiles

S7.9 Jadhav SC, Parate BA, Moon LC, Design of Cartridge Signal Flare 26mm
Dr. Virendra Kumar by using Aluminum Case

S7.10 Das T, Das PS, Chakraborty A, Ti-based Gate Dielectrics on

Majhi B, Hota MK, Mallick S, Verma S, Si-passivated n-GaAs for MOSFET
Maiti CK Applications

NCAAT 2008 V
Aerodynamics & Structures
Inbore, External & Terminal Ballistics


Peyada NK, Sen A, Dutta GG, Singhal A, Ghosh AK
IIT Kanpur – 208 016
Rajan KM, Anand Raj
Armament Research & Development Establishment, Pune 411 021
Wrap around fins (WAF) have been used on tube launched missiles and dispenser launched projectiles. Modern
advances in stealth technology have made the use of missiles equipped with WAFs desirable because they can be stowed to
reduce the radar cross section of the aircraft. Recent studies have identified several roll – and pitch – moment instabilities.
The rolling moment of the WASF is positive at subsonic velocities (defined here to means that the missile rolls towards the
concave side of fin). A roll reversal occurring at M = 1.0 indicates that the magnitude of the rolling moment decreases with
increasing Mach number and that a second rolling moment reversal may occur at high supersonic speed. Tilman et al and
Bowersox examined experimentally and numerically the flow structure in the vicinity of a single fin mounted onto blended
cylindrical body at Mach 3.0 and 5.0 Those studies indicated that the flow field was highly asymmetric about WAF, with a
stronger bow – shock structure on the concave side of the fin. This resulted in creating a very high-pressure region between
the fin and its center of curvature, which resulted in relatively high surface pressures near the mid span of the fin. In contrast,
on the convex side of the fin, surface pressure was relatively independent of location along the span.
This resulted in asymmetric pressure loading on the fin, which caused generation of out-of-plane moment at angle of
attack, it is suspected that this side moment is symptomatic of WAF configurations. When testing such configurations, the test
engineer should ensure that this side moment is obtained and stability boundaries are computed. This side moment can have a
dramatic effect on trajectory computation based on the conventional aerodynamic coefficients and derivatives.
Experiments have been conducted to study the effect of fin shapes on generation induced rolling moment and out-
of-plane moment coefficients. Four sets of WAF configuration were tested. The results were then compared with the flat fin
It was observed that addition of back sweep helps in reducing rolling moment as well as out of plane moment
substantially at supersonic Mach numbers.

Wrap around fins (WAF) have been used on tube launched missiles and dispenser-launched projectiles. Modern
Advances in stealth technology have made the use of missiles equipped with WAF's desirable because they can be stowed to
reduce the radar cross section of he aircraft1. Recent studies have identified several roll – and – pitch – moment instabilities2.
The roll reversal at transonic conditions is the most recognized instability. The rolling moment of the WAF is positive at
subsonic velocities (defined here to means that the missile rolls towards the concave side of fin)1. A roll reversal occurs at M
= 1.0 have indicated that the magnitude of the rolling moment decreases with increasing Mach number and that a second
rolling moment reversal may occur at high supersonic speed. Tilman et al3 and Tilman and Bowersox4 examined
experimentally and numerically the flow structure in the vicinity of a single fin mounted onto a blended cylindrical body at
Mach 3.0 and 5.0. Those studies indicated that the flow field was highly asymmetric about WAF, with a stronger bow – shock
structure on the concave side of the fin. This resulted in creating a very high-pressure region between the fin and its center of
curvature, which resulted in relatively high surface pressures near the mid span of the fin.

In contrast, on the convex side of the fin, surface pressure is relatively independent of location along the span.
This resulted in asymmetric pressure loading on the fin. This asymmetric pressure loading caused generation of out-of-plane
moment at angle of attack, It is suspected that this side moment is this side moment is symptomatic of WAF configurations,
when testing such configurations, the test engineer should ensure that this side moment is obtained and stability boundaries
are computed. Designers of such configurations should also consider the possibility of this side moment because it can have a
dramatic effect on trajectory computation based on the conventional aerodynamic coefficients and derivatives5.




Experiments have been conducted to study the effect of fin shapes on generation induced rolling moment and out-
of-plane moment coefficients. Four sets of WAF configurations were tested. The results were then compared with the flat fin
configuration. Configurations of the fin tested are shown in Fig. 1.

Configuration - I Configuration -II

Fig. 1 : Fin shapes tested for lateral characteristics

It is observed that addition of trailing edge sweep helps in reducing induced rolling moment as well as out of plane
moment substantially at supersonic Mach numbers. Figure 1 and 2 present comparison of induced rolling moment and out of
plane moment coefficient for these configurations. It is clearly seen that the fin with a larger back sweep resulted in lesser
Cl , Cy and Cna at supersonic Mach number. These values of Cn play important role in establishing dynamic stability. A
preliminary study shows that the design of fin must ensure lower value of C na (£ 6)

Fig. 2 : The comparative variation of Cl , Cy and Cn a

as function of Mach Number




Three sets of rockets having 20 degrees trailing edge sweep were fired. The trajectory coordinates (x, y, z, V) of
these rockets in flight were acquired by Doppler radar. The trajectory variables were used to estimate drag coefficients using
method proposed in Ref. 5. The trajectory variables used for estimation of CD is presented in Fig 3.

Fig. 3 : Trajectory variables used for estimation of CD



Estimated values of CD are graphically presented in Fig. 4

Fig. 4 : Estimated value of CD as function of Mach Number

It could be appreciated that the estimated CD is almost 8 – 12% higher tan the wind tunnel estimation. This information may
help in tuning the trajectory model for field Application.



The introduction of trailing edge sweep helps in reducing induce roll and out-of-plane moment coefficient. Such an
approach may help in avoiding any possibilities of dynamic instability associated with WAF rocket at supersonic speeds. The
estimation procedure presented in Ref. 6 may be used to estimate overall drag coefficient of the rocket using measured flight

The numerous suggestions provided by Shri D.K. Joshi, Joint Director, ARDE, Pune is highly acknowledged.

1.Mc Intyre T.C., Bowersox, R.D., and Larry, P.G., “Effects of Mach number on Supersonic Wrap Ariound Fin
Aerodynamics,” Journal of Space craft and Rocket, Vol. 35, No. 6, Nov – Dec 1998.
2.Abate, G.L. and Winchenbach, G.L., “Aerodynamic of Missiles with Slotted in Configurations,” AIAA paper 91-
0676 January 1991.
3.Tilmann, C.P., Huffman, R.E., Jr Buter, T.A. and Bowersox, R.D.W., “Experimental Investigation of the flow
structure Near a Single Wrap Around Fin,” Journal of Spacecraft and Rocket, vol. 34, No. 6, 1997.
4.Tilman, C.P. and Bowersox, R., “Characterisation of the flow field near a wrap around fin at Mach number 4.9,”
AIAA paper 98-0684, January 1998.
5.Winchenbach, G.L. and Randy, S.B., Whyte, T.A. and Hathway, W.H., “Subsonic and Transonic Aerodynamics of
a Wrap Around Fin configuration,” Journal of Guidance. Vol. 9, No. 6, Nov – Dec 1986.
6.G.G. Dutta, Ankur Singhal and A.K. Ghosh, "Estimation of Drag Coefficient using Real Radar Tracked Data of an
Artillery Shell," Proceedings AIAA, Paper No. 2006-6149, AIAA Atmospheric Flight Mechanics Conference and Exhibit-
06, Keystone, Colorado21-24 Aug, 2006.




Pushkaraj D. Sonawane
Vishwakarma Institute of Information Technology, Pune
Dr. Amarjit Singh
Defence Institute of Advanced Technology, Pune

The requirement of defense against ballistic missiles has led to a renewed interest in the aerodynamics of slender,
blunt-nosed, conical bodies for high speed armaments. Different slender aerodynamic bodies which can be used for the high
speed armaments are as follows:
Slender Bodies:
Jorgensen has shown that there are distinct aerodynamic advantages in using elliptic cones rather than circular
cones. Reggiori has demonstrated that a cone with anhedral wings can provide good high lift characteristics at high angles of
attack in hypersonic speeds.
Singh, Kumar, and Tiwari conducted a parametric study to determine the effects of nose bluntness on the entire flow-
field over slender bodies (blunted cones and ogives) under different hypersonic free stream conditions. Specific results
obtained for spherically-blunted cones and ogives demonstrate that there are significant differences in flow-field and surface
quantities between sharp and blunted bodies. Jorgensen investigated the aerodynamic advantages of elliptic cones over
circular cones experimentally to determine the force and moment characteristics for elliptic cones at high Mach numbers.
Power Law Bodies
Power-law bodies can be an alternative to the blunted cone configurations because a power-law body has a greater
internal volume than a blunted cone of same fineness ratio and secondly, various theories predict that a power-law body
represents the minimum drag at hypersonic speeds.
Spencer et al found from experiments on a series of 2/3-power law-wave-drag bodies at high Mach numbers that
increase in the lift, drag, and pitching moment coefficients and lift-drag ratios at positive angles of attack and at all test Mach
Jorgensen et al performed an experimental investigation to determine the effects of side strakes on the aerodynamic
characteristics of a body of revolution, at High Mach Numbers and incidences from 0° to 58°. The data demonstrated that the
aerodynamic characteristics for a body of revolution with side strakes can be significantly affected by changes in nose
fineness ratio, nose bluntness, Reynolds number, Mach number and angle of attack.
Amarjit Singh performed the force measurements for a one-half-power-law body and found that the lift, drag and
pitching moment coefficients increased smoothly with incidence. The strakes were found to improve the L/D ratio of the
configuration without any significant change in the pitching moment characteristics.
Bodies with elliptic cross-section exhibit higher lift to drag ratios than those of circular cross-section of the same
fineness ratio and volume. Power-law bodies can be an alternative to the blunted cone configurations because a power-law
body has a greater internal volume than a blunted cone of same fineness ratio and represents the minimum drag case at
hypersonic speeds. The additions of strakes to the slender bodies are found to improve the L/D ratio of the slender bodies
configuration without any significant change in the pitching moment characteristics.


M8 free-stream Mach number

Re8 free-stream Reynolds number
L lift
D drag
Cl lift coefficient based on body planform area
Cd drag coefficient based on body planform area



Cm pitching moment coefficient based on body length and planform area

Y the local radius of the body
Rb body base radius
d diameter
X axial distance from the nose
l body length
n the power exponent.
á incidence angle

The requirement of defence against ballistic missiles has led to a renewed interest in the aerodynamics of slender,
blunt-nosed, conical bodies flying at hypersonic speeds. In hypersonic flows, nose bluntness for leading edges of the aerofoil
is extremely useful to reduce stagnation heating. Many aerospace vehicles use blunt noses and leading edges not just because
it will be very difficult to keep them sharp in hypersonic flow but also because of the added advantage of reduced
aerodynamic heating and increased internal space. Theoretical and experimental investigations(1,2) have shown that an
elliptical cone may have significantly higher lift to drag ratios than a circular cone of the same cross sectional area per unit

Jorgensen(2) has shown that there are distinct aerodynamic advantages in using elliptic cones rather than circular
cones. Reggiori(3) has demonstrated that a cone with anhedral wings can provide good lift characteristics at high angles of
attack in hypersonic speeds. Singh, Kumar, and Tiwari(4) conducted a parametric study to determine the effects of nose
bluntness on the entire flow-field over slender bodies (blunted cones and ogives) under different hypersonic free stream
conditions. Specific results obtained for spherically-blunted cones and ogives demonstrate that there are significant
differences in flow-field and surface quantities between sharp and blunted bodies. Depending upon the flow conditions and
geometry, the differences are found to persist as far as about 300 nose radii downstream. For different angle of attacks and
nose bluntness the post shock flow field is studied(5) in detail from the contour plots of Mach number, density, pressure, and
temperature. The effect of nose bluntness for slender cones persists as far as 200 nose radii downstream.
Jorgensen(6) investigated the aerodynamic advantages of elliptic cones over circular cones experimentally to
determine the force and moment characteristics for elliptic cones at high Mach numbers. Results showed that bodies of
elliptic cross-section exhibited higher lift to drag ratio than those of circular cross-section.

Power-law bodies can be an alternative to the blunted cone configurations because a power-law body has a greater
internal volume than a blunted cone of same fineness ratio and secondly, various theories predict (7-8) that a power-law body
represents the minimum drag at hypersonic speeds. Spencer et al(9) found from experiments on a series of 2/3-power-law
wave-drag bodies at high Mach numbers that increase in the lift, drag, and pitching moment coefficients and lift to drag ratios
at positive angles of attack and at all test Mach numbers. The successive increases in major-to-minor axis ratio, with the major
axis horizontal, resulted in gains in lift-curve slope for bodies of the same fineness ratio.
The surface co-ordinates of a power-law body (Figure 1) are given by Y/Rb = (X/l)n, The n = 1 value represents a
sharp cone and a reduction in the value of n from 1 to 0
continuously increases the bluntness of the body. In general nose bluntness is useful to reduce the stagnation heating as well
as to increase the volume for a given fineness ratio (l/d) body.

Fig. 1 : Surface contour of power-law bodies



Spencer and Fox(10) compared the aerodynamic characteristics of power-law bodies, of circular and elliptic cross-
section with that of a body of minimum-wave drag shape determined under the constraint of prescribed body length and
volume. They found that the increase in the ellipticity ratio for a given power-law or the theoretical minimum-wave-drag
body results in an almost constant incremental increase in maximum L/D ratio, independent of body longitudinal contour.
Ashby(11) obtained experimental data for power-law bodies found that at the highest Reynolds number, the power-
law body for minimum drag is blunter (exponent n lower) than predicted by inviscid theory (n approximately 0.6 instead of n
= 0.667). Westby and Regan(12) conducted an experimental investigation of the longitudinal aerodynamic characteristics of
power-law bodies of revolution in a gun-tunnel also found the same reults.


Jorgensen et al(6) performed an experimental investigation to determine the effects of side strakes on the aerodynamic
characteristics of a body of revolution, at High Mach Numbers and incidences from 0° to 58°. The data demonstrated that the
aerodynamic characteristics for a body of revolution with side strakes can be significantly affected by changes in nose
fineness ratio, nose bluntness, Reynolds number, Mach number and angle of attack. Removing the strakes from the
cylindrical after section greatly decreased the lift and moved the center of pressure forward.
Levinsky et al(13) showed theoretically and experimentally that relatively large increases in lift with even the smallest
stakes sizes, i.e. 10% of the body radius, are possible.
Amarjit Singh(15) performed the force measurements for a one-half-power-law body is found to be in good agreement
with the other existing results. The lift, drag and pitching moment coefficients are found to increase smoothly with incidence.
The strakes are found to improve the lift to drag ratio of the configuration without any significant change in the pitching
moment characteristics.
Kontis et al(17) performed experiments and collected data for two pairs of elliptic cones with and without strakes at Mach
number 8.2 and unit Reynolds number 9.3 x 106 /m. The strakes produced a significant increase in the lift and drag
coefficients, in the incidence range of 0° to 20°. The addition of strakes moved the centre of pressure slight forward for the
right elliptic model and backwards for the power-law body at all incidences tested.
Both experimental and computational studies(18) showed that the addition of strakes caused a reduction of pressure on the
leeward side and an increase of pressure on the windward side. The paper concludes with suggesting that, in the design stage,
CFD can be used to identify important areas and important flow conditions before the model is cut, which can save
unnecessary waste in the experimental tests. After data collection from the experiments, detailed numerical simulation can be
conducted for the experimental conditions.

Experiments(15) were conducted to obtain lift, drag and pitching-moment coefficients for a one-half-power-law
body at the College of Aeronautics gun tunnel by one of the authors. The axisymmetric nozzle provided a useful jet of 15 cm
diameter at M8 of 8.2 and Re8 equal to 9.35 x 106/m. A three-component strain gauge balance was used to measure axial, and
normal forces and pitching moment. The balance was calibrated using a locally made test rig. The calibration of the balance
was repeated many times and showed insignificant difference.

Fig. 2 : Power-law elliptic body model with and without strakes



Figure 2 shows the basic dimensions of the wooden models. These models were187.5 mm long. The strakes are
approximately 1.67 mm square at the base. Dead weights were used at the base of the models to get the C. G. position at about
26 mm from the base. The mass of the models was 56 grams and 33 grams with and without strakes respectively. The models
had an elliptical cross section and a one half-power-law contour along the length. The models were rigidly screwed on to the
balance sting and fixed at zero roll angle.


A commercial solver has been used to calculate the flow field around the half-power-law elliptic body shown in
figure 2. Flow field results for incidence angle á = 0 , 30, 60, 90, 120,150 are obtained assuming a perfect gas and a laminar

boundary layer over the full length of the body. The free-stream conditions used are M8 = 8.2 and Re8 = 9.35x10 / m.
Approximately 800000 polyhedral cells are used for discretization of the flow field. The Navier-Stokes solution converged to
the experimental shock position after about 16,000 iterations on a Quadra core Linux workstation. Figure 3 shows the Mach
number contours which are reasonably good in agreement with the experimental results(15) . Model without strakes has been
run and model with strakes will be run soon to optimize the shape size and location of strakes to obtain maximum lift to drag
ratio without affecting the pitching moment characteristics.

Fig. 3(a) : Mach number contours of a half-power-law elliptic body



Fig. 3(b) : Mach number contours of a half-power-law elliptic body




Lift Coefficient
Figure 4 (a) shows the variation of lift coefficient with incidence. The computational results compare reasonably
well with the experimental values. The experimental values are slightly above the computational values. The gap widens
above = 5, the angle around which the leeward flow is expected to separate forming counter rotating leeward vortices.

Fig. 4(a,b,c) : Lift, Drag, Moment Coefficient vs Incidence angle



Drag Coefficient
Figure 4 (b) compares the drag values. The computational values are lower than the computational values. As the
incidence angle is increasing the variations in the results is more.

Pitching Moment Coefficient

The variation in pitching moment coefficient, figure 4 (c), follows the experimental results at lower incidences. At
higher incidences the difference in experimental results and the computational results is significant due to effect of base flow
and effect of balance string has not been modeled in the computational results.

Lift to Drag Ratio

The computational L/D ratio shows improvement over the experimental above 7 incidence due to increased suction
over the leeward side, figure 4 (d).

Fig. 4(d,e) : L/D ratio and Xcp/L vs Incidence angle




1. Bodies with elliptic cross-section exhibit higher lift to drag ratios than those of circular cross-section of the same
fineness ratio and volume.
2. Power-law bodies can be an alternative to the blunted cone configurations because a power-law body has a greater
internal volume than a blunted cone of same fineness ratio and represents the minimum drag case at hypersonic speeds.
3. The additions of strakes are found to improve the L/D ratio of the configuration without any significant change in the
pitching moment characteristics.
4. Mach number, lift and the pitching moment characteristics of for the computational and experimental results of a half-
power-law body gave a reasonably good agreement .However, the agreement is not so good for the drag and pitching
moment as effect of base flow and effect of balance string has not been modeled in the computational results.


1. Fraenkel, L. E. , “Supersonic flow past slender bodies of elliptic cross-section”, ARC A&M 2954, 1955
2. Jorgensen, L. , “Elliptic cone alone and with wings at supersonic speeds”, NACA TN-4045, 1957
3. Reggiori, A., “Lift and drag of wing-cone configuration in hypersonic flow”, AIAA Journal, vol. 9, No. 4 , April 1971,
pp 744-745.
4. Tiwari, S. N., Singh, D. J., Sehgal, A. K., “Combined effect of nose bluntness and angle of attack on slender bodies in
viscous hypersonic flow”, AIAA Paper 92-0755, January 1992.
5. Ashby, G. C., and Cary, A. M., “A parametric study of the aerodynamic characteristics of nose-cylinder-flare bodies at
Mach number of 6.0”, NASA TN D-2854, June 1965.
6. Jorgensen, L. H. , Nelson, E. R., “Experimental aerodynamics characteristics for a cylindrical body of revolution with
side strakes and various noses at angles of attack from 0° to 58° and Mach numbers from 0.6 to 2.0”, NASA TM X-
7. Eggers, A. J., Resnikoff, M. M., and Dennis, D. H., “Bodies of revolution having minimum drag at high supersonic
speeds”, NACA Report 1306, 1957.
8. Cole, J. D., “Newtonian flow theory for slender bodies”, Journal of Aeronautical sciences, vol. 24, no. 6, pp 448-455,
June 1957.
9. Spencer, B., Phillips, W. P. , “Supersonic aerodynamic characteristics of a series of bodies having variations in fineness
ratio and cross-section ellipticity, NASA TN D-2389, 1964.
10. Spencer, B., and Fox, C. H., “Hypersonic aerodynamic performance of minimum wave-drag bodies”, NASA TR R-250,
November 1966.
11. Ashby, G. C., Jr., “Longitudinal aerodynamic performance of a series of power-law and minimum wave drag bodies at
Mach 6 and several Reynolds numbers”, NASA-TM-X-2713, August 1974.
12. Westby, M. F., and Regan, J. D., “The aerodynamic characteristics of power-law bodies in continuum and transitional
hypersonic flow”, RAE Technical Memorandum Aero 2164, August 1989.
13. Levinsky, E. S., Wei, M. H. Y., “Non-linear lift and pressure distribution of slender conical bodies with strakes at low
speeds”, NASA CR-1202, 1968.
14. Singh, Amarjit, “Force measurement on a one-half-power-law elliptic cone with and without strakes”, Cranfield
University, College of Aeronautics Report No. 9505, August, 1995.
15. Singh, Amarjit, “Experimental study of slender vehicles at hypersonic speeds”, PhD thesis, Cranfield University, 1996.
16. Singh, Amarjit, Chauhan, Y. S., “A study of the effect of strakes on the aerodynamic characteristics of a power-law body
of elliptical cross-section at hypersonic speeds”, Proceedings of 51st AGM and Seminar on Advances in Aerospace
Technologies, Jan 21-22, 2000.
17. Kontis, K., Stollery, J. L., Edwards, J. A., “Hypersonic effectiveness of slender lifting elliptic cones with and without
strakes”, AIAA 97-521, pp1-11,1997.
18. Kontis, K., Qin, N., Stollery, J. L., “Computational and experimental investigation of hypersonic performance of a
lifting elliptic cone with and without strakes”, AIAA 97-2252, pp1-11, 1997.
19. Kontis, K., “Flow control effectiveness of jets, strakes, and flares at hypersonic speeds” Proc. IMechE Vol. 222 Part G:
J. Aerospace Engineering, pp 585-603, 2008.
20. Sonawane, P. D., Singh, Amarjit, Bhatkar, V. W. and Oak, S. M., “ CFD Analysis of Hypersonic Flow around a Power
Law Body” Engineering Design 2007 Conference, Indian Institute of Science, Bangalore, Aug 09-11, 2007.
21. Sonawane, P. D., Mittal, Nitika, Barpande, G. S., Jaware, V. B., Singh, Amarjit, “Slender Bodies at Hypersonic
Speeds”, IDST-MSAT 2007, Missile Complex, DRDL, Hyderabad, Oct 26-27, 2007.
22. Bertin J. J., “Hypersonic aerothermodynamics”. AIAA education series, 1994.
23. Cox, R. N. and Crabtree, L. F., “Elements of hypersonic aerodynamics”, Academic Press, New York, 1965.
24. Rasmussen, M., “Hypersonic Flow”, John Wiley & Sons, New York, 1994.




Misra A, Ghosh AK, Ghosh K
IIT, Kanpur 208016

Most guided bombs use planar fins that are mounted parallel to the body axes. These fins rotate back and forth to
generate forces in the horizontal and / or vertical planes that cause the bomb to yaw right and left, pitch up and down, or
roll as the bomb maneuvers towards the target.
Lattice or grid fins are a relatively new Aero-mechanical technology for tail controlled ordnances. One of the
advantages of grid fins is their capability to produce effective aerodynamic control forces at high angle of attack. Unlike
conventional planar fins, grid fins do not experience classical “stall” at high angle of attack. These characteristics lead to more
effective stability and control at intermediate and large angles of attack. In addition, the hinge moments produced by grid fins
are very small due to small chord. This leads to requirement of very small actuators for deflecting these fins. This advantage
becomes more and more significant as the bomb size increases. The major disadvantage of grid fins is high drag leading to
lower aerodynamic efficiency compared to conventional planar fins.
In order to address the issue of low aerodynamic efficiency of grid fins, a new category of fins called cascade fins is
proposed. Cascade fins consist of planar members placed parallel to each other at optimized gap to chord ratio. Unlike grid
fins, no cross members are present, only a base plate and an optional end plate (EP) is present. Planar members, in cascade fins,
are of symmetric airfoil cross section. Due to cascade effect the lift increases with increase in number of planar members for a
given angle of attack. These could also be flat plates. These fins produce lesser drag as compared to grid fins (better CL/CD).
Absence of cross members in a cascade fin does not alter its stall prevention characteristics. Cascade fins preserve both the
basic advantages of grid fins, namely delayed stall, and very low hinge moments. Flat plate cascade fins provide more stability
and better restoring moments, whereas Aerofoil cascade fins provide superior aerodynamic efficiency.
Some of the parameters affecting the aerodynamics of cascade fins are gap to chord ratio, number of planar members,
presence of end plate and the cross section shape of the planar members.

Most guided missiles use planar fins that are mounted parallel to the body axes. These fins rotate back and forth to
generate forces in the horizontal and / or vertical planes that cause the vehicle to yaw right and left, pitch up and down, or roll
as the missile maneuvers towards the target. These moving control fins are located aft of the weapon's center of gravity (C.G.)
in a tail control layout. At the CG in a wing control format, or forward of the CG in a canard control configuration. Missile
concepts with forward control fins or canard, have been used for many years. However, previous studies have shown that
concepts with canards can suffer from adverse induced rolling moments1-5.
Lattice or grid fins are a relatively new aero-mechanical technology for tail controlled missiles. The design of these
grid fins allows an effective aerodynamic control device stowed along with the body of a missile without increasing the
overall dimensions. Therefore, the fin configuration promises good stow ability for potential tube-launched and internal
carriage dispenser-launched applications. The internal grid structure, which forms webbing for a tail fin, provides a high
strength to weight ratio compared to planar fins that can be quite large. This in-turn, allows for the fins to have advantage of a
low hinge moment due to smaller chord and hinge control effectiveness. Therefore, small and light actuators can be used for
missiles with Grid / Lattice fin controls. The use of grid fins or “lattice controls”, for the tail control surfaces instead of
conventional planar fins was proposed by the U.S. Army Aviation and Missile Command (AMCOM) personnel as a possible
remedy for the roll control problem5.
A grid fin is an unconventional lifting and control surfaces of small chord6. One of the advantages of grid fins is their
capability to produce effective aerodynamic control forces at high angle of attack over wide Mach number ranges. Unlike
conventional planar fins, grid fins do not experience classical “stall” at higher angle of attack. These characteristics lead to
more effective stability and control characteristics at intermediate and large angles of atatck7. Grid fins are normally mounted
transverse to the longitudinal axis of the missile so that the flow passes through the grid numbers. The angle of attack of the fin
is achieved by simple rotation of the fin about its horizontal axis in much the same way a conventional fin control surface is
deflected. Although there is evidence8,9 that the lattice control concept has been around for some time, there use as missile



stability and control surfaces is a fairly recent innovation. Russian federation armed forces already use grid fins on the AA-12
medium range air-to-air missile. This missile is understood to be equivalent to the US AMRAAM. Grid fins are also used on
various ballistic missiles of the former Soviet Union and have even been used as emergency drag brakes on the SOYUZ TM-
22 spacecraft. United states uses grid fins on Massive Ordnance Air Blast (MOAB) bomb10,11. Thus, grid fine have found a
number of applications within the sphere of missile aerodynamics
The Research Development and Engineering Center of U.S. Army Aviation and Missile Command, Huntsville,
Alabama has investigated the aerodynamics of the grid fins since 1985. A total nine wind tunnel tests have been conducted in
order to gain greater insight into the aerodynamic characteristics of grid fins for Mach numbers ranging from 0.3 to 3.5. The
grid fin is a unique device that can be used as either an aerodynamic stabilizer or a control surface. It's unique design and
aerodynamic characteristics separate it from conventional planar fins.
Planar fins can generally be described with information about the root chord, tip chord, span, and thickness. Grid
fins require additional geometric parameters namely height, cell spacing element thickness12. The team continued their
research and tested the leading edge profiles (flat, convex and concave) with respect to the free stream directions and found
that effects of the curvature on the grid fin aerodynamics were negligible12. Next, the effect of frame cross-sectional shape as
well as web-thickness on the drag and other forces were studied12. Their tests results yielded significant changes in drag
characteristics at all Mach numbers tested, but the normal force (Lift) and hinge moment characteristics showed small
In early 1989, wind tunnel tests by Rockwell International Missile System Division, Georgia demonstrated several
interesting trends12. The results showed that horizontal slat configuration tend to provide higher normal-force characteristics
than open geometries13. Complex grid-like inner fin arrangements provided increase in normal force over open geometries
and comparatively larger axial force. The cross fin configurations produced roughly the same normal force as the single slat
fin. It was realized that the same fins because, of their low hinge actuators requirements, might be efficient as lifting surfaces
when flown in the normal transverse direction but could also be used as drag brakes if turned to a horizontal position13.
Interest in grid fins is primarily based on its potential use on highly maneuverable munitions due to their advantages
over conventional planar controls at high angle of attack and high Mach number. The fin design yields favorable lift
characteristics at high angle of attack and near zero hinge moments, which allow the use of small and light weight actuators1,2.
Aerodynamics of grid fins has been investigated since 1985 by the U.S. Army Missile Research and Development Center
(MRDEC). These investigations indicated that one of the advantages of grid fins is their ability to maintain lift at high angle of
attack since they do not have the same stall characteristics as those of conventional planar fins. The main disadvantage was
indicated to be higher drag compared to that of planar fins, although some techniques for minimizing drag by altering the grid
fin frame cross sectional shape were demonstrated. Any modification of grid fin configuration must necessarily provide
adequate lift at high angle of attack. This is possible if the stall is delayed and / or is of mildness nature compared to planar fin.
The drag reduction of a grid fin configuration could be achieved, to the some extent, by altering the grid fin cross sectional
shape. For performance enhancement such a configuration needs to generate higher values of aerodynamic efficiency.
Undamentally the delay in stall could be linked to delayed separation which could be further attributed to the cascade effect14.
A typical grid fin with criss cross members can be approximated as consisting of orthogonal slats4 (Fig. 1). One of the
important parameter effecting stall could be the gap to chord ratio of the lifting members of the grid fin14. A typical cruciform
configuration of grid fins as shown in Fig. 1 consists of two pairs.

Fig. 1: Schematic of the model used for design evaluatin and testing fo fgrid fins

One in elevator position and other in rudder position. In elevator position the lift producing members are horizontal
and cross members are vertical. In rudder position the lifting members are vertical and cross members are horizontal. The
cross members a part from providing structural rigidity to some extent, contribute to drag. In addition, elevator fins' cross
members contribute to directional stability and rudder fins' cross members contribute to pitch stability. The cross members do
not provide control in any case. In order to maximize we propose a new category of grid fins called cascade fins. The present
work is a study of such cascade fins which do not have cross members as is the case with grid fins. A cascade fin has fins
parallel to each other placed at optimized gap to chord ratio (g/c), forming a cascade of flow.
Most of the grid fins configurations available in open literature addresses issues related to grids fins with high aspect
ratio. The low aspect ratio fins, although induce large drag, are better suited for high angle of attack maneuvers. The selection
of test configurations was also influenced by the immediate requirement of a lattice fin tail unit for an existing aircraft bomb.



Accordingly the aspect ratio chosen for the planar member of cascade fin was 2.0. Further, to investigate the effect of cross
sectional shape of planar members, two cross sectional shapes, namely Flat Plate and Symmetrical aerofoil (NACA 0012)
were chosen. In order to eliminate any body fin interference, these cascade fins were tested in fin alone configuration with out
In order to study the effect of number of planar members of a cascade fin on its aerodynamics the number of such
planar members selected was 1, 3, 4 & 5. Here 1 planar member configuration refers to a single conventional planar fin.
Another important parameters to be studied was gap to chord ratio of a cascade fin. Five different gap to chord ratios
were tested (g/c = 0.5, 0.6, 0.7, 0.8, 0.9). A full scale model was tested at Re = 308,000. A sting (to mount the model in wind
tunnel) available in the tunnel was unsuitable to simulate high angle of attack in vertical plane. This was overcome by
mounting the cascade model on the floor of the wind tunnel as shown in Fig. 2.

Fig. 2: Cascade fin without and with end plate (Rear View at = 0)
0 0
The angle of attack range (0 to 50 ) was simulated in the horizontal plane by rotating the base turn table. A shroud
was placed over the balance front and rear end adapter to isolate the balance from direct wind loads.
A six component balance, mounted vertically on the floor, was used to measure the aerodynamic forces and
moments (CN, Cy, Cm, Cl, Cn). Both dry and wind runs were made in order to asses the value of the offset to be removed from the
measured data. The data obtained through these tests were processed to get numerical values of force and moment
coefficients. The reference area of 0.02 m2 and chord length of 0.1 m were used for nondimensiolizing the force and moment
As stated earlier, in order to address the issue of low aerodynamic efficiency of grid fins, a new category of fins
called cascade fins is proposed14. Cascade fins consist of planar members placed parallel to each other at optimized gap to
chord ratio. Unlike grid fins, no cross members are present, only a base plate and an optional end plate (EP) are present. Planar
members, in cascade fins, are of symmetric airfoil cross section. These could also be flat plates. These fins produce lesser
drag as compared to grid fins (better CL/CD). Absence of cross members in a cascade fin does not alter its stall prevention
characteristics. Cascade fins preserve both the basic advantages of grid fins namely delayed stall and very low hinge
Some of the parameters affecting the aerodynamics of cascade fins are gap to chord ratio, number of planar
members, presence of end plate and the cross section shape of the planar members. Exhaustive wind tunnel tests were
conducted to quantify the effect of these parameters on cascade aerodynamics. A brief discussion on the wind tunnel tests is
described next. The effect of gap to chord ratio and end plate effect has been documented in Ref 14. This paper briefly
presents the effect of number of planar member and cross sectional shape of planar member or cascade aerodynamics.


Figure 3 shows the effect of number of planer members in a cascade fin. As expected the lift increases with increase
in number of planar members for a given angle of attack. However, this increase in lift is not linear. This is due to cascade
Effect. It can be seen referring Fig. 3 that there is significant increase in the stall angle of the cascade fin. The stall angle for
single flat plate is found to be around 12 degrees. Whereas for cascade fins it lie near 22 degrees. The stall angle could easily
be increased by reducing the gap to chord ratio14-. The cascade fins could be used for high angle of attack operation without
facing conventional problems of stall.



Fig. 3 CL vs. Alpha for different n (n=no. of planar members); g/c = 0.5 (WT Data)

Next issue was to increase CL/CD of the cascade fins. One approach followed was to replace flat plate cascade by
aerofoil cascade. Figure 4 compares the aerodynamic characteristics of Flat Plate Cascade (FPC) and Aerofoil Cascade
(AFC). The lift curves show that upto 20 degrees angle of attack, FPC gives marginally higher lift. From 20 degrees to 30
degrees angle of attack, AFC gives marginally more lift. Around 32 degree angle of attack, AFC's lift drops abruptly. The drag
curves show that AFC is superior than FPC throughout the angle of attack range of 0 to 53 degrees. The aerodynamic
efficiency (CL/CD) of AFC is significantly higher than FPC from 4 to 32 degrees angle of attack. For bombs whose operable
range of angle of attack lie within 4 to 32 degrees, an AFC tail unit is a betterchoice from aerodynamic efficiency point of
The pitching moment characteristics of the two cascades show that the pitch stiffness provided by AFC is nearly
constant upto 22 degrees angle of attack. Beyond 22 degrees there is a sudden drop in pitch stiffness. FPC, on the other hand,
provides higher stiffness than AFC upto 20 degrees. Beyond 20 degrees angle of attack, upto 40 degrees, AFC gives
marginal stability. The reduction in pitch stiffness of cascade fins at higher angle of attack can advantageously be utilized to
design the suitable controller for highly maneuverable missiles / aircraft bombs.



Fig 4: Comparison of Flat Plate and Aerofoil Cascades' Aerodynamics

1. Allen, J.M., Blair, A.B., Jr., “Comparison of Analytical and Experimental Supersonic Aerodynamics Characteristics
of a Forward Control Missile,” Journal of Spacecraft and Rockets, 1982, 19(2), pp 155-159.
2. Blair, A.B., Jr., Dillon, J.L., Jr., Watson, C.B., “Experimental Study of Tail-Span Effects on a Canard-Controlled
Missile,” Journal of Spacecraft and Rockets, 1993, 30(5), pp 635-640.
3. Blair, A.B., Jr., “Supersonic Aerodynamics Characteristics of Maneuvering Canard Controlled Missile with Fixed and
Free-Rolling Tail Fins,” SAE paper 90-1993, Society of Automotive Engineers: Warrendale, PA, October 1990.
4. Burt, J.R., Jr., “The Effectiveness of Canards for Roll Control,” RD-77-8, U.S. Army Missile Command, Red Stone
Arsenel, AL, 1976.
5. James De Spirito, Milton E., Vaughn, Jr., and W. David Washington, “Numerical Investigation of Aerodynamics of
Canard-Controlled Missile using Planar and Grid Fins,” Part II: Subsonic and Transonic Flow, ARL TR 3162, March
6. Washington, W.D., Miller, M.S., “Grid Fins A New Concept for Missile Stability and Control ,” AIAA paper 93-0035,
American Institute of Aeronautics and Astronautics, Reston, VA, January 1993.
7. Washington, W.D., Miller, M.S., “Experimental Investigation of Grid Fin Aerodynamics: A Synopsis of Nine Wind
Tunnel and Three Flight Tests,” Symposium on “Missile Aerodynamics”, Sorrento, Italy, 11-14 May, 1998.
8. Belotserkovskiy, S.M., etal, “Wings with Internal Frame Work,” Machine Translation, FTD-ID (RS)-1289-86,
Foreign Technology Division, February 1987.



9. Simpson, G.M. and Sadler, A.J., “Lattice Controls: A Comparison with Conventional Planar Fins,” Symposium on
Missile Aerodynamics, RTO-MP-5, Italy, 11-14 May 1998.
10. Drab, J., and Patterson, B.C., “Massive Ordinance Air Blast Weapon Development,” AFRLS, Munition Directorate,
Elgin, AFBFL, 2004.
11. Fulghum, David A., “It is a Big One (MOAB Actually Fits in a B-2,” Aviation Week and Space Technology, March 16,
12. Washington, William D. and Miller, Mark S., “Curvature and Leading Edge Sweep Back Effects on Grid Fin
Aerodynamics Characteristics”, AIAA paper 93-3480, Applied Aerodynamics Conference, Monterey, CA, August
13. Miller, Mark S. and Washington, William D., “An Experimental Investigation of Grid Fin Configurations,” AIAA
paper 94-1914-CP, WL Technical Library.
14. Misra A., Ghosh A.K. and Ghosh K., “Cascade Fins - An Alternate Tail Stabilization Unit”, AIAA conference,
Hawaai, Aug 18-20,2008.




KK Chand
Proof & Experimental Establishment, Chandipur, Balasore, Orissa- 756 025

Trajectory modeling and simulation has been addressed in many contexts by researches and scientists in many
disciplines in its broad applications and usefulness. It is a computational tool to calculate the flight path of projectiles and its
elements. There is a wide span of trajectory modeling and computations differing with respect to their performance and
fidelity characteristics, which are used throughout the life cycle of a weapon system to improve the understanding of various
aspects for a variety of tasks. Trajectory modeling of a projectile system is a process and its important characteristics lie in its
mathematical representations. The applicability of a model is limited by the assumptions and the uncertainties in the
evaluation data, understanding the judgments associated with the modeling process. Its analysis under uncertainty
environments is an engineering discipline, and a powerful tool in any weapon systems developmental programmes. A
fundamental understanding is necessary to comprehend the factors that influence accuracy and how to account for them in the
determination of firing experimental data. Growing international competition in the technological and industrial applications
has necessitated that armament designers, managers and practitioners ensure a level of performance of their products before
releasing them for service use. The rational treatment of uncertainties in trajectory modeling has received increasing
attention in recent years. Loading and boundary conditions, material properties, geometry and various other parameters show
in some cases considerable variations at macro-scale or at micro-scale in a multi-levels physical context.
Observations and measurements of physical processes as well as parameters at different scales clearly indicate their
random characteristics. Uncertainty analysis has been an established engineering methodology, but is relatively new to
epidemiology for the quantitative assessment of biases, hence the ballistician's interested to apply it to the armament field too.
A systematic uncertainty analysis also provides insight into the degree of confidence of a measurement or experimental
results in the existing data and models to assess how various possible model estimates should be weighed. In view of above,
this paper will consist of two sections. Section one will discuss various types and sources of errors and uncertainties in
projectile trajectory modeling from physical, mathematical, experimental and computational aspects. Section two will
discuss a systematic approach to quantifying various parameters from a case study.

In our physical world, motion is ubiquitous and controlled by the application of two principal ideas, force and
motion. The study of these ideas is called dynamics. Through the study of dynamics, we are able to calculate and predict
the trajectory of a projectile system. Trajectory modelling is the process of identifying the principal physical dynamic
effects by analyzing a system, writing the differential and algebraic equations from the fundamental laws. Modelling
projectile motion and its simulation is nothing but as a computational tool to calculate the flight path and its
characteristics. There is a wide span of trajectory modelling and simulation differing widely with respect to their
performance and fidelity characteristics, such as specifying ammunition performance requirements, designing
ammunitions, optimizing the design parameters, assessing ammunition performance, teaching users the correct use of
weapon and fire control system.
The study of projectile motion is called ballistics, that is, it is, in essence, the study of projectile motion and the
conditions that influence that motion. As an applied and complex fields of science, ballistics has a wide meaning and a
wide scope of subjects, ranging from the behaviors of projectile inside a gun barrel (Internal Ballistics), the flight of
projectile through the air (External Ballistics) and the interaction between projectile and targets (Terminal Ballistics),
which shown in Figure 1. Its development is very much related to the evolution of military technology and particularly
in artillery projectile trajectory modeling system.
Similarly, the science of investigating projectile trajectory modeling and computation in space is known as external
ballistics or the science of flight dynamics. It existed for centuries as an art before its first beginnings as a science. For
understanding the inflight behaviour of the projectile system is to develop the mathematical models that can predict all
elements of the trajectory from launch to target. External ballistics variations are the main concern for artillery projectile



system in the field. Figure 2 depicts an exterior ballistics environment of a projectile system. Figure 3 depicts the global
motion of a projectile system in space.
Recent involvement in designing and developing a trajectory and simulation model to allow interaction between
physical, mathematical and computational processes has also led to interest in error and uncertainty propagation. The
effectiveness of a trajectory modelling is closely linked to the accuracy to which it may be delivered to the intended target
area. The trajectory of a projectile depends upon many factors, such as the components of initial/launch velocity, launch

Figure 3: Global Motion of a Projectile System



angle of the projectile, its mass/weight, shape, size, rate of spin, drag characteristics, the density and temperature of the air, the
direction and velocity of the wind at various altitudes, rotation-of-earth effects, forces and moments induced by projectile spin, and
the latitude of the gun and the azimuth of fire etc. Various models may be used, with varying complexities, to incorporate any of the
above effects. The main classes of trajectory models are known as: in-vacuo, point-mass, modified point-mass and six-degree-of-
freedom, universal and artificial neural networks.
In development of models, we often combine many parameters together to produce models with hundreds and even
thousands of parameters. The uncertainty in these models arises not only from the uncertainty in the model inputs but also due
to other more fundamental problems associated with our ability to model complex systems. Often in presenting model outputs
we fail to adequately represent these uncertainties. In this paper we intend to explore some of the issues associated with
uncertainty in projectile trajectory modelling.
Random firing vibration resulting projectile's initial disturbing having random error, which is got at the moment when they
fly out of the muzzle. Thereby, projectile deviate the predicted trajectory, which resulted in dispersion of a series of, fired projectiles.
So, dispersion of projectiles is decided by a few random elements. The following equation for muzzle velocity variation (MVV) is
valid for our purposes:

MVV (m/s) = Shooting Strength of Weapon + Ammunition Efficiency +Round to Round Variation;

Random nature of various parameters in ballistic test and evaluation demands probabilistic approach, which provides the
mathematical basis for evaluation of projectile's trajectory system performance via prediction of overall projectile's dispersion
accuracy or consistency, which are random variable and assumed a normal multidimensional distribution. It is essential parameters
for a projectile-weapon-ammunition system, which affect performance to simulate the firing process and predicting dispersion
In this paper, a general methodology for uncertainty analysis is presented that combines modeling, simulation and
experimental techniques for an understanding of projectile's trajectory modeling system performance and is in accordance to the
principles of the GUM. In view of the above, first, this paper discusses various types of errors and uncertainties, and its
classifications in general. In second, a case study has been discussed for quantifying various parameters from a test firing system.

Projectile Motion As A Dynamical System

A projectile is defined as “an object, projected by an applied exterior force or impulse and continuing in motion by virtue of
its own inertia over a very short period of time”. A projectile with its weapon and ammunition form a projectile system. It is term that
emerges from many disciplines and domains and has many interpretations, implications and problems associated with it. Generally,
modern artillery projectile systems have many features in common, which include a fuze, ogive, bourrelet, and rotating band.
Figure 4 shows the concept of a simple projectile system.
Similarly, linked with projectile motion, a trajectory can be defined as an imagined trace of positions followed by an object
moving through space. A projectile system, when subjected to propelled, fired, shot, launched, or thrown through space by the
exertions of a force by a suitable means, behaves dynamically. This is a dynamic system, because the level of the elements changes
in time. Mathematically the term trajectory refers to the ordered set of states (as a function of the time), which are assumed by a



dynamical system over time. Mathematically a trajectory may be described either by the geometry of the path (i.e. the set of
all positions taken by the object), or as the position of the object over time.
To describe a trajectory completely it is sufficient to specify the x, y and z coordinates of the center of gravity of the
projectile at any time t after the release by the projecting mechanism; in other words, a trajectory is regarded as defined by: x =
x(t); y = y(t); z = z(t); where x(t), y(t) and z(t) are functions of the time t which are equal to zero when t = 0. The trajectory is
determined by: (a) the position of the origin; (b) the conditions of projections;(c) the ballistic characteristics of the projectile;
(d) the characteristics of the air through which it passes;

Projectile Trajectory Modeling And Simulation

Trajectory modeling and simulation, in the present context, means computing the flight path and other parameters,
such as range, deviation, orientation, and angular rates of a projectile system from the start to the end of its motion. Trajectory
modelling deals with mathematical models of the behavior of projectile and its subsystems (if any) during its test &
evaluation and operations. The equations of motion determine the acceleration, velocity and position (range & deviation) of
the projectile resulting from forces and moments due to gravity, thrust and aerodynamics. If all characteristics of a projectile,
together with atmospheric conditions are exactly equal to a set of predicted values, the projectile will fly on a known
trajectory. This trajectory is called nominal trajectory. In practice, there are always some differences between the real and
predicted values. These differences make the projectile not to fly exactly on its nominal trajectory. Therefore, there are
always some errors between the positions of a desired and real impact point. Estimation of these errors is very important from
the operational point of view. Also, investigation of the error sources and their effects can help a projectile designer to
optimize design parameters for the lowest impact point error. The guidance and control models account for subsystems such
as the control system in case of a missile system.
Particularly trajectory simulation requires a customized approach. The level of sophistication of simulations varies
greatly depending on the applications ranging from unsophisticated two-dimensional models to very detailed six-seven
degree of freedom models as per shapes and sizes of projectile systems. The requirements of a given trajectory simulation are
derived from the objectives of the intended users requirements like, analysis, development, procurement and operation. The
modeling phase is a decisive stage because the analysis of the model depends strongly on the model. The choice of the model
depends on several criteria, the most important ones are; (i) data of the modeled system, (ii) system's complexity, (iii)
availability of a computer tool implementing the model, (iv) compatibility of the model with the analysis leading to expected

Mathematical Modeling
Let us first consider 2-D (3-DOF) case, the projectile as a particle and the motion of the projectile of mass m influenced by
gravity g and air resistance. Therefore the trajectory modelling of the projectile motion is described by:

d 2x d2y
ax = =
(1 / m ) F
× ×
cos a
q x = (1 / m) ×
- ( FD ×
sin q
+g ) (1)

dt 2
dt 2
Where , and
v +
v tan (
u= 2
v / vx )
y FD =
0.5 ×

v2 ×

by incorporating following basic limitations/assumptions:

(a) Ignoring of the earth's curvature or earth's flatness;

(b) Planar Motion of the projectile;
(c) No wind speed;
(d) Constancy of drag index and projectile's mass;
(e) Proportionality of drag force to squared velocity of the projectile;
(f) Constancy of drag coefficient and air density;
(g) Constancy of drag force or altitude independency of drag force;
(h) Neglecting of centrifugal, coriolis, magnus forces and its cross-effects;
The above equations are solved subject to the following state or initial conditions:

dx / dt =
v0 cos q
0;v =
dy / dt =
v0 sin q
0 v =
v0 ; x =y =0; at t = 0 (2)



Where the drag force FD acting on the projectile depends on air density ñ, velocity v, cross-section area
A= ð×r2, radius of the projectile r, and drag coefficient CD. If the drag force is neglected, the calculation of the projectile
trajectory becomes trivial. If however the drag force is taken into account, the analytical solution is not soluble due to
the drag dependency on the square of the velocity and presence of angle as argument of trigonometric functions.

Target Density Parameters

(D i
(i) Elevating probability error, Ba =
0.6745 1
(ii) Azimuth probability error, å
Z) i
Bb =
Where, Yi elevation coordination deviation from the center of clustering (meter), Yi=Yi - YcP ;, Zi azimuth
coordination deviation from the center of clustering (meter), Zi=Zi - ZcP ; YcP , ZcP clustering coordinate (meter)
n n
å Y
1 i
å Z , n - the number of the projectile; Yi and Zi elevation, azimuth coordinate of each projectile.
1 i
YcP = Z cP =
n n


Uncertainty is an integral part of almost every real-world problem due to its complex nature of the physical process.
The objective of uncertainty analysis or propagation is to characterize the uncertainties in the system output given some
knowledge of the uncertainties associated with the system parameters together with one or more computational models or
experimental data. For realistic systems one must include uncertainties of various types in the mathematical model of the
system. Therefore, researchers and scientists attempting to have a better understanding of the process have devoted
tremendous efforts. The appropriate incorporation of uncertainty into the analyses of complex systems is a topic of
importance and widespread interest. The particular context under consideration is shown in Figure 5.
Projectile trajectory modelling system analyses and designs uncertainties can arise from the various sources
including environmental uncertainties, model uncertainties, parameter uncertainties, data uncertainties, and operational
uncertainties. Environmental uncertainty is associated with the inherent randomness of environmental processes such as the
occurrence of wind speeds and humidity. A model is only an abstraction of the reality, which generally involves certain
degrees of simplifications and idealizations. Model uncertainty reflects the inability of a model or design technique to
represent precisely the system's true physical behavior. Parameter uncertainties resulting from the inability to quantify



accurately the model inputs and parameters. Parameter uncertainty could also be caused by change in operational
conditions of system structures, inherent variability of inputs and parameters in time and in space, and lack of sufficient
amounts of data. Data uncertainties include (1) measurement errors, (2) inconsistency and non-homogeneity of data, (3)
data handling and transcription errors, and (4) inadequate representation of data sample due to time and space limitations.
Operational uncertainties include those associated with construction, manufacture, deterioration, maintenance, and human.
Uncertainty analysis of measurement results is not a deterministic, but rather holistic and probabilistic in nature. Its
complexity and ambiguity contributes that the uncertainty analysis is often misunderstood, misrepresented or even avoided.
The process of developing uncertainty budgets requires a manufacturer to first find the causes of measurement uncertainty
and quantify them. A manufacturer or designer or evaluator must understand the measuring process being evaluated, and a
certain amount of experience, to identify and quantify uncertainty contributors. There are a couple of techniques that can
help in systematically searching for uncertainty contributors. When collecting uncertainty contributors, it is important to
know the areas where they may be lurking. These 10 areas are the most common a designer should use in looking for
uncertainty contributors. The "circle of contributors" (Figure 6) technique involves using a list of areas where uncertainty
contributors may be lurking as an aid in searching for them.

The process of generating and propagating the simulation values is shown in Figure 7. For each actual measured
coordinate point, (Xi, Yi, Zi), the associated errors, (Xi, Yi, Zi), are generated by sampling from the appropriate estimated
distribution. This process creates a set of new measurement points, (, which are then evaluated by the trajectory computation
software according to the feature being measured. Figure 7 shows an example of a trajectory path being measured; is the
final calculated trajectory of the projectile, and the function f represents the algorithm used to calculate the trajectory path
from the set of n data points. This process will be repeated a number of times according to how the experiment is designed,
and the statistics for will be used to estimate the combined standard uncertainty for the measurement of the trajectory path.

Definition Of Error And Uncertainty

The performance of a trajectory modelling system is adversely affected by error and uncertainty. According to the
scope of the Guide to the Expression of Uncertainty in Measurement (GUM), the error and uncertainty can be defined as:
(a) Error: A recognizable deficiency that is not due to lack of knowledge.
(b) Uncertainty: A potential deficiency that is due to lack of knowledge.

These two definitions imply that error is deterministic in nature and uncertainty is stochastic or non-deterministic in nature.



Types of Errors
Errors can be classified as either "acknowledged" or "unacknowledged":
(A) Acknowledged errors are those, which have been identified in a simulation, but the analyst decides not to
remove them because the results are adequate for the job requirements and budget (examples include round-off error and
discretization error, and convergence error in an iterative numerical scheme).
(B) Unacknowledged errors in a simulation are those the analyst is unaware of but they are recognizable. They
have no foolproof procedures for finding them but they may be detected by redundant procedures and double-checking
(examples are computer programming errors, or usage errors, including mistakes and blunders).
Our knowledge of how the physical system works is limited by at least five kinds of uncertainty: (1) inadequate
knowledge of all the factors that may influence something, (2) inadequate number of observations of those factors, (3)
lack of precision in the observations, (4) lack of appropriate models to combine all the information meaningfully, and (5)
inadequate ability to compute from the models. It is the incompleteness of the knowledge of the value of the measurand
associated with the given result of a measurement. It is an interval within which a true value of a measurand lies with
given probability.

The reasons responsible to introduce uncertainty in a model may be:

(a) The model structure, i.e., how accurately does a mathematical model describe the true system
for a real-life situation;
(b) The numerical approximation, i.e., how appropriately a numerical method is used in
approximating the operation of the system;
(c) The initial / boundary conditions, i.e., how precise are the data / information for initial and /
or boundary conditions;
(d) The data for input and/or model parameters;

Types of Uncertainties
The following three types of uncertainties can be identified:
Type-I: Uncertainty due to variability of input and/or model parameters when the characterization of the variability is
available (e.g., with probability density functions, (pdf ))
Type-II: Uncertainty due to variability of input and/or model parameters when the corresponding variability
characterization is not available;
Type-III: Uncertainty due to an unknown process or mechanism;

Type I uncertainty may be referred to as “aleatory” or “quantitative” (i.e., dependent on chance) or “random” or
“stochastic” or "inherent" or "irreducible" uncertainty. It can be expressed using numerical values. It is not due to a lack of
knowledge and cannot be reduced. Aleatory uncertainty, however, generally characterizes due to random processes or effects,



such as the wind speed and direction at an arbitrary point in space as a flying projectile passes.

Type II and III are referred to as epistemic or qualitative uncertainties. It can be expressed using qualitative terms and used
when the sources of uncertainty cannot be estimated precisely using numerical values. Epistemic uncertainty deals with our
state-of knowledge about portions of our model. This uncertainty has been called “subjective,” or "parameter," or “state-of-
knowledge” or "reducible" uncertainty. Examples are such things as inaccuracies and measurement error in the physical
burn-rate parameters characterizing a propellant. Epistemic uncertainty can be reduced or eliminated by an increase in
knowledge about the system being analyzed or measurement of the parameters.

Quantification of Uncertainty
It often happens in real life applications that all three types of uncertainties are present in the systems under study. The
goal of uncertainty quantification is to assign an appropriate mathematical model to real-world situation with respect to
objective and subjective uncertainty. The choice of an appropriate uncertainty model primarily depends on the
characteristics of the available information. That is, the underlying reality with the sources of the uncertainty dictates the
model. In each particular case this information must be analyzed and classified to be eligible for quantification.
Quantification or evaluation of uncertainties is very judgmental process, based on full scientific understanding,
knowledge and experience of measured quantities, properties and characteristics of instrumentation used, measurement
processes and procedures, and the data reduction procedures used to obtain the final measurement result.
Uncertainty quantification intends to work toward reducing type II and III uncertainties to type I. The quantification
for the type I uncertainty is relatively straightforward to perform. Techniques such as Monte Carlo are frequently used. Pdf
can be represented by its moments (in the Gaussian case, the mean and covariance suffice), or more recent techniques. To
evaluate type II and III uncertainties, the efforts are made to gain better knowledge of the system, process or mechanism.
Methods such as fuzzy logic or evidence theory are used.
In the planning phase of an experimental program, one focuses on the general or overall uncertainties. Consider a
general case in which an experimental result, f, is a function of n measured variables Xi:
f = f ( X , X ,... X ) (3)
1 2 n

Equation (3) is the data reduction equation used for determining f from the measured values of the variables Xi. The overall
uncertainty in the result is then given by

U2 =
¶f ö æ
U2 +
¶f ö
U2 +
¶f ö
... + U2
f ç
X1 ÷

x1 ç
X2 ÷

x2 ç
Xn ÷

è ø è ø è ø
where are the uncertainties in the measured variables Xi. It is assumed that the relationship given by Eq.(3) is
continuous and has continuous derivatives in the domain of interest, that the measured variables Xi are independent of one
another, and that the uncertainties in the measured variables are also independent of one another.

Uncertainties In Projectile's Trajectory Modeling Process

Uncertainty in projectile trajectory modelling system has become of interest in recent years and emerged from a
combination of various computational aspects in interior or exterior or terminal ballistics models. Decision-makers have also
expressed a desire to have probabilities assigned to each scenario so that there is a better sense of whether certain scenarios are
more likely than others. Difficulty in assigning probabilities to scenarios can stem from "epistemic" or "stochastic" sources of
uncertainty. Epistemic sources of uncertainty are those that can be reduced by further study of the system, improving various
states of knowledge, etc. Stochastic sources of uncertainty are those that are considered "unknowable"items such as variability
in the system, the chaotic nature of the projectile system, and the indeterminacy of human systems.
Generally, uncertainties in projectile trajectory modelling are considered are mainly attributed to ambiguity and
vagueness in defining the variables and parameters of the systems and their relations. The ambiguity component is generally
due to noncognitive sources, which include (a) physical randomness; (b) statistical uncertainty due to the use of limited
information to estimate the characteristics of these parameters; and (c) model uncertainties that are due to simplifying
assumptions in analytical and prediction models, simplified methods, and idealized representations of real performances.
Similarly, the vagueness-related uncertainty is due to cognitive sources that include (a) the definition of certain variables or
parameters, e.g., structural performance (failure or survival), quality, deterioration, skill and experience of construction
workers and engineers, environmental impact of projects, conditions of exiting structures; (b) other human factors; and (c)
defining the interrelationships among the parameters of the problems, especially for complex weapon systems.
The accuracy of a projectile's trajectory modelling system firing such as most tank or artillery systems relies on
understanding numerous sources of error and correcting for them. Random errors are addressed through projectile and its
ammunition design, quality production, effective maintenance, and firing processes that minimize the magnitude of these
error sources. Bias errors are different because their magnitude and direction can often be estimated and then compensated
for while aiming the weapon.



Characterizing Uncertainties Parameters

A common feature of engineering models is the need to provide values for model parameters that define properties,
or parameters, which appear in boundary or initial conditions. For example, according to drag law the combined effects of air
resistance can be reduced to a drag co-efficient for a projectile. It states that the drag force is proportional to the cross-
sectional or reference area of the projectile, air density in which the projectile moves and its velocity squared. The
proportionality constant is known as a drag coefficient (CD(a ) or CD(M )), which is widely used to represent a projectile's
aerodynamic efficiency. Its value depends on a function of the angle of attack, a or Mach number, M. Mathematically, it can
be written as Fdrag = CD(a) or CD(M )A(½ r V2), where CD(a ) or CD(M )= the coefficient of drag,;
A = cross-sectional or reference
area of the projectile; r = the density of air (~1.2 Kg/m3);and V = the velocity of the projectile relative to the air. The value for
this model parameter i.e. drag coefficient is estimated using experimental techniques utilizing samples of the projectiles. In
practical terms the drag co-efficient is a function of velocity, usually represented as the Mach number (M) that varies with air
temperature, which in turn reduces with altitude. The drag co-efficient is estimated during ammunition design, or since
computer simulations became possible, it can be modelled. These estimates can then be confirmed or adjusted from 'range &
accuracy' firing. Another widely used data is the ballistic co-efficient (CB), which reflects the rate at which velocity is lost as
the projectile penetrates air. The ballistic co-efficient is sometimes called the 'carrying power' of a shell, meaning how far a
given muzzle velocity will 'carry' it. It is expressed as CB=m/nd2, where CB = carrying power, m = projectile mass, d =
projectile diameter and n = êóô, where ê (kappa) is the co-efficient of shape, ó (sigma) the co-efficient of steadiness and ô
(tau) the co-efficient of air density. This means that mass is the main determinant of carrying power and is the reason that
artillery shells go further than rifle bullets even when the latter are fired at higher velocities. For a particular muzzle velocity
and firing elevation angle the heavy shell will always go further (unless its shape and steadiness are significantly inferior).
Generally, the parameters and co-efficients in models are not constant. The drag co-efficient varies with projectile velocity.
Particularly, important parameter, the density of the atmosphere varies with altitude. In flight, a projectile's velocity
decreases with gravity and air resistance until vertex is reached, when air resistance continues but gravity accelerates the
projectile as it descends (hence the elliptic trajectory). The rate of spin (revolutions per second) also decreases with velocity
throughout the shell's flight, which affects its stability. Gravity is not constant either; it decreases with distance from the
earth's surface (which is uneven).
Unfortunately, the methods used to characterize drag coefficient will always lead to errors in the measured coefficient due to
experimental technique and equipment, due to uncertainty associated with uncontrolled environmental conditions, and due to
uncertainty associated with variability from projectile to projectile because of the manufacturing process. A key question is
how do we quantitatively characterize the sources and nature of this uncertainty.
Looking at the parameters affecting a trajectory modelling, simulation and its analysis, six main categories can be
differentiated as:
(a) Parameters directly affecting the initial conditions or Fire Control: gun position, gun
height, muzzle velocity, azimuth and elevation or launch angle;
(b) Parameters indirectly affecting the initial conditions: propellant temperature and barrel
abrasion or wear;
(c) Parameters directly affecting the projectile: in flight containing all atmospheric properties
depending on height (temperature, air density, air pressure, wind, speed of sound) and all
projectile depending coefficients (drag, lift, Magnus and Coriolis forces, spin and yaw);
(d) Parameters indirectly affecting the projectile: case thickness, case strength and explosive
(e) Parameters directly affecting the target: impact angle, velocity, location, and angle-of-
(f) Parameters indirectly affecting the target: thickness and strength;
From a measurement point of view, all these parameters are error afflicted and therefore, every parameter can be
quantified by its mean and standard deviation. When we develop mathematical or simulations models we sometimes deal
with systematic uncertainties, which can originate from biases in measurements or biases in solutions methods. Examples of
systematic uncertainties include:
(a) Instrumentation system error when measuring gun tube elevation and azimuth angles;
(b) GPS error in target location;
(c) Measurement error in depth of penetration, range and deviation data;
(d) Prediction error because a physical phenomena is not accounted for in the model equations;
The gun itself also affects the projectile's flight due to 'droop' and 'jump'. When the axis of the bore at the breech
differs slightly from the axis of the bore at the muzzle the 'droop' generates. Similarly, the movement of the projectile's 'whip-
action', that means when the projectile leaves the axis of the bore at the muzzle (in the vertical plane) differs to its axis before



firing, the 'jump' generates. Similarly, a difference in height between a gun and its target also affects the trajectory.
Therefore, the angle of sight has to be added to or subtracted from the gun's firing elevation angle. However the slant range
to the target is greater than the horizontal range and applying the angle of sight does not compensate for this, which is called
non-rigidity of the trajectory is shown in Figure 8. The size of the correction for this depends on the difference in height
between the target and the gun, and the shell's angle of descent. The gun's muzzle velocity and the range to the target in turn
determine this angle. Furthermore, altering the elevation angle also puts the trajectory through a different 'slice' of the
atmosphere, where air resistance may be different. This leads to the subject of 'Ballistic Angles', or perhaps more correctly
'Gunnery Angles', the various elevation angles related to a barrel and a shell fired from it. At the other end of the trajectory
the Angle of Descent was the angle at the point of impact between the Angle of Sight and the tangent of the trajectory which
shown in Figure 9.

x sx

In measurement theory basically all measurands are error afflicted. These errors are quantified by two characteristic
values: mean value of the sample ( ) and standard deviation of the sample ( ). All errors afflicted measurands are assumed
Gaussian distributed. Trajectory modelling effects are inherently probabilistic in nature. Random and systematic
uncertainties exit in every phase of a projectile-weapon-target engagement.

A Case Study
The trajectory modeling used in this study has been applied to the 105mm caliber artillery projectile dynamics.
Table 1 shows a list of ten uncertainty parameters that have been used in this work. Therefore, in this present work, for the
sake of computational difficulty, the simple model with 3-DOF concepts is proposed to simulate the projectile trajectory
motion and efforts focused to find out the errors associated with the other parameter uncertainties. The limits of
uncertainties presented in Table 1 are biased on a set of known observations and measurements. The simulated and
experimental results of the projectile trajectory modeling with its accuracy are shown in Table 2.

Table 1: Uncertainty Parameters and Ranges



In general uncertainty due to inherent randomness of a trajectory modeling processes cannot be eliminated. So,
understanding trajectory modeling is one thing, applying it in the field to predict a required trajectory elements is a different
matter altogether. Combining these to predict the required trajectory modeling is the challenge. Various trajectory models
have been developed, solved and refined by available state-of-the-art mathematical and computational techniques. The first
step is a mathematical model, a set of equations of motion, to describe a projectile's trajectory and allow the variables to be
represented. On the other hand, uncertainties such as those associated with lack of complete knowledge about the process,
models, parameters, data and etc can be reduced through research, data collection and careful test and evaluation and
manufacturing processes. Given the complexities and uncertainties involved, combined with difficulties in getting accurate
data, these models and their data contained many approximations and simplifications. The second step is to derive methods
and data from these models that can be used in the field.

The author is grateful to Maj Gen Anoop Malhotra, Director, Proof and Experimental Establishment (PXE),
Balasore, for his consent to publish this paper.

[01] Weaponeering: Conventional Weapon System Effectiveness- M. R. Driels, 1st Edition, AIAA Series, Inc,
Virginia, 2004.
[02] Measurement, Design and Analysis: An Integrated Approach- E.J. Pedhazur, and L.P. Schmelkin,, LEA
Publishers, London,1991.
[03] Engineering Design Handbook, Army Weapon System Analysis, DARCOM Pamphlet,Pt-I, 1977.
[04] Decision Making Under Uncertainty: What Every Engineer Should Know About John X. Wang, Morcel
Dekker, Inc, New York, 2002.
[05] Uncertainty Modelling and Analysis in Engineering and the Science BilalM. Ayyub & George J.Klir,
Chapman Stall/CRC, Taylor & Francis Group 2006.
[06] Przemieniecki, J.S., Mathematical Methods in Defense Analyses, AIAA Series, Virginia, 2000.
[07] Text of Ballistics and Gunnery-I and II, London: Her Majesty's Stationary Office, 1987.
[08] An Analysis of Engagement Algorithms for Real-Time Weapon Effects- John Mann, et al, JDMS,Volume 3,
Issue 3, July 2006, Pages 189-201.
[09] Measurement Theory for Engineers - Ila, G.,1st ed., Springer-Verlag, NY, 2003.
[10] Measurement Errors and Uncertainties: Theory and Practice- S.G. Rabinovich,, 2nd ed., Springer Verlag, NY,
[11] The Theory of Errors in Physical Measurements- J.C. Pal , 1st ed., NCBA(P) Ltd, Kolkota,1995.
[12] Guide to the Expression of Uncertainty in Measurement, ISO, Geneva, Switzerland, 1995.
[13] Unleashing Error or Uncertainty Analysis of Measurement Results M Kostic, NASA 2003 Faculty Fellowship
Program Lecture Series (http://www.kostic.niu.edu/Uncertainty-Analysis-of- Measurement-Results.pdf).
[14] Estimation of measurement uncertainties an alternative to the ISO Guide-M. Grabe, Metrologia,2001, 38, 97-
[15] Realization of a virtual experiment system of gun- Yi Hang, Xiaofeng Wang, Guolai and Yunsheng Chen,
World Journal of Modelling and Simulation, Vol.2 (2006), No. 4, pp-255-260.
[16] Modeling of Tank Gun Accuracy Under Two Different Zeroing Methods David W. Webb and Bruce J. Held,
Sixth Army Conference on Applied Statistics, Houston, TX, 2000.
[17] Fundamental of Ballistics-Gunther Dyckmans (http://users.telenet.be/gd.ballistic.htm).



[18] Military Ballistics: A Basic Manual-G.M.Moss,D.W.Leeming and C.L.Farrar, Brassy's Ltd series,UK, London,
Vol-I, 1995.
[19] Mathematical Model to Simulate the Trajectory Elements of an Artillery Projectile Proof Shot - K.K. Chand,
and H.S. Panda, , Defence Science Journal , 57, No1, January, 139-148, 2007.
[201] An Ontology For Trajectory Simulation- U Durak, H Oguztuzun and SK Ider,Proceedings of the 2006 Winter
Simulation Conference, pp-1160-1167.
[21] System Analysis and Modelling DW Boyd, Academic Press, San Diego, CA 92101- 4495, USA, 2001.
[22] Applied Mathematical Modelling of Engineering Problems N Hritonenko and Y Yatsenko, Kluwer Acdemic
Publishers, USA, 2003.
[23] Applied Modelling and Simulation: An Integrated Approach to Development and Operation-Cloud and Rainey,
The McGraw-Hill Companies, Inc, 1998.
[24] Trajectory Correction Flight Control System using Pulsejet on an Artillery Rocket SK Gupta et al, Defence
Science Journal , 58, No1, January, 15-33, 2008.
[25] Modeling and Problem Solving Techniques for Engineers L Horvath and IJ Ridas, Elsevier Acdemic Press,
UK, London, 2004.




SN Giri, KK Chand
Proof & Experimental Establishment, Chandipur, Balasore, Orissa - 756 025

Applied Physics, a generic term, is intended for various engineering and technological applications, which deals
with the properties and interactions of matter and energy, space and time. Discovery via experiment provides an essential link
between physics and engineering system problems. "Applied" is distinguished from "pure" by a subtle combination of
factors such as the motivation and attitude of researchers and the nature of the relationship to the technology or science. It
usually differs from engineering in that an applied physicist may not be designing something in particular, but rather is using
physics or conducting physics research with the aim of understanding or developing new technologies or solving engineering
problems. This approach is similar to that of ballistics, which is the name of the applied scientific field. Competence in
Applied Physics and Ballistics (APB) is important interdisciplinary research areas in armaments. Because of their
interdisciplinary nature, the APB is inseparable from physical, mathematical, experimental and computational aspects. In
this context, this paper will provide a brief review into the APB, their important role in the armament research and also
summarize some of the current APB activities in academia, armament research institute and industry.
Objectives And Scopes
Applied Physics is a branch of physics that concerns itself with the applications of physical knowledge to other
domains. It is the discipline that deals with concepts such as matter and energy, space and time. It is the most fundamental
science, and enabling discipline underpinning much of our engineering and technology. Applied (or perhaps Applicable)
Physics and Ballistics consist of physical laws and experiments and results, including those from "pure" physical areas which
are used to assist in the investigation of problems or questions originating outside of physics. Applied Physics and Ballistics
(APB) are closely related disciplines, with ballisticians using physical principles to understand various ballistics problems in
armaments. Studying applied physics is important in weapon technology undergoing rapid technological change, and
requiring innovative problem solving. This is because applied physics studies develop critical thinking skills and the
understanding of fundamental scientific principles of wide applications and lasting value.
As over the last two millennia, physics has sometimes been synonymous with philosophy, chemistry and certain
branches of mathematics and biology but it emerged as a modern science in the 16th century as an applied physics. Applied
physics is now generally distinct from these other disciplines, even though its boundaries remain difficult to define
rigorously, especially in areas such as mathematical physics, computational physics, experimental physics, numerical
physics, biophysics, chemical physics, engineering physics and so on.
Applied Physics and Ballistics are both significant and influential in part, because advances in its understanding
have often translated into new technologies in armaments, but also because new ideas in applied physics often resonate with
the other sciences, mathematics, chemistry, computer science and biology. For example, advances in the understanding of
electromagnetism led directly to the development of new products that have transformed armament field (including
electromagetic gun, high speed computers and so on); advances in thermodynamics led to the development of advanced
electrothermal gun and high energy matereials or propellant; and advances in mechanics inspired the development of the
various trajectory modelling, and so on.
Similarly, the understanding and use of acoustics principle will result in better blast wall for safe testing
environment ; similarly, the use of optics creates better optical devices like high speed cameras. An understanding of applied
physics makes for more realistic trajectory modelling simulators, as well as in forensic investigations (what do we know and
when do we know it; what did the subject know and when did the subject know it).
Studying a multidisciplinary field is challenging. Not only must we learn more than one discipline, we must also
work with the separate languages and styles the different disciplines. In particular, our objective is to study and develop



various physical systems and its mathematical models e.g. thermodynamics system for high energetic materials like
propellant or gyroscopic system for stability behavior of projectiles and so on. In view of above, this paper discusses a review
of various aspects of role of Applied Physics and Ballistics in armaments.

Disciplines Of Applied Physics In Ballistics

There are many fields of Applied Physics which have strong branches, as well as many related and overlapping fields from
other disciplines that are closely related to Ballistics.
· Acoustics, the study of sound waves, is used for various shock wave phenomenon of a projectile system, and
muzzle devices for intermediate ballistics;
· Biophysics is the interface of biology and physics for studying various wound ballistics phenomena;
· Chemical physics studies the structure and dynamics of ions, free radicals, polymers, clusters, and molecules,
using both classical, chemical and mechanical viewpoints for internal ballistics;
· Engineering physics such as optics, nanotechnology, control theory, aerodynamics, or solid-state physics for
various ballistic intrumentations;
· Fluid mechanics is the study of fluids (liquids and gases) at rest and in motion. The Navier-Stokes equations are
used in supercomputers to model internal ballistics phenomena;
· Aerophysics is study of the properties and charactristics and the forces exerted by air in motion over a body.
· Lasers and radar are used for various projectiles tracking purposes;
· Materials science is the systematic study of the properties of materials for terminal ballistics;
· Medical physics includes the standards for radiation exposure and infrastructure for radiology for wound ballistics;
· Nanotechnology studies the creation of machines less than a micrometer in size for advance weapon systems;
· Optoelectronics creates devices which use light rather than current for development of ballistic instruments;
· Physics of computation used for solving of various nonlinear problems of internal, intermediate, external, terminal
and wound ballistic phenomena;
· Solid state physics, including the material properties of semiconductor devices and integrated circuits for
development of ballistic instruments;


Ballistics, as a science, that deals with the motion, behavior, and effects of projectiles. It relates to a great variety of
phenomena that occurs from the moment as object (projectile) is fired until its effects are observed on a target. A ballistic body
is a body which is free to move, behave, and be modified in appearance contour, or texture by ambient conditions, substances,
or forces, as by the pressure of gasses in a gun by rifling in a barrel, by gravity, by temperature, or by air particles. Ballistic
studies include applications as varied as the study of the structural and control behaviour of rockets and satellites; strikes on
aircraft, terrorist attacks and automobile crashworthiness modeling, to name but a few. Many of the basic problems of
ballistics are similar to those in other fields of applications, such as combustion, heat conduction, in-flight structural
behaviour, trajectory related issues, contact, impact, penetration, structural response to shock waves and many others.
Ballistics as a human endeavor has a long history. Ballistics, rooted as it is in the study of motion and indispensable
as it has become in the affairs of nations, has held the interest of scientists, engineers, generals, and rulers. From the earliest
developments of gunpowder in China more than a millennium ago, there has been an intense need felt by weapon developers
to know how and why a gun works, how to predict its output in terms of the velocity and range of the projectile it launched,
how best to design these projectiles to survive the launch, fly to the target and perform the functions of lethality, and the
destructions intended. Physicists and mathematicians have made fundamental contributions to this science, including many
of the most significant contributions.
Ballistics is inherently a branch of engineering physics. Not only are there a distractingly large number of physical
factors present in the problem, but the number of independent variables that are actually taken into account in firing is so
large, and these variable factors are to a great extent so uncontrollable, that simple accurate formulae are universally regarded
as out of the question. Rules based on the normal probability curve are used in actual field conditions, and the numerical
tables upon which original computations are constructed are themselves obtained by statistical methods involving more or
less arbitrary assumptions. The only methods used in the past have been to relegate the less important variations to a
secondary place (not to be ignored, but to be treated as linear differential perturbations) and to incorporate only the most



important features into the fundamental differential equation' for a trajectory. What this equation is, depends in part on the
particular problem, but is in no case as simple today as it was before the World War. In the sense of "rational mechanics", there
is no rational ballistics.
For ballistics, as for every other developing branch of engineering, there is a group of present-day problems
involved on the one side in seeking to test present-day assumptions, and if possible to establish these on a rational basis, and
on the other side in devising means for application of these accepted principles with an accuracy consistent with the data and
yet with the greatest possible simplification concordant therewith.
An excellent survey of the history of ballistics, replete with illustrative comparisons, explicit data, and critical
comments on physical assumptions, is given by the long classical work of C. Cranz. The only criticism is that despite
supplementary notes in later editions, the work closes essentially with the period before the World War, which changed the
perspective even on old problems.
The discipline over the centuries, ballistics has divided itself into three natural regimes internal ballistics, external
ballistics and terminal ballistics, but subsequently it is added with transitional ballistics, wound ballistics or forensic
ballistics and hydro or underwater ballistics as per understanding and various applications and usefulness. Let us discuss each
branch of its physical concept briefly as:

Physics of Interior Ballistics

Interior ballistics deals with the temperature, volume, and pressure of the gases resulting from combustion of the
propellant charge in the gun; it also deals with the work performed by the expansion of these gases on the gun, its carriage, and
the projectile, which is shown in Figure 1. Some of the critical elements involved in the study of interior ballistics are the
relationship of weight of charge to the weight of projectile; the length of bore; the optimum size, shape, and density of the
propellant grains for different charge system in guns; and the related problems of maximum and minimum muzzle energy.
Gun system, which is one of the propulsion units, has remarkable capacity to accelerate projectile by supersonic
velocity instantaneously. In gun systems, a projectile is propelled by pressurization of the chamber caused by combustion of
solid propellant. For this analysis, the interior ballistics has been used. The interior ballistics predicts the phenomena in gun
firings by means of physical modeling and simulations of solid propellant combustion and gas generation. The calculation
starts with ignition of solid propellant, and the end of calculation is the time when a projectile reaches the gun muzzle. These
phenomena in gun system should be analyzed on the basis of fluid dynamics, thermodynamics, thermo chemistry and

Figure 1: Interior Ballistics Environment

Physics of Transitional Ballistics

Transtional Ballistics, also known as intermediate ballistics, is study of a projectile's behavior from the time it leaves
the muzzle until the pressure behind the projectile is equalized, so it lies between internal ballistics and external ballistics,
which is shown in Figure 2. Transtional ballistics is a complex field that involves a number of variables that are not fully
understood; therefore, it is not an exact science. What is understood is that when the projectile leaves the muzzle, it receives a
slight increase in muzzle velocity from the expanding propellant gases. Immediately after that, its muzzle velocity begins to
decrease because of drag.



Figure 2: Transitional Ballistics Environment

Physics of External Ballistics
Exterior Ballistics is a generic term used to describe a number of natural phenomena, which is founded on the
physics of a projectile, that tend to cause the projectile in flight to change direction or velocity or both as it moves through air,
which is shown in Figure 3. In external ballistics, elements such as shape, caliber, weight, initial velocities, rotation, air
resistance or drag, gravity, wind, drift (when considering spin-stabilized weapons) and Coriolis effect help determine the
path of a projectile from the time it leaves the gun until it reaches the target. These phenomena are gravity, drag, wind, drift
(when considering spin-stabilized weapons) and Coriolis effect.

Figure 3: External Ballistics Environment

Physics of Terminal Ballistics
Terminal ballistics, a special domain of the third branch/a sub-field of ballistics, is the study of the behaviour of a
kinetic energy projectile-when it impacts with its target. Terminal ballistics is relevant for both small caliber projectiles as
well as for large caliber projectiles. The study of extremely high velocity impacts is still very new and is mostly applied to
spacecraft design. The study of the bullet that causes impact is the main aim of this branch of ballistics. There are three basic
classes of bullet: one that are designed for maximum penetration of the target, one that are designed to penetrate a specific
depth and stop, and one that are designed specifically for short range target shooting. It involves many empirical formulae.
Theoretical investigations and experiments, however, are carried on in penetration, fragmentation, detonation, shape of
charge, and related blast phenomena, including combustion and incendiary effects. Ballisticians derive the principles
governing the elements such as number, size, velocity, and spatial distribution of fragments produced by detonations of cased
high-explosive charges, which is shown in Figure 4.



Figure 4: Terminal Ballistics Environment (Hard Target: Amours)

Physics of Wound or Forensic Ballistics
In broad terms, Wound or forensic ballistics is the study of the interaction of wounding agents (such as bullets and
fragments from explosive weapons) with tissue. The laboratory aspect of wound ballistics is the simulated and measurable
physical interaction of wounding agents with tissue. Ballistic trauma, which overlaps with wound ballistics, includes the
pathophysiological reaction of the body to the physical process, which is shown in Figure 5. Therefore, ballistic trauma
includes blood loss, shock, wound infection and death.

Figure 5: Wound Ballistics Environment (Soft Target: Tissue)

Physics of Hydro or Underwater Ballistics

Hydro or Underwater ballistics is a branch of continuum mechanics, which deals with the laws of motion of an
incompressible fluid and of the interaction of the fluid with its boundaries in systems of water-borne vehicles like ships, boats
and submarines. Because of scientific curiosity, the subject has achieved at high state of practical, analytical and
computational developments, which is shown in Figure 6.



Figure 6: Hydro or Underwater Ballistics Environment

Physics of Ballistic Measurements

The development of high-speed photography and other imaging and measuring devices have led to greater
understanding in all six branches of ballistics. By means of such devices quantitative and qualitative behaviors of any
projectile can be photographed and measured in flight, thus permitting accurate studies not only of its position to study and
determine degree of stability and integrity and even of the shock waves it produces, which is shown in Figure 7.

Figure 7: Shockwave of a Projectile by High Speed Imaging

In few decades, the most important recent development in ballistics is the use of computers and the state-of-the-art
computational techniques. The calculus of exterior ballistics generally involves sets of second-order partial differential
equations. Solving such a set of equations typically involves hundreds of thousands of computations. To find the position of
the projectile at various points along the trajectory, dozens of such solutions are required. For each of various elevations of the
gun, the entire process must be repeated. Even with the aid of slide rules and ordinary calculating machines, such an operation
would take a mathematician an inordinate amount of time. Electronic computers compile complete solutions within a few
seconds. Computers are used also for simulation of missile flights.
The design, development, and calibration of a wide variety of highly specialized optical and electronic equipments
in recent years have furthered considerably the advance of all ballistics research, particularly with respect to the performance
of guided projectiles and missiles. Examples of these instruments are long-focus tracking telescopes, photogrammetric
cameras, and miniature radio transmitters and receivers installed in projectiles and missiles.



It is well established that mathematical models and simulation techniques along with the high speed imaging
techniques is able to offer cost effective quantitative and qualitative solutions to a multitude of various internal, intermediate,
external, terminal ballistic requirements, retaining and often improving competency levels whilst reducing design costs. At
the heart of any simulation are the simulation-processing methodologies (not the CPU, but the reality models and processing
engine that process all inputs and transform to meaningful outputs in time). This is the domain of the computational physicist.
The modeling of real world systems requires the ability to formulate the numerical or analytical equations/solutions required
to describe the system being simulated to the level of fidelity required. Simulations will one day be the norm for the research
and development that develop complex and expensive systems which are required to interact within complex environments.


There is a tendency to insist on the affinity between these two sciences and still more an idea of utilizing them
together in a sort of collaboration. These ideas are recent, owing their rise essentially to the various applications and to the
consequent diffusion of ballistical knowledge among physicists or related technical people. From a scientific point of view, it
is enough to enunciate the idea of such a comparison to understand the possibilities opened up. Ballistics is a science
hundreds of years old and during its long life it has accumulated a valuable stock of theoretical and experimental materials.
This has now to be considered under a new aspect and placed at the disposal of a new science which has not much material of
its own because, in former times, there was not much interest in making physical experiments, computer simulation and also
because the knowledge of applied physics was mathematical rather than physical.
The subject of Ballistics draws upon several multi facets scientific disciplines for its developments. "Ballistics" is
meant that part of the subject, which deals with the application of physical concepts along with its physical models to the
solution of various ballistics problems. By its very definition, therefore, applied physics may fairly and closely related to
Ballistics. Let us discuss this a little more elaborately.
One of the various applications of applied physics is ballistics where the concepts and methods of physics are used to
develop and analyse the fields of ballistics. Further information on this application of applied physics can be obtained from
the armament related laboratories and organizations. The kind of information needed to explore the flight of a projectile
crosses several scientific disciplines. One is the classical dynamics of Isaac Newton; another is the field of fluid dynamics,
more specifically aerodynamics. Ballistics is the name of the specific scientific field that encompasses the areas of science
that we need to address.
Before doing so it might perhaps be helpful to consider a preliminary question, what should be the physical and its
mathematical characteristics and its computational aspects of a subject in order that it be treated as a proper branch of applied
physics? Is the mere fact that some physical and mathematical models (mathematical physics) in applied physics are
employed in the formulation and solution of some of its problems sufficient to label a subject as "applied physics"? Applied
Physics, for all its extensive employment of a variety of physical concepts along with mathematical models and its techniques
is still "Applied Physics" and one would not like to think of it as "Mathematics" of any kind. Similarly, classical mechanics,
starting from the problem of determining the trajectory of a given dynamical system, has led to the extensive mathematical
theory of differential equations and the associated theories transformations, etc. Many more such related examples can be
given in support of the criterion stated above, viz. a subject can qualify as a branch of applied physics only if it is capable of
generating problems which are of significance physically as well as mathematically. The thought of ballistics should
therefore be looked at from this point of view.
As in all applications of physical and its mathematical models to concrete problems, we can discern three stages of
developments. First, with non-availability of reliable experimental data, empirical methods suffice to yield crude
approximations to the desired solutions. Second, with the accumulation of more and more data and a better understanding of
the phenomena, more sophisticated physical and mathematical models with its computational methods are invoked leading
to more satisfactory solutions. Third, the subject has reached a level of maturity where it throws up genuine physical
problems, which may lead to new developments in applied physics itself.
This is well illustrated in the developments of applied physics. Indeed we can discern a strong parallel between


applied physics and ballistics in this regard. While both are intensely practical subjects, professional physicists have been
primarily responsible for the theoretical developments in both the subjects.
The principal problem of ballistics (whether internal or intermediate or external or terminal or wound or hydro) is
the solutions of a system of nonlinear differential equations, not merely a solution but a whole family of solutions
corresponding to a. range of values of the occurring parameters (propellant parameters, gun dimensions, M.V., ballistic
coefficient, Met. conditions etc.). The entire emphasis in the application of physical concepts and its mathematical models
has therefore been on the development of suitable approximation procedures for reducing the labour of computing such
extensive tables of solutions.
The advent of the high-speed computer has done away with the necessity of making such simplifying assumptions
and has even radically altered the structure of the numerical computation procedure itself. The computer can turn out in a very
short time thousands of solutions corresponding to a large variety of prescribed conditions. The ultimate in this line of
development is the guided missile equipped with guidance and control systems, which provide mid-course corrections again
with the help of the computer to keep the missile on a collision trajectory. It is clear that sophisticated mathematical methods
are no longer as relevant as before for providing ballistic solutions. The present day ballistician will look to the physicists
with computer background rather than to the mathematician to solve his problems.
A classic example, going back to the very beginnings of ballistics, is provided by Lagrange's ballistic problem, viz.
the determination of the pressure and density distribution in the moving mass of gas behind the shot in its motion along the
gun barrel. Various approximate solutions to this problem based on more or less arbitrary assumptions have been proposed
and used with appropriate empirical corrections, but the basic gas-dynamic problem of the coupled motion of the gas and
shot, taking into account the presence of the rarefaction wave, the reflected shock waves and the boundary layer at the wall, is
a physical problem which, we think, still awaits a complete theoretical analysis. A generalization of the above problem is
obtained if we replace the combustion process by a detonation process.
The motion of a supersonic projectile after it leaves the gun barrel gives rise to a similar flow problem, viz. the
analytical determination of the resulting flow pattern at the base of the projectile and a turbulent wake of nearly constant
cross-section, there is an attached curved front shock and a main rear shock which possibly arises by the running of several
shocks. The mathematical analysis of such a complex shock system and the derivation of an expression for the drag is a most
fascinating problem, which has been studied, by different scientists, physicists and others. The extension of this work to
include the effects of gravity and to the region of hypersonic velocities constitutes interesting physical and mathematical
The problem of computing the trajectory of a projectile is the central problem of external ballistics. The problem of
the dynamics of a spinning artillery shell has a famous counterpart in classical dynamics - the motion of the spinning top. The
motion relative to the C.G. of the spinning shell is a natural generalization of the spinning top problem, with the addition of
new forces and moments (of Magnus type), the forces and moments all being dependent on the instantaneous state of motion.
Various mathematical models like point mass, modified point mass, six-degree of freedom and universal models have been
developed by different academicians and ballisticians for such problems. The motion of the C.G. itself is far more
complicated than in the case of the spinning top moving on a smooth horizontal plane, it will be recalled that even in the latter
case the equations can be integrated only in terms of hyper elliptic functions. Apart from constructing approximate particular
solutions and deriving criteria of stability, it would be of interest to investigate general theorems.
Apart from determining the trajectory, the shape of the projectile itself has been the subject of mathematical
investigation. The problem is the determination of the optimum shape of projectile, viz. to determine that projectile contour
which minimizes the total drag. Many studies of shapes of minimum drag using the techniques of the calculus of variations
have been carried out and even extended to the region of hypersonic flight. The calculus of variations has also been used in the
study of the optimal design of multistage rockets. These new applications of a classical discipline give promise of a rich
harvest of results not only regarding the specific problems but also in regard to the mathematical technique itself.
Another interesting problem that may be mentioned, here is that of "Scabbing" of a plate by an impulsive load
applied on one face of the plate. The compression wave produced by the applied load when reflected at the opposite face gives
rise to a tension wave and as a result of the interaction of the two wave systems, the tensile stress along a certain surface within
the plate may fall below the critical yield stress of the material. A corresponding portion of the plate is then thrown off as a
'scab'. Different national and international experts have carried out the complete analysis of the elastic wave-system



generated in the plate by the applied load.

In case of forensic science, wound ballistics is very important in the criminal investigation of an offence that
includes the use of firearms. With wound ballistics one can study those emerging trends of the firearms use, nature of
firearms, different aspects of a firearm that causes destructions etc. In ballistics one can study those emerging trends of the
firearms use, nature of firearms, different aspects of a firearm that causes destructions etc. This helps a lot in the criminal
investigations as it finds out the type of firearm used by the offender, nature of its impact etc specially when the firearm in
question is not located or has been hidden by the offender, it also helps in the determination of the range of fire, from which
angle the trigger has been pulled, when was the shot fired, what kind of bullet was used etc. These things can also be found out
by studying the wounds of the target or the victim. The evidence related to firearms offences usually centers around fired
bullets and shells, which are examined by wound ballistics for the purpose of determining the type or make of the guns from
which they fired, or for the determining whether or not they were fired from a particular weapon. Apart from this, with
ballistics studies one can even improve the quality of the firearms by improving certain aspects of interior and terminal
Similarly, by referring to the subject of hydro or underwater ballistics, viz. the problem of the underwater trajectory
of a projectile accompanied by its cavitations bubble. The size and shape of the bubble are determined by the motion of the
projectile' and on the other hand the motion of the projectile is itself profoundly affected by the accompanying bubble. The
cavitations: bubble has already been the subject of some high-grade mathematical research. But the dynamical problems of
the projectile motion under water such as those of cavitational and fully wetted motion, conditions for ricochet etc. have
hardly been touched upon, except under drastically restrictive assumptions.

Possibilities and Limits of Ballistics Simulations

The outstanding development of the computing techniques allowed not only the description of the behavior of
physical systems in different conditions, but also the simulation of some physical and technical processes produced in special
(difficult) conditions. Because this method allows a considerable decrease of the expenses necessary to experiment various
technical models, as well as the simulation of some phenomena produced in inaccessible conditions, it presents a
considerable interest both from the didactic and technical point of view. Taking into account the future improvement of the
computing abilities by means of the parallel computers, there appeared also some special simulation techniques as the Local
Interaction Simulation Approach, intended especially to such computing techniques. The main difficulty met by the
simulation techniques refers to the appearance of some numerical phenomena: instability, divergence, and dispersion,
distortions, which lead sometimes to considerable inaccuracies of some obtained numerical simulations.
Among these numerical phenomena, we consider that the most misleading is that of pseudo convergence, because it leads to
apparently correct simulations (their shape is even qualitatively correct), but considerably wrong from the quantitative point
of view. That is why the present work considers that the numerical simulations must be studied in strong connection with the
basic results of the data processing of the experimental results. A correlation between computational aspects of applied
physics and ballistics is shown in Figure 8.

Figure 8: Correlations between Applied Physics and Ballistics



Applied Physics and Ballistics: Interdisciplinary Studies

Applied Physics is a relatively young discipline than ballistics. During the last few decades, researchers in the
academic field of applied physics and computing have brought together a varierty of scientific disciplines and
methodologies. The resulting science i.e. applied physics, offers a variety of unique ways of explaining phenomena, such as
computational physics or computational ballistics.
"Interdisciplinary studies" is an academic programme or process seeking to synthesize broad perspectives,
knowledge, skills, interconnections, and epistemology in an educational setting.
Although interdisciplinary and interdisciplinarity are frequently viewed as twentieth century terms, the concept has
historical antecedents, most notably Greek philosophy. Julie Thompson Klein attests that "the roots of the concepts lie in a
number of ideas that resonate through modern discourse-the ideas of a unified science, general knowledge, synthesis and the
integration of knowledge" to further understand their own material.
Interdisciplinary programmes sometimes arise from a shared conviction that the traditional disciplines are unable or
unwilling to address an important problem. For example, social science disciplines such as anthropology and sociology paid
little attention to the social analysis of technology throughout most of the twentieth century. As a result, many social scientists
with interests in technology have joined science and technology studies programs, which are typically staffed by scholars
drawn from numerous disciplines. They may also arise from new research developments, such as nanotechnology, which
cannot be addressed without combining the approaches of two or more disciplines. Examples include computational
ballistics, an amalgamation of applied physics and ballistics combining fluid dynamics with computer science.

Possibilities of Futuristic Emerging Principles and Trends in Applied Physics and Ballistics
Due to advancement of various multidisciplinary fields or interdisciplinary fields globally, it has been realized that
the understanding of the following subjects are primary need to study the various ballistic problems via applied physics for
futuristic armament programmes.
(a) Principles, Ideas and Methods in Mathematical Modelling (b) Fundamentals of Aero-ballistics design (c) Principles of
combustion processes (d) Design philosophy of long rod penetrators (e) Qualitative studies of effective lethality and (f)
Understanding of uncertainty models.
In addition to above, the understanding of following topics are necessary:
(a) Conservation of mass, energy (kinetic and heat);
(b) Conservation of linear and angular momentum;
(c) Continuum relations for the motion of fluid/gas;
(d) Equation of State of the fluid/gas;
Due to increasing of complex nature of information networking warfare strategies, the following emerging
principles are to be considered.
(a) Conventional Principles - It has been to increase weapon power to increase lethality;
(b) Peripheral Principle - Here one destroys population by enemy like groups (militants), economics and natural
(c) Select Target Principle - Here one defeats the enemy militarily and politically. Weapon systems are designed
which are precise to eliminate specific targets by zeroing on them and avoid collateral damage. The philosophies for
Select target principles are:

·Directed energy weapons;

·Precision automation;
·Energetic explosives;
·Super high-speed data processing;
·Electronic warfare equipments;
·Networking pace wars;



Futuristic Problems In Applied Physics And Ballistics

Internal Ballistics
(a) Mathematical modelling of gas discharge and heat exchanges in a gun barrel after the
projectile leaves the barrel;
(b) Estimation of force constant, density function and instabilities of gas burning in an
ETC gun;
(c) Development and mathematical modelling of insensitive high energy propellants for
advanced gun concepts;
(d) Development and mathematical models for high/ low pressure chambers;
(e) Identification of parameters affecting the accuracy in internal ballistics such as:
· chamber volume and shape;
· shape and location of ignition charge;
· shape and location of main charge;
· shot mass;

Launch Dynamics
(a) Development of mathematical models needed to estimate the structural parameters. (examples are extrusion
force, incision force, spin up force, normal force to the barrel wall, friction force and force due to curvature of the barrel,
forces acting on the subcomponents of the launcher).
(b) Development of mathematical models to study the sabot discard dynamics of the FSAPDS projectile fired from a
ETC gun, the sensitivity of motion due to change in muzzle velocity is quite significant and the effect due to it during
discarding phase is to be effectively parameterized.

Intermediate Ballistics
The state-of-the-art Computational Fluid Dynamics (CFD) models are the need to overcome the limitations to
increasing its accuracies levels in subsequent motions for implementation. Parametric estimation of aero-coefficients during
this flight is highly desirable.

External Ballistics
(a) Development of Modeling in FAE and FSAPDS (non-conventional aspects of ballistics);
(b) Free flight aspects in ICBM and space related bodies;
(c) Parametric estimation of aero-coefficients during flight trajectories;
(d) Projectile design along with booster for extended range;
(f) Development of Neural network models for trajectories modeling and range tables;
(g) Development of CFD for estimation of various aero-ballistic coefficients.
(h) High Speed Ballistic Computers for long rage range table complilation;
(i) Development of virtual reality modeling language for trajectory modeling and simulation;

Terminal Ballistics
(a) Multi point initiated chemical energy warheads;
(b) Flow parameters of a shape charge;
(c) Development of Hyper codes;
(d) Technology for HE warheads;
(e) Ceramic armours;
(f) Endgame Technologies;

If applied physics is a unique field separate from ballistics; if applied physics poses unique questions, comes up with
unique explanatory models, utilizes unique methods, and leads to unique realistic physical problems; then it is easy to argue
that those unique things are probably not answered in the philosophy of ballistics, hence, there is a need for field that
addresses those unique things- the philosophy of applied physics. The aim of the Master of Science like programme in



Applied Physics and Ballistics is to give a broad and solid introduction to the theoretical, experimental computational and
physical concepts that is the basis for much of today's high armament technology. The philosophy of applied physics is an
important part of understanding. We should know what is applied physics, why it is necessary, what are various methods of
applied physics. From the above discussions we conclude that the applications of Applied Physics and Ballistics demonstrate
the potential for physicists for industrial benefits from such degree that is unique to a physics graduate. Also, we opine that:
“Applied Physics and Ballistics" is and will continue to be an active and fruitful branch of armaments as two sides of a coin.
This fascinating but difficult subject presents an open challenge to the applied physicists and ballisticians.

The authors are grateful to Maj Gen Anoop Malhotra, Director, Proof and Experimental Establishment (PXE),
Balasore, for his consent to publish this paper.

1. Text of Ballistics and Gunnery-I & II, Her Majesty's Stationary Office, London,1987
2. McCoy, R.L. Modern Exterior Ballistics. Atglen, PA, Schiffer Military History, 1999.
3. Rinker, Robert A., Understanding Firearm Ballistics, Basic to Advanced Ballistics Simplified,
Illustrated and Explained, 1999, Mulberry House Publishing, Corydon, IN, ISBN 0964559846.
4. Crosby, Alfred W., Throwing Fire: Projectile Technology Through History, Cambridge
University Press, UK, 2002.
5. Driels, M. R., Weaponeering: Conventional Weapon System Effectiveness, 1st Edition, AIAA
Series, Inc, Virginia, 2004.
6. Przemieniecki, J.S., Mathematical Methods in Defense Analyses, AIAA Series, Virginia, 2000.
7. An Analysis of Engagement Algorithms for Real-Time Weapon Effects- John Mann, et al,
JDMS, Volume 3,Issue 3, July 2006, Pages 189-201, The Society for Modelling and Simulation
8. Computational Ballistics-III, WIT Transactions on Modelling and Simulation, Vol-45, © 2007
WIT Press, , Southampton, Boston, 2007.
9. Realization of a virtual experiment system of gun- Yi Hang, Xiaofeng Wang, Guolai and
Yunsheng Chen, World Journal of Modelling and Simulation, Vol.2 (2006), No. 4, pp-255-260.
10. Physics for future Presidents- Richard A. Muller (muller.lbl.gov/teaching/Physics10/.../PffP-00-TOC-
11. Hatcher, Julian S., Hatcher's Notebook, 3rd edition, June 1962, Stackpole, ISBN 0811707954
12. Lowry, E. D., Exterior Ballistics of Small Arms Projectiles, Research Dept., Winchester-Western
Division, Olin Mathieson Chemical Corporation, January 1965, ASIN B0007FPINM
13. Mann, F. W., The Bullet's Flight From Powder to Target-- Ballistics of Small Arms, 1980,
Wolfe Pub Co; ISBN 0935632042.
14. McShane, Kelley and Reno, Exterior Ballistics, University of Denver Press, 1953, ASIN:
15. Hayes, T.J. Elements of Ordnance. New York; John Wiley and Sons, 1938.
16. Ballistics- DE Carlucci & SS Jacobson, CRC Press, 2008.
17. Handbook of Ballistics-C Cranz, His Majesty's Stationary Office, London, 1921.
18. Ballistics: An Overview and Advances-A Lecture notes, Dr. KC Sharma, Ex-Prof, DIAT, Pune, 2005.
19. Gun Shot Wounds- VJM Di Maio, CRC Press, NY1999.
20. Handbook of Physics, McGraw Hill Book Company, NY, 1967.
21. Military Ballistics: A Basic Manual-G.M.Moss,D.W.Leeming and C.L.Farrar, Brassy's Ltd series,UK,
London,Vol-I, 1995.
22. Applied Modelling and Simulation: An Integrated Approach to Development and Operation-Cloud and
Rainey, The McGraw-Hill Companies, Inc, 1998.




Raju Venkatanarayana, S Varadarajan
Electronics & Radar Development Establishment (LRDE), Bangalore – 560 093

The advent of long range weapon system and mechanization of land forces have extended the area of operation
much beyond the visual range. Deployment of electronic surveillance devices in the battlefield will serve as a force
multiplier to enhance the combat potential of our forces and optimize the effectiveness of our weapon systems. With the
advent of modern electronic means, enemy guns can be located with required accuracy and can be silenced or neutralized by
means of direction of own artillery fire.
Weapon Locating Radar (WLR) being phased array radar, has the capability of Electronic Beam Steering in both Azimuth
and Elevation, tracking the entire projectile trajectory.
The challenges faced by Radar are:
1. Detection and tracking of projectile with very low radar cross section in the background of other aerial objects, weather
and ground clutter.
2. Estimation of launch / impact point of projectiles fired in low and high angles accurately within a very short time from
detection of projectile in radar coverage.
This paper discusses how these objectives are met by the radar and also about how this radar will be useful for improving
armament technology. It will aid the designers of guns to refine the trajectory models and prepare highly accurate range tables
of weapons.
The Field Artillery Weapon Locating Radar provides fast, automatic (without operator intervention) and accurate
locations of all Guns, Mortars and Rockets firing within its effective zone of coverage and will be able to handle
simultaneously multiple projectiles fired from different weapons at different locations. The system is also capable of
adjusting the fire of our own Artillery Weapon. Thus the equipment has two major goals to achieve
·Weapon Location for enemy Artillery
·Adjustment of Fire for our own Artillery
Optionally the system can also provide impact point of the enemy weapon in Weapon Location Mode and the launch
point of the friendly weapon in Adjustment of Fire mode with degraded accuracies. However the Weapon Location
and the Adjustment of Fire Modes are mutually exclusive and cannot be performed simultaneously.
The main functions to be performed by equipment in order to achieve the defined objectives are :
·Classification as projectile / non-projectile targets
·Trajectory Estimation & Launch/Impact Point Estimation
·Height Correction for LP & IP
·Storing and displaying the trajectory / LP & IP data
·Communication with higher echelons (ACCCS/BSS)
The system operation as enemy weapon locating and as friendly fire adjustment is shown in fig. 1 & 2 respectively.


WLR configuration includes mainly two assemblies mounted on two different TATRA VVL vehicles, named as:
· Radar System Vehicle
·Power System Vehicle
Besides these two main assemblies, the following two optional elements/assemblies also form the composition of
WLR System:
·Radar Target Simulator
·Remote Radar Display (RRD)



Figure 1 Operation of Radar in Weapon Locating Mode

Figure 1 Operation of Radar in Fire Direction Mode



The Radar vehicle assembly comprises of a high mobility TATRA vehicle mounted with all the Radar subsystems.
The detailed block diagram of Radar and its functionality is explained in next section.
The WLR Power System Vehicle includes two (for redundancy) 75KVA/60KW, 50 Hz AC Diesel Generators along
with power conditioning equipment mounted inside an acoustic shelter.
System Built In Test Equipment (BITE) facility is provided mainly for System maintenance & functional testing.
Besides it can also be used for day-to-day system performance checks. It consists of a hundred feet mast with small radiating
horn and support-ropes and a BITE-rack containing associated electronics. All these items are mounted on a small-
containerized vehicle along with associated power and signal cables.
Functionally the Remote Radar display is similar to the Primary Radar Display. The system consists of a portable
rugged computer preinstalled with the operating system and associated application software. It is connected to Radar via a
300 meters long LAN (optical) cable.

Figure Configuration of WLR system

The WLR radiates a search fence which is a set of beams radiated at specified azimuths and elevations decided based
on the terrain at deployed site either automatically using digital map or manually using some other means and entering the
data at radar console. The Radar Controller selects the next beam to be scheduled with user specified waveform and programs
Receiver-Exciter, Signal Processor, Beam Steering Unit & Analog Display subsystems with required beam parameters. The
Radar Controller also computes the phase gradients for the given azimuth and elevation angle incorporating pitch/roll
corrections at the specified frequency. The Beam steering Unit computes the phase of each PCM using the phase gradients
and the array geometry, and loads the corresponding phase commands for each PCM in accordance with the temperature at
different points within array. Once the Radar Controller issues the Switch Beam command at scheduled time, the Antenna
looks into the desired direction the Exciter generates the Radar Waveform which is pre-amplified by a Solid State Driver to a
suitable level before applying to TWT amplifier in the Transmitter. The Transmitter output illuminates the antenna through
feed horn and the Antenna concentrates the RF power into the direction where it is looking. Any object present in that
direction within Antenna beam widths reflects a part of the transmitted energy back towards Radar and a part of the reflected
energy is intercepted by the Antenna.



BSU Array


Front En d Sensor


ADSB Synthetic


Remote Display

Fig. 4 : Block Diagram of WLR

This received energy is divided into three mono-pulse channels after passing through the multimode horn & mono-
pulse comparator which in turn is processed by the three identical gain-phase matched receivers. The Front-end Receiver
brings down the RF echo signals to IF level (C-Band to UHF to IF) and the IF receiver performs the matched filtering and I-Q
demodulation to give the Sum I-Q, Delta Azimuth I-Q and Delta Elevation I-Q signals to Signal Processor subsystem for
further processing. The Signal Processor subsystem digitizes these six analog inputs and passes it through MTI channels
(programmable 2 or 3 or 4 pulse/AMTI cancellers) before integrating (Linear/Binary) all the pulses of the dwell and going
for CFAR (Cell averaging, GO, SO) processing. For Track Dwells normalization (D/S) of difference channels (Delta
Azimuth and Delta Elevation) is done and the normalized difference signals (D/S) are smoothed over the number of pulses in
the “dwell” before being used to extract the off axis angular offset. Operation is performed in sine space co-ordinates based
on scan invariance property of the D/S function. Finally for every Radar Dwell, the output of the signal processor is collated
from various channels and Radar Dwell reports are prepared and sent to the Radar Computer. The Radar finds its own north
and pitch and roll of deployment are measured automatically in addition to its own location.
At Radar Computer these reports are processed by Radar Data Processing CSCI for tracking function. The tracker
uses a seven state Extended Kalman Filter with drag (unknown parameter to be estimated) as the seventh state. For all the
targets classified as projectile, the Weapon Locating Processor uses the stored measurement data of the tracking phase to
compute the drag and finally the Launch/Impact point of the projectile. A projectile model is run backwards till the initial
measurements captured, and then the Point Mass model is used for extrapolating the trajectory till the altitude of the projectile
matches that of the Radar.
To facilitate the online monitoring of the Radar performance, Analog Display is provided which displays the
scanned area with Raw/CFAR video super imposed on the screen in different form viz. Sector PPI, B-Scope, RH-Scope & E-
Scope. Target data after filtering and extrapolation is displayed along with Launch/Impact points on synthetic display which
supports different display modes including 3-D display and 2-D Map overlay for sector PPI mode. Height correction on the
extrapolated LP/IP is performed by synthetic display using the 3-D Map data. Synthetic Display also acts as the Human
Machine Interface for setting up the Radar and monitoring the Radar subsystems health status.
To validate the complete signal and data chain of the Radar, the ADSB subsystem generates simulated multiple
target triggers moving radially with programmable parameters like start range, speed, Doppler, number of targets etc. These



triggers are up converted at exciter and is fed to either RF front end (called as RF BITE) or fed to a horn in the bite mast (called
as field BITE) from where it is received through the antenna.


It is clear from the previous paragraphs that the prime function of WLR is to detect and track the projectiles. In its
role as WLR, the measured data is used for computing the ballistics of the projectile to get the launch/impact point.
On the other side, in Armaments we develop, the projectiles to strike deep behind the enemy's frontline from a standoff range
with the main focus on 'surgical strike' (accuracy with consistency) capability.
During the development phase of such projectiles the WLR can play a vital role in characterising the projectile
behaviour wherein one can have the complete trajectory as actually traversed by the projectile at desired data rate (up to 20
Hz) from close to the barrel nose and till impact. Rocket Trajectories from boost phase onwards can also be captured using
WLR. This will help in understanding the power phase characteristics of different rockets.
In addition, effect of MET parameters on trajectory motion can also be studied in depth. The parameters recorded
by the radar are positional measurements and filtered positional, velocity & drag parameters. Hence in a nut shell, the
valuable information provided by the radar will be very useful for designing the ballistic trajectory models for various
weapon calibres.
The fire direction mode of operation can be very useful for establishing the impact point of the projectile and
subsequently applying the necessary corrections to home our ammunition towards the desired location very accurately.

The radar is capable of tracking trajectories up to 15-20 Km altitude and 30 Km of range. If the terrain permits it is
possible to capture the full trajectory from almost launch to impact point. If full trajectory capturing is not desired, facility is
available to capture the desired segment. The trajectories can be viewed in various display modes namely

1. PPI – Plan position indicator

2. RHI – Range Vs Height
3. Time Vs Height
4. 3D plot
Snapshots of actual projectile trajectories of various calibres as captured by the radar are shown in fig. 5


The information of trajectory available from radar are measured positions (Range, Azimuth and Elevation) with
respect to radar at every 100 milliseconds. This can be given at every 50 milliseconds if the object of interest is a single
trajectory at a time. In addition filtered trajectory data in terms of position, velocity and drag estimates (X, Y, Z, VX, VY, VZ,
drag) are also available. There is a facility for logging the trajectory data in real time which can be extracted for all the
information mentioned above. The logged data can be replayed. In improving the armament technology this feature of the
radar can be exploited to the maximum. Snapshots of data captured are given below.

Measurement data available :

Range Azimuth Elevation Time SNR
11.85 110.56 10.39 3968.674 37.886

11.83 110.7 10.61 3968.878 38.742

11.82 110.71 10.78 3968.974 39.013

11.82 110.8 10.94 3969.082 41.537



155mm Bofors Shell (Range Vs Height)

GRADBM 21 Rocket (Time Vs Height)

3D – Plot
105mm Shell - (Time Vs Height)

Pinaka Rocket (Range Vs Height)

81mm mortar (Time Vs Height )

Fig.5 : Actual Projectile Trajectories of various calibers as captured by RADAR



Filtered data available :

X Y Z VX VY VZ Drag Time

10830.295 -4072.97 2279.67 -125.18 41.08 279.38 2 4055.626

10818.638 -4069.07 2308.73 -123.77 40.67 277.28 2.00 4055.734

10806.912 -4064.11 2333.24 -123.52 41.52 274.33 2.000135 4055.83

10794.409 -4059.51 2357.41 -124.05 42.09 271.03 2.00035 4055.926

The radar has been evaluated for different kind of projectiles at different firing angles and firing ranges. The logged
data of tracking (both raw and filtered) at data rate of better than 10Hz rates with high level of accuracy are expected to aid /
refine modeling of the trajectories, which will be helpful in characterizing the projectile behavior under various conditions

1.Introduction to Radar System by Skolnik.I.Merrill
2.Radar Handbook by Skolnik.I.Merrill
3.Radar – Principles, Technology, Apllications by Byron edde
4.Project Documents




LK Gite, M Padmanabhan, DK Joshi
Armament Research & Development Establishment, Pune- 411021
Weapon Locating Radars are one of the primary means of locating enemy indirect fire systems. The functionalities
include generating artillery target intelligence and fire missions, registering and adjusting friendly artillery fires and
validating the location of friendly fires. They play a critical role in battlefield by contributing effectively in the fire finder
mission of targeting process methodology to aid in decision making.
Parameter estimation for non-linear motion model of a foreign ballistic target involves higher degree of uncertainty
due to noise in system and measurements. Accurate determination of the transient parameters will aid in improved tracking
and hence measurements and vice-versa. The accuracy in estimate dictates prediction probability during forward and
backward extrapolation. A study of parameter identification and estimation carried out for the bimodal operational
environment of weapon locating radar is reported here. Various parameter estimation techniques were explored and
evaluated on the basis of their appropriateness to system dynamics, accuracy of prediction, rate of convergence and
The investigation includes approaches based on drag coefficient and ballistic coefficient. An augmented Extended
Kalman Filter with Dual Traversal, coupled with a MET scheme led to accurate ballistic parameter estimation. Multiple
estimates computed are evaluated for degree of appropriateness, using correlation filter on measured and computed
trajectories. The estimate with highest correlation peak qualifies for predicting forward impact point and backward launch
point. Results showed that prediction probabilities obtained were 90% for high angle trajectories and 80% for low angle
trajectories. Possible enhancements to the algorithm using particle filters are likely to minimize the quantitative
requirement on measurements and might offer better estimates in near real time.

Artillery indirect fire systems offer combat and cover operations in battlefield scenario. Weapons locating radars
(WLR) are one of the primary means of locating enemy indirect fire systems. The functionalities of WLRs include
generating artillery target intelligence and fire missions, registering and adjusting friendly artillery fires and validating the
location of friendly fires. Indirect fire is achieved by deployment of artillery weapon systems like Mortars, Field guns and
Rocket launcher systems. The projectiles fired from these systems are characteristically distinct from other objects due to
their negligibly smaller radar signatures and ubiquitous non-stationary noise patterns. WLR Plays a critical role in battle
field by contributing effectively in the fire finder mission of targeting process. The radar signature qualifies for the detection
of the projectile in space. The radar measurements include the spatial coordinates of the present position of the object in
space , the time stamp and characteristic velocity derivatives. The real time data is processed for predicting the future course
of motion of the projectile. Successive data obtained over a period of time constitute a time series data of the projectile
trajectory. Further processing of the time-series data using an underlying dynamic representation of the projectile dynamics,
followed by parameter estimation and backward or forward extrapolation leads to launch point or impact point.


Processing of the data involves simulation of the dynamical phenomena using a mathematical model and
estimation of appropriate parameter characterizing the particular situation. Simulation of real-life phenomena through
mathematical models is important for analysis of system behavior, optimization and simulation of extreme situations. These
mathematical models are often nonlinear, non-Gaussian and non-stationary. In most cases ubiquitous and noisy
measurements of the system at discrete time interval over limited duration are the only available information on system
observation. The model driven inverse problem which is generally insolvable can be reduced to numerical data fitting of



experimental data on relaxing the problem using iterative methods like Least Mean Square Estimate( LMSE ). LMSE
approach is weak for systems of physical world as it cannot tolerate noise. If the system is very noisy then the resulting fit
may be far from the real model. Further more if the objective function has many local minima the optimization process will
be very sensitive to initial values. Standard least squares algorithms compute local optimum, whereas the real interest
surrounds in finding a global optimum, for which adoption by LMSE would require the examination of the entire feasible
set. A possible approach to overcoming ill-posed ness lies in regularization methods, yielding smooth solutions. For
nonlinear inverse problems, these techniques are based on global optimization. However, it is in general difficult to provide
indicators of reliability for solutions. A possible approach will be the data driven inverse problem, which considers
measurement data errors with some prior error bounds and a general model fairly representing the system. Then the problem
simplifies to finding the parameters in the which characterize the system. Uncertainties develop and grow in greater strength
due to factors like inadequacies prevailing in obtaining accurate measurement data, assumptions and approximations
considered in modeling the system and impreciseness in understanding, handling and processing of the data and the model.
Generally for nonlinear parameter estimation problem, no exact solution can be obtained, or the solution is infinite-
dimensional, and hence various numerical approximation methods come in to address the intractability. In this case motion
model based on force dynamics and ballistic parameter estimation is adopted.

Mathematical Models
Mathematically the problem can be represented as a discrete time system. State-space and frequency domain
approaches are two major approaches for discrete- time system modeling. The requirement is to adopt a discrete-time
filtering model so as to make one step prediction and parameter estimation. The Generic stochastic filtering problem in a
dynamic state-space form can be given by

xt = f (t, xt, ut, dt) … state equation

yt = g(t, xt, ut, vt) … measurement equation

and xt represents the state vector, yt is the measurement vector, ut represents the system input vector (as driving
force) in a controlled environment; f : RNx _? RNx and g : RNx _? RNy are two vector valued functions, which are
potentially time-varying; dt and vt represent the process (dynamical) noise and measurement noise respectively, with
appropriate dimensions. As the radar measurement data is a time-series data a more practical discrete time form is ideal for
the situation.
Discrete-time filtering problem in a dynamic state-space form

xn+1 = f (xn, dn) … state equation

yn = g(xn, vn) … measurement equation

where xn, yn represent the measurement vectors and dn and vn are white noise random sequences with unknown
statistics in the discrete-time domain. The state equation characterizes the state transition probability p(xn+1|xn), whereas
the measurement equation describes the probability p(yn|xn) which is further related to the measurement noise model.
These equations reduces to the following special case on considering a linear Gaussian dynamical system :

xn+1 = Fn+1,nxn + dn
yn = Gnxn + vn

the analytic filtering solution is given by the Kalman filter, in which the sufficient statistics of mean and state-error
correlation matrix are calculated and propagated. In the above equations , Fn+1,n, Gn are called transition matrix and
measurement matrix, respectively.

Problem Nature
The filtering problem which is a conditional posterior distribution (density) estimation problem, which is known to
be stochastically ill posed especially in high-dimensional space. Filtering problem is an Inverse problem, in the sense that
given a set of collected yn at discrete time steps (hence y0:n), provided f and g are known, one needs to find the optimal or
suboptimal ˆxn. This inversion learning problem is one-to-many since the mapping from output to input space is generally
non-unique. In this context, filtering problem is an ill-posed problem. Since the ubiquitous presence of the unknown noise
corrupts the state and measurement equations given limited noisy observations, the solution is non-unique.



Some Important Terminologies

Kalman Filtering
Kalman Filter consists of an iterative prediction-correction process. In the prediction step, the time update is taken
where the one-step ahead prediction of observation is calculated; in the correction step, the measurement update is taken
where the correction to the estimate of current state is calculated. In a stationary situation, the matrices An, Bn, Cn, Dn ,
Kalman filter is precisely the Wiener filter for stationary least-squares smoothing. In other words, Kalman filter is a time-
variant Wiener filter. Kalman filter is also optimal in the sense that it is unbiased E[ˆxn] = E[xn] and is a minimum variance
estimate. In Bayesian interpretation Kalman filter reduces to a MAP solution. The derivation is somehow similar to the ML
solution. For simplicity, it can be assumed that the dynamic and measurement noises are both Gaussian distributed with zero
mean and constant covariance. In order to use the Kalman filter to estimate the internal state of a process given only a
sequence of noisy observations, one must model the process in accordance with the framework of the Kalman filter. This
means specifying the matrices Fk, Hk, Qk, Rk, and sometimes Bk for each time-step k as described below.

Model underlying the Kalman filter. Circles are vectors, squares are matrices, and stars represent Gaussian noise
with the associated covariance matrix at the lower right.
The Kalman filter model assumes the true state at time k is evolved from the state at (k - 1) according to

Fk is the state transition model which is applied to the previous state xk-1;
Bk is the control-input model which is applied to the control vector uk;
wk is the process noise which is assumed to be drawn from a zero mean multivariate normal distribution
with covariance Qk.

At time k an observation (or measurement) zk of the true state xk is made according to

where Hk is the observation model which maps the true state space into the observed space and vk is the observation
noise which is assumed to be zero mean Gaussian white noise with covariance Rk.

The initial state, and the noise vectors at each step {x0, w1, ..., wk, v1 ... vk} are all assumed to be mutually
independent.Many real dynamical systems do not exactly fit this model; however, because the Kalman filter is designed to
operate in the presence of noise, an approximate fit is often good enough for the filter to be very useful. Variations on the
Kalman filter described below allow richer and more sophisticated models.

Filtering is an operation that involves the extraction of information about a quantity of interest at time t by using
data measured up to and including t.
Prediction is an a priori form of estimation. Its aim is to derive information about what the quantity of interest will
be like at some time t + ô in the future (ô > 0) by using data measured up to and including time t. Unless specified otherwise,
prediction is referred to one-step ahead prediction in this paper.



Smoothing is an a posteriori form of estimation in that data measured after the time of interest are used for the
estimation. Specifically, the smoothed estimate at time t_ is obtained by using data measured over the interval [0, t], where
t_ < t.

Sufficient Statistics is referred to a collection of quantities which uniquely determine a probability density in its

Well-posed problem is one which satisfies three conditions: existence, uniqueness and stability, otherwise it is said
to be ill posed.

System Identification
System identification is an statistical inference problems. Sometimes, identification is also meant as filtering in
stochastic control realm, especially with a driving-force as input. However, the measurement equation can admit the
feedback of previous output, i.e. yn = g(xn, yn-1, vn). identification is often more concerned about the parameter estimation
problem instead of state estimation..

Filtering can be viewed as a sequential linear/nonlinear regression problem if state equation reduces to a random
walk. But, regression differs from filtering in the following sense: Regression is aimed to find a deterministic mapping
between the input and output given a finite number of observation pairs {xi, yi}_ i=1, which is usually off-line; whereas
filtering is aimed to sequentially infer the signal or state process given some observations by assuming the knowledge of the
state and measurement models.

Missing Data Problem is concerned about probabilistic inference or model fitting using limited data. Statistical
approaches (e.g. EM algorithm, data augmentation) are used to help this goal by assuming auxiliary missing variables
(unobserved data) with tractable (on-line or off-line) inference.

Density Estimation is a dependency estimation problem like filtering. Generally, filtering is nothing but to learn
the conditional probability distribution. However, density estimation is more difficult in the sense that it doesn't have any
prior knowledge on the data and it usually works directly on the state. Most of density estimation techniques are off-line.

Nonlinear Dynamic Reconstruction arise from physical phenomena in the real world. Given some limited
observations (possibly not continuously or evenly recorded), it is concerned about inferring the physically meaningful state
information. In this sense, it is very similar to the filtering problem. However, it is much more difficult than the filtering
problem in that the nonlinear dynamics involving f is totally unknown (usually assuming a nonparametric model to estimate)
and potentially complex (e.g. chaotic), and the prior knowledge of state equation is very limited, and thereby severely ill-
posed. Likewise, dynamic reconstruction allows off-line estimation.

Criteria For Optimality are a vital tool in evaluation of suitability of a filtering algorithms. The following are few
of them which have found widespread utility. Minimum mean-squared error (MMSE), which can be defined in terms of
prediction or filtering error (or equivalently the trace of state-error covariance) which is aimed to find the conditional mean.
Maximum a posteriori (MAP) is aimed to find the mode of posterior probability, which is equal to minimizing a loss
function. Maximum likelihood(ML) reduces to a special case of MAP where the priori is neglected. Minimax criteria is to
find the median. Minimum conditional inaccuracy, Minimum conditional KL divergence, and Minimum free energy are also
can be used for error estimation of prediction and filtering.

There are many statistical approaches to algorithmic development and performance. Posterior (Bayes)
probabilities is an optimal form of information for decision making processes. Under some characteristic circumstances
these probabilities and related quantities can be estimated using probability and Maximum Likelihood theories. Algorithms
which deliver information which is of practical use only need to be used. The first rule of any algorithmic development
process is the specification of the required information that one wishes to obtain from the measurement data.
It's generally expected that once this information of choice has been obtained it will be used for further processing
and will act as the basis for subsequent action resulting from a decision based on this data. The decision process requires
complex evaluation of data. An optimal way of utilizing the data is always the question. The successful out come will be



affected in two ways, first by a failure in the algorithmic process and the second an error in the data. As a consequence, an
algorithm must not only deliver an estimate of the measured data but also estimates of data reliability and estimate
reliability. Anything less than this information will make subsequent decision formation unreliable and inaccurate. And the
algorithm could then never form part of a practical system.
The most common form of such information is error covariance measures. computation and manipulation of these
quantities is very essential The most direct information regarding the successful outcome of a particular decision is the
posterior (Bayes) probability. This is defined as the probability that a particular event will be true given a particular
observation. Knowledge of Bayes probabilities of outcomes on a given a set of alternative states allows a direct assessment to
attempt alternative actions. Probability theory is regarded as the only self-consistent computational framework for all data
analysis and decision making.
It is therefore not surprising that it forms the basis of all statistical analysis processes. Compromises have to be
made, generally in assumptions about the statistical form of the processed data, and it is the adequacy of these compromises
which will determine the success or failure of a particular algorithm. Clearly, therefore, understanding these assumptions
and compromises is an important part of algorithmic development. We can conclude that algorithms which will work best on
a particular application will be those which model most closely the underlying statistics of the measurement process and
correctly propagate the effects of these through the output of the algorithm.
Algorithmic robustness goes hand in hand with getting this process correct and is something that can rarely be
compromised in the interest of computational speed. Two important concepts that has been widely used for parameter
estimation and target classification are bayes theorem and Maximum likelihood estimator.

Bayes Theorem
The basic foundation of probability theory follows from the following intuitive definition of conditional
P(AB) = P(A|B)P(B)
In this definition events A and B are simultaneous an have no (explicit) temporal order we can write
P(AB) = P(BA) = P(B|A)P(A)
This leads us to a common form of Bayes Theory,
The equation : P(B|A) = P(A|B)P(B)/P(A)
which allows us to compute the probability of the interpretation of an event in terms of observations and prior

Maximum Likelihood
Starting with Bayes theorem one can extend the joint probability equation to three and more events
P(ABC) = P(A|BC)P(B|C)P(C)
For n events with probabilities computed assuming a particular interpretation of the data Y (for example a model)
P(X0,X1,X2, ...Xn|Y )P(Y ) = P(X0|X1,X2, ...Xn, Y )P(X1|X2, ...XnY )...P(Xn|Y )P(Y )
Maximum Likelihood statistics involves the identification of the event Y which maximises such a probability. In
the absence of any other information the prior probability P(Y) is assumed to be constant for all Y. For large numbers of
variables this is an impractical method for probability estimation. Even if the events were simple binary variables there are
clearly an exponential number of possible values for even the first term in P(XY ) requiring a prohibitive amount of data

Parameter Estimation
Often describing the cost function for an algorithm is not the only part of the algorithmic solution. We also need a
way of searching a space of possible models in order to find one that optimises this function. If we take the view that the
parameters (or variables) in our system model correspond to unique, perhaps even physical, meaningful values then the
algorithmic goal becomes that of parameter estimation (along with the associated covariances so that we can make practical
use of the result). There are many parameter estimation techniques, the commonest being those that seek to optimise a well
behaved function E(a) subject to change in the model parameters a. A common approach here assumes that the functional
form of E(a) is approximately quadratic.
This type of a solution suggests an iterative update scheme for estimates of a close to a quadratic minimum and this
idea forms the basis of 'quasi-Newton' methods which provide iterative methods for approximating. The popular Kalman
filter is effectively an iterative form of such an algorithm which provides a mechanism for optimal combination of new data
into the current estimate of the parameters. These methods, and the associated 'conjugate gradient' scheme require accurate
derivative information which may not always be available. Under these circumstances methods which require only function



evaluations such as the 'downhill simplex' method may be used. This method can be recommended as a starting point for
getting results quickly from a particular minimisation problem. Sometimes the minimisation function is not only badly
behaved but so too is the parameter space a. Under these circumstances the whole concept of functional optimisation needs to
be reconsidered. At this point algorithmic research switches from an evaluation of the best optimisation functions to
designing new optimisation methods. Such approaches are best typified by algorithms such as simulated annealing and
genetic algorithms. The main difference between these approaches is that the former operates on what are effectively single
point trials of the optimisation function while the latter operates with a set of interacting solutions. These interactions provide
a very useful means for controlling the range of locations searched by the algorithm which is lacking from simulated
annealing. These algorithms can cope not only with multiple, discontinuous functions and parameter spaces but also non-
stationary (even stochastic) evaluation functions. These properties are not however, obtained freely, the first obvious cost is
the computation requirement. The success of simulated annealing is heavily influence by the choice of a problem specific
annealing schedule. Genetic algorithms require that the parameters be cast in an appropriate representation (generally
binary), suitable for the processes of cross-over and mutation which will drive the search efficiently through variants of the
parameters in its search for the optimum. For these reasons these techniques are often looked upon as methods for solving
specific long standing problems rather than automatic algorithmic solution to a class of problem. The main point to be made
on the process of function minimisation is that the choice of minimization method (if adequate for the task) will not alter the
performance of an overall algorithm. The main effect on algorithmic performance in the final evaluation lies with adequate
definition of the cost function itself, which must have its roots in probability theory. If a set of data are insufficient to
determine a set of parameters with sufficient accuracy for the task in hand then changing the method of optimisation will not
improve the situation.

Ballistic Parameter Estimation and Point prediction

Parameter estimation for non-linear motion model of a foreign ballistic target involves higher degree of uncertainty
due to noise in system and measurements. An approach involving augmented Extended Kalman Filter(EKF) for kernel
smoothing enabling the prediction process and an appropriate trajectory model guiding the correction was adopted. The
enhanced accuracy was noticeable on fine tuning. Which has led to improved state estimates and measurement updates.
Accurate determination of the transient parameters by guiding Point Mass Model Trajectory Computation (PMMTC) has
aided in improved tracking and hence measurements and vice-versa. Seven state EKF with 3 -spatial coordinates, 3 velocity
components ,1- unknown state transition matrix, and a force dynamics matrix corresponding to a simple point mass rigid
body motion was considered. Difficulties in convergence and instabilities were noticed during coefficient of drag (CD) based
approach due to the characteristic nature of CD profile. Without loss of generality it can be assumed that the CD profile
uniquely characterises the projectile trajectory. The non-linear nature of which imposes inaccuracies in estimates for
trajectory segments which are not fairly close to the CD values of the tracked segment. Adding to the above the unknowns
like projectile geometry and mass parameters which are not decipherable from the negligible radar signature dominant with
An alternative approach based on ballistic and a lumped parameter( grouping all unknowns) model, which
estimates ballistic coefficient (C0 ), an in-variant parameter fairly approximating the projectile trajectory was employed. The
estimate is a factor of adjustment to a more generalized aero dynamic behaviour representing the air resistance experienced
by any projectile in motion. Ballistic coefficient (C0 ), being a constant for a trajectory, prediction of forward or rearward
trajectory based on radar track segment estimate more closely represents the case. The accuracy in estimate dictates
prediction probability during forward and backward extrapolation. Further improvements to the estimates to ballistic
coefficient (C0 ) were obtained by dual Traversal in the forward and rearward direction to ensure absorption of any variances
due to initial state, parameter assumptions which are likely to influence the convergence, stability, optimality and accuracy
considerations. The C0 estimated is corresponding to non-standard ( influenced by meteorological and other random
variations affecting the projectile path) trajectory segment, engulfing the variations. Multiple approaches based on
meteorological (MET) extraction from trajectory segment and inclusion of them during extrapolation were explored for
accuracy in point (Launch point, Impact point) predictions. An appropriate MET scheme was identified to yield relatively
closer estimates was implemented. Simulation and field trials revealed the dependability of point prediction, estimate
probability on estimate's deviation from the true value. Accuracy in estimation of C0 directly enhances the accuracy of point
prediction. Also the range of values assumed by the different projectile classes like Mortar bombs, Artillery shells and
Rockets was used for classification of tracked projectiles in to mortar bombs, artillery shells and rockets.
Further enhancements to accuracy was achieved by constructing multiple segments of the time-series data and
adopting the processing repeatedly to the track segments. Multiple estimates computed are evaluated for degree of
appropriateness, using correlation filter on measured and computed trajectories. The estimate with highest correlation peak



was adopted for predicting forward impact point and backward launch point.


The algorithm was validated using simulation and field trials. The results showed prediction probabilities being
close to 90% and 80% for high angle trajectories and low angle trajectories respectively. Some results and cases illustrating
the algorithmic effectiveness are presented in the following sections.

Launch and Impact point estimation from tracked data

The WLR supposed to track ascending proton of trajectory. The maximum track duration is ten seconds. Using this
tacked data the launch and impact point are estimated by point mass dynamics based Extended Kalman Filter.

Launch Point Impact Point Tracked Segment


Height, m



0 5000 10000 15000 20000 25000
Range, m

Ballistics coefficient Co from tracked data using EKF

To estimate the launch point the main difficulty was how to obtain the drag pattern. Here the grouping all
unknowns in one called ballistic coefficient (C0 ) of that trajectory is estimated from tracked data as seventh state. Following
fig shows the estimation of C0 as seven state in EKF conversing to actual C0 with in 6 to 8 seconds

Estimated Co Actual Co


0 2 4 6 8 10 12
Time, sec



Actual Launch point and estimated launch point in simulation

The simulation of typical run. Twelve trajectories on constant launch point. The 9 times launch point is predicted
within desired accuracy.

Simulation Summary
For every weapon the combination of charge (muzzle velocity), fire angle, azimuth and MET variable. Fire angle 10
to 30 deg considered as low angle and above 30 deg high angle. The nature of trajectory that is incoming or outgoing in Radar
direction are varied by changing co-ordinates of weapon and fire azimuth. The noise added to every trajectory is different.
Suitable MET conditions are used accordingly. For each case 1000 run has carried out. Following table summarized the



Possible enhancements to the ballistic algorithms for launch and impact point prediction using Unscented Kalman
Filters(UKF), particle filters are being studied and is hoped that investigations using other optimal methods like genetic
algorithms, network architectures will also cater to the increasing demand on accuracy of estimates arising due to faster and
higher computing and signal processing capabilities. Though the decreasing algorithmic efficiency in estimating launch
point from the descending part of the trajectory instead of ascending part and impact point estimation from the ascending
part in the absence of fire data brings forth limitations, such a situation is not generally expected in the battle field offers
some leniency. But the domain specific-ness of the algorithm is also one important reason requiring further studies and
algorithmic development. A more generic algorithm is likely to offer more flexibility with increased target acquisition
capability and is likely to minimize the quantitative requirement on measurements and might offer better estimates in near
real time.
1. Kwakernaak, Huibert and Sivan, Raphael (1972). Linear Optimal Control Systems. First Edition. Wiley-
2. Sontag, Eduardo (1998). Mathematical Control Theory: Deterministic Finite Dimensional Systems. Second Edition.
3. Roweis, S. and Ghahramani, Z., A unifying review of linear Gaussian models, Neural Comput. Vol. 11, No. 2,
(February 1999), pp. 305-345.
4. Bucy, R.S. and Joseph, P.D., Filtering for Stochastic Processes with Applications to Guidance, John Wiley & Sons,
1968; 2nd Edition, AMS Chelsea Publ., 2005.
5. Jazwinski, Andrew H., Stochastic processes and filtering theory, Academic Press, New York, 1970.
6. Kailath, Thomas, "An innovation approach to least-squares estimation Part I: Linear filtering in additive white noise",
IEEE Transactions on Automatic Control, 13(6), 646-655, 1968
7. Robert Grover Brown and Patrick Y. C. Hwang. Introduction to Random Signals and Applied Kalman Filtering, 3rd
Edition. Prentice Hall,1996.
8. K. Ito and K. Xiong. Gaussian filters for nonlinear filtering problems. Kazufumi Ito and Kaiqi Xiong, Gaussian Filters
for Nonlinear Filtering Problems, IEEE Transactions on Automatic Control, vol. 45, no. 5, pp. 910927, May 2000.
9. Joseph J. and LaViola Jr. A comparison of unscented and extended kalman filtering for estimating quaternion motion.
Proceedings of the 2004 American Control Conference, IEEE Press, 2190-2195, June 2004.
10. Simon Julier and Jeffrey Uhlmann. A new extension of the kalman filter to nonlinear systems. Int. Symp.
Aerospace/Defense Sensing, Simul. And Controls, Orlando, FL, 1997.
11. Simon Julier, Jeffrey Uhlmann, and Hugh Durrant-Whyte. A new approach for filtering nonlinear systems. Proceedings
of the 1995 American Control Conference, IEEE Press, June1995.
12. Ben Quine, Jeffrey Uhlmann, and Hugh Durrant- Whyte. Implicit jacobians for linearised state estimation in nonlinear
systems. Proceedings of the 1995 American Control Conference, IEEE Press, June 1995.
13. Rudolph van der Merwe and Eric Wan. The square-root unscented kalman filter for state and parameter-estimation.
Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Salt Lake City,
Utah, May 2001.
14. Girish Chowdhary and Ravindra Jategaonkar. Aerodynamic parameter estimation from flight data applying extended
and unscented Kalman filter. Proceedings of the AIAA Atmospheric Flight Mechanics Conference and Exhibit ,
Keystone, Colorado, August 2006.
15. Text Book of Ballistics and Gunnery . HMSO Publications.




M Padmanabhan, DK Joshi, LK Gite

Armament Research & Development Establishment, Pune- 411021

Tracking and targeting is an important function in military decision making process. Field artillery target acquisition is
crucial in targeting. Weapon Locating Radars (WLR) are one of the primary means to achieve targeting. The process of
targeting, fire registration, adjustment and direction carried out successively neutralizes the target. Weapon locating mode,
the primary operational mode of WLR is the hostile mode operation, wherein WLR estimates enemy fire location-launch
point of indirect fire weapons like mortars, artillery weapons and rocket launchers.
The hostile mode operation requires radar to scan a larger terrain in order to ensure the capture of all penetrating
projectiles and acquire and track their flight to extrapolate the weapon location. Fire direction mode, the secondary
operational mode, is for registration of impact or burst points of friendly fire. The estimate for impact points needs to be
more accurate and timely than launch point since the fire prediction scenario is clearly known in the former case. An
optimization involving narrowed search fence and trajectory for improved parameter convergence rate leads to reduced
response time than the hostile mode. These two modes employed vis-à-vis enable the Fire Direction Centre in its mission.
This report stems out of the work carried out during the development process for WLR. The discussions here pertain to
ballistic aspects in various sub-modules such as tracker training, search fence narrowing, detection and identification,
tracking, estimation and prediction process. The role played by the peripheral value-added tools like position analysis,
diagnostics, run time testing and simulators is also described. The need and concepts involved in arriving at specification
and detection summaries, prediction probability and probability of location are also covered. Future extensions in
methodology and techniques for performance improvements of the WLR system with relevance to ballistics is briefly

In any battle-field, mission success entirely depends on quality of decisions. Military decision making
process(MDMP) involves multiple steps. Mission analysis and Course of Action(COA) assessment are two major steps in
MDMP and involves acquisition and processing of vital information about enemy deployment. A knowledge of enemy's
deployment in terms of layout, expanse, content and strength is a backbone on which all steps and guidelines of MDMP are
built around. Acquiring information on hostile deployment is an essential gamut of operations carried out in every battle-
field. Battlefield visualization is the most important feature of Commander's Critical Information Requirements(CCIR) in
assessing the enemy's position and strength, in order to carry out covering and combat operations. Target tracking is an
essential activity for enemy location estimation. Tracking is a means to deduce the COA of the object. Targeting is a
process of selecting targets and matching the appropriate response to them, taking account of operational requirements and

Target tracking
The classical methods of target tracking are employable only to larger, slow-moving, active targets. The tracking filters
employed in these systems have been rudimentary, typically featuring fixed gain alpha-beta filters with nearest neighbour
association logic. This has been acceptable because cooperative targets carry beacons and transponders, and an operator is
available to resolve any problems. Where as the case is not the same in situations dealing with smaller, fast-moving, passive
targets. Sophisticated filtering algorithms are required for tracking these targets whose dynamics is non-linear, ubiquitously
noisy. The type of target tracking may call for dedicated data acquisition and processing systems which use refined
dedicated algorithms and may be implemented on special purpose high speed processors.

* Flight Dynamics Division, ARDE, Pune-411021



Weapon Locating Radar

Weapon Locating Radars(WLR) are a kind of surveillance equipments belonging to the class of fire-finding systems.
WLRs have recently gained increased acceptance among artillery command control centers world over. These radars can
track multiple foreign ballistic targets which penetrate in to the region covered by radar beam scan, and offer timely
estimates on the launch location of the artillery shell with greater accuracy than conventional methods of target acquisition
and tracking. Apart from providing increased capability in target tracking and acquisition of hostile projectiles, they can also
track and predict the impact points of projectiles fired from own troops. Though this operational mode is secondary to the
hostile mode operation, it provides a means of accurate fire registration for friendly troops. Enhancements to fire direction
process achieved by procedural adoption of fire registration and fire adjustment, ensures timely annihilation of enemy

Role of WLR in Battle-field

Target acquisition and fire direction are two modes of operations that WLR carries out in coordination with Fire
Direction Centre(FDC). In the primary mode, electronically steered radar beams creates a screen like effect over the whole
terrain to ensure detection of any penetrating hostile projectile. On detection a verification beam is initiated. If verification
is positive the tracking beam is prompted to track further on. Measurements obtained are continuously used in an
appropriate parameterized mathematical model of the system. The estimated parameter will be used to back estimate the
target location (launch point) of indirect fire weapons like mortars, artillery weapons and rocket launching systems.
Availability of firing data in Direction Of Own Artillery Fire (DOOAF) mode enables appropriate trajectory prediction,
thereby providing a means to arrive at a fixed location in space, to initiate tracking of own artillery fire. DOOAF mode
tracking is to get prediction of the impact point. The estimate for impact points needs to be more accurate and timely than
launch point since, the fire prediction scenario is clearly known in the former case. Optimization obtained through
narrowing the search space enables allocation of maximal resources to hostile mode operation to achieve the fire-finding
mission .

General Problem
Target spectrum of WLR is restricted to mortar bombs, shells and rockets. Smaller surface area and higher speeds are
distinct characteristic features of this type of targets. Although non-maneuvering ballistic path of these projectiles likely to
provides some advantage, noisy measurements, non-availability of vital projectile data, the influence of meteorological
factors and aero-dynamical forces makes modeling projectile dynamics complex. Projectile dynamics can be completely
and accurately specified by Six-DOF Trajectory Models(6-DOFTM). Accuracy of these complex models essentially
depend on accuracy of aero-dynamic and other important parameters influencing the projectile motion. In the hostile mode
scenario essential parameters for computation are not available. Accurate estimation of the launch point of the projectile
using limited non-uniformly distributed radar tracked data, characterized by noise in measurements and system, imposes
great difficulties in appropriate system modeling for tracking and prediction. This prompts for robust and noise tolerant
tracking and parameter estimation algorithms. The discussions here focus on the fact that meeting the performance
requirement of WLR in turn translates to solving the inverse ballistic problem. Existing inherent intractability in the solution
process forces us to adopt an approach which looks in to the various ballistic aspects affecting individual sub-modules of

Every rigid body of finite mass 'm' hurled in to space with a velocity 'v' and an elevation 'è' follows a parabolic path.
Analytic solutions exits only for, most simplest of representations to this problem. The complete description of projectile
motion (3-translatory and 3-rotational) assuming existence of gravitational, aerodynamic forces and their coupled moments
would require a 6-DOFTM. Even in the modest case, incorporating just the second order variations of aero-dynamic
coefficients in the model would necessitate accurate estimation of a dozen aerodynamic coefficients. Accurate and
complete characterization of projectile's aero dynamic parameters using flight-trials, wind tunnel, CFD are still to date,
constrained by infrastructure, time and cost. The 6-DOFTM is not robust, in the sense that any deficiencies due to lesser
number of coefficients or their accuracies will not be tolerated. Moreover computational time required is also
comparatively higher. All these factors force the modelers to adopt less complex, more reliable and fast models like
modified point mass trajectory model (MPMTM) or point mass trajectory model (PMTM). Calibration of these models
require accurate measurements of all trajectory elements from the muzzle to target. Well calibrated models only can be used
to arrive at accurate prediction. In the light of foreseen discussions we investigate the requirements and procedural aspects to
be adopted in the various sub-modules of WLR for achieving the required performance.




The performance of WLR is entirely dictated by its ability in accurately predicting the location of the launch point (fire
source - weapon system) or the impact point (fire sink - intended target) using radar tracked data. The hostile mode
operation(fig-1) uses the ascending part of the trajectory to back calculate launch point. Whereas the friendly fire mode
tracks the descending part of the trajectory for impact point prediction(fig-2). It can be claimed that the WLR system
performance solely depends on Target Location Estimate(TLE).

Accuracy of TLE is in turn depends on efficiencies of individual sub modules. Each sub-module design should
consider the unique nature of the ballistic problem and envisage the effects of various factors influencing the projectile
dynamics. Failure to meet the minimum required performance criteria in any of these sub-modules may adversely affect the
sequence which produces the TLE.

Object Detection and Identification

Radar measurements R = {(R, è, Ø) / R-Radial range from radar to target, è Beam elevation, Ø Beam azimuth}
obtained at any instant of time denotes reflections from multiple objects illuminated by radar beams. From the moment of
projectile's penetration in to the “curtain of beams”( created by the rapid scanning of the terrain by the multiple beams), the
radar measurement is likely to contain a distinct signature pertaining to the target which will be absent in previous
measurements. This stage calls for the detection of the target. A general approach for detection would be by gradient
analysis of horizontal and vertical components motion parameters. Inaccuracies and resolution of measurements increases
the minimal data requirement for detection and identification. Ambiguities arise when the projectile motion is characterized
by lower speeds. In addition as the fired elevation (high angle fire or low angle fire) determines the nature of vertical and



horizontal gradients, a multi-dimensional approach to discriminant analysis needs to be adopted. Enhancements can be
achieved by constructing scenes using radar measurements from successive time stamps, followed by scene analysis
using correlation filtering methodologies. The data set which is a time stamp of spatially distributed objects in space, on
correlation filtering with an appropriately computed trajectory leads to the segregation of projectile tracked data from the
irrelevant back ground. Though this method does not improve on the cardinality of the data set, this is likely to very well
address the problem of ambiguities and missing of targets by improper identification. Improvements in identification
algorithm not only enhances the probability of prediction by minimizing misses, but also economizes time and data
storage. Every detection triggers a verification beam which in turn continues to track (most liberal sense) and on
identification as a target of interest prompts a track beam.

Filtering & Tracking

Tracking is the most vital process of the WLR system. Well established tracking always guarantees higher
probability of prediction leading to estimates close to the true value. Advancements in signal processing hardware and
algorithms have offered a possibility for finer resolution, thereby putting higher demand over more sophisticated and
accurate tracking algorithms. In the absence of accurate filtering (smoothing the potentially noisy data by mathematical
modeling and using it for next state prediction) algorithms the radar beams are more likely to miss the target than to score a
hit. Un-guided projectile dynamics is a particular case of non-maneuvering, non-linear system. Adding to the complexity
are the non-stationary noise pattern in measurements and the system. Kalman Filter (KF)can accurately estimates the states
of Linear dynamical systems from a time-series data of noisy measurements. The filter exploits the dynamics of the target
for estimation & prediction (multiple step ahead future state ). The Extended Kalman filter (EKF) (non-linear process
dynamics linearised using differentiable functions & their Jacobians for state & measurement prediction), a variant of
simple KF is well suited to model projectile motion. An augmented EKF with unknowns is widely and successfully
employed for tracking and parameter estimation. Though filtering and tracking by EKF offers robustness in dealing with
time series measurement data; convergence, difficulties in parameter tuning, increase in complexity with dimensionality
and parameter estimation are certain issues which need to be resolved. The method of propagating true mean and
covariance adopted in case of Unscented Kalman Filter (UKF), against average mean and covariance propagation in EKF
for predict state could improve the tracking process.

Estimation & Prediction

Process Model
Enhancement of the simple alpha-beta filter (a particular case of KF in its primal form) to EKF or UKF just alone
cannot fully solve the target location problem. The state transition is sensitive to parameters influencing the motion of
projectile in air. An incomplete representation of projectile motion like point mass trajectory model describing only
translational motion,itself requires estimates of unknowns like CD - drag coefficient, projectile parameters m mass of the
projectile, d reference diameter and non-stationary and uncertain atmospheric parameters like ñ density, (Wx ,Wy,Wz)
wind components, t temperature, p pressure prevailing over the whole range of trajectory. Sensitivity analysis reveals the
need for accurate estimation of these parameters and their influence over the target location estimate(TLE ). CD profile of a
projectile has a non-polynomial descript function over the varied mach regimes. There fore the CD estimate made using a
trajectory segment characteristic by a specific mach regime is only applicable in that regime and can to an immediate
neighbour-hood of that regime.

Point Mass Model PMM



Prediction Model
Where as the LP & IP prediction requirement fore sees the applicability of the estimate to a much broader
spectrum. Significant deviations in TLE have been noticed in certain cases where the mach regimes of projectile
tracking and prediction by extrapolation differed significantly. An artillery projectile undergoes drag force variation due to
the effect of compressibility of air and dissipation of energy. Every projectile in general would transit through a velocity
regime from high super-sonic to low sub-sonic passing though the transonic region of rapid variation, during its course
from muzzle end of weapon to the aimed target point. And in an environment of hostile mode operation as both aero-
dynamic and projectile parameters are not-known, grouping them in to one parameter and making an estimate for the same
would resolve the problem of multiple parameter estimation. This parameter would characterize trajectory of the projectile
in the specific regime. The extrapolations to impact points and launch points with the velocity prediction, falling within a
specific interval about the corresponding velocity of the estimate gives fairly close prediction to the true value. The
prediction error increases with increase in difference between the mach regimes of tracking and prediction.

A prevalent procedure employed by ballisticians to compile range tables using C0, ballistic coefficient (carrying
capacity of a projectile ) very well approximates the projectile behavior. In this approach unlike CD profile which is
unique(loosely) to every projectile, a profile of air resistance experienced by a class of projectiles(having geometric
similitude) is adopted as a reference profile (C0, the standard ballistic coefficient for the reference projectile ) and every
projectile belonging to the class is assumed to have a form-factor (the factor of adjustment to the reference profile), by which
the projectile behaviour can be approximated with the reference law. In spite of its simplicity and naïve-ness this approach
yields comparatively better estimates. C0, mostly an invariant for a trajectory, and varies within a small interval about C0 for
the specific shell. The combination involving bundling the parameter and ballistic parameter estimation, with a
meteorological(MET) data extraction during tracking and infusion of the same during extrapolation instead of standard
MET scheme offers improvements in prediction.

Launch point Prediction

Uncertainties and unknowns in hostile fire mode operations makes TLE less accurate. The MET factors have
pronounced effect on the tracked and extrapolated trajectory. MET data used will be inappropriate due to differences time
and location of the weapon. Also the error due to termination point along the extrapolated trajectory severely handicaps the
prediction. The height correction is based on data derived from terrain maps. Shallow fire ( low angle fire - like the firing
from Field Guns) is more sensitive to height error than the vertical fire (high angle fire like in case of Mortar and howitzer ).
Wind effect also becomes critical when the extrapolation of the rockets are carried out as their response to the wind (in
characteristic boost phase ) is different from standard projectiles. As proper implementation is only possible on right
identification and classification of the target. This addresses the need for classification of projectiles in to classes from the
negligibly smaller and noisy signatures. Target classification also helps in MDMP by providing more target intelligence.
Sophisticated and real time algorithms are needed to optimize radar resources and enable real time response.

Impact point Prediction and Search Fence

The advantages existing in terms of fire data availability in DOOAF mode puts higher demands on optimization of
resources, estimate accuracy and response time. The fire data can be effectively used to predict trajectory and to focus the
tracker to a point in space to initiate tracking. An exclusive Artillery Ballistic Kernel (ABK) developed for in service
artillery weapons (Mortars, Field Guns, Howitzer and rockets) is used for trajectory prediction. The fire data along with
MET data is used by the ABK to estimate a probable optimal region in space to focus the radar to ensure capture of own
artillery fire. A restricted simulation carried out within pragmatic variations and consolidation leads to narrow the search
fence. This way of focusing the radar beams in a much narrower region, and more accurate tracking optimizes the resource
spending in DOOAF mode and makes maximal resource allocation to hostile mode. The optimization achieved enhances
the systems ability to handle multiple hostile targets and friendly projectiles.
Accuracy improvements
Apart from the influence of meteorological parameters like humidity, temperature, barometric pressure and wind,
the inaccuracies due to altitude of met data station, geographical, chronological variations and earth rotation effects, earths
curvature effects are also likely to affect the accuracy in estimates. Sensitivity analysis with above parameters and also using
radar track data based on factors like height of initiation, height of termination, mach regime , aspect angle, search sector and
range, vertical scan, velocity requirements, time duration, track volume, sparseness of data and gradient analysis would aid
in prioritizing factors and for allocation of proper weights to each in the prediction model. Site selected for radar
deployment should be evaluated based on slope, area available in the foreground, screening crest, electronic line of sight.
Location averaging can effectively minimize the random error component. Automatic Height correction form terrain
information derived from digital maps and GPS can also improve the accuracy of estimate.


In order to facilitate the system development process and to aid in the process of testing and evaluation of various
hard ware and software sub systems various sub-modules related to ballistic aspects were developed and successfully
implemented. The ballistic work related to WLR carried out includes Radar tracked data simulation which was mandatory
for tracker training, algorithm verification and pruning. Ballistic algorithms for launch point estimation and impact point
estimation which are the core for WLR's prediction performance. Search fence narrower software. Single track simulator for
real time system testing. Target classification algorithm which enhances the radar's capability in assessing the enemy target
. Position analysis tool a means by which the target acquisition battery can decide the optimal geographical location (site
selection ) to track and target from a safe and survivable position. The prior analysis tool available will prove handy and
effective during real battle field missions. Diagnostic tools a sub module developed to replay the track to find and rectify the
problems for prediction probability, accuracy of estimates and reasons for misses. Multiple track simulator for offering off-
line user training for improved preparedness and hands-on experience.
The Position analysis and diagnostics software developed were exhaustively used in combination over a range of
projectiles involving mortar bombs, artillery shells, rockets in all possible variations of fire prediction. The conditions for
which the study was carried out includes geographical location, angle of bearing, elevation angle, aspect angle propellant
temperature, mass of projectile, muzzle velocity, MET parameters, height of track initiation and termination, range to target,
regime of tracking, volume of track, variance levels for the measurement noise, ascending and descending part of trajectory.
The results were analyzed using criteria based on probability of prediction and probability of accuracy of estimate falling
within the specified CEP. Out come of the analysis were used to tabulate the limitations of the system by the specification
summary. Which spells out the ranges of various parameters within which the system is guaranteed to operate. And the
detection summary details the various parameters corresponding to the track to throw light in to possible future directions in
which the tracking could be carried out to obtain the best performance and the factors influencing the accuracy and factors
leading to performance deterioration.

This paper has addressed the ballistic aspects in solving the Inverse Ballistic Problem of finding the fire data from
a segment of a trajectory measured using weapon locating radar. Though EKF based algorithms very well accomplish the
requirement for variety of domains. The domain of interest here exhibits non-linearity in process and in measurement
hence adoption of UKF model for specification nonlinear process and measurement is likely to provide better estimates than
EKF which has a linear measurement model. Further with possible improvements suggested in each of the sub-modules, if
carried out in a combined fashion after isolated evaluation of each methodology's score over the other, is likely to bring
forth enhancements to the system performance.

1. B. V. K. V. Kumar, A. J. Lee, J. M. Connelly, "Estimating Object Rotation and Scale using Correlation Filters",
Opt.Eng. 28(5), 474-481 (1989).
2. Y. Sheng and H. H. Arsenault, "Object Detection from a Real Scene Using the Correlation Peak Coordinates of
Multiple Circular Harmonic Filters", Appl. Opt. 28(2), 245-249 (1989).
3. Hosking, J.R.M. 1990. L-Moments: Analysis and estimation of distributions using linear combinations of order
statistics, J.R. Statist. Soc., Ser.B, 52 (1), 105-124, 1990.
4. Berger, J. 1980. Statistical Decision Theory, Springer Verlag.
5. Rudolph van der Merwe and Eric Wan. The square-root unscented kalman filter for state and parameter-estimation.
Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Salt Lake
City,Utah, May 2001.
6. Eric Wan and Rudolph van der Merwe. The unscented kalman filter for nonlinear estimation. Proc. of IEEE
Symposium 2000 (AS-SPCC), Lake Louise, Alberta, Canada, October 2000.on Julier and Jeffrey Uhlmann.
Unscented filtering and nonlinear estimation. Proceedings of the IEEE, March 2004.
9. Chi-Tsong Chen. Linear System Theory and Design, third edition. Oxford University Press,1999.
10. Robert Grover Brown and Patrick Y. C. Hwang. Introduction to Random Signals and Applied Kalman Filtering, 3rd
Edition. Prentice Hall,1996.
11. Fredrik Orderud. Comparison of Kalman Filter Estimation Approaches for State Space Models with Nonlinear
12. FM 3-60 (FM 6-20-10). Tactics, Techniques, and Procedures for the Targeting Process. 8 May 96.
13. FM 3-09.12 (FM 6-121). Tactics, Techniques and Procedures for Field Artillery Target Acquisition. 25 Sep 90.
14. Text Book of Ballistics and Gunnery . HMSO Publications.


Non Lethal Weapons


Rajeev Prakash
Armament Research and Development Establishment, Pune 411 021
There are many conditions under which too much lethal force is not recommended due to many factors. Today there
are many low intensity operations that carry the threat of violence and may cause severe damages to soldiers or relief units in
the field. The need of a Non Lethal Weapon (NLW) is becoming more and more necessary to balance between lethality and
effectiveness. The ideal NLW should incapacitate the threat to the extent that it is not a threat anymore. This incapacitation
must be absolutely reversible and temporary.
Requirement of clearing facilities and incapacitating personnel via non-lethal means in combatant/non-combatant
environment has been recognized by Navies of various countries. The need of NLWs for Navy can be categorized into two
missions - defensive and offensive. The defensive missions include force protection, particularly for vessels. For force
protection, the Navy should have non-lethal weapons in three capacities: counter-personnel, counter-material and counter-
capability. The force protection measure for a ship or vessel is to generate three layer architecture – an outer zone or assessing
and warning of approaching vehicles and personnel (above, on or below the surface); a middle zone for initial engagement to
turn away threat if it is still approaching, at which point non-lethal means may offer the only reasonable alternative to deter the
threat; and a third, inner zone in which lethal force could be employed. The offensive missions include ship interdiction,
blockades and strikes.
The non-lethal-weapons technologies are grouped in six categories: (1) kinetic-energy technologies (2) chemical and
materials technologies, (3) directed-energy technologies, (4) acoustic technologies, (5) electrical technologies, and (6)
barriers and entanglements. Non-lethal options could offer valuable means of facing the challenges of intercepting suspect
vessels. A running gear entanglement system (RGES) is proposed to protect Navy ships in port, and other waterside assets such
as museums and marines. The entanglement device will foul the propeller of unauthorized vessels attempting to approach
restricted areas. Also included in the NLW are pulsed-energy projectile (PEP)/flash-bang grenades, intense lights for
battlefield illumination (and for dazzling the opponent) and laser dazzling devices that temporarily shield a person or group
from sniper fire. Thermobaric technology is a nonlethal weapon, in development, that causes extended flash, sound,
temperature, and pressure conditions to disorient and/or temporarily incapacitate individuals. Some of the anti-material goals
of nonlethal weapons may be achieved by lethal weapons capable of precision attack. Their effect on material is destructive but
in some cases with very limited unintended damage. Such is the case, for instance, with the use of a laser-guided bomb (or one
guided by GPS) to destroy underground fiber optic cable or on occasion concrete-filled bombs can demolish a small structure
with minor damage to neighboring facilities. Endeavours are being made in development of newer NLW technologies that
include high-power microwaves for countering equipment containing electronics (including some vehicles), counter-
personnel lasers, and counter-materiel lasers.
These weapons for meeting the warfighting needs are still in embryonic stage but have made their need strongly felt
globally. This paper addresses the progress made and technologies being developed for NLWs with an attention to Naval

Less-lethal weapons, less-than-lethal weapons, non-lethal weapons, non-deadly weapons, or, more recently, compliance
weapons are weapons intended to be unlikely to kill or to cause great bodily injury to a living target. There are many conditions
under which too much lethal force is not recommended due to many factors. Today there are many low intensity operations that
carry the threat of violence and may cause severe damages to soldiers or relief units in the field. The need of a Non Lethal
Weapon (NLW) is becoming more and more necessary to balance between lethality and effectiveness. The ideal non lethal
weapon should incapacitate the threat to the extent that it is not a threat anymore. This incapacitation must be absolutely
reversible and temporary. NLWs, while development is currently focused on tactical level applications, offer the operational
commander flexible new tools that can be employed for warfare to achieve operational objectives. These can be used at
operational level to limit civilian casualties and collateral damage. This in turn will allow for more rapid war termination and
will minimize the destruction associated with conventional warfare.
Requirement of clearing facilities and incapacitating personnel via non-lethal means in combatant/non-combatant
environment has been recognized by Navies of various countries. The need of NLWs for Navy can be categorized into two
missions - defensive and offensive.



The defensive missions primarily include force protection, particularly for vessels. For force protection, the Navy should
have non-lethal weapons in the three capacities: counter-personnel, counter-material and counter-capability. Protection of
personnel and ships at anchor or in port is essential. A strategy emerging is the enforcement of a layered defense with three
zones around the perimeter of a vessel. The force protection measure for a ship or vessel is to generate three layer architecture –
an outer zone designed to alert personnel and to assess and warn of an incoming platform, approaching vehicles and personnel
(above, on or below the surface); a middle zone for assessing or affirming intent if the incoming platform or individual has not
somehow acknowledged the warning given in the outer zone and for initial engagement to turn away threat if it is still
approaching, at which point non-lethal means may offer the only reasonable alternative to deterring the threat and non-lethal
options are desirable in such situations because assessed threats may involve unintended intrusion or non-combatants; and a
third, inner zone in which lethal force could be employed. Establishing zones around the perimeter of a ship requires
delineating the zones with barriers that have visible markings and warning devices, such as flashing lights and lasers. Once the
outer barrier is penetrated, assessment of intent and tactics to delay or deter the threat are required. If a potential threat crosses
into the middle zone, the option is open for delivering a warning with a non-lethal deterrent such as “flash bangs” or electronic
vessel stoppers. Entry into the inner zone requires the use of tools that engage the threat directly through either lethal or non-
lethal means. The options selected must be effective on a time scale that allows the employment of lethal means if the NLW
fails to deter the threat.
The offensive missions include ship interdiction, blockades and strikes. Given the inherent capabilities of the Sea Strike
land attack concept, operational NLWs offer the potential for enhancing the effectiveness of lethal forces by, for example,
enabling the commander to engage early, immobilizing enemy equipment without permanent damage to roads or bridges, and
turning moving targets into stationary targets. Non-lethal capabilities can also establish areas of denial and restrict an
adversary's sea space and air space. Non-lethal means can deny an enemy the use of equipment and facilities. For instance, non-
lethal directed-energy weapons can disrupt enemy air defenses and early warning detection sensors; they can also neutralize

The non-lethal-weapons technologies are grouped in six categories: (1) kinetic-energy, (2) chemical and
materials technologies, (3) directed-energy technologies, (4) acoustic technologies, (5) electrical technologies, and (6)
barriers and entanglements.

Kinetic Energy
Kinetic-energy (KE) NLWs were among the first non-lethal weapons developed. KE weapons have a long history
of use by police and military forces. Nonlethal projectiles of various kinds have been developed to stun, confuse, and disperse
individuals and crowds. Rubber projectiles can be fired from standard issue shotguns, either singly or in clusters of balls, with a
range of up to about 30 meters. A grenade launcher weapon can dispense rubber balls for dispersing large crowds and
achieving site security. The sponge grenade can knock down an individual at 50m. A new subcaliber ammunition from an
Italian weapons manufacturer, the ART (Ammunition with Reduced Time of Flight), has been developed as a non-lethal
weapon for use in naval 76mm cannons. One of the proposed uses put forward is for disabling the rudder of a vessel without
sinking it. For intercepting speeding boats, developmental work has been carried out on a 6.25 inch diameter non-lethal
torpedo capable of carrying a 27kg payload.

Chemical and Materials Technologies

A large array of chemicals and materials have been suggested as candidates for use as NLWs. They fall into two
broad categories: (1) antipersonnel and (2) anti-material. Chemical antipersonnel NLWs or incapacitants are intended to
dissuade, temporarily inhibit, incapacitate, or otherwise impede—with no lasting side effects-individuals and crowds from
taking certain actions. Incapacitants are biological and chemical substances that have a calmative effect on humans, and can
cause symptoms such as nausea, disorganized thinking, and hallucinations. Chemical anti-material NLWs are intended to
disable, neutralize, or otherwise prevent the operation of electronics, engines, networks, and so on, in vehicles, vessels or
infrastructure. Anti-material agents are microbes which can be genetically engineered to produce acids or enzymes that have



the ability to degrade a variety of substances, including cement, polyurethane, paint, lubricants, and fuel.

Directed-Energy Technologies
NLWs utilizing directed energy may be divided into three categories for the purposes of understanding
applications and effects: (1) low-energy lasers and incandescent devices; (2) high-energy lasers; and (3) high-power
millimeter-wave and microwave devices.
The category of low-energy lasers and incandescent devices includes laser dazzlers and flash grenades that use
intense visible light to temporarily blind or disorient a person. Thermobaric technology is a nonlethal weapon, in development,
that causes extended flash, sound, temperature, and pressure conditions to disorient and/or temporarily incapacitate
High-energy laser refers to a system with sufficient energy (and/or power) to ablate, melt, or burn material. Their
use as NLWs is intended for applications such as bursting automobile tires, rupturing fuel tanks, selectively cutting through
electrical or communications lines, or setting fires. A second class of high-energy laser systems for antipersonnel application is
designed not to cause damage directly, but rather to produce a kinetic shock through a laser-induced plasma. One such
proposed system is the pulsed-energy projectile (PEP). PEP would utilize a pulsed deuterium-fluoride (DF) laser designed to
produce an ionized plasma at the target surface. In turn, the plasma would produce an ultrasonic pressure wave that would pass
into the body, stimulating the cutaneous nerves in the skin to produce pain and induce temporary paralysis.
NLWs bases on high-power microwave (HPM) and millimeter-wave technology can be grouped into two
subcategories: (1) those designed to disrupt electronic systems, such as communications and computer networks; and (2) those
designed to produce a physiological effect on an individual. Applications in the first category (electronic disruption) include
the capability of disabling or destroying electronic equipment. All sensitive electronics—including computers, cell phones and
radios, Global Positioning System (GPS) receivers, and engine ignition systems—are potential targets. HPM systems would
provide this capability without the accompanying blast effects, physical damage, or death to nearby personnel, characteristic of
explosive or other high kinetic-energy devices.
Some systems designed to produce physiological effects operate at frequencies corresponding to millimeter waves in
the range designed to be absorbed by the skin or at lower frequencies designed to produce resonance inside the body. Potential
applications are crowd control or perimeter protection around a vessel, airfield or other sensitive area. The Active Denial
System (ADS) is a non lethal, counter-personnel directed energy non-lethal weapon which can be used against human targets at
distances beyond the effective range of small arms. ADS projects a focused millimeter wave energy beam which induces
intolerable heating sensation on an adversary's skin and cause that individual to be repelled without injury. One application of
ADS is considered for airborne applications, from platforms such as a gunship.

Acoustic Technologies
A variety of nonlethal acoustical weapons have been proposed and evaluated. Some of these are little more than
fancy loud-speakers, while others involve more subtle or sophisticated processes and truly deserve the designation of acoustic
weapon. Simple high-intensity sound causes the inner ear to generate nerve impulses that register as sound. High-intensity
low-frequency sound may cause other organs to resonate, causing a number of physiological results, possibly including death.
Acoustic weapons pose the hazard of being indiscriminate weapons, potentially imposing the same damage on friendly forces
and noncombatants as on enemy combatants or other targets.
Underwater applications present a potentially more promising scenario due to the increased coupling of acoustic
energy. Past investigations have considered the use of ship sonar against underwater threats. Also being investigated are
underwater acoustic sources as warning or non-lethal options against such threats.

Electrical Technologies
Electrical shock weapons are designed to cause Electro Muscular Disruption (EMD) which, when affecting an
unprotected human completely overrides the central nervous system and directly control the skeletal muscles. Unlike bullet
impact or chemical agents, that are most effective when hitting specific body organs (respiratory system, heart or head), EMD



weapon is effective wherever direct contact is made with the subject's skin (even through few layers of clothing). Employed
either as a direct contact of from stand-off distance, as a stun gun, the EMD effect causes an immediate uncontrollable
contraction of the muscle tissue. Existing EMD weapons require physical contact with the target using hand held shockers or
stun guns, such as the Taser. These guns utilize compressed nitrogen to project two small probes up to 15, 21 or 25 feet at a
speed of over 160 feet per second. These probes are connected to the Taser device by insulated wire. An electrical signal is
transmitted through the wires to where the probes make contact with the body or clothing, resulting in an immediate loss of the
person's neuromuscular control and the ability to perform coordinated action for the duration of the impulse. For military
applications Taser gun is now enhanced for integration with assault rifles. Using Taser guns with rifles enable troops to
selectively use the firepower, and respond to various levels of threats with minimum risk of casualties to innocent civilians.

Barriers & Entanglements

Barriers can be used to form a line of demarcation, to separate adversaries from friendly forces, to delay adversaries
from gaining access to an area, to secure facilities, to stop vehicles, to disable boats, and to serve in many other applications
where delaying an adversary's action is required. Rapid deployment is often the major challenge for effective barrier use. For
vehicle barriers or entanglements, rapidly deployable systems are necessary for putting barriers in place before a vehicle can
enter a secure zone or for allowing timely interdiction of moving land vehicles or watercraft, precision delivery of barriers may
be required. Non-lethal options could offer valuable means of facing the challenges of intercepting suspect vessels. An
example is the running gear entanglement system (RGES), which has proven to be effective in temporarily stopping small, fast
watercraft. It delivers an entanglement device or rope that floats in the water ahead of the target vessel and fouls with the
propeller of unauthorized vessels attempting to approach restricted areas. As the vessel runs over it, the rope becomes
entangled around propeller/rudder. A challenge with a RGES is to integrate the barrier with capable delivery systems providing
rapid and accurate delivery. Under development is a helicopter-based deployment system to accurately and safely emplace the
RGES in order to stop fast boats.

Many of these weapons for meeting the warfighting needs are still in embryonic stage but have made their need
strongly felt globally. However, current and emerging technologies of NLWs will enable to impose one's will on enemy while
limiting the civilian casualties and collateral damage. Also, Non-Lethal Weapons can be the strongest tool of meeting the
defensive need of facing force protection challenges and offensive need of implementing sea strike concepts by Naval forces as

1 Graham T. Allison, Paul X Kelley and Richard L. Garwin (2004), “Non Lethal Weapons and Capabilities”. Report by an
Independent Task Committee Sponsored by the Council on Foreign Relations.
2 Committee for an assessment of non-lethal weapons Science and Technology (2003), “An Assessment of Non-Lethal
Weapons Science and Technology”. Project report under Naval Studies Board, US.
3 Robert T. Durkin (2000), “The Operational Use of Non-Lethal Weapons”.
4 Neil Davison and Nick Lewer (2004), “Research Report No. 6”. Bradford Non-Lethal Weapons Research Project
(BNLWRP), University of Bradford.
5 M. Micheletti (2001), “Non-Lethal Weapons”. 1st European Symposium on Non-Lethal Weapons.
6 Defence Update Issue 1 (2005), “Electro-Muscular Disruption (EMD) Weapon”. Online at http://www.defense-
7 Colonel George Fenton, “Proactive Naval Force Protection: Applying Non-Lethal Dimension”.




V Anguswamy, AC Sahoo
Proof & Experimental Establishment
Chandipur, Balasore, Orissa-756 025

The warhead designer must derive unique and simple technique that will ensure lethal hits on the target. The
effectiveness of each fragment impact in the disablement of the payload is of major concern .All thre at warhead types must
be known, and the designer must develop a warhead to counter all potential payloads. It is proposed new fragment technique
to increase the fragmentation of HE warhead by providing a steel sleeve over the body of the shell leading to decrease in
C/M ration and increasing the useable volume. The steel sleeve increases the degree of confinement and enhances case
breakup process. The better dynamic stress distribution on the inner shell assure better fragmentation. Thus to take on as many
as soft targets possible while outer sleeve will fragment less but would acquire higher velocity to take on hard target like
Tank,APC etc.
This paper dwels on the new technique of fragmentation to produce high velocity and high mass less fragmentation
for hard targets, while high velocity and more small fragmentation for soft targets for better destruction of area targets.

Today's warhead design process against required target is very complex from flyout to fuzing to actual warhead
detonation physics. The design also require examination of new trajectory shaping and guidance laws enable the projectile to
approach the target. Fuse technology need to be applied to determines optimum detonation point. The system should also
achieve high lethality. A fire set or safe and arm device must also be designed in conjunction with the warhead to ensure that it
can be initiated given proper environment. The warhead designer must derive unique and simple technique that will ensure
lethal hits on the target. The effectiveness of each fragment impact in the disablement of the target is of major concern. Key
conditions that define the warhead include warhead weight and space allocations target terminal kinematics including
approach angle and attitude of the projectile with respect to target fuzing logic, time allowed for deployment of warhead
mechanism and target characteristics.
The warhead weight and space allocation must be defined as initial starting condition for the designer. The explosive
density in approximately 22% that of steel, and the warhead volume will determine the warhead C/M ratio. The total warhead
weight and volume will determine the warhead C/M. Non fragment items such as detonator, end plates and fittings are referred
to as dead weight. The designer must made prediction of useful warhead weight relative to explosive and case weight. Actual
calculation of dead weight need to be performed to set total weight buyout. The first step is to approximate the warhead dead
volume where 1-2 inch should be allowed for fuze and fittings. The warhead C/M must be designed so that the velocity and
fragment mass are optimized to allow the fragments to impact as normal to the target as possible. The only function
performed by the warhead is to damage the target so that it is incapable of performing its intended mission . This paper
focuses on chemical energy warhead design details. A chemical energy warhead is a device that contains explosion that on
detonation create blast and fragmentation effects. The metal fragments are quickly accelerated to hypervelocity.


The isotropic or directional warhead concepts accelerated metal fragments in all directions about the center warhead
axis. The exact target location with respect to the warhead origin is not as vital as with a direct energy warhead. This type of
warhead device can be separated into two types, blast and fragmentation . Most blast warheads usually contain some
fragmentation capability ; however, these fragments are not considered the primary kill mechanism. Basically, a blast warhead
and explosive mass installed in a metallic container. When the warhead is detonated, it creates a high-pressure wave followed
immediately by negative pressure. This pressure wave moves outward at hypervelocity speed from the burst point. Damage
occurs when a target is impacted by the high-energy wave. Fragmentation warheads are similar to blast warheads except that
the metal case is the major kill mechanism. Explosive energy is used to accelerate metal fragments to high velocities that
impact and damage threat target components



A fragmentation warhead can be designed utilizing three different fragmentation methods. The first approach is
natural fragmentation . The explosive pressure expands a metal case 50-60% of its original diameter before metal failure
occurs. As the warhead case expands, the case begins to experience plastic strains. The metal case begins to fracture generating
many different size fragments. The mass of the fragments can be biased depending on the thickness, the explosive type, and
the case material properties. This failure of the metal case creates thousands of nonuniform size fragments that accelerate to
high velocities. The second approach strategically places machined grooves, known as stress raisers, which fracture the case
into predesigned uniform size fragments. Many different shaped groove have been developed over the years that can generate
specific fragment sizes and shapes. These grooves consist of v-cut, saw-thoothed, or rectangular shapes. Also, these cuts can be
inserted either on the inside of the case or both on the inside and outside. The last approach is known as a premade fragment
warhead. This warhead consists of fragments that have been fabricated to a specific weight and are bonded or embedded onto
the warhead case with epoxy. These warheads allow a designer to select the exact fragment size and shape that inflicts
maximum damage to targets.


The omnidirectional and directed energy warheads are the bases of most warhead payloads today. Given some
imagination, a design can combine or modify these concept to increase lethality performance. Most designer do not realize
that warheads are inefficient devices, however, in which the majority of the potential energy is wasted . With a warhead
detonation, its gas products accelerate through the vents/ endplates decreasing potential energy and, thus, the capability of
accelerating the metal case .Once the case fractures , there is extensive explosive venting through the metal case , resulting
in a loss of explosive energy . These warhead fragments accelerate to high velocities; however, only a few fragments actually
impact the target to cause component damage. The total amount of explosive energy used to accelerate fragments is on the
order of 10-20%, whereas only 10-15% of the total fragment mass actually impacts the target.
The potential energy of the water behind the dam is similar to a high-explosive charge. Once the dam is blown up, the
water flows towards the buildings and destroys them; nonetheless, all of the water behind the dam is not used to destroy all the
buildings. A warhead works in a similar manner by loading enough explosives or designing the warhead to compensate for
energy losses. Warhead designers need to design and utilize more chemical energy by developing new and novel techniques to
direct wasted energy into the target in an attempt to maximize warhead performance.
The isotropic concepts contain initiation points that are used to propel the fragments about the missile axes. However,
aimable, directional, or mass enhanced concepts better utilize the potential energy contained on the warhead to destroy the
target. When a warhead is detonated an explosive pressure is generated and imparted to the metal case creating 2-4 x 106 psi in
several microseconds. This pressure from the high explosive is much greater than the ultimate yield strength of the warhead
case material, and whether the case is scored or unscored determines how the case breaks apart. If a warhead designer wishes
to control fragments shape and mass, then the metal case is scored with predetermined machined grooves that act as stress
raisers. Conversely, if the case is unscored, then the fragment shape and mass are uncontrolled.
The effective warhead weight is the combination of explosive and case weight; other warhead components are not
considered. The fundamental principle behind the fragmenting warhead is to accelerate hundreds and even thousands of
metal fragments to high velocities. These high-velocity metal fragments have sufficient energy to penetrate hostile targets and
damage vulnerable components. Blast and impulse pressures waves created from warhead detonation are considered
secondary effect. Therefore, design emphasis is placed on fragment shape, material, weight and velocity. The warhead
designer must design the case to break into specific size fragment so that maximum damage is inflicted.


Naturally fragmenting warheads are classified as uncontrolled fragmentation devices. The analytical and
mathematical equations that predict uncontrolled case breakup help to describe and to model controlled case breakup
phenomena. The idea behind the natural fragmentation warhead is to break the case into a number of fragment masses giving a
bias toward a particular mass. Fragment mass bias is controlled by case molecular grain size, brittleness, toughness, case
thickness, confinement, and explosive fill. The natural fragmentation process model is a four-step process, which starts with
radial expansion of the case. The outer surface fractures develop into cracks that grow toward the inner case surface. The



explosive gas products start to flow through the cracks causing massive explosive venting. At this time the warhead case has
expanded to 50-60% of its original diameter. An explosive gas cloud is created with fragments coming out as the gas products
begin to decay. This four-step natural fragmentation breakup process model is shown in fig 1.

Fig. 1 Expansion process of natural fragmentation warhead

Fig. 2 Stress components on warhead case element




The warhead case expands until the internal energy exceeds the strength of the case. The warhead case can be
modeled as a cylindrical pressure vessel and loaded with a uniform static pressure. Upon warhead detonation, high pressure is
exerted on the inside of the case and produces a dynamic stress. The stress that is applied to the warhead case has three
components. The principle stresses are radial compressive stress (6r), tangential tensile stress(6?) and axial tensile stress (6z) .
The three principle stress components are shown in fig.2.


The shear stresses are determined from the principle stresses. Shear stress r/ is computed from principle stresses r and
whereas shear stress r /z is calculated from principle stresses r and.z The principle stresses and z calculate shear stress / z. A shear
plane element is shown in fig. 3. Principle stresses and shear stress can be computed, showing major modes of failure that are
caused by fracture stresses in the r/ plane and radial tensile fracture . One of these modes will dominate in the failure of the
warhead case and case thickness; metal properties and temperature are variables that influence such processes. Consider a
cylindrical warhead with an inside radius ri and case thickness t under an internal pressure (fig.4). the outside radius is
designated as ro whereas an inside pressure is Pi. A thin infinitesimal ring can be analyzed at radius r summing all forces in the
vertical plane and letting the equation equal zero gives


Fig. 3 Shear plane directions1

Fig 4 Cylinder subjected to internal pressure.




Shear fracture and radial fracture are the dominant modes resulting in case breakup and the fracture mode that is
predominate depends on case thickness, temperature, and material properties such as ductility. The locations of all of the shear
fractures on the warhead case is random, which results in the formation of many different size fragments. As the pressure in the
warhead case increases, the plastic region of the case extends outward creating fracture trajectories caused by the principle
stresses. These trajectories have direction and form a network of radii and circles. The maximum shear stress trajectories form
a grid of logarithmic spirals crossing the trajectories of principle stresses at 45-deg angles. This type of shear fracture occurs in
both the region of elastic deformation and in the region of plastic deformation. An illustration of a cylinder showing 45 degree
maximum shear trajectories is shown in Fig. 5

Fig. 5 Cylinder with 45-deg logarithmic spirals with fractured pieces from test.

The case thickness of the cylinder in Fig. 6. is thick steel and the photograph shows that the cylinder has broken into
many major fragments. The pressure generated from the explosive fill did not create a force great enough to fracture the entire
cylinder. The inner surface diameter increased near 77% of its original diameter


Consider a thick wall cylinder that experiences elastic-plastic stress distributions. The inner radius is ri and outer
radius is ro with an internal pressure Pi, The external pressure is expressed as Po but is negligible compared to the explosive
detonation pressure. If the cylinder is considered as plane strain, then åz = 0.0. The elastic hoop and radial stresses are expressed
as (óè, ór, whil ä is the radial case deflection.






Where v is Poisson's ratio. If the exterior pressure Po is equal to zero, then the principal stress occurs at the inner
surface of the cylinder. From Mohr's circle the inner surface of the maximum stress is (ó è - ór so that



Where j = ro / ri. The inner surface of the cylinder is where plastic yield first occurs. . If the difference between any two
principal stresses of the cylinder is greater than the case yield strength, then yielding has occurred. Plastic strain occurs first at
the bore when r =ri , and applying the equation gives



The internal pressure p/ is required to produce plastic yielding at the inside surface of the warhead. The elastic and
plastic fields are shown in Fig. 6. If the elastic plastic yield condition is employed, then Eq. (8) is now in terms of by where pip is
the inside pressure required to take the case to plastic yield. The plastic yielded material expands past the plastic zone and
infiltrates the elastic region when pip exceeds its plastic limits,



Fig 7 Warhead case expansion with fracture

This model can be expanded, and an equation can be derived that predicts internal pressure to cause case fracture. Assume the
warhead is initiated isotropically at and end after 10-20 us, the warhead would look as shown in fig.7

Figure 8 also shows the growth of surface cracks as a function of time. These cracks were initiated around 40-50 ìs
after detonation. The crack lengths were 10-20 mm, but as time increased they grew to 30-50 mm. These test provided valuable
data on number of fragment produced from a cylinder warhead. Fragment mass distributions were experimentally predicted to
be of the form


where N(> m) is the number of fragments with mass higher than m, No is the total number of fragments, ì is the fragmentation
size variable, and m is fragment mass. Fragment distributions were computed for each cylinder warhead test. Figure 9 plots the
number of fragments with mass greater then m vs 1 ft2 of fragment mass m.
The results in Fig. 9 show that experiments 5 and 6 did not produce a wider distribution of fragment sizes. One could
conclude that experiment's 5 and 6 warheads are brittle than the others. The explosive used in those tests was comp-B, and the
detonation front ran right to left.
The explosive provides a large pressure that is imparted to the metal cylindrical case, and this force causes the case to

expand with a high radial speed. The warhead case is

experiencing biaxial tension, and with increasing tension the
case material experiences a reduction in area. Once the critical
load is reached, then the material strength cannot compensate
for loss of reduced area. Several empirical equations for
uniaxial stress-strain can be applied for a strain-hardening
material. These equations are uniaxial stress-strain can be
applied for a strain-hardening material. These equations are
where B, B', C Y, N, P, and M are empirically calculated
constants.These equations help describe the fall in hoop force

(11) Fig. 8 Explosive test cylinder case

expansion as a function of time.
(Courtesty of Don Curran at SRI)



Fig 9 Fragment mass summary for explosive cylinder tests. (Courtesy of Don Curran at SRI)

around the circumference of the case where necking has formed. A case experiencing necking is in Fig.10
It is pointed out that if the radial speed is large, then a second and third neck will form. This is because the release
wave from the first waves has not passed the formation of the 'second and third portions of the necking case. It is important from
a engineering perspective to understand this necking process and to be able to calculate the initial speed of the warhead case that
causes static instability. Assume the elastic strains are small compared to the plastic strains. An equation of radial motion can be
written It is now easy to compute the instability strain as a function of V0. Before necking and fracture arise, the velocity that
will cause strains equal to the static instability






Fig 10 Necking of case when expanded at high radial speed.

(15 )


The charge to mass ratio C / M can be computed that would be required to obtain the velocity for the instability strain.
The radial velocity required to achieve instability strain is 688 F/s for steel while N, C, and B equal 0.50, 0.016, and 3.2/l.0E6.
The instability strain åè is 0.236 for a cylinder. The Gurney equation computes the peak ejection velocity of the fragment. This
peak velocity can be compared to the velocity for instability strain,

(16 )

(17 )

(18 )

(19 )

(20 )

(21 )




The preceding mathematical calculations computed the total number of fragments on a warhead given case thickness,
warhead diameter, warhead length, and explosive characteristics. An equation has been developed [Rosin, Rammier, and
Sperling (RRS distribution)], which computes number N of fragments whose mass is equal to or less than the value m, where M
is the total weight. This is

(22 )

(23 )

(24 )

(25 )

(26 )

(27 )

This equation was generated by collecting warhead arena test data and plotting Óm vs N. The data are plotted in Fig. 16
The dashed lines in Fig. 11 refer to Eq. (27) and the solid lines represent actual data. An attempt was made to derive an
equation for average fragment mass. The average fragment mass utilized 75% of the warhead casing mass, and Óm is 75% of the
total mass, which is divided by fragment number N. The average fragment masses for two different charge diameters are
plotted in Fig. 12. as a function of case thickness t.
The equation derived from Fig. 13

(28 )

(29 )

Fig 12 Average fragment mass m75%

Fig 11 Sum of fragment mass Óm vs N function of charge diameter. vs case thickness given steel case warhead



Fig 13 Expanded element of warhead showing geometric values

(30 )

(31 )

(32 )


(34 )

(35 )

(36 )

Equation (36) calculates the average fragment mass mf given a specific percentage of z where z is the ratio of
fragment mass Óm to total case mass M. The fragment mass can now be computed as a function of case thickness and explosive
diameter. The function z/[ -en (1 - z)] is plotted from 1 to 0, which corresponds to 100% and 0%.This warhead has expanded up
to 60% of its original diameter, and experienced elastic-plastic failure before fracture cracks occur. The equations can estimate
fragment size distribution as a function of C / M, L, D, and case thickness. several equations that were developed predict two-
dimensional case breakup and mass distribution of fragments. His equation is expressed as

(37 )

(38 )



Equation can be expressed assuming a specific confidence level (CL) of a particular size fragment

where the largest mass m f is achieved 95% of the time. The total number of fragments greater then m f is




The warhead case is considered thin walled when the thickness is less than or equal to 0.6 in. If the case thickness is thicker,
then a three-dimensional equation is employed,

The total number of fragments N o is equal to M /6ì, and 6ì, is the arithmetic average mass.


The selection of fragment mass and shape has always been difficult, and much time is utilized attempting to optimize
its geometric configuration. The warhead designer must analyze the vulnerability data of the target before any warhead is

The fragment mass can be computed by

MF = SFtpf /6=C1t3(C2+2)pf (44)

This derivation computes fragment thickness while maximizing fragment volume. The size, weight, and shape of the
fragment is critical in inflicting maximum damage to a target. One way to measure how lethal the fragments are is to compare
KE and momentum to penetration through target plates. The penetration of a fragment though a plate is a function of fragment
mass, obliquity, yaw, shape, and velocity. The velocity of a fragment when arranged on a warhead can be estimated

Fig. 14 C/M demonstration explosive and metal effects on peak ejection Velocity vs



by a simple Gurney equation

Vo= 2E CIM

Where 2E is aGurney explosive constant that has units of velocity. If ã½= Vo/ 2E and a curve ã ½ vs C/M is plotted, then the
effect of the C/M ratio on ejection velocity can be examined (Fig.14)

The curve shows that low fragment launch velocities are significantly affected by C/M whereas large C/M ratios contribute
little to enhancing velocity.
Based on the fragment shape and mass, an estimate of total momentum and KE of the warhead can be computed. This
analysis provides insight into maximum penetration performance. After that the case thickness t can be calculated as a function
of warhead design parameters. The fragment must be designed so that it does not shatter or fracture during initial launch. The
inner liner material is also an important part of the warhead design.
A great deal of care must be taken in choosing the liner material and thickness because it is possible that fragment
fracturing could occur. Fracturing of the

Fig. 15 percent of maximum vs (C/C+M)

Fig. 16 Description of pressure wave' s impact on different positions on case.



Premade Fragment is also a function of initiation geometry and fragment longitudinal location. If the fragments are near the
booster or near the warhead end, then nonuniform detonation pressures with reflected shock waves could be exerted on the
fragments. A half-barrel-shaped warhead is shown in Fig.(18), demonstrating that the C ratio and shockwave exerted on these
fragments differs depending on the fragment's position.
The fragments experience a lower ratio at location P1 whereas location p2 experiences increased with a normal
shock front. This increased pressure at p2
M is attributed to the location of the booster charge. When the warhead
M is detonated, the
highest pressure is thrusted forward, or normally in front, at nearly 300 kbars whereas lower pressure with rarefraction waves is
experienced near the booster side fragments. This occurs because the blast wave runs nearly tangent to the warhead case.
The inner thickness is designed to hold and sometimes mitigate some of the shock front, and to allow the fragment to
be ejected without major fracture cracks. The designer must be careful that the shock front does not fracture the fragment
during initial launch. A premade fragment after initial launch is shown in Fig.3.17

Fig . 17 Description of fragment stress wave during launch

The explosive creates an intense, high-pressure impulse wave that pushes on the fragment, causing it to accelerate to
high velocities. The explosive impulse acting on the fragment can be modeled as a triangular saw tooth pulse. If the fragment
undergoes fracture, then the tensile wave traveling though the fragment exceeds some critical stress value óf. When the wave
reaches the outside surface, it is reflected with a maximum tensile stress ót. The tensile stress is computed as

ót=óM - ói

where ói is the compressional incident stress and óM is the magniture of the incident compressive pulse.


Some warhead designers may want to design a warhead with an uncontrolled fragmentation case with premade
fragments. A designer may decide to make the premade fragments large because deep penetration into targets is required.
Smaller fragments, from an uncontrolled case, generally destroy thinner skinned threats.A warhead with premade fragments
on an uncontrolled case would generate a combined-effects warhead that would accelerate large mass fragments at heavy



Fig. 18 Illustration of pre-made fragments on uncontrolled fragmentation case

componented with small, light fragments that would kill thin component. A warhead with premade fragments with an
uncontrolled fragmentation case is shown in Fig 3.18

An equation for total warhead weight can be expressed as

WT = W e+ WUC + W f

Where We is the of the explosive, Wuc is the weight of the uncontrolled case, and Wf is the weight of the premade fragments.
This equation can be expanded to



From this equation the explosive diameter can be calculated based on initial conditions and desired fragment weights and

The equation for the explosive radius is




The explosive diamenter De is also the uncontrolled case inside diameter, and the outside diameter Di is computed by
This equation can be expanded to solve for uncontrolled case thickness. If
Then solving for tl gives



Where B is a constant.

The design of dual fragmentation aimed at effectively utilizing the maximum potential energy of the warhead. This is
achievable using two liners of different case thickness, with inner being thinner and allowed to expand maximum while
outter being thicker with stretching limited to just short of reaching plastic region. During such situation inner with
prefragmented is to spread maximum fragments and with high velocity. Where as the a outter with natural fragmentation
having bigger sizes and gaining tremendous momentum, would be able to destroy hard target like tanks , A PCs etc. where
as inner liners providing maximum number of fragments with equally high velocity would prove to be effective against soft


1. Richard M Lloyd, Conservational Warhead Systems Physics and Engineering Design, Progress in Astronautics and
Aeronautics, Volume 179

2. Roger Holmberg, Explosives and Blasting Technique, A..A Balkema Publishers, Lisse/Tokyo.

3. Charles L. Mader, Numerical Modeling of Explosives and Propellants, CRC Press, Taylor & Francis Group

4. Text Book of ammunition, 1936, HMS Office publisher

5. Artillery ammunition, Vol-3, The Ordnance School, Aberdeen Proving Ground, Maryland

6. Reger T Fernner, Mechanics of Solids, CRC Press, Washington DC

7. Liu Congdian and Roy W. Nichols, Design and Analysis, Vol-1, Pressure vessel Technology, Pergamon Press




Brig HSN Sastry (Retd)
Institute of Defence Scientists and Technologists, Pune

The paper brings out clearly the concepts of non- lethal weapons, their state of art, and the pressing
need for their use in the present global scenario. Security forces increasingly operate in challenging
environments other than conventional wars. Non lethal capabilities provide flexibility by allowing forces
to apply measured force with reduced risk of casualties. Political, economic and humanitarian
considerations dictate that wherever possible, casualties should be minimized while also limiting
collateral damage. First and second generation non-lethal weapons, which include rubber and plastic
bullets, bean bags, stun guns, tasers, electric shock weapons, incapacitant gases, batons, laser weapons,
water cannons etc have been described briefly. Emerging technologies in a range of mines, encapsulation
devices, thermobaric technology, electromagnetic directed energy weapons, melodorants, unmanned
aerial vehicles, acoustic beams, bio-technical and biodegrading microbes are briefly brought out in the
Suitable technologies to be developed in the country to meet the needs of our security forces, and
project to be undertaken by DRDO, in consultation with military, Para military, civil police etc are
brought out. Lethal, ethical and humanitarian aspects of non-lethal weapons, based on national policies,
public opinion, and the need of formulating suitable laws, commensurate with international standards are
The paper also bring out briefly, the role of IDST (Institute of Defence Scientist and Technologists), an
organization with a large number of experts with years of experience, who can assist in envisaging such
non lethal weapon projects, give necessary consultancy services and coordinate the progress of such





KD Dhote, KPS Murthy
Armament Research & Development Establishment, Pune 411 021

Gurney's formulae for estimation of velocity of metal liners in contact with detonating explosive are based on
conservation of momentum and energy. In spite of several limitations associated with these formulae, they are still in use for
prediction of ejection velocities of both controlled and pre made fragments, due to their simplicity in use and good agreement.
However, these formulae can't predict the ejection velocity of multi-layered pre made fragments.
The design of warhead is often dictated by the mass & space constraints. Hence, to achieve higher hit densities, the
pre-made fragmentation warheads are often designed with multiple layers of fragments. On detonation, fragments from
various layers are ejected out with differential velocities thereby generating number of fragment wave fronts. The separation of
fragments in multi-layers is affected by ringing and bouncing phenomenon associated with interaction of shock waves with
fragments, fragment motion through explosion products and subsequent fragment motion in free air.
To study the effect of fragment separation, authors have conducted experiments with single column of 3 fragments in a matrix
construction. In these studies, spherical tungsten alloy fragments were packed in the matrix and propelled out with high
explosive charges having C/M ratios from 0.16 to 0.64. It is observed that the velocities of inner fragment & middle fragment
are 86 % & 95% of the outer most fragment respectively, irrespective of C/M variations. Mathematical relations for estimating
the fragment velocity of each layer have been presented using conservation of momentum & coefficient of restitution. The
methodology of estimating the coefficient of restitution presented in this paper can be effectively used for different types of
fragment matrix constructions and derive the mathematical relations for estimating the velocity of the fragments from each
layer, before optimizing the design of warhead.

Gurney in 1940's developed the mathematical relations for estimating the velocity of metal liners in contact with
explosive, which are essentially based on conservation of momentum and energy [1]. Application of these relations for various
warhead configurations is limited due to the assumptions made in deriving the formulae. However, Gurney's formulae, are still
in use, due to simplicity in use and good agreement with experiment results for many configurations. One of the limitations is
that it can't predict the velocities of multi-layered pre made fragment configurations.

Warhead designers most often have to optimize the performance of kill mechanism with in the space constraints of
overall weapon system. In pre made fragmentation warhead configuration, as many as fragments are need to be packed to
achieve desired fragment hit density on target. Hence, fragments are packed in multiple layers. On detonation of warhead,
fragments are ejected outwards with differential velocities, thereby forming fragment wave fronts. Single layer pre-made
fragment launch dynamics is discussed in Reference [2]. The phenomenon of fragment separation in multi-layers is presented
in References [3] & [4].

The phenomenon of fragment layer separation is categorized in following three ways.

a) During shock waves transmissions
The separation of layer due to shock wave transmission occurs either by meeting of compressive & tensile shock waves at
layer interface or by long pressure pulse or by differential shock impedance characteristics of layer material [3].
b) During fragment motion in explosion gas products
Even after initial small separation of fragment layers by shock wave transmissions, high pressure gaseous products further
accelerate the inner fragment layer i.e. layer in contact with explosive. This fragment impacts on the fragment next to it
and next fragment impacts on the subsequent fragment ahead of it and so on. These multiple impacts between the
fragments are described as ringing bouncing phenomenon [4]. Thus, the fragments mainly separate due to momentum
transfer which is dependent on the coefficient of restitution 'e' for the materials in layer [5]. In addition to this, the
interaction of expanding hot gases with fragments also contributes in its separation.



c) During fragment motion in free air

After fragment comes out of explosion gaseous products with differential velocities they move in surrounding air. During
this phase, the drag force exerted on it further contributes in its separation.

The fragment separation phenomenon, thus depends on construction of fragment matrix and the material characteristics
predominantly in a given explosive system. Authors have conducted experiments with a single row of 3 layered, spherical
tungsten alloy fragments packed with resin harder in a matrix and propelled with an high explosive having C/M ratios varying
from 0.16 to 0.64 to study the effect of fragment separation. The velocities of inner fragment & middle fragment are observed to
be in the order of 86 % & 95% of outer most fragment respectively and do not depend on C/M ratio. Mathematical relations for
estimating the fragment velocity of each layer have been worked out using conservation of momentum & coefficient of
restitution. Based on experimental results, coefficient of restitution has been found to be in the order of 0.1 for the fragment
matrix, under experimentation.

On initiation of explosive, fragments get separated due to shock phenomenon, followed by its motion in explosive gas
products & in air. When the fragments are initially deployed there is an impact momentum transmission phenomenon that
occurs between neighboring fragments. The collisions that occur between the fragments happen within a short period of time.
When one fragment impacts another, the fragment are subjected to localized deformations under high strain rates, over a very
short period of time. This is followed by a short period of restitution. The coefficient of restitution is expressed as 'e' and is the
ratio of the integrals of restitution and pressure [5].
e = Rdt / p dt

The value of 'e' is between 0 and 1 and is dependent on fragment material, shape, surface condition, filler materials
(momentum traps) and impact conditions. For perfectly plastic impact, the coefficient of restitution 'e' will be 0 and the
fragments would stay together because there is no time period during which restitution occurs. However if e = 1, the impact is
said to be perfectly elastic, and the fragments move away from each other during initial launch itself with large difference in the

Let V1, V2 and V3 be the ejection velocities of the fragments of first, second and third layer fragment respectively. V is
the average ejection velocity of the fragments bundled together. V12 is the velocity of the fragment bunch of layer one and two.

The conservation of momentum yields

3 M V = 2 M V12 +M V3

Where, M is each fragment mass.

Thus, 3 V = 2 V12 + V3 ------------------------(1)
Similarly, 2 M V12 = M V1 + M V2
Therefore, 2 V12 = V1 + V2 -----------------------(2)
Thus, 3V = V1 + V2+ V3 -----------------------(3)

Consideration of the coefficient of restitution yields

e V = V3 - V12 ----------------------(4)
e V12 = V2 - V1 ---------------------(5)
The above equations (1), (2), (3), (4) & (5) can be solved for obtaining the initial velocities of fragments in
three layers i.e., V1, V2 and V3 as follows,
V1 = (1/6) (2-e) ( 3- e) V ------(6)
V2 = (1/6) (2+e)(3-e)V --------(7)
V3= (1/3)(3+2e)V ------- (8)




To study the fragment separation, a Fragment Launch Unit (FLU) of
steel has been used to launch three numbers of 9 mm spherical Tungsten Alloy
(TA) fragments in a single column by high explosive as shown in Fig. 1.
Fragments are packaged in resin hardener mix and assembled at one end of the
FLU. High explosive pellets are inserted from the other side. An electrical
detonator is used to detonate the high explosive. The trials were conducted
with C/M ratios of 0.16, 0.32, 0.48 and 0.64.

Trial set up is shown in Fig. 2 & Fig. 3. A high-speed camera having

recording speed 1,00,000 frames/sec was used. The camera speed was set to
10,000 frames/sec to record. A velocity screen was placed in front of the
camera to record fragment motion. On a velocity screen vertical lines having
distance 100 mm in between the lines were marked to measure distance of
fragment traveled. A firing circuit was used for positive recording of the event.
In addition to this, a screen is also placed in front of the FLU to record the
fragment spatial distribution & path of its travel.

The separation of fragments recorded in one of the trial is shown in

Fig. 4. The velocities recorded in various trials are given in Table 1. It is
observed that, the velocity of 1st fragment & 2nd (middle) fragment is around
86 % & 95% of 3rd fragment irrespective of C/M variations.

Based on experimental results given in Table 1 &

Eqn. 6 to Eqn. 8, coefficient of restitution for each trial has
been worked out, and found to be varying from 0.095 to
0.112 in various trials, for the 3 layered TA fragments
bundled in a column using resin hardener mix.




Shot no. C/M 3rd frag. vel.(V3) 2nd frag. vel. (V2) 1st frag. vel. (V1)
A1 0.16 310 m/s 295 m/s 265 m/s
A2 0.16 323 m/s 310 m/s 275 m/s
A3 0.16 318 m/s 302 m/s 270 m/s
Average [% wrt 3 frg. vel.] 317 m/s 302 m/s [95.3%] 270 m/s [85.2%]
B1 0.32 457 m/s 435 m/s 396 m/s
B2 0.32 450 m/s 430 m/s 390 m/s
B3 0.32 439 m/s 418 m/s 380 m/s
Average [% wrt 3 frg. vel.] 449 m/s 428 m/s [95.3%] 389 m/s [86.7%]
C1 0.48 522 m/s 496 m/s 449 m/s
C2 0.48 526 m/s 500 m/s 455 m/s
C3 0.48 540 m/s 510 m/s 460 m/s
Average [% wrt 3 frg. vel.] 529 m/s 502 m/s [94.9%] 455 m/s [86%]
D1 0.64 615 m/s 585 m/s 520 m/s
D2 0.64 593 m/s 564 m/s 515 m/s
D3 0.64 629 m/s 600 m/s 530 m/s
Average [% wrt 3 frg. vel.] 612 m/s 583 m/s [95.3%] 521 m/s [85.1%]

The paper presented the details of trials conducted for 3 layered Tungsten Alloy fragments bundled in a column using
resin hardener mix and mathematical relations for estimating the velocities. It is observed that the velocity 1st (fragment in
contact with explosive-inner) fragment & 2nd (middle) fragment is 86 % & 96 % of 3rd (outermost fragment) irrespective of
variation in C/M ratio and the coefficient of restitution is found to be in the order of 0.1 for the matrix under consideration. The
methodology presented in this paper can be effectively used for estimating the coefficient restitution for different types of
fragment matrix constructions and derive the mathematical relations for estimating the velocity of the fragments from each
layer before optimizing the design of warhead.

Authors are extremely thankful to Director, ARDE for his guidance, encouragement and motivation through out the
investigative trials, analysis of results and presentation of this paper.

1. Zukas W, Fundamentals of shape charges, 1989, pp. 45-71
2. O'Donoghue PE, Predebon WW, Anderson Jr. CE, “Dynamic launch process of preformed fragments”, J. Appl.
Phys.,63 (2), 15 Jan. 1988. pp 337 – 347.
3. Jones GE, “The Gurney equations for multilayered fragments”, J. Appl. Phys., 50 (5), May 1979, pp 3746-3747.
4. R. Lloyd, “Conventional Warhead Systems Physics and Engineering Design”, Progress in Astronautics and
Aeronautics, March 1998, Vol. 179, pp. 140-149.
5. Beer F. and Johnston E., Vector Mechanics for Engineers, Dynamics, 3rd ed., McGraw – Hill, New York, 1997. pp. 594 –





S Harikrishnan, KPS Murthy

Armament Research and Development Establishment, Pune 411 021

Tandem shaped charge warhead is one of the efficient methods to defeat Explosive Reactive Armour (ERA) protected Main
Battle Tanks (MBT). The concept involves two shaped charges mounted in the same missile are initiated one after the other
with certain time delay. First shaped charge jet would remove the ERA and the second jet would penetrate the bare armour.
Both these explosive charges are mounted close to each other most of the time due to severe space constraints in the system. It is
necessary to protect the second charge from the blast effects of first charge, during the time delay between the initiations of two
charges. Blast barriers are being used in the space available between the charges for this purpose. More often, addition of such
component might introduce excess weight penalty and deterioration in the performance of the second charge itself.
This paper deals with the design and evaluation of a lightweight barrier to protect the second charge. Shaping of the barrier
is carried out in such a way that collapse of the second shaped charge liner is unaffected. 2D analysis was carried out using
Autodyn to verify the same. Flash X ray taken at different instances after the initiation of first charge shows that the barrier is
intact when the second charge is set to initiate. Static experiments of tandem charges were conducted which shows that the
performance of second charge is unaffected by the effects of detonation of first charge.

Introduction of Explosive Reactive Armour (ERA) protection to Main Battle Tanks (MBT) necessitates the use of tandem
shaped charge warheads for anti tank missiles to defeat the MBTs. The concept of tandem shaped charge involves a precursor
charge (PC) and a main charge (MC) set to initiate with a predetermined time delay. Both these charges function based on
shaped charge principle. Hyper velocity jet from precursor charge would initiate ERA and the flying debris and high-pressure
detonation products emanating from ERA would create a disturbance zone for shaped charge jets. Hence the main charge jet is
allowed to reach the target only after the ERA products are removed from the main charge jet path. At the same time the
precursor charge detonation in the near vicinity of main charge would have detrimental effects on the performance of main
charge. Effects of precursor charge blast on main charge functioning are mainly in two ways. Firstly, direct blast effect and
secondly disturbance in the collapse process of main charge liner due to the flying components of missile system housed in the
space between PC and MC. Hence it is necessary to protect the main charge from the adverse effects of PC blast. This becomes
more relevant as presently larger precursor charges are being used in anti tank missles to defeat stronger ERAs.
Traditionally there are three different methods to protect the MC [1]. One of these methods is to use a massive bulk head to
shield the blast effects of PC. Its mass prevents the bulk head from accelerating towards MC. But the drawback of such a design
is that it would reduce the performance of MC. Moreover, the weight penalty offered by such a component might not be
acceptable to the missile system designers. A second method is to use components already in the missile as a shield. A third
approach is to use the space between the warheads to dissipate the blast energy before it reaches main charge.
This paper presents the design considerations and evaluation of a lightweight barrier for an existing anti tank missile
system. This system under consideration is a third generation anti tank missile with a massive front section, which houses the
precursor charge. Precursor charge consists of 250g of HMX based explosive. Main charge is housed at 200mm to the rear of
precursor charge. Space between PC and MC houses electronic systems consisting of printed circuit boards, connectors, cables
and their mounting frames. Precursor charge is initiated by a fuze, which has mass of 250g. A general arrangement of warhead
is shown in Fig: 1.



Main charge Fuze for Precursor charge Precursor charge

Electronic components

Fig: 1 General arrangement of a tandem warhead system for a third generation anti tank guided missile.


Requirement of the blast barrier was conceived from the inconsistent performance of Tandem warhead observed during
static trials. Systematic conclusions were drawn from the characteristics of main charge jet penetration hole observed on the
target material. Main charge jet was observed to be shifting to the upper plane from the intended point of hit on the target as
shown in Fig 2.

Shift in Jet

Jet impact

Fig 2 (a) Trial set up (b) Target plate

Shift in the jet penetration towards the upper plane can be explained by assuming an asymmetric collapse of the liner
elements. If collapse of liner elements above the liner axis is perturbed, jet flow will tend to shift towards the upper plane to
conserve momentum. Perturbation of collapse can be caused by the debris of precursor detonation entering into the collapse
Autodyn 2D analysis was carried out to assess the time at which the blast product from precursor enters the collapse volume
of main charge. An axisymmetric model of the precursor charge and blast barrier was employed for simplicity. Fuze for
precursor charge was modeled as equivalent thickness of Aluminium. It has been observed in the simulation the back blast from
the precursor charge arrives at the location of blast barrier close to the time at which main charge is set to initiate. It has been
observed the pressure wave deteriorates considerably since considerable amount of energy is absorbed by the deformation of
the fuze. Different events of the blast phenomenon of precursor charge is shown in Fig : 3. In the actual configuration, apart
from the fuze, several guidance electronic components are positioned within the space between the PC and MC. These
components are likely to enter the collapse volume and affect the process of jet formation much more adversely than observed
in simulation.



Main charge
collapse zone


Fuze is accelerating towards Fuze enters main

Main charge liner collapse charge liner collapse
zone zone
t = 300 m
Fig 3 Hydrocode simulation of Precursor blast


Location Of Blast Barrier

Objectives of blast barrier are to protect the main charge from direct blast effects of precursor charge and to arrest the flying
debris from entering into the collapse zone of the main charge. To meet these criteria, blast barrier has to retain its configuration
till the main charge is initiated and collapse process of liner is completed. It is not preferable to design a massive barrier to with
stand the blast loads since it only adds additional mass for the missile to carry. If the disc is too near the precursor charge, it will
attain high velocity causing damage to the main charge. Hence it is preferable to place the disc as far away from the precursor
charge as possible and allowed to receive less impulse and at the same time, the time taken for the movement of disc should be
such that the second charge is not disturbed. Distance at which barrier is to be located is worked out in an approximate way as

Pressure at the blast isolation disc is given by :

105 ´
20´ W 2/ 3
Pmax = 2

where Pmax = peak over pressure at the given distance, R

W = weight equivalent of explosive in terms of TNT

The pressure pulse duration is dependent on the distance R and the sound velocity in the gaseous products of detonation
t An estimation of is given by the ratio between the distance R and the sound
which is of the order of 2000m/s.

impulse per unit area = Pmax Dt
1 p
Total impulse = Pmax D́
t d2
2 4
Where d is the diameter of the blast barrier disc.

This total impulse is responsible for the movement of the blast barrier disc.



Momentum gained by the barrier, M = mdVd

Where Vd = Velocity of barrier,
md = mass of barrier

Distance available for the disc to travel = L-R

Where L is the distance between PC and MC

R is the distance from PC to barrier

Time taken for the disc to travel ta = L -

By equating ta time interval between initiation of PC and MC we get the value of R for a given mass of disc.As per the above
calculation, blast isolator would be positioned close to main charge.

Shaping of Blast barrier

Since the blast barrier is located close to the main charge, it has to be ensured that collapse process and jet formation of main
charge liner is not adversely affected. This is ensured by shaping the blast barrier in such a way that the barrier does not affect
the collapse process. Angle at which the liner elements from the base of the liner are projected to the axis can be calculated
using the principles of grazing detonation by Taylor. Details of this aspect can be found in reference [2].

Deviation angle is shown in Fig 2. can be computed from the Taylor formula;

V _ col
tan d
where V_col = collapse velocity of liner elements.

D = detonation velocity along the liner surface

Fig 2: Collapse geometry of Liner

According to the above approach, taking into consideration the mechanical integration aspects, a conical blast barrier is
envisaged whose included angle is 1000. Configuration and arrangement of barrier is shown in Fig:3.

Experiments were carried out to finalise the material and thickness of barrier. Initially, the tests were carried out with
barrier made of Aluminium. Inconsistent performance of warhead was recorded with Aluminium barrier. Further, Titanium
alloy Ti 6Al 4V was used as the material for barrier since it has high strength to weight to ratio and is capable of absorbing large
deformation energy per unit volume of the material. Warhead registered consistent performance with titanium barrier.
Subsequently, trials were conducted with dummy main charge and recorded the status of the blast barrier using Flash X ray
photography, at different intervals after precursor charge initiation. It is found that the barrier is intact and prevents precursor
debris from entering the collapse volume of main charge liner up to 350 microseconds. Details of these experiments are
explained in the forthcoming paragraphs. Nominal mass of the barrier design estimated to be 275g.

Fig 3: Configuration and mounting

arrangement of blast barrier



Further simulation studies were carried out to observe the collapse phenomenon of main charge. Main charge and blast
diverter were modeled in Autodyn using axisymmetric eulerian structured mesh. Simulation shows the smooth collapse of
liner elements. Some of the events in the collapse process are shown in Fig: 5

Fig: 5 Simulation of main charge liner collapse

Experiments were carried out wherein the precursor charge was detonated and Flash X ray photographs were taken at
different intervals till the time when main charge is set to initiate. Flash X- ray images show that the blast barrier is intact during
the time delay between initiation of precursor charge and main charge and it successfully arrests the debris of precursor blast
from entering the collapse volume of main charge. Main charge filled with non explosive material simulating weight and
dimensions were used. Trial set up is shown in Fig 6

X-ray film Cassettes

Test Article

Fig 6: Test set up

Blast barrier Precursor charge Blast barrier

Main Main
charge charge

T = 0m
s T = 250m

Main Main
charge charge
Blast barrier
T = 350m
s T = 500m

Fig: 7 Flash radiograph at different intervals from the initiation of precursor charge



As the present day explosive reactive armours incorporate highly insensitive explosives and thick metal covers, large
caliber precursor warheads with heavy fuzes are needed to defeat them. Moreover, third generation anti tank missiles
accommodate various electronic components within the space between main charge and precursor charge. Hence it becomes
imperative to protect the main charge for allowing a proper collapse of its liner. A light weight blast barrier has been designed
for a tandem warhead system and tested against a variety of targets. Shaping of the barrier is very important or else the barrier
itself can deteriorate the performance of the main charge. Above design of barrier weighs 270g and effectively protects against
a large precursor charge having 250g of explosive mass. It has been reported in literature [1] that an isolator weighing 300g has
been designed for protecting the main charge from a precursor charge having explosive mass of 230g. In comparison to this
design, the present isolator is lighter and effectively protects the main charge against stronger precursor charge.

The authors would like to express their gratitude to Shri Surendra Kumar, Director, ARDE, Dr. Satish Kumar, Director,
TBRL and Shri S.S. Mishra, Project Director-NAG for the support provided during the development of Tandem warhead
system and various diagnostic tests.


(1) M. Meier, P.VonAh, J. Sorensen, N. Ouye, Solutions to some design constraints in Tandem Warheads, Proceedings of
the 20th Inetrnational Symposium of Ballistics, Sept 2002, pp 663-668

(2) Ake Persson., A theoretical analysis of the mechanics of Tandem Shaped Charges and their interaction with different
targets, Proceedings of 7th Inetrnational Symposium of Ballistics, April. 1983, pp 251-260

(3) W.P. Walters and J.A. Zukas, Fundamentals of shaped charges, Wiley Interscience Publication, 1989.




Mohammadali Partanian, RN Pralhad
Defence Institute of Advanced Technology, Pune 411 025

Hollow charge HEAT projectile penetration into the target (Tank/Bunkers/Runways) has been
modeled from the concept of fracture mechanics and hydrodynamics theory. Initial development of the
model uses concepts of hydrodynamic theory for the part dealing with penetration of jet into the target,
concept of fracture mechanics has been used. The model has also been compared with the other basic
model developed by using the concept of Bernoulli's principle. The difference between the present and
earlier model has been highlighted. The model has also been computed for different target material and
different metallic liners. The results indicate that the present model yields better results.





PK Pradhan, PP Biswas, Dayanad, RN Bhattacherjee
Defence Research & Development Laboratory, Hyderabad
This paper describes the end game scheme for a Surface-to-Air Missile (SAM), capable of defeating wide range of aerial
targets (fighter aircrafts, Bombers, Helicopters, Cruise missiles, PTAs etc) in Air Defence (AD) role. Its kill zone envelop
covers low altitude - near boundary engagement to high altitude far boundary engagements. Ground based command guidance
system and powerful radars are ensuring the missile-target engagement in the entire kill zone envelop. However, it is seen that
classical engagement scheme of beam matching (i.e., fuze beamwidth to warhead beamwidth) is not adequate in defeating the
aerial targets even though minimum miss-distance is achieved during engagement. Following are the primary causes for
failure of target defeat due to ineffective delivery of WH fire power on to target.
i) Dynamic warhead beam width is varying as a function of missile velocity, angle of attack and its attitude.
ii) RPF beam is fixed beam depends upon antenna mounting.
iii) Targets are different types and its sizes are very much different from each other.
iv) Target velocity and its maneuvering state results different engagement geometry.
v) Different altitudes of engagement.
vi) Warhead characteristics
vii)Fuze characteristics
In End Game mission design, each and every factors listed above were addressed End Game algorithm formulated &
implemented in missile onboard towards ensuring kill of all types of maneuvering targets, bombers and cruise missiles etc. in
the entire kill zone envelope of Weapon System. This is a novel design.
SAM end game scheme is using an intelligent Radio Proximity Fuze (RPF) to aid the target kinematic information in real
time and End Game algorithm have been invoked in real time to detonate the warhead at an optimum time from the instant of
target detection by fuze.


End Game philosophy has been conceptualized based on the features of the following subsystem performances :
i) Possibility of resulting maximum miss-distance from ground based command guidance system.
ii) Radio proximity fuze (RPF) features viz., fuze beam and its lean angle.
iii) Pre- fragment MK-III warhead features (viz., beam width, hit density, lethal radius and its fragment kinematics).
iv) Maneuvering capability of targets.
v) Target sizes (starting from PTA to Bomber).
vi) Modes of engagement and kill zone envelope.
In past target engagement scheme has been worked out on warhead and RPF beam matching principle. It was anticipated
that all the time target will be in the beam of warhead at the instant of warhead detonation by RPF. However, it is seen that beam
matching is not feasible due to missile velocity variation, warhead fragment velocity variation over the range of engagement
and engagement geometry. In addition to the above, it is aimed to deliver the warhead firepower onto the vulnerable zone of
every target to maximize target kill. In turn, the SSKP of weapon system increases considerably.
Such engagement scheme revealed that there is a need of knowing the missile and target kinematics and its engagement
angle to choose the optimum time of warhead firepower delivery on to the vulnerable zone of target in real time. This calls for
modified guidance scheme and requirement of intelligent RPF having potential to provide target range and its velocity at the
instant of its detection.
Presently, SAM is using both the above features from guidance subsystem and RPF. Based on these features, end game
mission have been designed for the total kill zone envelope and an end game algorithm have been evolved and implemented to
choose the optimum warhead detonation time as per the following relation.
TOB = f (Rd, Vc)

Where, Rd : Target detection range by fuze

Vc : Classing velocity between missile and target.
TOB : Optimum warhead detonation time




Following salient features have been incorporated in various subsystems in order to achieve the desired end game

Engagement Geometry
Ground based command guidance system is ensuring the occurrence of desired engagement geometry in the total kill zone
envelope. A typical engagement scenario depicting the aspect angle is shown in Fig 1

Fig. 1: Typical Engagement geometry


Aspect angle

End game mission design ensures target interceptions for a maximum aspect angle upto 30 (i.e., the angle between missile
and target trajectory pertaining to missile terminal phase). This also ensures engaging maneuvering targets.

Fig.2 Target intercept geometry giving the warhead detonation position as per requirement.


30 F
Z [m ]

m T0 : Tgt. Detected by RPF
T1 : WH Burst Time
T2 : Target Interception

-5 0 5 10 15 20 25
X [m]

Radio Proximity Fuze (RPF)

The fuze is omni-directional with a width of 10 and maximum detection range of 45m. It is an intelligent fuze which detects
the target (as on when target is in its beam) and feeds the target position and velocity information to End Game algorithm.
Schematic figure of missile-RPF beam configuration is shown in Fig- 3.



Fig. 3 : Missile – RPF beam configuration

RPF Beam
RPF Unit
550 (a


Warhead (WH)
It is the pay load of SAM weapon system. It meets all following salient specifications required as per End Game mission
Weight : 55 kg
WH beam width : 70 to 100 (30)
Lethal radius : 50 m

Fig-4 Warhead


TA Cube
4. Warhead Fragment Kinematics
Warhead fragment velocity decay is considerable over lethal range. However, fragment has sufficient lethality to cause the
damage of target. Typical warhead fragment / kill-mechanisms velocity decay profile is shown in Fig-5.


Frag. Vel [m/s]

Fig-5 Warhead Kill Mechanism
1600 Frag. Velocity loss vs Decay
Velocity Decay Profile Actual


0 0.01 0.02 0.03 0.04 0.05

Time [sec]



5. SAM : Evolution of End Game Algorithm

End Game mission have been designed for the following parameters :
Ø Missile velocity : 500 – 750 m/sec
Ø Target velocity : 50 – 700 m/sec
Ø Engagement altitude : 0 to 16 kms
Ø Aspect angle : 30
Ø RPF characteristics : RPF detection range and closing velocity
Ø WH characteristics : Beam width, hit density, lethality
End Game algorithm have been evolved for all modes of target engagement (head on, receding & crossing) with consideration
of all the above parameters. The final algorithm have been worked out only in terms of two parameters, namely RPF detection
range and target closing velocity. The graphical representation of algorithm is given in Fig-6.

Fig.6: Graphical representation of end game algorithm

Closing velocity (RPF vs DTD[groupwise])
0.03 600
WH detonation time

0.025 1100

D T D (s )




Zone 1 Zone 2 Zone 3 Zone 4

5 10 15 20 25 30 35 40 45
RPF range(m)

RPF Detection range

Based on this algorithm a typical sea level target intercept simulation is shown in fig. 7.

Fig. 7 Sea level engagement, closing velocity: 900m/s Target maneuvering attitude = 15deg.

45 8

T2 6


30 T0
y [m ]
Z [m ]

20 0

15 -2


T1 -6
10 12 14 16 18 20 22 24 26 28
0 5 10 15 20 25 30 35 40 X [m]
X [m]




This algorithm is executed in real time by RPF as per End Game requirement at the terminal phase of missile trajectory. It
ensures target kill for all modes of target engagement and also enhances the SSKP of Weapon System.

1) Classical target engagement principle is not adequate
2) New End Game mission design principle ensures the following
a) Target kill in all modes of engagement.
b) Delivers WH lethality onto target vulnerable zone and enhances the target kill probability.
c) Target is defeated over entire range of miss distance.
In nutshell SAM is a modern-state-of-the-art weapon have all the advanced key features to defeat all varieties of aerial
target in Air-Defence roll.

1) R.M. Lloyd, “Conventional Warhead Systems Physics and Engineering Design”, AIAA, Vol.179, 1998.
2) J. Carleone, “Tactical Missile Warheads”, AIAA, Vol. 155, 1993.
3) R.M. Lloyd, “Physics of Director Hit and Near Miss Warhead Technologies”, AIAA, Vol.194, 2001.
4) TBRL Team, DRDL Team, “Evaluation of Fragment Kinematics for Cubical Projectile”, Internal Report
No.DRDL/DOS/WAD/512/5/07, 2007.




Kamlesh Kumar, Satwinder Singh, KPS Murthy

Armament Research & Development Establishment, Pune 411 021.

Development and deployment of thermobaric weapons have been increased in the recent past world over for defeating
tunnels or fortified underground targets. As the name implies thermobaric weapons are optimised for combined effect of heat
and over pressure on the target. Thermobaric weapons contain slow burning explosive that gets dispersed and detonated
simultaneously. Due to prolonged detonation time, total impulse on the target increases causing extensive damage.
Thermobaric weapons are much more effective against confined under-ground targets than over-ground targets due to re-
bouncing blast waves and sustained hot environment generated due to detonation of warhead over a significantly longer
For defeating under-ground targets the warhead has to be a penetrating type. The warhead penetrates the protective layer of
the target due to its kinetic energy and detonates inside the confined space.
This paper presents conceptual design of a 200 kg thermobaric warhead to defeat under-ground targets for a missile system.

Generally, conventional high explosive filled penetrating warheads / bombs are deployed against under-ground enclosures
like tunnels, caves, bunkers etc. are in vogue for holding war material / equipments, command and control centers, production
and storage of weapons of mass destruction (WMD), ballistic missile and artillery basing, military parking lots etc. These
warheads / bombs, produce only a localized damage effects even in confined volumes. In the recent past, Thermobaric
warheads, capable of causing damage over a large area in the confined volumes, are being pursued world over to defeat the
under ground fortified targets. US has developed a laser guided penetrating thermobaric bomb BLU-118/B with a weight of
2000lb having 560 lb of thermobaric composition and used in Taliban war, to neutralize several caves in eastern
The term “Thermobaric” is derived from two Greek words – 'Thermos' (means Temperature) and 'Baros' (means Pressure).
As the name implies Thermobaric warheads are optimized for combined effect of heat and pressure on the target. Thermobaric
warheads contain slow burning explosive compositions that keep their explosive impulse on the target for a longer duration.
Thermobaric compositions are essentially metal enriched and oxygen-deficient. Typical detonation velocities of thermobaric
explosives are in the order of 3–4 km/s. Unlike conventional HE warheads, the thermobaric warheads undergo simultaneous
detonation and dispersion. Only a part of the energy is released during the initial detonation phase, which generates “active
fuel-rich products” that undergo "after-burning" when mixed with the shock-heated air. The energy released through after-
burning increases the duration of blast overpressure along with fireball. The burning cloud is capable of penetrating into small
gaps, cracks, cleavages etc.

The comparative performance between high explosive detonation and Fig 1

thermobaric detonation are shown in Fig.1. The peak overpressure is much higher
for the conventional high explosive detonation (P2) than for the thermobaric
detonation (P1), but this pressure drops much more rapidly. The negative phase
results in a reversed-blast and is longer in thermobaric detonation than a high
explosive detonation. Therefore total impulse imparted to the target is much higher
than that of conventional explosive detonation. [2]

Thermobaric warheads for underground targets, are essentially Penetration

cum Blast (PCB) type of warheads containing fuel rich thermobaric explosives.
The target is neutralized by penetration of warhead into the target followed by
explosion within the confined volume. This paper presents conceptual design of a
200 kg thermobaric warhead to defeat under-ground targets for a missile system.



The thermobaric warhead system comprises of a casing filled with Thermobaric explosive composition and Post Impact
Delay Fuze. The design of warhead shall meet the following requirements.
w A sufficiently strong casing to withstand the stresses developed on impact and during the process of penetration.
w Accommodate maximum explosive contents for producing higher thermal and blast effects.
It is obvious that these two requirements are contradictory to each other. For strong warhead casing, thickness should be
considerably high, whereas the requirement of higher explosive content demands lesser casing thickness. At the same time
external configuration is to be selected in such a way that it will offer minimum resistance to penetration. Thus, the design
approach involves in (i) Design of casing (ii) Design of Burster charge and (iii) Selection of thermobaric explosive and its
A conceptual design of the 200 kg penetrating thermobaric warhead has been worked out for a missile system having
terminal velocity of 450 m/s with a space constraint of Dia 350 mm x 900 mm length. The warhead is required to penetrate a 1m
thick concrete covered with 1 m thick soil.

The design of casing is generally carried out using the empirical and semi empirical relations published in “Text Book of Air
Armaments Part- IV” and optimized using standard simulation models like LS Dyna, Dytran, Autodyn etc. Using this
methodology, the design of warhead casing (high strength steel alloy) has been worked out and the salient parameters are given
·Caliber - 310 mm
·Overall Length of warhead - 900 mm
·Total explosive contents - 63 kg with density of 1.75 g/cc
·Shoulder thickness - 20 mm
·Ogive length - 480 mm

The casing of warhead was analyzed for its integrity and penetration capabilities against targets having following
·Thickness of soil - 1000 mm
·Compressive strength of soil - 2.5 kg/sq.cm.
·Thickness of concrete - 1000 mm
·Compressive strength of concrete - 350 kg/sq.cm
From the penetration studies, it is observed that the warhead penetrates up to 2.1 m into the target and experiences a
maximum deceleration in the order of 12000 'g'. The time to penetrate the target is around 7.5 ms. The stress levels are around
880 Mpa.


The burster charge is placed centrally to detonate the warhead content and facilitate the dispersion of un-burnt metal
powders in surrounding zone. The burster charge breaks the warhead casing and disperses the thermobaric composition around
the point of burst. The volume covered by dispersed un-burnt explosive and metal particles will depend on quantity of burster
charge. Lower quantity of burster charge may not disperse the unburnt particles of explosive and metal over a wider zone. On
the other hand, higher quantity of burster charge will lead to low thermobaric explosive loading and result in lower
performance of warhead. Hence the quantity of burster charge will have to be optimized. Further the burster charge can be Al
rich which will under go completely anaerobic combustion and support setting up of initial hot environment required for “after
burning phase”. Based on the experience on burster charge designs for incendiary warheads, a 5 kg burster charge with the
following composition is finalised.




Thermobaric compositions are essentially fuel rich high explosives. Performance enhancement is primarily achieved by
addition of excess amount of micro-fine and nano size metal powders along with sufficient quantity of oxidizer to the explosive
composition. This combination will help the “after burning phase”, by sustaining high temperature ambience around the
warhead for a longer duration. The initial anaerobic combustion of micro-fine metal powder will serve the purpose of
prolonging the impulse time and sustain conditions for auto burning of micro fine and nano size metal rich explosive dust in
subsequent aerobic combustion stage.
In general there are two types of thermobaric compositions (i) Solid and (ii) Liquid based. Liquid based composition could
be either pure hydrocarbon fuels or gelled / slurried with metal powder and other ingredients. IPN is used as a liquid fuel for
liquid compositions due to its stability and ease of handling. Solid composition could be either cast or powdered Solid
compositions mainly comprise high explosives like RDX & HMX mixed with metal powders.
Liquid compositions have low filling density and thermal output compared to solid compositions. Powdered compositions
are not preferred for penetrating warhead systems due to density change / formation of discontinuities in explosive that occur
on impact and during the process of penetration. Therefore the cast compositions are the only choice for filling of penetrating
type of warheads. The filling density of cast compositions will be high with conventional filling techniques which can be
further improved by using advanced filling methodologies. The typical thermobaric cast composition contains following types
of ingredients -

Thermobaric cast composition – High Explosive + metal powder + Oxidiser + Binder.

High explosives for solid based composition could be RDX, HMX, TNT, PETN etc. RDX is preferred as compared to other
high explosives because it is readily available at low cost. The metal powders added to the basic composition should have low
burning temperature and produce high heat on combustion. Aluminium (Al), Magnesium (Mg), Boron (B), Zirconium (Zr) etc.
are the available options. Aluminium is preferred as it is readily available at low cost.
Generally the solid thermobaric compositions contain very high percentage of metal powders varying from 35 to 60%
[3] However the complete combustion of the metal powder present in the composition has been an issue, especially when
the percentage of metal powder in the thermobaric composition is on higher side. Poor combustion efficiency of metal
particles causes severe ineffectiveness of the weapon. This is due to high ignition temperature, 2000K, typically required
for proper combustion of metal (Al) powder. Low explosive quantity may not produce hot environment required for
complete combustion. The tests conducted by A Apparao, et al [4] using cast explosives with varying percentage of
micron-size Al powder, also reveal that the optimum percentage of Al powder is in the order of 40% only.
The hot environment required for full combustion of metal particles can be best maintained if it is supported chemically by
the combustion of oxidizers. 40-50% high explosive (RDX) mixed with 10-15% of oxidizers could be used for creating hot
environment required for complete combustion of metal particles. Further, the finer size of metal particles (nano) is preferable
for complete and easy combustion of metal particles. However, the use of nano size metal particles will result in low filling
density and shorter after-burning duration, compared to micro-fine metal particles. Therefore the size of metal particles will
affect the warhead performance. A mixture of micro-fine and nano size metal particles is recommended for complete
combustion and longer after-burning stage. Percentage of both size metal particles is to be optimized by conducting
In confined spaces, transition to full detonation is not required for enhanced blast, if the solid fuel is ignited early in the
dispersion process. A series of reflected shock waves generated by the detonation mixes the hot detonation gases with metal
particles and compresses the metal particles at the same time. These, actions provide the chemical kinetic support to maintain a
hot environment, causing more metal to ignite and burn. Hence the percentage of metal powders that can be mixed in
thermobaric composition for under-ground application could be substantially high and is in the order of 50 to 60% with
sufficient quantity of oxidizer.

The percentage composition of ingredients for thermobaric cast composition is given below –
Thermobaric Composition
RDX Al Binder Oxidiser Other
25-30% 40-50% 10-12% 10% 1-2%
The proposed thermobaric and burster charge compositions are for preliminary study which can be further fine tuned to
maximize the warhead performance.




To defeat under-ground targets the warhead need to penetrate the protected layers and then detonate within the confined
space. The warhead content is dispersed in the surround by detonation of centrally located burster charge while undergoing
anaerobic combustion. The temperature produced due to detonation of burster charge and anaerobic combustion of
thermobaric composition combined with the reflected shock waves assist in sustaining the requisite hot environment to
complete combustion of dispersed metal and explosive particles during the “after burning” phase.
For confined space applications cast explosives are preferred world over as compared to powder, slurried or gelled
compositions. RDX and Al are used as basic ingredients for cast composition along with oxidizer and binder. The percentage of
micro fine & nano size Al particles in the composition and quantity of burster charge shall be finalized after conducting few
The conceptual design carried out to defeat under ground fortified target protected by 1 m of concrete and 1 m of soil,
indicates that it is feasible to design a 200 kg thermobaric warhead with 55 to 60 kg of thermobaric composition.

The authors are grateful to Shri Surendra Kumar, Director, ARDE, Shri JS Gharia, Sc'G'(Retd), HEMRL and Shri A Appa
Rao, Sc'G', HEMRL for continuous support, guidance and encouragement provided during the work.

1 Internet Website - http://www.globalsecurity.org / military/systems/munitions/blu-118.htm
2. 'Aspect of thermo baric weaponry' by Dr Anna E Wildeggar-Gaissmaier, ADF Health Vol 4, Apr 2003
3 Advanced Thermobaric Explosive Compositions, May L Chan & Gary W Meyers, US Patent no.6955732 B1,
18 Oct 2005'
4. Evaluation of castable thermobaric explosive compositions for enhanced blast and thermal effects' by A Apparao,
R Samudra, P P Vadhe, R S Punekar, N G Wagmare and S H Sidi, VI International High Energy Materials Conference
& Exhibit (HEMCE), 13-15 Dec 2007


Launch Technology
SAM & Fuze
Armament Electronics


Sen A, Peyada NK, Subrahmanyam S, Ghosh AK, Wahi P
IIT Kanpur 208016

In this paper, a novel dynamic test rig is proposed. The paper describes the development of a test rig dedicated to
tests of artillery type models, rotating missiles and aircraft, in order to simulate roll, yaw and pitch moments of the models. A
regular test rig in the wind tunnel would be used for the purpose of deriving the static flight maneuvers. However, for deriving
the dynamic parameters, one would require the model to acquire a certain rate of translation or rotation, which would be
impossible to impart by means of static rig. The dynamic test rig uses simple principle of allowing the model to execute all
sorts of maneuvers in the tunnel as it would during the real flight.
In case the model is given a movement, it would have an additional burden of the rig to carry with it, and that would
not give us the actual scenario of action of the model. Hence aerodynamic compensator surface with adjustable arm length
will be used. The compensator would be used for negating out all the forces and moment terms arising due to the rig.
Furthermore, a theoretical description based on a six-degree-of freedom approach is developed for the motion of the model.
Part of the investigation concentrates on the characterisation of the roll, yaw and pitch damping behavior of the model in use
for the calculation of the dynamic behavior of the models used during the spin-down phase of flight. Development of such a
dynamic rig will add immensely in updating database for missile/rockets and aircrafts.

V = Velocity of aircraft
a = angle of attack
q = pitch angle
q = pitch rate
b = angle of sideslip
y = yaw angle
f= roll angle
Ixx = Moment of inertia about the X-axis
Iyy = Moment of inertia about the Y-axis
I zz = Moment of inertia about the Z-axis
m = Mass
V = Velocity
CL = Lift coefficient
CD /CX = Drag/Axial coefficient
= Side force damping coefficient
= Yaw damping coefficient
= Pitch damping force coefficient
= Pitch damping moment coefficient
u, v, w =Velocity components
pqr = Angular velocity components

In all previous Wind Tunnel Test rigs, the aircraft model would be tested for static stability parameters only, whereas
the dynamic stability parameters could only be estimated in real flight. This is because, generating the rates of rotation for the
aircraft would not be possible for a fixed configuration of the rig in the Wind Tunnel.
In other words, the aircraft model inside a Wind Tunnel, on a regular test rig, would be restricted in movement, and
will not be allowed to have the complete dynamics of its own as it would in real flight.



Over the ages, attempts have been made1-7 to impart the elusive degrees of freedom to the aircraft model inside the
Wind Tunnel and replicate its exact 6-DOF flight inside Wind Tunnel. There is however an inherent disadvantage for this.
Most Wind Tunnels have a space constraint and hence cannot have the forward degree of freedom along the X-axis. The
forward velocity is usually fixed as per the capacity of the Wind Tunnel. Pitch and Yaw8-10 degrees of freedom have also
separately been imparted over aerospace vehicle models. Efforts have been concentrated on having 1 Roll degree of
freedom11-16 additionally, along with the forward velocity. However, these would only let the model achieve upto 2 or 3
degrees of freedom. A breakthrough in imparting several degrees of freedom to the test model was achieved in NASA17-21
where a total of 4 degrees of freedom was imparted to the model.
An enhancement of the number of degrees of freedom was achieved at Bristol by Prof. M. H. Lowenberg , where a
unique rig called the Pendulum Support Rig (PSR22-23) was developed to study the dynamic derivatives of the aircraft using
dynamic maneuvers, and a total of 5 degrees of freedom were achieved (Fig-1). However, the calculations for such a model
would be very resource intensive, and the manufacturing would be very difficult.
The present paper is inspired by the philosophy of the work at Bristol, where a novel attempt has been made to
estimate the dynamic derivatives by allowing the aircraft to have 5 degrees of freedom like the Bristol model, but with
enhancements, so as to make the manufacturing and calculations easier and inexpensive.

As described in Fig. 2, the model, christened H-model, due to its appearance, would allow the aircraft (a/c) scaled down
model to pitch and yaw about a 2 DOF gimbal G at its CG, and the rig arm would take a roll, along with the aircraft, about joint J.
The joint J would be collinear and coaxial with the CG of the aircraft model.
In addition, heave would be provided by the movement of the horizontal arm HH over the two vertical rails VV1 and
VV2. Sway would be provided by the sliding movement of the joint J along the horizontal arm HH through the slider rail
provided along it.
In case the aircraft is given a movement, it would have an additional burden of the rig to carry with it, and that would not
give us the actual scenario of action of the aircraft. Hence an aerodynamic compensator surface A with adjustable arm length C
will be used. The compensator would function as any normal aircraft control surface, with the capacity of being deflected by the
The compensator would be used for negating out all the force and moment terms arising due to the rig, along with its
Second Law effects.
The difference can be seen from Fig.1 and Fig.2, between the two models. The PSR22-23 model imparts the linear
velocities of heave and sway through the pitch and yaw degrees of freedom at the joint J. Also, the rig has a non-conventional
design, and cannot be procured off the shelf.
The differences in calculation can be found once the equations of motion are developed for both the cases.




From the above Fig. 2, the following can be seen:

1. u is given in the global coordinates( u e ) by the wind tunnel

2. ve and we is given by movement of J on HH over the slider rail, and movement of HH on VV1
and VV2 respectively
3. The velocities seen are of the global coordinate system because they are being imparted on NON
4. The rig is aligned such that its c.g., the point J, and the aircraft c.g. are all collinear.

é ép&
w êú
qú & ê
5. The aircraft sees its a/c =
ê and consequently w
a/c =
ê ú.

ëû r&
ë ú

é p&
é ù
the rig sees w
ê ú & ê

6. rig =

ê and consequently w
rig =
ê ú
ë ú
û ê
ë ú
ue ù
7. velocity of the aircraft is Va / c Ta¢
= ê
ve ú
w.r.t aircraft body axes, where Ta / c is the
/c ê ú
we ú
ë û
24 ï
transformation matrix from body to earth fixed axes, with all q
angles í
ý, and Trig is the
transformation matrix from body of rig to earth fixed axes with angles 0ý


8. Velocity of the rig will be taken also at the axes of the c.g. of the aircraft, with the assumption #3 in place. It would
be interesting to note that the rig will see all the velocities that the aircraft sees. However H will see only ue and we. Vv1 and
VV2 will see only ue, but they are fixed, so we will not consider their movements, and their aerodynamic forces will be
considered to be already taken care of. The aerodynamic forces of HH will also be considered as already taken care of. It is to
be noted that for HH and VV1 and VV2, Drag will be the primary aerodynamic force. The moments caused due to their
aerodynamic forces will be combined with the aerodynamic moments of the rig. This is an assumption for ease of
understanding of the rig equations and can be trivially added to the equations at any point of the calculation without any
One very interesting thing is to note that we shall not bother about the compensation of the drag in this case (Fx
terms). This is because they can be statically accounted for, and drag due to dynamic degrees of freedom is negligible.
Also we can see that moment due to drag terms will be very less as the z-arm of the moment is negligible in all cases
(zero, theoretically).
We can also safely assume that u& e=0 for all cases since it is very difficult to vary the tunnel velocity dynamically.
Now the forces24 on the entire system (rig, aircraft and horizontal HH)can be calculated.



The force acting on the aircraft body, rig and compensator is summarized in Eq. 1.
r é ù
Fa / cbody ùê ú
ê ú ê ú
ê úma / c (
ê V& + w ´ V ) ú

ê =ê
a/c a/c a/c (1)
ê ú
ê ú
ê ¢ ú
+ úmrig (
ê TrigV&a/c +w rig
Trig Va / c )
rú ê
+ ú
ë û
C úê ú
ê é u& ù
e ú
mH Ta¢ ê
0 ú ú
ê /c ê ú ú
ë ë &
w e
û ú

The first row of Eq.1 on LHS and RHS constitute the main aircraft equations which can be found in any standard text

However when we look into the 2nd and 3rd rows we find they are the “spurious” terms in a given aircraft equation of motion.
Hence these are the terms we would like to negate in order to let the aircraft fly in an environment akin to free air. So we can
é ù
ê & ú
éùê mrig ( ¢
Trig Va / c +
Trig Va / c )
ê úê
F + ú
êrig ú
ê ú
+ ú
= ê é u& ù ú (2)
r úê ê ú
FC ú
ë ûêmH Ta¢/c ê0 ú ú ú
ê ê
ë ë w&eû
ú ú
So from the above Eq. 2, we can see that since all the other quantities are known or are measurable, we can calculate .
So Eq. 2 is a design condition for the compensator.

Moment Equations24 follow the same philosophy similar to Force Equations. However, when only the aircraft rotates
about G, it does not see any effect of the rig. When the aircraft and rig as a whole rotates about the aircraft x-axis, about point J,
then affects the motion of the aircraft.
M a/c ù
êúé I a / c- &
cg w a/c + w a/c ´ ( I a / c- cg wa/c ) ù
ê ú T ¢Iê ú(3)
ê ê (
rig rig - &
w + w ´ (T ¢ I ) w ) ú
M rig ú
cg _ a / c rig rig rig rig cg
- _ a / c rig
êú ê ú
+ ú ë û
ê ú
MC ú
ë û



Where all Moments of Inertia and aerodynamic moments are calculated about the cg of the aircraft.

mrig ( x) 2 =
I rig rig _ cg + I rig -
cg _ a / c (4)

Where x is the horizontal distance between rig cg and aircraft cg.

The first row on LHS and RHS are usual terms of an aircraft equation of motion.
Hence the design condition for the compensator would be24

M rig ù
êúé T¢
rig I
= ê rig -
( &
cg _ a / cw
rig +
w ¢
rig ´ I rig -
cg _ a / cw
rig )
ù )
ê ú ú (5)
ê úêë ú
MC û
Hence Eq. 5 is the design criterion in addition to Eq. 2.

All cross product terms can be expanded in a manner similar to usual aircraft equations of motion.

The compensator design would thus depend upon:

1) dimensions of the rig involved
2) material properties of the rig
3) aerodynamic characteristics of the rig
4) dynamic roll

We can see that if there is no roll, it can be verified from the equations that the rig plugs on to the aircraft model like a
dumb mass.


In the above drawing, we can see that the only difference is at the end of the rig. Here, the ve , we are being given
by means of the rotation of the big joint J. So instead of ve , we we will write ve =a/c
y J ´Ra / c for a/c and ve =
J ´ Rrig for
rig, where is the vectorial distance from c.g. of the body to J. Similar is the case for we . This is a typical example of a system
which has higher velocity near the edge of the rotational radius and lower velocity closer by.
In addition, since it is quite impossible to put the compensator unit away from the body and/or keep it on the body
without making the compensator unit acquire a different velocity, the velocity of the compensator unit will also depend
on/change accordingly with the moment arm of its compensating surfaces as measured from J.
This indeed would prove to be tedious for our future plans of designing a control system suitable to operate the
Now with the supposition that we are in a position to accept such difficulties, we proceed to derive the subsequent

Fa / cbody [ f ( R 2 a / c )] ù
ê ú
ê ú ma / c (
é V&a/c + wa/c ´ Va / c ) ù
r 2
ê ú ê ú
F [ f ( R )] +
ê ú (6)
êrig rig =
+ ú mrig (
ê ¢
Trig &
Vrig +w rig Trig Vrig )
¢́ ú
ê ú ê ú
r 2 ê ú
FC [ f ( R compensator )]ú ë û
ë û



and for the moments24:

M a / cbody [ f ( R 2 a / c )] ù
ê ú
ê úé I a / c-cg w
& a/c + w a/c ´ ( I a / c- cg wa/c ) ù
ê ú ê ú
M rig [ f ( R rig )]
ú= T¢
ê ( rig I rig -cg _ a / c
wrig + w rig ´ ¢
(Trig I rig -
cg _ a /)c w
rig ) ú (7)
+ úê ú
ê ú ë û
ë C [ f ( R compensator ú)]û

Where all Moments of Inertia and all aerodynamic moments are calculated about the cg of the aircraft, as the above case.

mrig ( x) 2 =
I rig rig _ cg + I rig -
cg _ a / c (8)

Where is the horizontal distance between rig cg and aircraft cg

It is interesting to note the dependencies of each of the LHS terms w.r.t. the distances from point J.

The rotations at point J, except roll, which are meant for imparting velocities, will NOT be considered in any other

Hence after this comparative study, we wish to follow the H-model configuration for further study and development.
The present study would not be complete without a study of the dynamic controllability of the H-model, as would be provided
by the compensating surface. This is presented in the next segment.


When the rig is designed for static purposes, then it seems that the function of the compensator is redundant, and the
same can be easily achieved through a dead weight compensation.

However, the main advantage of the rig comes in a dynamic case, as illustrated hereunder.

Fig. 3: Diagram showing layout of compensator

Let us consider from figure-3 that the compensator has a horizontal fin and a vertical fin, both of which can be
individually or simultaneously deflected. In addition, the length Rc can be changes in order to provide suitable moment arm
to compensate the “spurious” moment terms.



So we have a total of 3 controls in our hands, the deflection of the horizontal fin of the compensator(dh), deflection of the
vertical fin of the compensator (dv) and Rc, the moment arm of the compensator.

So from eqn (2) we can see that:

é ù
éùê ú
ê úêmrig ( ¢
Trig V&a/c +
Trig Va / c )
Frig ú
ê = ê
+ ú
+ úê ú
r úê é u& ú
e ù
FC ú
ë ûê ê ú ú
mH Ta¢
ê /c ê 0 ú ú
ê ê
ë ë w& ú

Which we can further break down as:

é ù
ê ú
ê r ú
-Frig ú
ê ú
ê ú
rê ú
] mrig (
FC = ê ¢
Trig V&a/c +
Trig Va / c )
ú (9)
+ ú
ê ú
ê é u&
e ù ú
mH Ta¢ ê 0 úú ú
ê /c ê ú
ë ë ê w& ú

If we closely observe the RHS terms we shall see a term . w
Trig Va / c
This term implies that the DYNAMIC degree of freedom of the rig also has to be suppressed. So we have to select the
compensator airfoil profile judiciously such that the C Land CY terms take care of the rig dynamic characteristics.
q r

Further, for the moments, we can revisit Eq. 5.

M rig ù
êúé T¢ (
rig I rig - &
cg _ a / cw
rig +
w ¢
rig ´ I rig -
cg _ a / cw
rig )
ù )
êú ê ú
ê úëê ú
MC û
ë =
Which can be simplified as
-M rig ù
ê ÷
r ê ç
+ ÷
] ç
MC = ê
T¢rig I rig - & ¢
(Trig I rig -
è cg _ a / cw
rig +
rig ´ cg _ a / cw
rig ) ø
ê ú
ë û



So we have to select the airfoil profile of the compensator judiciously such that theCmand Cnterms
q r
take care of the rig
dynamic characteristics. In addition, dhmax and dvmax are to be determined as well.
The above statements give the outline for development of the aerodynamic compensator of the rig, along with the
mathematical models describing the H-model layout. When all these factors are compensated, the design of the aerodynamic
compensator would be taken as complete. However, arriving at a particular design would involve a lot of iteration cycles.

In this paper, a novel dynamic test rig has been proposed and the rationale has been portrayed. The rig has been analyzed
mathematically and has been compared to an existing dynamic rig (PSR) configuration. The advantages of the rig layout have
also been brought out. The compensator has been analyzed and its design constraints have been identified. As future work, the
complete structural design of the rig layout and the detailed aerodynamic design of the compensator and other controllers (if
any) has been identified.


1. Chaturi Singh and K. Poddar, ”PXI-based High-Performnance Versatile Instrumentation and Motion Control
System for Wind Tunnel Applications” National Wind Tunnel Facility, Indian Institute of Technology, Kanpur-
2080 16, India.
2. Lars E. Ericsson, “Conceptual Fluid/Motion Coupling in the Herbst Supermaneuver” JOURNAL OF AIRCRAFT,
Vol. 34, No. 3, May June 1997.
3. Scott M. Murman, “Reduced-Frequency Approach for Calculating Dynamic Derivatives” AIAA JOURNAL ,Vol.
45, No. 6, June 2007.
4. Mark F. Reeder and Walter Allen, “Wind-Tunnel Measurements of the E-8C Modeled with and Without Winglets”
JOURNAL OF AIRCRAFT, Vol. 45, No. 1, JanuaryFebruary 2008.
5. Peter Eliasson, “Investigation of a Half-Model High-Lift Configuration in a Wind Tunnel” JOURNAL OF
AIRCRAFT ,Vol. 45, No. 1, JanuaryFebruary 2008.
6. Martin E. Beyers, “Direct Derivative Measurements in the Presence of Sting Plunging” JOURNAL OF AIRCRAFT
,VOL. 23, NO.3, MARCH 1986.
FROM DYNAMIC TESTS” AIAA Atmospheric Flight Mechanics Conference and Exhibit 11 - 14 August 2003,
Austin, Texas.
Mechanics Conference and Exhibit 11 - 14 August 2003, Austin, Texas.
9. Brian D. Dykas and Daniel W. Tellier, “A Foil Thrust Bearing Test Rig for Evaluation of High Temperature
Performance and Durability ” ARL-MR-0692 April 2008.
10. Peter G. Hamel, ” Birth of Sweepback: Related Research at LuftfahrtforschungsanstaltGermany” JOURNAL OF
AIRCRAFT, Vol. 42, No. 4, JulyAugust 2005.
11. “Free-to-Roll Testing of Airplane Models in Wind Tunnels” NASA Tech Briefs LAR-17153-1.
12. “Identification of Dynamic Systems -Applications to Aircraft Part 2: Nonlinear Analysis and Manoeuvre Design”
AGARD 300 flight test technical series - Volume 9.
13. Khanh Nguyen and Benton Lau, Ames Research Center, Moffett Field, California “Dynamics of the McDonnell
Douglas Large Scale Dynamic Rig and Dynamic Calibration of the Rotor Balance ” NASA Technical Memorandum
14. LEWIS B. SCHIFF* AND MURRAY TOBAK, ” Results from a New Wind-Tunnel Apparatus for Studying Coning
and Spinning Motions of Bodies of Revolution ” AIAA JOURNAL VOL. 8, NO. 11, NOVEMBER 1970.
15. Mehmet Altun Ýbrahim Iyigun, ”Wind Tunnel Experiments for Maneuvering Aircraft” AGARD ADVISORY
NUMBER 3 (19-27).




17. Gerald N. Malcolm and Lewis B. Schiff, “Recent Developments in Rotary Balance Testing of fighter aircraft
Configurations at NASA Ames Research Center” NASA Technical Memorandum 86714 , July 1986.
19. Lars E. Ericsson, ” Wind-Tunnel Aerodynamics in Rotary Tests of Combat Aircraft Models” JOURNAL OF
AIRCRAFT Vol. 35, No. 4, JulyAugust 1998.
20. Martin E. Beyers, “Implications of Recent Rotary Rig Results for Flight Prediction” JOURNAL OF AIRCRAFT
Vol. 37, No. 4, JulyAugust 2000.
21. Gerald N. Malcolm, “New Rotation-Balance Apparatus for Measuring Airplane Spin Aerodynamics in the Wind
Tunnel” 264 J. AIRCRAFT VOL. 16, NO. 4.
22. Lowenberg, MH & Kyle, HL., 'Development of a pendulum support rig dynamic wind tunnel apparatus', AIAA
Atmospheric Flight Mechanics Conference, Monterey, USA, 2002.
23. C. D. C. Jones, M. H. Lowenberg, and T. S. Richardson, “Tailored Dynamic Gain-Scheduled Control” JOURNAL
OF GUIDANCE, CONTROL, AND DYNAMICS, Vol. 29, No. 6, NovemberDecember 2006.
24. Robert C. Nelson, “Dynamics of Flight”, Chapter 6, Published by Wiley and Sons.




MV Rao, VS Mahalingam
Centre for Artificial Intelligence and Robotics, Bangalore

Pinaka weapon system and Artillery Combat Command and Control System (ACCCS) have
been accepted by the Artillery, Indian Army for induction after extensive field trials. The
production/deployment is in progress.
The ACCCS enable the arty Commander to plan the fire mission using inputs like gun parameters, target
location, shell/fuse combinations, prevailing meteorological conditions, etc. the major functions
supported by ACCCS are Technical fire control, Tactical fire control, Fire planning, Deployment
management etc. the major effort of technical fire control involves calculation of Trajectory parameters
by Point Mass and Modified Point Mass models. ACCCS communicates with higher echelons for
decision making and approval procedures as per the established practices/policies in the Indian Army.
The fire plans are prepared using Geographical Information System (GIS) on a digitized military map
(including military symbols), which enables the commander to visualize the battlefield scenario,
analyze the situation, probable area for deployment of guns, etc. All existing weapon systems of
Artillery like 105mm, 105 IFG, 120mm, 130mm, 122 Grad-BM21 and 155mm etc. have been integrated
into ACCCS.
PINAKA weapon system has to be integrated into ACCCS as yet another weapon system. Suitable data
transfer formats and procedures were finalized for smooth integration. The ACCCS software was
revised accordingly to accommodate the data formats and procedures being used in the Pinaka system.
A Communication Interface Unit (CIU) designed and developed by CAIR provides the robust
communication over the noisy channels of Combat Net Radio (CNR). The communication between the
BCP and the Battery of six launchers were customized to suit the requirements keeping the ACCCS
Communication procedure with higher echelons in tact over the Radio as well as 2-Wire Line.





Salutagikar CS, Mukane SB, Pravaker RC, Pawale SO, Rajan KM
Armament Research and Development Establishment, Pune 411 021

The role of an Artillery Rocket is to neutralize enemy concentrations, to destroy POL depots and 'B' vehicles , with
minimum collateral damage. To meet these requirements the accuracy and consistency become important aspects for artillery
rockets. Dynamic balancing becomes imperative to ensure performance improvement for any free flight artillery rocket. In
reality achieving a desired level of dynamic balancing for artillery rockets becomes a challenging task due to various aspects
like manufacturing inaccuracies, assembly intricacies and different sources of manufacture of sub-systems . The paper
brings out an innovative methodology of dynamic balancing of an artillery rocket to improve its performance. It also gives the
constructional features for balancing mass and its integration in the rocket hardware. Finally, detailed flight data is presented
to highlight improvement in accuracy achieved at target end with known unbalance and also with unbalance mass corrected.
This methodology has enormously helped in smooth integration of rockets at production centers and faster delivery of
complete rocket system to the Users.

An Artillery rockets are generally free flight all weather rockets fired from a Multi Barrel Rocket Launchers
(MBRLS). A free flight rocket and a MBRL are shown in fig. 1(a) and fig. 1(b). The accuracy and consistency of free flight
rocket is generally in the range of 1.5 2 % of its maximum range. It has been seen that increase in unbalance mass in flight will
make the dynamic axis of the rocket different from its geometrical axis. The angle between these two is to be minimized for
reducing the lateral deviation. Also, higher unbalanced mass in the case of spinning rockets leads to undesirable stresses due
to centrifugal force.

The dynamic unbalance 'U' is the product of unbalance mass and it's radius at which it is located.

It is almost impossible to make a dynamically balanced rocket. The unbalance of any system is because of
any of the following three reasons: i) Inhomogeneity of material ii) Machine inaccuracy involved iii) Non-alignment in
integration of sub-assemblies.



There are three types of balancing methods, i) Static balancing i.e. unbalance measured in a plane keeping on roller,
ii) Couple or Dynamic balancing , i.e. unbalance measured in different planes and iii) Combined balancing i.e. combination
of static and dynamic balancing. In actual practice the combined balancing is followed, which is measured on balancing
machine. It has been established in the theory of balancing that combined static and dynamic unbalances of a rotor can always
be represented by two unbalances in any two desired planes. Static balancing by gravitational means, which is done by
rolling the unbalance part on it's mandrel along two parallel horizontal ways or knife edge, can only detect the static portion of
unbalance (for single plane separation). Static balancing can be justified only for slowly rotating rotors having L/D ratio less
than 1, because the dynamic unbalance if present, will be too small to be of practical significance. On the other hand when the
L/D ratio is more than 1, it will not be sufficient to balance static unbalanced forces but also dynamic unbalanced couple must
be balanced. As the length to diameter (L/D) ratio of an artillery rocket increases the dynamic unbalance increases because of
non-alignment of axes in integration of sub-assemblies and shift in radial direction of center of mass from geometric axis
thereby the unbalance angle between dynamic axis and geometric axis increases.


Dynamic unbalance is the result of the longitudinal axis of inertia (also called dynamic axis) not being the same as
the geometric longitudinal axis, i.e. the weight distribution is not uniform about the geometric axis. If the rocket is perfectly
dynamically balanced then geometric axis and dynamic longitudinal axis will be coincident and the Principal longitudinal
moment of inertia will be equal to longitudinal axis moment of inertia and all the product of inertia term will be zero. But it is
practically impossible to manufacture a complete dynamically balanced rocket. For dynamically unbalanced rocket, product
of inertia term will come in to picture. Due to dynamic unbalance of rocket there will be increase of structural load on the
rocket. The more the dynamical unbalance the more will be the angle between geometric axis and dynamic axis and so more
will be dispersion in line. To calculate the angle between geometric axis and dynamic axis, longitudinal axis moment of
inertia and product of inertia of rocket should be known. A solid model of Pinaka in IDEAS-11 is taken to calculate the
longitudinal axis moment of inertia and product of inertia of rocket with different unbalance masses at the two selected
planes. For dynamic balancing of rocket there should be at least two balancing planes. Two planes are selected such that one
lies left to centre of gravity and another right to CG as well as where one can add or remove mass for balancing. For the case of
Pinaka rocket both plane1 and plane2 are separated at a distance 1754 mm and the first plane is at a distance 1450 mm from
nozzle end. To keep the accuracy and consistency within the specified limit (1.5-2% of range) the specified value of dynamic
unbalance is maximum 350 gm at a radial distance of 107 mm based on performance analysis of the system.

Fig. 2 Balancing mass in specified planes

A Pinaka solid model is taken for simulation of dynamic unbalance mass. On the solid model two planes are selected
as shown in figure above. Two unbalance masses assumed in the form of very small cylinder (diameter 1 mm and height 1
mm) is taken. One unbalance mass in each plane is placed opposite to each other at a radial distance of 107 mm i.e. at
periphery of rocket. This is shown in fig. 2. Unbalance masses are placed to each plane since it will give the maximum product
of inertia and hence the maximum angle between geometric axis and dynamic axis for a given unbalance mass in both planes.
Simulation for different unbalance masses are done by varying the density of cylindrical unbalance masses. So different
unbalance is in both planes are found. We simulated for 200 g, 350 g, 500 g, 750 g and 1000 g of unbalance at a radial distance
of 107 mm in both the planes. The product of inertia found out for different unbalance masses in both planes are given in
Table 1.



Table-1: The Product of inertia values with respect to different unbalance in specified planes


Suppose X-axis is the longitudinal axis and Y & Z axes are the transverse axis of the rocket as shown in the fig. 3.

So X- axis will be the axis of symmetry and it is called Geometric axis. Y- Axis is so selected that unbalance
mass will put in XY plane. Now X-axis will be the axis of symmetry or geometric axis.

Ixx = Moment of inertia about X axis.
Iyy = Moment of inertia about Y axis.
Izz = Moment of inertia about Z axis.
Ixy = Product of inertia about X, Y axis.
Ixz = Product of inertia about X, Z axis.
Iyz = Product of inertia about Y, Z axis.

The free flight rocket rotates about one of principal moment of inertia axis called Dynamic axis. The principal
moment of inertia can be calculated by equation given below.




For the nominal mass condition the moment of inertia is calculated by solid modeling of Pinaka rocket. The
value of moment of inertia about longitudinal axis is 1.82 kg m2 and about transverse axis is 451.08 kg m2. The angle
between geometric axis and dynamic axis is calculated by equation (2) given below.

Ö = Ixy / (Iyy - Ixx) (2)

Applying the above formula we calculated the angle between dynamic axis and geometric axis for the Pinaka rocket
for different unbalance masses. The calculated angles are listed in the table-2.

Table-2: Angle between geometric axis and dynamic axis and deviation in line at range of 32 km with
respect to different unbalance masses.

The above data is represented in graphical form in fig. 4.




Dynamic balancing machine

There are mainly two types of balancing machines. They are, i) hard balancing machine ii) Soft balancing machine.
The soft bearing machines are sensitive to mass of the rocket and calibration will be required every time. Hence in balancing of
rocket hard balancing machine are used as not often calibration is required and the same calibration can be used for different
size and shape rockets with minor changes in position of supporting structure. In hard bearing machines the natural frequency
of supporting system is five times the natural frequency of balancing speed of the machine in order to ensure the bearing
bridges are kept rigid. For the present study hard balancing machine is used. Measuring system of the machine is given below.
i) Electrical measuring system using two watt meters and manual operation phase transmitting drums graduated in
3600 for knowing the amount as well as position of unbalance.
ii) Electronic measuring system incorporating entirely solid state circuitry : The measuring device is provided with
two unbalance indicating meters for two unbalancing planes and position of unbalance is indicated by 'stroboscopic light'
iii) Electronic measuring system using 'Analogue computer' incorporating entirely solid state circuitry with
integrated solid state circuitry. The measuring device is provided with one set of angle meters and unbalance indicating meter
for left plane and right plane.

Testing procedure
The main objective of the testing is to find out any unbalance mass of the rocket. Figure 5 explains parameters of
unbalanced rotor with supporting planes and correcting planes. Resultant unbalance masses (Ma and Mb) and their
orientations in two desired planes ( A and B) are shown in figure 5.
A rocket on dynamic balancing machine is shown in figure 6. Location of support rollers and correcting planes are
shown in figure 7
Correction plane for dynamic unbalance
In tube launched artillery rockets, attaching a correction balancing mass is always a challenging task, due to space
constraint, as the design of propulsion system is arrived at with optimum external configuration and thickness of components
with only sliding clearance of approximately 0.5 mm radial clearance between rocket body and its launch tube. Also, as the
rocket thrust chambers are high strength thin wall tubes, making provisions for dynamic unbalance correction is rather
difficult. So manufacturing of rocket within the geometrical tolerances is of utmost importance in order to reduce
misalignment at joints of rocket sub-systems to a lesser degree leading to low amount of unbalance mass and provision of



attaching correcting balancing mass for the same. The misalignment and unbalance lead to shift in the physical properties, i.e.,
principal mass moment of inertia in transverse and axial directions. The geometrical axis and the dynamic axis will be
differing considerably. These effects will contribute in wider dispersion of rockets at the target end.
It is really a challenging task to add the correcting balancing mass in the rocket system. Balancing by removing the
unbalance mass is not feasible in the rocket system as the airframe is designed with low structural factor and low level factor of
safety (1.4). The rocket propulsion system under study consists of a thrust chamber with twin motor coupled assembly tubes
with optimal thickness. The motor tubes are manufactured through flow forming technique wherein correction for unbalance
mass is not feasible. Also adding the correcting balancing mass in the aft stabilizing system is also not practicable because of
complexity of design and loading constraints. This creates limitation for adding the correcting balancing mass in some
particular portion only. In the present case study in the front adaptor of propulsion system, provision for the correction of
balancing mass is considered.
Attachment of balancing mass
Unbalance meters and angle meters on the balancing machine display the amount of unbalance mass in both
correcting planes and their orientation respectively. Fig. 5a shows unbalance mass reading and their orientation from
balancing machine. Resultant balancing force is calculated by force triangle as shown in fig. 5b. The correcting balancing
mass obtained is attached in the resultant orientation as derived from the unbalance masses and their orientations from both the
correcting planes. Methodology and constructional features of correction is shown in fig. 6. Out of two planes one correcting
plane is on thrust chamber, where no correction for unbalance is feasible. Other correction plane is at front end adapter where
inside the cavity, addition of correcting balancing mass is feasible and this is shown fig. 6. The maximum radius at which
balancing mass is to be fixed is constant, because of space constraint. The dynamic balancing rotational speed is also fixed and
the reference distance of plane in which correction is made is also fixed. So, considering these fixed parameters, it is simply a
parallelogram of unbalance masses in both the planes, which is shown in fig.8.




A typical result of the testing showed an unbalanced mass of 459 g at orientation of 170o in the left plane i.e. near
front adaptor and 394.3 g at 271o in the right plane (at the nozzle end). Correcting balance mass of 750 g at orientation of 2170
is calculated. Due to limitation of space in the hardware cavity the correcting mass (290 g at 217o) little lower than the
calculated one is placed in the hardware cavity of front adaptor of propulsion system at the left plane. After correction
dynamic balancing of the rocket was carried out. Result of the testing showed an unbalanced mass of 344.9 g at orientation of
186o in the left plane and 286 g at orientation of 316o in the right plane. Thus unbalanced masses at the both planes are kept
below the maximum acceptable limit of 350 g. Subsequently, during flight trial it was revealed that the rockets with
correction in unbalance mass showed lower lateral dispersion than the uncorrected one. For a typical results discussed,
dispersion of the rocket reduced to 350 m as against 700 m and 300 m gain in the range.



Deviation in line of a rocket is the effect of many parameters like unbalance mass, wind effect, launch errors, thrust
misalignment etc. Wind effect is taken constant for all the rockets fired in series in a short time interval. The aiming error of
the launcher is maintained below 3 arc minutes with the incorporation of Automatic Gun Alignment and Pointing system
(AGAPS). The dispersion due to residual launch errors may also be taken constant for a given launcher. Hence it can be
concluded that the amount of dispersion is directly proportional to the magnitude of rocket unbalance mass.
The rocket unbalance mass Vs dispersion in line from a fired series is given in table below:

Table 3: Rocket unbalance mass Vs dispersion



Artillery rockets with high unbalance mass experiences coning effect leading to high drag and subsequently reduced
range and increased dispersion. By incorporating suitable unbalance mass correction provision in the rocket hardware, the
original unbalanced mass could be brought down within acceptable limits thereby reducing the errors in accuracy and
consistency. The methodology presented has been found very effective in productionisation of Pinaka rockets at Ordnance
Factories which in the absence of unbalanced mass correction would have led to considerable rejection of rockets.


1. Dynamic Balancing of rotating machinery; Wilcox J.B.; Sir Isaac Pitman & sons Ltd., London; 1967.
2. The Theory of Machines; Thomas Bevan; CBS Publishers & distributors, Delhi; Third edition, 1984.
3. Balancing of high speed machinery; Marks S. Darlow; Springer- Verlag, New York Inc.; 1989.
4. Machinery vibration balancing; Wowk Victor, McGraw Hill Inc, 1995
5. Rotating machinery: Practical solutions to unbalance and misalignment; R. B. McMillan, PE, CEM; The
Fairmont Press Inc, Lilburn, 2004.
6. Roy, P. K., and Bankhele, S. D. (1981). On some aspects of dynamic balancing machine, The Engineer, pp. 2-14.
7. Dynamic machine manufacturer manual, ABRO.




Mahmoud Zarrini, RN Pralhad
Defence Institute of Advanced Technology, Pune - 411 025

Wound ballistics deals with wounding capacity of the bullet in the human body. In the proposed
study, impact of bullet on the human body has been investigated.
The model has been developed from the concept of underwater explosions with appropriate changes to
the present model requirements. The analysis has been restricted to the estimation of shock velocity and
pressure created by the bullet as a result of impact on the body. Basic concepts of fracture mechanics
have been incorporated while developing the model. The model has been compared with the other basic
model and its advantages over the previous model have been highlighted.





RP Pandey, Gaurav Verma, Aishwarya Dixit, SV Gade, AM Datar
Armament Research & Development Establishment, Pune 411 021

Small arms are personal defense weapons used by law enforcing agencies. When it is fired on an opponent, it is
done with the clear intention that the opponent will get incapacitated and will not retaliate. Lethality of a weapon is its
capability to knock down the opponent in shortest possible time. The weapon which performs the function of knocking
down the opponent in more effective way and shorter time is called as more lethal. It is the Ammunition which causes the
damage at the target end. Hence the design of Ammunition plays very important role in determining the lethality of the
The human body is a very complex target. It is far more difficult to predict a human target's composition and bullet design
that will be most advantageous. The combinations of muscle, bone, organs, skin, fat, and clothing create a staggering
number of target types which often require different lethal mechanisms. Physical condition, psychological state, size,
weight, and body form all play a factor in the body's ability to resist damage. The same bullet fired against a large, thick,
healthy person has a very different reaction than that fired against a thin, malnourished opponent. The paper presents
definition of incapacitation and various factors affecting the incapacitation of Human body. Since the subject under the
study is Human being, the trial and evaluation has to be done on simulated targets. Hence various experimental techniques
of evaluation of lethality of Small Arms Ammn are devised. The same are discussed in the paper. The Ballistic trials
conducted on wood and Ballistic Gelatine block to assess the comparative lethality of various Small Arms Ammunitions are
also presented in the paper.

Use of force to protect and defend a nation from external aggression, internal disorder and unlawful activity is
considered necessary and morally acceptable. Even though ultimate, lethal force is sometimes unavoidable the force should
always be in proportion with the prevailing threat. Firearms injuries caused in military conflicts, law enforcement and
civilian environment are numerous and of great social and humanitarian concern. Small arms are weapons designed for
personal use. In any armed conflict the objective is to neutralize the opposite force. While killing an enemy soldier certainly
accomplishes this task, incapacitating him achieves the same goal while putting more burdens on the opponent's medical
and logistic resources. As far as the weapon designer is concerned, it is the ability to incapacitate which is more important
rather than killing capability. At many points in the Army, the requirement arises for quantitative comparison of
performance of various competing weapons.
A police officer using a firearm must know what it will do and what not [UN 1990]. As a corollary the bullet must
perform consistently in various foreseeable circumstances. A law enforcement bullet must therefore have sufficient tactical
range and kinetic energy, yet avoid excessive injuries and especially danger to bystanders. Increasing bullet velocity gives a
longer tactical range. It will also increase kinetic energy. If most of the energy is not dissipated into the target or deformation
of the bullet, the result is increased danger of injury to bystanders due to excessive penetration and high residual velocity
and energy of the exiting bullet. If most of the energy is dissipated into the target, the result will very often be increased
injury to the target. If the bullet velocity is reduced the tactical range shortens and the bullet may not perform consistently.
This may in turn prolong the gunfight and significantly increase the danger to both bystanders and law enforcement
officials. An ideal law enforcement bullet should not fragment during glass or other light barricade and subsequent soft
tissue penetration. It should, however, lose most of its kinetic energy into deformation or disintegration on impact with a
hard substance like street surface or brick wall in order to minimise the danger from ricochets.

What Is Lethality?
The simple question requires a very complex answer. For the Soldier in combat, lethality equals death: the desire to have
every round fired result in the death of the opposing combatant, the so-called "one-shot drop." However, death--or lethality-



-is not always necessary to achieve a military objective; an enemy combatant who is no longer willing or able to perform a
meaningful military task may be as good as dead under most circumstances. Lethality ultimately equates to the potential of
the weapons system to eliminate its target as a militarily relevant threat.
Assessment of injury potential is central for determining the legitimacy of a weapon and its ammunition. It is essential to
make a distinctive difference between assessment of injury potential and assessment of effectiveness i.e. incapacitation
potential. Although greater injury usually also means greater incapacitation potential the relationship is not at all clear. As
injury deals with the physiological result, the incapacitation includes also the psychological and more difficult to measure
processes whose outcome is affected by determination, beliefs and even previous experiences.

The weapon designer uses incapacitation as a measure of weapon performance. Wound is a structural derangement
which may cause incapacitation. Incapacitation is a functional rather than a structural concept. Different tactical roles
involve different tasks hence the wounds are not always incapacitating. The same wound received by two individuals
performing different tasks may be incapacitation to one but not to other. In order to calculate incapacitation one must consider
biomedical degradation caused by the wound and biomechanical requirement that goes with soldier's role. It is important to
make a distinction between the specific effect that results from single projectile and the generic overall effect averaged over
the entire body.
The human body is a very complex target. It has a number of built-in mechanisms that allow it to absorb damage and
continue to function. Compared to a tank, it is far more difficult to predict a human target's composition and what bullet
design will be most advantageous. The combinations of muscle, bone, organs, skin, fat, and clothing create a staggering
number of target types which often require different lethal mechanisms. Physical conditioning, psychological state, size,
weight, and body form all play a factor in the body's ability to resist damage, and all add to the complexity of the problem. The
same bullet fired against a large, thick, well-conditioned person has a very different reaction than that fired against a thin,
malnourished opponent. The average area of the various human body parts and the vulnerability is given below:
a) Total Area : 4000 cm2
b) Head and neck : 12 %
c) Thorax : 16 %
d) Abdomen : 11 %
e) Upper Limbs : 22 %
f) Lower Limbs : 39 %
g) Vulnerable area : 40 % of the total area (Head chest and Abdomen)
h) In general the back side has better inbuilt protection to the vital organs by means of tissues, bones etc.

Incapacitation Mechanism
The physical mechanisms for incapacitation--causing the body to no longer be able to perform a task—is due to followings:
i) Destruction of central nervous system tissue so that the body can no longer control function.
ii) Reduction in ability to function over time through blood loss.
The closest things the human body has to an "off switch" are the brain, brain stem, and upper spinal cord, which are small and
well-protected targets. Even a heart shot allows a person to function for a period of time before finally succumbing to blood
loss. What the wound ballistics community at large has long known is that the effectiveness of a round of ammunition is
directly related to the location, volume, and severity of tissue damage. In other words, a well-placed .22 caliber round can be
far more lethal than a poorly placed .50 caliber machine gun round. Setting shot placement aside for the moment, though, the
challenge becomes assessing the potential of a given round of ammunition to cause the needed volume and severity of tissue
damage, and then relating this back to performance against a human target.
Upon impacting the target, the bullet penetrates tissue and begins to slow some distance into the target, the tissue
acting on the bullet also causes the bullet to rotate erratically or yaw; the location and amount of yaw depend upon speed of
the bullet at impact, angle of impact, and density of the tissue. If the bullet is moving fast enough, it may also begin to break
up, with pieces spreading away from the main path of the bullet to damage other tissue. If the target is thick enough, all of
these fragments may come to rest in the target, or they may exit the target. Meanwhile, the impacted tissue rebounds away
from the path of the bullet, creating what is known as a "temporary cavity." Some of the tissue is smashed or torn by the bullet
itself, or its fragments; some expands too far and tears. The temporary cavity eventually rebounds, leaving behind the torn



tissue in the wound track--the "permanent cavity." It is this permanent cavity that is most significant, as it represents the
damaged tissue that can impair and eventually kill the target, provided, of course, that the damaged tissue is actually some
place on the body that is critical.
This is where the balance of factors in bullet design becomes important. Volume of tissue damage is important--
which might suggest high velocities to enable the bullet to tumble and fragment sooner, materials that cause the bullet to
break up sooner, etc.--but it must also occur in critical tissue. If the bullet immediately breaks up, it may not penetrate through
outer garments to reach tissue, or it may break up in muscle without reaching vital organs underneath. The projectile must
have enough penetration capability to reach vital organs to cause damage. At the same time, it must not have so much
penetrating capability that it passes completely through the target without significant damage--resulting in a so-called
"through-and-through." Energy expended outside the target is useless (as it does not separate energy expended in damaging
the target from energy lost beyond the target). The ideal bullet would have enough energy to penetrate through any
intervening barrier to reach vital organs without significantly slowing, then dump all of its energy into damaging vital organs
without exiting the body. Unfortunately, design of such a bullet is nearly impossible in a military round, even if all human
bodies were uniform enough to allow for such a thing. When hitting a target the bullet might have to pass through a forearm,
exit, enter, the shoulder, then proceed down the trunk before striking heart or spinal cord. A flanking hit would engage the
same target through or between the ribs to strike the same vital regions. All these possibilities are encountered with the same
ammunition. Ultimately, bullet design is a series of tradeoffs complicated by the need to survive launch, arrive at the target
accurately, possibly penetrate armor, glass, or other barriers.

A bullet impacting the target has an impact mass of mi (g) and velocity vi (m/s). Its kinetic energy Ei (J) is defined as
Ei = 0.5 * mi * vi2 / 1000
Impact energy Ei is partially dissipated into the target and performs work upon it. From equation above, we can see that both
the bullet mass but more significantly its velocity determines the amount of kinetic energy. If the energy is not dissipated into
the target, it is used somewhere else. The wound ballistic energy equation can be expressed as:
Er = Ei – Edef – Ed
where Er is the residual kinetic energy, Ei the impact energy, Edef the energy used by bullet deformation and Ed the energy
dissipated into the target. Since Ei has to be significant, Edef and Ed must be maximized in order to minimise Er. The residual
energy is a significant factor describing the danger to bystanders when the bullet completely penetrates and exits the primary
target continuing its flight. The factor of Edef has often been overlooked in the literature. Kinetic energy dissipation (Ed) can
be increased by bullet instability, deformation and fragmentation. When a rigid tail-heavy bullet hits the target it tends to start
tumbling because the rate of spin is insufficient to maintain stability in dense medium like tissue. This increases the cross-
sectional area in the direction of penetration which increases the dissipation of kinetic energy. The process is, however,
somewhat out of control. The precise depth at which tumbling occurs is difficult to predict as it depends on the yaw angle on
impact, properties of the tissue encountered and internal instabilities of the bullet.
Controlled deformation can in principle be achieved by a cavity in the tip of the bullet. Upon impact these bullets
start expanding at the tip. This makes the cross-sectional area larger and increases Ed. It will also shift the centre of gravity of
the penetrating bullet closer to the tip making a long bullet in theory more stable in penetration. The dimensions and surface
angles of the cavity together with the bullet materials and construction determine the rate and type of expansion. Bullet
fragmentation can be controlled by jacket thickness, making pre fragmentation incisions in the jacket and by varying the
strength of bonding between the bullet core and jacket.
The exiting bullet makes an exit wound which is usually larger than the entry wound as the bullet has already
deformed or may be in an unstable state exiting base or side first. The result will in many cases be a star like rupture. The
difference between entry and exit wounds can also be hypothesized with different penetration mechanism. In an entry wound
the projectile presses the skin against subcutaneous tissue thus reducing the stretch whereas on exit the skin is stretched
outwards from the supporting tissue thus increasing the stretch. As the wound must anyway be opened (debrided) for tissue
excision and the skin injuries can relatively easily be corrected, the entry and exit wound sizes as markers for wound severity
cannot be held very significant. A large exit wound should, however, be treated as a warning sign for possible extensive
internal injury. In order to cause only the justified injuries the bullet must be inherently accurate and have a satisfactory
tactical range.



International Laws
In order to reduce the use of excessive force and un necessary sufferings from the bullet injuries, some international
rules are framed. Some are discussed below:
i) United Nations Universal Declaration of Human Rights 1948 declares that 'everyone has the right to life, liberty and
security of person' (article 3), and that 'no one shall be subjected to torture or to cruel, inhuman or degrading treatment or
punishment' (article 5).
ii) The UN Code of Conduct for Law Enforcement Officials (1979) defines standards for law enforcement practice that
are consistent with basic human rights. Article 3 states that: 'the use of firearms is considered an extreme measure.
iii) Every effort should be made to exclude the use of firearms, especially against children. In general, firearms should
not be used except when a suspected offender offers armed resistance or otherwise jeopardizes the lives of others and less
extreme measures are not sufficient to restrain or apprehend the suspected offender'.
iv) The UN Resolution on Small-Calibre Weapon Systems (Geneva, 1979) recognizes that bullet expansion is not the
only factor increasing the extent of injury and that rapid advance in weapons technology requires further research and
standardized assessment methodology.
v) A number of technical interpretations of the International Law and recommendations are presented as the basis for
assessing the legitimacy of firearm and ammunition types from the injury potential point of view:
vi) To avoid superfluous injury and unnecessary suffering to the offender
a) There must be a maximum for acceptable kinetic energy dissipation for the first 250 mm of penetration in
defined tests with accepted synthetic material, for example, gelatine validated to simulate muscle tissue.
b) There must be a maximum for an acceptable viscous criterion of non-penetrating kinetic impact projectiles
(baton projectiles) in defined tests.
c) There must be a maximum value for acceptable fragmentation in defined tests (including kinetic impact
d) There must be a maximum for acceptable 'crookedness' of the permanent wound channel in defined tests.
e) No fragmentation bullets shall be used directly against persons.
f) No explosive bullets shall be used directly against persons.
g) No incendiary bullets shall be used directly against persons.
h) Projectiles used directly against persons shall not contain X-ray undetectable components.
i) To avoid unwarranted risk and injury to uninvolved persons and other officials.
j) Both the weapon and the ammunition must function reliably as a combination in defined conditions.
k) Ammunition performance must be consistent; in particular, penetration, permanent and temporary cavity
formation must be consistent in defined tests.
l) Accuracy and selectivity of weapon and ammunition must provide for a sufficient tactical range and
accuracy in various defined conditions.
m) Penetration ability of the standard issue projectile must be controlled with minimum and maximum tissue
simulant penetration in defined tests with a defined maximum for acceptable residual kinetic energy after
penetration of 250mm of simulated tissue.
n) A personal side-arm standard issue projectile should not penetrate the body armour worn by the official.
o) Special circumstances may warrant the use of bullets with higher penetration ability.
p) There must be a maximum for acceptable ricochet deflection angle from standardised surfaces and a
maximum for acceptable kinetic energy of the heaviest fragment.
q) Functional safety of the weapon must be proven by a) standard test set and b) any necessary weapon
specific tests.
r) Visual and tactile markings shall be provided to assist in identification of various types of ammunition
under adverse light and weather conditions.
s) For ease of use under stress, the weapon should not require the performance of any fine motor functions in
a tactical situation except for loading the magazine and pressing the trigger.
t) To increase proficiency, the weapon systems should handle with reasonable similarity.
u) To increase accountability, it should be possible to trace the bullet back to the weapon that fired it and as a
matter of principle, all firearms should be assigned to named officials.
v) Exceptional and grave circumstances may justify deviation from the abov recommendations.



The tissue simulants and their preparation methods have not been standardized making comparison of the reported results
somewhat difficult and reducing their power of evidence. To reach the goal of comprehensive and surgery compatible wound
ballistic simulation a lot needs to be done. Validation of simulation materials is still at an embryonic stage with only muscle
and skin tissue simulants having been validated. Bone simulants have not been validated with human bones. As the purpose is
not to simulate wound ballistic events on a pig the simulant system must be brought as close to a human being as possible. It
has too often been taken for granted that a pig resembles a human being. An example of this is the specific gravity of muscle
tissue quoted to be 1.06. It actually is that of a 100 kg landrace pig. Human muscle tissue specific gravity is 1.02-1.04. A
comprehensive study on the visco-elastic properties of human tissues needs to be done in order to develop and validate
corresponding simulants. Wound ballistic testing with tissue simulants cannot be replaced with computer simulation. Power
of evidence and needs of quality assurance require that real bullets must be fired upon simulated targets. Computer simulation
and modelling are, however, important for learning more about the complex interaction between a projectile and a target. The
mathematical possibilities can far faster help to find better protection and bullet designs than experimental research only.
Tissue simulants can be divided into soft tissue, skin, bone and skull simulants. A good tissue simulant must possess the
following qualities:
a. similarity in the deceleration of the projectile between the simulant and the living tissue
b. the simulant has been validated for
c. similarity in the deformation behaviour of the projectile
d. similarity in the kinetic energy dissipation
e. kinetic energy dissipation measurability with reasonable accuracy
f. extrapolation of temporary cavity diameter
g. elastic behaviour similar to living tissue for observation and measurement of temporary
h. cavity formation and tissue compression
i. extrapolation of permanent cavity diameter
j. reproducibility
The above list of requirements means that the simulant does not need to possess exactly the same biomechanical properties as
living tissue as long as the results can be measured and appropriately extrapolated or scaled to reflect what happens in living
tissue. In fact a simulant with lower tensile strength than that of muscle tissue might allow for more accurate measurement of
the cavities and kinetic energy dissipation; hence the requirement for extrapolation instead of similarity.

Soft Tissue Simulants

Principally two types of tissue simulants have been used to simulate muscle tissue. Specially manufactured
glycerine soap and hydrogels prepared from water solutions containing 10-20 mass-% gelatine. Soap is plastic and does not
reproduce the pulsations of temporary cavity. The cavity left in soap reflects temporary cavitation. Permanent cavity cannot
be measured. Ballistic gelatine is usually used in 10 or 20% nominal concentrations. The actual concentration should be
adjusted i.e. calibrated for every batch of gelatine powder to give precisely the desired penetration and to compensate for
possible manufacturing variances. Gelatine blocks are prepared by dissolving the calibrated amount of usually 250A bloom
gelatine powder into warm water, adding possibly preservative and pouring the solution into moulds where it first cools off
and then solidifies. To avoid dehydration and unnecessary contamination by airborne microbes the moulds should be covered
with plastic foil. The gelatine blocks are then taken to refrigerator for storage and cooling into prescribed usage temperature.
It should, however, be noted that isotropic materials like gelatine and soap fail to take into account the
heterogeneous nature of tissues and their different densities.



Three 5.56 mm Ammns, having similar velocities and mass but different bullet constructions (Steel insert+Lead, Al
insert+ Lead and full Lead core) were fired upon the wood and Gelatine blocks. The trial results are as follows:

Trial On Wooden Target

Experimental trials were conducted to compute the energy transferred to 1 inch deal wood placed at range of 11m.
The details of the trial and the results are as follows



SV= Striking Velocity, SE= Striking Velocity, EE= Exit Energy, ET= Energy Transferred



Trial On Gelatine Block

The trials were conducted on 20 % Gelatine block of size 100 mm dia and 150 mm length placed at range of 5 m. The
energy deposited could not be computed since the bullets were breaking inside the block. The details of the trial and the
results are as follows



Summary Of Experimental Trials

i) The energy transferred to the wooden target by the bullet with Al insert and Lead is max as compared with the bullets
of different construction.
ii) In the Gelatine block;
a) The bullet with Al insert and Lead is getting destabilized very early as compared to bullet with Steel insert and Lead
and bullet with full lead core.
b) The Max hole dia created by the bullet with Al insert and Lead is max as compared to other two Ammns.
c) The exit hole dia of the bullet with Al insert and Lead is also max as compared to other two Ammns.
d) The amount of damage created on the Gelatine block by the bullet with Al insert and Lead is max as compared to
other two bullets.

Assessment of lethality of Small Arms weapon is very important aspect in selection of the weapon. The subject involves the
study from Biomedical and Biomechanical sciences. A thorough study of behavior of the bullet inside the human body will
immensely assist in designing an optimized bullet. Due this the morale of the law enforcing officers will boost and the
offenders will be neutralized without any causality to by stander. Standardization of experimental and simulation studies on
the simulants will help in deciding the Armies to select suitable weapon for a particular operation.

The authors are deeply indebted to Shri Surendra Kumar, outstanding scientist and Director ARDE for his continuous
encouragement and suggestions.

1.The Modeling and Application of Small Arms Wound Ballistics- Memorandum Report BRL-MR-3923, David N
Neades, Russel N Prather, US Army Laboratory Command, Aug 1991.
2. Wound Ballistic Simulation: Assessment of Legitimacy of Law Enforcement Firearms Ammunition by Means of
Wound Ballistic Simulation; Jorma Jussila, Helsinki, 2005




Aishwarya Dixit, Gaurav Verma, RP Pandey, SV Gade, AM Datar
Armament Research & Development Establishment, Pune 411 021

Small Arms are used to incapacitate protected and un-protected targets. The design of Small Arms Ammn is mainly
governed by the Range and type of target. Considering these requirements the Terminal, External and Internal Ballistics
studies are performed to decide the caliber, Bullet construction, muzzle velocity, rifling rate etc. The Small Arms bullets are
jacketed. Inside the jacket, either full Lead core, Steel insert and Lead core or other combinations are used to arrive at a
suitable bullet mass and stability.

The paper discusses in detail about the effect of bullet construction on the penetration of Mild Steel plate of 3.5 mm
thickness. To study the effect of bullet construction, two new types of Bullets have been developed. One bullet envelope is
filled with full lead core and other bullet envelope is filled with Al insert and Lead core. It has been ensured in the design that
the bullets are stable in the flight against any perturbation. The existing 5.56 mm Ammn with Steel insert and Lead core is
also used for the purpose of analysis. Experiments have been conducted by firing three different types of bullets of 5.56 mm
caliber at a range of 100 m on 3.5 mm thick MS plate. The striking velocity at this range was in the range of 800-850 m/s. It
has been observed that the bullet with lower density tip material has made bigger hole as compared to the denser material tip
bullet. The penetrations of bullets have also been simulated on ANSYS AUTODYN. It has been observed that the
simulation results are in good agreement with the experimental results. This will help in studying the performance of the
Small Arms bullets against hard targets and optimization of bullet design.

Small Arms are the weapons used for personal defence as well as disabling the opponents by various law enforcing agencies
and individuals. When the Ammn is fired upon, it is expected that the opponent will stop the current activities immediately
and not retaliate. This is achieved by incapacitating the opponent. Among other factors, incapacitation is also governed by
energy transferred by the bullet to the individual.
Various types of Ammns are under use by law enforcing agencies. In the same calibre different construction bullets are also
available. A Typical small arms bullet is encapsulated by a thin gilding metal jacket filed by lead core or in addition to some
lead the core may contain steel as shown in figure1.
5.56 mm calibre, three different construction bullets have been selected to study the effect of bullet construction on the
penetration capability of bullet on the hard target. They are as follows
a. Envelope filled with Al insert and Lead core
b. Envelope filled with full lead
c. Envelop filled with Steel insert and lead

The target selected for the purpose of the study is 3.5 mm thick MS plate.
The complete study was carried out in two stages. In the fist stage, the firing trials were conducted by placing the target at
range of 100 m and measuring the striking velocities and the hole dia on the target. During the second stage, theoretical
estimation of damage and deposited energy was estimated by using AUTODYN (hydro-codes) simulation which is widely
used in dynamic transient loading phenomena such as ballistic impacts. Hydro-codes simulations can accurately predict
penetration as well as impact crater geometry (depth and diameter) in given thickness of steel armor for different types of
bullets. To predict behavior of hard target on bullet impact, material models for target and bullet have been developed.

Bullet Construction
All projectiles used for Small Arms ammunition are of composite construction. Solid lead is not used because of its low



strength. If it is fired from rifled barrel, it results in the coating the bore of the barrel with layers of lead which soon fills the
rifling grooves. The most common constriction is lead antimony alloy covered with a copper alloy envelope. Steel jackets
may also be used to increase the strength of bullet and to reduce the costs.
The Ogive of the high velocity projectile is made to be well pointed to reduce aerodynamic drag. It is also usual for the base of
the bullet to be tapered (boat tail) to reduce base drag. To increase the stability in flight the ogive portion is often filled with a
light weight material to move the center of gravity of the bullet more towards base.

Fig1: General construction of a bullet


A series of firing trials were conducted with the three types of the bullets. The striking velocities were recorded with help of
Doppler radar set up and used in subsequent AUTODYN simulations. The results are tabulated below.
Table 1: Firing range 100 m, Target 3.5 mm MS Plate:

The AUTODYN computer code was employed in full Lagrange mode to simulate the all three bullets impacting on Mild steel
targets. In this study 5.56 mm Bullet with Envelope filled with Steel insert and Lead core bullet was used as a basis for
investigation of different core material configurations. Figure 2 shows the Lagrangian mesh for this bullet mirrored about its
horizontal axis of symmetry. The analyses were conducted with AUTODYN -2D using axial symmetry about the axis. The
model consists of more than 5000 grid cells that are filled with appropriate user defined materials models.



Factors Affecting the Penetration

A complete description of the dynamics of the impacting solid would demand that account be taken of the geometry
of the interacting bodies, elastic, plastic and shock wave propagation, hydrodynamic flow, finite strains, work hardening,
thermal and frictional effect and the initiation and propagation of failure in the colliding materials.
For low velocity regime (less than 250m/s) typically loading response times are in millisecond regime. Material behaves
primarily plastic.
For the striking velocity range 500-1000m/s the response of the structure mainly focused with in a small zone
(typically 2-3 times of projectile diameter) of the impact area. Viscous material strength plays significant role. A wave
description of the phenomena is appropriate and the influence of velocity, geometry, material constitution, strain rate,
localized plastic flow and failure are manifest at various stage of impact process. Typically the loading and reaction times are
on the order of microseconds.
Higher impact velocities (2000-3000m/s) resulted in localized pressure that exceeds the strength of material by
multiple orders of magnitude. In effect the colliding solids can be treated as fluids in the early stages of impacts.

Modeling of Material used in AUTODYN Simulation

In general, materials have a complex response to dynamic loading and the following phenomena may need to be
modelled.The modelling of such phenomena can generally be broken down into two main components.

Equation of State
An equation of state describes the hydrodynamic response of a material. This is the primary response for gases and
liquids, which can sustain no shear. Their response to dynamic loading is strictly hydrodynamic, with pressure varying as a
function of density and internal energy. This is also the primary response for solids at high deformation rates, when the
hydrodynamic pressure is far greater than the yield stress of the material.
A general material model requires equations that relate stress to deformation and internal energy (or temperature). In most
cases, the stress tensor may be separated into a uniform hydrostatic pressure (all three normal stresses equal) and a stress
deviatoric tensor associated with the resistance of the material to shear distortion. Then the relation between the hydrostatic
pressure, the local density (or specific volume) and local specific energy (or temperature) is known as an equation of state. In
the present study due to the loading condition the used equation of state is shock.
Shock EOS:
The Rankine-Hugoniot equations for the shock jump conditions can be regarded as defining a relation between any pair of the
variables r, p, e, up and U. In many dynamic experiments making measurements of up and U it has been found that for most
solids and many liquids over a wide range of pressure there is an empirical linear relationship between these two variables: If
U is shock velocity and up is material velocity behind the shock, C is the velocity of sound, p is hydrostatic pressure, r is

The Gruneisen Gamma G

, defined as:

Where J is the mechanical equivalent of heat

density, m compression = r/r0 - 1, e is specific internal energy, S is specific entropy.
It is then found convenient to establish a Mie-Gruneisen form of equation of state based on the shock Hugoniot. Where U is
shock velocity,

Material Strength Model

Solid materials may initially respond elastically, but under extreme shock loadings, they usually reach stress states
that exceed their yield stress and deform plastically. Material strength equations describe this non-linear elaso-plastic



The original formulation of material strength effects considered that materials were elastic -perfectly plastic. However it is
possible to generalize the approach by making the yield function Y a function of material properties such as strain, strain rate,
energy, temperature etc. without excessively complicating the resultant calculations. Model used to solve present problem
are Von Mises and Johnson-Cook.

Von Mises Model:

This model uses the original von-Mises premise that the yield stress has a constant value. Consequently the von-Mises
cylinder has a fixed radius. States lying inside the cylinder are elastic. States on the surface of the cylinder are plastic.

Johnson-Cook Model:
Use this model to represent the strength behavior of materials subjected to large strains, high strain rates and high
temperatures. Such behavior might arise in problems of intense impulsive loading due to high velocity impact.
With this model, the yield stress (and the radius of the von-Mises cylinder) varies depending on strain, strain rate and
The model defines the yield stress Y as

Where ep = effective plastic strain, ep* = normalized effective plastic strain rate, TH = homologous temperature = (T -
Troom) / (Tmelt - Troom ).
The five material constants are A, B, C, n and m. The expression in the first set of brackets gives the stress as a function of
strain when ep* = 1.0sec -1 and TH = 0 (i.e. for laboratory experiments at room temperature). The constant A is the basic yield
stress at low strains while B and n represent the effect of strain hardening. The expressions in the second and third sets of
brackets represent the effects of strain rate and temperature, respectively.
Material properties of lead, copper and steel employed in these simulations are given in table 2.
Table2: Material Properties for impact Simulation
The impact velocity was given on the basis of experimental data recorded previously. Fig 3, 4 and 5 show the details of
simulation study conducted on different types of bullets. Simulated penetration at various time steps are shown for each
bullet. Energy absorbed by target with reference to time is shown in graph. Final dimensions of AUTODYN simulated bullet
crater are measured and compared with actual cross-section cut.




Simulation of bullet with Al insert and Lead core is shown in Fig 3 at 12, 32 and 35 micro sec. Similarly for bullet
with full Lead core and bullet with Steel insert and Lead core are shown in figures 4 and 5 respectively. Dimensions of
predicted impact crater are also measured and given in Table No 3. The actual holes made by the bullets and the predicted
crater dia are also shown in the respective figures.
Measured and predicted crater dimensions are in good agreement. This shows that this method of simulation is
validated and could be used for other bullet design exercises. In case of bullet with Steel insert and Lead core, mesh tangling
was more and program stopped in the middle. This problem was resolved by redefining mesh geometry and giving proper
erosion criteria. Energy absorbed by the target due to impact of the three types of the bullets is given in table 4.



Fig 3e: Bullet at 4.5e-2 ms



Fig 3g: energy absorbed by the

Fig 4d: Bullet at 3.56e-2 ms



Fig 4f: Actual cross section cut

Fig 5b: Bullet with Al insert just before striking the plate



Fig 5d: Bullet at 2.60e-2 ms

Fig 5f: Actual cross section cut



Simulations studies and firing trails have been conducted to predict the effect of bullet compositions on penetration of hard
targets (MS plate 3.5 mm thick). These studies have shown how core material properties such as density and deformability
influence both bullet penetrability and its residual potential for damaging the hard targets. Results show that the projectile
with less density material as insert has better damage effect at the target. Good agreements have been observed among
experimental results and AUTODYN outputs. Energy transferred by the bullet with Al insert and Lead core was max (325 J)
as compared to bullet with full Lead core (250 J) and bullet with Steel insert and Lead core (80 J).
The Von-mises strength model for Lead and Copper material provided satisfying results. For the purpose of Steel and Al,
Johnson Cook strength model was successfully used. Mesh tangling problem with Lagrange solver during processing was
solved by redefining the mesh geometry and giving suitable erosion criteria. It was observed that the Eulerian treatment
results were less accurate as compared to Lagrange solver.

The authors are deeply indebted to Shri Surendra Kumar, Outstanding Scientist and Director ARDE for his continuous
encouragement, suggestions and permission to publish this paper.

1.Century Dynamics, AUTODYN User and theory Manual
2. Journal of Impact engineering (Elseiver)
3. By Dale S Preece and Vanessa S Berg” Bullet Impact on Steel and Kevlar/ Steel Armor- Computer Modeling and
Experimental Data” published in proceedings of the ASME Pressure Vessels and piping Conference Symposium Under
Extreme Loading, July 25-29, 2004, San Diego, CA.
4. RE Wellard “Cannon and Small Arms Ammunition and Terminal Ballistics” proceedings of the 11th Euopean Small
Arms Symposium, Royal Military college of Science, Shrivenham, 1997.




TB Khilare
Armament Research & Development Establishment, Pune 411 021

Increase in range by insertion of rocket mortar in Artillery shell or bomb reduces the capacity of explosives in the
bomb. Consequent to this, overall lethality of bomb or shell is reduced in compensation for increased range. Therefore, ways
are required to be thought to increase the lethality over standard HE bomb or shell, or at least equalize the lethality to the
standard HE bomb or shell. Selection of high fragmentation yielding material for body, optimum metal to explosive ratio /
proper distribution of mass of explosives with respect to thickness of steel casing of bomb or shell , use of high VOD
explosives and use of pre-notched body are the major contributors which will increase lethality of bomb/shell. In this paper, a
study has been carried out for 120mm rocket assisted bomb to increase the lethality against standard HE bomb. After design
of rocket assisted bomb and selection of high fragmentation yielding material, prototypes have been developed. After filling
of high explosive, prototypes have been trial evaluated statically for lethal radius.
Results show that lethal radius of rocket assisted bomb has been enhanced 1.5 times over standard HE bomb.

RAB - Rocket Assisted Bomb
HE - High Explosive
VOD - Velocity of Detonation
- Probability of kill
- Number of hits
Lethal Radius A radius from the point of burst of projectile within which a chance of incapacitation is 50 per cent.
Lethal fragment A fragment is said to be lethal if it perforates 1.5 mm thick steel plate.

The principle cause of casualties is invariably fragmentation from high explosive ammunition and statistics from
the Second World War, Korean War, Vietnam War and many other wars verify this. The principle of high explosive
ammunition is very simple: detonation of the explosive filling causes the casing in which it is held to break up, propelling
fragments outwards at high velocity. The way in which the casing breaks up is influenced by several factors, the most
important of which are the velocity of detonation of explosive, the distribution of the explosive in relation to the thickness of
casing, and the metallurgical properties of casing. The fragments are hurled outward, and penetrate or perforate objects
within certain distance from the point of burst. The ability to penetrate the target or to damage the object is characterized by
the concept of the “effective fragment” which is defined as: A fragment is effective if it perforates a steel sheet of 1.5mm thick.
The number of hits of effective fragments on the distance from the point of burst, are determined in a fragmentation test area.
Probability of kill is computed from the number of hits and lethal radius is worked out. A chance of 50 % kill is achieved at a
certain radius from the point of burst, a bomb or shell is said to have that much lethal radius. A literature study has been shown
that Mn renders a material fracture prone because it reduces grain size whereas Si increases frictional stress resisting motion
of dislocation. Similarly Carbon and Nitrogen raise stress needed to unlock dislocations held up at grain boundaries.
Rocket assisted bombs are used to engage targets at higher ranges. However lethal radius of bomb reduces due to
incorporation of rocket motor into the bomb body of the same calibre and nearly same mass of bomb body due to reduction in
explosive and explosive mass in cross section in relation to thickness of casing. However lethal radius may be increased by
various means like use of high yielding fragmentation material for casing, use of high brisance explosive, use of
predetermined fracture points. It has been observed from previous study that high carbon content material yields more
number of small fragments. However, Designer has to keep in mind that the casing of bomb body should have adequate



strength to withstand firing stresses during gun launch. Material Spheroidal Graphite Iron which has high percentage of
carbon and also contain graphite in the form of spheroid which helps in formation of large number of small fragments. While
selecting high VOD explosive, factors like ease of processing during mass production, insensitivity of explosive during
handling from safety point of view have been considered.
In the present study, with above background, material AISI 9260 ( En 45A) which has higher carbon content ( 0.6 %)
and higher Si ( 2.0 % ) and also has moderate per cent of Mn and adequate strength to withstand firing stresses has been used
for rocket assisted bomb in comparison to material grade SS F/44 to specification JSS 9510 with carbon of 0.3 0.4 per cent
was used for standard HE bomb. Pre-fractured points were made inside the casing of rocket assisted body as comparative
study to assess its effect on enhancement of lethal radius.

Sketch of design of rocket assisted bomb body for 120mm calibre is shown in figure 1 and sketch of standard HE
bomb is shown in figure 2. Bomb body was made of forgings made from AISI 9260 material having obtained proof strength of
800 MPa with 10 % elongation after heat treatment which was adequate for withstanding firing stresses for designed casing
thickness. Hardness range is between 300-350 HB. Standard HE bomb of 120mm calibre made from forging material of steel
grade SS F/44 of JSS 9510 specification having proof strength of 590 MPa with 8 % elongation with hardness of 250-300 HB.
Bodies of rocket assisted bomb were also made with pre-fractured points internally with 0.5- 0.7 mm depth at 60° angle. Each
square was varying size from 5 x 5 mm to 8 x 8 mm due to contour of body. Thickness of body was varying from 6mm to
8.5mm. Empty rocket casing made from aluminium alloy was placed inside as shown in figure 1. Standard HE bomb bodies
were filled with TNT and rocket assisted bomb bodies were filled with RDX/TNT (60/40). Qty 5 rocket assisted bodies with
pre-fractured points and qty 5 bodies without pre-fractured points were manufactured. Explosive Charge to metal ratio for
both the bombs are shown in Table 1. It is seen from the table1 that explosive charge to metal ratio has been slightly improved
in the rocket assisted bomb
Metal mass, explosive mass & charge mass ratios are tabulated below.

Trial Set-up & Static Evaluation

The Bomb was suspended 1.2 m above the ground vertically with nose down position and fired remotely. Steel
plates of 1.5mm thickness with 2.4m height were erected at distances of 20m, 30m, 40m and 50m from the centre of burst
point of bomb. Average velocity of fragments was measured with high speed photography. A sketch of trial set-up is shown in
fig 3. Qty 3 numbers of bomb each with and without pre-fractured points were test fired for evaluation of lethal radius. Qty 2
numbers of bomb each were fired in sand pit to assess the total number of fragments.


It is seen from the Table-2 that lethal radius (kill probability 50%) of Rocket assisted bomb is 30 m whereas lethal
radius of standard service bomb is 20 m.



Total numbers of fragments of rocket assisted bomb are 3120 which is much higher as compared to standard service
HE bomb of having only 1780 number of fragments.
If the comparison is made between pre-fractured and non pre-fractured rocket assisted bombs, it is seen that average
number of fragments in both the cases are almost same and there is no significant improvement in lethal radius as well. This
may be due to insufficient depth of pre-fractured points and appropriate distance between the pre-fractured points. However it
may be noted that sufficient wall thicknesses are required to be kept to withstand the firing stresses. Moreover manufacturing
of pre-fractured points in the bomb body are time consuming and not cost effective.
Thus, therefore it is concluded that by increasing charge to metal ratio and use of high fragmentation yielding
material and high VOD explosive, lethal radius can be increased significantly in rocket assisted bomb.
The results of trials are tabulated below.
1. No of lethal hits & lethal hits per square metre are given for average of 3 bombs.
2. Probability of kill is worked out based on lethal hits per square metre.
3. The equation for Probability of kill is given below.

Author would like to express his sincere gratitude to Shri Surendra Kumar, Director ARDE for his kind permission
to publish the results. Author would further wish to thank Shri Anil Datar, Associate Director for his encouragement. Thanks
are also due to Dr. Pardeep Kumar, Sc 'E' from TBRL, Chandigarh for his support in conducting the static evaluation of trials.
Author is also grateful to Shri RD Wasnik Scientist E and his team from HEMRL, Pune for their support for explosive filling
of bombs.

1. Handbook of Weaponry- Rheinmetal Gmbh 1982
2. Brassey's “ Ammunition for the land battle “ by PR Courtney- Green
3. TBRL Report No.427/96 on, “Study of Fragmentation behaviour of plain Carbon & Low alloy steels.”
4. TBRL Report No.70/72 on, “Fragmentation and lethality of 120mm Brandt mortar HE Ammunition”
5. ARDE Closing Report NO. 1179 of July 2007 on, “To Evaluate the Lethality of 120mm Rocket Assisted Bomb for
Existing Service Mortar E1”



Integrated G-hardened Telemetry System

Cannon Launched Munition Testing
Vaibhav Saini,Vinay Kumar,DS Saini, S C Rana, Kapil Deo

Scientists, Armament R&D Establishment, DRDO, Pune, India

Email: sainids 1(a),gmail. corn

32 Analog Inputs
Abstract-The Telemetry group at ARDE actively develops
telemetry systems for testing of Artillery rockets, munitions and
Aircrafts bombs. The integrated g-hardened telemetry system
utilizes new capabilities in transmitter , power supplies, electronic
packaging and sensors and PCM encoder incorporating FPGA Signal Conditioning
based architecture for operational flexibility. Each module has
Transmitting Antenna
been individually test qualified to the required shock and vibra-
tion level This PCM/FM telemetry system has been demonstrated
to function in shock environment up to 700g and survives impact Serial Link 1
shock up to 2000g. Design criteria are presented for multi- Pre-
S Band
channel re-configurable PCM telemetry system to monitor on Serial Link 2 PCM Encoder modulation
board flight computer digital data, 3-axis body acceleration, roll

rates and all other digital system parameters and telemeter it Serial Link 3
to ground receiving station . The complete telemetry system with
g-hardening to survive severe launch acceleration is described Supply
and results of its flight performance are also presented. swn^n^ng
Internal Management

Key words : g-hardened telemetry system , miniaturization, g- eaneries Motlule

hardened PCM Encoder, Telemetry frame, Microstrip antennas.

Enernal Power



This paper introduces a development of highly integrated Fig. 1. Telemetry Block Diagram
low cost telemetry system for munitions testing that can

measure analog and digital signals. This telemetry system

was required to acquire and collect telemetry data such as design was verified with a series of thirteen full missile system
vibration, temperature, accelerations, pressure, and so on. flight tests.
The telemetry system synchronizes with digital data sources
of various on-board subsystems such as Flight control Unit

(FCU), Inertial Measurement Unit (IMU), Actuator control

A. PCM Encoder

Unit (ACU) and collects health monitoring and real time The PCM encoder receives analog and digital data from
calculation data. In addition,it withstand Knock Out Engine several sources, organizes the data into a data frame, and

(gas generator) set back acceleration of 700 g, vibration levels sends a randomized serial data stream to the transmitter for

of 0.06 g2/h,_- and capture data at 1 111Samp1es/s. modulation and transmission. The PCM encoder is designed

The telemetry systems consists of four modules, Power for data rates upto 2 Mbps and provides 32 analog channels

source and Switching Module, Pulse Code Modulation (PCM) that are sampled at IMS/s, with 12 bit resolution. There

Encoder, Sensor and signal conditioning module (SCM), are three serial RS-422 interfaces that can process data at

Transmitter and Transmitting Antenna. The system block 115kbps, PCM encoder directly acquires high sampling rate

diagram is shown in figure 1. Field Programmable Gate Array analog channels after signal conditioning. One of the key

technology was incorporated into the PCM Encoder design. features of encoder is synchronization with digital burst data

An innovative modular packaging and interconnect scheme and PCM frame construction.The PCM Encoder is coded

allowed high survivability. This modularity also allows card by VHDL (Hardware Description Language), synthesized by

level repairs and easy system level reconfiguration for alternate Xilinx ISE, and implemented on an ALTERA CycloneT100.

uses. The post flight data processing was also simplified and PCM Encoder block diagram is shown in figure 2.
enhanced by using MATLAB to process captured data. The The PCM Encoder has features as shown in Table I.

final result is a compact, modular telemetry system ruggedized In order to interface with RS422, Three UART (Universal

to survive ultra high setback and impact shock levels. This Asynchronous Receiver and Transmitter) is implemented in


Data Data D . t. Data Data Data Data Data Daa Data Data
word word ward Word Nord word word word word word word
1 2 3 , S 5 1 a It 1`41 N

Fnm Fram
Sync ya. sFID Dara
Minor Fr.me I
Word Word Word Al A2 D,t D,2 D,3 Al D, 1 D,5 D ,1 Word Al A10
F- Fram

Minor Frame 2 Sync Sy- SFID

ward word wwd Al A3 DA D,2 D3 At Dt2 D,5 D,2 At A10
1 2
Fram Fram
Mirror Frame 3 wsynrdc s.yn SFID
Al A4
wont were D, l D,2 D,3 Al Dt3 Dt5 113 ... Al A10
1 2
Fram Fr am
sync syn. SFID At A2
Ed w«a word D 1 D,2 D3 Al D,4 D,5 DA .. Al A10
1 2

Fra Fram
sync syn. SFID
ward wwd
Al A3 DA D,2 D,3 At D,3 Dt5 D >M-1 - Al A10


sm. SFID
Word word
Al A4 D,1 D,2 D,3 Al Dr4 D,5 ,M ... Al At()
...................... ..' .................................. - Digital Data Words Analog Data Words - Al, A2. A3.. SFID-Sub Frame Identifier
RS422 Chi - DA, D,2, D3..
RS422 Ch2-DA, D,2, Dt3..
RS422 Ch3 - D,l, D,2, D,3..

Fig. 3. Telemetry Frame

Encoder in receive mode only. After power on reset, the UART

waits for its individual synchronization data words. Upon
receipt of sync words, UART starts storing the corresponding also implemented in FPGA, which outputs BiPh-L or NRZ-L
asynchronous words in RAM bank which resides in the FPGA. coded randomized data stream as per output requirement.
When finished receiving block of data, UART waits until it The parameters sampled at high rate and digital data from
again receives digital frame sync words. Independent sync three sources are packetized to form a telemetry frame which

words can be programmed for each RS 422 interfaces. The is simply a matrix with rows representing minor frames and

serial interface has features as shown in Table It. columns as subframes as shown in Fig 3. . All the channels

The analog and digital channels can be commutated as with required sampled rate and digital data without any con-

normal, super and sub commutation. Encoder uses a state gestion should be accommodated within the given data rate.
machine block that generates addresses according to the PCM To get the best fit for all data the following unique identifiers

format. Additionally, it controls and initiates the A/D conver- for a data word are used

sion. The sampled data read from ADC and stored RS-422 Starting Word No, Word Increment, Starting Frame No. , Frame
serial data, is routed and interlaced in telemetry frame as per increment.

the commutation scheme. PISO (parallel in and Serial Out) is These four constants completely identifies occurrence of a

data word in telemetry frame [1]..By using simple memory

management techniques the vacant data block in minor frame

is filled with data words. Telemetry frame is finalized once the

Sampling criteria for all data words are met. The randomized
Analog channels( 0 to 5 V) 32 PCM serial data stream is then filtered through 5-pole Bessel
A/D converter 12 bit(1 MSPS)
pre-modulation filter to limit the transmitted RF bandwidth to
Frame programmability Full (including sub-commutation
and super commutation.) 1.4 times the bit rate for Bi-phL and 0.7 times for NRZ-L as
Baud rate . upto 2Mbps Programmable per IRIG 106-01 [2]. The Bessel filter characteristic is chosen
Digital Configurable Serial Input 3ports (RS422) for its maximally flat delay characteristics over the filter pass-
PCM output (TRIG-106) NRZ-L,M,S
band Figure 5. presents the modulated RF output Spectrum
or BiPh-L,M,S
Operation voltage 5 Volt Single Supply after pre-modulation filter. The circuit is operational on single

power supply.
B. Transmitter and Transmitting Antenna

number of Interfaces 3 twisted shielded RS-422 The transmit configuration is Single Transmitter -single
protocol Asynchronous mode Antenna. In this configuration a single telemetry transmitter,
communication mode burst mode
data quantity 4800,96W 19200, operating on a S Band (2255.5MHz) carrier frequency, is
38400.1 15.000bits/sec connected to a single telemetry antenna using co-axial cable.
Sync Words 4 byte sync data from source to PCM Encoder The telemetry transmitter selected for the requirement was de-
Digital Frame 256 bytes (Programmable) )telemetry data
veloped by M/A-COM,Inc. which is available as unpackaged
including 4 bytes of frame header
from source to PCM Encoder. electronic die. These transmitters are rugged, small, low cost
Other Details I byte consists of start Ibit, data Sbits modules and outputs power level of 500mW.
I and stop I bit
The efficiency of a modern FM telemetry transmitters



Fig. 6. Microstrip Antenna

Fig. 4. Transmitter housing

Fig. 7. Telemetry Housing

head compartments. The axial and lateral acceleration were

measured by second generation of MEMS accelerometers

(ADXL210, 150,190). The MEMS Accelerometer ADXL210

was used to measure Lateral acceleration of Spin Stabilized

Munition but care was taken to locate it along the spin axis

(so as to avoid large DC off-sets due to centrifugal force).

Fig. 5. RF Spectrum
The launch environment is very severe with axial accelerations

of 1000 g's, these sensor showed cross-axis sensitivity of

generally of the order of 20 percent, approximately 80 percent nearly 10`%, thus are saturated during launch phase It is of

of the power consumed by a transmitter must be dissipated interest to note that the recovery is quick and no occurrence

as heat. This overheating drastically shortens the life of the of Stiction (the nonrelease of MEMS structures subsequent

output power amplifier stages,thus optimum output power level to large over shocks). These MEMS Sensors can be used for
taking sufficient margin should be selected, smaller the size of future munitions where lesser space and low cost will be the
transmitter greater are problems of heat dissipation at higher requirement.

power levels. This transmitter has thermal vias which are Nontraditional navigation sensors (Automotive Grade) were

connected to ground plane for heat dissipation. Our novel used for angular rate measurements in the pitch and Roll axes.

approach was to design a small, lightweight, rugged mounting These sensors (from Murata) are piezoelectric in nature and

platform for the transmitter chip die as shown in figure 4. are available in miniaturized and rugged package.. The PCB

The die was encapsulated tightly using Teflon machined rings Piezotronics 3901 F31IB2000G accelerometer was chosen for

maintaining clearance of components from the housing. A low setback force measurement.This sensor was chosen for several

dropout, linear regulator generating +3.3 V was also mounted key reasons. First, it has high mounted resonance, offering

inside the casing. the best opportunity to collect wide bandwidth data without

Micro strip conformal antenna as shown in figure 6 is the exciting transducer resonance. Second, it is rugged, small

obvious choice when space and weight are prime considera- and light weight.and is proven to survive in extreme shock

tions.The antenna was mounted on the outer wall of warhead environments. Third, since it is piezo-resistive, its frequency

casing. It was made up of four rectangular patch sections response extends to DC.

that are fed through a power divider network. After mounting The transducers utilized in instrumentation system limits

on the metal cylinder, it behaved as cylindrical - rectangular the quality of the data, but equally important is its mounting

antenna and the desired radiation pattern in Pitch plane was and electronic interfacing. The key to survive high 'g' is

achieved. miniaturization and ruggedization thus minimizing mechanical

electrical impact.To ensure High Survivability following key

C. Sensor and Signal Conditioning
factors were taken care of. All interconnections between

The telemetry system is often placed in mother system sensors and signal conditioning were done within the module

where unused space is found such as in ogives and war- itself and the power and conditioned outputs of sensors were



Telemetry Trial Launch Acceleratlon Vs Time

o.o, o.ox o .aa o.os o.ae m

Time (sec)

Fig. 8. Telemetry Block

Fig. 9. Measured Shock

brought out from module using high density D Connectors.

The circuit boards are stacked using teflon spacers to the walls III. CONCLUSION
of aluminum housing box. All internal wiring is completed and This paper presented design and implementation of g
secured prior to Mounting The circuit boards are coated with a telemetry system for Cannon launched Munitions. Inorder
Polysulfide rubber sealing compound to protect surface mount to collect telemetry data and health monitoring data from
components.The potting compound( GE Bayers Silicon) is other units , the system adopts RS422 serial data link . A
applied ensuring to hold supply capacitors, Mounting nuts, Pcm Encoder , is designed and implemented on Xilinx FPGA
raised components and is left for the curing. The High g coded by VHDL. To qualify the instrumentation system for
accelerometers are mounted in channels cut in aluminum the flight tests, a qualification test was conducted in which
housing and layer of epoxy (Araldite and M-Seal Compound the instrumentation system was tested for 700g in launch axis
60:40) is applied to the underside and top of the accelerometers and subjected to random vibrations tests . The instrumentation
before finally securing with screws. system functioned properly during the test and was, therefore,
accepted for use in the flight test . Thirteen Tube-launched
instrumented flight tests were conducted at ITR Balasore using
D. Power Switching Module the developed system . Telemetry data was obtained from all
tests and provided test data for analytical simulations.
The power Switching module allows the telemetry system
to be operated from external power supply thus preserving

the telemetry power source.MOSFET Switching regulators The authors is thankful to all the members of Telemetry
were implemented for switching. The components and sensors Group at ITR,Chandipur for their support, Specially to the
in the design operated from either +5.0 VDC or +9 VDC. Telemetry Group at ARDE for their valuable contribution.
Low dropout, linear regulator regulator were used to provide Special thanks goes to Director ARDE.
these voltage outputs. Rechargeable Lithium-Ion technology
From Ultralife Inc. was used as primary power source. These
[I] M M Samman and S C Cook Configuration of Flight Test Teleme-
batteries available in laminated packaging are small in size,
tr), Frame Format. International Telemetry Conference Pg 643-
weight are high g resistant. The Battery module is potted using 650,USA,Nov 1995.
the flowable Silicon Rubber (3140 RTV, Dow Corning between [2] Secretariat Range Commanders Council, Telemetry Standards, IRIG
White Sands Missile Range,New Mexico.1993
each individual Li-ion battery cell, this act as a cushion
[31 M M Samman and S C Cook Configuration of Flight Test Teleme-
minimizing shock and vibration thus securing the batteries. tn Frame Format, International Telemetry Conference Pg 643-

Polyethylene Packaging material is inserted between the gaps 650,USA,Nov 1995.

of the aluminium housing. All internal wiring is completed

and secured prior to encapsulation.




P.K.Saha, B.K.Sahu, P.Dhande, Ms C.P.Mahajan
Armament Research & Development Establishment, Pune-411021

In terms of modern day's war perspective, accuracy, lethality and economic use of explosive is playing an
important role. We not only have to kill the target with maximum lethality but also with minimum resources. There is also a
need to counter the enemy's effort to deceive our system's intelligence. For that an intelligent and optimum rearrangement of
munitions and specific sensors is required. This paper presents the design concept of intelligent actuation of scattered
munitions. It is required for effectively handling the countermeasures applied to scattered munition by enemy to blast it
prematurely or erroneously. ECCM feature is to be provided so that enemy cannot deceive the munition by only wirelessly
transmitting the signature of the target for which the munition is made. It is also required to maximize the lethality for the
type of target chosen by actuating the selective munition(s) situated at the optimal distance from the target. If it is a soft
target like a human being, only one munition will suffice, but if it is massive like combat vehicle having thick armor then
plurality of munitions can be selected for blast. This paper presents a practical approach to solve all these problems.

In the present paper the configuration considered for scattered munition is in random orientation with varying
distance between successive rows and columns in general. The maximum distance between any two neighborhood
munition is considered to be 50 meters(Maximum). Each of the munitions will know its position with respect to other in
terms of coordinates. One RLG (Ring Laser Gyro) unit attached with each of the munition will dynamically determine the
position of it during the handling. The munitions will keep on communicating with each other about the position of the
target. Each munitions nearer to the target will receive varying Seismic energy and acoustic energy. A crude method is
applied to resolve those signals and determine the velocity, direction and position of the target. Based on the information
taken from the distributed network one local server (one of the munition) will manage the data and will give command for
actuation of the specified munition(s). Each munition will have both client mode and server mode. If one server is destroyed
new server will be elected by the nodes on priority basis.
The power supply will be in semi-passive mode activating only when it is required implementing the power
harvesting circuits inside (to elongate the operating life of the cell). The communication will be a blending of spread
spectrum and frequency hopping (with high processing gain for anti-jamming) based. The modulation scheme will be
GMSK(Gaussian Minimum Shift Keying) based to avoid ISI (Inter Symbol Interference).The networking protocol
followed will be CSMA-CD(Carrier Sense Multiple Access with Collision Detection) based to make the mines to
communicate each other on a single channel.


The munition will receive the seismic and acoustic data through miniaturized geophones and microphones
respectively, purify the signal by equalization and noise elimination and give it to the microprocessor unit. The RLG
unit will keep on communicating with the processor assembly about the co-ordinate of the system.
The antenna assembly is interfaced with the main microcontroller unit by matching circuits. It will connect the munition
with the rest of the network by a secure RF link. X-band will be used keeping in mind the smaller size of the antenna and
higher data rate. The link will be spread spectrum modulated to avoid any jamming.
The storage battery power supply unit will provide the power to all the active circuit. Power harvesting unit is
attached to enhance the storage battery limit. The power harvesting circuit will harvest power from acoustic, seismic and RF



The shaped charge with copper liner will be in a spherical container which in turn is able to rotate in both azimuth
and elevation direction. It can be pointed in any direction where the maximum lethality is expected. For that both azimuth
and elevation motors has been provided driven by electrical power and controlled by microcontroller. The pre-fragmented
interface consists of n number of steel balls to provide the blast a certain level of lethality.

Network Initialization
Network initialization will be done during the idle time when none of the target is approaching. It may be right after the
installation. The munitions are assumed to be air delivered at the probable areas of enemies and hence will be scattered
randomly with no predefined position. Each system will have RLG (Ring Laser Gyro) system which is initialized
previously will dynamically determine the changing co-ordinate of the system. Before the air delivery the power supply of
each unit will be switched on by some automated process. Each munitions will have its unique factory identification number
engraved in some particular format.



After being scattered each munitions will flood its coordinate along with its identification number in the network
with the highest power level available to the system. Maximum power level possible is assumed to be 30 dB. The
transmission will maintain unique packet format. Packet will contain some preamble bits (NRZ type), identification
number, its current global co-ordinates Fetched from RLG unit and end bits. The networking protocol will be CSMA-CD
After sharing the co-ordinate information a predefined data burst is generated by all the nodes to maintain

Now each node will transmit different levels of calibrated power (5dB, 10dB, 15dB), FSK modulated by its own
identification number in binary form. The network capturing for transmission of power will be done again by CSMA
In this way the different power level will be received by all the nodes, transmitted by all the nodes. The relative
distance will be calculated by all the nodes with the help of general radar equation (Friis equation)in all combinations and
the calculated distance will be tallied with the previously received global co-ordinates.

Based on the received power different groups (each including 5 to 6 munitions) are formed. Nodes with received
power below some threshold level (-50 dBm) will be rejected for that particular group. One munitions can exist in one
group. One munition for server will be selected for each group based on the maximum power reception by all its clients.
Now the network is said to be initialized and ready for operation.

The logical flowchart is shown in Figure 2



Target Sensing and Actuation: A logical flowchart has been shown here followed by its detailed description.

Array of
Seismic Transmitting Calculation of absolute
sensors Calculation of
angle target position
approximate position
(Algorithm -1) information to
Array of
acoustic Determination of Calculation of locus of
No signature target(Algorithm -3)

Determination of
severity of target in Is the
Confirmation of signature
terms of known signature
known? from all its clients

(Algorithm -4)

Determination of
Height of the target
from the stored

Selection of nodes from Giving actuation Giving azimuth and

geometrically optimized command to elevation data to the
algorithm particular node(s) required node(s) Is server
involved in
one the
Upgrade the

Flood the
Is delay required message for
(future predicted network re -
position is found

Yes Actuation of



Array of seismic & acoustic sensors

Array of seismic and acoustic sensors are used for sensing the target. The arrays are proposed to be circular with a
fixed radius and each of the sensors are properly indexed and uniquely defined by its exact co-ordinates with reference to
munitions global co-ordinates. Sensors are positioned in a periodic and fixed angular distance with respect to centre of the
circle referred.
Sensor will be passive type and can be anything starting from geophones to miniaturized MEMS sensors. Each
sensor will be properly interfaced with the main processor of the system.

Calculation method for approximate target position (Algorithm-1)

On the approach of the target each sensors will receive some power. The received power will be varying in nature for
each of the sensors in the circular array depending on the relative position of the sensors (because all the waves undergoes
some attenuation with the distance traveled).The phase information is also considered. For each node a mathematical model
has been prepared to approximately calculate the angle of approach of the target and its approximate position in any
particular time frame. It is described as below.

Let us suppose O be the centre of any such circular array and T is the position of the target at any particular time
frame. Now at this time instant the electrical parameters (amplitude and phase) are measured for the seismic wave.
Each of the measured result is compared. In this case maximum and minimum amplitude value of the seismic wave received
will be measured at point D and A respectively (because it is in the direction of the maximum dimension, diameter, of the
circle). The difference of the value will give us the rate of attenuation of the waves. The line joining the points A and D will
give the direction of the target.
This difference in value of amplitude will keep on decreasing for the points like B and C (for all other chords in that
direction). With the attenuation factor known (at the position of the munition, generally it varies slightly from position to
position) the algorithm will search for all the amplitude values to find the combination like B & C.
Now the line joining BC and AD will intersect at point T. So T is the expected position of the target. In such a way all
intersecting points will be found out and finally one intersecting point will be chosen by the method of average. All the
erroneous noisy data leading to incorrect points and direction will be eliminated by the algorithm. Detailed mathematics is
given below;



Direction (angle) finding

Suppose any two adjacent sensors subtends an angle á at the centre of the circle and let point 'D' is Kd-th sensor of the
array. So direction of the target will be è=Kdá

Approximate position finding

In the figure the triangles TDC and TBA is similar. So from the property of similar triangle we can say that

Now AD, DC, BC and AB are all can be expressed in terms of radius(r) of the circle and the angle of the points A,B,C
and D(integer multiples of á) which are all known.
So from this the length TD and TC can easily be calculated which directly defines the position of the target.
Expressions for AD, DC, BC and AB is given below


The final expressions of TD and TC are given below in terms of DC, AB, BC and AD.

Now knowing the coordinates of all the sensors and the angle è, the co-ordinate of the point T (the target)
can be found out.



Calculation method for absolute target position (Algorithm-2)

Now all these angle information which is calculated at each munition's end is transmitted to the server of the group. Now
each server will calculate the absolute position of the target as described under.

Each of the nodes will have its own global co-ordinates and angle information from previous processing. Suppose the t-th
node has global co-ordinates (Xt,Yt) and angle ãt with global reference. So equation of the straight line can be obtained as
(X- Xt)=tan(ãt)(Y- Yt).
In this way in the server straight lines for all its clients are calculated as shown in the figure-4 for the nodes
Ideally all these straight lines should be meeting at a single point. But in practical cases due to some measurement
error or noise or data corruption during transmission and reception these meeting points will be scattered somewhat. In the
figure these are i, j, m, n, o. It is obvious that smaller the area of the polygon 'ijmno' greater is the accuracy of the system.
The decision whether to actuate the system or not depends on the area of the polygon. If it is too large engulfing all the mines
it will postpone the decision of actuation to avoid a certain misfire. In this process the server will also consider the
approximate calculated co-ordinates of individual munitions.
If the area of the polygon is below a certain tolerance value the final co-ordinate of the target is determined by
simply averaging the co-ordinate points of the polygon.



Calculation of approximate locus of the target (Algorithm-3)

It is the responsibility of the server to note down the co-ordinate points of the target with respect to changing time. So in
this way an approximate function of ordinates with change of time is obtained. Upto third order terms of Taylor series is
considered to convert the assembly of discreet co-ordinate points to a continuous curve.
Finally the functions are obtained in the form X(t) and Y(t).
Now needless to say a first derivative of the function will give the velocity and the second derivative will produce the

Signature identification (Algorithm-4)

With varying time frame all the nodes will keep on sensing the signature of the target and keep on tallying the existing
signatures. A block diagram is shown to elaborate this.

The server will try to compare the retrieved signature with the existing ones with all combination of gains (some
discreet value between 0 to 1) as shown in the figure. If a match is found the server will try to determine the severity and
height of the target from its existing database. Accordingly it will calculate the number of munitions required to kill the
target. Say for human only one munition is sufficient. For normal vehicle some multiple numbers. For combat vehicle even
more. Exact figure will depend on the amount of the charge the individual munition is having.
If no match is found then it will try to determine range between which existing signature(s) does the current one fall
and act according to that range of signature(s) using a fuzzy logic based technique.



Selection of nodes for actuation (Algorithm-5)

The server in some particular group will try to select the nodes in such a way that the total distance between the
target and the selected nodes will be optimized. It calculates the distance from co-ordinate reference of each node as
shown in the figure like D's. Find its sum and try to optimize the expression
F= ÓDi
By selecting or rejecting the nodes. For the scarcity of suitable nodes it tries the next best possible combination.

Final actuation & network re-initialization

With the proper determination of the locus of the target the future position of it can be predicted by extrapolating
the existing data. Once the position, time of actuation and munitions for blast is selected the command is sent to the
particular nodes with all the information of the target. This information is;
1) The current location of target.
2) The future predicted location for the blast (Should be very near to the current location).
3) The target information like its severity, height, and massiveness uniquely determined from the signature.
From these each node will calculate the elevation and azimuth motor data and feed it to its steering mechanism,
which will accurately align the copper liner towards the target and wait for the proper time to commence. Sufficient latency
period will be provided to complete these jobs. If any server is involved in the blast it will send a command to all the nodes
for network re-initialization before destroying itself. In this way the complete cycle repeats.

The aforesaid concept will be very useful because it will intelligently actuate the munition matching the human
thoughts. To initiate the explosive at the right time, correct point and in the proper direction it takes the help of its inherent
intelligence implemented through different algorithms with sharpness and precision. The distributive processing feature
enables it to process numerous sensed data and bring out its essence with a speed which a single processor can't do. It can be
deployed organically and is autonomous in nature. In that way there is a significant reduction of the time required to lay the
munitions. No preplanning and human intervention is required. It assures minimum logistic burden and hence reduces



complexity to a great extent. It poses a robust obstacle to disrupt enemy target and can be used both for offensive and
defensive action. In that sense it becomes a suitable candidate for an effective future weapon..

Authors are thankful to Director ARDE Surendra Kumar and Associate Director A.M.Datar for their valuable
guidance and permission to present the paper.


1. Wireless Underground Sensor Network : www.sciencedirect.com

2. Wireless Communication Technologies By Narayan Mandayan (Rutgers University)

3. RF & Microwave Handbook By John Michael Golio

5. RF interferometer/Doppler target location system. US patent No. 4509052

6. Method of improving seismic resolution US patent No. 741491

7. Seismic-acoustic low flying aircraft detector. US patent No. 4630246.




P Dhande, BK Sahu,PK Saha,Ms. CP Mahajan

Armament Research & Development Establishment, Pune-411021

This paper describes the design concept for setting the time data (in hours/days) remotely in munitions. The wireless distance is
up to 100 cm. The systems (Transreceiver modules) for setting the time data in remote munitions utilize the concept of
Magneto-induction coupling method. It consists of miniaturized time retention transreceivers (Integrated Circuit, fitted inside
the munitions) including an antenna using a local time setter system (Hand-held transreceiver, in the hand of user).

KEYWORDS: Handheld Time Setter Transreceiver, Miniaturized Time Retention Transreceiver

In the earlier days the time data (in hours/days) had been set in the sub-munitions at the time of manufacturing itself
and have no flexibility to change it into the field by the users. This technique forced the users to maintain different inventories
of sub-munitions for different days setting. Remote time data setting concept will not only help the user to eliminate the need to
maintain the inventories of munitions for different days but at the same time also give flexibility to set the time (according to
the war scenario) in individual munitions before its deployment.


MAGNETO-INDUCTION coupling method uses various frequency bands to communicate with the time setter.
Figure 1 shows the most commonly used bands. The main frequencies are 125 kHz LF, 13.56 MHz HF, UHF (860-
930MHz), and 2.45 Ghz.

Different frequency bands are needed for various applications in rugged environments. For instance, metal and lossy
material effects as well as water and human body absorption are more detrimental to system performance at UHF and higher
frequencies than the lower frequencies. At 125 KHz frequency there are so many AM broadcast signal present in the
environment that will triggered our system even if no actual power is transmitted. Higher data rate is also achieved with higher
frequencies and anti-collision speed is limited at lower frequencies as well. Read range is also higher at higher frequencies (~3-
5 m in free space) than at the lower frequencies (~30 cm) That's why there is a trade-off between higher data rate, higher anti-
collision speed, and higher read range at UHF and 2.45 GHz compared to the better performance in rugged environments in the
presence of metals, lossy materials, and humans at the lower 125 kHz and 13.56 MHz bands.
Based on above study it is proposed that operating frequency of 13.56 MHz (HF band) is very much suitable for our
application. The following are some of the benefits and limitations of High Frequency (HF) Remote Time Setting System:




1. HF Penetrates most materials well including water and body tissue.
2. HF Time Retention Transreceivers can easily be embedded into non-metallic items.
3. HF does not affected by electrical noise that may be generated by electric motors in a surrounding environment.
4. Higher data transfer rate, the higher the frequency, and the faster the communication.
5. HF Time Setter Transreceiver can communicate with multiple Time Retention Transreceivers simultaneously in short
6. HF Time Retention Transreceivers have larger memory capacity.

1. Not as effective as LF in the presence of metal and water.
2. Read range is less than two meters.


HF Band (13.56 MHz) generally utilize the concept of Magneto-induction coupling method, in that the data is transferred
remotely through magnetic field (Fig 2), which required a simple coil antenna as compared to the modulated backscatter
coupling method in which the signal is transferred remotely through electric field, which required larger antenna at HF band.

Magneto-induction is a phenomenon through which a change in the magnetic field of a source such as a time setter
transreceiver creates a voltage level in a remote circuit such as a time retention transreceiver. A parallel circuit used to tune the
time retention transreceivers operating frequency to that of the time setter transreceiver is formed by a capacitor in parallel to
the time retention transreceiver antenna coil, which behaves as an inductor. At 13.56 MHz frequency, the receiver's antenna
interchanges energy at a particular rate, called resonant frequency, which depends on several design parameters such as size of
the coil, number of turns or distance between the capacitance plates, among others. At this time high currents are generated in
the time setter transreceiver antenna coil by the method known as resonance step-up. Resonance step-up occurs when the
frequency of the transmitted signal becomes similar to the resonant frequency of the circuit. At this point a portion of the energy
stored as magnetic form in the inductor is transferred into electrical form in the capacitor. Before achieving a stable state, a
temporary adjustment in the energy levels at the time retention transreceiver inductor and capacitor occurs; this whole process
explains the before mentioned resonance step-up effect. After reaching the stable state at resonant frequency, the received
energy is “stored” in the time retention transreceiver by transfer back and forth from the capacitance to the inductance, in
accordance with the required field strengths for the operation of the remote time retention transreceiver. A constant current is
required for IC's operations; this can be achieved by using a rectifier, a built-in component which transforms the alternating
energy arriving at the time retention transreceiver into a constant current. As a general principle it is possible to say that
inductively-coupled systems are based upon a transformer-type coupling between the primary coil in the time setter
transreceiver and the secondary coil in the time retention transreceiver.




The working schemes described as:
1. Handheld Transreceiver (as shown in Fig.3) transmits RF energy (signal) towards the munitions consisting of time
retention transreceiver (as shown in Fig.3).
2. This RF signal power up the munitions transreceiver. After powering up, munition transreceivers generates unique
Identification Number (data) and transmit towards handheld transreceiver.
3. Handheld transreceiver captures ID data and displayed it on LCD display for the identification of the munitions.
4. Handheld transreceiver then transmit time data (in hours/days) into individual munition transreceivers. As soon as
the time data is set into the munition transreceivers, these transreceivers send feedback signal towards handheld
transreceivers for user conformation.


In complete remote time setting system, design and development of antenna coil and tuned it to 13.56MHz frequency
is the critical issue. The antenna coil has been designed, developed and tested at ARDE as shown in Fig.4. It will consist of 30
turns of copper wire wrapped on antenna former (Fig.4) having a diameter of 75mm.

Experiment Conducted in Laboratory:

Dimension of coils:
Radius = 3.75cm, No. of turns = 30, radius of copper wire = 0.0225cm
Coil height = 1cm, conductance of copper (ócu) = 5.813õ107 Ù-1 m-1
Capacitance at Tx side = 22pF
Fig.4 Antenna coil
Capacitance at Rx side = 12pF (Copper conductor wrapped on former)

Operating frequency = 13.56MHz

RF measurement and test equipment used Rx coil
Tx coil
RF signal generator: To provide power to transmitter antenna coil (23 dBm)
RF spectrum analyzer: To observe the received power spectrum of receiver
antenna coil 100 cm
Oscilloscope: To measure voltage and current on transmitter and receiver
antenna coil.
Fig.5 Transmitter and Receiver coils



Observation table for received power Vs distance between Tx & Rx

1) Received power when both coils are in same plane (Fig.5),
Transmitted power = 16 dBm (39.81mW)



2) Received power when coils are in same plane and are moved according to
Fig.7 Transmitted power = 16dBm (39.81 mW)

Distance between Received power (in

transmitter & receiver dBm)
(in cm)
0 -20.19
5 -19.70
10 -19.97
15 -20.45
20 -22.10
25 -23.60
30 -23.70
35 -24.21
40 -25.95
45 -27.30
50 -27.95

Table 2



3) Received power when plane of coils are perpendicular to each other and receiver coil is moved according to Fig.9
Transmitted power = 16 dBm (39.81mW)





4) Received power when both coils are in same plane and receiver coil is covered with aluminum sheet of thickness 0.19mm.
Transmitted power = 16dBm (39.81mW)

By comparing Table 1 and Table 4, it is clear that if time retention transreceiver is covered with Aluminium sheet, received
power reduced by nearly 7dBm.



The above said concept for setting the time data remotely in munitions is very efficient which will not only help the
user to eliminate the need to maintain the inventories of munitions for different days but at the same time also give flexibility to
set the time (according to the war scenario) in individual munitions before it deployment.

Authors are thankful to Director and Associate Directors of ARDE for their valuable guidance and permission to
present the paper.


1. RFID Technical Tutorial and Threat Modeling Version 1.0 by Neeraj Chaudhry, M.S., Dale R. Thompson, P.E., Ph.D.,
Craig W. Thompson, Ph.D.
2. Design and characterization of RFID Modules in multilayer configurations by Sabri Serkan Basat.
3. A Radio-oriented Introduction to Radio Frequency Identification by By Daniel M. Dobkin, Enigmatics and Titus
Wandinger, WJ Communications
4. Radio Frequency Identification based Manufacturing and Distribution: An Exploratory Overview by Arunava
Banerjee and Ashok Kumar Jha, UG Student, Assistant Professor, Department of Production Engineering and
National Institute of Technology, Jamshedpur.
5. http://www.rfidjournal.com/article/articleview/1338/1/129.
6. K. Finkenzeller, RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification, 2
ed. New York: Wiley, 2003.
7. S. Chen and V. Thomas, “Optimization of inductive RFID technology,” in Proc. IEEE Int. Symp. Electronics and the
Environment, 2001, pp. 8287.




Ms Savita, Ms Prachi Patel, Pramod Dhande, P K Saha, Ms C P Mahajan
Armament Research & Development Establishment, Pune-411021

The new trend for next generation of underground munitions deals with the wireless operation of these systems. This
provides flexibility of operation to the user (i.e. man in loop) , safety during recovery and passing of our own troops and
optimization of laying the static field considering the system constraints, for establishment of any remote wireless
communication link, selection of frequency is the very first and foremost requirement.
To make wireless underground communication possible, the Electromagnetic (EM) waves need to penetrate to the
buried munitions. Depth of penetration of EM wave depends on the dielectric properties of the medium. According to field
requirement, munitions can be buried in marshy land, sandy land, rocky or normal soil, each having different dielectric
properties and hence different depth of penetration. Depth of penetration of EM wave also depends on the frequency and
polarization of EM wave. Since attenuation increases with frequency, high frequency radio waves do not penetrate far
through soils and rocks, while at low frequency, electromagnetic fields no longer travel as waves but diffuse, which is the
realm of inductive EM or eddy current measurements. Hence frequency should be optimized. It is also important to note that
the dielectric may not be isotropic and may contain a mixture of mineral particles. In general, the larger the difference in
dielectric constant of soil layers, the larger the reflection. One can increase exploration depth by increasing the transmitter
power. If transmitted signals are of high power, then they may interfere with nearby communication equipments present in the
This paper includes theoretical study of penetration of electromagnetic waves, the absorption losses in soil and the
reflection losses at the interface of the air and soil. Based on this study, the paper also describes the parameters to be
considered for selection of frequency for underground application along with the constraints such as size, penetration depth,
signal attenuation, dielectric constant of various soils, etc.

The concept of remote activation/deactivation supports the activation and deactivation of buried munitions
remotely from a safe distance (1km or more). This requires a wireless communication link from air to underground munition.
For any wireless communication system, selection of frequency is main criteria. In this system, selection of the frequency is
very crucial because medium is changing from air to ground (1 km in air, and upto 15 cm in ground), several factors such as
free space path loss, reflection at air-ground interface, depth of penetration in soil, absorption losses in soil(which depends on
soil properties) affects the communication link. All these parameters should be considered during selection of frequency.
In this paper we discuss how these parameters affect the communication link and based on this we optimize the frequency.

For remote activation and deactivation of buried munitions selection of frequency depends on several factors. These are:
a) Free space path loss
b) Depth of penetration in soil
c) Absorption loss in soil
d) Vegetation scattering loss
e) Size of antenna and availability of components

a) Free space path loss: Electromagnetic waves (EM) suffer from path loss due to spherical wave front spreading
while traveling in free space. Path loss is given by the formula (4ðd/ë)2 ,where 'ë' is the wavelength of the EM wave and 'd' is
the distance traveled in free space. Path loss is directly proportional to square of distance and square of frequency; hence path
loss increases with frequency and distance. If we go for high frequency path loss is more and more power is required to
overcome these losses. But transmitter power cannot be increased above a certain level in order to prevent interference from
other operating systems. Repeaters can be used to increase distance.



b) Depth of penetration in soil: Depth of penetration (ä) of EM wave in any medium is given by
ä = 1/v (ðfµó)
Depth of penetration in any medium is inversely proportional to square root of frequency. For high frequencies depth of
penetration is less and for low frequencies depth of penetration is high. So we have to select operating frequency that will
travel at least 15 cm into the ground. Fig.1 shows graph of penetration depth Vs frequencies for different medium. From graph
it is clear that for a depth of around 15 cm under the earth for the humid soil (worst case) the operating frequency comes out to
be around 1 Ghz.

c) Absorption losses in soil: Since munitions are buried into the ground, EM wave should have sufficient power to
overcome the absorption losses in soil to reach up to munitions. Attenuation due to soil absorption is the major concern in
underground wireless communication. Predicting path losses for an underground channel is a challenge due to the
inhomogeneous nature of ground in which munitions is to be buried.
Characteristic of soil can be described by its complex permittivity given by:
ç = å - j 60*óë
ç = Complex permittivity
å = Permittivity of soil
ó = Conductivity of soil
ë = Wavelength
In case of earth surface, å and ó entirely depend on the degree of humidity. Presence of humidity in the ground
produces solutions of salts which can considerably increase conductivity at low frequencies.
From the above equation it can be seen that for any given soil there is a transition frequency for which real and imaginary part
of ç are equal. If the frequency is lower than transition frequency then the ground behaves as a conductor for all practical
purposes. If the frequency is greater than the transition frequency then the ground behaves as the dielectric for all practical
For any non-maritime medium (soil) the transition frequency lies between 0.6MHz and 6MHz.

Several factors such as composition of soil, including its water content, particle size, density and temperature
directly affects the attenuation of EM waves passing through the soil.



Now we discuss the effects of these parameters on signal attenuation.

Water content: Soil water content is the most significant parameter when predicting signal loss through a given type of
soil. Any increase in the water content of a soil will make the channel significantly more lossy. At a frequency of 1 GHz as
water content rises from dry to 13%, volumetric loss/meter increases about 137dB. The effect of rise in water content is
highly dependent on type of soil, for example sandy soil show less attenuation as water content increases than do clay soil.
Attenuation caused by soil water content is also dependent on frequency being used. Lower frequencies experience less
attenuation at a given water content than higher frequencies.

Particle size: Soils are classified by the diameter of their particles and are generally described as some variation of sand,
silt or clay. Sandy soil produce the least amount of loss and clay soils the most. In addition different soil particle types
respond differently to changes in water content.

Density: Increasing soil density increases path loss. The denser a soil, the greater the signal attenuation.

Temperature: Increasing the temperature of soil; changes its dielectric properties and will increase signal attenuation.
Additionally changes in temperature will affect the dielectric properties of any water present in the soil.

d) Signal attenuation in Vegetation: When EM wave reaches the ground it may pass through vegetation, it suffers
additional losses due to vegetation. This happens due to scattering of energy out of the radio path. These losses depend on the
species and density of vegetation. Figure 2 shows the attenuation due to vegetation over the frequency range of 30 MHz to 30
GHz for vertically and horizontally polarized wave. Below about 1 GHz, there is a tendency for vertically polarized signals
to experience higher attenuation than horizontally polarized signal.



e) Size of antenna and availability of components: Size of antenna depends on frequency of electromagnetic wave. For
efficient radiation and reception of EM waves, antenna size should be comparable to the wavelength of EM wave. Due to
size constraint we cannot go for low frequency. At high frequencies size of antenna is small as compared to low frequencies.
Also at high frequencies components are easily available.


In air to underground wireless communication, to calculate received power at buried munitions link budget equation
is as follows: Pr = Pt +Gt +Gr 20 log (4ðd/ë) -Lm
Pr represents signal power at the receiver in dBm, Gt represents the gain of the transmitting antenna in dB, Gr
represents gain of receiving antenna in dB, 20 log(4ðd/ë) represents free space path loss due to spherical wave fronts
spreading , and Lm represents signal attenuation in dB due to absorption by soil and water.
During calculation of received power, polarization of signal arrived at receiver (munitions) also matters. When EM
waves penetrate/enter the ground, polarization of EM wave may change. Change in Polarization of EM wave depends on the
properties of medium in which wave is entering and angle of incidence. If polarization of EM wave changes during
communication, then receiving antenna can not extract all of the incoming power. So during design of system this factor
should be considered.

Selection of optimized frequency will only make the concept for remote activation and deactivation of buried
munitions efficient and successful.
From the above discussion it is clear that at high frequencies free space path loss is more and depth of penetration is
less, and at low frequencies size of antenna is large. So there is a tradeoff between losses and size of antenna while selecting
frequency. Considering all the factors discussed above, frequency in L-band (~1GHz) is best suited for this application.

Authors are thankful to Director and Associate Directors of ARDE for their valuable guidance and permission to
present the paper.

1. RF & Microwave Handbook By John Michael Golio
2. Wireless Underground Sensor Network : www.sciencedirect.com
3. Wireless Communication Technologies By Narayan Mandayan (Rutgers University)
4. Elements of electromagnetic By Matthew N. O. Sadiku
5. Electromagnetic Waves By R K Shevgaonkar




KD Deodhar
Armament Research & Development Establishment, Pune 411 021

This paper outlines the development efforts of state–of–the–art Hybrid Composite Technology for Gun Barrel
Application for the first time in the country. To meet the Infantry requirements for Light Weight Guns especially for high
altitude mountain warfare, ARDE has concentrated the efforts on reducing the mass of the gun. Composites with carbon fibres
as reinforcement and epoxy resins as the matrix materials offer very good mechanical and physical properties like very high
strength – to – mass ratio as well as very high tensile modulus-to-mass ratio. The advantage of composite materials is that the
properties can be tailor-made, and when reinforced fibres are oriented in the hoop direction, thickness could be optimized for
lower mass.
An attempt has been made to implement the technology to realize the 84 mm Light Weight Launcher (LWL) shoulder
fired anti-tank infantry weapon. 84 mm LWL has undergone a series of User and troop trials at high altitude and extreme cold
climatic as well as at desert & hot climatic conditions. Recently, the technology has been transferred to OFB for bulk

There was a requirement for an anti-tank infantry gun system from army, which is light in weight, man portable,
capable of shoulder firing, safe, reliable and robust to be deployed in varied terrains and operated in all weather conditions. The
then COAS asked us to take up the project on priority to make light weight rocket launcher. Feasibility study was carried out
and a project titled “Design and development of 84 mm Light Weight Launcher” was undertaken. It was successfully
completed within a record time of thirty months and the same technology has been handed over to OFB for indigenous bulk

Fig.1 : 84 mm Light Weight Launcher (LWL)

Last few decades, the use of composite materials in the industry has been growing exponentially, mainly because of
the reasons that the material properties can be tailor made as per demand of the product. Variety of combinations is available
for the composites. A vast spectrum of materials & manufacturing processes is available for selection giving flexibility for
wide choice.
The composite technology is being used mostly in aerospace industries, Automobile industries, in a big way. Defence
industry is also not behind and right from structures, armour plates, missile casings, Naval boats etc. are made of composite



84 mm Light Weight Launcher barrel has been designed & developed with state-of-the-art hybrid composite
technology for the first time in the country. This has resulted in drastic mass reduction, without compromise on the parameters
like stress & strain due to pressure experienced by the weapon. The hybrid composite barrel technology has been developed by
using carbon fibers as reinforcing agent and epoxy resins are as binder matrix material. The carbon composites exhibit
superior mechanical properties such as high specific strength and high specific modulus i.e. Strength-to-weight ratio as well
as elastic modulus-to-weight ratio is very high. This property gives the component required strength/stiffness still very low in
weight. That is how the mass of the weapon, 84 mm LWL is below 11 kg was achieved, still it will withstand the firing stresses
generated due to pressure to the order of 100 MPa. The carbon composite has been built up over a thin riffled steel tube by
“hoop over wound” technique. The high strength, high modulus carbon fibers impregnated through epoxy resin system have
been wound by filament winding technique and cured in controlled atmosphere by autoclave technique.
After in-situ winding of composite barrel on a mandrel, it has been put for curing cycle in an autoclave. The temperatures, rate
of rise of temperature and cooling, air pressure, vacuum etc. are under control to culminate the properties as desired. It will
provide good structural integrity, robustness etc.

Why hybrid composite?

To induce certain properties required, the composite barrel has been designed with hybrid reinforcement of carbon
and glass in appropriate proportion. The composite layers will be laid over a thin riffled steel liner in various discrete layers of
carbon and 'E' glass alternately. The carbon will provide fatigue resistance, high tensile modulus and partial conductivity for
heat dissipation. The E-glass reinforcement helps to have high resistance to crack growth rate, high strain carrying capacity,
high tensile strength and higher percentage elongation.
The matrix material used is epoxy resin which is possible to be selected from wide variety available, since large
number of starting materials, curing agents, and modifiers are available.
Absence of volatile matters during cure, low shrinkage during cure, excellent resistance to chemicals and solvents
and excellent adhesion to a wide variety of fibers, fillers & other substrates are the advantages of the epoxy resins for selecting
as a matrix material. Carbon composites provide high performance, strength having low density, fatigue resistance, wear
resistance, corrosion resistance, and high temperature resistance having unique damping characteristics. Tensile strength of
carbon fibers increases with temperature.

Fig. 2 : Construction of Composite Barrel

Out of various manufacturing processes, filament winding technology has been selected as we are utilizing the continuous
fibers as reinforcement, which has been oriented in the maximum stress direction to get optimum wall thickness. The thin steel
liner has been used to provide all the internal features of the barrel like chamber and bore with a set of helical grooves to impart
spin to the projectile for stability in fight with out changing the internal ballistics. Other advantages are:
i) protection from direct contact of very high pressure & high temperature propellant gases, so that charring of
composite material is avoided.
ii) Maintain mirror finish of the chamber & bore.
iii) Reusability of the launcher.
An important aspect of carbon composite with steel liner is the interface. Special coupling agent has been employed to
avoid de-bonding, so that the composite barrel can be deployed for use in negative temperatures say -40°C at Himalayan
ranges as well as high temperature say +50°C like at Pokharan ranges.



The steel liner, mounted on a suitable mandrel, wound on a filament winding machine and the barrel in wet condition kept
rotating with low speed to allow it for semi-cure and complete polymerization of matrix material, say about 20-30 minutes, is
then put into an autoclave. Then mandrel, liner wound with barrel will be put into a vacuum bag and placed into a chamber of an
auto clave where the hydrostatic air pressure is applied over the vacuum bag. The outlet of vacuum bag is attached to vacuum
pump. The temperature inside the chamber also can be monitored including rate of heating and cooling.
Generally the barrel has been kept for better curing at constant pressure and temperature for specific time to get full
polymerization and cross-linking called as thermosetting to induce the properties. If some air traps, voids or volatile gases
come out of chemical reaction, they will be taken away by vacuum pump to give flawless, uniform, sound and strong structure.


As a normal practice, right from raw material, quality checks were carried out, however specific Non-destructive and
destructive tests were included into quality assurance plan. Since the weapon is a shoulder-fired weapon, the safety of Jawan is
of paramount importance, hence, ultra sonic C-Scan test, acoustic emission test (AET) has been introduced during
manufacture itself to ensure defect-free product. Additionally environmental tests as extreme hot and cold conditions have
been included in QAP. Destructive tests like tensile strength test by split disk method, hydrostatic pressure test, fibre content
test etc. have been thrusted to ensure safety. Test coupons for these destructive tests will be manufactured along with each
barrel simultaneously to have same raw materials, same process, same proportions and same layer morphology.
Finally each and every composite barrel will be assembled to a full weapon and subjected for dynamic proof firing test at
20 % higher than service pressure to reveal out unknown defects if any.

Fig. 3 : Set-up for C-Scan



Fig. 4 : Firing at Leh

State-of-the-art hybrid composite barrel technology has been established first time in the country. The technology has
been validated by various users, troops at varied terrains including high altitude viz. Tangatse, Leh and Pokharan desert by
subjecting these composite barrels to dynamic firing trials and recommended for induction. Compatibility with advanced
ammunition has also been validated. This technology has been transferred to Ordnance Factory Board for indigenous bulk
production at Gun and Shell Factory, Kolkata.
The technological spin-off is being explored for higher caliber guns and mortars for the futuristic weapon systems and
automated salvo firing platform for naval application.




Sachin Sood, G Narasimha Rao, G Chandramouli, Dr RR Panyam
Defence Research & Development Laboratory, Hyderabad

Indigenously developed medium range surface to air (SAM) missile system Akash, has completed
all development trials, user trials and being inducted in to the Indian Armed Forces. The weapon system
consists of number of surveillance radars, guidance radars, control centers , missile launchers, supersonic
missile Akash and many other ground support equipments. Each weapon system element is a system of
systems consisting up to sixty multi disciplinary subsystems. Since, first time such complex and varied
multi discipline weapon system is developed in India, the type of system integration and testing
requirements were to be evolved abinitio, in the absence of precedence and published literature.
Two critical technologies ie Integral rocket ramjet technology and multi function phased array technology
were developed as a part of the programme. A close monitoring was required for stage wise development,
integration, testing, clearance and revision of design parameters/test setups. Frequent interactions with
developing agencies, users, manufacturing agencies were required to evolve the performance and test
methodologies. For the first time a Defence R&D project is executed in parallel with Technology
The typical integration and testing tasks involve interfacing Radars, Command Control Centres,
Launchers and Missiles. The typical functions include establishing secured communication, Muti-target
tracking, fusion of target track data from various radars, display of a single integrated air picture, generate
fire control commands, automated missile launch, missile acquisition, missile tracking, guidance and kill
assessment. These functions are automated for ease of combat decisions.
A number of onboard and ground elements are required to work in tandem for the multiple missile launch,
control and guidance. The methods used, setups built and evaluation techniques evolved etc are an
experience by themselves. The DRDO teams and production teams could learn many techniques during
the course of development.
As a part of user trials, the weapon system elements were subjected to mobility and performance trials in
the desert terrains. The missile system performance is evaluated by flight testing against fast moving
targets as per user specified mission profiles. In these trials, all the equipments including support
equipment have been transported to user specified locations and tested for deployment and performance.
This paper discusses how the technologies were developed and how they were integrated, tested and
evaluated for Indigenous SAM system before being subjected to field tests and user trials.



Test & Evaluation
Guidance & Control


Jaywant P Kolhe, Johni Bhasha Shaik,Geeta A Bokil, SE Talole

Defence Institute of Advanced Technology, Pune -411 025

In this paper a robust roll autopilot design based on Extended State Observer (ESO) approach is proposed. The effect of the
disturbances acting on the system as well as the states of the transformed model are estimated by ESO and are used for design
and augmentation of a state feedback controller. Closed loop stability of the autopilot is established. Simulation results are
presented to demonstrate the effectiveness of the ESO in estimation of the uncertainties and states and in meeting the
objectives of the design

Many cartesian controlled missiles are required to be roll position stabilised in their flight as the rolling motion results into
several undesirable effects [1]. Roll motion causes cross-coupling of pitch and yaw channels which can lead to instability apart
from causing incorrect execution of guidance commands. Command guided missile will need to carry resolvers on board if
allowed to roll freely. High roll rate may lead to loss of target during terminal phase. Also it can be shown that high roll rate
coupled with servo lag may cause instability of the guidance loop. The missile may roll in flight due to various undesirable
rolling moments/disturbances acting on the missile such as the moments arising due to aerodynamic imbalances on the body of
the missile, rigging errors and external disturbances etc. Whenever such moments act, the missile will tend to roll. In such
circumstances, it becomes necessary to stabilise the missile at desired roll position in order to overcome the undesired effects
resulting from the roll motion. To achieve this, the missile is equipped with roll autopilot that controls the roll attitude of the
missile at desired roll angle thereby arresting unwanted rolling motion of the missile.
The roll autopilot can be designed by employing various approaches. In classical approach, compensators are designed by
considering conservative stability margins [1]. Such designs may offer inferior performance due to reduced closed loop
bandwidth. On the other hand, the modern control approaches require exact mathematical model. Further, many of such
approaches need some information on characteristics (such as bounds) of the disturbance acting on the system. Controllers
designed using simplified models may become unstable when un-modelled dynamics are considered as shown in Ref. [2].
One effective strategy to address these issues is to design a controller based on disturbance estimation. In this approach, the
effect of the disturbances acting on the system is estimated on line and the same is cancelled by augmenting a controller
designed for nominal system. The disturbance may represent effect of all uncertainties and inaccuracies arising in practical
control system such as the effect of un-modelled dynamics, parametric variations, and external disturbances acting on the
system. Many approaches such as the Uncertainty and Disturbance Estimation (UDE) [3], Unknown Input Observer (UIO)
[4], Disturbance Observer (DO) [5], exist in literature that can be used for estimation of the disturbance acting on the system.
Some other approaches such as the Time Delay Control [6] are also used for the estimation of the effect of the uncertainties.
These approaches usually offer estimation of the disturbances only acting on the system. Frequently one requires the state
vector also to be available for implementation of controller necessitating a design for separate state observer. The issue of
estimation of uncertainty as well disturbance in an integrated manner is addressed by state and perturbation observers such
as the Extended State Observer [7].
In this paper, a robust autopilot for roll position stabilisation of a cruciform (Cartesian controlled) missile is designed using
the Extended State Observer approach. As the mathematical model considering the real disturbance does not satisfy matching
condition, state coordinate transformation is performed in such a manner that an equivalent disturbance acts in the same
channel as that of control signal. The effect of the disturbances acting on the system as well as the states of the transformed
model are estimated by ESO and are used for design and augmentation of a state feedback controller. Closed loop stability for
the system under the proposed controller is established. Simulations are carried out by considering significant external
disturbance acting on system to demonstrate the effectiveness of the approach and the results are presented. The remaining
paper is organised as follows: In section 2 the description of the problem and design of the state feedback controller is
presented. In section 3 the ESO theory is briefly reviewed and applied for state and uncertainty estimation of the present
problem. Simulation results using ESO based controller are presented in section 4 and section 5 concludes this work.




This work concerns with the design of a roll position stabilisation of a cruciform missile. The aim of the autopilot is to
stabilize the roll position in spite of the external disturbances acting on the missile. It is expected that the design offers desired
robustness, stability and performance over its operational flight envelope.

A Mathematical Model
The roll dynamics of a cruciform, tail controlled missile, in the presence of disturbances is given by [1]
= p
=+lp p lx
x =
y =
f [1]

where f (rad) is the roll angle, p (rad/s) is the roll rate, are the non-dimensional aerodynamic derivatives, d is the
lumped disturbance acting on the system, x (rad) is the aileron deflection, and u is the input, i.e., commanded aileron
deflection. The servo is modelled by a first order dynamics with bandwidth of k
From Eq. (1) it can be noted that the disturbance and control input are acting in different channels and thus the matching
condition is not satisfied. For this purpose state coordinate transformation is performed obtained as follows. To this end,
differentiating the output of Eq. (1) till the input appears explicitly, one gets

=(l p - &
k )f
+ &
(l p k )f( klx
++ d&
)u ( kd +) [2]

Now denoting equivalent disturbance d e as de = d&

kd +Eq. (2) can be re-written as

f=(l p - &
k )f
+ &
( l p k )f
+( klx
)u +

Defining transformed state variables as x1 =

f &
, x2 =
f &
, x3 =
fEq.(3) can be written in state space model as



x& (lpk)x2 (lp -

3 =+ k)x3 +
)u +
de [4]

Being linear dynamics, conventional linear design methodologies can be employed for the system of Eq. (4) for design of the
controller such that the missile roll attitude is regulated as desired with satisfactory transients and steady state accuracy. To
achieve the roll position stabilisation, i.e., to ensure that the roll angle, f
,remains always zero in the presence of disturbances,
the following pole placement controller is proposed

1 [5]
u= l pkx2 -
(- (l p -
k ) x3 -
k1x1 -
k2 x2 -
k3 x3 -
de )



Obviously by choosing the values of the gains appropriately, one can obtain stability and roll position stabilisation. The design
is robust as it caters for the disturbances as well as for the model uncertainties. However implementation of the controller will
need availability of the states as well estimation of the equivalent disturbances. The same has been achieved by using the ESO
as presented in the next section.


As stated earlier, the ESO is an observer which can estimate the effect of the uncertainties along with the states of the
system enabling disturbance rejection or compensation. The ESO regards all factors affecting the plant, including the
nonlinearities, uncertainties and the external disturbances as the total uncertainty to be observed and in that sense, it can be
viewed as an unknown input observer or disturbance observer. Since the observer estimates the uncertainty as an extended
state of the original system, it is designated as Extended State Observer. Its merit is that it is relatively independent of the
mathematical model of the plant, performs better and is simpler to implement. The robustness is inherent in its structure as will
be obvious in next section. A comparison study of performances and characteristics [8] of three advanced state observers
namely High Gain Observer, ESO and sliding Mode Observer is presented and it is shown that over all the ESO is much
superior in dealing with uncertainties, disturbances and sensor noise. Several diverse applications of ESO based control
strategies have appeared in literature such as the control of induction motor drive [9] and aircraft attitude control [10]. In this
section a brief review of this approach and its application to state and uncertainty estimation for the present problem are

3.1 Concept of ESO

Consider an nth order, single input-single output dynamical system described by
(n) ( n-
a ( y , y&
y= ,K
, y )+
bu +
where, a(.) represents the dynamics of the plant (linear or nonlinear), w(t) is the unknown disturbance, u is the input and y
measured output. Let a (.) = a and b =
a0 (.) +
D b0 +
Db where a0 (.) and b0 are the best available estimates of a and b respectively
and D a and D b are their associated uncertainties. Defining the uncertainty to be determined as d @ Da+ D bu and
designating it as an extended state, n+ 1, the dynamics of Eq. (6) can be written as

x& x2
1 =

x& x3
2 =

x& xn+
n = a0 +
1 + b0 u
x& h
1 =
n+ [7]
y =
Where h is the rate of change of uncertainty, i.e. h = d& and it is assumed to be unknown but bounded function. By making
d as a state, however, it is now possible to estimate it by using a state estimator. To this end, consider a linear ESO of the form,
ˆ1 =
xˆ2 +
ˆ =3xˆ +b.e

ˆ =
xˆ +
1b.e +
n a +
bu 0 0

ˆn +
1 =

where e = y- yˆ = x1 - xˆ1 and xˆn +1 is an estimate of uncertainty. The quantities b i are the observer gains. If one chooses
the bi properly, the estimated state variables xˆi are expected to converge to the respective states of the system xi i.e. xˆi ®
xi ;
i= ,n +



3.2. Observer Error Dynamics

Having designed the ESO, it is necessary to prove the stability of observer error dynamics. For linear dynamics,a (.) in (8) is
linear in its arguments, i.e., a (.) =
-an x1 - 1 x2 L
an - - a1 xn Since the dynamics of the present problem is linear one, we continue
to represent a (.) as linear function of states and following this, the dynamics of the extended order plant is re-written as,

x& Ae xe +
e = Beu +
Bh h
Cxe [9]

where [
xe =L 1]
x1 , x2 , , xn +and
0 1 0 ¼ 0ù
ê ú
ê 0 1 ¼ 0ú
ê Oú
Ae =
ê ú
-ano - a( n-1)o ¼
ê - a1o 1 ú
0 0 ¼ 0 1ú
ë û
Be = 0 b0 0 ]
0 ¼ h [
; B =¼
0 0 0 1]
; C =
e [ 0]
10 0 ¼ [10]

Here one can note that the additional term involving h is present in the state equation. Now assuming linear gains, we get a
linear form for the observer (8) as,

ˆe =
Ae xˆe +
Beu +
y=Cxˆe [11]

Subtracting the state equation of Eq. (11) from the state equation of Eq. (9), the observer error dynamics can be obtained as

e& (A-
0 = B¢
LC)e0 +h [12]

where e0 @xˆ and

b1 b
2 ¼
n b1]

Since the pair ( A, C ) is observable, one can compute the observer gain, L and so the ESO gains, b i such that the observer
error dynamics is stable and has desired transient behaviour. As the eigen-values of the observer error dynamics given by
Eq. (12) are placed at the desired locations, it can be stated that under the assumption of boundedness of h bounded input-
bounded output stability for the linear system of Eq. (12) is assured. A special case arises when h = 0 i.e., when the rate of
change of uncertainty, d is zero. In this circumstance, the errors will go to zero asymptotically, i.e. lim e0 ( t ) =
0 The similar
result can be expected if the rate of change of the uncertainty is reasonably small.

3.3. Eso Based Controller

Now consider the transformed roll dynamics given by Eq. (4) and the controller of Eq. (5). As stated earlier, to implement the
control law, one requires the value of uncertainty, d e apart from the states and for this purpose an ESO can be designed. To this
end following the procedure outlined in earlier section, an ESO of the form of Eq. (8) with n = 3 is designed for the dynamics
of Eq. (4) by taking the total uncertainty as an extended state, x̂4 Having the estimated states, the controller (5) takes the
following form 1
u= (- l pkxˆ2 - k ) xˆ3 -
(l p - k1xˆ1 -k2 xˆ2 -
k3xˆ3 -xˆ4 ) [14]



where x̂4 is an estimate of the uncertainty d e The controller (14) is designed as the ESO based controller and is implementable.


Simulations are carried out for the autopilot, thus designed, under nominal condition (without disturbances) and also
by considering disturbances. The plant parameters of l p andlx and servo bandwidth k, are taken as -38.8542, -14062.5
100, respectively. The controller gains k ' s required in Eq. (14) are obtained by placing the poles at æ ç1 + s
÷ with a time
i è 3 ø 4

0.05 The ESO linear gains, b i ' s are obtained by placing the poles at ç
constant of t æt ö
= 1+
è4 ø
s÷with a time constant of
t = 0.01 The initial conditions of the plant, are taken as x (0) = é10 0
0 0 ù while the initial conditions for the observer are
T ë û
taken as xˆ e (0) = [0 0 0 0] Two models of disturbances are used to study the effectiveness of the approach. First the
disturbance is assumed constant with d = 1000 N - m and secondly it is assumed time-varying as d = m The
1000sin(t ) N -
corresponding equivalent disturbance for second case works out to be d e = 100000 sin(t ) + 1000 cos(t ) With these data
and disturbance models, simulations are carried out and the results are presented. Simulation results for d = 1000 N - m are
presented in Fig. 1 and from Fig. 1(a) it can be noted that while roll angle goes to 0 for ESO based controller, it does not go for
the nominal one (i.e. controller without uncertainty compensation). The corresponding control input histories are given in
Fig. 1(b). In Fig. 1(c) , the plot of the actual and estimated disturbance are presented and it can be noted that that disturbance is
estimated quite accurately. Next simulation results for time-varying disturbance d = 1000sin( t ) N - m are given in Fig. (2)
and one can note that the ESO based controller offers satisfactory performance. Note that in this case d& is not equal to zero.
Thus in the presence of the disturbance, while the nominal control is unable to keep the roll angle ( f ) to zero, the ESO based
control accomplishes its objective quite satisfactorily.

x 10
30 6 12
ESO based
aileron deflection (deg)

disturbance and its estimate

5 nominal 10
20 ESO based
8 actual
nominal 4 estimated

10 6
-10 1 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
time(s) time(s) time(s)
[a] [b] [c]

Fig. 1: Performance of Nominal and ESO based controller with d =

1000 N -

25 4 x 10
disturbance and its estimate

ESO based ESO based actual

aileron deflection (deg)

20 nominal nominal 8 estimated


15 6

10 4
5 2

0 0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
time(s) time(s) time(s)
[a] [b] [c]

Fig. 2: Performance of Nominal and ESO based controller with d=

1000sin( t ) N - m



In this work an ESO based controller is proposed for design of robust roll autopilot. The effect of the disturbance
acting on the system is estimated and a controller designed for nominal system is augmented by using the estimated
disturbance to compensate the same. Simulation results show that the ESO based controller offers satisfactory performance in
presence of large constant as well as time-varying disturbances.


[1] Garnell, P., “Guided Weapon Control Systems”, Pergamon Press, Oxford, England, 2nd Edition, 1980.
[2] Nesline, F. W., and Zarchan, P., “Why Modern Controllers Can Go Unstable in Practice”, Journal of
Guidance, Control, and Dynamics, Vol. 7, No.4, Jul-Aug 1984 pp. 495-500.
[3] Zhong, Q.C., and Rees, D., “Control of Uncertain LTI Systems Based on an Uncertainty and Disturbance
Estimator,” Transactions of the ASME, Journal of Dynamic Systems, Measurement, and Control, Vol. 126,
No. 4, Dec 2004, pp. 905–910.
[4] Takahashi, R.H., Peres, P.L.D., “Unknown Input Observers for Uncertain Systems: A Unifying Approach
and Enhancements”, Proceedings of the 35th Conference on Decision and Control, Vol. 14, No. 12, Dec
1996, pp. 1483–1488.
[5] Chen, W.H., “Nonlinear Disturbance Observer-Enhanced Dynamic Inversion Control of Missile”, Journal
of Guidance, Control, and Dynamics, Vol. 26, No. 1, Jan-Feb 2003, pp.161-166.
[6] Youcef-Toumi, K., and Ito, O., “A Time Delay Controller for Systems with Unknown Dynamics”,
Transactions of the ASME, Journal of Dynamic Systems, Measurement, and Control, Vol. 112, No. 1, 1990,
pp., 133–142.
[7] Gao, Z., “Scaling and Bandwidth-Parametrization based Controller tuning”, Proc. of the American Control
Conference, Jun 2003, pp. 4989-4996.
[8] Wang, W., and Gao, Z., “A Comparison Study of Advanced State Observer Design Techniques”, Proc. of the
American Control Conference, Jun 2003, pp. 4754-4759.
[9] Feng, G., Huang, L., and Zhu, D., “A Nonlinear Auto-Disturbance Rejection Controller for Induction
Motor”, IEEE IECON, Vol. 3, 1998, pp. 1509-1514.
[10] Huang, Y., Xu, K., Han, J., and Lam, J., “Flight Control Design Using Extended State Observer and Non-
smooth Feedback”, Proc. of the 40th IEEE Conference on Decision and Control, Dec 2001, pp.223-228.





Mallikarjuna Bommu, Venkata Rama Subbaiah B., B. Srinivas Kalyan and

C. Arjun Rao and S. E. Talole
Defence Institute of Advanced Technology, Girinagar, Pune 411 025

In this paper, a dynamic inversion based controller is employed for missile autopilot design. As the controller lacks
robustness, fuzzy logic compensator is introduced to robustify the controller. Simulation results are presented to show the
effective of the controller in presence of large uncertainties in system parameters. Further the performance of the proposed
controller is compared with some of the existing formulations and it is shown that the proposed controller offers superior

Missile autopilots are traditionally designed using the classical linear control approaches. In general, the missile
dynamics is linearized around certain finite number of operating points in the flight envelope and then controllers are designed
by employing well-known linear control strategies. For each linearization, a linear time invariant controller is obtained, which
in the local region of the fixed operating point delivers the desired performance characteristics. Gain scheduling, is then
employed to deliver the performance in the complete flight envelope of the missile. To this end, the linear control techniques
have dominated missile autopilot design over the past several decades[1-2]. Lately, to cater for large uncertainties, various
robust control techniques such as H-infinity [3] have been used to design robust autopilots.
Although design of autopilots by linearizing the missile dynamics is one of the most prominent approaches, the
controller thus designed may offer unsatisfactory performance especially when the dynamics is highly nonlinear and
undergoes large motion. In such circumstance, use of various nonlinear approaches such as the Quantitative Feedback Theory,
Variable Structure Control, Adaptive Control, Predictive Control, Fuzzy Logic Control, the Feedback Linearization and
Control strategies based on Neural Networks have been proposed for designing of the missile autopilots.
In Ref. [4], a nonlinear autopilot design based on Dynamic Inversion (DI) is proposed. The design is based on two
time scale separation of the missile dynamics wherein the pitch dynamics is the fast dynamics whereas the angle of attack
dynamics is the slower one. Since the performance of dynamic inversion technique is not guaranteed to be satisfactory in
presence of uncertainties, a PI controller is employed in outer loop for robustification and the authors have shown that the
performance of the controller is superior in comparison with the pure DI based controller.
In present work a Fuzzy Logic Control (FLC) based compensator is introduced in place of the PI compensator to
obtain robustification and to achieve better performance in presence of the large uncertainties in aerodynamic parameters. The
effectiveness of the proposed controller is demonstrated by comparing its performance with the existing controllers such as the
dynamic inversion (fast as well as slow) with PI control and fast dynamic inversion with PI control and the results are
The paper is organized as follows. In section 2, a nonlinear missile model is described. The dynamic inversion based
controllers, i.e., dynamic inversion based, dynamic inversion with PI and fast dynamic inversion with PI are presented in
section 3. The formulation of fast dynamic inversion with Fuzzy Logic compensator is the subject of section 4 and simulation
results giving comparison of the proposed controller along with the rest are given in section 5. Finally section 6 concludes this

As is well-known, the equations that represent the dynamics of missiles are highly nonlinear. The nonlinearity is more
pronounced in the case missile encountering high angle of attacks. A representative pitch plane dynamics for a tail controlled
missile as taken from [1] is given by



where the various parameters and their values are:

á(t) : angle of attack(deg),

q(t) : pitch rate(deg/s)
Po = 46627.30069 newton/mt2(static pressure at 20,000 ft)
S = 0.040877338 mt2 (surface area)
m = 204.022722 kg (mass)
vs = 315.89472 mt/sec (speed of sound at 20,000ft)
d = 0.2286 mt(diameter)
M = 3(Mach number)
Iy =247.4367 kg- mt2 (pitch moment of inertia)

Ká = Ká = , Kq =

The normal force and pitch moment aerodynamic coefficients are given by

(t),M ) = sgn(á)[a n |á(t)| 3 +b n |á(t)| 2 +c n (2 -
C n ( á(t),d )|á(t)|]+ d n d(t)
d p 8M
(t),M ) = sgn(á)[a m |á(t)| 3 +b m |á(t)| 2 +c m (-7 +
(3) C m ( á(t),d )|á(t)|]+ d m d
(t) (4)
d c

a n = 19.374 rad -3 , a m = 40.44 rad -3 ; b n = -31.0225 rad -2 , b m = -64.014 rad -2

c n = -9.7174 rad -1 , c m = 2.9221 rad -1 ; d n = -1.9481 rad -1 , d m = -11.803 rad -1 .

These approximations are valid for á≤20 deg.

The tail fin actuator model is represented by 2nd order transfer function as
p wn
= 2 2
c s +wn s +
2V wn
where p and c are the actual and commanded tail fin deflections (deg) respectively and damping ratio, V
= 0 .7 and natural
frequency of wn = 20Hz The input is commanded control surface deflection, i.e., u=c. The limits on control surface
deflection are c ≤±30 deg and its rate, ≤300deg/sec. The objective of the controller in missile autopilot design is to force
the missile to track the commanded acceleration profile.

The Feedback Linearization [5] is one of the most widely studied techniques for nonlinear control system design. In
this approach, a given nonlinear system is transformed into an equivalent linear system through nonlinear state coordinate
transformation and static state feedback resulting in a linear input-state relation and is designated as state-space linearization,
input-state linearization or simply feedback linearization. When applied to output control, only nonlinear state feedback is
involved and the technique is referred to as input-output linearization, also known as Dynamic Inversion (DI) in aerospace



literature. One can find large number of papers wherein the DI approach is used for design of controller for missile autopilot.
For example an approximate input-output linearization based design for missile control by taking angle of attack as output is
presented in Ref. [6]. However the DI based controller suffers from robustness as the design assumes perfect knowledge of the
plant dynamics. As this requirement is rarely satisfied in practical systems, it becomes necessary to robustify the DI
controllers. Applications of the robustified DI controllers for missile autopilot design have also been reported in literature. For
example, an angle of attack tracking controller for missile based on input output linearization is presented in Ref. [7]. To
provide necessary robustness the controller is augmented by adding a term which is based on the assumed bound of
In Ref. [8] the input-output feedback linearization approach is employed wherein the output is taken as a linear
combination of angle of attack and pitch rate. Further the controller is robustified by using a disturbance observer. In what
follows is development of three controllers based on the DI theory. Subsequently a Fuzzy Logic Control based compensator is
presented for robustification of the DI controller.

Slow And Fast Dynamic Inversion

As the dynamics associated with the angle of attack, á, operate on a slower time-scale than that of the body rate dynamics,
time scale separation can be used in design of the pitch autopilot. Following this line, a two time-scale separation is used for
autopilot design in Ref. [4] wherein the pitch dynamics (inner loop dynamics) is the fast subsystem and the angle of attack
(outer loop) is the slow subsystem. Dynamic Inversion is then employed separately to these loops separately resulting into
Fast Dynamic Inversion (FDI) and Slow Dynamic Inversion (SDI) controllers. A functional block diagram of this scheme is
shown in Fig. 1. The inner loop inversion uses the control surface deflections to control the fast state q. This loop calculates
control surface deflection commands from the rate commands d, given by the slow inversion to achieve the desired dynamic
response. The desired dynamics given by FDI are

The slow inversion replaces the actual ác dynamics with the desired dynamics

where wq and wa are the design parameters. The values taken in this study for wq and wa are 3Hz and 1Hz respectively.

The inversion control laws are given by



Dynamic Inversion With Pi Control

As the DI controllers lack robustness, a PI compensator is employed for achieving robustness for the autopilot. In this
control scheme, the dynamic inversion is used along with PI compensator as the outer loop of dynamic inversion. The PI gains,
Kp and Ki are calculated to satisfy closed loop bandwidth of 1Hz and an overshoot of 5%. The block diagram of nonlinear
dynamic inversion with PI is shown in the Fig. 2.

Fast Dynamic Inversion With Pi Control

In this control scheme, fast dynamic inversion and PI control law is used and the controller structure is as shown in Fig. 3.


Since the introduction of the concept of fuzzy reasoning, its use in engineering disciplines has been widely studied.
The main attraction undoubtedly lies in the unique characteristics of fuzzy logic systems. They are capable of handling
complex, nonlinear, and sometimes mathematically intangible dynamic systems using simple solutions. Very often, fuzzy
systems may provide better performance than conventional non fuzzy approaches, with less development cost [9-11].
It is well known that fuzzy systems have the ability to make use of knowledge expressed in the form of linguistic rules
without completely resorting to the precise plant models. In control applications, fuzzy logic approaches using if-then rules
can solve complex and practical problems. In recent years, researchers have also attempted to apply it on missile guidance
designs [12-13].
However, obtaining an optimal set of fuzzy membership functions and rules is not an easy task. It requires time,
experience, and skills of the operator for the tedious fuzzy tuning exercise. In present work the fuzzy logic is employed for
robustifying the DI controller. The functional block diagram using FLC as compensator is shown in Fig. 4. The procedure
followed in this design is discussed in the subsequent sections.



Selecting input/output control variables:

In this work, fuzzy controller is used instead of the PI controller to improve performance while achieveing
robustness. The input variables are
Error = error between alpha command and actual alpha = ac - aa (t )
ac= Alpha command, aa (t ) = actual angle of attack
CError = change in error of angle of attack = Error (t)-Error (t-1)

Output variable: The output of fuzzy controller is the command pitch rate qc

Defining Universe of Discourse (UOD):

The generalized universe of discourse [-1 1] is selected for both input and output variables with proper scaling Variable UOD
Error [-1 1]; CError [1 -1]; Pitch rate [-1 1]

Scaling (I/O):
With control the maximum Error and CError are [-40 40] with attacking angle (á) varies from +20 to -20 So the
scaling factor for input variables is 1/40 Scaling factor for output pitch rate 5

Selecting membership functions and fuzzy sets:

The triangular and trapezoidal memberships are widely used because of their simplicity in representing in
programming and ease of tuning. The triangular membership can simply be represented with three parameters a, b, c and
mathematically it is represented with linear straight line equation. a, b base of triangle, c centre of triangle. The member ships
are shown in Fig. 5 and Fig. 6. And Linguistic are given in Table.1 and Table.2.

Table.1. Error, Change in Error

Linguistic table
Table.2. Pitch rate linguistic table



Fuzzification, Defuzzification
In fuzzification, And method min, Or method Max, Implication Min, Aggregation Max is used while In
Defuzzification the centroid method is employed. The rule table with 25 rules is shown in Table. 3. The inner view of fuzzy
logic controller is shown in Fig. 7.

Table.3. Fuzzy logic rule table


Simulations are carried out for the three controllers namely (1) PI+DI controller (2) PI+FDI controller and (3)
Fuzzy+FDI controller. First simulations are carried out without any unce- rtainty in the system parameters and the results are
presented in Figs. 8-11. Next simulations are carried out by considering -20% uncertainties in the aerodynamic parameters and
the results are given Figs. 12-15. Lastly simulations are carried out by considering +20% uncertainties in the aerodynamic
parameters and the results are given Figs. 16 -19. The results consist of the time histories of angle of attack, control surface
deflection, pitch rate command and pitch rate response.

A ttackin g ang le p lot w ith o ut u n certainity D elta w ith o u t u n certain ity

30 150
A lp ha-c DI
20 P I+ D I P I+ D I
P I+ FD I P I+ F D I
10 fuz z y + fdi
d e lt a ( d e g )
alp h a(d eg )

s t eady erro r= FU Z ZY +F D I FU ZZY +FD I

3. 75% 50
ts = 0.41s ec
-5 0

-30 -10 0

-40 -15 0
0 1 2 3 4 0 1 2 3 4
p it c h r a t e c o m m a n d ( d e g / s e c )

tim e(s) tim e(sec )

Fig.8. á (deg) vs time Fig.9. (deg) vs time
p it c h r a t e ( d e g /s e c )

p itc h ra te c o m m a n d w ith o u t u n c e rta in ity p itc h ra te w ith o u t u n c e rta in ity

P I+ D I 100 P I+ D I
500 P I+ F D I P I+ F D I
Fuzzy+FD I 50 Fuzzy+FD I
0 0

-5 0
-5 0 0
-1 0 0
-1 0 0 0
-1 5 0

-1 5 0 0 -2 0 0
0 1 2 3 4 0 1 2 3 4
tim e (s e c ) tim e (s e c )
Fig.10. qc (deg/sec) vs time Fig.11. q (deg/sec) vs time



A lp h a w ith -20 % u n c ertain ity d e lta w ith -2 0 % u n c e rta in ity

40 200
a t t a c k in g a n g le ( d e g )

com m anded DI
D I+ P I 150 P I+ D I
F D I+ P I
P I+ F D I
Fuzzy+FD I

d e lt a ( d e g )
F u z z y+ F D I 100


0 0

-5 0

-2 0 -1 0 0

-1 5 0

-4 0 -2 0 0
0 1 2 3 4 0 1 2 3 4
tim e (se c ) tim e (s e c )
Fig.12. á (deg) vs time Fig.13. (deg) vs time
p itc h ra t e c o m m a n d ( d e g /s e c )

p i tc h ra t e c o m m a n d w i t h -2 0 % u n c e rt a i n i ty p itc h ra te w ith -2 0 % u n c e rta in ity

P it c h r a t e ( d e g /s e c )
DI 150
P I+ D I P I+ D I
100 P I+ F D I
P I+ F D I
Fuzzy+FD I
F u z z y+ F D I 50

-5 0 0 -5 0

-1 0 0
-1 0 0 0
-1 5 0

-1 5 0 0 -2 0 0
0 1 2 3 4 0 1 2 3 4
ti m e ( s e c ) tim e (s e c )

Fig.14. qc (deg/sec) vs time Fig.15. q (deg/sec) vs time

A lp h a w ith + 2 0 % u n ce rta in ity d e lta w ith + 2 0 % u n c e rta in ity

30 150
a t t a c k in g a n g le ( d e g )

com m anded DI
D I+ P I P I+ D I
P I+ F D I
d e lt a ( d e g )

F D I+ P I 100
F u z z y+ F D I F uzzy+FD I


-1 0 0
-2 0
-5 0
-3 0

-4 0 -1 0 0
0 1 2 3 4 0 1 2 3 4
t im e ( s e c ) tim e (s e c )
Fig. 16. á(deg) vs time(sec) Fig. 17. (deg) vs time(sec)

p i t c h ra t e c o m m a n d w i t h + 2 0 % u n c e r t a i n i t y p itc h ra te w ith + 2 0 % u n ce rta in ity

p it c h r a t e c o m m a n d

p it c h r a t e ( d e g /s e c )

1000 150
P I+D I 100 P I+ D I
500 P I+F D I P I+ F D I
F u zzy +F D I 50 F u zzy+F D I
0 0

-5 0 0 -5 0

-1 0 0
-1 0 0 0
-1 5 0

-1 5 0 0 -2 0 0
0 1 2 3 4 0 1 2 3 4
tim e(se c) tim e (s e c )

Fig. 18. qc (deg/sec) vs time(sec) Fig. 19. Q (deg/sec) vs time(sec)



Performance Comparison
From simulations, comparison is carried out of all the controllers and the results are summarized as shown in Table 4..
From the table it can be observed that the Fuzzy+DI controller offers the least settling time for all the cases while giving
satisfactory steady state accuracy as well as peak overshoot.
Case I: Performance without uncertainty
(Alpha=20, step time=2, Kp=6.85, Ki=21.64)
Control Steady Settling Peak
Scheme state time(ts) overshoot
error(ess) % ( normalized) (Mp) %
DI 30 1.88 -
PI+DI 0 6.77 40
PI+DI 0 7.3 -
Fuzzy+DI 2 1 -

Case II: Performance with + 20 % uncertainty

Alpha=20, step time=2, Kp=6.85, Ki=21.64
Control Steady Settling Peak
Scheme state time(ts) overshoot
error(ess)% (normalized) (Mp)%
DI 33.75 0.8247 -
PI+DI 0 4.06 42
PI+DI 0 3.97 -
Fuzzy+DI 4 1 -
Case III: Performance with -20 % uncertainty
Alpha=20, step time=2, Kp=6.85, Ki=21.64

Control Steady Settling Peak

Scheme state time(ts) overshoot
error(ess)% (normalized) (Mp)%
DI 27 1.875 -
PI+DI 0 8.75 45
PI+DI 0 6.25 -
Fuzzy+DI 6 1 -

Table 4. : performance comparison of controllers

A Fuzzy Logic based compensator is proposed to robustify the DI controller for missile autopilot design. The
performance of the proposed controller is compared with some existing controllers appeared in literature and it is shown that
the proposed controller offers performance benefit in comparison with the considered controllers.




[1] Garnell, P., Guided Weapon Control Systems, Pergamon press, Oxford, England, 1980
[2] Horton, M. P., Autopilots for Tactical Missiles: An Overview, Proc. Instn Mech. Engrs, Journal of Systems and
Control Engineering, Vol. 209, 1995, pp. 127-139.
[3] Nichols, R. A., Reichert, R. T., and Rugh, W. J., Gain Scheduling for H-Infinity Controllers: A Flight Control
Example, IEEE Transactions on Control Systems Technology, Vol. 1, No. 2, June 1993, pp. 69-78.
[4] Cho, S., Kim, S.-H., and Choe, D.-G., Robust Missile Autopilot Design using Dynamic Inversion and PI Control,
Proc. of the International Control Conference (ICC2006), Glasgow, Scotland, UK, 30 Aug-01 Sep 2006.
[5] Slotine, J.-J. E. and Li, W., Applied Nonlinear Control, Prentice-Hall, Englewood Cliffs, New Jersey, 1991.
[6] Devaud, E., Siguerdidjane, H., and Font, S., Some Control Strategies for a High Angle of Attack Missile
Autopilot, Control Engineering Practice, Vol. 8, 2000, pp. 885-892.
[7] Hull, R. A., Schumacher, D., and Qu, Z., Design and Evaluation of Robust Nonlinear Missile Autopilots from a
Performance Perspective, Proc. of the American Control Conference, Seattle, Washington, June 1995, pp.
[8] Chen, W.-H., Nonlinear Disturbance Observer-Enhanced Dynamic Inversion Control of Missilee, Journal of
Guidance, Control, and Dynamics, Vol. 26, No. 1, Jan-Feb 2003, pp. 161--166.
[9] Cox, E., Fuzzy Fundamentals, IEEE Spectrum, Vol. 29, No.10, 1992, pp.58-61
[10] Albertos,P.,Martinez, M.,Navarro, J. L., and Morant, F., Fuzzy Controllers Design: A Methodology, IEEE
Conference on Control Applications, 1993, pp. 69-75
[11] Chia-Nan Ko, Tsong-Li Lee, Yu-Yi Fu, Chia-Ju Wu, Simultaneous Auto-Tuning of Membership Functions and
Fuzzy Control Rules Using Genetic Algorithms, IEEE International Conference on Systems, Man and
Cybernetics, 8-11 Oct. 2006, pp.1102-1107
[12] Mishra, S. K., Sarma, I. G., and Swamy, K. N., Performance Evaluation of Two Fuzzy-Logic-Based Homing
Guidance Schemes Journal of Guidance, Control, and Dynamics , Vol. 17, No. 6, 1994, pp. 1389--1391.
[13] Shieh, C. S., Nonlinear Rule Based Controller for Missile Terminal Guidance, IEE Proc. on Control Theory
Applications, Vol. 150, No. 1, 2003, pp. 45--48.





N Nayak and T. K.Biswal

Proof and Experimental Establishment,
Chandipur , Balasore – 756 025
High speed imaging system covering three major types of instruments like 16mm film based camera (NAC-HD),
still video recorder ( SVRII) and high speed video camera ( Speed Cam Visario) have been used to record different
performance parameters like sabot separation of fin stabilized armor piercing discarding sabot (FSAPDS) ammunition, spin
of rockets and sequence of functioning of high explosive squash head (HESH) shell when impacted on armor . The system
configuration and the analysis of image output to derive ballistic and performance parameters have been presented. In view of
the fast events and short duration of exposure, appropriate choice of imaging system has been crucial to record the desired
parameter. Trajectory of long range projectile and rockets, functioning of fuzes and ejection of child stores in case of
remotely delivered minelets (RDMS) and cargo ammunition have been studied by using tracking doppler radar . Different
plot formats viz velocity-time, velocity –distance, acceleration-distance, range-azimuth, horizontal range-vertcal
components have been used to derive desired information. Ideal plot format to study specific parameters like all burnt point
(ABP) of rockets, time and height of fuze functioning and malfunctioning in trajectory has been identified. Combination of
imaging and radar based measurement to support failure investigation by recording fast event of fuze premature has been

Ballistics and performance parameters obtained from dynamic firing evaluation of armaments are gainfully utilized
by the designer , producer and quality control personnel for design validation and optimization during the developmental
stage and also for assessing its performance for user acceptance1. Range instrumentation plays a vital role in recording the
design and performance parameters involving internal , external and terminal ballistics of any ammunition. Parameters like
pressure , muzzle velocity, range , deviation , height of fuse functioning , integrity of projectiles on exit from muzzle are some
of the routine dynamic evaluation requirements for conventional Armaments2. With enhanced emphasis on large caliber, long
range, pin pointed accuracy and force multiplier ammunition , there has been a greater emphasis on recording total
trajectory, separation of child store in case of cargo ammunition , base bleed functioning and spin of rockets among others1.
Over and above, the recording of additional parameters to understand fast and transient phenomena associated with dynamic
firings evaluation and related failures/malfunctioning provides additional advantage to designers and evaluating agencies3.
The recording of ballistic and performance parameters are crucial but its analysis to deduce the desired parameter
vis-a-vis correlation with the actual end results is more challenging. Ideal range layout coupled with selection of
appropriate instrumentation is therefore key to the successful performance evaluation of any Armament store.
A wide variety of parameters are required to be determined for assessing performance of a broad spectrum of
conventional Armaments and therefore, no single set up of instrument or its category can record all the desired parameters .
Moreover all ballistic phenomena occur at high speeds and great number of complex interaction occur. The major range
instrumentation essential for dynamic test and evaluation can therefore be broadly grouped into three categories viz. optical,
radar and sensor based measurements. High speed imaging and doppler radar based measurements covering select class of
ammunition which provide information on fast and transient phenomena along with the governing principle are presented in
this paper. Importance of combinations of both optical and radar based measurement for typical case for detection of
malfunctioning has also been highlighted.


High speed photography is used directly in all branches of ballistics. Few of the techniques which have been well
established and put to practice in the author's laboratory are given in Table1. High speed photography plays a vital role for



evaluation of ammunitions parameters like projectile stability, integrity , impact studies , armor defeat mechanism , sabot
discarding, spin measurement and defect analysis . To study these phenomena it is necessary to use specialized measuring

Table1: Parameters measured by imaging technique

Type of Armament Parameters

Sabotted projectile Sabot separation

(APDS/FSAPDS) Penetrator status & velocity
Rockets Spin
Fin opening
Thrust & acceleration
Dip angle
Base Bleed projectile Ignition of base bleed mortor
Ordnance Recoil &Run out
HESH Scabbing effect
Pyro technique Ammunition Smoke emission, Ignition time
spread of Red Phosphorous pellets
Fuzes Arming Delay
Height respect to target
Armor Penetration & perforation
Impact & residual projectile velocity

In general all events in ballistics take place quickly and in order to obtain clear pictures exposure are usually very
short. As object movement become faster, exposure must decrease and illumination must increase or image must be artificially
enhanced electronically. Therefore, the realm of high speed photography is generally determined by the exposure times used
or the picture repetition rate. Knowledge of the duration of actual exposure or shutter speed is important in HSP to ensure
clear and desired picture . Three types of instruments like still video recorder ( SVRII), film based camera (16mm NAC-HD),
and high speed video camera (Speed Cam Visario) to record ballistic phenomena like sabot separation of FSAPDS
ammunition, sequence of functioning of HESH shell and spin and dip angle of free flying rocket are presented below. Specific
instruments have been chosen based on the phenomena to be recorded to derive desired performance parameter.

Sabot discarding & study of penetrator in FSAPDS type ammunition by SVRII

The integrity of penetrator , separation of three sabot petals, interaction of sabot with penetrator as well as the status
of tail unit are some of the parameters required to be ascertained in intermediate ballistic zone after exit of projectile from
muzzle. Separation of tail unit or breaking /bending of penetrator due to very high chamber pressure, non uniform sabot
discarding are among the defects which can be inferred from the high speed photographs instantly. The film based camera can
be more handy to study sequence of sabot separation at a very high frame rate. But inherent disadvantage like very small
picture size and development of film continue to persists. On the other hand, HSV camera with 10000 fps can also provide a
sequential pictures covering 2-4m of travel in flight. However to record specific information like status of penetrator and its
velocity at the instant of sabot separation, SVR can be the ideal tool. In this method the image (fast event of penetrator in flight)
can be captured in single or multiple exposures ( in a single frame ) from a minimum of 200ns to 1ms . A schematic



diagram of capturing the motion of an FSAPDS ammunition is shown in Fig. 1a and that of the out put of obtained
photograph is given in Fig. 1b.

a b
Fig. 1.a) schematic layout of SVR II b) Sabot discarding of FSAPDS Ammunition

.The status of penetrator along with tail unit is clearly seen in Fig 1b. The velocity of the penetrator can be accurately
determined by taking pictures in multiple exposure (instead of single one ) in a single frame with appropriate time mark.

Study of functioning of HESH shell by film based high speed camera

High explosive squash head shells functions with the principle that the nose portion squashes on impact on target,
followed by padding of inert material and pasting of high explosives. The explosive is detonated by the initiation of a base fuze
with select delay to ensure proper pasting of explosives. The detonation wave sets into the target results in scab formation from
rear side. The sequence of impact of HESH on to target and initiation of detonation has been studied by a film based high
speed framing camera (16mm NAC-HD). This is a film based mechanical rotating prism camera with a speed of 32000
pps . A schematic diagram for recording functioning of HESH shell on impact on armor and the recorded photographs for
four rounds are shown is Fig. 2 a and 2b respectively..
Two distinct pattern of functioning i.e one type for Fig.2b & d and different for 2c &e is clearly visible. A delay time
of the order of 250 µs for the former and 170 µs for later type was obtained from frame to frame analysis. It was of interest to

a b c d e
Fig. 2 a) schematic for deployment of HSP, b-e) HSP of functioning of a HESH shell



observe clear scabbing in shells representing the sequence as in b & d where as non scabbing was observed in c& e where
delay was 170µs. Criticality of delay time of detonation after impact vis-a-vis scab formation was inferred from this
recording. This technique has proved to be an ideal tool for defect investigation of such fast phenomena.

Spin of Rockets and projectile by High Speed Video camera (Speed cam visario)
The spinning of rockets and its fin opening in case of wrap around ones on exit from launcher are among the crucial
launch parameters sought after by the designers to correlate ultimate range and accuracy. Due to comparatively lower launch
velocity and high angle of fire, high speed video camera is the preferred one for recording such parameter. Speed cam Visario
is a color high speed digital camera with CMOS sensor capable of recording a dynamic event at 1000 fps for 4s at its
maximum resolution 1536 X 1024 .In windowing mode it is capable of recording at a maximum of 10,000 fps at 512 X 196
resolution The flight of 122mm rocket captured by the camera is shown in Fig. 3. The Spinning of the rockets can be
measured on exist from launcher. The external surface of warhead portion is suitably paint marked with appropriate
dimension and contrast as reference for spin measurement.


Muzzle velocity (MV) of any projectile fired from
tank or gun is one of the crucial parameter to correlate internal
ballistic parameters and predict the range by considering
ballistic co-efficient and other meteorological considerations
. Due to the introduction of the large caliber projectiles,
rockets, base bleed type ammunition with enhanced range,
the requirement of accurate measurement of impact range,
deviations from predicted line and height of functioning is on
the rise. Traditional technique of deploying survey
instruments has not been always been met with success.
Hence radar based measurements gaining more importance
where both MV and range capabilities of projectiles can be
Fig.3 Rocket in flight for spin measurement
simultaneously measured.

Table2: Parameters measured by Doppler Radar

Type of Armament Parameters

Sabotted ammunition Muzzle velocity
(FSAPDS) Velocity - time &velocity-distance
Rockets ABP Time, Velocity & distance
Carrier Projectile Range & height of functioning
(Cargo/RDMS) Time of fuze functioning
Ejection of child store
Pyro techniques Separation in flight
Ejection of pay load
HE Projectile Spin & spin decay
Drag coefficient



Doppler Radar is one of the most effective and useful tool in artillery range instrumentation4 . Though it works in the
basic principle of doppler shift, the doppler radar technology has gone through improvements over the years 4-5 . CW
Doppler with mono pulse receiver can accurately measure velocity , azimuth and elevation of a moving target. But practical
use of such instruments need to have an in depth knowledge of their working principle, projectile design parameters and
correct interpretation of signal obtained with appropriate format.
Doppler radar has been extensively used in the authors laboratory to measure the external ballistics of a series of
ammunition viz. high explosive shell cargo ammunition, rockets, and base bleed projectile (Table 2). The data generated has
been validated by physical range observations in some cases. Though the details of working principle is available in
literature 5 and that of the commercially compact radar set up , the generation of desired data in appropriate format and its
analysis needs special attention . These aspects have been presented in subsequent paragraphs.

Carrier Projectiles

Remotely Delivered Minelets System ( RD M S )

Evaluation of Rockets with submunition warhead require a set of parameters to be measured simultaneously .
The rocket parameters like all burnt point, (ABP) all burnt time, all burnt velocity, range , height and time of fuze
functioning and separation of child store are measured simultaneously. Visual detection and recording of height and range
of fuze functioning, separation of minelets by normal observer becomes difficult due to little smoke/flash produced by fuze
functioning at a height of 300-500 m height and at a distance of 1.5 km from observer. Even if the fuze functioning is located,
exact range and dispersion of minelets and the empty warhead casing are not possible to locate. Under such circumstances the
tracking radar is quite useful to generate all the above parameters in addition to complete trajectory.


c d

Fig. 4. a) Schematic of RDMS functioning, b) Trajectory, c) velocity-time plot, d) range vs azimuth



A schematic of RDMS functioning is shown in Fig. 4a. Its trajectory (X-Y plot) and the variation in azimuth in flight
(X-Z plot) is shown if Fig.4b &d. A functioning height of 500m above ground at a range of 96000m is obtained from the plot in
fig.4b. The rocket functions with an azimuth which deviates 300m from the line of launching (fig.4d). Further, crucial
information can be deduced from the velocity-time plot (Fig.4c). The discontinuity of V-T plot and multiple radar signals at 30
sec of flight results due to fuze functioning which matches with the fuze set time. Separation of warhead from the rocket and
subsequent ejection of minelets are represented as distinct and simultaneous lines arising from Doppler shift. These
observations have been corroborated by the simultaneous recordings from electro optical tracking system (EOTS).

Cargo Ammunition
Cargo ammunition with 24-56 bomblets is required function at a predefined height by setting of electronic time
(ET) fuze. During its evaluation, it is not always not possible to observe the functioning aspects visually due to safety
constraints. Further the separation of bomblets and its dispersion zone is also desired to assess the functioning and recovery of
components during design trials.. Tracking radar becomes handy to record the entire sequence of functioning including the
range, time of fuze functioning and at the same time the ejection of bomblets. A sectionized view of typical cargo shell with
bomblet and fuze is shown in fig.5a. Experimental data recorded during dynamic firing evaluation of this shell by tracking
radar DR-6700 is shown in fig.5b-d. The complete trajectory covering a range of 8000m is seen in X-Y distance plot format
(Fig. 5b).


c d

Fig. 5 a) schematic of an cargo shell, b) total trajectory (X-Y), c) velocity - time d) velocity - distance plot

A break in trajectory corresponding to a range of 6500 in descending trajectory represents the fuze functioning and
ejection of bomblets. Velocity-time plot (Fig. 5c) indicates the presence of multiple objects after 17 sec of flight which
corresponds to the actual fuze set time. .The upper curve after the break corresponding to this time represents the movement of
empty body and lines below indicate movement of child stores. The range corresponding to this event can be obtained from a
velocity-distance plot format (Fig.5d) An analysis of the above three plot format in combination with change in azimuth in
flight can provide information on the range and ground location of dispersion of bomblets as well as the mother container and
can be the guiding factor for their recovery.




The individual application of imaging and radar based measurements are presented in above paragraphs. At times,
these measurements supplement each other to derive more meaningful and conclusive results. Detection of flight premature
(fuze mal functioning) of a 155mm ERFB-BB projectile is a case in point for illustration.
The projectile launched for functioning at a range of 20km did not complete the trajectory even though the ignition
of base bleed and forward movement after muzzle exit could be seen from on line video camera. The velocity-time out put of
doppler radar showed multiple doppler signals at about 300m ahead of muzzle and also subsequent flight up to the range of
8000m. The signals at 300m was indicative of some separation of portions from the projectile and continuation of subsequent
flight (Fig.6a). It was rather difficult to draw any conclusion from such data. A careful study of CCD camera pictures of the
flight recorded in different look angles showed the functioning of only fuze at a distance of 300m and subsequent flight of
projectile with base bleed functioning (Fig.6b). A sharp drop in velocity (when compared with normal projectile at
corresponding distance) after the fuze functioning was the direct effect of induced drag due to change in ogive shape of shell.
Thus the fuze premature in flight and non initiation of booster to ignite main charge filling of shell could be conclusively
established which is a unique recorded observation in experimental ballistics.

Fig.6 a) Velocity- distance plot b) still video out put of ERFB-BB shell in flight

Range instrumentation is the backbone of any dynamic test activity . Ideal range layout coupled with integrated
range instruments is vital for study of fast events associated with dynamic firings. Interpretation of recorded data has been
found to be the key to identify failure modes and also for defect investigation. Appropriate choice of recoding system both
imaging and Doppler based ones can provide vital clue to the designers and measure performance parameters ; few of which
have been highlighted in this article.

The authors express their gratitude to Maj Gen Anoop Malhotra, Director, PXE, Chandipur for his encouragement
and permission to publish this article.



1. N Nayak & G C Das “Dynamic test & Evaluation of Armaments: Recent trends “ Proceedings of Seminar on Test &
Evaluation of Ammunition organized by Deptt of Defence Production, DGQA, May 2005

2. C L Farar & D W Leening 'Military Ballistics: a basic manual, first edition, Brassys Publishers limited, Oxford, 1983

3. N Nayak & G C Das “Identifying Failure Modes and Defect Investigation by Dynamic Test & Evaluation: Case Studies
“ Proceedings of Seminar on Test & Evaluation of Ammunition organized by Deptt of Defence Production, DGQA,
May 2005

4. G C Das, N Nayak & A Bose “ Doppler radar Based Measurements for Evaluation of Conventional armaments”
Proceedings of National conference on Range technology, NACORT-2006, ITR, Chandipur.

5. Merril I Skolnik, Introduction to Radar systems, second edition, McGraw-Hill Book company, Singapore, 1981





T K Biswal & N Nayak

Proof & Experimental Establishment, Chandipur, Balasore-756 025
Three different types of test and evaluation methodologies pertaining to performance of proximity fuze against an
aerial target along with its miss distance studies in case of blind, estimation of rapid rate of fire of AK630 weapon and attitude
of a projectile before its impact on armor with adoption of innovative range techniques have been reported in this paper. High-
speed imaging as well conventional videography have been used to record the performance parameters. Certain features in
digital imaging like high resolution capabilities in space and time domains, synchronous imaging, concept of circular buffer
memory in the camera head leading to pre- and post-recording capabilities with various types of software/hardware triggers
and almost of real time display have been exploited in order to solve emerging test and evaluation requirements. In the absence
of standard testing techniques, the methodologies adopted have ultimately helped the users for acceptance of the armament

Performance evaluation of any armament store by dynamic firing is the ultimate acceptance criteria before its issue to
users. At the same time dynamic firing data is crucial for design validation and optimization during the development stage.
Laboratory testing and simulation studies, many a times, do not represent the true performance as in actual condition. It is also
difficult to simulate the actual condition in the laboratory due to the magnitude of firing stress and environmental effect.
Hence, dynamic test and evaluation is of paramount importance both from the designers and users point of view. Conventional
armaments cover a broad spectrum of ammunitions, armors and weapons and its test and evaluation is system specific and no
open literature is available. In addition, evaluation methods adopted also vary depending on the design philosophy, intended
use and facilities available. Test range facility is one of the guiding factors for developing the evaluation techniques.
In recent past the transition of film-based high-speed cameras to digital high-speed imaging devices has got a
tremendous impact in the complete domain of high-speed imaging and its applications in the field of ballistics. Certain features
in digital imaging like
high resolution capabilities in both space and time domains, synchronous imaging, concept of circular buffer
memory in the camera head leading to pre- and post- recording capabilities depending on various types of software/hardware
triggers, almost real time display etc are add on and very much useful to the experimentalists to devise various techniques in
order to solve emerging test and evaluation requirements.
However, the physical problem needs to be understood thoroughly to devise a correct solution within the limitations
of the imaging devices. Few such problems of special significance both from the designers and users point of view will be
presented here. To mention some of the techniques evolved during recent years, this paper primarily focuses three different
aspects of test and evaluation methodologies pertaining to performance of proximity fuze within a close vicinity of an aerial
target along with its miss distance studies, estimation of rapid rate of fire of AK630 weapon and attitude of a projectile before
its impact on armor. In all the techniques, high-speed imaging as well conventional videography play a crucial role which have
been brought out in greater details in the following paragraphs in the form of case studies.

Recent developments in conventional armaments focus more on pin-pointed accuracy, rate of fire and reliability of
the product for absolute performance. Appropriate technique to evaluate the product becomes a challenge to range
technologists due to a variety type of stores and its intended use. Since no single technique or method can meet the evaluation
requirement, there is a need to develop methodologies for deriving performance parameters from case to case basis. Few of the
methodologies are presented in subsequent paragraphs. Efforts have been made to outline the principles involved,



experimental techniques adopted and determination of the final parameters. This has been illustrated with sketches and
photographs for ease of understanding of the readers.

Co-ordinates of functioning of Proximity Fuze against low flying target & recording of miss distance of blinds
Measurement of miss distance is a vital parameter for the fuze designer to know the performance of
ammunitions/warhead in the close vicinity of enemy target. Spatial measurement accuracy of this parameter is very much
important to judge ammunitions lethality and its defeating capabilities. In the evaluation process of fuze ammunitions, it is
mandatory to know about miss distance of the fuze-based projectile when it fails to function with respect to some aerial target.
Estimation of proximity fuze performance in the close vicinity of the target is also crucial for its acceptance. Here the fuze is
associated with a 40mm projectile traveling at 1000m/s and passing over an aerial target located at 1000m along the line-of-fire
and designed to function within 3m miss distance zone with respect to the target. One standard exocet dummy missile target
with photograph of fuze functioning and the definition of functioning zone (A&B) is shown in Fig1.(a) and (b) respectively.
Non-functioning round passing in zone A is called a blind and identified as a defect from user point of view. It is therefore
crucial to measure the miss distance in case of its non-functioning.
In real scenario where the target is dynamic in nature, measurement of miss distance is based on microwave and
image based devices. In case of microwave devices, miss distance indicator (MDI) is incorporated with one of the two flying
bodies and it is based on Doppler principle. At closest approach Doppler signature (frequency) is zeroed giving rise to a dc
signal from which MDI is calculated based on prior information (calibration). In imaging both the target and the ammunition
trajectories are calculated from which the closest approach between the two flying bodies is estimated. In both cases real time
data acquisition is possible.


Fig.1 a) Exocet dummy missile target & fuze functioning, b) defined accepted zone of functioning



However, in this paper conventional imaging and high-speed imaging have been used to obtain both point of
functioning as well miss distance1.

The missile target (MT) is erected at 5m heights with respect to seabed and it simulates a low flying missile. It is an
extended target of cylindrical shape with 0.2m radius, 5.0m length and made of aluminum sheet. Proper distance is maintained
between the target and gun by use of distometer/laser range finder and the line of fire is chosen considering the range safety in
to account. MT is erected in a horizontal way such that its body axis and gun axis lie in a vertical plane. DGPS is used to
measure the altitude of target location and that of camera location so that camera can be adjusted to match its axis to that of MT
and both axes lie in one horizontal straight line. Measurement accuracy down to 0.01m is achieved in this DGPS. It helps to
know the exact separation between camera and target and also to ascertain both are at same height.
Camera with telephoto lens is positioned on a high-precision tripod to meet the above requirement and horizontality
of camera axis is ascertained through a digital clinometer. High-speed digital CMOS camera, Speed Cam Visario, is used
where a maximum resolution of 1536x1024 pixel is achieved at 1000pps.Same camera is also used in windowing mode at
1024X768 pixel resolution at 2000pps. Image acquisition at these rates is decided based on ambient light condition. Higher the
frame rate better is the time resolution that results in less blur and better measurement accuracy. Sensor has got square pixel of
size 11µm x11µm.Image acquisition is performed through an external trigger, a photo-trigger, that picks up muzzle flash of the
gun during firing and initiates image acquisition. This trigger time is considered as reference time t0 at which projectile leaves
the muzzle. Camera records the event in the post-trigger mode to facilitate determining time to burst of fuze functioning over
MT which is obtained through off-line video analysis. In a series of functioning rounds, burst time is determined for each round
and then mean burst time is calculated. This is probable because in a firing series, parameters like charge mass, charge
temperature, gun elevation etc are maintained constant through out. Minimum distance of the projectile with respect to MT
needs to be calculated during flight for miss distance calculation. Ideally in 40mm proximity ammunitions, shell is designed to
function within 3m radius with respect to MT body axis. So a field-of-view (FOV) of more than 3.0m radius around MT
longitudinal axis is chosen using suitable optics.

Experimental Set-up
The experimental lay out is shown in Fig.2..The in-line high-speed digital camera C1 and MT are separated by
1000m and a telephoto lens of 800mm focal length is used which attains a spatial resolution of 13mm/pixel in the object plane.
It ensures a 3x3 pixels image point of 40mm proximity fuze at the target plane that can be well resolved. Camera axis is
matched to that of AMT adjusting the height difference measured through DGPS and zeroing of camera is ascertained using a
digital clinometer. Since imaging is done in the longitudinal direction, the extent of blur caused due to projectile motion, both
translation and rotation (roll, pitch & yaw), during exposure time gives rise to motion blur which is considered for
measurement accuracy of miss distance down to single pixel. Exposure time down to 100 microseconds is set to reduce the
extent of blur.

C1 Gun MT
C1– in-line high-speed digital camera;
C2 –range camera;
PC – camera control computer;
T- optical trigger; T
MT- missile target
To ascertain that shell is functioning within the miss PC

distance zone of the target, one conventional camera C2,

called as range camera, is placed at right angle to MT and also Fig.2 Schematic set up of experimental lay out
maintaining co-planarity to have equal height with respect to
MT. In case of functioning, the co-ordinates of point of
functioning is determined from both C1 and C2 data which
read two orthogonal planes of functioning.



Initially two fuze shells are fired to function over MT that is confirmed by range camera C2.Orthogonal views of
inline and range camera are shown in Fig. 3.
Burst time is calculated for both shells and mean time to burst is found out (typically in the order of 976ms). The
functioning of the two rounds is affirmed well within three meter radius with respect to MT. Subsequently two inert shells are
fired which simulated non-functioning ones. These shells are also captured in flight. Separation between MT center and the
shell at 976ms from reference time is calculated from offline image analysis and found out to be 0.371m as shown in Fig.4.
Miss distance is calculated down to single pixel accuracy.

(a) (b)
Fig. 3 Functioning of fuze (a) Orthogonal view (b) Perpendicular view

Matching axes of both high-speed camera and MT is bit difficult if one thinks of in terms of pixel accuracy. This can
be done making use of PC-based pan & tilt gimbal as a platform for camera. Due care has to be taken to match image center of
camera field-of-view (FOV) to that of MT. In this process it is assumed that camera optic axis is passing through sensor image
center, which is lying at the center of the image plane. However, there is an offset both in x- and y- directions in the image plane,
which needs to be determined very accurately by any of the methods as suggested by R K Lenz & R Y Tsai 2 to have better
measurement accuracy.

Fig. 4 Measurement of miss distance with respect to MT center

Determination of Rate of Fire of AK630 Naval Weapon

AK-630 weapon of Russian origin has six barrels of 30mm caliber and determination of its high rate of fire is crucial
for its acceptance. Rate of fire of the order 4500 rounds/min has been determined by an innovative layout by firing a salvo of 30
rounds on to a witness target and simultaneously recording target hit points by a conventional video camera. The cumulative
impact on the target is calculated over time through frame by frame analysis leading to estimation of rate of fire. The same has
also been attempted by acoustics method and the results are in very close agreement and reproducible.



One witnessing target; a ply board of size 2.5mX2.5m, is erected at 100m from muzzle in the transverse plane with
respect to the line of fire. Conventional video camera is placed adjacent to the gun to image the witnessing target and to find out
the number of shots crossing the very target plane over time from their signatures i.e. impacts on target. Measurement of
cumulative impact points over time (interframe separation time 40ms) gives rise to the magnitude of rate of fire. In addition, it
also provides both consistency and accuracy parameters of the ammunitions and weapon as well.

Experimental Set up
Schematic set up for rate of fire studies of AK-630 weapon is shown in Fig.5. Here an interline transfer ½” CCD
camera is used with an object lens of focal length 180mm. It covers a field of view of 3.5mX2.66 m in the target plane. Sensor
has got a resolution of 640 x 480 pixels which corresponds to 5mm resolution in the target plane.


CCU-camera control unit,
T-witnessing target

Fig.5 Schematic set-up for recording Rate of Fire of weapon

Camera is powered form a camera control unit (CCU) which is connected to a PC for display and image analysis. Camera
is positioned on a platform to match with the line of fire and its image center coincides with the witnessing target center.
The main aim is to look at the cluster of impacts from a right angle position. Passive imaging is incorporated in the total
process and due care is taken to image impact signatures.
The impact points are well defined and accounted for its quanta unless otherwise there is a complete superimposition
of one above the other.


(a) (b)
Fig.6. a) Witnessing target showing impact points, b) graph showing impact points vs. time

Cluster of 30 munitions is fired in a salvo mode and their impacts are shown on the witnessing target in Fig.6.(a).Then
offline analysis is performed to obtain rate of fire which has been depicted in Fig.6b. Here cumulative impacts over time is
plotted which shows a linear curve .It means a uniform rate of fire is obtained in this weapon and from its slope rate of fire is
calculated. This is found out to be 4852 shots/min and has got a very close match with the one taken by normal acoustic



Determination of Attitude of Projectile before Impact on Target

The attitude of a projectile at the point of impact on armor affects the depth of penetration3. Variation of angle of
impact other than the selected one can lead to inconclusive penetration results. Methodology for recording the attitude of the
projectile before its impact on target has been developed by single shot capture of images of the projectile in two orthogonal
planes simultaneously. It provides projectile attitude both in vertical and horizontal planes leading to measurement of both yaw
and pitch resulting its determination of attitude at the very instance of space and time.

ºv º

Fig. 7 Vector diagram for defining angle of attack º of the projectile in terms of vertical and
horizontal angles º v and º h representing yaw and pitch respectively

Imaging the projectile in two orthogonal planes simultaneously, both horizontal and vertical, determines both pitch
and yaw angles from which the attitude angle (angle of attack) is determined as shown in Fig.7.
Let º be the angle of attitude which is defined in 3-D as shown above. It is resolved in two orthogonal planes resulting
º v and º h which are yaw and pitch angles respectively considering paraxial firing4 as practiced during armor testing. Here the
angle of attitude º is expressed as
º = tan-1 √(tan2 º v + tan 2 º h )
Further tan º v and tan º h are slopes of the projectile image body axis captured in the vertical and horizontal planes
Experimental Set up
The mirror set up is positioned along the line of fire such that longitudinal axis of the mirror and the line of fire lie in
the same vertical plane of camera FOV. Here the projectile is allowed to pass within a hollow cylindrical object containing the
very mirror at 45º with respect to the vertical plane. Both direct and mirror images are recorded simultaneously by a single
ICCD camera. In this case SVR II Ballistic Range Camera is used as shown in Fig.8.

Gun Mirror


Fig. 8 Schematic set up for projectile attitude studies: T- optical trigger (sky screen); S-Xenon strobe



Optical trigger, a sky screen, placed under line of fire and close to the mirror set up, is used as external trigger for
capture of shot image within the camera FOV after incorporating appropriate delay to the Xenon strobe, used as illuminating
source. The attitude of a 40mm AP shot before impact on armor at 50m has been generated and image is shown in Fig.9.

Fig.9 Orthogonal images of 40mm shot captured in SVR II:

Top-direct image (vertical plane) & bottom-mirror image (horizontal plane)

In this case º v and º h are found to be 1.1º and 2.81º respectively and º is found out as 3.01º. Three shots were fired
and in all the cases º was found to be less than 5º.This technique can be used for different types of AP shots. During set up
perpendicularity of the camera to the line of fire as well to the longitudinal axis of the mirror has to be done very carefully. To
achieve this, simple pendulum can be used as a test image where center of circles of both the direct and mirror images should
lie in one vertical line.

Technique for estimation of miss distance measurement of proximity fuze FB-40 for 40mm ammunitions has been
developed and implemented. However, present technique is based on passive imaging which can be further improved upon by
the use of either active imaging or imaging in the infrared band adopting high-speed IR cameras. Latter will be more effective
overcoming environmental conditions like low visibility, foggy weather etc. Determination of rate of fire from multi barrel
weapon has been established. The results are in close agreement with acoustics method adopted simultaneously. Attitude of
projectile before impact is crucial for study of penetration phenomena and it will help in assessing actual depth of penetration.

The authors wish to express their sincere thanks to Maj Gen Anoop Malhotra , Director, Proof & Experimental
Establishment, Chandipur for his encouragement and permission to publish this paper.


1. T.K. Biswal et al. “ Miss distance studies of 40mm proximity fuze with respect to anti-missile target using high-speed
videography”, NACORT-2006, ITR (DRDO), Balasore.

2. Reimar K. Lenz & Roger Y. Tsai, “Techniques for Calibration of the Scale Factor and Image Center for High Accuracy 3-
D Machine Vision Metrology,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 10, no.5, pp713-
720, 1998.

3. Manfred Held “Impact parameters of projectiles”, Propellants, Explosives, Pyrotechnics 24,211-216, 1999.

4. Donald E. Carlucci & Sidney S. Jacobson “Ballistics: Theory and Design of Guns and Ammunitions”, CRC Press, Taylor
& Francis Group, FL 33487-2742, 2008.




Subhas Saha, DR Kolhatkar

Armament Research & Development Establishment, Pune 411 021

Highly-Accelerated Life Testing (HALT) and Highly-Accelerated Stress Screening (HASS) testing are the next
generation techniques for Environmental testing. The aim is to cause failures that identify weak design elements. Failure-
prone elements are then re-designed and tests are continued at higher levels. Highly-Accelerated Life Testing (HALT) and
Highly-Accelerated Stress Screening (HASS) are starting to supplement traditional vibration and thermal testing to meet these
new quality targets.
Highly-Accelerated Life Testing (HALT) exposes the product to a step-by-step cycling in environmental variables
such as temperature, shock and vibration. HALT involves vibration testing in all three axes using a random mode of
frequencies.. The goal of HALT testing is to break the product. When the product fails, the weakest link is identified to improve
product quality. HALT testing must be performed during the design phase of a product to make sure the basic design is reliable.
Highly-Accelerated Stress Screening (HASS) is an reduced form of HALT testing. HASS testing is an on going
screening test, performed on regular production units. The idea is not to damage the product, but rather to verify that actual
production units continue to operate properly when subjected to the cycling of environmental variables used during the HASS
The purpose of testing, computer modeling, engineering analysis, probability studies, Failure Modes and Effects
Analysis, Fault tree Analysis and good old fashioned thinking is to generate and evaluate information so that decisions can be
made. Hence how important is 'information', can be gauged from Mr. George Herbert's little rhyme

All for want of a nail….

For want of a nail a shoe was lost
For want of a shoe' a horse was lost
For want of a horse a rider was lost
For want of a rider, a message was lost
For want of a message, a battle was lost
For want of a battle, a kingdom was lost
All for want of a nail.

Without the right piece of information at the right time, the battle and the kindom war were lost. The right information
is critical to making any plan successful.In our R& D work, do we get 100% information? No, So we are forced to have a timing
formula P= 40 to 70 in which P stands for probability of success and the number indicate the percentage of information
acquired. We should not act if we are having less than 40 % information and should not wait until 100% information.

What is Reliability?
Reliability is the best quantitative measure of the integrity of a part, component, product, or system. Reliability
engineering provides the theoretical and practical tools where by the probability and capability of parts, components,
equipment, product, subsystems, and system to perform their required functions without failure for desired periods in specified
environments, that is their desired optimized reliability, can be specified, predicted, designed in, tested, demonstrated,
packaged, transported, stored, installed and started up and their performance monitored and fed back to all concerned



Traditional Testing and Its Limitations

Traditional vibration and temperature testing has played an important role in the genesis of today's reliable and
sophisticated electronic and electro-mechanical products. The core philosophy of this testing method is to define a set of
specifications, usually minimum or maximum temperatures and/or vibration levels, and to conduct the tests by changing only
one variable at a time. Vibration testing is performed on one axis at a time. If the device is still functional after being tested
according to the test specs, it is considered to have "passed." A "passing" result is a positive outcome. However, thinking about
this further, it becomes clear that a "pass" result does not help identify the weakest link in the product. In other words, the
traditional test cannot help the designer or scientist make the product any more robust.
Furthermore, with the "one-at-a-time" change in environmental variables, and the one-dimension vibration testing,
the test specs are not similar to real-world operating environments. As a result, this kind of testing does not provide an accurate
indication of how the product might perform in the field. This critical look at traditional environmental testing is not intended
to be a blanket condemnation of that process. After all, this kind of testing has played a key role in the evolution of today's
highly reliable products. Instead, this examination of certain weaknesses in classical environmental testing can be helpful in
understanding how new testing methods, in particular HALT and HASS testing, can lead to even greater levels of product
quality and reliability.

Accelerated Life Testing

Life cycle testing is a test procedure applied to a small percentage of products, subjecting them to stresses similar to
those found in actual service conditions. Accelerated Life Testing (ALT) is a test activity during product development in which
prototypes are subjected to stress in terms of temperature and vibration at levels much higher than those anticipated in actual
use. The aim is to cause failures that identify weak design elements. Failure-prone elements are then re-designed and tests are
continued at higher levels. Generally till to day, product quality has been determined through environmental testing such as
vibration tests, thermal cycling, mechanical shock, thermal shock, and more. Recently, there has been a significant trend in
today's globalized competitive market to improve product quality even further.

Highly Accelerated Life Testing (HALT)

Highly Accelerated Life Testing (HALT) and Highly-Accelerated Stress Screening (HASS) testing are the futuristic
techniques for Environmental testing. All external and internal environmental conditions in terms of temperature, humidity,
radiation, magnetic and electric fields, shock and vibration and natural, man-made or self-induced conditions influence the
form, function, and reliability of a product. Environmental testing involves subjecting a sample of products to a simulation of
anticipated storage, transport, and service environments in terms of vibration, shock, temperature, altitude and humidity. As
such Highly-Accelerated Life Testing (HALT) and Highly-Accelerated Stress Screening (HASS) are starting to supplement
traditional vibration and thermal testing to meet these new quality targets.
Highly Accelerated Life Testing (HALT) exposes the product to a step-by-step cycling in environmental variables
such as temperature, shock and vibration. HALT involves vibration testing in all three axes using a random mode of
frequencies. HALT testing can include the simultaneous cycling of multiple environmental variables, for example,
temperature cycling plus vibration testing. This multi-variable testing approach provides a closer approximation of real-world
operating environments. Unlike conventional testing, the goal of HALT testing is to break the product. When the product fails,
the weakest link is identified. Designers identify exactly what needs to be done to improve product quality. After a product has
failed, the weak component(s) are upgraded or reinforced. The improved product is then subjected to another round of HALT
testing, with the range of temperature, vibration, or shock further increased to make the product fail again. This identifies the
next weakest link. By going through several iterations the product can be made quite robust. This informed approach, only
identifies the weak spots for improvement. This type of testing provides so much information about the construction and
performance of a product, that it can be quite helpful for newer designers assigned to a product with which they are not
completely familiar. HALT testing must be performed during the design phase of a product to make sure the basic design is



Highly Accelerated Stress Screening (HASS)

Highly Accelerated Stress Screening (HASS) is a reduced form of HALT testing. HASS testing is an on going
screening test, performed on regular production units. The idea is not to damage the product, but rather to verify that actual
production units continue to operate properly when subjected to the cycling of environmental variables used during the HASS
test. The limits used in HASS testing are based on a skilled interpretation of the HALT testing parameters.
The components, materials, and processes can and do change over time, thereby affecting the quality and reliability of the final
product. The best way to ensure that production units continue to meet reliability objectives is through HASS testing.

Objectives of HASS
Some of the objectives of HASS are to:
1. Precipitate relevant defects from latent to patent at minimum cost and minimum time.
2. Detect as many defects as possible at minimum total cost and in a minimum time in order to reduce feedback delay
and cost.
3. Provide the start of a closed loop failure analysis and corrective action program of all defects found in the screens.
4. Increase field reliability by reducing the total no of faults sent to the field.
5. Decrease the total cost of production, screening, maintenance and warranty.

Purposes of HALT and HASS

General purpose of applying accelerated stress conditions in the design phases is to find and improve upon design and
process weaknesses in the least amount of time and correct the source of their weaknesses before production begins. It is
generally true that, robust products will exhibit much higher reliability than non- robust ones and so the ruggedization process
of HALT in which large margins are obtained will generate products of high potential reliability. In order to achieve the
potential, defect- free hardware must be manufactured or, atleast the defects must be found and fixed before shipment. In
HASS, accelerated stresses are applied in production in order to shorten the time to failure of the defective units and therefore
shorten the corrective action time and the number of units built with the same flaw. Each weakness found in HALT or in HASS
represents an opportunity for improvement. The application of accelerated stressing techniques to force rapid design maturity
(HALT) Results in pay-backs that far exceed these from production stressing (HASS). Nonetheless, production HASS is cost
effective in its own right until quality is such that a sample HASS or Highly Accelerated stress Audit (HASA) can be put into
place. The use of HASA demands excellent process control since most units will be shipped without the benefit of HASS being
performed on them, and those units in the selected sample will be screened for defects.
The stress used in HALT and HASS include, but are not restricted to, all-axis simultaneous vibration, highrate, broad-
range, temperature cycling, power cycling, voltage and frequency variation, humidity, and any other stress that may expose
design or process problems. No attempt is made to simulate the field environment. One only seeks to find design and process
flaws by any means possible. The stress used generally far exceeds the field environments in order to gain time compression;
that is, shorten the time required to find any problem areas. When a weakness is discovered, only the failure mode and
mechanism is of importance, the relation of the stress used to the field environment is of no consequence at all.
'Mechanism' here means the conditions that caused the failure, such as melting, exceeding the stable load or
exceeding the ultimate strength. The corresponding failure mode could be separation of a conductor, elastic buckling and
tensile failure, considering the margin instead of the failure mode is a major mistake, which is made by most engineers used to
conventional test techniques. In HALT and HASS one uses extreme stresses for a very brief period of time in order to obtain
time compression in the failures. In doing so, one may obtain the same failures as would occur in the field environments, but
with a different stress.
In the HALT phase of product development, which should be in the early design phase, the product is improved in
every way practicable bearing in mind that most of what are discovered in HALT as weaknesses will almost surely become
field failures if not improved.
HALT and HASS are not restricted to electronics items, but apply to many other technologies as well. Some of the
diverse examples include products like shock absorbers, airframes, auto bodies, exhaust systems and power steering hoses etc.



Equipment Required
The application of the highly accelerated stress techniques is very much enhanced by, if not impossible without, the use of
environmental equipment of the latest design such as all –axis exciters and combined very high –rate thermal chambers (60 °
C/min or more rate). All- axis means three translations and three rotations.
A single-axis, single-frequency shaker will only excite modes in the particular direction of the vibration and only those nearby
in frequency. A swept sine will sequentially excite all modes in the one direction being excited. A single – axis random shaker
will simultaneously excite all modes in one direction. A Six axis system will simultaneously excite all modes within the
bandwidth of the shaker in all directions .If all modes in all directions are not excited simultaneously, then defects can be
missed. Obviously, the all- axis shakers are superior for HALT and HASS activities since one is interested in finding as much as
possible as fast as possible.

Some General Comments on HALT and HASS

The successful use of HALT or HASS requires several actions to be completed. In sequence, these are precipitation,
detection, failure analysis, corrective action, verification of corrective action and their entry into a database. All the first five
must be done in order for the method to function at all. Adding the sixth, results in long-term improvement of the future

1. Precipitation means to change a defect, which is latent or undetectable to one that is patent or detectable. A poor solder
joint is such an example. When latent it is probably not detectable electrically unless it is extremely poor. The process
of precipitation will transpose the flaw to one that is detectable; that is, crack. This crack joint maybe detectable under
modulated excitation. The stress used for the transportation maybe vibration combined with thermal cycling and
perhaps electrical over stress. Precipitation is usually accomplished in HALT or in Precipitation sreen.
2. Detection means to determine that a fault exits. After precipitation by whatever means, it may become patent, that is,
detectable. Just because it is patent does not mean that it will actually be detected since it must first be put into a
detectable state, perhaps using modulated excitation, and then it must actually be detected. Assuming that we actually
put the fault into a detectable state and that the built in test of external test setup can detect the fault, we can then
proceed to the most difficult step, which is failure analysis.
3. Failure Analysis means to determine why the failure occurred. In the case of the solder joint, we need to determine
why the joint failed. If doing HALT, the failed joint could be due to a design flaw; that is, a extreme stress at the joint
due to vibration or may be due to a poor match of thermal expansion coefficients. When doing HASS, the design is
assumed to be satisfactory and in that case the solder joint was probably defective. In what manner it was defective
and why it was defective. In what manner it was defective and why it was defective need to be determined in sufficient
detail to perform the next step, which is corrective action.
4. Corrective action means to change the design or process as appropriate so that the failure will not occur in the future.
This step is absolutely essential if success is to be accomplished. In fact, corrective action is the main purpose of
performing HALT or HASS.
5. Verification of corrective action needs to be accomplished by testing to determine that the product is really fixed and
that the flaw, which caused the problem, is no longer present.
6. The last step is to put the lesson learned into a database from which one can extract valuable knowledge whenever a
similar event occurs again. Designers and manufacturers, who practice correct HALT and utilize a well-kept
database, soon become very adept at designing and building very robust products with the commensurate high

The HALT process would be a good tool to determine the answer to the following tongue –twister riddle- ' How much
wood would a woodchucker chuck if a woodchucker could chuck wood?'
The basic principle of highly accelerated life testing is the margin discovery process. The goal of this process is to
determine the margins between the service conditions of the product, the functional limits and the destruct limits. In other
words, how much wood would the wood chucker chuck, or how much temperature can the circuit board handle, or how much



vibration, or how cold, or how much voltage.

HALT is a test method usually applied to solid state electronics that determines failure modes, operational limits and
destruct limits. This test method differs significantly from reliability test. Halt does not determine a statistical reliability or
determine any estimated life. HALT applies once stress source at a time to a product at elevated levels to determine the levels at
which the product stops functioning but is not destroyed (operational limit) .The levels at which the product is destroyed
(destruct limit) and what failure modes cause the destruction of the product.
HALT has a significant advantage over traditional reliability test in identifying failure modes in a very short period of
time. Traditional reliability tests take a long period of time (from a few days to several months). A HALT typically takes two or
three days. Also, a HALT will identify several failure modes providing significant information for the design engineers to
improve the product. Typical reliability test will provide one or two failure modes, if the product fails at all.
HALT typically uses three stress sources: temperature, vibration and electrical power. Each stress source is applied
starting at some nominal level (for example 30 deg C ) is then elevated in increments until the product stops functioning. The
product is then brought to the nominal conditions to see if the product is functional. If the product is still functional, then the
level at which the product stops functioning is labeled the operational limit. The product is then subjected to levels of stress
above the operational limit, returning to the nominal levels each time until the product does not function. The maximum level
the product experiences before failing to operate at nominal condition is labeled the destruct limit. This process is repeated for
hot temperature, cold temperature, temperature ramp rate, vibration and voltage. The process is also repeated for combined

"Classic" vibration and temperature testing have definitely helped improve product quality over the years, but today's
very high standards for product quality and reliability require tests with a better ability to reduce or even eliminate the failures
in the field conditions. HALT testing provides a controlled, repeatable method of determining product quality under conditions
comparable to field operating conditions, and is critical for proving the basic design of a product. HASS testing is a quick,
effective screening process, which can be used to ensure that the production units continue to meet quality standards. While it
is true that HALT and HASS testing can add to the short-term manufacturing cost of a product, the increment is surprisingly
small in most cases. In the long run, the cost of the testing is much less than the cost of failures in the field conditions arising due
to reliability problems.
The challenge in front of young scientists / technocrats of DRDO is how to make use of HALT and HASS testing
methodologies in making more dependable, reliable Armament Stores, that which is explosive filled and potentially hazardous
in nature. The big question remains— Can we go upto the 'breaking point ' in case of explosive filled stores in terms of
temperature and vibration levels etc. Since the futuristic testing is HALT and HASS, a solution to the question has to be found
out in the near future as traditional testing methodologies become outdated.


1. “Accelerated testing & Validation” by Alex Porter ,London, Elsevier, 2004.

2. Notes on “Highly Accelerated Reliability Testing” , Short term Course, IIT, Kharagpur.




Niteen P Bhange, AK Ghosh
IIT, Kanpur -208016

In an effort to develop the capability to predict sabot separation dynamics within the framework of a simple physical
model, computational code is developed and named as MK-Code. MK-Code computes pressure distribution on the sabot's
front scoop and underside in high supersonic flow speed using simple analytical aerodynamic model. The experimental test
data available for IAT HVP_94_016 sabot configuration and firing test results on MK-II package are used to investigate the
accuracy of the MK-Code for predicting pressure distribution on sabot during its discard


V = Velocity of aircraft
a = Angle of attack
q= Pitch angle
q= Pitch rate
= Oblique shock angle
= Angle between sabot and penetrator
= Maximum turning angle
P2 = Pressure at the tip of sabot
g= Specific heat ratio
M = Mack number
u 0 = P-M angle

Anti tank kinetic energy projectiles are generally launched with initial velocities of around 5 – 6 Mach. The
projectiles are normally sabot launched. A sabot is a device usually made of two or more petals that encases the projectile,
providing needed structural support during launch. Because most projectiles are sub-caliber, i.e. projectiles diameter is smaller
than the barrel diameter, the sabot fills the space between the projectile and barrel wall, preventing the muzzle gases from
blowing by the projectile. However, after the launch, the sabot becomes a parasitic mass and must be removed as quickly as
possible to allow low-drag flight of the projectile to the target.
This is normally accomplished by incorporating a scoop into the front of the sabot so that aerodynamic forces lift and
rotate each petal away from the projectile. A good knowledge of the pressure distribution on the sabot is essential to predict the
dynamic behavior of sabot during the discard process. Researchers1-4 have presented various methods to compute pressure
distribution over sabot using experimental and numerical methods.
Guillot et al in Ref. 1 have presented results pertaining to pressure distribution on sabots in hypersonic flight. Dick et
al2 presented method to compute pressure distribution on sabot during discard. In Ref. 3, sabot separation studies have been
carried out experimentally using wind tunnel tests. Guillot and Reinkey4 presented exhaustive work wherein both numerical
and experimental methods have been applied to analyse sabot separation dynamics.
In Ref. 4, experimental data pertaining to pressure distribution over example sabot nomenclatured as IAT
HVP_94_016 have been presented.The sabot configuration of IAT HVP_94_016 has been presented in Fig. 1.



Figure 1:- IAT HVP_94_016 four petal sabot Model

Figure2:- MK-II Three petal Sabot Model

Pressure Modeling
Theoretically, one could compute the pressure distribution on the sabot using the full Navier–Stokes equations5 and
couple this solution to the dynamic equations governing the motions of the sabot petals to compute sabot trajectories
accurately. Such intensive computations during sabot design are not feasible, so we need analytical model for quicker
simulations, than solving the entire problem using a CFD based approach. However, such a simplified model based on
analytical approach must capture the important physics dominating the sabot separation process.

Figure 3 : Diagram showing Overall pressure distribution during sabot discard



Fig. 3 presents a summary overview, which highlights essentially all of the local flow elements previously used to
describe the overall surface pressure/aerodynamic force pattern. The key features involve impingement of the sabot bow shock
upon the projectile surface resulting in a shock-reflection process. The reflected shock then impinges upon the sabot petal
underside creating another separated zone of elevated pressure. An code (MK-code) has been developed which can be used as
practical tool for sabot configuration design. The code uses 2-D aerodynamic model to compute aerodynamic forces and
moment on the sabot during its launch to discard.

Front-scoop Pressure Distribution

The MK-code can be used as practical design tool. Basic assumption made here is that the aerodynamic forces acting
on the sabot petals are approximated simpler 2-D aerodynamic models. Front scoop geometry needs to be divided in to
segments for application of 2-D gas dynamics. For our purpose, we have considered 6 segments. It was observed by us that a
higher number of segments gave better results. On front scoop, the pressure distribution depends on the angle of attack of
sabot. If turning angle of the flow being encountered on any segment on front scoop is small enough for an attached oblique
shock to exist, the pressure on that segment is assumed to be constant at the value behind the oblique shock.

If turning angle is greater than that for an

attached shock, the pressure on the front scoop is
assumed to vary linearly from stagnation at the outer
radius to sonic at inner radius. Occurrence of bow shock
in any one of the segments of the sabot front scoop
depends upon the angle of attack seen by the sabot, pitch
rate of the sabot and angle between sabot segment and
penetrator. These situations are at small angle of attack
and attached shock exists on the outer segments whereas
detached shock exists ahead of remaining segments.

Figure 4 : Diagram showing Front-Scoop pressure distribution during sabot discard

During compression oblique shock angle is found by solving Eq. 1.

æM 2 sin 2 b
w -1 ö
tan q =
2 cot b
wç ÷ [1]
w ç2
M ( 2÷
è g+ cos 2b w)+ ø

Pressure behind oblique shock is obtained through Eq. 2.

æ2 g2 ö
1 +M sin 2 b
P1 =
ç (
w -

÷P ) [2]
èg+1 ø



Mach number behind oblique shock can be calculated through Eq. 3

+ 1 2 sin b w sin q
ç M +

ç cos ( ÷
è2 bw - q ø [3]
M1 =
sin 2 bw

During shock the expansion of flow Mach number after expansion is calculated by solving Eq. 4

1 -
1 g
0 = tan (M 2 - tan -
1) - 1
M 2-
1 [4]
1 g

Pressure after expansion of flow can be calculate through Eq. 5.

æg- 1 2ö(
ç M1 ÷
è 2 ø´
P1 = -g
P [5]
æg- 1 2ö(
ç M ÷
è2 ø

Condition for maximum deflection angle for given Mach no. is as follows:

4 M - 1 () 2 2
Max = 2
3 3(g1)
+ M

If turning angle is greater than that for an attached shock, the pressure on the front scoop is assumed to vary linearly from
stagnation at the outer radius to sonic at inner radius. Sonic pressure behind normal shock is given by

æ2g2 ö
Psonic =
P1 ç
1 +M -
1÷ () [7]
èg+1 ø

Stagnation pressure behind normal shock is calculated by Eq. 8.

æ ö
ç æ g
+ 1 2ö g -
g 1)
ç ç
M ÷ ÷
Pstag P0 ç è
= ø
1 ÷ [8]
ç M1 -
2g (g 1)
- g
- 1 ÷
ç ÷
è ø
Underside Pressure Modeling
The beginning of the pressure pulse is determined by projecting the line from forward tip by making an angle with
penetrator equal to oblique shock angle formed by contracting flow when it is turned by angle á. These serve as the beginning
of the pressure pulse. A mach wave is then reflected back up to the sabot underside at the mach wave angle. The end of the
pressure pulse is determined by reflecting a Mach wave from the same location on the projectile back up to the point where it
intersects the underside of the sabot. The base pressure level on the underside of the each sabot petal is computed by assuming
that underside incoming flow first passes through a strong shock with a turning angle equal to d max
then it gets expanded by an
angle d max - a s
by Prandtl-Meyer (PM) expansion .



Figure 5 : Diagram showing Underside Pressure Distribution during sabot discard

However the pressure along the underside of sabot increases in the axial direction from tip to tail due to shock
reflection and shock interaction This depends on the Mach number in previous zone, which decides the condition for d maxin the

next zone, and accordingly it decides whether shock will be oblique or normal. Geometric distances over sabots are calculated
by following equation,

Ai ´-
cos( b
1 qrel )
Sa =
sin b1

Ai ´-
sin( b1 qrel )
Sb =
sin b1 tan( b2 +
q rel )

To calculate distances after nth reflection, the value of Ai is computed using Eq. 10.
An -

sin( b 1-
rel )
An i = n-
sin bn-1

As the flow progressively gets reflected, the Mach number goes on decreasing.

The flow gets choked since the gap between the petals is not large enough to accommodate the incoming mass flow
rate. The interaction is characterized by a single normal shock upstream of the sabot petals and essentially a stagnant volume of
air within the scoops. The subsonic flow behind the normal shock is accelerated to supersonic Mach numbers as it turns
through a series of expansion waves and aligns itself with the petal interfaces bounding the core flow. (This is similar to the
case of subsonic flow in a converging channel, which has to accelerate.) However, since the area between the petals continues
to decrease, the supersonic flow will choke again, resulting in a second normal shock called the throttling shock. The location
of the throttling shock depends on the area ratio and the Mach number of the supersonic flow between the petals. The area ratio
can be calculated from the radial position and angle-of-attack of the petal at that instant in time. Estimating the local static
pressure, using hypersonic relations based on blast wave theory, and assuming isentropic expansion from stagnation
conditions behind the first normal shock upstream of the entire package, we can calculate the Mach number. The flow through
the launch package remains choked until the area between the petals is large enough to accommodate the incoming mass flow



Model Validation Hvp_94_016 Sabot Test Data

It was necessary to perform validation to verify that the inviscid 2-D gas dynamics used for finding pressure
distribution on sabot surface. Experimental data available from wind tunnel test conducted in the Mach 5 blow down wind
tunnel of the University of Texas at Austin on IAT HVP_94_016 sabot geometry (developed by US army ballistic research
laboratory) was taken as reference. During the test, trailing edge distance from penetrator maintained during course of the
experiment is Y (Table-1), whereas AOA is determined relative to the free stream velocity vector.

Table : Sabot discard point used

in experimental study

Aerodynamic Model Investigation

The experimental conditions were M ¥ = 4.95 , P¥= 0.6715 psi and V¥ =2734 ft / s . The experiment was performed at
four distinct configurations, which are identified by the sabot's angle of attack (AOA) and trailing edge gap (Table 1).

a) 3-Deg AOA b) 7-Deg AOA c) 14-Deg AOA d) 23-Deg AOA

Figure 6. Experimental and computed pressure distribution along centerline of sabot front scoop and underside

All the experimental data show sharp pressure increases at the tail of the sabot. Although it is reasonable to expect that
flow choking down through a reduced area would cause a pressure increase, MK-Code was used for finding pressure
distribution on HVP_94_016 sabot surface. Also these results were compared with AVCO code prediction of pressure
distribution. The AVCO code incorporates a pressure model based on two-dimensional inviscid gas dynamics applied to the
centerline of the sabot. Figure 6 presents a comparison of pressure distribution (computed using simulated (MK-code),
Experimental and AVCO code). A fairly goo