Criteria of Design Improvement of Shaped Charges

CRITERIA OF DESIGN IMPROVEMENT OF
SHAPED CHARGES
USED AS OIL WELL PERFORATORS

A Thesis submitted to The University of Manchester for the degree of
Doctor of Philosophy

The Faculty of Engineering and Physical Sciences

2012

Tamer Abd Elazim Elshenawy

School of Mechanical, Aerospace and Civil Engineering

2

(This page is intentionally left blank)

Table of Contents
3
TABLE OF CONTENTS
Table of Contents ................................................................................................................. 3
List of Figures ..................................................................................................................... 10
List of Tables ...................................................................................................................... 18
Nomenclature ..................................................................................................................... 21
Abstract ............................................................................................................................... 26
Declaration ......................................................................................................................... 27
Copy Right Statement ....................................................................................................... 28
Dedication ........................................................................................................................... 29
Acknowledgements ............................................................................................................ 30
List of Publications ............................................................................................................ 31
CHAPTER.1 INTRODUCTION ....................................................................... 32
1.1 Research Background ............................................................................................ 32
1.2 Originality of Research ......................................................................................... 33
1.3 Objectives and methodology of research ............................................................. 34
1.4 Thesis structure ...................................................................................................... 34
CHAPTER.2 LITERATURE REVIEW ...........................................................36
2.1 Overview ................................................................................................................. 36
2.1.1 Shaped charge phenomenology .................................................................................................. 36
2.1.2 Shaped charge applications ......................................................................................................... 37
2.1.2.1 Development of the shaped charge and its wide usage in military applications .................. 37
2.1.2.2 Civil applications of shaped charge ...................................................................................... 39
2.1.3 Application of shaped charge in the oil field .............................................................................. 39
2.1.3.1 Introduction ........................................................................................................................... 39
2.1.3.2 Description of the shaped charge used as perforator ............................................................ 40
2.2 Factors affecting the shaped charge used as oil well perforators (OWP) ......... 42
2.2.1 Stand-off distance ........................................................................................................................ 43
2.2.2 High explosive ............................................................................................................................ 44
2.2.3 Liner geometry ............................................................................................................................ 45
2.2.4 Detonating wave form ................................................................................................................. 47
2.2.5 Symmetry .................................................................................................................................... 48
2.2.6 Liner materials ............................................................................................................................ 49
Table of Contents
4
2.2.6.1 Introduction ........................................................................................................................... 49
2.2.6.2 Liner material grain size ....................................................................................................... 52
2.2.6.3 Liner crystal shape ................................................................................................................ 56
2.2.6.4 Liner impurities ..................................................................................................................... 57
2.2.6.5 Strain-rate .............................................................................................................................. 58
2.2.7 Liner manufacturing .................................................................................................................... 59
2.2.7.1 Flow turning (spinning or shear forming) ............................................................................. 60
2.2.7.2 Deep drawing ........................................................................................................................ 60
2.2.7.3 Cold forging .......................................................................................................................... 61
2.2.7.4 Warm forging ........................................................................................................................ 61
2.2.7.5 Hot forging ............................................................................................................................ 62
2.2.7.6 High energy rate fabrication (HERF) ................................................................................... 62
2.2.7.7 Electroforming copper .......................................................................................................... 62
2.2.7.8 Infiltrating technology .......................................................................................................... 63
2.2.7.9 Press moulding or powder metallurgy technique ................................................................. 64
2.2.8 Applied pressure on perforators .................................................................................................. 65
2.2.8.1 Under-balance technique ...................................................................................................... 65
2.2.8.2 Overbalance (traditional) drilling technique ......................................................................... 67
2.3 Oil well perforator testing according to API-RP43 ............................................ 67
2.4 Rock material properties ....................................................................................... 68
2.4.1 Introduction ................................................................................................................................. 68
2.4.2 Rock stresses and penetration of jet into its material .................................................................. 70
2.5 Summary ................................................................................................................. 71
CHAPTER.3 SHAPED CHARGE JET FORMATION AND PENETRATION
MODELS ......................................................................................................72
3.1 Introduction ............................................................................................................ 72
3.1.1 Steady state Birkhoff theory for jet formation ............................................................................ 72
3.1.2 Unsteady state PER theory .......................................................................................................... 74
3.1.3 Modifications to PER theory ....................................................................................................... 78
3.1.4 Jet elongation behaviour ............................................................................................................. 80
3.2 The Gurney velocity approximation .................................................................... 80
3.2.1 Introduction to Gurney formulae ................................................................................................ 80
3.2.2 Determination of Gurney energy and Gurney velocity ............................................................... 81
3.2.3 Formulae for different configurations ......................................................................................... 82
3.2.3.1 Open faced sandwich ............................................................................................................ 82
3.2.3.2 Symmetrical sandwich [74] .................................................................................................. 83
3.2.3.3 Asymmetrical sandwich [74] ................................................................................................ 83
Table of Contents
5
3.2.3.4 Infinitely tamped sandwich ................................................................................................... 83
3.2.3.5 Cylinderical shell .................................................................................................................. 84
3.2.3.6 Spherical shell ....................................................................................................................... 84
3.2.3.7 Formulae for the Gurney approximation in the shaped charge ............................................ 84
3.3 Visco-plastic model and jet coherency ................................................................. 86
3.4 Breakup time models ............................................................................................. 88
3.4.1 Empirical formulae ...................................................................................................................... 88
3.4.2 Hydrocode Simulations ............................................................................................................... 91
3.4.3 Analytical Models ....................................................................................................................... 92
3.5 Shaped charge jet penetration models ................................................................. 93
3.5.1 Uniform velocity jet .................................................................................................................... 93
3.5.1.1 Jet breakup effect .................................................................................................................. 95
3.5.1.2 Standoff distance effect ....................................................................................................... 95
3.5.1.3 Target material effect ............................................................................................................ 96
3.5.2 Variable velocity jet .................................................................................................................... 97
3.5.3 Particulated jet ............................................................................................................................. 98
3.5.4 Target strength ............................................................................................................................ 99
3.6 Crater growth process ......................................................................................... 101
3.7 Summary ............................................................................................................... 102
CHAPTER.4 HYDROCODE SIMULATION .................................................103
4.1 Introduction .......................................................................................................... 103
4.2 Studied parameters .............................................................................................. 104
4.3 Autodyn jetting analysis description .................................................................. 105
4.4 Autodyn jet formation model description ......................................................... 106
4.5 Material modeling description ............................................................................ 108
4.5.1 Description of the used explosives ............................................................................................ 108
4.5.2 Explosive initiation and wave propagation ............................................................................... 108
4.5.3 Description of the liner materials .............................................................................................. 110
4.5.4 Description of the charge case .................................................................................................. 112
4.5.5 Description of the concrete material ......................................................................................... 113
4.5.5.1 General .................................................................................................................. 113
Table of Contents
6
4.5.5.2 Equation of state of the concrete material ......................................................... 114
4.5.5.3 Strength model for concrete ................................................................................................ 116
4.5.5.3.1 The failure surface: ........................................................................................................ 116
4.5.5.3.2 Elastic limit surface: ....................................................................................................... 117
4.5.5.3.3 Strain hardening ............................................................................................................. 117
4.5.5.3.4 Residual failure surface .................................................................................................. 118
4.5.5.3.5 Damage ........................................................................................................................... 118
4.5.6 Description of the layer of the steel gun carrier and the wellbore casing ................................. 119
4.5.7 Description of the water layer ................................................................................................... 120
4.6 Solution stability .................................................................................................. 120
4.7 Output of numerical modeling ............................................................................ 121
CHAPTER.5 PARAMETRIC ANALYSIS RESULTS ..................................... 122
5.1 Introduction .......................................................................................................... 122
5.2 The main features of the jetting analysis and the jet formation and penetration
solvers .............................................................................................................................122
5.2.1 Standard shaped charge jetting analysis model......................................................................... 122
5.2.2 Shaped charge jet formation and penetration ............................................................................ 125
5.3 Mesh sensitivity study .......................................................................................... 128
5.3.1 Mesh sensitivity for the jetting analysis .................................................................................... 128
5.3.2 Mesh sensitivity for the jet penetration ..................................................................................... 130
5.4 The verification and Validation (V&V) of the hydro-code results .................. 131
5.4.1 The jetting analysis and the jet formation Validation ............................................................... 131
5.4.2 The validation of the hydro-code penetration modeling ........................................................... 133
5.5 Effect of the surrounding medium on the jet characteristics .......................... 134
5.6 Shaped charge parametric study results ........................................................... 136
5.6.1 High explosive effect on jet performance ................................................................................. 136
5.6.2 Liner wall thickness effect ........................................................................................................ 139
5.6.2.1 Uniform liner wall thickness ............................................................................................... 139
5.6.2.2 Varied liner wall thickness .................................................................................................. 142
5.6.3 Cone apex angle ........................................................................................................................ 142
5.6.3.1 Standard jetting analysis ..................................................................................................... 143
5.6.3.2 Jet formation and penetration calculations ......................................................................... 144
5.6.3.3 The optimization of the cone apex angle and liner thickness parameters .......................... 146
5.6.4 Liner shape and its geometry .................................................................................................... 151
Table of Contents
7
5.6.5 Explosive amount and head height ........................................................................................... 152
5.6.6 Water stand-off distance ........................................................................................................... 153
5.6.7 Degree of confinement effect on the jet parameters ................................................................. 155
5.6.8 Effect of the initiation point on jet characteristics .................................................................... 157
5.7 Liner portioning into jet and slug ...................................................................... 158
5.8 The Gurney velocity approximation .................................................................. 161
5.9 Summary ............................................................................................................... 163
CHAPTER.6 INFLUENCES OF TARGET STRENGTH AND CONFINEMENT
ON THE PENETRATION DEPTH OF AN OIL WELL PERFORATOR .......... 164
6.1 Introduction .......................................................................................................... 164
6.2 Experiments .......................................................................................................... 166
6.3 Numerical models ................................................................................................ 167
6.3.1 Hydro-Code Algorithms ........................................................................................................... 167
6.3.2 Mesh sensitivity ........................................................................................................................ 167
6.3.3 Material models ......................................................................................................................... 169
6.4 Results and discussion ......................................................................................... 169
6.5 Summary ............................................................................................................... 176
CHAPTER.7 PERFORMANCE OF ZIRCONIUM JET WITH DIFFERENT
LINER SHAPES................................................................................................. 177
7.1 Introduction .......................................................................................................... 177
7.2 Critical angle calculations conditions for the zirconium jets ........................... 180
7.3 Experiments .......................................................................................................... 186
7.4.1 Methodology ............................................................................................................................. 188
7.4.2 Mesh sensitivity ........................................................................................................................ 188
7.4.3 Material models ......................................................................................................................... 190
7.5 Results ................................................................................................................... 190
7.5.1 The
c
-V
2
calculations ................................................................................................................ 190
Table of Contents
8
7.5.2 Jet analysis and penetration of different liner shapes ............................................................... 193
7.5.3 Penetration ................................................................................................................................. 197
7.6 Summary ............................................................................................................... 199
CHAPTER.8 A MODIFIED VIRTUAL ORIGIN MODEL FOR SHAPED
CHARGE JET PENETRATION WITH NON-UNIFORM DENSITY DISTRIBUTION
.................................................................................................. 200
8.1 Introduction .......................................................................................................... 200
8.2 Penetration analytical model .............................................................................. 202
8.3 Liner materials and penetration experiments ................................................... 207
8.4.1 Hydrocode algorithms of the jetting analysis, the jet formation model and the jet penetration
model ..................................................................................................................................................211
8.4.2 Mesh sensitivity for the jet formation model ............................................................................ 212
8.4.3 Mesh sensitivity for the jet penetration into target ................................................................... 214
8.5 Results ................................................................................................................... 214
8.5.1 The Jetting analysis results ....................................................................................................... 214
8.5.2 Jet density distribution .............................................................................................................. 215
8.5.3 The penetration depth calculations ........................................................................................... 216
8.6 Summary ............................................................................................................... 219
CHAPTER.9 ZIRCONIUM SHAPED CHARGE JET BREAKUP TIME .. 220
9.1 Introduction: ........................................................................................................ 220
9.2 Determination of V
PL
........................................................................................... 221
9.3 Calculation of the J-C constitutive equation parameters for zirconium ........ 223
9.4 The jet temperature estimation .......................................................................... 225
9.5 Results ................................................................................................................... 226
9.5.1 Calculations of the necking strain and V
PL
using J-C constitutive equation ............................ 226
9.6 Summary ............................................................................................................... 231
Table of Contents
9
CHAPTER.10 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE
STUDIES ................................................................................................. 232
10.1 Introduction .......................................................................................................... 232
10.2 Conclusions ........................................................................................................... 232
10.3 Future work .......................................................................................................... 233
References ......................................................................................................................... 235
Appendices ........................................................................................................................ 246

Word count: 59,473

List of Figures
10
LIST OF FIGURES
Figure 2-1 A schematic drawing of a shaped charge configuration. ................................... 36
Figure 2-2 Collapse of the liner and the formation of the jet [6]. ........................................ 37
Figure 2-3 Different effects of shaped charge on target, (a) unlined cavity effect, (b) lined
cavity effect, and (c) lined cavity with stand-off distance. ............................... 37
Figure 2-4 OWP fitted inside the gun carrier facing the cement and concrete materials. ... 41
Figure 2-5 Schematic drawing of the location of down hole gun perforators, steel carrier,
well fluid, well casing, cement layer, and hydrocarbon rock [17]. ................... 42
Figure 2-6 Depth of penetration versus stand-off [20]. ....................................................... 43
Figure 2-7 Penetration versus liner cone angle [26]. ........................................................... 46
Figure 2-8 Shaped charge jet profile at different cone angle [6]. ........................................ 46
Figure 2-9 Jet and tail velocities as a function of cone angle [26]. ..................................... 47
Figure 2-10 The original and modified liner shapes [17]. ................................................... 47
Figure 2-11 The shape of the DW travelling inside CSC explosive charges with and
without wave shaper. ........................................................................................ 48
Figure 2-12 The shaped charge in Wang and Zhu [33]. ..................................................... 51
Figure 2-13 Anomalous behavior of Cu-W jet at large stand-off distance [33]. ................. 52
Figure 2-14 Variation of penetration depth with stand-off distance for Cu-W and Cu jets
[33]. ................................................................................................................... 52
Figure 2-15 The radiographs of the jet formed by ECu/W powder after 50 and 90 s from
initiation. Lengths of jets: 292 and 572 mm corresponding to instantaneous jet
velocity of 7.25 and 7.0 km/s, respectively [15]. .............................................. 52
Figure 2-16 Stress-strain curves of 3N, 6N and 8N copper samples at strain-rate of 4.210
-5

s
-1
[37]. .............................................................................................................. 54
Figure 2-17 Dogbone samples of the nano copper [41]. ...................................................... 58
Figure 2-18 The stress-strain curve at different strain-rates [41]. ....................................... 59
Figure 2-19 Flow turning technique [31]. ............................................................................ 60
Figure 2-20 Deep drawing technique [31]. .......................................................................... 61
List of Figures
11
Figure 2-21 Cold forging technique. .................................................................................... 61
Figure 2-22 Warm forging technique [31]. .......................................................................... 62
Figure 2-23 Sketch diagram of infiltrating technology [33]. ............................................... 63
Figure 2-24 The penetration hole; the damaged and the crushed area profiles for both the
balanced (left) and the 300-psi under-balanced perforation (right). ................. 66
Figure 2-25 Test setup of the perforating gun [17]. ............................................................. 68
Figure 2-26 A schematic diagram indicating the concrete API testing configuration [17]. 68
Figure 3-1 The collapse process according to the steady state theory [49]. ........................ 73
Figure 3-2 A schematic drawing of non steady state jet formation according to PER theory
[50]. ................................................................................................................... 75
Figure 3-3 Geometry showing parameters in unsteady state theory [50]. ........................... 76
Figure 3-4 Relation among collapse, flow and stagnation velocities. ................................. 77
Figure 3-5 The acceleration of the liner element with the time, (a) infinite acceleration, (b)
linear, (c) exponential. ...................................................................................... 79
Figure 3-6 The inverse velocity gradient [6]. ...................................................................... 79
Figure 3-7 The open faced sandwich configuration [6]. ...................................................... 82
Figure 3-8 The symmetrical configuration. ......................................................................... 83
Figure 3-9 The asymmetric configuration. .......................................................................... 83
Figure 3-10 The Infinitely tamped sandwich configuration. ............................................... 83
Figure 3-11 The cylindrical configuration [6]. .................................................................... 84
Figure 3-12 The spherical configuration [6]. ....................................................................... 84
Figure 3-13 A schematic diagram of the collapsing liner under explosive load [67]. ......... 86
Figure 3-14 Comparison between hydrocode simulation of jet necking due to instability
and flash radiograph of a jet at approximately the same time [89]. .................. 92
Figure 3-15 Hydrodynamic penetration of jet into target [6]. ............................................. 94
Figure 4-1 The flow chart of the different stages and input and output results from the
jetting analysis and the two solvers (Euler and Lagrange) of the Autodyn
hydrocode. ....................................................................................................... 104
Figure 4-2 The jetting model of OWP with cone apex angle 60
o
and liner thickness
List of Figures
12
1.74mm under fixed apex node boundary condition. ..................................... 106
Figure 4-3 The Euler part in 2-D Visualizer showing the geometry and the boundaries of
the jet formation model. .................................................................................. 107
Figure 4-4 The produced jet obtained from Euler solver and remapped to Lagrange
processor for penetration analysis. .................................................................. 107
Figure 4-5 The produced jet impacting on steel gun casing, water wellbore fluid, steel
casing and concrete material. .......................................................................... 108
Figure 4-6 The different stages of detonation wave spherical propagation through the
explosive charge inside CSC. ......................................................................... 110
Figure 4-7 The shock velocity against particle velocity for the EOS of the liner material
[116]. ............................................................................................................... 111
Figure 4-8 The pressure-porosity curve for the concrete material [116]. .......................... 114
Figure 4-9 Stress loading curve for the RHT strength material model [116]. ................... 117
Figure 4-10 The concrete strain hardening curve according to RHT model. .................... 118
Figure 5-1 The different stages of the jetting analysis of the OWP 60
o
cone apex angle and
liner wall thickness of 1.74mm. ...................................................................... 123
Figure 5-2 The stagnation, the flow and the jet velocities of the OWP calculated using
jetting analysis. ............................................................................................... 125
Figure 5-3 The different stages of the detonation of CSC at different times indicating the
start of the jet breakup at 54.91 s. ................................................................. 126
Figure 5-4 Velocity vectors of the shaped charge jet indicating the velocity gradient. .... 127
Figure 5-5 Grid plot of the shaped charge jet remapped into Lagrange processor. ........... 127
Figure 5-6 OWP remapped jet penetrating the gun wall, water wellbore fluid, steel tube
casing and concrete (35MPa). ......................................................................... 127
Figure 5-7 The collapse angle at different distances from the liner apex using different
meshes. ............................................................................................................ 129
Figure 5-8 The elemental jet velocities at different distances using different mesh sizes. 129
Figure 5-9 The penetration depths into the concrete using different mesh sizes and the
relevant time consumption. ............................................................................. 130
Figure 5-10 The flash x-ray trial setup, 1: the tested OWP, 2: the aluminium foil layers, 3:
List of Figures
13
the x-ray heads. ............................................................................................... 132
Figure 5-11 The real x-ray jet and the numerical Euler jet at 34s and 122s from the
moment of detonation. .................................................................................... 132
Figure 5-12 The numerical jet velocity at different distances from the liner apex and the
real tip velocity estimated experimentally. ..................................................... 133
Figure 5-13 The experimental (upper) and the numerical (lower) penetration depths into
40MPa concrete. ............................................................................................. 134
Figure 5-14 The located fixed gauge points used to predict the surrounding medium effect
on the tip velocity. ........................................................................................... 135
Figure 5-15 The jet tip velocities at different gauges for the air and void media. ............. 135
Figure 5-16 The velocity-time histories for the three jets stretching through air and void
mediums. ......................................................................................................... 136
Figure 5-17 The dependence of jet tip velocity and Gurney velocity on the explosion heat
of explosives. .................................................................................................. 137
Figure 5-18 The relation between the jet tip velocity and the detonation velocity of the
used explosive. ................................................................................................ 138
Figure 5-19 The ratio of the jet tip velocity to the detonation velocity and the Gurney
velocity of the used explosives. ...................................................................... 139
Figure 5-20 The penetration depth of 46
o
conical OWP for different liner thicknesses with
HMX explosive charge and steel casing thickness. ........................................ 140
Figure 5-21 The penetration as a function of the jet momentum ....................................... 141
Figure 5-22 The jet tip velocities as a function of cumulative jet mass for different liner
wall thicknesses according to standard jetting analysis algorithm. ................ 141
Figure 5-23 Liner walls with varied (tapered) thicknesses. ............................................... 142
Figure 5-24 The jet tip velocity as a function of cone apex angle with uniform liner wall
thickness of 1.77mm. ...................................................................................... 144
Figure 5-25 The calculated penetration depth for OWP at different cone apex angles. .... 145
Figure 5-26 A comparison between the jet tip velocity for different apex angles at both the
same mass ratio and the same thickness. ........................................................ 146
Figure 5-27 2-D contours of the desirability with the liner angle and its thickness. ......... 149
List of Figures
14
Figure 5-28 3-D surface of the calculated desirability for the optimization problem. ...... 149
Figure 5-29 The penetration depth 2-D contours with the optimization parameters (angle
and liner thickness) using the optimum (minimum) explosive mass 26.25gm
..............................................................................................................150
Figure 5-30 3-D surface of the calculated penetration depth for the optimization. ........... 150
Figure 5-31 The three liner shapes; (a) the conical liner, (b) the trumpet liner, and (c) the
bi-conical liner, all with uniform liner wall thickness. ................................... 151
Figure 5-32 The four OWP with different explosive masses. ........................................... 152
Figure 5-33 The damaged areas around the penetration path using different explosive
masses. ............................................................................................................ 153
Figure 5-34 The OWP charge fitted inside the gun carrier and water stand-off distance
measured from gun casing wall. ..................................................................... 154
Figure 5-35 The penetration depth and hole diameter for OWP detonated at different water
stand-off distance. ........................................................................................... 154
Figure 5-36 Different fixed target points along the liner axis to predict the P-t history on
the explosive-charge interface using 8mm casing wall thickness. ................. 155
Figure 5-37 The predicted pressure and impulse-time histories for both 1mm casing
thickness (left) and 8mm casing thickness (right). ......................................... 156
Figure 5-38 Shaped charge with side point of initiation at time 0s. ................................ 157
Figure 5-39 The detonation pattern of the side initiation at 1.26s. .................................. 157
Figure 5-40 The twelve colours of the liner material used to track the liner portioning into
jet and slug. ..................................................................................................... 158
Figure 5-41 Multi-coloured copper liner of OWP of 46 deg. cone apex angle and 1.4mm
liner wall thickness at time 0s. ...................................................................... 159
Figure 5-42 The multiple-colours contours of the collapsed liner indicating the jet
formation from certain liner regions at different times. .................................. 159
Figure 5-43 The selected moving target points on the liner axis to illustrated the liner
portioning into jet and slug. ............................................................................ 160
Figure 5-44 Absolute velocity-time history plot for the nine moving gauge points used to
illustrate the liner partition into jet and slug. .................................................. 160
List of Figures
15
Figure 5-45 A schematic diagram illustrating the jet and slug potions based on the
simulation results. ........................................................................................... 161
Figure 5-46 The Gurney velocity as a function of and the (P
CJ
/I
SP
o
) relation. ................ 162
Figure 6-1 The shaped charge used in the concrete strength study (left) and a cross-section
of the liner (right). ........................................................................................... 166
Figure 6-2 The layout and the experimental test setup according to API-RP43. .............. 167
Figure 6-3 The cumulative jet mass versus the jet axial coordinate obtained from the jetting
analysis using different mesh sizes. ................................................................ 168
Figure 6-4 The numerical penetration into concrete using different mesh sizes. .............. 169
Figure 6-5 The jet generation and stretching at different times. ........................................ 170
Figure 6-6 The concrete penetration stages. ...................................................................... 171
Figure 6-7 The concrete damage contours history. ............................................................ 171
Figure 6-8 The penetrated tested witness concrete targets and the steel discs. ................. 171
Figure 6-9 The effective jet length and the time relation for virtual origin model. ........... 172
Figure 6-10 The penetration depth dependence on the concrete equivalent strength based
on Eq.(6-5). ..................................................................................................... 175
Figure 7-1 A schematic drawing illustrates the collapse process path from the initial liner
position to its axis. .......................................................................................... 177
Figure 7-2 Velocity vectors in a moving coordinate system [50]. ..................................... 178
Figure 7-3 The flow configurations in the supersonic regimes detached and attached shocks
[83]. ................................................................................................................. 181
Figure 7-4 The calculated critical angles for different liner materials at different flow
velocities [3]. ................................................................................................... 181
Figure 7-5 The flow configuration Autodyn 2-D model used to estimate the critical angle
of jetting. ......................................................................................................... 183
Figure 7-6 The relation between the critical angle and the flow velocity calculated
numerically (i.e. method 2). ............................................................................ 184
Figure 7-7 Variations of
c1
and
c
with collapse velocity. ............................................... 185
Figure 7-8 The zirconium solid cylinder (left) and the manufactured liners; 1: conical; 2:
hemispherical; 3: bell and 4: bi-conical shape. ............................................... 187
List of Figures
16
Figure 7-9 Dimensions of the test setup and the experimental test configuration ............. 187
Figure 7-10 The impact area of the jet-test layers modelled by jet solvers using three
different mesh sizes. ........................................................................................ 188
Figure 7-11 The damage contours near the impact surface for the three mesh sizes at 40s.
........................................................................................................... .......... 189
Figure 7-12 The Numerical penetration using different mesh sizes and experimental
penetration. ...................................................................................................... 190
Figure 7-13 (a) the cross-sections of the collapsed zirconium jet impacting on the
symmetrical axis at collapse angle of 12 degree and flow velocities of 3, 5 and
6km/s for cases I, II and III respectively; (b) the corresponding regions on the
c
-V
2
curve. ..................................................................................................... 191
Figure 7-14 The analytical and numerical
c
-V
2
curves for zirconium liner with the
numerical
c
-V
2
curve for copper liner as reference. ...................................... 192
Figure 7-15 and V
2
values for four zirconium liner shapes from jetting analysis, in which
different regions of zirconium jet formation and coherency are shown. ........ 192
Figure 7-16 Jet velocity profile along the liner axis with and without tip correction. ....... 194
Figure 7-17 The jets shapes for the different liner geometries right before the impact on the
test layers. ....................................................................................................... 194
Figure 7-18 The collapse velocity histories for the different liners. .................................. 195
Figure 7-19 The x-momentum histories for the four perforators with different liner
shapes..................................................................................................196
Figure 7-20 The penetration depth dependence on the concrete strength. ........................ 197
Figure 7-21 The damage contours of the concrete penetrated by the bi-conical jet at
different times. ................................................................................................ 198
Figure 7-22 The damage contours of the concrete penetrated by the hemispherical EFP at
different times. ................................................................................................ 198
Figure 8-1 The hydrodynamic jet penetration; [2]. ............................................................ 203
Figure 8-2 The relationship between the scaled density ratio and the scaled jet velocity. 205
Figure 8-3 A sketch of the designed shaped charge well perforator. ................................ 208
Figure 8-4 A sketch of the punch, the die, the ejector and the produced powder liner. .... 209
List of Figures
17
Figure 8-5 The three liners studied in the work. ................................................................ 209
Figure 8-6 The measured densities of the liner elements at different distances from the
cone apex point. .............................................................................................. 210
Figure 8-7 Dimensions of the test setup and the experimental test configuration ............. 211
Figure 8-8 Location of the fixed gauge point used to predict the density and the velocity
histories for the mesh sensitivity study. .......................................................... 212
Figure 8-9 The recorded density-time histories for the fixed gauge point using five
different mesh sizes. ........................................................................................ 213
Figure 8-10 The recorded velocity-time histories for the jet material particles moving
through the gauge point. ................................................................................. 213
Figure 8-11 The penetration depths into laminated target using different mesh sizes and the
relevant time consumption. ............................................................................. 214
Figure 8-12 (a): Density, (b): compressibility and (c): velocity of the copper jet just before
the jet tip impacts the target; (d): a picture of the recovered copper slug. ...... 215
Figure 8-13 Jet velocity and density histories recorded at a fixed gauge point. ................ 216
Figure 8-14 Jet density and velocity distributions along the jet axis for copper liner. ...... 216
Figure 8-15 The fan plot of the copper jet showing the original and the modified effective
jet length due to the presence of the laminated test layers. ............................. 217
Figure 8-16 Comparison among experimental result, numerical simulation and the virtual
origin model predictions for the penetration of jets with three different liners
.. ............................................................................................................... 219
Figure 9-1 The measured and the calculated stress-strain curves for the four zirconium test
specimens. ....................................................................................................... 224
Figure 9-2 The zirconium jet temperature contours at the moment of jet formation for the
zirconium liner with a wall thickness of 1.7mm. ............................................ 226
Figure 9-3 The velocity difference between the jet fragments for different T
L
/CD values for
both copper and zirconium jets. ...................................................................... 229
Figure 9-4 The specific breakup time (1/V
PL
) as a function of the scaled value (T
L
/CD) for
zirconium and copper [5]. ............................................................................... 230
List of Tables
18
LIST OF TABLES
Table 2-1 Explosive properties for some high explosives. .................................................. 44
Table 2-2 Penetration potential ranking of the different liner materials [31]. ..................... 49
Table 2-3 The produced jet characteristics using different liner materials [13]. ................. 50
Table 2-4 Breakup time and effective jet length for nine different copper samples. .......... 55
Table 2-5 The constants of the ZerilliArmstrong model [13]. ........................................... 55
Table 2-6 The impurity presence of the tested samples and its concentration in ppm
measured by chemical analysis [7]. .................................................................. 57
Table 2-7 The dependence of the jet breakup time on the sulphur content [7]. ................. 58
Table 2-8 The mass percentages of the OMNI powder pressed liner composition [44]. .... 64
Table 2-9 The testing configuration according to API-RP43-API target [17]. ................... 68
Table 2-10 Tunnel characteristics in both liquid and gas saturated Berea sandstones. ....... 69
Table 3-1 Explosive characteristics and Gurney velocity for some common explosives
[72]. ................................................................................................................... 82
Table 3-2 The yield strength of some liner materials [6]. ................................................... 90
Table 4-1 Input data to the code for the used explosive materials. ................................... 109
Table 4-2 The mechanical properties of liner materials [116]. .......................................... 112
Table 4-3 Input data to the code for the charge casing material [116]. ............................. 113
Table 4-4 The input parameters for P- and the polynomial EOS for the concrete targets
[116]. ............................................................................................................... 115
Table 4-5 The input parameters for the RHT strength and failure model for concrete
materials [116]. ............................................................................................... 119
Table 4-6 The input parameters for the A-36 steel material [116]. ................................... 120
Table 5-1 The jetting summary of OWP 60
o
with liner wall thickness of 1.74mm .......... 124
Table 5-2 The jet mass and its kinetic energy for different mesh sizes. ............................ 128
Table 5-3 Effect of the explosive type on the jet characteristics of 46
o
conical copper liner
of wall thickness of 1.4mm and 29.32g liner mass with 4.5 mm steel casing
thickness. ......................................................................................................... 137
List of Tables
19
Table 5-4 The jet output data and penetration results of CSC with 46
o
cone apex angle,
1.4mm liner of thickness using different filling explosive charges and 4mm
steel casing into 35 MPa concrete target. ........................................................ 138
Table 5-5 The produced jet characteristics and its penetration for 46
o
conical OWP for
different liner thickness with HMX explosive charge. ................................... 140
Table 5-6 The jet output data and the penetration results for OWP with two different liner
wall thicknesses of nearly the same weight. ................................................... 142
Table 5-7 Effect of the cone apex angle on the jet characteristics at the same explosive to
metal mass ratios (C/M = 1.069, RDX to Copper liner mass ratio). .............. 143
Table 5-8 Effect of the cone apex angle on the jet characteristics for the same 1.77mm liner
wall thickness and 6mm steel casing. ............................................................. 144
Table 5-9 The jet characteristics and penetration results of OWP of 1.77 mm liner
thickness for different cone apex angles. ........................................................ 145
Table 5-10 The jet output data and penetration results for OWPs with different liner cone
apex angles and the same explosive to metal mass ratio ( C/M = 1.069). ...... 146
Table 5-11 The input factors and their response values for the optimization study. ........ 147
Table 5-12 The input constrains, the governing limits and the response importances. ..... 147
Table 5-13 The optimum solutions and their corresponding desirability calculated by the
steepest slope optimization. ............................................................................ 148
Table 5-14 The jet and penetration characteristics of the three different shaped charge
liners. ............................................................................................................... 151
Table 5-15 The amount of the explosive and its impact on the jet and the penetration depth.
.............................................................................................................152
Table 5-16 The jetting analysis data obtained from the jetting analysis of OWP using RDX
main charge with different casing thicknesses. ............................................... 156
Table 5-17 The Chapman-Jouguet pressure, the specific impulse, the calculated and the
measured Gurney velocities and the deviation between them for various
explosives. ....................................................................................................... 162
Table 6-1 Jet characteristics based on the standard jetting analysis .................................. 170
Table 6-2 The jet tip exit velocity and the relevant effective jet length for the test layers.173
List of Tables
20
Table 6-3 The penetration results into concrete materials with different strength values. 175
Table 7-1 The condition for the jet formation and the state of its cohesion at different
collision velocities and collapse angles [83]. .................................................. 179
Table 7-2 The elemental percentage of impurities in the zirconium material. .................. 186
Table 7-3 The liner shapes and their jets characteristics. .................................................. 196
Table 7-4 The numerical and experimental penetration depths using different liner shapes.
.............................................................................................................. ......... 197
Table 8-1 The values of parameters a and b in Eq.(8-13). ................................................. 207
Table 8-2 The elemental percentage of impurities presented in the zirconium material. .. 208
Table 8-3 The mass percentage of the powder liner composition. .................................... 208
Table 8-4 The different liners and their jet characteristics. ............................................... 215
Table 8-5 The effective jet length and the jet exit velocities of the three test layers. ....... 218
Table 8-6 Comparison among experimental result, numerical simulation and the virtual
origin model predictions for the penetration of jets with three different liners..
..............................................................................................................218
Table 9-1 Shaped charge parameters related to the jet breakup for the studied zirconium
liners with different liner wall thicknesses ..................................................... 228
Table 9-2 The jet tip radius and jet breakup time for zirconium OWPs. ........................... 230
Nomenclature
21
NOMENCLATURE
ACRONYMS
AD Autodyn
API-RP American petroleum institute research procedure
bcc Body centred cubic
CD Charge diameter
CSC

Conical shaped charge
CFE Core fluid efficiency
cg Coarse grain
DW Detonation wave
ERA Explosive reactive armour
EOS Equation of state
EMP Electromagnetic pulse
EFP Explosive formed projectile
fcc Face centred cubic
HERF High energy rate fabrication
HB Brinell hardness
HD The dynamic hardness of the target material
HE High explosive
HEAT High explosive anti-tank
hcp Hexagonal close packed
JWL John-Wilkins-Lee equation of state of high explosives
J-C Johnson-Cook constitutive model
K.E. Kinetic energy
OWP Oil well perforator
OFHC Oxygen free high conductivity copper
OFEC Oxygen free electrolytic copper
PER Pugh-Eichelberger-Rostoker unsteady state jet theory
PM Powder metallurgy
PURE Perforating for ultimate reservoir exploitation
PVE Pressure-volume-energy relation
QC Quality control concrete target
Re Reynolds number
RHT Riedel-Hiermaier-Thoma brittle material constitutive model
RA Reduction of area
RHA Rolled homogeneous armours
SOD Stand-off distance
TOW Tube launched optically tracked wire-guided missile
VO Virtual origin
V&V Validation and verification
Nomenclature
22
NOTATION
A Parameter in Held crater radius formula
A Constant in explosive JWL equation of state
A Failure surface constant
a Tensile strain-rate factor
a Hall-Petch constant
a Biots constant
B Parameter in Held crater radius formula
B Constant in explosive JWL equation of state
B
max
The target Brinell hardness beneath the penetrating projectile
B
max
Brinell hardness
B
o
Grunisen Gamma Constant in Mie-Grunisen equation of state
C
o
Shock velocity
C
L
Collapse velocity limit for coherency criteria
C Explosive mass per unit area
c Speed of light in vacuum
c Sound wave velocity
D
a
Detonation wave speed
D Projectile diameter
D Compressive strain-rate factor
d
jo
Initial jet diameter
d Grid spacing
d Average grain size
E
t
Youngs modulus
E Gurney constant related to explosive
e
H
Initial Hugoniot energy
f
c
Compressive strength
g
i
The gap distance between two subsequent broken-up jet elements
g
o
Empirical constant in particulated jet penetration formulae
g
ave
The average gap distance between jet fragments
H
o
Height of interaction region between the jet and the target
H
D
Dynamic hardness of the target
I
FG
Shape index
I
sp
Specific impulse of the explosive
Nomenclature
23
k
T
The body shape factor of the target in the Alekseevskii penetration model
k
j
The body shape factor of the jet in the Alekseevskii penetration model
k Hall-Petch constant
k Number of broken-up elements
L, l Projectile length
M Liner mass per unit area
m Jet mass
m Thermal softening exponent in Johnson-Cook model
m Hardening exponent in Zerilli-Armstrong constitutive model
m
j
Jet mass
m
s
Slug mass
N Failure surface exponent
P
*
Normalized pressure
P Depth of penetration
P Pressure
P
e
The elastic pressure of the brittle material
P
s
The fully compaction pressure
P
CJ
Chapman-Jouguet pressure of the explosive
P
-
Penetration in particulated jet
P
o
Penetration at zero stand-off distance
p
H
Initial Hugoniot pressure
Q
v
Explosion heat
R Liner radius in jet radius formula
Re Reynolds number
R
o
The explosive outer radius
R
I
The explosive inner radius
R
x
The radius of the stationary surface
R
t
The target strength factor defined in Tate model
r Jet radius
r
1
, r
1
Constants in explosive JWL equation of state
r
o
Initial radius of the jet
r
c
Crater radius
S The distance between the VO and the target surface
s Constant representing the slope in shock EOS
T Temperature in Zerilli-Armstrong constitutive model
Nomenclature
24
T
L
Liner thickness
T
H
Homologous temperature in Johnson-Cook model
T
L
Liner thickness
T
t
The thickness of the perforated target in VO penetration model
T
e
Explosive thickness
t Current time
t
o
Detonation wave arrival time in Randers-Pehrson relation
t
o
Arrival time of the jet to the target surface
t
b
Breakup time of the jet
t
f
Final cratering time
U
D
Detonation velocity of the explosive
U Penetration velocity
U Shock velocity
u
p
Particle velocity in shock equation of state
V
j
Jet tip velocity
V
c
Jet cut-off velocity
V
o
Collapse velocity in jet formation theories
V
1
Stagnation velocity in jet formation theories
V
2
Flow velocity in jet formation theories
V
tail
Jet tail velocity
V
min
The minimum jet velocity for penetration
V
PL
Plastic particle velocity
Y Yield strength
Y
p
The projectile strength factor defined in Tate model
Z
o
The effective jet length in virtual origin model
Z The jet coordinate measured along the liner axis
z
o
The initial position of the jet element
GREEK SYMBOLS
Half cone angle
Porosity
Constant in the Haugstad breakup time equation
Constant in the Pack and Evan penetration formulae
Collapse angle in jet formation
Constant representing the jet spreading.
Nomenclature
25
c
Critical collapse angle
Deflection angle in jet formation
Flow stress
j
Jet dynamic yield strength
eff
Effective stresses
t
Compression strength of the target material
y
Yield stress
e Strain-rate
Strain
Explosive thickness
*
p
Normalized effective plastic strain
p
Effective plastic strain
The square root of the ratio of the target to jet density
Time constant in Randers-Pehrson relation
Constant in the particulated jet penetration formula
Viscosity coefficient
Metal to explosive mass ratio
Compressibility
Plate bending angle (=2)
Non-dimensional number
Constant in the JWL equation of state of explosives
Constant depending on the velocity gradient

The kinematic viscosity
t Time step for finite difference calculations
j

Density of jet material
Exp.

Density of explosive material
l

Density of the liner material
The angle between the normal to the detonation wave front and the liner
surface.
Abstract
26
ABSTRACT
In addition to its various military applications, shaped charges have been used in oil
industry as an oil well perforator (OWP) to connect oil and gas to their reservoirs. The
collapse of the liner material under the explosive load produces a hypervelocity jet capable
of achieving a deep penetration tunnel into the rock formation. The achieved penetration
depends on the OWP design, which includes the geometry and the material of the
explosive and the liner as well as the initiation mode and the casing of the shaped charge.
The main purpose of this research is to assess the performance of OWP with different
design aspects in terms of its penetration depth into concrete material.
This research employed the Autodyn finite difference code to model the behaviour of
OWPs in the stages of liner collapse, jet formation and jet penetration. The design
parameters of OWPs were studied quantitatively to identify the effect of each individual
parameter on the jet characteristics and the jet penetration depth into concrete material
according to the API-RP43 standard test configuration. In order to validate the Autodyn
jetting analysis, this research compared the jetting simulation results of copper OWP liners
with those obtained from flash x-ray measurements while the numerical jet penetration into
the laminated concrete target was validated experimentally by the static firing of OWPs.
Above-mentioned experiments were designed and performed in this project.
The validated hydrocode was implemented in this research to study the effects of the
concrete target strength, the liner material and the liner shape on the jet penetration depth
into concrete targets.
For the target strength, the traditional virtual origin (VO) penetration model was modified
to include a strength reduction term based on Johnsons damage number and the effect of
the underground confinement pressure using Drucker-Prager model. The VO analytical
model is also implemented in the liner material study to account for the jet density
reduction phenomena and its induced reduction of jet penetration capability. The jets
obtained from machined copper and zirconium liners and from copper-tungsten powder
liner all exhibited the density reduction phenomena. The modified VO model considers the
non-uniform distribution of jet density based on the jet profile analysis using Autodyn and
the experimental soft recovery for some tested liners. The results lead to a modified VO
penetration model including the non-uniform jet density effect.
For zirconium liner material, numerical and analytical studies were conducted for different
flow velocities and different collapse angles in order to determine the boundaries between
the jetting and non-jetting phases and whether a coherent or a non-coherent jet will form.
This study indicated that the suggested four different liner shapes (i.e. the conical, the
biconical, the hemispherical and the bell) will produce coherent jet when the zirconium is
used as OWP liner.
The validated Autodyn hydrocode is also used in this thesis to calculate the velocity
difference between two neighbouring zirconium jet fragments. The velocity difference is
related directly to the breakup time of an OWP jet, and thus, it is calculated for a range of
zirconium liners with different liner wall thicknesses. The calculated values of velocity
difference gave a clear insight for the breakup time formulae for zirconium jet in terms of
the liner thickness and the charge diameter.
Declaration
27
DECLARATION
I hereby declare that no portion of the work referred to in the thesis has been submitted in
support of an application for another degree or qualification of this or any other university
or other institute of learning.

Sign ..

Copy Right Statement
28
COPY RIGHT STATEMENT
I. The author of this thesis (including any appendices and/or schedules to this thesis)
owns certain copyright or related rights in it (the Copyright) and he has given
The University of Manchester certain rights to use such Copyright, including for
administrative purposes.
II. Copies of this thesis, either in full or in extracts and whether in hard or electronic
copy, may be made only in accordance with the Copyright, Designs and Patents
Act 1988 (as amended) and regulations issued under it or, where appropriate, in
accordance with licensing agreements which the University has from time to time.
This page must form part of any such copies made.
III. The ownership of certain Copyright, patents, designs, trade marks and other
intellectual property (the Intellectual Property) and any reproductions of
copyright works in the thesis, for example graphs and tables (Reproductions),
which may be described in this thesis, may not be owned by the author and may be
owned by third parties. Such Intellectual Property and Reproductions cannot and
must not be made available for use without the prior written permission of the
owner of the relevant Intellectual Property and/or Reproductions.
IV. Further information on the conditions under which disclosure, publication and
commercialisation of this thesis, the Copyright and any Intellectual Property and/or
Reproductions described in it may take place is available in the University IP
Policy (see http://www.campus.manchester.ac.uk/medialibrary/policies/ intellectual-
property.pdf), in any relevant Thesis restriction declarations deposited in the
University Library, The University Librarys regulations (see
http://www.manchester.ac.uk/library/aboutus/regulations) and in The Universitys policy on
presentation of Theses.

Dedication
29
DEDICATION

For my parents,

My dear father Abd Elazim Elshenawy
My loving mother Samia Elshenawy

and

My wife and children

Acknowledgements
30
ACKNOWLEDGEMENTS
First and foremost, I wish to give all the praise to Almighty Allah for giving me the
strength and time to complete this research.

I wish to express my deepest gratitude to my supervisor, Dr Qingming Li, for his constant
encouragement, inspiration, guidance, various helpful advices and comments on the thesis
during this research. He provided me with all kinds of support during my PhD study.

I would like to express my gratitude to Dr Steve Burley for his assistance, support, helpful
advices and time. My past and present colleagues in the research group have always been
an unfailing source of comfort. I would like to thank all of them for their support and help.

Special thanks go to Charlie, Ambrose and Tom, the technical staffs of the Event Horizon
Company and to Eric Palmer and technical staffs of COTEC for their help during my
experimental field testing.

This study would not have been possible without the financial support of my sponsor,
Technical Research Centre of the Egyptian Armed Forces.

My special gratitude goes to Dr. Abd Elhameed Mostafa, Dr. Gamal Abdo and Dr. Hussien
Elwany for their moral support. Besides, I cannot forget Dr Mohamed Ismail and Prof.
Ahmed Reyad for their tremendous effort in the beginning of my research career.

Finally, I would like to express my deepest gratitude to my wife, my son, my daughters,
and my brothers for their unflinching support, encouragement and love. Without them, this
would not have been possible.

Tamer Elshenawy
December, 2012
List of Publications
31
LIST OF PUBLICATIONS
T. Elshenawy, Q.M. Li, Parametric analysis of the penetration performance of shaped
charges used as oil well perforators, 41
st
International ICT-Conference, Karlsruhe,
Germany, June 29 July 2, 2010.
T. Elshenawy, Q.M. Li, Penetration performance of oil well perforators into rocks, The
First International Conference of Protective Structures; ICPS, Manchester, 2010.
T. Elshenawy, Q.M. Li, Perforator charge casing effect on the wave shape and jet
characteristics, The 9
th
New Models and Hydrocodes for Shock Wave Processes, NMH
Conference, Imperial College London, 2012.
T. Elshenawy, Q.M. Li, Design of shaped charge as oil well perforator, 23
rd
International
Congress of Theoretical and Applied Mechanics (ICTAM2012) XXIII ICTAM, Beijing,
China, 19-24 August 2012.
T. Elshenawy, Q.M. Li, Influences of target strength and confinement on the penetration
depth of an oil well perforator, International Journal of Impact Engineering, 2012, (in
press).
T. Elshenawy, Q.M. Li, A Modified virtual origin model for shaped charge jet penetration
with non-uniform density distribution, submitted to International Journal of Impact
Engineering, August, 2012.
T. Elshenawy, Q.M. Li, Performance of zirconium jet with different liner shapes,
Submitted to International Journal of Mechanical Sciences, November, 2012.
T. Elshenawy, Q.M. Li, Zirconium shaped charge jet breakup time, submitted to the
Journal of Propellant, Explosives and Pyrotechnics, November, 2012.

Chapter 1: Introduction
32
CHAPTER.1 INTRODUCTION
1.1 Research Background
In addition to its military applications such as anti-tank munitions and missile warhead,
shaped charge has been used in civilian applications. For example, oil well perforator
(OWP) based on shaped charge technology is used to connect the wellbore to the oil and
gas reservoir by creating deep holes. When a shaped charge is detonated, it produces a
hypervelocity jet with a tip velocity around 10km/s and a slug velocity around 2km/s. Due
to the velocity gradient between its tip and slug, the jet stretches until its breakup. The
hypervelocity stretching jet can produce a deep penetration into the target in front of the
shaped charge jet. For the OWP application, the shaped charge jet penetrates multi-layered
laminated targets.
The performance of the OWP in terms of its penetration depth for a given target depends
mainly on the type and the amount of the explosive used, the liner geometry and liner
material. Other parameters such as the charge confinement and the mode of initiation also
have evident effects on the jet characteristics and its penetration capability. The penetration
performance of an OWP is characterised by the static firing of the OWP against the
standard laminated targets representing the geological medium, the gun carrier, the well
steel wall and the wellbore fluid. The standard laminated target is steel/water/steel/concrete
with respective thicknesses of 3.2/17.2/9.5/1000mm according to the American Petroleum
Institute test section (API-RP43).
This thesis is aimed to identify the dominant parameters that influence the OWP jet
formation and its penetration capability into concrete target with the assistance of the
Autodyn hydrocode package. This hydrocode is validated by the experimental static firing
of OWP against the standard test configurations in field tests and by measuring some
designed liner aspects (e.g. velocity and breakup time) using flash x-ray facility. The
suitability of Autodyn hydrocode for the simulation of shaped charge jet formation and
penetration has also been demonstrated by many previous works.
Small diameter OWP (i.e. 36mm) is selected as a baseline in this study, while the design
improvements include the liner material and its shape. The design of the liner material
based on a range of new material candidates, such as zirconium and copper-tungsten
33
powder mixtures in addition to the traditional copper. The liner geometries include conical,
hemispherical, bell and bi-conical shapes.
1.2 Originality of Research
The wide usages of shaped charges in the military and civilian sectors have demonstrated
that the increase of penetration depth is one of the main objectives of the shaped charge
designers. Thus, the penetration enhancement is one of the key issues to be resolved, and
therefore, is the focus of this research.
The majority of existing research work on shaped charge penetration employed the
assumption that the jet has a uniform density with the same value as that of the solid liner
material, where the effects of the jet heating, compressibility and metallurgical changes are
usually ignored. Zernow [1] is one of the few researchers who discussed the unexplained
jet density reduction and density deficit phenomena, but he did not discuss the effect of
density reduction on the penetration depth. Thus, the virtual origin penetration model
introduced initially by Allison and Vitalli [2], where constant jet density was assumed, was
modified in this research to account for the penetration decrease due to the density
reduction phenomena. The influences of concrete strength and its underground
confinement pressure on the penetration depth are also studied in this research where a
modified Allison-Vitalli equation is proposed to include a target strength correction term.
On the other hand, Cowan and Holtzman [3] calculated flow velocities of jet elements and
the corresponding collapse angles for some common liner materials except zirconium.
These parameters are important because they give the different regions defined by
jetting/non-jetting and coherent/non-coherent boundaries for a given shaped charge liner
material. These regions were determined by analytical and numerical studies for the
zirconium liner material in this study. The numerical modeling of the jet formation was
presented to understand the features of jet formation in different regions for four zirconium
liners of different shapes. The bell, hemispherical, conical and bi-conical liners were
successfully designed and tested based on these analytical and numerical results.
Hirsh [4-5] suggested a unique way to determine the jet breakup time based on the
maximum plastic particle velocity (or the velocity difference between neighbouring jet
fragments) for shaped charge copper jet, which is a critical parameter to distinguish
continuous and particulated jets that have significant influence on the jet penetration. This
research calculated the plastic particle velocity for shaped charge jet of zirconium liners of
different wall thicknesses. A simple analytical relation between the ratio of liner thickness
34
to charge diameter and the plastic particle velocity is introduced and used directly for the
calculation of the breakup time of the zirconium jet, which demands huge computing cost
if numerical approach is employed to obtain the plastic particle velocity over the entire
range of liner thicknesses.
1.3 Objectives and methodology of research
The objectives of this study are
- To develop a modified virtual origin model with the considerations of the jet density
reduction and the target strength enhancement due to the confinement pressure effect in
order to improve the prediction of the jet penetration depth.
- To test the penetration performance of OWP using different liner materials.
- To investigate the conditions of jet formation and its coherency for different zirconium
liner shapes.
- To present an independent formula for zirconium jet, by which the plastic particle
velocity can be easily estimated and used to predict breakup time directly.
1.4 Thesis structure
Chapter 2 introduces shaped charge, the general parameters affecting its performance, the
different liner materials and their manufacturing techniques.
Chapter 3 describes the shaped charge jet formation process, the jet breakup models and
jet penetration models.
Chapter 4 introduces Autodyn hydrocode, equation of state (EOS) and material
constitutive/strength models.
Chapter 5 presents a general parametric study including the validation and the verification
of the codes used in this project.
Chapter 6 discusses the effects of concrete strength on the penetration reduction, in which
four concrete targets with different strengths are tested for the identical OWP design.
Chapter 7 studies the liner shape effect on shaped charge jet performance for four
different shapes of zirconium liners, i.e. the conical, bell, hemispherical and bi-conical
35
liner shapes. The theoretical calculations in this chapter include the jet formation and jet
cohesion conditions for the zirconium liner material. The performances of the OWP with
these zirconium liner shapes are verified against concrete targets.
Chapter 8 studies the liner material effect, including copper, zirconium and copper-
tungsten un-sintered powder mixture liners. The effect of the jet density reduction
phenomena on its penetration capability is discussed, in which the traditional virtual origin
model was modified to account for the reduced penetration depth due to the jet density
reduction.
Chapter 9 calculates the velocity difference between neighbouring zirconium jet
fragments. Parameters of the Johnson-Cook constitutive equation are obtained for a range
of strain, strain-rate and temperature, which are then used to calculate the characteristic
plastic particle velocity and jet breakup time for zirconium jet.
The thesis is concluded in Chapter 10 where the main findings from this research are
presented together with recommendations for the future study in this field.

Chapter 2: Literature Review
36
CHAPTER.2 LITERATURE REVIEW
2.1 Overview
2.1.1 Shaped charge phenomenology
The hollow charge is a cylinder of explosive with a hollow cavity in one end and a
detonator at the opposite end. The hollow cavity causes the gaseous products formed from
the detonation of explosive at the end of the cylinder to focus the energy of the detonation
products. The focusing of the detonation products creates an intensive and localized force
when it is directed against a target. This concentrated force is capable of creating a deeper
cavity than a cylinder of explosive without a hollow cavity; even though more explosive is
available in the latter case. This phenomenon is known in USA as Munroe effect, and in
Europe as the Vonfoester or Neumann effect [6], as illustrated in Figure 2-1.

Figure 2-1 A schematic drawing of a shaped charge configuration.
If the hollow cavity is lined with a thin layer of ductile metal, glass, ceramic, or any solid,
the liner may form a jet when the explosive charge is detonated.
After the detonation of a shaped charge, a spherical wave propagates outward from the
point of initiation. The extremely high pressure resulting from the explosive detonation
pushes out the liner material causing it to collide with other collapsed liner elements and
form a hyper velocity jet with high strain-rate of the liner in the range of 10
4
to 10
7
s
-1
[7].
Under this extremely high pressure, the succeeded jet will move with a tip velocity around
37
9 km/s, while the tail of the jet called a slug moves with a velocity around 2km/s as
illustrated in Figure 2-2.

Figure 2-2 Collapse of the liner and the formation of the jet [6].
When this extremely energetic jet strikes a metal target at a distance from the shaped
charge (i.e. stand-off distance), a deep cavity is formed, exceeding that caused by a hollow
charge without liner [6], as illustrated in Figure 2-3.

Figure 2-3 Different effects of shaped charge on target, (a) unlined cavity effect, (b) lined
cavity effect, and (c) lined cavity with stand-off distance.
The cavity produced in the metal plate due to the jet-target interaction is not due to the
thermal effect but due to the hydrodynamic flow of target material by extremely high
pressure.
2.1.2 Shaped charge applications
2.1.2.1 Development of the shaped charge and its wide usage in military applications
Shaped charges are extremely useful for penetrating armours or piercing barriers in the
field of military applications. It can be used as a part of torpedoes, missiles or particularly
38
as an anti-tank ammunition. Its military application started from World War II when the
so-called hollow charge projectiles were proposed.
The effect of shaped charge with unlined cavity on metallic target was firstly recognized
by Max Forester in 1883. This effect was also discovered by Charles Munroe in 1888 and
by Neumann in 1910. The effect of shaped charge with unlined cavity was called the
Munroe effect in the USA and UK and the Neumann effect in Germany [6]. Munroe used
this device to print symbols on metal plates. Moreover, he discovered that placing the
hollow charge at a distance from the target surface could increase the depth of crater in
target [6].
The US army used the lined shaped charge invented by Mohaupt [6] to produce the first
shaped charge grenades. Moreover, Mohaupt used his invention to produce rifle grenades
and mortar bombs up to a calibre of 100 mm. After the end of World War II, UK started a
development program of lined shaped charge, whereas USA produced the 2.26 in. high
explosive anti-tank (HEAT) machine gun grenade and the 75 mm and 105 mm HEAT
projectiles. Later, the machine gun round was modified to include a rocket motor and a
shoulder launcher, which was named Bazooka. In 1941, the Bazooka was firstly used by
UK in North Africa.
During the 1950s, tremendous efforts were done toward the understanding of the
phenomenon associated with the shaped charge jet formation. Analytical models were
developed to calculate the liner collapse characteristics and the penetration depth.
In 1973, when Manfred Held invented the explosive reactive armour, which can easily
defeat the shaped charge jet, tandem warheads were developed to defeat modern armour.
Each warhead consists of two shaped charges, placed one after the other. The idea was that
the first jet would make the penetration easier for the second charge by initiating the high
explosive or the reactive materials in the reactive armours or explosive reactive armour
module (ERA) [8].
In 1990, new developments of shaped charges were performed regarding the type and the
shape of liner material leading to the development of advanced warheads such as EFP
(explosive formed projectiles) installed in rocket warhead TOW (Tube Launched Optically
Tracked Wire Guided) missile.
Currently, shaped charge research continues in order to countermeasure the advanced
armours. Studies that originated in the nineteenth century and developed in the twentieth
39
century still continue, notably, torpedo applications of shaped charge rounds, multi-staged
or tandem warheads, long stand-off rounds, non-conical and non-copper liners, etc. Also,
metallurgical and chemical aspects of the liner material as well as methods of liner
fabrication remain important [6].
2.1.2.2 Civil applications of shaped charge
In addition to its wide usage in military fields, shaped charges can also be used in different
civil fields such as:
- Oil industry as oil well perforator
- Explosive ordnance disposal
- Cautious blasting and demolishing works
- Break, crack or form holes in rocks
- Earthmover in large constructions (e.g. tunnels)
- Cutting of steel tubes and railways with large diameter and wall thickness [9]
- Explosive welding
- Generation of transient antennas to countermeasure the use of electromagnetic
pulse (EMP) weapons [10].
2.1.3 Application of shaped charge in the oil field
2.1.3.1 Introduction
Oil well completion involves the drilling of a hole with the designated surface depth using
pit cutting element, which crushes the rock efficiently as it rotates and initiates fluid out to
loosen and carry out debris to the surface. When the hole reaches the designated depth, the
logging information enables the oil company to determine if the well is a producer or not.
If the well is characterized as a producer, a steel pipe is inserted back into the hole to
ensure that it is still intact and circulate mud through it to test the casing. If everything tests
positively, the pipe will be removed and the last casing pipe is inserted into the hole and
cemented.
The perforating gun is lowered into the hole to the production depth using a thin metal
cable called wireline and an electrical signal is sent down the wireline to fire the gun and
ignite the explosive charges. These charges create holes through the cement, casing and
formation connecting the well bore to the reservoir. To stimulate the flow of hydrocarbons,
sometimes it is necessary to pump air, sand and fluids under high pressure through the
perforations to increase the cracks in the formations. The remaining particles will hold the
40
cracks opened releasing the oil or gas. Once the pressure is released, the hydrocarbons are
allowed to escape and flow into the well bore.
2.1.3.2 Description of the shaped charge used as perforator
In addition to its wide usage in anti-tank warheads and explosive disposal devices, shaped
charge as oil well perforator (OWP), has been extensively used and developed in the field
of oil extraction. In 1946, Mclemore [11] firstly used the shaped charge in the field of oil
industry. According to Ref. [12], Rinehart et al revealed a design of shaped charge capable
of perforating oil well casing, well bore fluid and tubing.
As the shaped charge detonates, the detonation energy liberated from the explosive charge
detonation will collapse the metallic liner towards its axis to form a jet and a slug. The jet
which represents about 20 % of the liner mass is moving in the front with a velocity
ranging from 5000 to 10000 m/s, while the heavy tail representing about 80% of liner mass
called the slug is moving with a velocity around 1000 m/s. Due to this velocity gradient
between the jet tip and its slug, the jet length increases and its diameter decreases with both
time and travelling distance. The jet under extremely high pressure behaves much like a
fluid although it is still in its solid state. The produced jet is acting as a fluid due to severe
plastic deformation by intensive shock loads from explosive detonation, rather than due to
thermal melting although the temperature of explosion may exceeds 3000 or 4000
o
C [13],
[14].
The elongated jet has a very high kinetic energy and has the ability to create deep holes
into different hard target materials. The shaped charge used in the oil completion referred
to as OWP should perforate steel casing, wellbore fluid and cement to achieve a deep
penetration depth into rock formation to connect oil and gas reservoirs to the wellbore.
Figure 2-4 presents a schematic drawing illustrating the fitting of the OWP inside the gun
carrier, all centralized in the pipe tubing inside the rock formation.
The common features of the oil well perforators are:
- The stand-off distance is limited but a concave gap should be maintained in order to
reduce the clogging with the wall of the casing [11].
- The shaped charge designed for rock penetration should cause both large penetration
depth and large diameter hole.
41
- The OWP should be designed to withstand deep well pressures exceeding 150MPa
[12]. The housing or casing of the shaped charge is made of pulverable material
(Alumina Ceramic) having a very high compressive strength in order to resist the
high pressure in the very deep wells for oil extraction and to create small fragments
upon detonation of the OWP. Also, the charge casing design should consider the
interference between the charge case fragments and the other perforators to avoid
premature explosion.
- The selection of high explosive in the design of OWP should consider not only the
explosive performance but also its sensitivity because the temperature of the
downhole can be greater than 260
o
C [11], which is close to the ignition temperatures
of some high explosives.
- The main problem for the use of shaped charge as OWP is that the resultant useless
slug can clog the aperture in the borehole; therefore, the oil productivity can be
affected by this clog. To overcome this problem, research on powder metallurgy or
powder pressing technique has been conducted in order to create a jet with high
percentage of mass and low density porous slug from the pressed powder liner [15].
It has been shown that 30% of the clogged geological boreholes in the well
production were caused by the heavy massive slugs, which is the main factor to
affect the well productivity [15].
The most common type of gun perforator is the casing gun, in which all the perforators are
fixed on a wireline and conveyed by steel tubing, which protects perforators from impact
and isolating them from the well fluids [16]. Figure 2-4 illustrates the plan view of an
OWP fitted inside the gun with the main elements of gun carrier.

Figure 2-4 OWP fitted inside the gun carrier facing the cement and concrete materials.
42
Figure 2-5 represents a side view of down-hole gun perforator (i.e. the steel carrying the
perforators), other elements and the detonating cord required for instantaneous detonation,
as well as the well and the surroundings. The perforators are fixed inside the gun at
constant angles in a top-view plane (Phasing). These perforators are held in a hollow steel
carrier to protect them during operation. The thickness of the steel carrier is sufficient to
protect the perforators from the downhole conditions of heat and pressure even at a thin
scalloped area, through which the perforator is fired [17]. The cement layer is pumped into
the annulus between the tube casing and the bearing rock in order to prevent contamination
of the water around the well casing by the produced oil [17]. After the well is completed,
the oil and the gas are allowed to travel up to the surface, stored and refined.

Figure 2-5 Schematic drawing of the location of down hole gun perforators, steel carrier,
well fluid, well casing, cement layer, and hydrocarbon rock [17].
2.2 Factors affecting the shaped charge used as oil well perforators
(OWP)
The shaped charge geometry design and the liner thickness are the most effective
parameters governing the performance of an OWP [11]. Apart from its cone diameter;
conical shaped charge (CSC) liner performance in terms of its breakup time is governed
and controlled by the following factors [18]:
- The production method of the liner material,
- Quality of both the inner and outer surfaces of the copper liner,
- Purity and quality of the copper material,
- The adhesive material between copper and high explosive materials,
- The type of the high explosive, and
43
- The presence of air cavities in the liner material [19].
2.2.1 Stand-off distance
The shaped charge jet does not become fully formed until it has travelled a certain distance
from the target. This distance is called the stand-off distance, which is proportional to the
cone diameter. In general the optimum stand-off distance is between two and eight times
that of the cone diameter depending on the cone diameter and the geometry [16]. A proper
stand-off distance can increase the penetration depth by 50 % in comparison with zero
stand-off distance [16]. Figure 2-6 illustrates the relation between the depth of penetration
and the stand-off distance. If this stand-off distance is too large, the coherent unidirectional
jet does not exist. Instead, tumbled, deflected and particulated columns of jet are observed.

Figure 2-6 Depth of penetration versus stand-off [20].
Since the penetration depth depends on the length of the penetrator, the design of the
shaped charges used in the oil industry should consider the limited clearance between the
liner base of the shaped charge perforator and the casing wall. These perforators need to be
fitted inside the casing of the gun leaving limited free space to allow the jet to stretch [11].
Thus, the achieved penetration depth may be lower than that of the shaped charges with
suitable stand-off distances [16]. In a practical application, the stand-off distance between
the liner base and the gun wall is about 1cm while the well bore fluid gap between the gun
and the casing walls is about 2-2.5cm [11].
44
2.2.2 High explosive
Theoretically, more energetic explosive produces faster jet, greater jet kinetic energy and
deeper penetration [6]. The energy obtained from the high explosive during its detonation
is related to Gurney velocity of this explosive, which is the energy liberated from the high
explosive and transformed into mechanical work imparted to the liner element. Gurney
velocity increases with the detonation velocity and/or the detonation pressure of the
explosive which leads to the increase of the jet tip velocity. As a result, the jet kinetic
energy and its penetration potential into target will be enhanced.
Table 2-1 illustrates the explosive properties of some high explosives. It is expected that
shaped charges filled with HMX, which has the highest Gurney velocity, will produce
higher penetration depth, as shown by Tamer and Li [21].
Table 2-1 Explosive properties for some high explosives.
Parameter

H.E.
Density
(g/cm
3
)
Detonation
velocity
(m/s)
Gurney
velocity
(m/s)
Explosion
heat
(kJ/kg)
Detonation
pressure
(GPa)
Ignition
temp.
(
o
C)
HMX 1.891 9100 2960 5553 420 280
LX-14
(HMX/Estane)
95/4.5 %

1.835 8800 2800 5559 370 NA
RDX 1.730 8489 2870 4118 330 210
Cyclotol
(RDX/TNT) 75/25
1.754 8250 2790 5245 320 NA
PETN 1.720 8142 2920 5770 220 202
TNT 1.600 6913 2390 3681 210 227
The Gurney velocity for Cyclotol was obtained from [22], while the ignition temperature
was obtained from [23].
It is also known that the penetration depth of the shaped charge jet into concrete material
increases with the increase in the amount of high explosive used in the shaped charge,
which also causes the increase of the damage of the crushed region around the penetration
path [24].
The selection of high explosive in the design of gun perforator is very important for both
its performance and sensitivity issues. The temperature of the down-hole can be greater
than 260
o
C [11], which should be considered because it is close to the ignition
45
temperatures of some high explosives. Therefore, care should be taken in the design of the
main explosive charge and the degree of casing confinement.
Another important issue related to the high explosive filling of the OWP is the
manufacturing technique. The explosive density, the presence of air bubbles and cracks
inside the explosive also affect the performance of OWP. The explosive should be pressed
under vacuum to remove air bubbles and to increase its density, as shown by Renfre et al.
[12]. Moreover, shaped charges used in military purposes should be checked by flash x-ray
for air voids and cracks. Other parameters such as grain size and homogeneity of high
explosives should also be considered [6]. Moreover, it has been claimed that the shaped
charge warhead may be expected to perform much more effectively and efficiently when
the filling explosive has a particle size less than 200m [25].
2.2.3 Liner geometry
Liner is considered as the most critical element affecting the dynamic characteristics of the
shaped charge jet and its penetration capability into target materials. There are many liner
shapes, which could produce different jet charateristics. These shapes include conical,
hemispherical, Tulip, trumpet (or bell shape) and bi-conical liners [6].
The liner shape determines the characteristic of the produced jet. For example, the conical
liner produces deeper penetration with small hole diameter. On the other hand, the bell
shape liners produce shallow depth penetration with greater hole diameter [11]. In general,
the geometry of the cone is determined by the cone apex angle. If this angle is small, the jet
is long, thin and more penetrative. As the cone angle widens, the jet becomes shorter,
thicker, and less penetrative [11], as illustrated in Figure 2-7 and Figure 2-8.
Since the OWP performance is represented not only by the penetration depth but also the
crater diameter, a balance must be established between the small cone angles which
produce large penetration depth and the wide cone angles which produce large crater
diameter. Figure 2-9 illustrates the relation between jet and slug velocities and liner cone
angle.
Various improvements of the liner elements have been done in the past thirty years. In
1998, Davinson and Pratt [17] proved that modifying the liner shape design of shaped
charge can increase the jet kinetic energy by 10% and hence can improve the penetration
depth by 28%. The newly improved design includes the replacement of old conical liner of
tapered (linearly increasing) thickness with a new bell-shaped one of variable thickness
46
maintaining the same explosive mass (39g) constant. Figure 2-10 illustrates the original
and modified liner. The diameter of the baseline liner was 4.06 cm while the modified bell
liner was 4.39 cm. The improvement of the liner design is attributed to the greater surface
area of the improved bell shape which increases the absorption of the detonation energy,
and in turn, increases the collapse velocity of the liner elements leading to the increase of
the jet velocity. However, the increase in the collapse velocity attributes not only to the
greater surface area in the bell shaped liner, but also to the greater space allowing the liner
elements to accelerate than that of the conical baseline liner, as illustrated in Figure 2-10
[17].
In 2001, Lee [11] suggested a varied thickness liner of a hemisphere shape. The
thicknesses were 1.41mm at the liner base and 0.52 mm at the apex section, which has a
hole of 9.27mm diameter. The purpose of this design was to generate a double velocity
gradient jet with a bulged section capable of creating small diameter hole in the well casing
and big hole in the rock formation with maximum penetration depth in the rock layer.

Figure 2-7 Penetration versus liner cone angle [26].

Figure 2-8 Shaped charge jet profile at different cone angle [6].
47

Figure 2-9 Jet and tail velocities as a function of cone angle [26].

Figure 2-10 The original and modified liner shapes [17].
The optimum thickness of a liner has been shown experimentally to be about 2% of the
cone diameter [26], [27]. In addition, divergent profile of varied thickness liner was
designed according to desired jet characteristics. Renfre et al. [12] had performed an
experimental testing of shaped charge of varied liner thickness, in which the liner thickness
at the apex is 10-40% greater than that at the liner base. The thickness of the liner between
the apex and the liner base is tapered smoothly. This shaped charge was tested against
concrete target according to Standard API-RP43 (section II), where the resultant achieved
penetration depth was 12 inches (4.5 times the charge diameter) and one inch in diameter.
2.2.4 Detonating wave form
The velocity, the length and the cohesion of the jet depend on the manner, in which the
liner collapse, which is strongly influenced by the shape of the detonation wave (DW)
48
when it meets the liner. The DW travels inside the explosive in the form of hemispheres.
The angle between the tangent to these hemispheres and the liner defines the value of the
deflection angle, which has a great effect on the jet and the slug masses and velocities [28].
Moreover, it will determine the magnitude of the collapse angle, which is the key
parameter to determine the jet formation. In general, a more cohesive jet is formed if a
smaller angle is induced. This improvement can be achieved using a spacer (inert or active)
in the explosive. This spacer is a barrier embedded in the explosive charge between the
cavity and the rear initiation point in order to delay the DW. Such spacer has been referred
by various researchers as wave-shaper, wave-former or explosive lense [26].
Figure 2-11 shows the plots of detonation wave of two conical shaped charges (CSC)
indicating the shape of the DW as it meets the liner. The left shaped charge is without
wave shaper, while the right one has a wave shaper of a spherical shape. Smaller incidence
inclination angle is preferred for a shaped charge design improvement as it will increase
the real collapse velocity of liner elements and therefore the jet element velocities will be
increased. Further analysis of shaped charge jet formation will be discussed in detail in
Chapter 3.

Figure 2-11 The shape of the DW travelling inside CSC explosive charges with and
without wave shaper.
2.2.5 Symmetry
Any change in the shaped charge symmetry will produce a weak jet, in which curved path
and radial velocity components are observed leading to the decrease of the penetration
depth [29], [30].
49
2.2.6 Liner materials
2.2.6.1 Introduction
Recently, researchers have shown an increased interest in the different liner materials and
their manufacturing techniques. Held [31] showed different materials that could be used as
liners and their ranking according to the predicted penetration performance in terms of
liner density and jet velocity, as illustrated in Table 2-2. In general, the characteristics of a
good candidate material for shaped charge liner include [6]:
- High density
- High melt temperature
- High bulk speed of sound
- Fine grain and proper grain orientation
- Availability and cost
- Easiness of fabrication
- High dynamic strength and ductility
Table 2-2 Penetration potential ranking of the different liner materials [31].
Liner Material Al Ni Cu Mo Ta U W
Density (g/cm
3
) 2.7 8.8 8.9 10.0 16.6 18.5 19.4
Bulk sound speed (km/s) 5.4 4.4 4.3 4.9 2.4 2.5 4.0
V
jo,max
(km/s) 12.3 10.1 9.8 11.3 5.4 5.7 9.2
Jet performance
(kg/m)
0.5
/s
20.2 30.0 29.2 35.7 22.0 22.0 40.5
Ranking 7 3 4 2 6 5 1
Because tungsten has great density exceeding 19 g/cm
3
, high melting point of 3410
o
C,
high sound speed and great ductility, it has been widely used in anti-armour technology as
a shaped charge liner material [32]. However, the most commonly used liner material is
copper. It flows easily to produce a coherent jet when it is deformed by the detonation
wave. This copper material should be oxygen-free with high conductivity and low impurity
according to ASTM standard C101OO LAW F68-77 temper 070 [12]. Gold is denser and
has greater dynamic ductility than copper. In theory, it should achieve better penetration
performance than other materials [26].
In 2001, Bourne et al. [13] used zirconium, silver, titanium and depleted uranium to study
their shaped charge jet characteristics when same liner masses were used to compare the
j jo
V
max ,
50
cumulative jet length produced from these metal liners. They designed a hemispherical
liner to test the liner material performance. This design produced a jet containing 80% of
the full liner mass. The characteristics of these jets for different liner materials including
copper are listed in Table 2-3 [13].
Table 2-3 The produced jet characteristics using different liner materials [13].
Metal
Jet length
(mm)
V
tip

(km/s)
V
tail

(km/s)
Breakup
time (s)
Material
ductility
factor (Q)
Silver (Ag) 1456.0 6.48 3.01 419.6 181.90
Zirconium (Zr) 2058.9 6.75 2.34 603.8 246.10
Titanium (Ti) 1327.4 6.34 2.99 396.2 175.70
Depletted
Uranium (DU)
1700.0 6.40 3.30 548.4 217.67
Copper (Cu) 1130.5 5.90 2.56 338.5 161.53
It was concluded that silver, zirconium and depleted uranium liner materials can produce
more ductile jets than copper with larger breakup times and effective jet lengths than those
of the copper liner material. However, toxicity of the depleted uranium prevents its usage
in liner manufacturing [13].
Bimetallic liners or multi-material alloys have been fabricated and successfully tested by
many researchers. In 1996, Wang and Zhu [33] stated that the liner alloys could be
manufactured from Copper (Cu), Tungsten (W), Nickel (Ni) and Tellurium (Te). They
showed that the penetration capability of a conical shaped charge in Figure 2-12 was
increased by 30% of copper by using copper tungsten alloy liners due to the increase in
both the breakup time and the jet material density. The alloy was produced by infiltrating
technology using W-Cu alloy 80-20% weight ratios with traces of Nickel material (about
0.5%). The produced alloy liner has a density of 15.2g/cm
3
and a hardness HB number of
182. The main difference between a copper-tungsten jet and a copper jet is that the copper
tungsten jet produces fragmentation or disintegrating spray particles rather than segmented
jets in the case of copper liner. It has been shown that the copper-tungsten jet produced
larger penetration depth than copper jet especially at the short stand-off distances of 3CD,
where CD is the shaped charge calibre) under the same experimental conditions. However,
the copper-tungsten jet exhibited an anomalous behaviour at large stand-off distance (3
CD) as shown in Figure 2-13, due to the incoherency of the jet tip and the radial movement
of the fragmented elements near the particulated jet tip, as illustrated in Figure 2-14.
51
In 1998, Davinson and Pratt [17] modified the design of well perforator to increase its
penetration depth into concrete target from 105 cm to 126 cm. The improved bell shape
liner was manufactured by low cost powder pressing technique using copper and tungsten
powders with average density of 11.4 g/ccm.
In 2001, Lee [11] used a bronze (90% of copper and 10% Sn) in the lower portion of the
liner near the apex of 1 cm hole diameter; while the other part of the hemispherical liner
was made of copper. This configuration demonstrated an efficient design to generate a
bulged jet with large crater diameter into the Westerly granite.
In 2001, Glenn et al. [16] used tungsten alloy liner to study the effect of surrounding
medium on the jet characteristics and the penetration potential of such perforators. The
surrounding media was pressurized by inert gases hydrogen and helium. The compacted
powder liner consists of 45.2% tungsten (by weight), 11.05% tin, 43.19% copper, 0.53%
graphite and 0.03% lubricating oil. The produced density of the liner was 11.19 g/cm
3
. It
was found that the increase of the surrounding inert gas pressure increases the coherency of
the jet and hence increases the penetration depth.
In 2008, Bogdan and Zenon [15] used electrolytic copper (ECu) and tungsten (W) powders
to produce some liners of (ECu/W) by the matrix press moulding method. The final
dimensions of the liner were achieved by further processing, such as low temperature
sintering and machining of the pressed metal powder liners. It has been found that the
liners made from (ECu/W) exhibited a lower jet tip velocity than that of the monolithic Cu.
However it exhibited a higher depth of penetration due to its high density (12.5 g/cm
3
)
[15]. Figure 2-15 illustrates the coherent jet profile produced from ECu/W alloy at
different times.

Figure 2-12 The shaped charge in Wang and Zhu [33].
52

Figure 2-13 Anomalous behavior of Cu-W jet at large stand-off distance [33].

Figure 2-14 Variation of penetration depth with stand-off distance for Cu-W and Cu jets
[33].

Figure 2-15 The radiographs of the jet formed by ECu/W powder after 50 and 90 s from
initiation. Lengths of jets: 292 and 572 mm corresponding to instantaneous jet velocity of
7.25 and 7.0 km/s, respectively [15].
2.2.6.2 Liner material grain size
The jet cohesion, breakup time and effective jet length are the predominant governing
parameters affecting the penetration depth of a shaped charge into target material, which
depend on the grain size and crystal shape of the liner material [34]. Many papers have
been published to discuss the effect of grain size of liner material on its mechanical
properties and the validity of Hall-Petch relation over wide range of copper grain size from
nano meter to hundred micrometres. For example, Gertsman et al. [35] measured the yield
strength of copper with different grain size particles and compared the measured yield
strength with that obtained by the Hall-Petch relation:
2-1
a
o y
kd
+ =
53
where
y
is the yield stress,

d is the average grain size,
o
and k are material constants and
a = 0.5 [35].
The apparatus used to determine the yield stress was miniaturized disk bend test (MDBT).
The copper sample with micrometer grain size was produced from copper rod of diameter
2mm and 4N purity (99.99). The produced sample was 0.2mm thickness and annealed at
300-600
o
C for 30 minutes to produce different grain sizes. The nano copper was produced
by the evaporation of pure copper from tungsten boat under 1kPa pressure of helium and
then compacted under vacuum to produce a pellet of 0.3mm thickness [35].
It was found that the yield stress of the coarse grain size could be approximated by Eq.
(2-1), where
o
=92(12) MPa, k=399(61) MPa/m
-0.5
and a=0.5.
However, the classic HallPetch relation could not be applied to nano-crystal copper
because of the lattice dislocation that can move across the crystallite of a polycrystalline. It
was difficult to deduce global equation governing the dependence of yield strength on the
entire grain size range of the copper material. But, an empirical relation based on
experimental test was suggested by Gertsman et al. [35] :
y
= 104.9+111.8e
(-d/10.3)
+54.9 e
(-d/135.6)
+235.6 e
(-d/0.13)

2-2
where
y
is in MPa and d in m.
Another study of the relationship between average grain size and mechanical properties of
copper used as shaped charge liner was investigated by Meyers et al. [36]. They performed
an experimental investigation on pure copper OFHC (4N purity) in order to correlate the
relation between the average grain size of copper and the resulted mechanical strength
under severe plastic deformation [36]. The experimental work was performed by a flyer
plate of 4.7mm thickness stainless steel accelerated by PBX 9501 explosive to an impact
velocity of 2.2 km/s. The purpose of this experiment is to create exactly the same
conditions of the high pressure and strain-rate as those during the shaped charge liner
collapse mechanism. The impact pressure of the flyer plate was approximately 50 GPa,
while the pulse time duration was only 2s [36]. It has been shown that the Hall-Petch
relation is not applicable in the Nano-scale grain size [36].
In 1995, Fujiwara and Abiko [37] tested the mechanical properties of three copper samples
of 3N, 6N and 8N with average grain size of 30 m, 50 m and 100 m, respectively,
54
under strain-rate of 4.210
-5
s
-1
. It was found that both yield and maximum stresses in 3N
sample were higher than those of both 6N and 8N samples except that the ductility of 3N
(82%) is lower than that of the others (91% and 96% for 6N and 8N respectively) due to
the effect of grain size on the ductility as shown in Figure 2-16.

Figure 2-16 Stress-strain curves of 3N, 6N and 8N copper samples at strain-rate of 4.210
-5

s
-1
[37].
In 1993, Bourne et al. [34] used copper liner manufactured by shear forming in order to
investigate the effect of both grain size and texture severity on the jet length and its
breakup time. They used both Defence Research Agency (DRA) analytical model JETPEN
and flash x-ray to determine the fragmentation of shaped charge jet and particulation time
as well as effective jet length. In 1996, Renfre et al. [12] confirmed that the spinning or
flow turn machining of the copper material affects not only the liner performance during
detonation, but also the grain shape orientation, which has a direct relation to breakup time.
In Table 2-4, nine copper liner samples were used to record both effective jet length and
breakup time of the jet. The used shaped charge has a calibre of 102mm and height of
151mm. The cone apex angle is 60
o
and the liner wall thickness is 2mm. The 3mm
thickness casing is made of aluminium material [34].

55
Table 2-4 Breakup time and effective jet length for nine different copper samples.
Liner code
Average grain
diameter (m)
Breakup time
(s)
Effective jet
length (mm)
ME1A 10 195 1450-1500
IE2C 15 172 1300-1250
IE2B 20 172 1270-1280
IE1B 22 174 1248-1330
IE1A 26 182 1400-1330
E175A 42 161 1175
IE1C 43 172 1190-1350
IE2D 43 149 1000-1140
IEE1A 48 126 870-880
In addition to the well-known Hall-Petch relation between average grain size and liner
mechanical properties, ZerilliArmstrong model [38], as discussed by Bourne et al. [13],
describes the relation among the deviatoric flow stresses (), plastic strain (), strain-rate
( ), temperature (T) and grain size (d).
The general form of ZerilliArmstrong model is [38]:
1
n (-C
2
1+C
3
1 Ins )
4 5
-0.5
6
m

2-3
where parameters C
1
, C
2
, C
3
, C
4
, C
5
, C
6
, m and n are constants given in Table 2-5, d is the
average grain size in (mm) and is the strain-rate in (s
-1
).
The first term represents the effect of the thermal activation on the motion of dislocations.
The second and the third terms represent the additional stress due to the grain size effect
(i.e. Hall-Petch effect), while the last term represents the strain hardening. This equation
describes the stress-strain behaviour of the bcc (body centred cubic), fcc (face centred
cubic) and hcp (hexagonal close packed) materials.
Table 2-5 The constants of the ZerilliArmstrong model [13].
Metal
C
1
(MPa)
C
2

(K
-1
)
C
3

(K
-1
)
C
4
(MPa)
C
5

(MPa.mm
0.5
)
C
6

(MPa)
n m
Cu 980 0.0028 0.000115 46.5 5 0 0.5 0
Ta 1125 0.00535 0.000327 0 19 310 0 0.44
W 16500 0.591 0.000279 0 25.6 860 0 0.443
Mo 937 0.0036 0.000107 0 22.65 647 0 0.401
Zr 600 0.0024 0.000132 21 7.9 76 0 0.51
Ti 1100 0.00226 0.00017 54 14.86 300 0 0.5
Fe 1033 0.00698 0.000415 0 22 266 0 0.289
56
As a direct measure of the jet efficiency and its dependence on stress, strain and strain-rate,
the breakup time model developed by DERA [39] is:
t
b
=
nr
o
I
PL
-
1
o

2-4
where r
o
is the radius of the jet and
o
is the strain-rate of the jet material, V
PL
is the
maximum plastic wave velocity in the metallic liner (i.e. the velocity difference between
two neighbouring jet fragments), which is defined as:

2-5
where
o
is the original density of the liner.
Both the breakup time of the jet and the cumulative jet length (breakup time multiplied by
jet tip velocity) are inverse functions of the plastic wave velocity [13].
In a separate study, Tian et al. [40] found that changes of the liner microstructure and grain
size influence the dynamic behaviour of liner material. Hence, it affects the penetration
depth into target materials.
2.2.6.3 Liner crystal shape
For the fine crystal structure, it is expected that particulation time is longer and the
transverse movement of the particulated jet elements can be avoided [31]. It was found that
the sharpness or severity texture of the liner material has less influence than the grain size
effect on the jet breakup time and effective jet length [34].
It was also found that the crystal shape and its deformation due to manufacturing process
affect the particulation behaviour in the later time when the jet is fully stretched causing
transverse movement of the particulated elements or even jet tumbling [31]. Held verified
that the shaped charge jet is very sensitive to the small deviations of the liner structure,
which can be amplified in the stage of jet collapse and formation [31]. As a result, a
tumbling and spinning particulated jet elements around the jet axis can decrease the jet
coherency, and therefore, decrease the penetration performance of the shaped charge jet.
Moreover, the shaped charge jet undergoes a dynamic recrytallization due to large
deformation and dislocation movement of the grain [32].
d
d
V
o
PL
1
=
57
2.2.6.4 Liner impurities
Recently, researchers have shown increased interest in the effect of copper material
impurities on the ductility of the copper used as shaped charge liners. In 2003, Schwartz et
al. [7] described the dependence of cooper ductility on the total type and number of
impurity atoms. The copper used was 4N (99.99%) purity and this liner was manufactured
by cold forging technique to extrude it to a hollow cone shape. After the cold forging
process, the produced liners are annealed at 315
o
C for one hour or 400
o
C for 10 minutes to
stabilize the microstructure of sulphur doping. Table 2-6 illustrates the impurities
percentage in the tested sample in ppm, while Table 2-7 indicates the effect of sulphur
content on the breakup time of the shaped charge jet at constant grain size of 40m [7]. It
has been shown that the total number of impurities decreases the ductility of the copper
due to the segregation of the impurities at the grain boundaries [7].
In 1995, Fujiwara and Abiko [37] performed experiments on the ultra high purity copper in
order to investigate the effect of impurity presence and operating temperatures on the
copper ductility. In this study, the ultra high purity copper was produced by electronic
beam refining and vacuum melting technique. The tensile test was performed on the ultra
high purity copper 6N, 8N and compared with commercial purity copper rod 3N (99.9%)
under high vacuum of 710
-4
Pa at a strain-rate of 4.210
-5
s
-1
. The average grain size for
the three copper specimens was 30, 50 and 100 m for 3N, 6N and 8N, respectively [37].
This implies that the copper impurities have a significant effect on its mechanical
properties and performance as a shaped charge liner.
Table 2-6 The impurity presence of the tested samples and its concentration in ppm
measured by chemical analysis [7].
Impurity
Concentration
(ppm)
Impurity
Concentration
(ppm)
H 0.9 Ni 1
C 5 As 0.4
N < 0.1 Se 0.3
O 6 Ag 6.4
Si 0.2 Sb 0.3
P 0.4 Pb 0.2
S 4 Bi 0.2
Fe 2

58
Table 2-7 The dependence of the jet breakup time on the sulphur content [7].
Sulphur concentration (ppm) Breakup time (s)
3 186
4 185
7 147
2.2.6.5 Strain-rate
It has been demonstrated by Lu et al. [41] that the fracture strain of nano-crystalline copper
increases with increasing strain-rate from 610
-5
to 1.810
3
s
-1
. This may be attributed to
the creep rate and super-plasticity that have been found in the nano scale metals and alloys
at much lower temperatures. The governing deformation mechanism of the nano-scale
copper at low temperatures is the grain boundary mechanism rather than lattice dislocation
mechanism.
In Ref. [41], nano copper was produced by electro-deposition technique using electro-
discharge machining, where the produced copper has an average grain size of 20nm, a
purity of 99.993% and oxygen content of 24ppm. Two dog bone samples of the nano
copper in Figure 2-17 were prepared for the tensile test at both low and high strain-rates.
The low strain-rate test at 610
-5
to 610
-1
s
-1
were conducted using standard uniaxial
tensile Shimadzu servo-hydraulic test machine (1 kN). The high strain-rate test at 1.8 10
3
s
-1
was conducted using rotating disk-bar tensile impact apparatus [41].

Figure 2-17 Dogbone samples of the nano copper [41].
It was found from the experiments that the fracture strain increases from 15% to 39% when
the strain-rate increases from 610
-5
to 610
-1
s
-1
and increased to 55% at the high strain-
rate of 1.810
3
s
-1
.

This is different from the behaviour of cg (coarse grain) copper, in
which the fracture strain decreases slightly at higher strain-rates [41].

59
The general relation between material stress and strain-rate:
m

2-6

where m is the strain-rate coefficient. For nano copper, m= 0.036 within the strain-rate
range of 610
-5
to 1.810
3
s
-1
. For cg copper, this coefficient was 0.011 in the same strain-
rate range [41].
The nano grain copper exhibits much sensitivity of its mechanical properties to the strain-
rate because of lattice dislocation activities, grain boundary effects and high resistance to
crack nucleation [41].

Figure 2-18 The stress-strain curve at different strain-rates [41].
2.2.7 Liner manufacturing
There are many methods that can be used to manufacture the shaped charge liner element.
The manufacturing technique is determined according to the applications of the shaped
charge. For military warhead applications, high precision and accuracy liners are required,
therefore high cost precision forging and flow turn techniques are normally applied. In the
oil industry, the low cost manufacturing is the predominant feature of liner production, and
thus most liners are made by powder metal technology and low precision forging technique
[16]. These methods are briefly introduced in the rest of this section.
60
2.2.7.1 Flow turning (spinning or shear forming)
The liner plate is deformed by the turning roller against a core as shown in Figure 2-19.
The produced liners are fully annealed at 450
o
C (i.e. for copper) for one hour to reduce
strain hardening, hence improving its ductility [34].

Figure 2-19 Flow turning technique [31].
Advantages:
- The production is completed in one-step,
- Both internal and external surfaces are smooth,
- Small grain size texture is produced,
- There is lower symmetrical deformation around liner axis.
Disadvantages:
A rotational component is observed due to shear process involved in this manufacturing
process, which may result in the spinning of the jet during formation.
2.2.7.2 Deep drawing
This is a cheap method for producing small liners in large numbers. The process is shown
in Figure 2-20. It has following disadvantages:
- Different crystal structure along the liner height due to the existence of different
drawing ring zones,
- Intermediate annealing steps are required for large cones to reduce strain hardening,
which is a high cost technique.
61

Figure 2-20 Deep drawing technique [31].
2.2.7.3 Cold forging
Cold forging can produce very fine crystal structure. The final liner wall thickness is
produced by machining. The process is shown in Figure 2-21.

Figure 2-21 Cold forging technique.
2.2.7.4 Warm forging
This technique is characterized by the use of lower step distance than those used in cold
forging technique. It can produce very fine crystal structure without spin effect when the
temperature is controlled below the re-crystallization temperature of the liner material.
This technique, illustrated in Figure 2-22, is still a research topic [31].
62

Figure 2-22 Warm forging technique [31].
2.2.7.5 Hot forging
One step liner is produced by the forging of the heated material up to 800
o
C. The oxidized
layer can be removed by machining; hence, it is considered a cheap technique. However,
forging above crystallization temperature will produce coarse crystal structure of the
product, and therefore, this process could no longer be used for the high precision shaped
charge liners.
2.2.7.6 High energy rate fabrication (HERF)
A special impact machine is used, in which the liner rod is pressed at a velocity of around
20m/s, where a very fine crystal structure is produced.
2.2.7.7 Electroforming copper
The very fine grain liner is produced by the anode electron deposition of pure copper on a
polished mandrel, i.e. the electroforming technique, in which the electrolytic solution
CuSO
4
.7H
2
O (300g/l) was used. The anode material is 4N pure copper; while the cathode
is titanium bar. The substrate, on which the copper ions will be deposited is a stainless steel
conical shape molud. The surface of this substrate is mechanically polished in order to
allow the separation of the copper liner from it. The produced grains, which are columnar
shape parallel to the direction of the growth, could be finer and more equi-axial if the
substrate has a high rotational velocity. The average grain size of the copper material
produced by this technique is 1-3 m [40]. Advantages of this technique are the produced
small grain size and rotational symmetric structure. This technique could also be applied to
produce other liner materials such as Nickel and Cobalt materials. However, this method
63
has some disadvantages, such as early jet breakup time and brittle jet is produced if smooth
surface inhibitors are used. It is a very expensive technique and is time consuming.
2.2.7.8 Infiltrating technology
In 1996, Wang and Zhu [33] manufactured shaped charge conical liner by infiltrating
technology using W-Cu alloy of 80:20% mass ratio with traces of Nickel material (about
0.5%). First, the Tungsten powder with average particle size of 6m and Nickel powder of
5m average particle size were mixed together with some rubber and copper (3%) in a
mixer. The homogeneously mixed powders are pressed under high pressure (200MPa) to
form a cylinder of diameter 35mm. The pre-sintered cylinder at 1200
o
C will be infiltrated
by copper powder at 1150
o
C under protective atmosphere to form the blank, which
eventually can be machined to the required liner dimensions. The produced alloy liner has
a density of 15.2g/cm
3
and a hardness of HB number of 182. The whole production steps
are illustrated in a schematic drawing of Figure 2-23.

Figure 2-23 Sketch diagram of infiltrating technology [33].
This technique exhibits better coherent jet at short stand-off distances and longer breakup
times, hence the penetration capability of the produced shaped charge liner by Wang and
Zhu was increased by 30%, especially in the short stand-off distances (3 times calibre) due
to the increase in both the breakup time and the jet material density [33]. But, at long
stand-off distances, the penetrability dramatically decreases due to the anomalous
behaviour of the jet tip incoherency and the radial movement of the fragmented elements
near the particulated jet tip.
64
2.2.7.9 Press moulding or powder metallurgy technique
The metal powder technology is extensively used nowadays to manufacture the liner
material of the OWP in order to overcome the problem of the traditional solid liners
formed by cold working, whose slug is a carrot-like, which clogs the hole and prevents the
hydrocarbon from reaching the well bore [42]. Besides, it is a very low cost manufacture in
comparison with the traditional manufacturing techniques [24].
In 2010, Liu and Shen [43] used a Copper-Tungsten powder liner against a steel target. The
used shaped charge had a calibre of 36mm and the stand-off distance was 30mm. It
exhibited an improved penetration depth at short stand-off distances in comparison with
the traditional copper liner.
The powder metallurgy technique was also used by Bogdan and Zenon [15] to produce
liners using electrolytic copper (ECu) and tungsten (W) powders. The final dimension of
the liner was achieved by further processing, such as low temperature sintering and
machining of the pressed metal powder liners. The average grain sizes were 10 m for the
electrolytic copper and 3 m for the tungsten powders, respectively, and the used liner has
base diameter 33.3 mm and a cone angle of 45

degree. It has been found that the liners
made from (ECu/W) exhibited a lower jet tip velocity than that of the monolithic Cu.
However, it exhibited a higher depth of penetration due to its high density (12.5 g/cm
3
)
[15].
In 2001, Halliburton energy services located in Alvarado, TX, USA used this technique to
produce OMNI perforator charge liner with material composition shown inTable 2-8.
Table 2-8 The mass percentages of the OMNI powder pressed liner composition [44].
Material Copper Tungsten Tin Graphite
Mass ratio % 43 45 11 1
Function Binder Main powder Binder coating Lubricant
The maximum penetration was obtained when a 3CD stand-off distance was applied where
the total penetration was 203mm (ie: 6 CD into Rolled Homogeneous Armor (RHA)).
Advantages of powder metal liners are [44]
- Optimum performance at short stand-off distance
- Short charge length and short head height
- Low charge cost

65
Disadvantages of powder metallurgy include
- The material creeping, which is the slight expansion of the powder pressed liner,
after assembly and storage [42].
2.2.8 Applied pressure on perforators
The penetration of oil well perforators can be enhanced by at least 25% by applying high
pressure light gas atmosphere (hydrogen or helium) during the detonation of shaped
charge. This may be attributed to the gas that confines the shaped charge jet before
breakup. As a result, the breakup time increases and consequently, the effective jet length
increases [16]. Similarly, the well bore fluid pressure also influences the jet penetration.
The well bore fluid pressure mainly depends on the well drilling techniques, which can be
described as under-balnce and over-balance techniques.
2.2.8.1 Under-balance technique
The detonation of the perforators produces a crushed damage zone of low permeability and
porosity due to the production of fine particles and detonation residual debris. For many
years, much extensive work has been done in order to minimize the damaged zone area,
hence to improve the well productivity. The under-balanced pressure technique is widely
used and offers optimizing approach in the oil well. This technique involves the
implementation of static wellbore pressure lower than the corresponding rock formation
pressure. In the conventional overbalance drilling, the hydrostatic drilling fluid pressure is
designed to exceed the pressure of the hydrocarbon fluids in the rock so that fluids and fine
particles are lost to the formation [45]. These losses cause damage near wellbore area
resulting in severe reduction in the productivity of the well. In under-balncing drilling
operations, the hydrostatic drilling fluid pressure is designed to be less than the reservoir
hydrocarbon fluids pressure. Thus, there is a continuous flow of hydrocarbon fluid into the
well during the drilling process and no near well bore damage occurs, which results in
ultimate production. The difference between balanced and unbalanced perforation hole
profiles is illustrated in Figure 2-24.
The advantages of under-balance drilling technique include:
a- The prevention of damage by:
- Increasing well productivity
- Decreasing clean-up time [45].
b- The reduction of drilling problems by:
- Eliminating differential sticking
66
- Increasing the rate of penetration
c- The increase reservoir knowledge by:
- Well testing while drilling
- Identifying prolific zones
- Production steering.
d- The continuation of well production.
Moreover, the under-balance technique known as PURE or (Perforating for Ultimate
Reservoir Exploitation) not only cleans the perforation tunnel; but also produces wellbore
pressure fluctuations that can causes the effective cleanup due to the surge flow of the
liquids. Under-balanced drilling uses a variety of drilling fluids to control the bottom hole
pressure, such as water and dizel aerated with light gas nitrogen or natural gas.
Core fluid efficiency (CFE) is defined as the ratio of the steady state flow through a
perforated core to theoretical flow through a drilled hole with the same dimensions as that
of the perforation [45]. For example, applying under-balance pressure of 16.5MPa for
Berea sandstone core is capable of completely cleaning the tunnel resulting in CFE up to
0.92 [45]. The under-balance pressure of 27.6 MPa can produces a clean tunnel with zero
perforation skin and permeability from 0.01 to 100 mD (milli Darcy) in the Nugget
sandstone rock [45].
The most important factor in the PURE technique is the sharp drop of the wellbore
pressure, which results in a few hundreds oscillations a second. These oscillations produces
an instantaneous surge flow of the fluids [45].

Figure 2-24 The penetration hole; the damaged and the crushed area profiles for both the
balanced (left) and the 300-psi under-balanced perforation (right).
67
Moreover, the difference between the fluid pressure and the rock confining pressure has a
significant effect on the penetration depth (i.e. the effective stresses) as discussed by Grove
et al. [46].
2.2.8.2 Overbalance (traditional) drilling technique
In 2001, Glenn [16] found that imploding the liner of the perforator into a pressurized
(from 1500 to 5000 psi) light inert gas such as hydrogen or helium can improve the
penetration potential of the perforator into the reservoir rock by 40% compared with the
air medium. It was found that the pressurized light gas does not influence the liner collapse
but helps the jet to be confined especially in the latter stages of jet formation; hence the
total depth of penetration is increased [16].
The well completion can be accomplished by pressuring fluid into the perforated hole in
the rock reservoir to expand the cracks. In this way, the hydrocarbons can be easily
pumped to the surface and the well productivity will be increased [16].
The stand-off distance of well gun perforator is limited because of the limited space inside
the tube casing. Thus, the decrease in the penetration due to the inadequate stand-off
distance can be compensated by the technique provided by Glenn [16] when high pressure
light gas (hydrogen or helium) was applied to compensate the reduction of the stand-off
distance. The light gas applied to the perforator caused the jet to be confined to its axis,
especially in the latter stages of jet instability to improve the total penetration [16].
2.3 Oil well perforator testing according to API-RP43
The recently revised API (American Petroleum Institute) Recommended Procedures for
Evaluating Shaped-Charge Perforators (RP43, 5
th
edition), which includes procedures for
testing the penetration of gun perforating system into concrete and measuring well
productivity through core fluid efficiency [11]. This edition contains two tests to evaluate
the OWP. The first one is the Quality Control (QC) concrete target, while the second is the
API target. The test setup according to preliminary QC testing is illustrated in Figure 2-25,
while Figure 2-26 illustrates the orientation and dimension of the concrete material to be
tested according to API target. A complete description about the mechanical properties of
the layers of steel, water and API-concrete is illustrated in Table 2-9.
68

Figure 2-25 Test setup of the perforating gun [17].

Figure 2-26 A schematic diagram indicating the concrete API testing configuration [17].
Table 2-9 The testing configuration according to API-RP43-API target [17].
Layer Material Density (g/cm
3
) Strength (kbar) Thickness (cm)
Gap Air 0.0013 - 1.575
Scallop 4140 Steel 7.86 10.3 0.318
Fluid Water 1 - 1.727
Casing L80 Steel 7.86 6.2 1.151
Concrete ASTM C33-67 2.2 0.37 140
2.4 Rock material properties
2.4.1 Introduction
When the OWP is detonated in front of rock target, it forms a metallic jet together with the
explosive detonation products create a region of mechanically deformed rock with lower
permeability and porosity around the tunnel border [47]. As the jet stretches and penetrates
the rock, it loses its kinetic energy in the later stages, and therefore, the rock damage
decreases with the distance from the entrance [47]. The reservoir rock has different
characteristics that affect its resistance to the penetration by OWP. These characteristics
are the material strength related to the confining pressure, the volume fraction of void
69
filled with pore fluid in the well termed as porosity, and the permeability which is the
ability of the pore fluid to flow through the pore cracks network [47].
To study the jet penetration in sandstone saturated by gas and liquid, Karacan et al. [47]
used a gun perforator containing 6 g of HMX explosive to face a Berea sandstone core of
10.2cm in diameter and 18.5 cm in length with a porosity around 20% and permeability
ranging from 100 to 600 md. They used liquid to saturate the sandstone was a mixture of
heavy silicone oil and 1-iododo-decane, which is used as a tracer in the tomography. On
the other hand the nitrogen gas was used in the case of gas saturated cores. An under-
balance condition of 5.2 MPa was applied during the perforation flow tests, in which the
porosity and permeability tests were performed according to API-RP43 standard
procedures in order to obtain the core fluid efficiency via the measurement of perforation
tunnel dimensions [47].
It was found that both the total penetration depth and average tunnel diameter are higher in
case of liquid saturated Berea sandstone compared to that of gas saturated one. Also, the
core fluid efficiency for the liquid saturated Berea sandstone showed better flow
performance of the core, which is five times better than that of gas saturated one when
applying 750 psi under-balance as illustrated in Table 2-10. This good enhancement for the
liquid saturated cores may be attributed to the following factors, i.e.:
- The gas saturated cores have a lower cleaning-up capability of fine fragments in the
perforated tunnel, thus a lower permeability cores will be obtained,
- The drag force that can cause cleaning-up to the damaged regions is much greater in
case of liquid than that of gas because of its high viscosity,
- The liquid has lower compressibility compared to gas. Therefore, the core damage is
much lower than the gas case. As a result, the damaged rock and perforation debris
are excluded on small area of the core boundary, thus more severe permeability
damage in case of gas saturated cores is obtained.
Table 2-10 Tunnel characteristics in both liquid and gas saturated Berea sandstones.
Test layer
Average tunnel
diameter (cm)
Total
penetration
depth (cm)
Core fluid
efficiency
(CFE)
Berea (liquid saturated) 1.2-1.4 16 0.45
Berea (gas saturated) 0.5-0.7 14 0.11
70
2.4.2 Rock stresses and penetration of jet into its material
Grove et al. [46] performed some experimental testing on the stressed rocks to investigate
the effects of the confining stress and pore fluid pressure on the performance of OWP in
terms of the penetration depth into the stressed rock. Both the radial (
r
) and the axial (
a
)
stresses were considered to study the penetration of OWP into both stressed and unstressed
Berea sandstone cores. The applied pressure value was 69 MPa on the Berea sandstone of
fine to medium grain with ultimate compression strength of 55MPa, a porosity of 20% and
a permeability of 200 md. The standard saturation liquid was a brine solution containing
3% potassium chloride. The purpose of this work was to create exactly the same conditions
in down-hole rock at depth of 3km below ground level.
Following assumptions were made in test:
- The pressure in the front of the jet is composed of two components, i.e. dynamic
component, which is related to the kinetic energy density of the jet and the static
component due to the geologic stresses.
- The penetration of the perforator jet into geologically stressed rock is related to shear
strength, which depends on the confining pressure in the rock.
- The principle stresses that characterize the rock are the vertical and the horizontal
stresses. Generally the vertical component is much greater than those of the horizontal
components. This may be attributed to the greater depth of the rocks near the well
pore. At this depth, the confining pressure and the rock overburden make the rocks
much stronger than the rocks near the surface [46].
The effect of the well-bore fluid pressure on the effective stress was described by
eff
= P
c
- aP
p
2-7
where
eff
is the effective stress or the general measure of net stresses, P
c
is the rock
confining pressure
,
P
p
is the pore fluid pressure, a = 0.5 is a constant. a is noted as the
Biot's poro-elastic constant, pore pressure multiplier or ballistic pore pressure coefficient
for high pore pressure. This constant is an intrinsic property of the rock material and varies
from 0 to 1 depending on both permeability and porosity of the rock material. For rocks
with high porosity and permeability, the parameter "a" tends to be one. Eq.(2-7) reprenents
more general situations than Hallec Equation that is a special case of Eq.(2-7) when a=1.
The former gives better prediction of penetration depth than the latter [46]. Grove et al.
71
[46] suggested a model that governs the relation between the penetration depth into Berea
sandstone and the effective stress for a wide range of pore pressure, i.e.
P = C
1

2
eff
+C
2
eff
+ C
3
2-8
where P is the normalized penetration depth, C
1
, C
2
, C
3
are constants calculated from the
P-
eff
curve fitting.
In addition to the perforation tunnel caused by the jet, the local damage termed as the
cumulative energy effect of the shaped charge liner into rock, is called Mohaupt effect.
This cumulative damage results from the gaseous products evolved from the shaped charge
detonation; these gaseous products, which exhibit a high pressure and temperature together
with the resultant metallic jet will cause the cracks to extend behind the initial guide
fracture; thus forming wedging action. These cracks into rock can propagate if the tension
stresses intensity exceeds the fracture toughness limit [48].
2.5 Summary
This chapter summarized the different applications of the shaped charge devices in both
civil and military fields. The application of shaped charge OWP in the oil industry to
complete the well was introduced. The different manufacturing techniques for both the
solid- and the powder-based liners were reviewed. Moreover, the main parameters
governing the performance of the shaped charge devices were discussed.
Chapter 3: Shaped Charge Jet Formation and Penetration Models
72
CHAPTER.3 SHAPED CHARGE JET
FORMATION AND PENETRATION MODELS
3.1 Introduction
Upon the detonation of a shaped charge, three different phases are observed. These phases
are jet formation, jet breakup and jet interaction with a target. For the first phase, Birkhoff
et al. [49] proposed the steady state theory, which was developed into the well known PER
(Pugh, Eichelberger and Rostoker) theory, the most commonly used unsteady state model
[50]. Also, Godunov et al. in 1975 [51] modified the steady state theory to include the
strain-rate effect, which was further developed by Walters [52].
In this chapter, a detailed discussion about the established theories of the jet formation will
be performed. The Gurney velocity approximation from the simplest explosive-metal
configuration, which may be applied to the conical liner shaped charge, will be discussed.
Since the efficiency of the shaped charge is characterized by its breakup time and its
penetration capability, a survey about the different empirical formulae of the breakup
models and hydrodynamic penetration models will be investigated and discussed in this
chapter.
3.1.1 Steady state Birkhoff theory for jet formation
The steady state theory for the shaped charge jet formation was established by Birkhoff et
al. [49], which had the following assumptions:
- The liner elements are accelerated instantaneously to the final collapse velocity at
the liner axis with the same value V
o
,
- A constant jet length equal to the slant cone height is assumed,
- Moreover, the pressure applied to the liner wall is assumed to be equal and the
collapse angle 2 is greater than the original cone apex angle 2,
- Both the velocities and the cross-sectional areas of jet and slug are constants.
Figure 3-1 presents a schematic drawing of the steady state jet collapse process, in which
represents the collapse angle and is the half cone angle. V
o
is the collapse velocity, V
1
is
the velocity of the moving coordinates (or stagnation velocity of point A), V
2
is the flow
velocity, U
D
is the detonation wave speed of the explosive.
73

Figure 3-1 The collapse process according to the steady state theory [49].
From the trigonometric relations, the angles in triangle PAB can be calculated as a function
of and as follow:
(PBA) =

2
-
(o +[)
2

3-1
(BPA) = =

2
+
(o -[)
2

3-2
An observer at point A will feel point P approaches him by a velocity V
2
, while point A
itself moves towards right by a velocity V
1
, therefore the velocities of both the jet and the
slug can be calculated from

V
jet
=V
1
+V
2
3-3
V
slug
=V
1
-V
2
3-4
where V
1
and V
2
are the stagnation and the flow velocities, respectively. They can be
estimated according to the sin rule from:
I
o
sin [
=
I
1
sin_
2
+
(o -[)
2
]
=
I
2
sin _
2
-
(o +[)
2
]

3-5
Thus, V
1
and V
2
can be expressed as
1
v
c
cos |([-u)2]
sn[
3-6
I
2
= I
o
_
cos |([ -o)2]
tan [
+sin _
[ -o
2
] _ 3-7
74
in which the collapse angle can be calculated from
u =
I
o
cos |([ -o)2]
sin([ -o)
.
3-8
where U is the deonation wave velocity along the PP' and can be calculated by u =
u
cos o, where U
D
is the detonation velocity of the used explosive.
Thus, the jet and slug velocities can be calculated as follow:
3-9
3-10
The masses of the jet (m
j
) and the slug (m
s
) can be calculated using the mass and the
momentum conservation equations for the jet and the slug. Thus, their masses can be
estimated according to the following equations:
3-11
3-12
where m is the original mass of the liner.
This model over-predicts the jet tip velocity. The calculated jet length is greater than the
slant height of the cone (original length of the liner), which contradicts the initial
assumption that the jet length is the same as the slant hight of the cone. Moreover, the
steady state model does not consider the velocity gradient, which is the main reason for the
jet breakup phenomenon [49].
3.1.2 Unsteady state PER theory
The basic principle of this theory is the same as that of the steady state theory except that
the collapse velocity of a liner element is different from other elements depending on its
original position on the liner material [50]. The collapse velocity has its maximum value at
the apex but decreases gradually toward the base of the cone. It is assumed that the
collapse angle increases towards the liner base, therefore, the jet velocity decreases when
the new liner elements are added into the jet. PER model is illustrated in Figure 3-2 based
on an assumption of constant thickness and cone angle of the liner material [50]. In
)]
2
tan( )[csc sin(
cos

+ + = ctn
U
V
D
Jet
)]
2
tan( )[csc sin(
cos

= ctn
U
V
D
Slug
) cos 1 (
2
1
= m m
j
) cos 1 (
2
1
+ = m m
s
75
addition, the strength of the liner material may be neglected because of the extremely high
pressures on liner during the collapse process. Moreover, the formed jet has a velocity
gradient, in which the tip travels much faster than the tail leading jet elongation and
eventually breakup.

Figure 3-2 A schematic drawing of non steady state jet formation according to PER theory
[50].
The element P would have reached N when element P reached J if their collapse velocities
were identical, and therefore, QNJ remains a straight line. However, in PER model, P has
a slower collapse velocity than P, the collapsing liner has a curved contour QMJ as shown
in Figure 3-2. As a result, the unsteady state collapse angle is greater than the steady state
collapse angle
+
. This assumption is based on the assumption that each liner element is
thin and will not be affected by its neighbours. The liner element is not moving
perpendicular to its original surface but has a small (Taylor) deflection angle () with the
normal to the liner surface as shown in Figure 3-3. In Ref. [6] Richter proposed a formula
to determine the Taylor deflection angle ()
1
26
1
c
1
L
Kp
l

1
c
3-13
where
l
and T
L
are the density and the thickness of the liner wall, respectively. K and
o

are constants that are determined from the type of the used explosive and the angle of
incidence, which the detonation wave makes with the liner. T
e
is the thickness of the
explosive that drives the liner.
The geometry that shows different angles according to the unsteady state PER theory is
76

Figure 3-3 Geometry showing parameters in unsteady state theory [50].
= sin
-1
(V
o
/2U) 3-14
where V
o
is the collapse velocity of the liner element in a stationary reference system and
u = u
cos o . 3-15
The relation among the collapse velocity, the stagnation velocity and the flow velocity is
illustrated in Figure 3-4, where V
1
and V
2
can be calculated using V
o
from the sin rule as
follow:
I
o
sin [
=
I
1
sin[
2
-([ -o -o)
=
I
2
sin _
2
-(o +o)]

3-16
Thus,

3-17

3-18
where V
1
is the velocity of the moving reference (stagnation velocity), V
2
is the element
flow velocity in the moving reference and V
o
is the collapse velocity of liner elements in
the stationary reference.

sin
) cos(
1

=
o
V
V

sin
) cos(
2
+
=
o
V
V
77

Figure 3-4 Relation among collapse, flow and stagnation velocities.
The collapse angle for the unsteady state theory, can be calculated according to Hirsh
[53] and Liu [54] by
=
+
+ 3-19
where
+
is the collapse angle for the steady state theory, where
+
is given by
+
= o +2o ,
3-20

= tan
-1
_
-(xsinu)
cos(o +o)coso
_
v
o
i
v
o
__
3-21
where x is the distance along the liner axis from the apex and prime denotes differentiation
with respect to x.
Thus, the jet and the slug velocities can be calculated by:

3-22
3-23
These equations are valid in both steady state (V
o
is constant) and unsteady state (V
o

varies) theories.
The masses of the jet and slug satisfy:
3-24

3-25
)
2
cos(
2
csc

+ =
o Jet
V V
)
2
sin(
2
sec

+ =
o Slug
V V
2
sin
2

=
dm
dm
j
2
cos
2

=
dm
dm
s
78
3.1.3 Modifications to PER theory
Allison and Vitalli [55] obtained good agreement between their experimental results and
the PER theory predictions. However, Eichelberger [56] found some discrepancies
between the experimental results and predictions from PER model because PER model
assumed that the acceleration to the liner axis is instantaneous (i.e. infinite acceleration), as
shown in Figure 3-5 (a). Eichelberger [56] suggested that the acceleration of the liner
collapse is a constant, which could be calculated by Eq. (3-26) and Figure 3-5 (b). This
assumption was used by Carleone et al. [57], where the acceleration is given by
o = c
P
C]
I
L
p
I

3-26
where P
CJ
is the Chapman-Jouguet pressure of the used explosive, T
L
and
l
are the
thickness and density of the liner, respectively, and c is an empirical constant.
The more realistic equation describing the liner acceleration was recommended by
Randers-Pehrson [58], as illustrated by Eq. (3-27) and Figure 3-5 (c).

3-27
where is the time constant and could be calculated from the following equation:

3-28
wher M is the original mass per unit area of the liner and c
1
and c
2
are empirical constants.
Theoretically, the apex portion of the liner should have its maximum velocity because it
has the maximum explosive-liner mass ratio. However, this is not the case because the
liner material near the apex does not have sufficient time to reach its theoretical collapse
velocity, and therefore, the first collapsed elements do not posses the maximum velocity.
Instead, the elements that collapse after the apex elements will have the maximum
velocity. This piling up of the velocity will cause a phenomenon called the inverse velocity
gradient as illustrated in Figure 3-6.

] 1 [ ) (

o
t t
O
e V t V
=
2 1
c
P
MV
c
CJ
O
+ =
79

Figure 3-5 The acceleration of the liner element with the time, (a) infinite acceleration, (b)
linear, (c) exponential.

Figure 3-6 The inverse velocity gradient [6].
Many authors have attempted to account for the dependence of jet velocity on the shape of
the detonation wave when it meets the metallic liner. Behrmann and Birnbaum [59] and
Carleone [60] modified the PER jet formation model to consider the effect of the
detonation wave on the jet formation.
The detonation velocity of the detonated explosive along the liner surface (U) is replaced
by u
cos (x) where U

D
is the detonation velocity of the used explosive and (x) is the
80
angle between the normal to the detonation wave front and the liner surface. Therefore, the
Taylor deflection angle, Eq.(3-14) is generalized to be:
. 3-29
3.1.4 Jet elongation behaviour
Since there is a velocity gradient between the jet tip and its tail, the jet elongates (or
stretches) during its flight. To study the jet stretching behaviour, the position of each jet
element at any time is expressed in terms of its initial position (x) and time. The element
jet position at certain time (t) can be estimated by:
3-30
where z is the coordinate measured along the central axis of the shaped charge with an
origin of users choice. (i.e. the Lagrangian liner element position). t
o
is time when the jet
element just reaches the jet axis. When t=t
o
then z=z
o
. Using the above expression and
assuming the jet incompressibility, time dependent jet length can be determined [6].
3.2 The Gurney velocity approximation
3.2.1 Introduction to Gurney formulae
The model of an explosively-driven metal is used to predict the fragmentation velocities,
the flyer plate motions and the collapse velocities of the shaped charge liners. Many
authors have attempted to deduce simple relations governing the driven metal velocity
under the effect of detonation gaseous products. Gurney [61], Thomas [62] and Sterne [63]
tried to identify the chemical energy liberated from the detonation of high explosive that
could be imparted to the metal in contact with the explosive, causing its acceleration and
attaining terminal velocity [6]. This energy is called Gurney energy (E) and is considered
as an intrinsic property for each explosive. It was defined as the part of the total chemical
energy of explosive that are released during detonation and converted to the kinetic energy
of the metal. The final kinetic energy from the detonation of explosive was partitioned
between the kinetic energy of the driven metal and the gaseous product by an estimated
linear velocity profile. To simplify the calculations of the terminal velocity of the metal,
following assumptions were normally made:
- The detonation products are assumed to expand uniformly with constant density;
D
o
U
V
2
cos
sin

=
) ( ) ( ) ( ) , ( x V t t x z t x z
j o o
+ =
81
- Rarefaction and shock wave effect within the solid metals due to shock waves are
neglected;
- The total Gurney energy is divided into the kinetic energy of the gas expansion and
the kinetic energy of the driven elements in contact with the explosive.
The Gurney model was essentially based on the principles of momentum and energy
conservations. It could be applied to any one-dimensional explosive-metal interaction
system. The Gurney approximation exhibits high accuracy for the prediction of the final
metal velocity over the range of mass ratio between metal (M) and explosive (C) from 0.1
to 10 [6].
Many investigators introduced their analytical formulae used to determine the Gurney
velocity [64]. Kennedy [65] and Jones et al. [66] discussed the most well-known
configurations of metal and explosive. But for the common shaped charges, the liner
collapse velocity was determined using the formulae derived by Chou and Flis [67], Duvall
et al. [68] and Shushko et al. [69]. Hirsch [70] extended the Gurney model to small shaped
charges, whereas he also deduced another formula considering the effect of the
confinement of the charge on the collapse velocity [54, 71].
These studies show that the terminal velocity of the metal depends on the configurations of
the metal-explosive interactions, the explosive Gurney energy and the mass ratio M/C.
3.2.2 Determination of Gurney energy and Gurney velocity
Kennedy [65] provided an easy approximation to determine the Gurney energy for some
explosives. This approximation is:
E = 0.7 Q
v
3-31
where Q
v
is the heat of explosion of the explosive. However, for most commonly used
explosives, E varies between 0.61 Q
v
and 0.7Q
v
.
In 2006, Keshavarz and Abolfazl [22] extended this definition to include the detonation
gaseous and solid products and their heat of formation. The proposed new relationship
between E and Qv is similar to those calculated by the existing approximations for a range
of different exolosives. In 2002, Koch et al. [72] used the law of energy conservation to
get a relation between the explosive detonation velocity and its Gurney velocity ( ).
They applied the law of energy conservation to different explosive metal configuration,
symmetrical plates, cylinder filled with explosive core and hollow metallic sphere filled
E 2
82
with a solid explosive. For common explosives, it was found that the Gurney velocity
= U
D
/3.08, where U
D
is the detonation velocity of the explosive, which generally
agrees with existing estimations [72]. Furthermore, Gurney velocity has been
experimentally determined for certain explosives, which are listed in Table 3-1. The
average value calculated for these explosives is 3.19 with a standard deviation of 0.2,
which means that this formula presents a reliable tool to estimate the Gurney velocity for
certain explosives based on their detonation velocity.
Table 3-1 Explosive characteristics and Gurney velocity for some common explosives
[72].
Explosive Density (g/cm
3
) U
D
(km/s) (2E)
1/2
(km/s) U
D
/(2E)
1/2

Comp B. 1.717 7.89 2.35 3.36
HMX 1.89 9.11 2.97 3.07
LX-14 1.835 8.83 2.80 3.15
PETN 1.76 8.26 2.93 2.82
RDX 1.59 8.25 2.45 3.37
TNT 1.63 6.73 2.04 3.32
Tetryl 1.63 7.50 2.27 3.30
3.2.3 Formulae for different configurations
Many authors investigated the terminal velocity of different configuration models in order
to deduce a relation connecting the metal velocity as a function of the Gurney velocity and
the mass ratio between explosive and metal. In 1943, Gurney [61] deduced the well-known
formulae for the cylindrical and spherical configurations. In 1967 Henry [73] made a
complete review of Gurney approximations. In 1970, Kennedy [65] independently got the
same formulae. The different formulae for the open faced sandwich, the symmetrical
sandwich the cylindrical and spherical shells are listed below with the schematic diagram
for each configuration.
3.2.3.1 Open faced sandwich

Figure 3-7 The open faced sandwich configuration [6].
E 2
83

3-32
3.2.3.2 Symmetrical sandwich [74]

Figure 3-8 The symmetrical configuration.

3-33
3.2.3.3 Asymmetrical sandwich [74]

Figure 3-9 The asymmetric configuration.

3-34

3-35
3.2.3.4 Infinitely tamped sandwich

Figure 3-10 The Infinitely tamped sandwich configuration.
2
1
3
1 6
2 1 1
2
(
(
(
(
+
|
\
|
+
|
\
|
+ +
=
C
M
C
M
C
M
E V
2
1
3
1
2 2
+ =
C
M
E
M
V
2
1
3
1
2
+
+
=
c
N M
N
M
E
M
V
2
1
3
1
2
+
+
=
c
N M
M
N
E
N
V
84

3-36
3.2.3.5 Cylinderical shell

Figure 3-11 The cylindrical configuration [6].

3-37
3.2.3.6 Spherical shell

Figure 3-12 The spherical configuration [6].

3-38
3.2.3.7 Formulae for the Gurney approximation in the shaped charge
As the detonation wave passes the liner, it forces the liner material to collapse towards the
axisymmetrical axis of the liner, after which the stretching of the collapsed elements of the
liner start to form jet and slug. The collapse velocity is so important that it can be used
directly to predict the jet and slug velocities.
The Gurney equations mentioned previously are much simplified and cannot be applied to
conical and other liner geometries. Therefore, several Gurney velocity approximations
have been derived for shaped charge analysis as discussed in details by Walters and Zukas
[6]. One of these approximations was developed by Chou and Flis in 1986 [67], where they
summarized the different models used to calculate the terminal collapse velocity starting
from the general form of:
2
1
3
1
2
+ =
C
M
E
M
V
2
1
2
1
2
+ =
C
M
E
M
V
2
1
5
3
2
+ =
C
M
E
M
V
85
3-39
where is the M/C ratio and f() is a function of depending on the metal and explosive
configuration geometry. They showed that the collapse velocity V
o
of the shaped charge
liner can be estimated by the approximation used to calculate a single flat plate backed by a
slab of explosive:
.
3-40
For the same flat plate, Duvall et al. [68] used hydrodynamic theory to get a formula
provided that the detonation wave propagation is in a direction tangent to the liner. They
obtained the following equation by hydrodynamic theory:
I
o
= u
_1 +
27
16
_1 -
_
1 +(
S2
(27p)
,
)]_ . 3-41
In 1972, Deribas [75] concluded a similar equation:
.
3-42
Kleinhanss [76] presented an empirical equation, which, however, does not account for the
curvature in geometry such as shaped charge liner, i.e.,
o
-1
2
3
.
3-43
Kleinhanss [76] also showed experimentally that the liner collapse velocity depends upon
the radius of the cylinder explosive charges, i.e.
I
o
= u
_
r
-e(2r
-e)
r
-e
.
1
|C
o
+e(b)]
_ 3-44
where is the metal liner thickness, b=r
o
-r
i
is the explosive thickness, r
0
and r
i
are the outer
and inner explosive radii, repectively. C
o
and f(b) are empirical parameters that depend on
the used explosive and metal liner.
Chou and Flis [67] reused Gurney model to analyze the liner element velocity under the
detonation of explosive cylinder. They included the impulse generated from the explosive
into the velocity equation, i.e.
) ( 2 f E V
o
=
5 . 0
2
]
1 5 4
3
[ 2
+ +
=

E V
o
( )
( )
(
(
+ +
+
=
1 27 / 32 1
1 27 / 32 1
2 . 1
D o
U V
86
o
o
I
3-45
where I is the specific impulse, which could be obtained from 2-D hydrocode, R
o
and R
I

are the outer and inner radius of the explosive, respectively, as illustrated in Figure 3-13.

Figure 3-13 A schematic diagram of the collapsing liner under explosive load [67].
3.3 Visco-plastic model and jet coherency
The visco plastic model according to Godunov et al. [51] assumes that the jet flow is an
incompressible flow and its stress is dependent upon the strain-rate and viscosity
coefficient i.e.
o = o
+pe 3-46
where is the stress,
y
is the yield stress, e is the strain-rate and is the dynamic
viscosity coefficient. Many experiments have been conducted in USSR [77], [78] and USA
[52], [79] in order to determine the values of , which depends on temperature, pressure
and strain-rate. It was found that the dynamic viscosity coefficient ranges from 10
2
to 10
6

poise (i. e. g/cm.s).
Godunov et al. in 1975 [51] calculated the strain-rates for certain shaped charge liners and
found that they are in the range of 10
6
s
-1
. However, Bauer and Bless [80] showed that the
strain-rate in shaped charge jet is in the order of 10
4
s
-1
based on the measurements of
copper cylinder deformed at extreme loading due to explosion. In addition, Chou and
87
Carleone [81] calculated the strain-rate for a 81mm calibre shaped charge, which was
found to be 10
5
s
-1
for copper liner.
The visco-plasticity is an important model to describe the jet coherency. The coherent jet
should have no radial velocity component. Otherwise, after certain travelling distance of jet
tip, some particles start to deviate from its travel axis, decreasing the effect of jet on the
target. This incoherent jet is called overdriven or diverging jet and could happen when
Reynolds number is greater than 2 according to the following equation based on a visco-
plastic model proposed in USSR [6]:
Rc =
I
L
I
2
sin
2
[
v(1 -sin[)
3-47

3-48
where T
L
is the liner wall thickness, is the kinematic viscosity, is the jet viscocity, is
the collapse angle and V
2
is the inviscid flow velocity. This equation could be used to
determine the critical flow velocity of the jet.
The stagnation velocity V
1
is given by:
I
1
= u
sin([ -o) -sin([ -o -)

sin[
3-49
where U
D
is detonation wave speed, is the half of the conical liner apex angle, = 2 is
the plate bending angle. The flow velocity V
2
is given by:
I
2
= u
sin(o +) -sino
sin[
. 3-50
The jet and slug velocities can be obtained according Eqns.(3-3) and (3-4), respectively.
The visco-plastic model predicts a lower jet velocity than that calculated by PER theory,
however, its slug velocity is higher than that of PER theory [6].
To characterize the cohesion of the jet, Walsh et al. [82] concluded that, the jetting occurs
only when the fluid is incompressible or if the flow velocity is subsonic [6]. Besides, Chou
et al. [83] presented the criteria conditions for jetting formation and jet cohesion i.e.,
a. The flow velocity is subsonic, then a solid coherent jet can always be formed,
=
88
b. In the supersonic regimes, the jetting occurs only if collapse angle, is greater than
the critical collapse angle,
c
and incoherent jet is produced.
c. In the supersonic collisions when is lower than
c
, the jet will not be formed.
As a general rule, for shaped charge with a copper liner, the produced jet will be coherent
if the flow velocity of the liner is below critical Mach number 1.2 for the copper material
[83].
3.4 Breakup time models
Since shaped charge jet elements have a velocity gradient from its tip to its tail or slug, the
shaped charge jet breaksup into small elements at large travelling distances. In the breakup
stage, the penetration efficiency of the shaped charge starts to decrease steadily due to the
decrease of the effective jet length prior to impact and the presence of the air gaps between
the jet segments or particles. Therefore, the understanding of the jet breakup phenomenon
and the methods of delaying the onset of jet breakup are the major interests of the shaped
charge designer. Recently, many investigators studied this phenomenon, e.g. Cowan [84],
Hirsch [4], and Hennequin [85]. They used empirical formulae, hydrocode simulations and
one-dimensional analytical models to determine the jet breakup time.
3.4.1 Empirical formulae
A few empirical formulae are presented herein. Hirsch [4] suggested a phenomenological
formula for the jet breakup time. This formula calculates the breakup time (t
b
) as
b
d
]c
v
PL
, 3-51
where d
jo
is the initial diameter of jet element when the elongation starts and V
PL
is the

characteristic plastic velocity or the velocity difference between successive fragments.
Hirschs formula for breakup time based on Eq. (3-51) and the liner geometry is
t
b
=
1
I
PL
8RI
L
sin_
[
2
] 3-52
where T
L
is the original liner thickness, is the collapse angle and R is the radius of each
element from the liner axis.
Hirsch [5] further used SCAN code and a set of experiments with charges of varying liner
thicknesses to study the breakup time, in which V
PL
was found to be a function of liner
89
thickness and charge diameter. 1/V
PL
was named as specific breakup time of the liner and
was given by:

3-53
where C
D
is the charge diameter and T
L
is the liner element thickness.
Eq. (3-53) predicts reasonable jet breakup times for certain shaped charges. However, its
application is limited because it is independent of the stretching rate of the formed jet
[86].
Pfeffer [87] deduced an empirical equation to determine the jet breakup time using the
results of a two-dimensional hydrocode simulation of jet stability, i.e.
t
b
=
1.4
s
+8.
c
C
c

3-54
where is the initial strain-rate of the jet material, r
o
is the initial jet radius, and C
o
is the
sound speed in the jet material. This model is also limited because it is independent of the
jet strength [6].
Haugstad [86] presented an empirical formula based on Eq. (3-46). For the case where
tends to zero, the breakup time equation can be predicted by

b
ud
c
p)
1
s
,
3-55
where is an empirical constant, is the liner material density and d
o
is the initial
diameter of jet.
For >>0, the empirical equation has the following form:
t
b
= [
p
o
-
1
e
3-56
where is an empirical constant.
Haugstad [86] pointed out that the exact determination of the jet breakup time might not be
obtained by this formula since it does not account for material microstructure. Haugstad
[86] also concluded that the increase in the breakup time of a shaped charge jet could
correlate to the decrease in
y
and the increase in .
D
L
PL
C
T
V
49 . 101 886 . 13
1
=
90
Chou and Carleone [88] deduced a formula predicting the breakup time. The formula had
good agreements with the experimental measurements and has the following expression:
t
b
=
r
o
C
o
_S.7 -.12
r
o
o
C
o
+
C
o
o
r
o
_ 3-57
where C
o
=

o
p,
y
is the yield strength,
o
is the initial deformation rate of the jet
material, which equals to V
PL
/l
jo
, where l
jo
is the initial length of the jet material and V
PL
is
the velocity difference between neighbouring jet fragments and r
o
is the initial jet radius
when the jet elongation starts. Table 3-2 lists some values of jet yield strength for some
common liner materials.
Table 3-2 The yield strength of some liner materials [6].
Liner material Jet yield strength:
y
(MPa)
Copper ETP 200
Copper OFHC 270
Aluminum 100

Hennequin [85] developed a new formula providing a better estimation of the jet breakup
time and the accumulated length of the jet. To determine the time t
b
, only one parameter is
needed, which is the fragment shape index I
FG
that can be calculated by the initial jet
length and the final length of the fragment. This parameter characterizes the type of jet
material in terms of ductility or brittleness. It was taken to be 1.46 in Hennequin
experiments [85]. The equation has the form:

b
2
c
v
PL

1
c
v
PL
2
c
3-58
The comparison among the above mentioned equations shows that the breakup time of jet
predicted by Eq.(3-58) is similar to that predicted by Eq.(3-57), but the former gave a more
accurate results than the latter [88].
Held [18] defined the average breakup time of several shaped charges using flash x-ray to
calculate the fragments length. He defined the breakup time by
t
b
=

I
,o
-I
,c
, 3-59
91
where is the summation of broken-up jet elements, V
j,o
and V
j,cut
are the velocities of
the jet tip (V
jo
) and the cut off element (i.e. the velocity of last penetrating element),
respectively.
Held also described the quality of the copper liner material by the scaled breakup time,
which is defined by

bs

3-60
where t
bs
is the scaled breakup time and C
D
is the charge calibre or liner diameter at the
base.
3.4.2 Hydrocode Simulations
Using the Lagrangian code HEMP, Chou and Carleone [89], and Karpp and Simone [90]
determined the jet breakup time by following the jet profile changes with time. Karpp and
Simone [90] demonstrated that a jet with a uniform initial radius under continuous
stretching eventually developed necking, which depends on the wavelength of the initial
surface perturbation. They also estimated the strength of copper under dynamic conditions,
which was found to be 0.1 GPa. This is needed for the prediction of the jet breakup time.
Chou and Carleone [89] used the same code to predict the effects of the yield strength, jet
density, the initial disturbance wavelength and its amplitude on jet breakup time. They
showed that the perturbation in jet strength or velocity causes plastic instability. The
critical wavelength seems independent of the perturbed physical quantity that initiates the
instability. Figure 3-14 shows a comparison of their hydrocode simulation with the flash
radiograph of a typical jet.
In 1981, Miller [91] used the two-dimensional hydrocode named PISCES in order to
predict the breakup time of copper jet. Miller used Steinberg-Guinan constitutive equation
in the hydrocode in order to account for the strain hardening and thermal softening by
considering the effect of temperature, pressure and large plastic strain. Miller [91] used
unconfined BRL-105mm diameter 42
o
conical shaped charge to compare the experimental
results with the numerical results. It was found that the predicted breakup time for the
subsequent elements is longer than that calculated by the hydrocode. Later in 1982, Miller
[92] concluded that the difference between the hyrocode simulation and the experimental
test was attributed to the random necking of the jet due to imperfect formation of the jet.
92

Figure 3-14 Comparison between hydrocode simulation of jet necking due to instability
and flash radiograph of a jet at approximately the same time [89].
Osborn used the two-dimensional code named TOODY to study the jet breakup problem
while Pfeffer simulated the jet breakup using the STRESS-2 to study the same phenomena
[6].
3.4.3 Analytical Models
In 1976, Chou and Carleone [89] studied the jet breakup phenomenon using a one-
dimensional model. They focused on the influences of jet material strength and its inertia
force and showed that the ratio of jet flow stress to its material density controls the growth
of the instability. They predicted that the breakup time increases with the decrease of this
ratio.
The one-dimensional model was extended by Carleone and Chou in 1977 [93] in order to
include the stress concentration at the jet necks. The solution of their theory showed that
the critical wavelength was independent of the jet-stretching rate, where it had a value of
2.22 in terms of the jet diameter at the beginning of the jet instability. However, a two-
dimensional hydrocode simulation by Carleone and Chou [93] predicted that the critical
wavelength was a function of the jet-stretching rate. The correct number of jet segments
was also determined from the two-dimensional solution.
In 1982, Miller [92] also developed a one-dimensional model to study the jet necking
problem. The model was based on the separation of variables and Fourier integral
93
technique. Miller [92] assumed that a long and nearly cylindrical jet with a small neck at
the centre of the jet has a linear velocity gradient. A perfectly plastic constitutive equation
was used. Although the predicted results were in good agreement with experiments, the
initial material conditions such as temperature and flow stress had to be assumed. The
results obtained by the one-dimensional model in [92] using perfectly plastic constitutive
equation were similar to that obtained using Steinberg-Guinan constitutive equation [6].
Walsh [94] developed an analytical model to perform a detailed analysis of the effects of
surface roughness, the non-uniform initial velocity gradients and the non-uniform yield
strength on jet breakup. The predicted results were similar to those of Carleone and Chou
[93]. The model predicted that the breakup time was only mildly dependent on the
amplitude of the initial disturbance. Walsh [94] also concluded that the jet breakup time
could be delayed by reducing the shaped charge fabrication tolerances or increasing the
homogeneity of the shaped charge elements.
In 1993, Backofen [95] used an analytical model to calculate the different parameters of
the produced jet. These parameters include the virtual origin, the breakup time and jet
penetration capability into different target materials. It was found that the used analytical
model gives reasonable results to the breakup time estimated by using flash x-ray during
the detonation of the shaped charge.
3.5 Shaped charge jet penetration models
Shaped charge penetration models were initially proposed based upon Bernoulli equation,
which was subsequentially modified to account for jet particulation, compressibility and
strength effect. This section will discuss the fixed and variable velocity jets and the effects
of stand-off distance, target material strength and breakup time on jet penetration.
3.5.1 Uniform velocity jet
As the first approximation, the Bernoulli equation assumes that the jet is an inviscid and
incompressible fluid. The pressure generated during the impact of jet on the target is much
higher than the yield strength of many materials. The penetration process is also assumed
to be a steady state process and the length of the jet is assumed to be constant [6].
Diagrammatic schemes of the initial length of jet (l) prior to impact and the jet penetration
into target are shown in Figure 3-15.
94

Figure 3-15 Hydrodynamic penetration of jet into target [6].
The hydrodynamic model assumes that pressure on both sides of the interface is in
equilibrium, i. e.
1
2
p
(I -u)
2
=
1
2
p
1
u
2
.
3-61
The consumption of the jet (
-dI
d
) can be linked to the penetrator velocity and the
penetration velocity U, i.e.
I -u = -
t
.
3-62
The relation between penetration depth (P) and penetration velocity is
dP
d
. 3-63
Therefore,
dP
d
= -_
p
]
p
.
dI
d
, which can be integrated from the start of impact (i.e. l=L at
t=0) to the moment when the jet is completely consumed (i.e. l=0 at t=t
f
),
P = -
p
p
1
. =
p
p
1

0
L
. 3-64
Eq.(3-64) gives a first order approximation of the shaped charge jet penetration. However,
the hydrodynamic model neglects several important factors, which may influence jet
penetration. These factors include [6]:
- Strength of both jet and target materials, which become important when the jet
velocity is low.
- Secondary penetration due to the crater inertia after the jet is completely consumed,
95
- Velocity gradient in the jet, which requires suitable stand-off distance to maximize
the jet penetration,
- The jet tip velocity affects the penetration depth and
- Other factors such as jet-target interactions, jet alignment, jet compressibility,
aerodynamic drag and variable-area of jets.
3.5.1.1 Jet breakup effect
To account for the jet breakup effect, Evans in 1950 [6] developed the hydrodynamic
theory by applying the dynamic pressure produced by jet particles, which is the jet force
(rate of the change of jet momentum with time) divided by the total cross sectional area of
the jet. This term equals to the pressure generated in the target during the impact.
Therefore,
p
(I -u)
2
= p
1
u
2
3-65
where p

is the average jet density including gaps between the particles and

is a
constant, which equals one for a continuous jet and two for a particulated jet. Thus, the
Bernoullis equation primary penetration is given by:
P = _
p
p
1
_
0.5

3-66
This model also neglected those factors ignored in the hydrodynamic model for fixed
velocity.
3.5.1.2 Standoff distance effect
Another model had been suggested by Birkhoff et al. [49] in order to account for the stand-
off distance effect on the penetration depth. The suggested semi-empirical formula for the
continuous jet is:

3-67
where S is the stand-off distance, P
o
is the

penetration at zero stand-off (i.e. S=0), is a
constant depending on the jet velocity gradient and is a constant representing the jet
spreading. Constants and can be determined from curve fitting of the penetration-stand-
off curve.

S
S P
P
o
+
+
=
1
) 1 (
96
For the broken-up jet, the penetration formula is:

3-68
where P
o
is the

penetration depth from the previous equation (P) for the continuous jet
and is a constant depending on the velocity gradient.
3.5.1.3 Target material effect
Many investigators attempted to improve the accuracy of the simplified theory. Pack and
Evan [96] considered the importance of the strength of target material on jet penetration.
They modified Eq.(3-64) by multiplying it with a correction term as follows:
P =
p
p
1
_1 -
o
p
I
2
_ 3-69
where is a constant and Y is the yield strength of the target material. They showed that
for steel, the correction term op
I
2
is 0.3, which means that the penetration is reduced
by 30% due to the effect of the target strength.
Eichelberger [97] made extensive measurements of jet penetration histories. It was shown
that the hydrodynamic formulae Eq. (3-64) could not be used in the later stages of
penetration as the jet velocity decreases. The later stage penetration mechanism is
somehow different from the earlier stage penetration mechanism, during which both the
target and the jet behave hydrodynamically. It was also found that when the jet was
brokenup, in Eq. (3-66) would be less than unity if
j
was taken as the same as the
original density of the liner material. Thus, Eichelberger [97] proposed a formula that
includes the strengths of jet and target materials, i.e.
1
2
p
(I -u)
2
=
1
2
p
1
u
2
+o
3-70
with =
T
-
j
3-71
where
T
and
j
are the resistances to plastic deformation for target and jet materials,
respectively. Each resistance term was taken as one to three times the value of its static
uniaxial yield stress [97]. The importance of the strength term effect on the penetration was
demonstrated by Pugh [98] and Klamer [99] when they found that the penetration into
armoured steel is 15% ~ 20% less than that into mild steel.
S
S
P P
O
+
+
=
1
) 1 ( 2
'
5 . 0
97
3.5.2 Variable velocity jet
The jet has non-uniform velocity distribution and its length is increasing with time,
therefore, the fixed length jet should be modified. Abrahamson and Goodier [100]
extended the hydrodynamic penetration model to include non-uniform jet velocity
distribution and stand-off distance. This model strated from an arbitrarily selected initial
time and required the initial jet length at this moment to be given, which makes this model
difficult in practical use. Allison and Vitali [2] developped a penetration model based on
following assumptions:
a. Existence of a virtual origin, from which each jet element is emitted at its own
velocity that remains constant during its travelling between the virtual origin and
target.
b. Negligible strength of the jet and target materials: To ensure the validity of this
assumption, a minimum jet velocity for penetration, V
min
, must be defined to
represent the termination of penetration by slow moving jet elements.
c. Negligible compressibility of the jet and target materials,
d. Simultaneous breakup of the entire jet, and
e. That each broken jet segment penetrates as a continuous jet.
Allison and Vitali [2] derived the following penetration equation for the continuous jet:

3-72
where t
o
is the time at which the jet moving with a tip velocity V
o
arrives the target and
is the square root of the density ratio, i.e .
.
3-73
Eq.(3-72) is only valid before the jet breaks up because the penetration depth predicted by
Eq.(3-72) is independent of stand-off distance. In reality, penetration decreases with the
increase of stand-off distance after the jet is broken.
Allison and Vitali [2] model is still a useful model for the study of jet penetration. Dipersio
et al. [101] and Schwartz [102] presented explicit formulae based on Allison and Vitali [6]
model, for the following three cases:
a) Penetration before jet breakup (Tt
b
)
b) Jet breakup during penetration (t
o
t
b
T)
o o
o
o
V t
t
t
t V t P =
+1
) ( ) (

j
T
=
98
c) Jet breakup before reaching the target (t
b
t
o
T),
where T is the total time at the end of penetration, t
b
is the jet break-up time and t
o
is the
time when the tip reaches the target.
For case (a), the total penetration depth P is:

3-74
where S is the distance from the virtual origin to the target and V
min
is the minimum jet
velocity for penetration.
For case (b), the depth of penetration is
P =
( +1)(I
o
t
b
)
1
+1

+1
-I
mn
t
b
-
3-75
Finally, for case (c), the depth of penetration is
.
3-76
Eqns. (3-74 to 3-76) are called DSM (Dipersio, Simon and Merendino) model, which can
also be used to obtain the exit velocity V
P
of a continuous jet after perforating a finite
thickness (T
t
) [103] i.e.
I
P
= I
o

+1
. 3-77
It was found that at larger stand-off distance, these formula, give larger penetrations than
the measured values due to the occurrence of asymmetric wavering of the jet, which is not
considered in the model [103]. This over-prediction may also be caused by the occurrence
of the tumbling or the deceleration due to air friction [103].
3.5.3 Particulated jet
For the particulated jet, the gaps between the resulted jet elements decrease the penetration
capability of the jet. The penetration depth of a particulated jet, P, can be calculated using
a formula developed by Carleone et al. [6]:

3-78
(
(
(
|
|
\
|
= 1
1
min
V
V
S P
o
( )
b o
t V V
P
min
=
|
|
\
|
=
o
i
g
g
P P 1
'
99
where g
i
is the gap distance between two subsequent broken-up jet elements
nondimensionalized with respect to the increment of the jet length, g
o
is an empirical
constant equals 6.5 for precision shaped charges (relatively large calibre) and 4~6 for small
charges (non-precised). An average value of the gap distance g
ave
, may be used to replace
g
i
in Eq. (3-78);

3-79
where K is the number of broken-up elements.
Hence, to calculate the jet penetration at a given stand-off distance, it is necessary to
consider the jet traveling time and to check if the the jet is continuous or broken-up. If the
jet reaches the target after it has been broken-up, its penetration is first calculated using the
same procedures for continuous jet and then it is corrected using the Eq. (3-78).
3.5.4 Target strength
Various penetration models have considered the target strength effect on the penetration
depth. The most frequently used model for the shaped charge jet penetration was suggested
by Alekseevskii [104] and Sanasaryan [105] based on the modified Bernoulli equation [6]:

3-80
where HD is the dynamic hardness of the target material (Vickers hardness), k
T
and k
j
are
the body shape factors of the target and jet respectively, Alekseevskii [104] takes both of
them as 0.5 and
SD
is the dynamic yield stress of the jet material.
The cutoff velocity was calculated from Eq. (3-80) when U=0, therefore,
.
3-81
Christman and Gehring [106] developed a model with the consideration of four different
penetration phases according to the generated pressure, i.e. the transient, the primary, the
secondary, and the recovery phases. This model was applied to long rod penetrator in the
velocity range of 2-6.7km/s, i.e.,
P
= _1 -
] _
p
p
p
1
]
0.5
+2.2 _
] _
p
p
p
1
]
23
_
p
1
I
2
B
mux
_
13

3-82
where B
max
is the Brinell hardness of the target material. The first term represents the
hydrodynamic penetration in the primary phase, while the secondary penetration phase is
K
g
g
k
1 i
i
ave
=
=
2 2
) ( U V k U k HD
j j SD T T
+ = +
j j
SD
k
HD
V

=
min
100
represented by the second empirical term. This model was modified by Doyle and
Buchholz [107] to predict the penetration of EFP (Explosively formed projectiles). The
modified formula is:
P
= _1 -
] _
p
P
p
1
]
0.5
+
.1S
] _
p
p
p
1
]
13
_
E
1
B
mux
]
13

3-83
where P is the target penetration depth, L is the projectile length, D is the diameter of
projectile.
P
and
T
are the densities of the penetrator and target materials, respectively. E
1
is the energy in last part of projectile (Joule) and B
max
is the Brinell hardness of the target
(kg/mm
2
).
Semi-hydrodynamic model with the consideration of material strength was presented by
Alekseevskii in 1966 [104], and Tate in 1967 [108].
The dynamic equation of the projectile is
3-84
and the geometrical relationship is
I -u = -
t
,
3-85
where L is the current length of the rod. The explicit solution of this model was suggested
by Tate [109], where the values for
p
and
T
are taken to be the Hugoniot elastic limit of
the penetrator and 3.5 times that of the target material, respectively. However, Walters et
al. [103] suggested different value for
T
to be 2.5 times the Hugoniot elastic limit and
p

to be the uniaxial yield strength of the penetrator, which led to a good agreement with
their experimental work.
Tate 1986 [110] has provided methods to estimate R
t
and Y
p
based on dynamic yield
strength [111].
1
2
p
P
(I -u)
2
+
P
=
1
2
p
1
u
2
+R

3-86
where Y
p
and R
t
are the projectile and target strength factors and are defined by Tate [112]
to be:

3-87
p p
dt
dV
L =
yp p
Y ) 1 ( + =
101

3-88
where is a constant independent of the jet velocity (taken to be 0.7 for steel material),
yp
is the dynamic yield strength of the jet material and E
t
is the its Youngs modulus.
In 1982, Matuska [113] presented the steady state jet penetration using the HULL software
to empirically determine the modified Bernoulli equation parameters, i.e.
U
2
p
(I -u)
2
+[o
=
1
2
p
1
u
2
+oo
.
3-89
It was found that =1, =0.3 and
0 = .7 +.28p
+.86p
2
+.72 lnI
3-90
where
j
is in g/cm
3
, V in km/s and 0 is the deviation parameter which accounts for the
decrease from unity due to the resistance of the target material to the radial flow.
3.6 Crater growth process
Like penetration formulae discussed previously, many models have been proposed to
describe the radius of a crater created by shaped charge penetration [114]. One of these
models was proposed by Held and Kozhushko [111] where the hydrodynamic equation
was used to calculate the maximum crater radius (r
cm
) in aluminium and glass fibre
reinforced plastic targets by shaped charge jet. The experimental crater radius had a good
agreement with model prediction according to the following equation:
i
cm
=
r
v
j
_
p
T
2R
,
_1 +_
p
T
p
j
, ]

3-91
where r
j
is the jet radius,
T
is the target density and R
t
is the target resistance to radial
crater growth, which is different from the target resistance term used in the penetration
depth formulas. R
t
can be determined by the simultaneous measurement of the projectile
and the penetration velocities. It was found that the resistance of target to radial cratering is
lower than to the axial penetration.

)) 4 )(
2
ln(
3
2
(

+ = e
E
R
yt
t
yt t
102
3.7 Summary
Models of shaped charge jet formation were presented in this chapter. The steady state
Birkhoff model assumes constant collapse velocity along the liner surface, while the
collapse velocity in the unsteady state PER model varies along the liner. The analytical,
numerical and semi-epempirical breakup models of a shaped charge jet were presented.
Finally, jet penetration models with the consideration of the breakup time, the stand-off
distance, the jet and target material strengths and the jet velocity gradient were described.
Chapter 4: Hydrocode Simulation
103
CHAPTER.4 HYDROCODE SIMULATION
4.1 Introduction
The numerical simulation is performed using AUTODYN, which can be used to solve non
linear problems related to impact, penetration, perforation and explosion and has built-in
mathematical models such as shaped charge jetting analysis [115]. Autodyn hydrocode is
based on mass, momentum and energy conservation equations, where the materials can be
defined by its equation of state and its strength model [116]. This hydrocode is capable of
performing the shaped charge jetting analysis, jet formation and penetration into concrete
materials, which are briefly described below.
The jetting analysis is based on both the numerical finite difference technique to calculate
the collapse velocities and the analytical unsteady PER theory [50] to calculate the jet and
the slug velocities and masses as well as the collapse and the deflection angles of the liner
elements. In this algorithm, the liner is described as a thin shell composed of a series of
nodes having the real thickness of the liner, while its apex point should be fixed by a
boundary condition to prevent its motion [115]. An interaction between the Lagrangian
shell liner and the Euler explosive and charge casing is defined by Euler-Lagrange polygon
surface. The jetting points are defined for all the liner nodes except the first one fixed by
the boundary condition. Once the jetting is completed, the jetting summary including the
concerned angles, velocities and masses of all liner elements will be obtained.
The jet formation is simulated using Euler method based on continuum mechanics to
obtain the jet profiles at different time stages. In this scheme, the explosive, the charge
casing and the liner materials are filled into the global Euler multi-material part [116]. This
processor is suitable in the early jet formation stages, where large distortions will be
caused by extremely high strain-rate in the order of 10
7
s
-1
[7, 31]. These distortions will
cause the solver to stop working if a Lagrange solver is selected for the jet formation. The
Euler multi-material processor describes the detonation wave propagation inside the charge
and shows the jet profile as it elongates with time. The jet is allowed to move on the Euler
grids up to the moment when it just impacts the target. At this moment, the formed jet will
be remapped as a Lagrangian mass having non-uniform velocity distribution. The output of
this scheme will be used as the input of next scheme.
104
The jet interaction (penetration) with the laminated target layers is simulated using
Lagrange method. In this scheme, the jet obtained from the jet formation Euler solver is
remapped to Lagrange moving grids and impacts the multilayered target. To overcome the
mesh distortion problem in Lagrange solver, a mesh discard option or erosion strain is
applied to the jet and the target materials. The erosion strain does not represent a physical
phenomenon, but a numerical algorithm to prevent the mesh distortions [117].
The input, the solver type and the output data from the jetting analysis and the two
different modeling schemes (solvers) are illustrated in the flow chart in Figure 4-1.

Figure 4-1 The flow chart of the different stages and input and output results from the
jetting analysis and the two solvers (Euler and Lagrange) of the Autodyn hydrocode.
4.2 Studied parameters
- High explosive type: six different explosives were used to study the effect of
detonation characteristics of explosive on the formed jet characteristics and its
105
effects on the penetration capability. These explosives were TNT, RDX, Cyclotol,
HMX, LX-14 and PETN.
- Cone apex angle: the used design cone apex angles were 22
o
, 32
o
, 40
o
, 46
o
, 56
o
,
60
o
and 70
o
with the same explosive charge RDX. The liner thickness was also
changed with different cone apex angles in order to maintain the same explosive
to metal mass ratios. Hence, the jet characteristics are presented as a dependence
of the cone apex angle. Furthermore, numerical simulations were conducted for
constant liner thickness of 1.77mm with four different cone apex angles.
- Liner thickness: the liner thickness is varied for the same explosive, RDX. The
varied liner thickness was 0.8, 1, 1.2, 1.4, 1.6, 1.8, 2mm and 3mm with constant
cone apex angle of 40
o
.
- Liner material: the selected materials for the liner were OFHC solid copper, solid
pure zirconium and copper-tungsten powder.
- Degree of confinement: 1, 2, 4, 6 and 8mm steel casing thickness were used for
RDX OWP of 1.4mm liner wall thickness with cone apex angle of 46
o
.
- Detonation point: the behavior of the detonation wave inside the explosive charge
was studied by selecting two different initiation methods, i.e. a central point on
the charge axis and a point on the side of charge.
- Target material effect: four different target materials were selected to investigate
the effect of target material strength on the penetration depth of perforators. These
were concrete targets with compression strengths of 26, 40, 47 and 55 MPa.
- Water stand-off distance: the performance of an OWP is tested for penetration
after its jet penetrated through 0.5, 1.7, 2, 4 and 6cm of water layers.
4.3 Autodyn jetting analysis description
Autodyn-2D and 3D finite difference codes have a built-in jetting routine, which is
included in the code, where PER theory calculations (explained in chapter 3) is performed
to estimate the jetting parameters for every liner element [115]. In the jetting analysis, the
explosive and the casing parts can be modeled as a Lagrangian or Eulerian grid, but the
liner must be modeled as a shell, in which specified mass points are defined as jetting
points. The boundary condition applied to the liner is Fix because The first node is on the
axis of symmetry and is therefore not able to jet. Instead, the boundary condition fixes this
node in its starting position and prevents its motion [115]. After the shaped charge is
detonated and the jet is formed, the following jet data will be obtained:
106
- Initial X coordinate
- Initial Y coordinate
- Initial liner mass
- Time of jet formation
- X coordinate of jet formation
- X component of collapse velocity at jet formation
- Y component of collapse velocity at jet formation
- Liner collapse angle () at jet formation
- Deflection angle () in the jetting equations
- Collapse speed at jet formation (V
o
)
- Jet velocity (V
j
)
- Velocity of the stagnation point (V
1
)
- Velocity of jet relative to stagnation point (V
2
)
- Jet and slug masses
- Cumulative jet mass and its kinetic energy.
Figure 4-2 illustrates the OWP assembly including the Lagrangan casing, the Eulerian
explosive and the Shell liner with fixed apex node as a complete representation of jetting
model.

Figure 4-2 The jetting model of OWP with cone apex angle 60
o
and liner thickness
1.74mm under fixed apex node boundary condition.
4.4 Autodyn jet formation model description
The jet formation model was established in order to obtain the jet profile, the contours of
different jet parameters and the jet breakup phenomena, which were needed to test the
performance of the perforator. The model uses Euler solver with outflow boundary
107
condition, which allows the detonation gaseous products and the casing material to expand
smoothly towards the Euler part boundary and prevents them from returning back to avoid
their effect on the jet and the slug formation as shown in Figure 4-3. The large deformation
occurs inside the liner, and a continuous jet is formed while the fixed meshes allow the jet
to moving over it without being stopped because of mesh distortion. The code output from
the Euler solver will create jet and slug profiles according to the Conical Shaped Charge
(CSC) perforator design. This jet (Figure 4-4) will be remapped to a new Lagrange model,
which is suitable for simulating the jet penetration into laminated configuration of the
target layers, as shown in Figure 4-5. The physical parameters of the jet, such as its kinetic
energy will be almost unchanged during the remapping and exporting process.

Figure 4-3 The Euler part in 2-D Visualizer showing the geometry and the boundaries of
the jet formation model.

Figure 4-4 The produced jet obtained from Euler solver and remapped to Lagrange
processor for penetration analysis.
108

Figure 4-5 The produced jet impacting on steel gun casing, water wellbore fluid, steel
casing and concrete material.
4.5 Material modeling description
4.5.1 Description of the used explosives
The explosive required for shaped charges must have high velocity of detonation and high
density to provide a high detonation pressure, which results in fast jet tip velocity and
larger depth of penetration [6]. The explosive materials used for filling CSC perforators
were TNT, HMX, PETN, Cyclotol, RDX and LX-14. The equation of state for the used
explosives is Jones-Wilkins-Lee (JWL) equation, which is a simple pressure, volume,
energy (PVE) relation that has been developed to describe the adiabatic expansion of the
detonation products of explosives [118].
p = A_1 -
u
i
1
v
] c
-r
1
v
+B_1 -
u
i
2
v
] c
-r
2
v
+
E
v
4-1
where p is the pressure, v is the relative volume (1/), A, B, r
1
, r
2
, C and are constants
[119]. The values of the experimental constants for some explosives have been determined
from sideways plate push dynamic test experiments [120]. These values were determined
experimentally by the cylinder expansion test. For the listed explosives, the values of the
above mentioned constants are available in the material library of Autodyn. The input data
to Autodyn hydrocode for the explosive materials are listed in Table 4-1.
4.5.2 Explosive initiation and wave propagation
The detonation wave is assumed to travel at the prescribed detonation velocity U
D
and its
path from the predefined initiation point can be determined. The detonation wave
propagates in the spherical direction to engulf the whole un-burnt explosive meshes. The
109
use of JWL constitutive model assumes that the detonation wave is strong enough
to completely detonate the explosive and an instantaneous transition to the CJ state is
achieved. At this state, the full reaction of the explosive is completed and the full energy of
the explosive is liberated, after which the detonation gaseous products will start to expand.
The different stages of a CSC detonation are illustrated in Figure 4-6, where the
propagation direction of the contours shape spreading from the initiation point source can
be observed.
Table 4-1 Input data to the code for the used explosive materials.
Parameter
Explosive Type
TNT HMX Cyclotol LX-17 PETN RDX
Density
(g/cm
3
)
1.630 1.891 1.754 1.900 1.500 1.600
Parameter A
(kPa)
3.74010
8
7.78210
8
6.03410
8
4.46 10
8
6.25310
8
6.53910
8

Parameter B
(kPa)
3.74710
6
7.07110
6
9.92310
6
1.33910
6
2.329 10
7
7.29310
7

Parameter r
1
4.15 4.2 4.3 3.85 5.25
4.83

Parameter r
2
0.9 1 1.1 1.03 1.6 2.24
C-J detonation
velocity (m/s)
6930 9100 8250 7600 7450 8100
C-J energy
per unit
volume
(kJ/m
3
)
6.0010
6
1.0510
7
9.210
6
6.910
6
8.5610
6
5.6210
6

C-J pressure
(kPa)
2.110
7
4.210
7
3.210
7
3.010
7
2.210
7
2.610
7

Parameter 0.3 0.3 0.35 0.46 0.28 0.3
110

Figure 4-6 The different stages of detonation wave spherical propagation through the
explosive charge inside CSC.
4.5.3 Description of the liner materials
The materials that have been used for liner element were solid copper-OFHC, solid
zirconium, and copper-tungsten powder mixture. The equations of state (EOS) of these
materials were shock, while their strength models were neglected because shock pressure
is much higher than material strength [116]. This EOS is suitable for both monolithic liner
material manufactured by spinning or drawing and powder mixture liner manufactured by
powder metallurgy technique. In 2010, Liu and Shen [43] used Autodyn hydrocode to
model the Copper-Tungsten liner manufactured by powder pressing. Liu and Shen used
shock equation of state to model the solid liner. The mixture theory was used to determine
111
the parameters of the shock EOS, in which each parameter was calculated according to the
mass fraction of its material in the mixture. Good agreement between the numerical and
the experimental results were obtained.
It has been shown experimentally that, for most solids and liquids that do not undergo a
phase change, the shock Hugoniot values of shock velocity (U) and material velocity
behind the shock (up) can be adequately fitted to a straight line ( Figure 4-7).

Figure 4-7 The shock velocity against particle velocity for the EOS of the liner material
[116].
u = C
o
+su
p
. 4-2
This is valid up to shock velocities around twice the initial sound speed C
0
and shock
pressures in the order of 100 GPa. For materials where a linear fit is not adequate, a
quadratic relation between U and u
p
has been used. Generally, piecewise linear or
piecewise quadratic relations (U, u
p
) can be applied [116].
The Mie-Gruneisen EOS based on the shock Hugoniot is expressed as:
p = p
H
+Ip(e -e
H
)
4-3
where is the Gruneisen Gamma coefficient and equal to where B
o
is a
constant, =
o
o
=constant is assumed; is the density. p
H
and

e
H
are the Hugoniot
pressure and energy, respectively, given by
) 1 /( +
o
B
112
4-4
and 4-5
where =(/
o
)-1 is the compressibility, C
o
is the sound speed in the material and s is a
constant giving the slope of shock velocity-particle velocity relationship. The mechanical
properties of these materials are illustrated in Table 4-2, where the constants in the
previous equations were taken from the material library.
Table 4-2 The mechanical properties of liner materials [116].
Parameter OFHC Copper Tungsten Alloy Zirconium
Equation of state Shock Shock Shock
Reference density (g/cm
3
) 8.90 17.00 6.51
Gruneisen Coefficient 2.02 1.54 1.09
Parameter C (m/s) 3940 4029 3757
Parameter s (non) 1.489 1.237 1.018
Ref. temperature (K) 300 300 300
Strength
None None None
4.5.4 Description of the charge case
Unlike shaped charge, oil well perforator has a thick confinement in order to afford the
high pressure in the well bore of about 68 MPa at depth of 3km below ground level and
also the elevated temperature of 300
o
C at this depth, which is close to the ignition
temperature of some high explosives [11]. The confinement degree does not affect the jet
velocity of the liner elements near the charge axis. However, collapse velocities associated
with the liner elements near the charge edges (i.e. base of the conical liner) are strongly
affected by the degree of confinement. The material used for the charge case was steel
4340. The equation of state for the steel is shock EOS which has been described previously
for the liner material, while its strength model was Johnson-Cook. This constitutive model
aims to model the strength behavior of materials subjected to large strains, high strain-rates
and high temperatures [116]. The model defines the dynamic yield stress Y [121] as:
o = (A +Be

n
)(1 +Cne
-

)(1 -I
H
m
) 4-6
2
2
] ) 1 ( 1 [
) 1 (

+
=
s
c
p
o o
H
|
|
\
|
+
=
1 2
1
o
H
H
p
e
113
where is the dynamic flow stress, is the effective plastic strain, A is the yield strength,
B is the hardening constant, n is the hardening exponent, C is the strain-rate constant and
m is the thermal exponent constant.
-

is the normalized effective plastic strain-rate (i.e.
the applied true strain-rate divided by the reference strain-rate). T
H
is the homologous
temperature that can be calculated by:
.
4-7
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
*
equal to1.0 sec
-1
and T
H
= 0 for laboratory
experiments at room temperature. The expressions in the second and third sets of brackets
represent the effects of strain-rate and temperature, respectively. In particular, the third
term represents the thermal softening so that the yield stress drops to zero at the melting
temperature T
melt
. The constants in these expressions determined by means of material tests
over a range of temperatures and strain-rates. The input data to Autodyn for the case
material are listed in Table 4-3.
Table 4-3 Input data to the code for the charge casing material [116].
3
) 7.83
Tensile strength (MPa) 744
A (MPa) 792
B (MPa) 510
n (non) 0.26
C (non) 0.014
m (non) 1.03
Gruneisen coefficient 1.93
Parameter C1 (m/s) 4569
Parameter S1 (non) 1.4
Ref. temperature (K) 300
4.5.5 Description of the concrete material
4.5.5.1 General
The concrete material is modeled by P- EOS, which was presented by Herrmann [122].
This model provides a good description of matreial behavior at high stresses and a
reasonable description of the compaction process at low stress levels. This model was
validated by Heider and Hiermaier [123]. The strength model used with P- was the RHT
(Riedel-Hiermaier-Thoma) brittle material constitutive model [124]. This model describes
room melt
room
H
T T
T T
T
=
114
the dynamic resistance of concrete and other brittle material such as rock and ceramics by
the combined plasticity and shear damage, in which the deviatoric stress in the material is
limited by the generalised failure surface. Further details of RHT model can be found from
reference [125]. The P- porous model and the RHT brittle material model were validated
by Berg and Preece [117] and Leppnen [126], who implemented a bi-linear softening law
that modifies the strain-rate dependency in tension, hence improves the spalling, cracking
and scabbing of concrete impacted by a K.E. projectile or fragments. Hayhurst et al. [127]
validated these models using SPH, Lagrange and Euler solvers for hard penetrator
impacting on ceramic armour. The difference between the measured and calculated
Lagrange penetration depths was small (6.7%).
4.5.5.2 Equation of state of the concrete material
According to API-RP43, the concrete material used for testing oil well perforator is ASTM
C33-67 concrete. This concrete contains agglomeration sand conditioned to simulate the
down-hole hardness, porosity and compressive strength of the rock formation [128].
The EOS used to simulate the concrete material is P- for the porous material, while the
fully compacted one will be modeled by polynomial EOS [122], which will be described
below. The complete parameters of both the two used EOS are illustrated in Table 4-4,
while the general behavior of the concrete material is illustrated in Figure 4-8.

Figure 4-8 The pressure-porosity curve for the concrete material [116].
The porosity is given by:
v
v
s
p
s
p

4-8
115
where, V=1/ and V
s
=1/
s
are the specific densities of the porous and solid materials,
respectively.
The P- relation for the porous materials was suggested by Herrmann [122].
P
s
s c
n
4-9
where n is the pressure exponent (n=3 is normally used). P
e
and P
s
are the elastic pressure
and the fully compaction pressure, respectively.
P
is the material porosity at the beginning
of the plastic deformation.
The pressure exponent in this model was modified by many authors in order to fit the
experimental date to this model.
The general formula of the polynomial EOS for the compacted material is:

4-10
where A
1
, A
2
, A
3
, B
0
and B
1
are constants; is the compressibility; e is the specific internal
energy per unit mass. The parameters of this equation are listed in Table 4-4 for the
concrete materials.
Table 4-4 The input parameters for P- and the polynomial EOS for the concrete targets
[116].
26MPa 35MPa 40MPa 47MPa 55MPa
Porous EOS P-
3
) 2.75 2.75 2.75 2.75 2.75
Porous sound speed (m/s) 2892 2920 2935 2957 2981
Initial compaction pressure (MPa) P
e
17.3 23.3 26.6 31.3 36.6
Solid compaction pressure (GPa) P
s
6 6 6 6 6
Compaction exponent n 3 3 3 3 3
Solid EOS Polynomial
Bulk modulus A
1
(GPa) 35.27 35.27 35.27 35.27 35.27
Parameter A
2
(GPa) 39.58 39.58 39.58 39.58 39.58
Parameter A
3
(GPa) 9.04 9.04 9.04 9.04 9.04
Parameter B
0
(none) 1.22 1.22 1.22 1.22 1.22
Parameter B
1
(none) 1.22 1.22 1.22 1.22 1.22
e B B A A A P
o
) (
1 0
3
3
2
2 1
+ + + + =
116
4.5.5.3 Strength model for concrete
Over the past decades, extensive number of papers have been published on experimental,
analytical and computational methods to study penetration mechanics of hypervelocity
projectiles into concrete and rock materials [129]. Adel [130] used AUTODYN 3-D in
order to select the optimum strength model and failure criteria related to limestone and
concrete targets penetrated by kinetic energy penetrator moving with a velocity up to
1500m/s. The strength models that have been tested were RHT- brittle material model
[131], Von Mises model and Druker-Prager model. It was found that the RHT brittle
material constitutive model can be used efficiently in characterizing the non-linear
behaviour of the rock during penetration especially when RHT failure damage model is
taken into consideration. It demonstrated a correct physical mechanism for the penetration
process proved by crater profile and a good agreement between the experimental and the
calculated penetration depths.
The RHT brittle material constitutive model is an advanced plastic model proposed by
Riedel, Hiermaier and Thoma at the Ernst Mach Institute (EMI). This model describes the
dynamic resistance of concrete and other brittle material such as rock and ceramic by the
combined plasticity and shear damage in which the deviatoric stress in the material is
limited by the generalised failure surface. Generally, the RHT model could be divided into
five parts, which are [125]:
4.5.5.3.1 The failure surface:
The failure surface, Y
fail
, is defined as a function of hydrostatic pressure (P), lode angle ()
and strain-rate (e).
]uI
(P
-
, 0, e) =
1XC
(P) R
3
(0) F
RA1L
(e) 4-11
where Y
TXC
is the compressive meridian and is given by:
1XC
=
c
|A(P
-
-P
spuII
-
F
RA1L
)
N
] 4-12
where f
c
is the unconfined uniaxial compressive strength, A is the failure surface constant,
N is the failure surface exponent, P
*
is the pressure normalized by f
c
, P
spuII
-
=
c
, f
t
is
the uniaxial tensile strength and F
RATE
is the dynamic increase factor and is defined by:
117

RA1L
s
s
c

for P> f
c
/3 (compression) 4-13
RA1L
s
s
c
u

for P< f
c
/3 (tension) 4-14
where D and a are the compressive and tensile strain-rate factor exponents, respectively.
R
3
() defines the third invariant dependency of the model through the tension/compression
meridian ratio.
4.5.5.3.2 Elastic limit surface:
The elastic limit surface is scaled from the failure surface using:
Y
cIastIc
= Y
IaII
F
cIastIc
F
CAP(P)
4-15
where F
elastic
is the ratio of elastic strength to failure surface strength based on the tensile
elastic strength (f
t
) and compressive elastic strength (f
c
) shown in Figure 4-9. F
CAP(P)
is a
function that limits the deviatoric elastic stresses under hydrostatic compression. The
model provides an option to close the elastic limit surface towards high pressures to ensure
the consistency between the deviatoric and inelastic volumetric stresses.

Figure 4-9 Stress loading curve for the RHT strength material model [116].
4.5.5.3.3 Strain hardening
Linear hardening is used in RHT model prior to the peak load. During hardening, the
current yield surface, Y
*
, is scaled between the elastic limit surface and the failure surface.
118

4-16
The values of
pl
and
pl(pre-softening)
are shown in Figure 4-10.

Figure 4-10 The concrete strain hardening curve according to RHT model.
4.5.5.3.4 Residual failure surface
The residual frictional failure surface is defined by:
csd
-
= BP
-
M

4-17
where B is the residual failure surface constant and M is the residual failure surface
exponent.
4.5.5.3.5 Damage
The plastic straining of the material leads to accumulated damage and strength reduction
i.e.,

4-18
,
4-19
where D
1
and D
2
are the damage constants and
f
min
is the minimum strain to failure (input
parameter). The post-damaged failure surface is interpolated by:
.
4-20
) (
) (
*
elastic fail
softening pre pl
pl
elastic
Y Y Y Y + =

=
failure
p
pl
D
min * *
1
2
) (
f
D
spall
failure
p
P P D =
* * *
) 1 (
residual failure fractured
DY Y D Y + =
119
The post-damaged shear modulus is interpolated by:
,
4-21
where G
residual
is the residual shear modulus fraction (input parameter). The strength model
parameters are listed in Table 4-5 for concrete materials.
Table 4-5 The input parameters for the RHT strength and failure model for concrete
materials [116].
Concrete strength (MPa) units 26 35 40 47 55
Equation of State P-
Porous density g/cm
3
2.30 2.31 2.32 2.34 2.35
Porous sound speed m/s 2892 2920 2935 2957 2981
Initial compaction pressure MPa 17.3 23.3 26.6 31.3 36.6
Solid compaction pressure GPa 6 6 6 6 6
Compaction exponent - 3 3 3 3 3
Strength RHT Concrete
Shear Modulus GPa 16.2 16.7 17.0 17.3 17.7
Compressive Strength (fc) MPa 26 35 40 47 55
Tensile Strength (ft/fc) - 0.1 0.1 0.1 0.1 0.1
Shear Strength (fs/fc) - 0.18 0.18 0.18 0.18 0.18
Intact Failure Surface Constant A - 1.6 1.6 1.6 1.6 1.6
Intact Failure Surface Exponent N - 0.61 0.61 0.61 0.61 0.61
Tens./Comp. Meridian Ratio (Q) - 0.68 0.68 0.68 0.68 0.68
Brittle to Ductile Transition - 0.01 0.01 0.01 0.01 0.01
Elastic Strength / ft - 0.70 0.70 0.70 0.70 0.70
Elastic Strength / fc - 0.53 0.53 0.53 0.53 0.53
Fractured Strength Constant B - 1.60 1.60 1.60 1.60 1.60
Fractured Strength Exponent M - 0.61 0.61 0.61 0.61 0.61
Compressive Strain-rate Exp.
-
0.034 0.032 0.031 0.029 0.028
Tensile Strain-rate Exp.
-
0.038 0.036 0.035 0.033 0.032
4.5.6 Description of the layer of the steel gun carrier and the wellbore casing
The selected material for both gun carrier and wellbore casing is Steel A-36 according to
the previously mentioned API standard. The equation of state for the steel is shock model
while the selected strength model is Johnson-Cook model. The input parameters for the A-
36 steel material are listed in Table 4-6.
residual fractured
DG G D G + = ) 1 (
120
4.5.7 Description of the water layer
Water is the standard wellbore fluid used to test the oil well perforator performance. The
selected equation of state of the water layer is linear with no strength model. The reference
density of water is 1 g/cm
3
and its bulk modulus is 2.23 GPa [116] with reference
temperature of 300 K.
Table 4-6 The input parameters for the A-36 steel material [116].
Parameter Steel A-36
3
) 7.85
A (MPa) 250
B (MPa) 477
n (non) 0.18
C (non) 0.012
m (non) 1
Gruneisen coefficient 2.17
Parameter C (m/s) 4569
Parameter S (non) 1.49
Ref. temperature (K) 300
4.6 Solution stability
Since the numerical algorithm used in Autodyn is an explicit scheme, there is an optimum
time step of integration, which must be determined to obtain a reasonable representation of
solution. The local time step ensuring stability is calculated for each mesh point. The
minimum value of all these local values multiplied by a safety factor (a default value of 2/3
is built into the code) is chosen as the time step for the next update. In Lagrangian mesh,
the time step must satisfy the Courant condition [116], i.e.,
t d / c 4-22
where d is the typical length of a mesh (defined as the area of the mesh divided by its
longer diagonal) and c is the local sound speed. This ensures that a disturbance does not
propagate across a mesh in a single time step.
The minimum value of t must be found for all zones and this value will be used for all
meshes for the next time step of integration.

121
4.7 Output of numerical modeling
In the following, the predicted parameters associated with the different simulation studies
herein are listed. The histories of these parameters during their simulation processes will
also be presented in the next chapter. The histories of the following parameters are
predicted from the jetting analysis solver in Autodyn:
- Jetting points parameters (collapse velocity, collapse angles, elemental velocity
history, jet tip velocity, jet momentum and kinetic energy, jet and liner masses)
- Cumulative jet mass and length.
- The time at which each liner node point will be jetted on the jet axis
Output of jet formation model (Euler)
The histories of the following parameters from the jet formation model are predicted:
- Jet profile at different times
- Jet breakup phenomena
- Different jet contours (pressure, temperature, sound speed, velocity, etc.)
- Energy history plots (momentum, kinetic and internal energies).
In addition to the selected gauge point histories at the specified spatial locations and the
Lagrangian jet that could be obtained from the remapping model, outputs of jet penetration
into laminated layers configuration (all Lagrange layers) include:
- Jet penetration into concrete at different times,
- Crater profile along the penetration path,
- The damage contours accompanied with the penetration process, and
- The history plot of gauge points at different times.

Chapter 5: Parametric Analysis Results
122
CHAPTER.5 PARAMETRIC ANALYSIS
RESULTS
5.1 Introduction
This chapter presents parametric analyses for the numerical algorithms and models used in
this project. The Autodyn hydro-code package is used to perform the shaped charge jetting
analysis, the jet formation modelling and the modeling of the formed jet interaction with
A-36 steel, water and A-36 steel layers backed by concrete targets, according to the
standard API-RP 43 (quality control target). Generally, this chapter presents the obtained
parametric analysis results on the following main issues
- General features of the shaped charge jetting analysis, the jet formation and jet
penetration models,
- The mesh sensitivity study for the jetting analysis and the jet penetration,
- The verification and validation of the hydro-code software,
- The effect of the surrounding medium (air or void) on the jet velocity,
- The parametric analysis of the OWP including the liner, the explosive and the
charge design as well as the detonation point effect,
- The liner portioning into jet and slug portions, and
- The Gurney velocity approximation.
5.2 The main features of the jetting analysis and the jet formation and
penetration solvers
5.2.1 Standard shaped charge jetting analysis model
A series of the standard jetting analysis at different times is shown in Figure 5-1 for the
OWP of 1.74mm liner wall thickness and 60
o
cone apex angle. The model stopped at
19.6s from the moment of detonation. The time represents the total time, at which the
entire liner elements arrive at its axis and take part in the jet and slug portions.
123

Figure 5-1 The different stages of the jetting analysis of the OWP 60
o
cone apex angle and
liner wall thickness of 1.74mm.
The standard jetting analysis output is a HTML file and contains the jetting data of all the
liner elements. The data may be used for further calculations such as in virtual origin
model and in the determination of the breakup time. They also can be used directly to
predict the jet characteristics such as its kinetic energy and its momentum. The jetting
solution summary is listed in Table 5-1. The data file gives a range of jetting parameters of
124
jet elements with respect to their original distances from the liner apex, as shown in Figure
5-2, as an example.
Table 5-1 The jetting summary of OWP 60
o
with liner wall thickness of 1.74mm
liner mass (g) 28.3 Jet mass (g) 5.35
Liner momentum (kg.m/s) 24.3 Jet momentum (kg.m/s) 16.7
Liner kinetic energy (kJ) 31.9 Jet kinetic energy (kJ) 30.2
J
X
o
(mm)
Y
o
(mm)
Lin.
mass
(g)
X-jet
(mm)
T-jet
(s)
V
o

(m/s)
Angle

V
1

(m/s)
V
2

(m/s)
V
jet

(m/s)
Jet
mass
(g)
2 2.72 3.66
0.355
34.9 2.30 576.0 32.2 1080.2 941.9 2022.1 0.0273
3 3.58 4.17
0.403
35.9 2.96 761.9 34.7 1337.6 1146.0 2483.6 0.0358
4 4.44 4.67
0.452
37.0 3.50 883.1 35.6 1515.1 1284.9 2800.0 0.0422
5 5.30 5.17
0.501
38.0 4.01 985.1 36.2 1664.5 1404.4 3068.9 0.0483
6 6.16 5.67
0.549
39.1 4.49 1077.0 36.7 1799.3 1512.4 3311.6 0.0544
7 7.02 6.17
0.598
40.2 4.94 1172.5 37.0 1945.4 1631.6 3577.0 0.0601
8 7.88 6.67
0.646
41.3 5.38 1256.0 37.5 2057.3 1718.1 3775.4 0.0668
9 8.74 7.17
0.695
42.4 5.82 1331.9 38.3 2142.8 1781.6 3924.5 0.0748
10 9.60 7.67
0.743
43.5 6.21 1429.1 38.1 2310.3 1917.5 4227.8 0.0792
11 10.46 8.18
0.792
44.6 6.59 1515.9 38.0 2459.8 2035.8 4495.5 0.0837
12 11.32 8.68
0.840
45.7 6.97 1572.8 38.3 2530.9 2087.1 4618.1 0.0906
13 12.18 9.18
0.889
46.8 7.31 1611.6 38.0 2615.4 2156.5 4771.9 0.0940
14 13.04 9.68
0.937
47.9 7.66 1643.5 37.8 2677.5 2207.1 4884.6 0.0983
15 13.90 10.18
0.986
49.0 8.02 1664.5 37.7 2714.8 2236.9 4951.8 0.1032
16 14.76 10.68
1.035
50.1 8.37 1680.1 38.1 2717 2237.0 4954.1 0.1103
17 15.62 11.18
1.083
51.2 8.73 1681.9 38.6 2692.1 2213.9 4906.1 0.1181
18 16.48 11.68
1.132
52.3 9.09 1681.1 38.5 2696.3 2219.5 4915.8 0.1229
19 17.34 12.19
1.180
53.5 9.48 1679.6 39.6 2626.1 2160.2 4786.3 0.1354
20 18.20 12.69
1.229
54.6 9.87 1664.9 40.7 2539.9 2090.0 4629.9 0.1487
21 19.06 13.19
1.277
55.8 10.27 1648.1 41.5 2471.8 2038.3 4510.1 0.1602
22 19.92 13.69
1.326
56.9 10.73 1627.7 43.1 2358.4 1950.5 4308.9 0.1789
23 20.78 14.19
1.374
58.1 11.24 1601.9 45.7 2201.9 1831.5 4033.3 0.2069
24 21.64 14.69
1.423
59.3 11.82 1571.3 49.0 2024.1 1701.5 3725.5 0.2443
25 22.50 15.19
1.472
60.6 12.51 1535.2 52.8 1838.8 1572.0 3410.8 0.2909
26 23.36 15.69
1.520
62.0 13.41 1493.1 58.4 1617.8 1424.2 3042.0 0.3615
27 24.22 16.20
1.569
63.7 14.72 1451.2 64.5 1420.6 1292.9 2713.5 0.4462
28 25.08 16.70
1.617
66.6 17.33 1413.1 87.7 949.1 1085.1 2034.2 0.7767
29 25.94 17.20
1.666
68.8 20.56 1240.3 108.1 526.06 971.6 1497.6 1.0917
In Table 5-1, X
o
, Y
o
are the initial position of the liner element; X-jet is the element jet
position, T-jet is the time to jet formation, V
1
, V
2
are the stagnation and flow velocities,
respectively; V
o
is the collapse velocity and is the collapse angle.
125

Figure 5-2 The stagnation, the flow and the jet velocities of the OWP calculated using
jetting analysis.
5.2.2 Shaped charge jet formation and penetration
In the following part, the jet profile for each conical shaped charge (CSC) OWP after
detonation is obtained as illustrated in Figure 5-3 and Figure 5-4. Figure 5-3 shows the
sequence of events from detonation to the breakup of the formed jet. The jet tip velocity,
cut-off velocity (velocity of the last penetrating element), jet breakup time (the time after
which the jet elements starts to particulate and hence its penetration capability decreases
dramatically) and the penetration depth into concrete target were obtained and analyzed.
Figure 5-4 illustrates the jet, slug and cut-off elements; while Figure 5-5 illustrates the grid
plot of the same jet remapped from the jet formation using Euler solver and imported to the
Lagrange solver. This jet will hit a laminated target (steel/water/steel) layers backed by the
tested concrete as illustrated in Figure 5-6.

126

Figure 5-3 The different stages of the detonation of CSC at different times indicating the
start of the jet breakup at 54.91 s.
127

Figure 5-4 Velocity vectors of the shaped charge jet indicating the velocity gradient.

Figure 5-5 Grid plot of the shaped charge jet remapped into Lagrange processor.

Figure 5-6 OWP remapped jet penetrating the gun wall, water wellbore fluid, steel tube
casing and concrete (35MPa).
128
5.3 Mesh sensitivity study
5.3.1 Mesh sensitivity for the jetting analysis
It is well known that the shape and the density of the mesh affect the simulation results.
Generally, simulation with fine meshes produces an accurate solution; but it consumes
longer time than that needed for coarse meshing simulations. The mesh sensitivity study
for the jetting analysis is performed on a conical OWP of apex angle of 46 degree and
copper liner with tapered profile. The Euler grids containing the explosive charge had four
uniform square cells with different sizes of 0.3, 0.6, 1.4 and 2mm. The four jetting models
with different mesh sizes were allowed to run until the jetting analysis is completed for the
entire liner elements. Table 5-2 lists some of the jetting summary output data obtained
from the four models. Little change can be observed in the jet mass for the four models,
but significant difference was shown for the jet kinetic energy. The difference in the jet
masses is explained by variation of the jet collapse angle for the four models as shown in
Figure 5-7. The collapse angle has a direct effect on the mass ratio of the liner that flows to
form the jet and the slug [50]. The variation among the four models in their kinetic energy
is mainly caused by the different velocities of jet elements for the four meshes as indicated
in Figure 5-8. The jet elemental velocities at the liner apex and its base are almost
independant of the mesh size, but the velocity difference for different mesh sizes become
obvious at certain middle part of the liner, at which most of the jet will be formed. Thus,
the velocity drift in this area is the main reason for different kinetic energies. The jet
velocity curves seem to be convergent to the curve of 0.3mm mesh model curve, which
means that this mesh size is expected to be close to the asymptote limit. On the other hand,
the computational time for the model using the finest mesh size of 0.3mm is only 25%
longer than that needed for the coarse mesh of 2mm due to the semi-analytical nature of
the jetting analysis. Therefore, the mesh size of 0.30.3mm was used for the rest of
parametric studies of jetting models.
Table 5-2 The jet mass and its kinetic energy for different mesh sizes.
Mesh size(mm) 0.30.3 0.60.6 1.41.4 22
Jet mass (g) 6.02 6.14 6.29 6.69
Jet K.E. (kJ) 49.30 47.40 43.20 41.20

129

Figure 5-7 The collapse angle at different distances from the liner apex using different
meshes.

Figure 5-8 The elemental jet velocities at different distances using different mesh sizes.

130
5.3.2 Mesh sensitivity for the jet penetration
Mesh sensitivity is also an important issue in jet penetration. In order to find how the
penetration depth into the concrete material is related to the mesh size, five different
uniform square mesh sizes of 0.5, 1, 2, 3 and 4mm were used for the laminated target i.e.
the steel, the water and the concrete targets, while the copper jet mesh density of
0.50.5mm remains unchanged for the five models. The penetration depth into the
concrete using different mesh sizes is depicted in Figure 5-9. This figure shows a
convergent penetration depth to the value of 65.3 cm corresponding to the mesh size of
0.5mm; however, the time consumption for this mesh size is eight times more than that
needed with the mesh size of 4mm. On the other hand the penetration depth using mesh
size of 1mm has only 0.3% (i.e. 0.2cm) difference in comparison with the result obtained
for finest mesh size of 0.5mm, but its time consumption is less than half the time needed
for the 0.5mm mesh size. Thus the mesh size of 1mm1mm is used for the rest of
simulations.

Figure 5-9 The penetration depths into the concrete using different mesh sizes and the
relevant time consumption.

131
5.4 The verification and Validation (V&V) of the hydro-code results
Autodyn hydro-code has been verified by the code developer during their development
process and by numerous applications of the code. Thus, only validation issue will be
addressed in this section.
5.4.1 The jetting analysis and the jet formation Validation
The numerical Autodyn jetting analysis algorithm was validated by Century Dynamics for
a shaped charge of 90mm liner diameter and a cone angle of 18
o
. The results of the
Autodyn numerical jetting analysis agreed with the experimental results for this shaped
charge and other analytical models (i.e. HEMP and PISCES) [115]. In the present research,
the flash x-ray is used to measure the jet tip velocity and to depict the jet profile at different
times to compare with the numerical results in this study. The flash x-ray trial was
performed in COTEC (Cranfield Ordnance Test and Evaluation Center) field. Two heads
were used to capture photos of the jet profile at different times. The aluminium foil layers
were used to trigger the time when the jet tip penetrates through them. Figure 5-10 shows
the setup of the x-ray trial field test, while Figure 5-11 shows the jet shapes from x-ray
photos and the numerical jet formations at 34s and 122s, respectively. A curved shape is
observed for the real x-ray jet, which may be caused either by some asymmetries in the
liner positioning during the manual filling of the charge or due to the non-uniform
explosive mass distribution inside the charge cavity. The main reason for these possible
defects is due to the COTEC safety regulation requirements, which demand the charge
filling at the test location. Nevertheless, the general aspects of the jet shapes are similar.
Besides, the jet tip velocities were found to be 6100m/s and 6182m/s from the x-ray
measurements and the numerical simulation, respectively, as shown in Figure 5-12. This
means that the error percent is only 1.34% in terms of jet tip velocity, which implies that
the numerical hydro-code can be used effectively to model and calculate the shaped charge
jet characteristics.

132

Figure 5-10 The flash x-ray trial setup, 1: the tested OWP, 2: the aluminium foil layers, 3:
the x-ray heads.

Figure 5-11 The real x-ray jet and the numerical Euler jet at 34s and 122s from the
moment of detonation.
133

Figure 5-12 The numerical jet velocity at different distances from the liner apex and the
real tip velocity estimated experimentally.
5.4.2 The validation of the hydro-code penetration modeling
The Autodyn simulations for penetration into concrete materials using Lagrange solver
have been validated experimentally by many authors [117, 124, 130, 132-133]. In this
research, the validity of the numerical hydro-code penetration model will be demonstrated
by several penetration tests of OWP into API-RP43 configurations. A sample of these tests
for a copper liner OWP is shown in Figure 5-13 with the crater profile obtained by the
Autodyn penetration modeling. The experimental and the numerical penetration depths
were 64cm and 65cm, respectively. The penetration craters are almost identical in
experiment and simulation. Thus, the Lagrange numerical model can be used effectively to
predict the penetration depth into concrete target with sufficient accuracy (i.e. a small
difference of 1.6% was observed between experimental and numerical results).

134

Figure 5-13 The experimental (upper) and the numerical (lower) penetration depths into
40MPa concrete.
5.5 Effect of the surrounding medium on the jet characteristics
The medium surrounding the jet during its flight has a direct effect on the jet velocity as it
travels through this medium [134]. To study this effect numerically, two identical shaped
charge OWP were fired in air and void mediums, while the jet tip velocities in both media
were tracked using the fixed gauge point facility. The gauge points were located at a
distance of 1 CD (i.e. charge calibre of 36mm) from each other. A sketch of the gauge
points and their locations is shown in Figure 5-14. Figure 5-15 shows the measured
maximum jet tip velocities for the three gauges in both media. This figure indicates that
the jet tip velocity slightly decreases as it travels short distances (i.e. 3CD). The rates of
velocity decrease are 50m/s and 53m/s per 1CD for the jet tip travelling through air and
void materials, respectively. Generally, the difference in the jet tip velocity in both cases at
short stand-off distance is negligible, which means that the void medium can be used
instead of air because it has some advantages to largely reduce the calculation time.
Besides, most of the modelled OWPs have to be tested against the laminated target at short
stand-off distance (i.e. 1CD), which means that this medium can be used effectively
without major changes in the velocity of the jet elements. On the other hand, the jet tip
velocity measured in air medium is higher than that in a void medium. This can be
explained by the jet velocity-time histories in both mediums as shown in Figure 5-16 for
the three gauges. Both the histories are similar except for their peak values that have very
small difference between them. Such difference may be attributed to the air motion in front
of the jet that can cause little increase in the recorded peak velocity values, but the overall
velocity shapes and their arrival times are almost identical in both media.

135

Figure 5-14 The located fixed gauge points used to predict the surrounding medium effect
on the tip velocity.

Figure 5-15 The jet tip velocities at different gauges for the air and void media.
136

Figure 5-16 The velocity-time histories for the three jets stretching through air and void
mediums.
5.6 Shaped charge parametric study results
5.6.1 High explosive effect on jet performance
Gurney energy or Gurney velocity is a measure of explosive power or its efficiency. The
higher the Gurney energy, the higher the velocity of the produced jet, and hence the higher
the penetration capability of the shaped charge. Figure 5-17 shows the dependence of the
jet tip velocity and the Gurney velocity on the explosion heat (Qv) of the used explosive
charge. The details of the used explosives and the produced jet characteristics obtained
from standard jetting analysis are listed in Table 5-3. The relation between the jet tip
velocity and the detonation velocity of the explosive is illustrated in Figure 5-18. It shows
that the most powerful explosive is HMX, which has a Gurney velocity of 2960m/s and
detonation velocity of 9100 m/s. This explosive produces a jet tip velocity of 7103m/s and
a jet mass ratio of 17.76%. This result was confirmed by the jet formation model and
penetration model tests where the OWP filled by HMX produced the largest penetration
depth of 74.88cm. Table 5-4 lists the penetration depths, jet tip and tail velocities and exit
hole diameter of the different OWP obtained from jet formation and penetration models
using different explosive charges.
137

Figure 5-17 The dependence of jet tip velocity and Gurney velocity on the explosion heat
of explosives.
Table 5-3 Effect of the explosive type on the jet characteristics of 46
o
conical copper liner
of wall thickness of 1.4mm and 29.32g liner mass with 4.5 mm steel casing thickness.
Explosive properties Output
Explosive

o

(g/cm
3
)
D
(m/s)
Qv
(kJ/kg)

(m/s)
Jet
mass
(g)
Jet %
Jet tip
vel.
(m/s)
Jet K.E.
(kJ)
TNT 1.63 6930
3681 2390 5.45 16.47 6108 38.19
PETN 1.50 7450
5707 2920 5.33 16.10 6605 49.70
LX-17 1.90 7600
6900 2680 5.73 17.30 6046 36.06
RDX 1.73 8100
4118 2870 5.68 17.16 6813 44.59
Cyclotol 1.75 8250
5245 2790 5.82 17.59 6652 43.35
HMX 1.89 9100
5553 2960 5.89 17.76 7103 44.00
Qv is the explosion heat of the explosive material.

E 2
138
Table 5-4 The jet output data and penetration results of CSC with 46
o
cone apex angle,
1.4mm liner of thickness using different filling explosive charges and 4mm steel casing
into 35 MPa concrete target.
Parameter
Explosive Type
TNT Cyclotol RDX HMX PE TN LX-17
Jet tip velocity (m/s) 6108 6652 6813 7103 6605 6046
Jet momentum (kg.m/s) 15.46 16.64 16.78 15.03 17.72 15.05
Jet tail velocity (m/s) 722 744 656 709 815 674
Penetration depth (cm) 60.96 64.38 71.20 74.88 72.90 66.80
Exit hole diameter (mm) 12.04 16.24 18.80 14.00 19.80 15.52
Note: LX-17 is a mixture of 92.5% TATB and 7.5% Kel F binder.

Figure 5-18 The relation between the jet tip velocity and the detonation velocity of the
used explosive.
The scaled jet tip velocity to the detonation velocity and the Gurney velocity of the
explosive are shown in Figure 5-19 as a function of detonation velocity of the used
explosive. It can be concluded that these ratios are nearly constant for the used six
explosives. The scaled jet tip to explosive detonation velocity ratio is 0.82, while the scaled
jet tip to the Gurney velocity ratio is 2.38. This indicates a nearly constant ratio of jet
velocity to the explosive detonation characteristics over a wide range of explosive
materials.
139

Figure 5-19 The ratio of the jet tip velocity to the detonation velocity and the Gurney
velocity of the used explosives.
5.6.2 Liner wall thickness effect
5.6.2.1 Uniform liner wall thickness
According to Zukas [6], the optimum liner wall thickness is 1-4% of the charge calibre.
The liner thickness values in this study range between 0.8mm and 3mm for the same OWP
of calibre 36mm, which are 2.2% and 8.3% of the charge calibre, respectively. The details
of copper liners and the jet output data together with their penetration results are listed in
Table 5-5 and illustrated in Figure 5-20. This table illustrates the dependence of jet
characteristics on the liner wall thickness. It can be observed that the decrease of the liner
wall thickness will reduce the liner mass. Hence, the corresponding jet velocity will
increase depending on the mass ratio between the liner and the explosive. The variations of
the jet velocity with its cumulative jet mass for the different liner wall thicknesses are
illustrated in Figure 5-22. It can be observed that the smallest liner thickness 0.8mm
exhibited the highest tip velocity, but has the lowest mass despite its highest jet to liner
mass ratio. This thickness does not give the maximum penetration depth, which can be
directly related to the jet momentum. On the other hand, the liner thickness of 1.4mm (i.e.
about 4% of the charge calibre) has achieved the largest penetration depth, which supports
Zukas [6]s recommendation of optimum liner wall thickness. The jet momentum is
correlated with the jet penetration depth as shown in Table 5-5.
140
Table 5-5 The produced jet characteristics and its penetration for 46
o
conical OWP for
different liner thickness with HMX explosive charge.
HMX
Mass
(gm)
M/C

Liner Geometry Output
Thick.
(mm)
Mass
(g)
Jet
mass
(g)
%
jet
Jet
tip vel.
(m/s)
Jet
momentum
(kg.m/s)
Pen.
(cm )
Hole
diam.
(mm)
50.74 0.37 0.8 18.9 3.9 20.6 8135 17.22 42.7 14.4
50.74 0.46 1.0 23.7 4.6 19.4 7619 18.27 56.1 15.0
50.74 0.55 1.2 28.3 5.3 18.5 7251 19.00 62.9 16.4
50.74 0.65 1.4 33.1 5.9 17.7 7103 19.56 74.8 14.0
50.74 0.81 1.8 41.5 6.9 16.6 5852 20.32 73.9 16.6
50.74 0.91 2.0 46.1 7.4 15.9 5533 20.52 71.0 10.2
50.74 1.29 3.0 65.5 9.1 13.8 4331 20.30 72.4 18.0

Figure 5-20 The penetration depth of 46
o
conical OWP for different liner thicknesses with
HMX explosive charge and steel casing thickness.
The penetration of a shaped charge OWP with different liner thicknesses as a function of
jet momentum is shown in Figure 5-21. The penetration relation seems to be directly
propotional to the jet momentum upto a certain value, after which the penetration decreases
due to the massive jet and the large diameter of jet produced from large thickness liners.
17.0
17.5
18.0
18.5
19.0
19.5
20.0
20.5
21.0
30
35
40
45
50
55
60
65
70
75
80
0.5 1 1.5 2 2.5 3 3.5
J
e
t

m
o
m
e
n
t
u
m

(
k
g
.
m
/
s
)
P
e
n
e
t
r
a
t
i
o
n

(
c
m
)
Liner wall thickness (mm)
Penetration
Jet momentum
141

Figure 5-21 The penetration as a function of the jet momentum

Figure 5-22 The jet tip velocities as a function of cumulative jet mass for different liner
wall thicknesses according to standard jetting analysis algorithm.
30
35
40
45
50
55
60
65
70
75
80
17.0 18.0 19.0 20.0 21.0
P
e
n
e
t
r
a
t
i
o
n

(
c
m
)
Jet momentum (kg.m/s)
142
5.6.2.2 Varied liner wall thickness
Effects of the tapered liner wall thickness on jet performance were also studied, in which
the wall thickness at the cone apex is different from that at the liner base. The two studied
shapes are illustrated in Figure 5-23 .

Figure 5-23 Liner walls with varied (tapered) thicknesses.
The jet output data and the penetration results for the two liner shapes are illustrated in
Table 5-6. Case (b) exhibited higher jet tip velocity but lower jet mass than that of case (a).
However, its total momentum is 15% lower than that of case (a), which explains the
difference between them in the achieved penetration depth.
Table 5-6 The jet output data and the penetration results for OWP with two different liner
wall thicknesses of nearly the same weight.
Varied thickness
1.1-1.6
Case (a)
Varied thickness
1.6-1.1
Case (b)
Mass of jet (g) 6.25 4.17
Jet % 18.80 14.63
Jet tip velocity (m/s) 6213 7050
Momentum (kg.m/s) 19.30 16.38
Penetration depth (cm) 61.40 57.00
r Hole diamete ) mm ( 13.80 14.76
5.6.3 Cone apex angle
The characteristics of the shaped charges jet mainly depend on the explosive to metal mass
ratio and the liner geometry (i.e wall thickness and its cone apex angle). The mass ratio
will normally be changed as the cone angle changes, which should be considered in the
143
study of the cone angle effect on the jet characteristics. To consider this effect, seven
shaped charge models with different cone apex angles, but a constant explosive to metallic
liner mass ratios (C/M) and a constant liner wall thickness were studied in order to
differentiate the effect of cone apex angle on the jet characteristics from other effects. The
seven different cone angles are 22, 32, 40, 46, 56, 60 and 70
o
, which were used to estimate
the produced jet characteristics and its efficiency using the standard jetting analysis, the jet
formation and penetration codes in Autodyn. The constant C/M ratio and the liner
thickness are 1.069 and 1.77mm, respectively.
5.6.3.1 Standard jetting analysis
The jet characteristic data according to the jetting analysis for different cone angles at
constant mass ratios are listed in Table 5-7. The jetting analysis shows that the jet tip
velocities for the 22
o
and 70
o
cone apex angles were 8243 and 5538 m/s, respectively.
However, the jet mass percentage in the case of 22
o
is less than that for 70
o
, which explains
the big difference on the jet kinetic energy between these two designs. Such difference is
attributed only to the cone apex angle effect because their mass ratios are kept constant. On
the other hand, the calculated maximum collapse velocities of both models were 2136 and
1755 m/s, respectively, which have direct influence on the jet formation. This comparative
study supports the theory, which states that narrow cone angles produce fast jet but with
lower jet mass.
Table 5-7 Effect of the cone apex angle on the jet characteristics at the same explosive to
metal mass ratios (C/M = 1.069, RDX to Copper liner mass ratio).
Apex
angle

Liner
thick.
(mm)
Liner
mass
(g)
Jet
mass
(g)
% jet
Jet tip
velocity
(m/s)
Jet K.E.
(kJ)
22 2.72 24.53 1.54 6.27 8243 18.25
32 2.34 45.43 1.87 4.12 6674 38.38
40 1.87 43.53 4.04 9.27 6500 52.62
46 1.79 36.94 3.82 10.34 6088 41.36
56 1.77 29.62 3.37 11.36 5855 39.66
60 1.74 26.62 3.40 12.77 5603 36.30
70 1.47 27.63 4.30 15.56 5538 44.58
For the constant liner wall thickness of 1.77, a liner with cone apex angle of 22
o
is
expected theoretically to achieve a deep penetration depth due to its high velocity; but the
mass ratio of the produced jet is only 6.15% of the total liner mass as shown in Table 5-8.
The model with cone apex angle of 40
o
produces a jet with a tip velocity of 6995 m/s and a
144
jet mass ratio of 9.89%. As a result, the total kinetic energy of this jet is more than twice
that of the former one. The dependence of the tip velocity of the jet resultant from the same
liner wall thickness of 1.77 for the entire apex angles is illustrated in Figure 5-24.
Table 5-8 Effect of the cone apex angle on the jet characteristics for the same 1.77mm liner
wall thickness and 6mm steel casing.
RDX
mass
(gm)

C/M

Apex
angle

Liner
thick.
(mm)
Liner
mass
(g)
Jet
mass
(g)
% jet
Jet tip
velocity
(m/s)
Jet
K.E.
(kJ)
26.25 1.33 22 1.77 19.67 1.21 6.15 9531 25.37
66.04 1.72 32 1.77 38.30 2.54 6.63 7629 48.29
56.47 1.33 40 1.77 42.34 4.19 9.89 6995 54.77
48.47 1.29 46 1.77 37.66 4.05 10.75 6097 42.59
31.66 1.38 56 1.77 29.62 3.37 11.36 5855 39.66
37.99 1.49 60 1.77 25.43 3.05 11.99 5554 33.03
29.53 0.89 70 1.77 32.97 5.02 15.22 4862 43.10

Figure 5-24 The jet tip velocity as a function of cone apex angle with uniform liner wall
thickness of 1.77mm.
5.6.3.2 Jet formation and penetration calculations
For uniform liners, as the cone angle widens, the jet becomes shorter, thicker, and less
penetrative [26], which can be illustrated by Table 5-9 and Figure 5-25 obtained from the
jet formation and penetration results. The smallest apex angle gives the fastest jet tip
145
velocity as illustrated in Figure 5-26; but with the smallest jet mass. However, for the
wider angles, the charge performance is better than the small apex angles when they have
the same mass ratio between explosive and liner as illustrated in Table 5-10. This result
illustrates the effect of the liner thickness on the penetration and jet characteristics.
However, this parametric study is not sufficient to judge the best design, therefore, an
optimization study for the cone apex angle and the liner thickness will be performed to
achieve the largest penetration depth with the lowest mass of explosive.
Table 5-9 The jet characteristics and penetration results of OWP of 1.77 mm liner
thickness for different cone apex angles.
Cone Apex
angle. (deg.)
Liner
thickness
(mm)
Code output
Jet tip
velocity
(m/s)
Jet tail
velocity
(m/s)
Penetration
depth (cm)
Hole
diam.
(mm)
22 1.77 9531 4254 50.48 12.8
32 1.77 7629 2614 69.72 12.4
40 1.77 6995 2560 64.72 14.4
46 1.77 6097 1023 62.88 17.8
56 1.77 5855 1172 60.96 18.3
60 1.77 5554 1921 58.00 15.0
70 1.77 4862 1753 54.64 13.6

Figure 5-25 The calculated penetration depth for OWP at different cone apex angles.
30
35
40
45
50
55
60
65
70
75
15 25 35 45 55 65 75
Cone apex angle (deg.)
P
e
n
e
t
r
a
t
i
o
n

d
e
p
t
h

(
c
m
)
Same mass ratio
Same liner thickness 1.77mm
146

Figure 5-26 A comparison between the jet tip velocity for different apex angles at both the
same mass ratio and the same thickness.
Table 5-10 The jet output data and penetration results for OWPs with different liner cone
apex angles and the same explosive to metal mass ratio ( C/M = 1.069).
Cone apex
angle.
(deg.)
Liner
thickness
(mm)
Code output
Jet tip
velocity
(m/s)
Jet tail
velocity
(m/s)
Penetration
depth (cm)
Hole
diam.
(mm)
22 2.72 8243 2774 38.36 8.2
32 2.34 6674 2896 47.40 21.4
40 1.87 6500 1724 51.80 17.8
46 1.79 6088 1400 59.04 16.4
56 1.77 5855 1172 60.96 18.3
60 1.74 5603 2033 59.00 18.5
70 1.47 5538 2030 58.16 19.4
5.6.3.3 The optimization of the cone apex angle and liner thickness parameters
The objective of this optimization is to obtain the maximum penetration depth into
concrete with the convenient explosive mass and apex angle of the cone. Table 5-11 lists
the input parameters for the design expert software used to do the optimization
calculations. The selected effective design parameters in the optimization are the cone apex
angle, its liner wall thickness and the masses of both the liner and the explosive material.
The response parameters that will be considered are the penetration depth (to be
maximized) and the explosive mass (to be minimized).

147
Table 5-11 The input factors and their response values for the optimization study.
Factors
Responses
Apex angle
(deg.)
Liner thick.
(mm)
RDX mass
(gm)
Penet. depth
(cm)
22 2.72 26.25 38.36
32 2.34 66.04 47.40
40 1.87 56.47 51.80
46 1.79 48.47 59.04
56 1.77 41.06 60.96
60 1.74 37.99 59.00
70 1.47 29.53 58.16
22 1.77 26.25 50.48
32 1.77 66.04
69.72

40 1.77 56.47 64.72
46 1.77 48.47 62.88
56 1.77 41.06 60.96
60 1.77 37.99 58.00
70 1.77 29.53 58.64
The goals, the importance and the boundary constrains of the studied factors are listed in
Table 5-12, in which the penetration depth is set to be the most important objective design
response, while the next one to be considered is the explosive charge mass.
Table 5-12 The input constrains, the governing limits and the response importances.
Name Goal Lower limit Upper limit Importance
Cos ()* is in range 0.819 0.982 ++
Liner thickness (mm) is in range 1.47 2.72 ++
Explosive mass (g) Minimize 26.25 66.04 +++
Penetration depth (cm) Maximize - - +++++
Note: is half of the apex angle of the conical liner
Table 5-13 summarizes the results that were obtained from the optimization run. The thirty
solutions in this table are arranged according to their desirability, which is observed to be
almost unity for the whole range. This represents a high degree of accuracy between the
expected response calculated by the statistical objective function based on the fitting data
of the input factors and that presented as a real experiment. In general, desirability value of
zero respresents a completely undesirable response, while the desirability value of unity
represents an ideally desirable response. However, the manufacturing capability of the
suggested optimum designs should be considered from the manufacturing point of view.
148
For instance, the suggested designs that have a liner wall thickness greater than 2mm
should be ignored if the spinning is the method that will be used to manufacture the liner
with the small perforator dimensions. This is because of the difficulty of machining the
thick liners with small details and high precesion. Generally, the first two designs are
considered as the optimum designs that can be easily manufactured due to the facts that
their angle and their liner thickness can be done easily without further manufacturing
limitations. The combinations between the two columns, which are the liner thickness and
the Cos () will be changed if the used exolsive amount is greater than 26.25g. This means
that different varieties for liner thickness and Cos () will be obtained depending on the
used explosive mass. Similarly, different desirability values corresponding to different
explosive charges will be obtained.
Table 5-13 The optimum solutions and their corresponding desirability calculated by the
steepest slope optimization.
Number
Angle
(2)
Cos ()
Liner thick.
(mm)
Explosive
mass (g)
Penetration
depth (cm)
Desirability
1 56.96 0.879 1.49 26.25 100.23 1.00
2 43.13 0.93 1.58 26.25 102.50 1.00
3 58.15 0.874 2.05 26.25 76.05 1.00
4 54.25 0.89 2.41 26.25 132.49 1.00
5 26.69 0.973 1.54 26.25 128.67 1.00
6 63.36 0.851 2.49 26.25 219.49 1.00
7 32.11 0.961 2.67 26.25 71.41 1.00
8 39.90 0.94 1.63 26.25 90.19 1.00
9 55.99 0.883 2.1 26.25 75.94 1.00
10 69.63 0.821 2.66 26.25 351.02 1.00
11 61.37 0.86 1.53 26.25 83.51 1.00
12 52.21 0.898 2.21 26.25 78.43 1.00
13 43.44 0.929 1.59 26.25 98.63 1.00
14 36.75 0.949 2.6 26.25 80.28 1.00
15 55.50 0.885 2.1 26.25 73.64 1.00
16 34.12 0.956 1.55 26.25 118.76 1.00
17 67.60 0.831 2.14 26.25 123.05 1.00
18 36.75 0.949 1.56 26.25 115.06 1.00
19 52.98 0.895 2.28 26.25 92.61 1.00
20 37.48 0.947 2.56 26.25 74.35 1.00
21 63.36 0.851 1.5 26.25 81.22 1.00
22 68.21 0.828 2.25 26.25 156.10 1.00
23 59.55 0.868 2.01 26.25 75.00 1.00
24 67.18 0.833 2.63 26.25 314.34 1.00
25 66.14 0.838 2.45 26.25 219.98 1.00
26 52.47 0.897 2.42 26.25 125.44 1.00
27 64.01 0.848 2.69 26.25 321.42 1.00
28 59.31 0.869 2.54 26.25 208.74 1.00
29 43.75 0.928 1.5 26.25 125.65 1.00
30 60.92 0.862 2.21 26.25 114.21 1.00
149
To illustrate the desirability of the solution with the selected parameters, a graph of the
tested factors was selected with the relevant desirability. The graphs illustrated in Figure
5-27 and Figure 5-28 show the two areas, in which the desirability could be very high (i.e.
close to the unity). The preferred two areas of the highest desirability are represented by
areas A and B, while any combinations of the liner thickness and the apex angle that lead
to regions C and D should be avoided.

Figure 5-27 2-D contours of the desirability with the liner angle and its thickness.

Figure 5-28 3-D surface of the calculated desirability for the optimization problem.
The two optimum areas of this design are shown in Figure 5-29 and Figure 5-30
considering the penetration response, where the penetration contours in the upper area
150
indicate that the increase in the penetration capability of the perforator demands the
increase in both the liner thickness and the cone angle. On the other hand, in the lower
area, the the penetration depth increases with the decreases in the liner thickness and its
angle, provided that the same amount of explosive remains unchanged (i.e. 26.25 gm). On
the other hand, the liner design including angle and thickness that can produce a
penetration in both the blue (right) and green (left) regions in Figure 5-29, should be
avoided.

Figure 5-29 The penetration depth 2-D contours with the optimization parameters (angle
and liner thickness) using the optimum (minimum) explosive mass 26.25gm.

Figure 5-30 3-D surface of the calculated penetration depth for the optimization.
151
5.6.4 Liner shape and its geometry
In order to investigate the effect of liner geometry, three different geometries of shaped
charge liners with nearly the same explosive and liner masses were used. One of these
geometries is the conical liner with apex angle of 46
o
with a liner wall thickness of 1.4mm.
The second shape is the trumpet liner with the same liner thickness of 1.4mm. The third
one is a bi-conical shape with a uniform liner wall thickness of 1.1mm. Figure 5-31 shows
the shapes of the three different liners and Table 5-14 lists the calculated jet characteristics
and their penetration depths into concrete targets.

Figure 5-31 The three liner shapes; (a) the conical liner, (b) the trumpet liner, and (c) the
bi-conical liner, all with uniform liner wall thickness.
Table 5-14 The jet and penetration characteristics of the three different shaped charge
liners.
Liner shape Conical Trumpet Bi-conical
Explosive mass (g) 50.74 50.74 50.74
Liner mass (g) 29.32 28.93 28.78
Jet mass (g) 3.30 4.96 5.40
Jet to liner mass (%) 11.26 17.14 18.76
Jet tip velocity (m/s) 7103 7853 8244
Jet K. E. (kJ) 44.00 51.40 54.11
Penetration depth (cm) 74.88 81.00 87.00
Exit hole diameter (mm) 14.0 11.0 6.6
152
The changes in the jet velocity and kinetic energy indicate why the bi-conical liner has
achieved the maximum penetration depth. Nevertheless, other factors may need to be
considered in the practical design such as the charge length and the manufacturing cost.
5.6.5 Explosive amount and head height
The jet velocity and the damage caused in the rock formation depend mainly on the
amount of the explosive used in the shaped charge. Minimizing this damage is a key
objective when completing the well using OWP. This effect is studied using four similar
perforator designs, but with different explosive masses. The liner materials were copper
with the same design. A sketch of the four charges is illustrated in Figure 5-32. The jet
characteristics and its penetration capability into the standard concrete for the four targets
are listed in Table 5-15.

Figure 5-32 The four OWP with different explosive masses.
Table 5-15 The amount of the explosive and its impact on the jet and the penetration depth.
Case A Case B Case C Case D
Explosive mass (g) 50.74 46.30 40.00 24.57
Liner mass (g) 29.32 30.15 32.04 33.00
Jet mass (g) 3.30 4.06 6.05 5.5
Jet to liner mass (%) 11.26 13.47 18.89 16.66
Jet tip velocity (m/s) 7103 6628 4851 4539
Jet K. E. (kJ) 44.00 38.88 19.8 17.20
Penetration depth (cm) 74.88 69.00 61.00 41.00
The damaged areas around the perforated tunnels in concrete targets are shown in Figure
5-33 for different perforators. The damaged areas along the crater profiles near the impact
153
areas are similar, but the overall crushed zone thicknesses exhibit different values for the
four cases, which implies that the flow productivity will be affected by the amount of the
used explosive. However, these qualitative simulations are not sufficient to calculate the
well productivity because it demands further permeability calculations, which are not
available in the Autodyn hydro-code.

Figure 5-33 The damaged areas around the penetration path using different explosive
masses.
5.6.6 Water stand-off distance
In order to properly fit the gun carrying the shaped charge perforators inside the casing, as
shown in Figure 5-34, and since the jet travelling distance has a great effect on its
penetration capability, the water stand-off distance effect was studied to find its influence
on the jet performance and the depth of penetration.
To study this effect, the jet produced from OWP 46
o
cone angle, HMX main explosive and
liner thickness of 1.4mm is studied. This jet penetrates concrete after passing through
water layers of different thicknesses of 0.5, 1.7, 2, 4 and 6cm. The penetration depths and
the hole diameters relevant to the different water stand-off distances are illustrated in
Figure 5-35.
154

Figure 5-34 The OWP charge fitted inside the gun carrier and water stand-off distance
measured from gun casing wall.

Figure 5-35 The penetration depth and hole diameter for OWP detonated at different water
stand-off distance.
From this figure, it can be found that at 1.7cm stand-off distance, the perforator achieves
the maximum penetration depth. This may be attributed to the fact that at this short stand-
off distance, the jet is not fully stretched, which affects its penetration capability. After
3cm stand-off, the jet has to travel long distance in water, which causes the jet to be
155
particulated into small fragments or to be eroded and thus, its penetration capability
decreases beyond this distance.
5.6.7 Degree of confinement effect on the jet parameters
Unlike shaped charges, OWP casing imposes a thick confinement, which affects the jet
parameters, such as the jet tip velocity, especially in the region close to the liner base. This
effect may be attributed to the reflection of the detonation waves on the casing surface
back into the explosive, which may meet the liner with different incident angles between
the detonation front and the liner wall. The subsequently reflected waves can produce
regions of high pressure on the liner surface resulting in a jet with higher velocity. This
was verified by adding 7 gauge points to the explosive-liner interface as shown in Figure
5-36 and using OWP with cone apex angle of 40
o
, liner thickness of 1.4mm and casing
thicknesses of 1, 2, 4, 6 and 8 mm.
It can be observed that both the obtained pressure-time histories of the two cases (i.e. 1 and
8mm casing thicknesses) have nearly the same pattern. However, the impulse-time
histories in Figure 5-37 explained the reason why the obtained collapse velocities of all the
relevant jet elements in the entire five models are different from each other, which in turn
gave different jet velocities of their elements. Therefore, the jet tip velocity for 8mm case
is higher than that for 1mm case. All the jetting analysis data for the casing thicknesses
study are listed in Table 5-16.
All the resulting jets are coherent because their maximum flow velocities are lower than
the bulk speed of sound of the copper material which is 3940 m/s [135]; where the
maximum calculated flow velocity is 3748 m/s for 8mm casing. This means that all the
selected casing thicknesses are suitable to produce a coherent jet. But, an optimization
should be done based on the lowest casing thickness that is capable of confining the OWP
explosive charge and protecting it against premature explosion.

Figure 5-36 Different fixed target points along the liner axis to predict the P-t history on
the explosive-charge interface using 8mm casing wall thickness.
156

Figure 5-37 The predicted pressure and impulse-time histories for both 1mm casing
thickness (left) and 8mm casing thickness (right).
Table 5-16 The jetting analysis data obtained from the jetting analysis of OWP using RDX
main charge with different casing thicknesses.
Casing
thick.
(mm)
Apex
angle
(deg.)
Thick.
(mm)
Mass
(g)
Jet
mass
(g)
% jet
Max
flow
vel.
(m/s)
Max.
coherent
flow vel.
(m/s)
Jet
velocity
(m/s)
Jet
K.E.
(kJ)
1 40 1.4 36.64 5.65 15.41
3115 3940
6489.4 35.6
2 40 1.4 36.64 5.66 15.45
3214 3940
6540.3 39.1
4 40 1.4 36.64 5.69 15.54
3549 3940
6790.5 43.3
6 40 1.4 36.64 5.72 15.61
3671 3940
7035.5 49.2
8 40 1.4 36.64 5.76 15.71
3748 3940
7232.8 54.1

157
5.6.8 Effect of the initiation point on jet characteristics
The shape of the detonation wave when it meets the liner is so important that it can
determine the amount of the produced jet velocity and its degree of coherency after it
travels a certain distance. Therefore, an alternative side point of initiation was selected to
an OWP as shown in Figure 5-38. The corresponding detonation wave pattern at 1.2 s
from the moment of detonation is illustrated in Figure 5-39.
From the standard jetting analysis of 40
o
OWP with liner thickness 1.4mm and side
initiation point, the produced jet has a mass of 3.01 g and a velocity of 8018.5 m/s. The
same perforator with a normal central initiation point gives a jet mass of 2.325 g and a
velocity of 7812 m/s. Thus, the whole kinetic energy of the jet for the side initiation is
50.472 kJ, which is 13.3% greater than that of the OWP with normal initiation point, which
is 44.542 kJ. Thus, the predicted penetration depth when the detonation wave shape is
modified is expected to be better than that without wave shape modification. However, this
technique is difficult to be applied industrially, as the whole perforators cannot be
instantaneously detonated along their circumferential line.

Figure 5-38 Shaped charge with side point of initiation at time 0s.

Figure 5-39 The detonation pattern of the side initiation at 1.26s.
158
5.7 Liner portioning into jet and slug
In order to properly investigate the liner material portioning into jet and slug, two different
techniques were implemented. First one is done by dividing the copper liner material into
different colour tracers. The OWP used for this study has 46
o
liner cone apex angle and
1.4mm wall thickness filled with HMX explosive. The twelve tracer regions are illustrated
in Figure 5-40. Figure 5-41 illustrates the coloured tracers for the same liner material using
Autodyn jet formation simulation, while Figure 5-42 shows the different coloured contours
of the liner flowing into jet and slug portions at different times. The liner collapse figures
show that most of the first three tracer portions at the liner apex flow into the slug part,
while the twelfth portion and the liner base do not collapse down on the jet axis, and
therefore, they will not actually take part in the jetting. The jet is formed from other tracer
portions (i.e. from portion four up to eleven), but with different percentage from each
individual portion. The percentage of material flow into jet increases from the apex toward
the liner base, while their velocities decrease with their position in the same direction. The
jet tip is mostly composed of the four tracer regions (i.e. four up to seven), while the pile-
up or inverse velocity gradient part is observed near the jet tip. This part is formed because
the collapsed points from these regions do not have enough space to be accelerated to their
theoretical maximum values.

Figure 5-40 The twelve colours of the liner material used to track the liner portioning into
jet and slug.
159

Figure 5-41 Multi-coloured copper liner of OWP of 46 deg. cone apex angle and 1.4mm
liner wall thickness at time 0s.

Figure 5-42 The multiple-colours contours of the collapsed liner indicating the jet
formation from certain liner regions at different times.
The second method used to investigate the portioning of the liner material to the jet and the
slug is done using massive moving Lagrangian target points, which are located on the liner
material to facilitate its tracking. The first gauge is placed at the bottom (liner air interface)
and the next gauges are placed at selected spacing from each other, (e.g. 0.1 mm distance)
as illustrated in Figure 5-43. The output absolute velocity-time histories exactly specify the
profiles of collapsing velocities of the gauges forming both the jet and the slug. Figure
5-44 illustrates the absolute velocity-time histories of the selected nine gauges that depict
the material flow to form the jet, the slug and the infliction or collision point. This figure is
so important that it can be used with the multilayered or laminated liner material research,
where a coaxial or outer liner material can be added to delay the breakup of this jet, and
therefore, to increase its efficiency [136]. A sketch for the liner material portioning based
on the absolute velocity history is shown in Figure 5-45. This figure shows that by moving
160
from the liner apex to its base, the mass of the liner flows to form a jet increases, while its
velocity decreases. This conclusion can be implemented to study the effect of non-uniform
liner densitiy distribution produced by powder pressing technique on the jet characteristics
(i.e. velocity and mass distribution).

Figure 5-43 The selected moving target points on the liner axis to illustrated the liner
portioning into jet and slug.

Figure 5-44 Absolute velocity-time history plot for the nine moving gauge points used to
illustrate the liner partition into jet and slug.

161

Figure 5-45 A schematic diagram illustrating the jet and slug potions based on the
simulation results.
5.8 The Gurney velocity approximation
The Gurney velocity of an explosive is so important that it contributes directly to the
analytical calculations of the shaped charge jetting parameters. However, this value is not
known for all the well-known explosives, thus, it is linked directly to the Chapman
Jouguet-pressure-explosive impulse ratio as depicted in Figure 5-46 for several explosives.
The following relation was obtained from the fitting of the previously calculated Gurney
velocity against the P
CJ
/I
SP.
o
ratio.
V2E = .29 [
P
C]
I
SP
p
c
+9.7 5-1
where P
CJ
is the Chapman Jouguet pressure (Pa), I
sp
is the specific impulse of the explosive
used as a monopropellant (Ns/kg) and
o
is the explosive density (kg/m
3
).
If the impulse of the explosive is not known, it can be calculated using the detonation
velocity-impulse relation [137]:
I
sp
p
o
= 1
u
-198
1.S

5-2
where U
D
is the detonation velocity of the explosive in m/s,
o
is in kg/m
3
and I
sp
in Ns/kg .
The various explosives with their pressure, impulse and densities as well as the calculated
and the measured Gurney velocities and the deviation between them are listed in Table
5-17. The greatest deviation between the measured and the calculated Gurney velocities
based on Eq. (5-1) is -5.48%, which means that this approximation can be used accurately
over a wide range of explosives.
162

Figure 5-46 The Gurney velocity as a function of and the (P
CJ
/I
SP
o
) relation.
Table 5-17 The Chapman-Jouguet pressure, the specific impulse, the calculated and the
measured Gurney velocities and the deviation between them for various explosives.
Explosive
P
CJ

(GPa)
I
SP
(N.s/g)
o
(kg/m
3
)
V2E
Eq. (5-1)
(m/s)
V2E
(m/s)
Dev.
(%)
FEFO 25.00 2.389 1590 2555.4 2435.0 [72] -4.94
H6 24.00 2.147 1760 2497.6 2425.0 [72] -2.99
A3 30.00 2.636 1650 2634.7 2630.0 [72] -0.18
DIPAM 18.00 2.096 1550 2294.2 2175.0 [72] -5.48
C4 28.00 2.671 1600 2547.9 2660.0 [6] 4.21
HMX 42.00 2.614 1890 3037.0 2970.0 [22] -2.26
DATB 25.10 2.139 1788 2550.7 2560.0 [138] 0.36
NG 25.30 2.543 1590 2474.0 2548.7 [72] 2.93
OCTOL 34.20 2.500 1809 2801.4 2800.0 [22] -0.05
Cyclotol 31.60 2.508 1743 2717.8 2790.0 [22] 2.59
Comp B 28.70 2.434 1713 2631.1 2700.0 [22] 2.55
PBX-9011 34.00 2.425 1767 2894.9 2820.0 [6] -2.66
PBX-9501 37.00 2.579 1841 2859.3 2900.0 [139] 1.40
PBX-119 24.40 2.422 1635 2450.0 2509.7 [72] 2.38
PBX-9404 37.50 2.583 1844 2879.4 2900.0 [140] 0.71
LX-04 35.00 2.423 1865 2847.4 2776.0 [72] -2.57
LX-10 37.50 2.596 1860 2852.6 2922.1 [72] 2.38

163
5.9 Summary
Parametric analysis of OWPs is performed in this chapter, in which the used numerical
hydro-code is validated against standard jetting analysis, while the shaped charge jet
formation and penetration models are validated using the flash x-ray facility and the static
firing of OWPs against laminated target, respectively. The shaped charge design
parameters that include the explosive fill, liner thickness, charge casing and the mode of
initiation are studied using the jetting analysis and the jet formation algorithms. Besides,
the effect of the water layer stand-off distance that simulates the wellbore fluid on the
depth of penetration into concrete was also considered. Moreover, a simple relation among
the explosive inpulse, its Chapman-Jouguet pressure with the Gurney velocity is presented
to give a good approximation for the characteristic Gurney velocity of the explosive
materials.

Chapter 6: Influence of Target Strength and Confinement on The Penetration Depth of an OWP
164
CHAPTER.6 INFLUENCES OF TARGET
STRENGTH AND CONFINEMENT ON THE
PENETRATION DEPTH OF AN OIL WELL
PERFORATOR
6.1 Introduction
Oil well perforator (OWP) uses a shaped charge to open deep hole into the rock formation
in a productive oil field. Upon the detonation of an OWP, the high velocity metallic jet
perforates the carrier gun wall, wellbore fluid, pipe casing, and finally reaches the rock
formation that contains crude oil [11]. The productivity of the oil well increases with the
penetration depth. The penetration depth of an OWP depends on the design of a perforator
and the strength of rock material. Researches have been done to understand the effect of
the target strength on the penetration depth of a shaped charge jet into target material. Pack
and Evan [96] introduced a correction term related to the target strength in the
hydrodynamic formula of penetration depth, i.e.
2
(1 )
j
T j
Y
P L
V

=

6-1
where is a constant;
j
and
T
are the densities of jet (rod) and target materials,
respectively; L is the length of the rod penetrator; Y is the dynamic yield strength of the
target; V is the penetrator velocity. The correction term can be linked to an important non-
dimensional number in impact dynamics, i.e.,
2
D j
Y
J V

= where
2
j
D
V
J
Y
= is Johnsons
damage number [141-142]. It shows that the influence of the target strength on penetration
depth decreases with the increase of Johnsons number. For a steel target, this correction
term is around 0.3, which means that the penetration depth can be reduced by 30% due to
the effect of the target strength. This approximation has some limitations for jet penetration
as it was developed for a continuous rod projectile. Other parameters, such as the stand-off
distance and jet tip velocity, may also influence the penetration depth.
Extensive experimental results on shaped charge jet penetration were reported by
Eichelberger [97]. It was shown that the simple hydrodynamic equation is not valid in the
165
later stages of penetration when the jet velocity decreases. During the later stages of
penetration, strengths of both jet and target materials become relatively important.
Eichelberger [97] added two strength terms to the hydrodynamic equations to account for
their influences. The importance of the strength term effect on penetration was further
verified by Pugh [98] and Klamer [99] when they found that the penetration depths into
armoured steel are respectively 15% and 20% less than those in a mild steel target. Allison
and Vitalli [2] deduced three different models of shaped charge jet penetration based on the
assumption of the existence of virtual origin (VO) for a shaped charge, in which the
penetration depth for a continuous jet can be described by:
1
[( ) 1]
j
c
V
P Z
V
=

6-2
where Z is the effective jet length measured from VO to the target surface; V
j
and V
C
are
the jet tip and rear velocities, respectively; is the square root of the target-jet density ratio
(i.e. /
T j
= ).
Predicted penetration depths based on Eq.(6-2) agree well with the experimental results of
a 105mm shaped charge against monolithic metallic targets [2]. However, Eq.(6-2)
neglected the influence of the target strength on the penetration depth, and therefore, may
not be suitable for shaped charges used as OWP because the initial jet tip velocity of an
OWP may decrease after the perforation of multiple material layers before the jet reaches
the main target. The feature of the multi-layer target in the application of OWP is reflected
in the testing standard of American Petroleum Institute (API) [143]. Consequently, Eq.(
6-2) may over-predict the penetration depth due to the neglecting of target strength.
Therefore, it is necessary to understand the influence of target strength on the penetration
depth of a shaped charge jet in an OWP test. On the other hand, the main target in an OWP
application is subjected to large underground confinement pressure and the compressive
strength of the concerned quasi-brittle materials (i.e. rock, concrete) can be largely
enhanced by the confinement pressure, which will also be studied in this chapter.
Section 6.2 describes the experimental set-up and configurations of the shaped charge as
well as the standard OWP specimen. Section 6.3 introduces the numerical models, material
models and material parameters used to simulate the shaped charge jet and penetration.
Results will be presented in Section 6.4 with further analysis, which is followed by
conclusions in Section 6.5.
166
6.2 Experiments
The liners of the OWP used in this study were electrolytic Copper of grade C10100 OFEC
(Oxygen Free Electrolytic Copper). This material has a high purity (99.99%) and very low
oxygen and phosphorus contents for relatively high ductility, which is needed for the jet
material to sustain longer breakup time and have a better coherent performance [7]. The
copper liners were manufactured using the deep drawing technique, which is suitable for
OWP because this manufacturing method is economical and efficient for producing large
quantities of small calibre liners with a reasonable accuracy [31]. It starts with cutting a
circular copper disc and applying five steps of drawing by hydraulic press with an
intermediate annealing of 1000
o
C (two minutes) to decrease the strain hardening and
maintain the material ductility [31]. The liner has a small base diameter of 33mm, a cone
apex angle of 46 degree and a wall thickness of 1.4mm as illustrated in Figure 6-1. The
charge casing is steel with an average wall thickness of 4.5mm, while the main explosive
charge is PE4 with a total average mass of 40.0g and a standard deviation of 1.3g.

Figure 6-1 The shaped charge used in the concrete strength study (left) and a cross-section
of the liner (right).
Concrete targets with four different strengths were poured and cured according to the test
evaluation of the well perforator [144]. These concrete targets were tested according to the
167
standard OWP testing configuration and requirements in the Section-II of API-RP43 [143].
The configuration of the target layers and their dimensions and a picture of the
experimental setup are illustrated in Figure 6-2. The strengths of the standard concrete
cubes corresponding to concrete targets are 26.0, 40.0, 47.0 and 55.0 MPa with standard
deviations of 0.9, 0.9, 1.7 and 0.9 MPa, respectively, measured at 28 days after their
casting [145].

Figure 6-2 The layout and the experimental test setup according to API-RP43.
6.3 Numerical models
6.3.1 Hydro-Code Algorithms
The hydrocode algorithms were presented in details in Chapter 4.
6.3.2 Mesh sensitivity
It is well-known that the shape and the density of the mesh may affect the simulation
results. Generally, simulation with fine meshes produces more accurate solution with the
cost of longer time consumption in comparison with coarse meshing simulations. When the
erosion criterion is applied, effect of the mesh density on simulation results may increase.
168
In order to study the mesh sensitivity on the jet penetration, nine different mesh densities
were proposed for the concrete target material, while the Lagrangian jet meshes remain
unchanged (i.e. 0.5mm0.5mm). Uniform square meshes of 0.2, 0.3, 0.5, 0.8, 1, 1.5, 2, 2.5
and 3mm are selected for concrete target. The mesh sensitivity study was also performed in
the jetting analysis where five different Euler mesh sizes of 0.3, 0.6, 1.4, 2 and 4mm were
applied to PE4 explosive to examine the variation of the jet characteristics with the mesh
density.
Mesh sensitivity for Euler jetting analysis is shown in Figure 6-3, where the relationship
between the cumulative jet mass and its axial X-position is shown. It can be observed that
the predicted curves for five different mesh sizes have nearly the same shape at the
beginning of the jet formation. Then, noticeable variations among five curves occur for
different mesh sizes. However, with the decrease of mesh size from 4 mm to 0.3 mm, the
convergence of the solution is observed. Thus, 0.30.3mm cell is used in all jetting
analyses.

Figure 6-3 The cumulative jet mass versus the jet axial coordinate obtained from the jetting
analysis using different mesh sizes.
Figure 6-4 shows the mesh sensitivity for Lagrange penetration analysis where the
variation of the predicted penetration depth with mesh size is shown for a 40 MPa concrete
target with nine mesh sizes of unity aspect ratio. Penetration depth converges to a value of
0
1000
2000
3000
4000
5000
6000
7000
30 40 50 60 70 80 90 100 110
Jet X- coordinate (mm)
C
u
m
.

j
e
t

m
a
s
s

(
m
g
)
0.3x0.3mm
0.6x0.6mm
1.4x1.4mm
2x2mm
4x4mm
169
68cm using square shape element of size 0.2mm, however, the element size 0.5mm gave a
penetration depth of 67.5, which is only 0.7% different from that of the finest mesh, but it
save more than half the time needed to do the simulation with the element size of 0.2mm.
Thus, the Lagrange mesh size of 0.5mm0.5mm was applied to all penetration simulation
calculations considering its reasonable accuracy and time consumption.

Figure 6-4 The numerical penetration into concrete using different mesh sizes.
6.3.3 Material models
The material models were presented in detail in Chapter 4.
6.4 Results and discussion
The standard jetting analysis of the studied OWP indicated that the produced jet from this
perforator is coherent because the flow velocity satisfies the stability condition [83]:
,max
1.23
flow o
v C
6-3
where v
flow
,
max
is the maximum flow velocity of all liner material points and C
o
=3940m/s is
the sound speed in the copper material. The maximum flow velocity was found to be
3161m/s, which means that the produced jet will be coherent during its stretching. A
summary of the jetting analysis output is listed in Table 6-1.

20
30
40
50
60
70
80
0 0.5 1 1.5 2 2.5 3 3.5
N
u
m
e
r
i
c
a
l

P
e
n
e
t
r
a
t
i
o
n

d
e
p
t
h

(
c
m
)
Concrete meh i!e (mm)
170

Table 6-1 Jet characteristics based on the standard jetting analysis
Liner mass (g) 33.20
Jet mass (g) 6.02
Jet tip velocity (m/s) 6698
Jet tail velocity (m/s) 2054
Jet kinetic energy (kJ) 49.29

The jet elongation at different times is illustrated in Figure 6-5, which shows the jet
formation up to 18s (measured from the moment of detonation), at which the jet starts to
interact with the first steel layer of the laminated multi-layer target. The penetration stages
of the jet into 55MPa concrete target are illustrated in Figure 6-6, while Figure 6-7
illustrates the contours of concrete damage at different times due to the jet penetration. It
can be observed that the jet caused a radial damage along its penetration path into the
concrete, thus the penetration depth is measured experimentally based on the remaining
witness part of the concrete as shown in Figure 6-8.

Figure 6-5 The jet generation and stretching at different times.
171

Figure 6-6 The concrete penetration stages.

Figure 6-7 The concrete damage contours history.

Figure 6-8 The penetrated tested witness concrete targets and the steel discs.
172
The jetting data were used to calculate the effective jet length (Z), which is defined as the
distance from the source point or the virtual origin to the target surface. It can be calculated
by plotting the reciprocal of the jet velocity of each liner element against time, and
applying the back projection on the horizontal distance axis at the real interaction time as
illustrated Figure 6-9. The projected effective jet length was 127mm at 18s, at which the
jet impacts the first steel layer. However, this value can not be used directly with Eq.(6-2),
because the effective jet length and the jet tip velocity have to be modified considering the
thicknesses of the laminated steel and water layers. The jet tip velocity was corrected based
on the following equation for the exit jet tip velocity perforating a finite thickness target
[103]. This correction was derived from Eq.(6-2), in which the penetration P=T (i.e.
perforation of a finite thickness T), where V
C
is replaced by V
jex
.
i
i
jex jin
i i
Z
V V
Z T
=
`
+
)

6-4
where V
jex
and V
jin
are the exit and the input jet tip velocities respectively, Z
i
is the
effective jet length at the front of the target surface, T
i
is the target thickness and i refers to
the index of the target layer to be perforated. The values of the exit jet tip velocity and the
relevant effective jet length for the testing layers are illustrated in Table 6-2.

Figure 6-9 The effective jet length and the time relation for virtual origin model.
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 100 200 300 400 500
T
i
m
e

(
m
s
)
Distance from virtual origin to the target Z (mm)
"
t
e
e
l
t
a
r
#
e
t

l
o
c
a
t
i
o
n

173
Table 6-2 The jet tip exit velocity and the relevant effective jet length for the test layers.
Air Steel Water Steel
T
i
(cm) - 0.3 1.7 0.9
Effective jet length Z

(cm) 12.7 13 14.7 15.6
- 0.936 0.335 0.936
Exit jet velocity V
j
(m/s) 6698 6556 6320 5997

It is found that the real jet tip velocity just before impacting the concrete layer, V
j
, is
5997m/s while the corresponding effective jet length from the VO point to the concrete
target is 15.6cm, based on which the penetration depth can be calculated according to
Eq.(6-2) (89.78cm in this case). However, Eq.(6-2) is unable to consider the influence of
target strength on the penetration depth. In Eq.(6-1), Pack and Evan [96] introduced a
target strength correction term in hydrodynamic penetration model. According to Taylor
expansion, when 1
j
c
V
V
,
1
1 1
1 1 1
j j j j
c c T c
V V V
V V V

| | | |
+ = +
| |
\ \
when the
quadratic and higher order terms are neglected, which reduces Eq.( 6-2) to
( )
1
( ) 1 1 1 1
j j j j
j c
c T c c T
V V
Z
P Z Z V t V t
V V V t

(
(
| |
( = + = (
|
(
( \

where
j
Z V t = is the distance of the jet tip to the virtual origin and
j c
V t V t L = is the current
length of jet. When 1
j
c
V
V
, the jet length L is a constant and Z . Therefore,
1
c
Z Z
V t Z L
=
and Eq.(6-2) can be reduced to the hydrodynamic equation, which is

Eq.(6-1) when the target strength is ignored. This simple analysis for the link between
Eqs.(6-1) and (6-2) implies that a same strength correction term can be introduced into the
Allison-Vitalli equation [Eq.(6-2)], i.e.
P = Z
0
__
I
I
c
]
1

-1_ (1 -

c
i
p
I
c
)

6-5
where
'
c
f is the compression strength of concrete, which is a function of the applied
hydrostatic pressure if a confinement is present; V
j
is the jet tip velocity corrected after
perforating steel-water-steel layers of the testing specimen (5997 m/s in the present case);
is a constant determined from the real experiments, which was found to be 200.31 as the
average of the four experimental tests.
174
When underground confinement is considered [46], the Drucker-Prager equation can be
used to calculate the equivalent compressive strength. When uniform lateral confinement is
considered (i.e.
2 3 r
= = , positive in compression), the compressive strength of
1

stress can be derived as:

'
2 2 2tan
tan 1
3 9 3
c c H
f f P
| | | |
= + +
| |
\ \
,

6-6
according to Drucker-Prager equation [146] where
c
f is the unconfined uniaxial
compression strength; P
H
is the applied hydrostatic pressure; is the frictional angle,
which was found to be 50 degree for concrete [146].
The second bracket in the modified Allison-Vitalli equation [Eq.(6-5)] represents the
penetration reduction due to the target strength effect. This equation was used to calculate
the penetration depth theoretically for the four concrete materials using the jet velocities
and the effective jet lengths predicted from jetting analysis. The predicted results from
Eq.(6-5), the measured penetration depth and the numerical simulation results based on
Autodyn are listed in Table 6-3 and illustrated in Figure 6-10. It shows that the maximum
difference between the analytical and the experimental results was 7.5%, which
demonstrates the validity of the analytical model. Also, the maximum difference between
the numerical and the experimental penetration is 8.8%. Therefore, the numerical
prediction of the penetration depth when underground confinement exists could be used to
assess the validity of Eqs.(6-5) and (6-6).
Both numerical results and Eqs.(6-5) and (6-6) indicate that the target material strength has
a significant effect on the penetration depth of the OWP jet. When lateral confinement is
absent, Eq.(6-5) gives reasonably good predictions when compared with experimental
results where the maximum difference between the measured and the calculated
penetration depths was 7.52%. With the correction term in Eq.(6-5), the effect of the
underground confinement can be taken into account through Eq.(6-6), which agrees with
numerical predictions. In addition, the experimental penetration depth was found to
decrease about 0.73cm per 1 MPa increase of the compressive strength of the target, which
can be extended to situations when the underground confinement of oil field is considered.
Besides, the rate of penetration decrease according to the curve fitting of the data obtained
from Eq.(6-5) is 0.48cm per 1 MPa increase of the compressive strength of the target.

175
Table 6-3 The penetration results into concrete materials with different strength values.
Concrete
unconfined
strength
(MPa)
*Hydrostatic
pressure
(MPa)
Equiv.
Strength
Eq. (6-6)
(MPa)
) cm ( Penetration depth
The
correction
term
(
'
2
c
j C
f
V
)
Simulation Exp.
Cal.
Eq.
(6-5)

26 - 26.0 81.0 78 77.32 0.139
40 - 40.0 68.0 73 70.62 0.213
47 - 47.0 62.0 68 67.27 0.251
55 - 55.0 57.0 59 63.44 0.293
26 40 82.2 50.8 - 50.40 0.439
40 40 87.9 50.0 - 47.71 0.469
47 40 90.7 45.0 - 46.36 0.484
55 40 93.9 40.0 - 44.82 0.501
26 68 132.5 34.3 - 26.33 0.707
40 68 138.1 31.2 - 23.64 0.737
47 68 140.9 29.5 - 22.29 0.752
55 68 144.1 24.0 - 20.75 0.769
*The hydrostatic pressure (i.e. 68MPa) was taken from Ref [46] at a depth of 3km.

Figure 6-10 The penetration depth dependence on the concrete equivalent strength based
on Eq.(6-5).

176
6.5 Summary
The strength and confinement effects on the OWP jet penetration into standard laminated
specimen are studied experimentally and numerically. It is found that the strength of the
target can largely reduce the penetration depth of OWP jet. Allison-Vitalli formula of jet
penetration depth is modified to include the target strength effect using Johnsons damage
number. Furthermore, the effect of the underground confinement pressure on the target
compressive strength is considered using Drucker-Prager model and is introduced into the
modified Allison-Vitalli equation, which can be easily applied to estimate the OWP jet
penetration depth in an underground oil formation.

Chapter 7: Performance of Zirconium Jet with Different Liner Shapes
177
CHAPTER.7 PERFORMANCE OF
ZIRCONIUM JET WITH DIFFERENT LINER
SHAPES
7.1 Introduction
Oil well perforator (OWP) has been used in oil and gas wells to connect them to the
reservoir [11]. When the OWP is detonated, the hypervelocity jet can achieve a very deep
penetration depth into the geological formation material. The velocity and the diameter of
the jet depend mainly on the design of the perforator, specifically the liner shape, which
has a direct influence on the elemental velocities and their collapse angles. The collapse
velocity is an explicit function of the mass ratio between explosive and liner element [6,
68, 147], but its real value may be limited by the short distance available between the liner
element and its axis near the apex potion, as shown in Figure 7-1. Figure 7-2 depicts the
flow velocity (V
2
) in a moving coordinate system with a stagnation velocity (V
1
) and their
relationship with the collapse velocity (V
o
).

Figure 7-1 A schematic drawing illustrates the collapse process path from the initial liner
position to its axis.
178

Figure 7-2 Velocity vectors in a moving coordinate system [50].
According to the unsteady state Pugh-Eichelberger-Rostoker (PER theory) analysis,
stagnation (V
1
) and flow (V
2
) velocities can be determined by [50]:

1
v
c
cos([-(u+6))
sIn[

7-1

and
2
v
c
cos (u+6)
sIn[
,
7-2
respectively. The jet velocity V
j
is determined by:
I
= I
1
+I
2
=
I
o
sin[
(cos([ -o -o) +cos(o +o))
7-3
where 2 is the cone angle, is the deflection angle and is the collapse angle, as shown
in Figure 7-1 and Figure 7-2. Based on PER model [50], the deflection angle is given by
o = sin
-1
(
I
o
2u

)
7-4
where U=U
D
/cos, U
D
=8200m/s is the detonation velocity for PE4 explosive charge.
The mass of the jet element m
j
has also a direct relation with the collapse angle [50], i.e.
m
=
1
2
m(1 -cos [)
7-5
179
According to Eq.(7-3), the jet velocity increases with the increase of V
o
and with the
decreases of the collapse and cone angle and 2. On the other hand, according to
Eq.(7-5), the jet mass increases with the increase of . Thus the collapse angle has
opposite influences on the jet velocity and jet mass. Furthermore, when is very small and
the flow velocity V
2
is relatively high, the jet may not form according to [83]. Therefore, a
combination of V
2
and determines whether a jet can be formed for each liner element.
Meanwhile, the flow velocity V
2
also determines the coherency of the formed jet. The
jetting conditions had been studied by many reaerchers over the past decades. Walsh et al.
[82] concluded that the jetting always happens if the jet material is incompressible or if the
collision (flow) velocity in the moving coordinates system (V
2
) is subsonic. Cowan and
Holtzman [3] presented another overview for the jetting condition criteria in the explosive
welding applications. Chou et al. [83] summarized the jetting conditions and the cohesion
characteristic of the produced jet as shown in Table 7-1, where
c
is the critical collapse
angle for an attached oblique jet at a given flow velocity.
Table 7-1 The condition for the jet formation and the state of its cohesion at different
collision velocities and collapse angles [83].
Flow regime Impinging Angle Jet formation
Jet
coherence
Supersonic (V
2
> C
L
)
c
No No
>
c
Yes No
Subsonic (V
2
C
L
) All values Yes Yes
In Table 7-1 C
L
is the longitudinal sound speed in the solid liner material,
L
1-
1+
o

7-6
in which, Co = _
K
p
o
, where
o
is the jet density, K =
E
S(1 -2:)
,
is the bulk modulus,
E is Youngs modulus and is the Poissons ratio. The conditions in Table 7-1 have been
confirmed by Harrison [148], and Walker [149] experimentally where flash x-ray was used
to show that a coherent jet is formed when V
2
<C
L
. Therefore, different regions on -V
2

domain can be determined for the jet formation/coherency for a given shaped charge,
which is useful for the design of shaped charge.
Liner shape has significant influence on the shaped charge jet performance. It has been
shown that the conical liners with small apex angles produce relatively deep crater with
180
small diameter. On the other hand, the hemispherical liners produce shallow crater with
large diameter [11]. Various improvements of the liner design have been done in the past
fifteen years. For example, Davinson and Prat [17] and Lee [11] proved that modifying the
liner shape design can increase the jet kinetic energy and hence the penetration depth. Held
[150] used a special flat liner to obtain a superfast jet of tip velocity of 25km/s. Therefore,
it is necessary to understand the jet characteristics for different liner shapes and their
corresponding penetration performance.
This chapter will study the jet formation, coherence and penetration of four commonly-
used liner shapes, i.e. conical, hemispherical, trumpet (or bell shape) and bi-conical shapes,
in which explosive mass and outer diameter are kept constant. Conical liner will be treated
as a baseline. The enhanced flow velocities and collapse angles for these zirconium linear
shapes are discussed based on the conditions of jet formation and coherency in Table 7-1.
The performance of the formed jets is characterized by their penetration capability into the
standard target in comparison with the penetration capability of the conical liner jet. These
liners are tested experimentally against the laminated steel-water-steel-concrete standard
target according to API-RP43 (Section II) [143]. Calculations of jetting, jet formation and
penetration of four liner designs are performed using the hydro-code Autodyn.
Section 7.2 gives the conditions of jet formation and coherency for the zirconium liners.
Section 7.3 describes the liner manufacture and the experimental set-up. Section 7.4
introduces the numerical models, material models and material parameters used to simulate
jet formation and penetration of the shaped charges. Results with further analyses are
presented in Section 7.5 followed by conclusions in Section 7.6.
7.2 Critical angle calculations conditions for the zirconium jets
It is well known in gas dynamics that for a flow of free stream velocity V
2
impinging on a
solid wall, there is a maximum angle
c
, above which an attached shock wave cannot exist
as depicted in Figure 7-3 [83]. This mechanism is also applicable to the shaped charge jet
formation [82-83].
Relationships between the critical collapse angle and the flow velocity for some liner
materials apart from zirconium have been determined analytically in [3]. The
c
-V
2

relationship is important because it defines the boundary between jetting and no jetting
181
regions on -V
2
domain, as shown in Figure 7-4. Together with the C
L
limit, three regions
can be defined on -V
2
domain, which is explained in Figure 7-4 for the copper (Cu-Cu)
liner (C
L
=4.84 km/s), i.e. (a) Region-I: the region on the left of C
L
limit where coherent jet
is formed; (b) Region-II: the region on the right of C
L
limit and above the
c
-V
2
curve
where non-coherent jet is formed; (c) Region-III: the region on the right of C
L
limit and
below the
c
-V
2
curve where jet cannot be formed.

Figure 7-3 The flow configurations in the supersonic regimes detached and attached shocks
[83].

Figure 7-4 The calculated critical angles for different liner materials at different flow
velocities [3].
182
Two methods will be used in this chapter to calculate the critical collapse angles at a given
flow velocity.
(i) Method 1:
The analytical model presented in [3, 82], employed the momentum balance to obtain the
critical collapse angle. According to [82], the critical collapse angle is given by
tan
2
[ =
P jp
o
I
2
2
[
p
p +1
-P

[
(p
o
I
2
2
-P)
2
,
7-7
where P is the pressure,
o
is the initial liner density, is the compressibility (i.e. =/
o
-
1).
The maximum angle can be determined for a given impinging velocity V
2
from the
condition [ p / = at =
c
according to [82]. Thus, for =
c,
dP
d
P|P-p
c
v
2
2
]
(+1)|p
c
v
2
2
-P(+2)]

.
7-8
Eq. (7-8) together with the equation of state (EOS) of the liner material can be used to
calculate the critical angles
c
at different values of V
2
. Shock EOS takes the form of [116]:
pC
0
2
(+1)
(1-(-1))
2

7-9
where C
o
is the sound speed of the liner material and S is the slope of the shock speed-
particle velocity line. For the zirconium material, C
o
= 3757m/s and S=1.018 [116].
Differentiating Eq.(7-9) with respect to gives
dP
d
(1-(-1))pC
0
2
[(1-(-1))(2+1)+2(
2
+)(-1)
(1-(-1))
4
.
7-10
For simplicity, assume S1, therefore Eq.(7-10) reduces to:
dP
d
o
2
.
7-11
The critical compressibility (i.e.
c
) and the corresponding critical pressure P
c
can be
obtained from Eqs. (7-8), (7-9) and (7-11). Therefore,
c
can be derermined from Eq.(7-7).

The detailed steps for the calculation procedures are illustrated in Appendix A.
183
(ii) Method 2:
The second method that was used to calculate the
c
-V
2
curve is the numerical method
using Autodyn hydro-code with the Euler solver, which was used to simulate the
impinging of the liner with its axis of symmetry, as shown in Figure 7-5. The jet formation
was validated by Refs. [133] and [135], where the obtained features of the jet using the
Autodyn Euler solver were supported by using the flash x-ray photograph. The problem
was approximated by solving a transient oblique impact model with proper initial
conditions assuming a steady-state flow configuration. The initial model is a zirconium
liner of wall thickness 2mm moving towards its axis at a uniform constant free stream
velocity V
2
impacting a rigid boundary at constant angle , as depicted in Figure 7-5. The
rigid boundary is the axis of symmetry for the axisymmetric 2D model [83] and the initial
pressure throughout the material is zero. Various combinations of flow velocities and
collapse angles were used to sufficiently cover all three regimes, i.e., subsonic, supersonic
jetting, and supersonic non-jetting. These calculations were performed for the
axisymmetric Euler configurations, in which the jetting and non-jetting criteria will be
identified. The flow velocities that were tested are 3, 4, 5, 6, 7 and 8km/s, while the tested
collapse angles are 2.5, 5, 7.5, 10, 12, 15, 17 and 20 degrees. The obtained
c
-V
2
curve for
the zirconium metal liner can be used to design the liner shape to form a coherent jet.

Figure 7-5 The flow configuration Autodyn 2-D model used to estimate the critical angle
of jetting.
According to Eqs.(7-2) and (7-4) and the jet formation condition V
2
C
L
, we have

[
c1
sin
-1
_
I
o
C
L
cos _o +sin
-1
(
I
o
2u

)]]
7-12
184
where
c1
is the theoretical critical collapse angle (i.e. the minimum collapse angle) for the
formation of a coherent jet.
For zirconium liner and PE4 explosive, C
L
=4566m/s, U=U
D
/cos with U
D
=8200m/s and
=23
o
, Eq. (7-12) becomes:
[
c1
sin
-1
_
I
o
66
cos _2S +sin
-1
(
I
o
17816
)]] .
7-13
On the other hand the
c
-V
2
curve can be calculated based on analytical or numerical
methods introduced before, which together with Eq.(7-2) can define another critical
collapse angle as a function of collapse velocity V
o
. As an example, we used the numerical
method (i.e. Method-2) to determine the
c
-V
2
curve empirically (i.e. from the fitting of the
c
-V
2
curve), for the zirconium liner as shown in Figure 7-6.

Figure 7-6 The relation between the critical angle and the flow velocity calculated
numerically (i.e. method 2).
The relation between V
2
and
c
is
sin[
c
= _
I
2
1
-.S621] .
7-14
185
Substituation the value of the V
2
from Eq. (7-2) into Eq. (7-14) gives
sin[
c
= _
I
o
cos(o +o)
1sin[
c
-.S621_.
7-15
Solving Eq. (7-15) for the value of
c
in 0<
c
< gives
[
c
= sin
-1
_
.S277 +
I
o
1
4
cos(2S +sin
-1
(
I
o
17816
)) -.181_ . 7-16
Plotting Eqns.(7-13) and (7-16) for different values of collapse velocity V
o
, three regions
were defined in Figure 7-7, which can be compared with those corresponding regions in
Figure 7-4. The advantage of Figure 7-7 is that it can be used to check the jet formation
and coherence directly from the jet collapse velocity V
o
, which is determined by the liner-
explosive mass ratio and the used explosive, without further need of calculating flow
velocities.

Figure 7-7 Variations of
c1
and
c
with collapse velocity.

186
7.3 Experiments
Amongst different tested liner materials, zirconium exhibited the longest cumulative jet
length and the highest ductility factor, which are necessary to sustain longer breakup time
and to achieve larger penetration depth [13]. In this study, the zirconium liners were
manufactured by high precision CNC machine (i.e. the precision of 5m). The row
material was a solid cylinder of pure zirconium 4N (99.9951) having a diameter of
46.17mm and a length of 99.89mm with a density of 6623kg/m
3
. The impurity percentages
of the zirconium material are listed in Table 7-2. The zirconium rod was annealed to 900
o
C
for one hour before machining in order to obtain a relative small average grain size, hence
to increase its ductility, which in turn increases its breakup time and improve the liner
performance [151]. The row material of the zirconium rod and the manufactured liners are
Table 7-2 The elemental percentage of impurities in the zirconium material.
Element Impurities amount (%)
Fe 0.005
Cr 0.0009
C 0.001
N 0.008
The zirconium material has two problems with its machining. The first one is related to its
high tendency to work hardening during machining, while the second one is the possible
ignition of the fine chips that accumulate near the machining equipment [152]. To avoid
these problems, slow speed and heavy feed were applied with a continuous coolant supply
of water soluble oil lubricant to reduce temperature and prevent flammability of the fine
chips.
The PE4 explosive was used with the four shaped charges. PE4 is a RDX-based powerful
explosive (i.e. mass composition of 88% RDX and 12% plasticizer and other additives)
having a detonation velocity of 8027m/s at 1590kg/m
3
density [153] and 8200m/s at
1600kg/m
3
density [154]. It was chosen for its high performance and low sensitivity to
different kinds of stimuli (i.e. friction and impact). The assembly and set-up procedures
include: (i) fill PE4 into the steel casing; (ii) press the liner slowly against the steel casings
containing high explosives to expel air gaps inside the perforator charge; (iii) attach the
shaped charge to the upper steel layer of the test configuration in Figure 7-9.
187

Figure 7-8 The zirconium solid cylinder (left) and the manufactured liners; 1: conical; 2:
hemispherical; 3: bell and 4: bi-conical shape.
The concrete cylinders with the designated strength were cast in 1mm wall thickness PVC
tubes and allowed to cure according to the test evaluation of the well perforator [155].
These concrete targets were tested according to the standard OWP testing configuration
and requirements in the Section-II of API-RP43 [143]. The measured average strength of
the standard concrete cubes was 40.02 MPa with a standard deviation of 0.92 MPa,
measured at 28 days from their pouring day [145].

Figure 7-9 Dimensions of the test setup and the experimental test configuration
(1: Detonator; 2: Boaster; 3: OWP; 4: Front steel disc; 5: Concrete; 6: Power supply).
188
7.4.1 Methodology
General description of the used Autodyn algorithms was presented in Chapter 4.
7.4.2 Mesh sensitivity
In order to study the mesh sensitivity on the jet penetration, five different mesh densities
were proposed for the concrete target material, while the Lagrangian bi-conical jet meshes
remain unchanged (i.e. 0.5mm0.5mm) due to its sufficiently small dimensions. Uniform
square meshes of 0.25, 0.5, 1, 2 and 3mm are selected for the concrete targets. Figure 7-10
shows a sample of three different mesh sizes at the impact area, while Figure 7-11 shows
the concrete damage contours relevant to these meshes at 40s from the moment of impact.
It can be observed that the damage areas of the three mesh sizes are similar, but the crater
profiles indicate the main difference in the crater shape due to the different mesh densities.

Figure 7-10 The impact area of the jet-test layers modelled by jet solvers using three
different mesh sizes.
189

Figure 7-11 The damage contours near the impact surface for the three mesh sizes at 40s.
All the five simulation models were allowed to proceed until the final penetration is
achieved. This happens either when the jet is completely consumed or eroded on the crater
walls, or when the jet velocity decays below a certain value, at which no change in the
penetration is remarked with time. The total penetration depth for the five mesh sizes is
depicted in Figure 7-12.
There is a large difference of the penetration depths for coarse and fine meshes between
1mm and 3mm indicating the sensitivity of the penetration depth to the mesh size.
However, when mesh size is smaller than 1mm, this sensitivity is largely reduced, and the
penetration depth approaches to an asymptote. According to Figure 7-12, the penetration
modelling using 1mm mesh size is 2.75% different from that of 0.25mm mesh size while
the latter model costs five times computational time of the 1mm size model. Thus, it was
decided to use 1mm concrete mesh size for the rest of modelling to maintain reasonable
computation accuracy and time.
190

Figure 7-12 The Numerical penetration using different mesh sizes and experimental
penetration.
7.4.3 Material models
A general description of the material models used in this chapter is shown in Chapter 4.
7.5 Results
7.5.1 The
c
-V
2
calculations
For zirconium material, C
o
=3757m/s and =0.34. Therefore, the longitudinal sound speed
of the zirconium material is C
L
=4567m/s according to Eq.(7-6).
Three cases with fixed collapse angle of 12 degree and different flow velocities, i.e. Case
(I): V
2
=3.0 km/s, Case (II): V
2
=5.0 km/s and Case (III): V
2
=6.0 km/s, were simulated using
Autodyn hydro-code model in Figure 7-5. According to Table 7-1, Cases (I)-(III) belong to
coherent jetting, non-coherent jetting and non-jetting situations, respectively. Figure 7-13
(a) shows the cross-sections of the collapsed jet impacting on the symmetrical axis for
these three cases, in which, jetting [Cases (I) and (II)] and non-jetting [Case (III)] cases can
be easily identified as depicted in Figure 7-13 (b). The jet in Case (II) has a large number
of radially dispersed particles representing a non-coherent jet.
191
The analytical and numerical
c
-V
2
curves for zirconium liner are shown in Figure 7-14,
where the numerical
c
-V
2
curve for copper liner was shown as a reference. This figure
illustrates the boundary between jetting and non-jetting and the boundary between coherent
jetting and non-coherent jetting cases. When V
2
and

are calculated, this figure could be
used to determine the jetting formation and behavior in the design of zirconium liner.

Figure 7-13 (a) the cross-sections of the collapsed zirconium jet impacting on the
symmetrical axis at collapse angle of 12 degree and flow velocities of 3, 5 and 6km/s for
cases I, II and III respectively; (b) the corresponding regions on the
c
-V
2
curve.
The and V
2
values were obtained for four zirconium liner shapes from jetting analysis,
which are shown in Figure 7-15. This figure confirms that the four liner designs can
produce coherent jets and indicates the abnormal high collapse angles for the
hemispherical liner, which help forming massive explosively formed projectile (EFP)
rather than the traditional thin jet. This can be explained by Eq.(7-5), where large collapse
angles were found to be common for all the hemispherical liner elements according to the
standard jetting analysis. Therefore, the jet mass, which is directly proportional to the
collapse angle, showed that the produced EFP mass is 68% of its liner total mass. The EFP
is characterized by its uniform low velocity massive slug, which produces a shallow
penetration depth but a large hole diameter.
192

Figure 7-14 The analytical and numerical
c
-V
2
curves for zirconium liner with the
numerical
c
-V
2
curve for copper liner as reference.

Figure 7-15 and V
2
values for four zirconium liner shapes from jetting analysis, in which
different regions of zirconium jet formation and coherency are shown.
193
7.5.2 Jet analysis and penetration of different liner shapes
The jetting analysis calculates the jet characteristics until the liner elements reach their axis
to form a jet element. Thus, it does not consider the jet elongation and the breakup time
resulting from this elongation [115]. In addition, the jet tip velocity correction is not
included. This means that the points near the apex region do not have sufficient space to
accelerate to its theoretical maximum collapse velocity, which results in a reduced jet tip
velocity and the pilling up of the jet mass [6]. Thus, the inverse velocity gradient needs to
be removed by adding the piled-up mass to the jetting element with the highest velocity.
Then, the jet tip velocity is corrected based on the momentum conversation. The corrected
jet tip velocity was calculated according to [156]:
p
] v
]
(dm
]
dX)dX
X
i
0
] (dm
]
dX)dX
X
i
0

7-17
where m
j
is the jet mass and X is the axial distance of the jet element.
Figure 7-16 shows the jet velocity as a function of the distance from the apex with and
without correction. It was predicted by the jetting analysis that the liner with the bell shape
has the highest tip velocity exceeding 9km/s. However, after the tip correction, the tip
velocity is reduced to 6.63 km/s. The pilled-up tip mass of the bell-shape liner is illustrated
in Figure 7-17, which can be compared with the characteristics of the jet mass generated
from other three liner shapes. On the other hand, the bi-conical liner shows less difference
between the theoretical and the corrected tip velocities, which are 8.6 and 8.4km/s,
respectively. The hemispherical liner produces an explosively formed projectile (EFP)
rather than the thin jet. As shown in Figure 7-17, most of the hemispherical liner mass
flows to form the EFP with the largest diameter and mass, but the slowest velocity
(3.8km/s). Therefore, the crater diameter resulted from the interaction of EFP with the
concrete targets is expected to be the largest one amongst the caretrs created by four liner
shapes.
194

Figure 7-16 Jet velocity profile along the liner axis with and without tip correction.

Figure 7-17 The jets shapes for the different liner geometries right before the impact on the
test layers.
The collapse velocity can be fully developed either by allowing sufficient acceleration
distance between the liner elements and the axisymmetrical axis of the shaped charge (e.g.
the bell shape liner) or by using a reduced cone apex angles (e.g. the bi-conical liner).
Actually, both methods work in the jet formation of bell shaped liner because in addition to
the increased liner distance from its axis, the bell shaped liner close to the base has a small
angle, which increases the jet velocity according to the unsteady PER theory [50]. The
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
0 20 40 60 80 100
$
e
t

%
e
l
o
c
i
t
&

(
m
'
)
(iner poition () *rom ape+)
,emipherical
Conical
-ell
-i.conical
/ithout tip correction
$et %elocit& 0ith tip
195
collapse velocity-time histories for the studied four liner shapes are shown in Figure 7-18.
The hemispherical liner apex angle decreases gradually from the apex to the liner base, and
the minimum angle () of the hemispherical liner is 29
o
, which is responsible for the
reduced achievable jet velocity. However, the collapse velocity of the hemispherical liner
is considered to be the highest among the four liner shapes in the apex region up to 40% of
liner position from the apex. This is due to the existence of sufficient space available for its
liner elements in the apex region to accelerate. The jetting analysis indicated that the
collapse angles of the hemispherical liner ranging from 95 to 120 degree are greater than
those of the other liner shapes, which lead to the greatest percentage (i.e. 67.57 %) of the
jet mass from the total liner mass according to Eq.(7-5).

Figure 7-18 The collapse velocity histories for the different liners.
The collapse velocities for both the conical and the bi-conical liners are similar in the
region near the liner apex because they nearly have the same inclination angle at that
region and only have small difference between their explosive-metal mass ratios. Beyond
this region, their collapse velocity difference begins to increase because the bi-conical liner
has smaller liner angle, which enhance the explosive-metal mass ratio due to the charge of
geometry. The bell shape liner, which has a long distance between its inner surface and the
axisymmetrical axis of the shaped charge, gains more collapse velocity than the conical
liner near the apex and the bi-conical liner at the liner base. Since the tip velocity of the jet
is produced from the liner elements near the apex, the bi-conical liner has the largest jet tip
velocity. In addition to the above-discussed factors of the liner shape, another important
factor for the remarkable difference of the collapse velocities is the liner surface area that
determines the transmission of explosive energy to the liner. According to Table 7-3, the
196
surface area of the bell and the bi-conical liners are 7% and 25% greater than that of the
conical one respectively, which means that both these shapes are capable of absorbing
more energy from the explosive although they have nearly the same total mass ratio
between the explosive charge and the metallic liner. The transmitted momentum can be
indicated by Figure 7-19, where the x-momentum-time histories of the four liner shapes
exhibit a similar exponential relation, but the difference between them accounts for the
difference in their jet velocities and explain why the liners with a greater surface area have
a better performance than those with lower surface area.
Table 7-3 The liner shapes and their jets characteristics.
Liner shape Hemispherical Conical Bell Biconical
Liner mass (g) 25.90 26.40 23.97 29.30
Explosive mass (g) 20.80 24.20 24.45 31.80
M/C mass ratio 1.25 1.03 0.98 0.95
Jet mass (g) 17.50 3.47 3.77 5.06
% jet to liner 67.57 13.14 15.73 17.27
Jet K.E. (kJ) 31.05 34.70 36.90 49.70
Liner surface area (mm
2
) 1785 2400 2568 3005
Jet velocity without corr. (m/s) 3815 5761 8780 8611
Jet velocity with corr. (m/s) 3815 5386 6630 8402
Tip distance from apex (%) 0 63.12 52.65 32.77

Figure 7-19 The x-momentum histories for the four perforators with different liner shapes.
197
7.5.3 Penetration
The numerical penetration depths into 40 MPa concrete targets for the four perforators
with different liner shapes and their corresponding experimental penetration depths are
illustrated in Table 7-4 and Figure 7-20. Samples of the numerical penetration stages of the
biconical jet and hemispherical EFP, as two examples representing very different shaped
charges, into concrete are illustrated in Figure 7-21 and Figure 7-22, respectively. The
difference between their crater profiles is significant. The shortest penetration path with the
biggest crater diameter for the hemispherical liner is a typical characteristic of EFP, which
has a short length, a large diameter, large mass and a very low tip velocity in comparison
with the other jets. On the other hand, the produced jet from the bi-conical liner has a small
jet diameter and a high tip velocity, therefore its kinetic energy per unit area of the jet
cross-section is considered the greatest one, and hence its penetration depth is the largest
one in comparison with the other three liner designs.
Table 7-4 The numerical and experimental penetration depths using different liner shapes.

Figure 7-20 The penetration depth dependence on the concrete strength.
0
10
20
30
40
50
60
70
80
90
,emiph. Conical -ell -iconical
P
e
n
e
t
r
a
t
i
o
n

d
e
p
t
h

(
c
m
)

Num. simulation
Experimental
Hemisph. Conical Bell Biconical
Penetration (Numerical) (cm) 42.3 55.9 64.7 74.1
Penetration (Exp.) (cm) 48.4 68 75 83
Difference bet. Num. and Exp. (%) 12.6 17.8 13.7 10.72
Num. exit hole diameter (mm) 29.6 16.0 14.5 12.0

198
The penetration depth of the OWP into the concrete material indicates the importance of
the liner shape design on the performance of these perforators. Thus, the longer the
penetration depth into the concrete, the greater the flow productivity of the well and the
better the performance of the perforator charge.

Figure 7-21 The damage contours of the concrete penetrated by the bi-conical jet at
different times.

Figure 7-22 The damage contours of the concrete penetrated by the hemispherical EFP at
different times.

199
7.6 Summary
The critical collapse angle as a function of flow velocity was calculated numerically and
analytically for the zirconium liner material in this chapter, which, together with the
longitudinal sound velocity of the liner material, defines the boundaries among jetting,
coherent jetting and non-jetting situations. The regions divided by these boundaries were
used to exam the jet characteristics of four zirconium OWP liners with different shapes. It
was shown that the jets of four different liners are coherent, but have different collapse,
flow and jet velocities because of their different geometries and surface areas. The
enhanced performance of the liner shape effect was confirmed by the static firing of four
zirconium OWPs against the laminated API-RP43 targets. The bi-conical liner exhibited
the largest penetration depth into target, which is 22% greater than that of the baseline
conical zirconium liner. The large collapse angles in the hemispherical shape increased its
jet mass, but decreased its velocity, and therefore, it is classified as an explosively-formed-
projectile (EFP). An EFP achieves the shortest penetration depth, but largest crater
diameter among the four OWPs with different liner shapes.

Chapter 8: A modified Virtual Origin Model for S. C. Jet with Non-uniform Density Distribution
200
CHAPTER.8 A MODIFIED VIRTUAL
ORIGIN MODEL FOR SHAPED CHARGE JET
PENETRATION WITH NON-UNIFORM
DENSITY DISTRIBUTION
8.1 Introduction
Hypervelocity jet of a shaped charge has excellent penetration capability into various
targets. Due to its penetration capability, shaped charge has been successfully used both in
the battle field to defeat armours and in the oil and gas wells to perforate tunnels to connect
the wellbore to the reservoir. In these applications, it is necessary to predict the depth of
penetration, which is an important parameter for the assessment of shaped charge effects
on a target.
Since the shaped charge jet travels at hypervelocity, the impact of the jet on target
produces much higher pressure than the strength of the target, and thus, the hydrodynamic
model [49, 103] can be applied to study the jet penetration. These original hydrodynamic
models assumed uniform distributions of jet density and jet velocity along the jet length
and applied Bernoulli equation at the interface between jet and target for the pressure
equilibrium
1
2
p
(I
-u)
2
=
1
2
p
1
u
2
,
8-1
where I
is the velocity of a continuous jet; u is the velocity of the jet-target interface or

penetration velocity; p

and p
1
are jet density and target density around the jet-target
interface, respectively. When the distributions of jet density and velocity are uniform, the
consumption of the jet is controlled by
I -u = -
t

8-2
where is the current length of jet.

201
The depth of penetration of the jet into target is determined by
u =
dP
d
oi P = ] ut
0
, 8-3
where t=0 is the time when the jet starts to hit the target. The maximum depth of
penetration is achieved when the jet is completely consumed at t = t
]
, or (t
]
) = . For a
jet with original length of
0
, the maximum depth of penetration is determined by Eqs.(8-1)
and (8-3), i.e.
mux 0
p
]
p
8-4
Eq. (8-4) is also applicable to solid rod penetrator.
For a particulated jet, Bernoulli equation cannot be used directly because the internal
pressure cannot be supported after the jet is particulated [49]. This study will only consider
continuous jet. Interested readers may refer to [103] for the penetration models of
particulated jet.
Since the early time of the jet penetration study, it has been realised that the spatial
distribution of jet velocity is not uniform [49]. Birkhoff et al. [49] extended the
hydrodynamic penetration model [Eq.(8-4)] to the jet with non-uniform velocity
distribution. However, this model introduced several parameters that cannot be easily
determined, and therefore, it has not been widely used. Abrahamson and Goodier [100]
also extended the hydrodynamic penetration model to include non-uniform jet velocity
distribution and stand-off distance. This model started from an arbitrarily selected initial
time and required the initial jet length at this moment to be given, which makes the model
difficult in practical use.
The concept of virtual origin was first proposed by Allison and Bryan [157] and then
developed by Allison and Vitali [2] for the penetration of continuous and particulated jets
with the consideration of velocity gradient and the stand-off distance between the virtual
origin and target surface. This model has been widely accepted, which can be used to
predict the depth of penetration before and after jet breakup [101, 103].
The virtual origin model keeps the basic equations in hydrodynamic model, i.e. Eqs.(8-1)
and (8-3) where the strengths and the compressibility of the jet and target materials were
202
neglected. Eq.(8-2) was abandoned because the concept of jet length cannot be applied
when it is lengthened as it travels forward from the shaped charge. In addition to these
assumptions, following conditions for the existence of a virtual origin need to be satisfied.
Existence of a virtual origin: All jet elements are formed simultaneously at a virtual origin
located at a distance Z
0
from the target surface. Each jet element is emitted from the virtual
origin at its own velocity that remains constant during its travelling between virtual origin
and target. The existence of a unique virtual origin location of the entire jet requires that
the spatial distribution of jet velocity is linear.
In the virtual original model and its applications, the density of the jet element is treated as
a constant, i.e. the density of each element remains constant during its travelling and the
spatial distribution of the jet density is uniform. However, it has been observed that there is
a density deficit based on flash x-ray measurements and the soft recovery of jet fragments
[1, 158]. Variable density distribution was also observed in the jets formed from powdered
metal liners [159-161]. Therefore, it is necessary to extend the virtual origin model to the
jet with non-uniform density distribution.
This chapter keeps the assumption that the density of each jet element remains constant
during its travelling, but considers the non-uniform jet density distribution to study its
effect on the penetration depth. The non-uniform jet density distributions along its axial
distance are estimated numerically using the Autodyn jet formation algorithm for the three
liners made from electrolytic OFHC copper, zirconium and copper-tungsten un-sintered
powder. An analytical approach is introduced to account for the penetration decrease due
to the non-uniform density distribution along its axis. The proposed model is validated by
experiments and numerical simulations using Autodyn hydro-code.
A modified virtual origin model with non-uniform distribution of jet density is proposed in
Section 8.2. Section 8.3 describes the liner manufactures and the experimental set-up
configurations. Section 8.4 introduces the numerical models used to simulate the shaped
charge jets and penetrations. Results are presented in Section 8.5 with further analysis,
which is followed by conclusions in Section 8.6.
8.2 Penetration analytical model
In this chapter, we will focus on the jet penetration before or without jet breakup. Figure
8-1 is a schematic drawing that defines the penetration parameters of a shaped charge jet
203
penetrating into an incompressible target. Z
o
is the stand-off distance from the virtual
origin point to the target surface, t is the penetration time, P(t) is the penetration depth at
time (t) and V
j
is the impinging velocity of the jet onto the target (observed at the jet/target
interface), which equals to the velocity of the jet element that impacts the target at the
same moment of time t.

Figure 8-1 The hydrodynamic jet penetration; [2].
Therefore, the depth of penetration P(t) at a given time t is determined by
P(t) = tI
(t) -Z
o
.
8-5
The depth of penetration increase monotonically with time, which requires the satisfaction
of following condition:
dP
d
for tI
(t) Z
o
. 8-6
This condition was not checked in previous publications. A proof of this necessary
condition will be given in Appendix B.
When hydrodynamic Bernoulli equation [Eq.(8-1)] is applied,
u =
v
]
+1

8-7
where = p
1
p
/ .
Following equation can be obtained from Eqs.(8-3 , 8-5, 8-7),
I
(t) +t
dv
]
()
d
=
v
]
()
+1
.
8-8
When the jet density is a constant, the solution of Eq.(8-8) predicts the jet velocity V
j
(t) as
204
I
(t) = I
0
_
t
o
t
]

+1

8-9
where I
0
is the jet tip velocity and t
o
is the time when the jet tip reaches the target surface
(i.e. t
0
I
0
= Z
0
).
From Eq.(8-9), time can be expressed as
t = t
o
_
v
0
v
]
()
]
y+1
y
,
8-10
and therefore, the depth of penetration at time t can be obtained from Eq.(8-5) when V
j
(t)
and t in Eq.(8-5) are substituted by those Eqs.(8-9, 8-10)
P(t) = Z
0
__
v
0
v
]
]
1
y
-1_ = Z
0
_[
1
y+1
-1_ .
8-11
The maximum penetration is achieved at time t
c
when the cut-off jet element (i.e. the last
jet element that has hydrodynamic penetration capability) hits the target at the cut-off
velocity (I
c
). Therefore, the maximum depth of penetration is
P = Z
0
_[
v
0
v
c
1
y
-1_ .
8-12
In the proposed model, it is assumed that the density of each jet element will be a constant
during its travel between virtual origin and target. However, since different jet elements
experienced different jet formation processes, their densities are different. Therefore, the
spatial distribution of the jet density is non-uniform. At the jet-target interface, the
observed jet density should be a function of time, i.e. p
= p
(t). Let p
0
represent the
original density of the liner material and the density of target p
1
is a constant, then
parameters
0
= p
1
p
0
/ and (t) = p
1
p
/ (t) are introduced. Thus,

()
0
=
_
p
]0
p
]
()
.
Based on jet formation analysis presented later, it is found that the normalised jet density is
directly related to the normalised jet velocity in a linear relationship, as shown in Figure
8-2. According to Figure 8-2, the density reduction at the jet tip is larger than that at the
rear jet. The maximum density reductions in the simulated examples are around 15.8% for
copper and zirconium liners and 21.7% for copper-tungsten liner, respectively. These
205
values are in line with the experimental observations by Zernow for the copper liner [1].
Details of the numerical simulation will be presented in Sections 8.4 and 8.5.
The linear relationship between
y(t)
c
and
V
]
(t)
V
0
can be described by
()
c
= o
v
]
()
v
0
+b
8-13
where a and b are constants to be determined from data fitting of numerical results and
analytical consideration, which will be given at the end of this section.

Figure 8-2 The relationship between the scaled density ratio and the scaled jet velocity.
Equation (8-8) can be rearranged as
-dv
]
V
]
(:)

-
dv
]
y(1)V
]
(:)
=
d:
:
, which can be integrated when
Eq.(8-13) is used, i.e. ]
-dv
]
V
]
(:)

v
]
()
v
0
-
1
c
]
dv
]
V
]
(:)[
a
v
0
V
]
(1)+b
v
]
()
v
0
= ]
d:
:
c
, or
t = t
o
_
v
0
v
]

()
]
1+
1
by
c
_
[
c
v
0
v
]
()+b
[
c
v
0
v
0
+b
_
1
by
c

.
8-14
This equation is reduced to Eq.(8-10) when
y(t)
c
= 1 or a=0 and b=1 in Eq.(8-13).
When t=t
c
, V
j
(t)=V
c
, the maximum penetration is achieved by the last penetrating element
at a cut-off velocity V
c
. t
c
can be determined by Eq.(8-14), i.e.
206
c o
v
0
v
c
1+
1
by
c

C
1
by
c

8-15
where
c
= p
1
p
c
/ =
0
[o
V
c
V
0
+b and
= p
1
p
/ =
0
(o +b) according to
Eq.(8-13), in which
jc
and
jt
are the densities of last penetrating element and tip element
of the jet, respectively.
From Eq.(8-14), the impact velocity of the jet is determined by an algebraic equation of
b _
v
0
v
]
]
b
0
+1
+o _
v
0
v
]
]
b
0
= (o +b) [
b
0
.
8-16
Eq.(8-16) reduces to Eq.(8-9) for constant jet density when a=0 and b=1.
The penetration depth at time t is determined by Eq.(8-5) when t is substituted from
Eq.(8-14) i.e.,
P(t) = Z
0
__
v
0
v
]
]
1
by
c

_
u
v
]
v
c
+b
u+b
_
1
by
0
-1_ ,
8-17
which can be reduced to Eq.(8-11) for constant jet density when a=0 and b=1. The solution
of Eq.(8-16) is needed to give an explicit expression of P(t) in Eq.(8-17).
The maximum depth of penetration is given by:
P = Z
0
_[
v
0
v
c
1
by
c

_
u
v
c
v
c
+b
u+b
_
1
by
0
-1_ = Z
o
_[
v
0
v
c
1
by
c

[
1
by
c
-1_

,
8-18
when V
j
=V
c
and t=t
c
. Eq.(8-18) is reduced to Eq.(8-12) for constant jet density when a=0
and b=1 or when
c
=
t
=
o
and b=1.
The derivation equations for the penetration depth calculation will be given in Appendix C.
The values of a and b in Eq.(8-13) were determined from the curve fitting of -V
j

relationship along the jet length in Figure 8-2, which are shown in Table 8-1.

207
Table 8-1 The values of parameters a and b in Eq.(8-13).
Curve fitting using least-square excel fit
from Figure 8-2.
a [Eq.(8-19)]
a b
0.077 1.0082 0.080
0.089 0.9983 0.086
0.107 1.0179 0.096
It was further found that parameter a is correlated with the density of the liner material
(
j0
), the stand-off distance (Z
o
), the total mass of the jet (m
jet
) from the standard jetting
analysis and the radius of the jet (r) from jet formation simulation or flash x-ray
experiment. A non-dimensional formula is recommended for the calculation of parameter
a, i.e.
m
]c
p
]
c

n
2
z
c

8-19
The values of a using Eq.(8-19) are also listed in Table 1 for three liner materials. It can
be seen that values of a predicted by Eq.(8-19) are very close to the corresponding values
determined by curve-fitting method. According to Table 8-1, the values of b can be
approximated to unity for the three liner materials.
8.3 Liner materials and penetration experiments
The three liners that have been used in this study were the copper, the zirconium and the
un-sintered copper-tungsten powder. The liner has a small base diameter of 33mm, a cone
apex angle of 46
o
and a varied liner wall thickness as illustrated in Figure 8-3.
The copper liner was OFEC (Oxygen Free Electrolytic Copper) of grade C10100 with
purity of 4N (99.99%). It was manufactured using the deep drawing technique with an
intermediate annealing of 1000
o
C (two minutes) to decrease the strain hardening and
maintain the material ductility [31].
The zirconium liner was manufactured from a solid pure zirconium cylinder 4N (99.9951)
having a density of 6623kg/m
3
using high accuracy CNC machine in order to guarantee a
high precision manufacturing (i.e. the precision of 5m). The zirconium rod was annealed
to 900
o
C for one hour before machining in order to obtain a relative small average grain
size, hence to increase its ductility, which in turn will increase its breakup time and
208
improve the liner performance [151]. The type and the percentage of the impurities present
in the zirconium material are listed in Table 8-2.
Table 8-2 The elemental percentage of impurities presented in the zirconium material.
Element Impurities amount (%)
Fe 0.005
Cr 0.0009
C 0.001
N 0.008
Powder metallurgy (PM) technique has been used to manufacture OWP liners. It has good
penetration capability especially at short stand-off distances [15, 24, 43-44]. The
composition of the powder mixture ingredients is listed in Table 8-3. Small average grain
size with irregular particles shapes are chosen for the liner powders. The powders are
mixed together with the designated mass ratio until the homogeneous mixture blend is
obtained, after which they are pressed using the punch, the die and the ejector, shown in
Figure 8-4. The applied pressure was 100MPa using hydraulic press at a low rate (i.e.
1MPa per second) to avoid trapping air voids inside the liner material. The product is a
brittle material in the pre-sintering state and is called the green, which is tested in this
state without sintering. All three tested liners are show in Figure 8-5.

Figure 8-3 A sketch of the designed shaped charge well perforator.
Table 8-3 The mass percentage of the powder liner composition.
Material Copper Tungsten Tin Graphite
Mass ratio % 43 45 11 1
Average grain size (m) 3 0.6:1 < 45 < 20
Function Binder Main powder Binder coating Lubricant
209

Figure 8-4 A sketch of the punch, the die, the ejector and the produced powder liner.

Figure 8-5 The three liners studied in the work.
The powder liner density is not uniform over the entire liner height because the force
distribution is not homogeneous due to the conical liner profile. Therefore, small parts of
the same powder liner specimen were cut off and used to measure their densities using the
gas pycnometer [162]. The measured densities for the testing specimens as a function of
the scaled distance from the cone apex to the liner height are shown in Figure 8-6, which is
taken into account in the description of liner physical properties in Autodyn hydro-code
simulations.
210

Figure 8-6 The measured densities of the liner elements at different distances from the
cone apex point.
The charge casings are steel with an average wall thickness of 4.5mm. The main explosive
charges for the three charges are PE4 with a total average mass of 24.5g and a standard
deviation of 0.8g. The PE4 explosive is a powerful RDX-based explosive (i.e. 88% RDX
in mass, 12% plasticizer and other additives) having a detonation velocity of 8027m/s at
1.59g/cm
3
density [153] and 8200m/s at 1.6g/cm
3
density [154]. The explosive charge was
filled into the steel casing first. Then, the liner was pressed slowly against the steel casings
containing explosives to avoid holding air gaps inside the explosive. The charges were then
attached to the upper steel layer of the test configuration as shown in Figure 8-7.
The concrete cylinders with the designated strength value were cast from the same mixture
and allowed to cure according to the test evaluation of the OWP [155]. These targets were
tested according to the standard OWP testing configuration and requirements in the
Section-II of API-RP43 [143]. The measured average strength of the standard concrete
cubes was 40.02 MPa with a standard deviation of 0.92 MPa, measured at 28 days after
their casting [145].
10.60
10.70
10.80
10.90
11.00
11.10
11.20
11.30
11.40
11.50
11.60
0 20 40 60 80 100
1
e
a
u
r
e
d

d
e
n
i
t
&

(
#
'
c
m
3
)
2itance *rom the ape+ ())
211

Figure 8-7 Dimensions of the test setup and the experimental test configuration
(2: Boaster; 3: OWP; 4: Front steel disc; 5: Concrete; 6: power supply cable).
In a separated test without target, the particulated copper jet fragments were recovered
using sand. The densities of two jet fragments were measured using a helium gas
pycnometer, which has an accuracy of 10
-4
g/cm
3
. The measured densities of the two jet
fragments were 7.4120 and 8.2300 g/cm
3
at the tip and the rear, respectively with a
standard deviation of 0.05 g/cm
3
. In comparison with the original density of copper liner
material (8.930 g/cm
3
), they represent 17.0% and 7.8% density reductions, respectively.
The density reduction at the tip of jet (17.0%) is very close to the maximum density
reduction predicted in Figure 8-2 (15.8%). This again gives evidence to support the
existence of density reduction in the formed jet of a shaped charge, which was first
observed based on x-ray measurements [1, 158].
8.4.1 Hydrocode algorithms of the jetting analysis, the jet formation model and the
jet penetration model
Hydrocode algorithms and material models were presented in Chapter 4.
212
8.4.2 Mesh sensitivity for the jet formation model
In order to study the mesh sensitivity on the jet characteristics, five different mesh
densities were proposed for the jet formation modelling. The jet formation was used to
identify the density and velocity of the jet elements as they pass the gauge point shown in
Figure 8-8. The fixed gauge point is located 100mm from the liner base (i.e. 3 time
calibre). The uniform square meshes of 0.17, 0.20, 0.25, 0.33 and 0.50mm of the Euler
grids were used for mesh sensitivity analysis in the jet formation model.

Figure 8-8 Location of the fixed gauge point used to predict the density and the velocity
histories for the mesh sensitivity study.
Figure 8-9 shows the recorded density history at the fixed gauge point using five mesh
sizes. The density histories for the five mesh sizes are detected between 14 and 16 s. It
shows that finer mesh sizes give higher densities at the beginning. But the density
corresponding to each mesh size convergences to the copper solid material density as the
jet tail passes through the gauge point. This means that the density of the jet material
increases gradually from the tip to its tail due to the existence of velocity gradient. The
maximum relative difference of density for the finest and coarsest meshes is about 7%.
The velocities of jet elements passing the gauge point are obtained by the same way used
for the calculation of density. The velocity histories for different mesh sizes are depicted in
Figure 8-10. It can be observed that the five meshes predict nearly the same shape of the
velocity history. The relative difference of the peak velocity between coarsest and finest
meshes is 14.8% while the relative difference of the peak velocity between the finest and
second finest meshes is reduced to 2.9%, which indicates the convergence with the
decrease of mesh size. These evidences ensure that a mesh size of 0.17mm is sufficient
while practically affordable, which will be used globally for the calculations of jet velocity
and density.
213

Figure 8-9 The recorded density-time histories for the fixed gauge point using five
different mesh sizes.

Figure 8-10 The recorded velocity-time histories for the jet material particles moving
through the gauge point.
214
8.4.3 Mesh sensitivity for the jet penetration into target
In order to find the influence of mesh size on the penetration depth into target, five
different mesh sizes were used for the laminated target consisting of steel, water and
concrete (their dimensions have been shown in Figure 8-7) while the copper jet mesh
density remains unchanged for all five models. The different uniform square mesh sizes of
0.5, 1, 2, 3 and 4mm were used for the laminated target. The penetration depths into the
target using different mesh sizes are shown in Figure 8-11. It is evident that the penetration
depth is convergent with the reduction of mesh size. The relative difference of the
penetration depth for 0.5 and 1.0mm meshes is only 0.3%. However, the simulation time
for 0.5mm mesh is doubled (about170 hours). Therefore, the mesh size of 1mm1mm is
used globally for the penetration analyses of three liners.

Figure 8-11 The penetration depths into laminated target using different mesh sizes and the
relevant time consumption.
8.5 Results
8.5.1 The Jetting analysis results
The outputs of the jetting analysis for the shaped charge perforators with three different
liners are summarized in Table 8-4. The kinetic energies of the three produced jets nearly
have the same value of 36 kJ because they have the same liner shape and the same amount
of explosive. However, the zirconium liner with the lowest density and mass has the
highest jet tip velocity of 6075m/s but with the lowest jet mass of 3.1g.
215
Table 8-4 The different liners and their jet characteristics.
Liner material Copper Zirconium Powder
C
o
(m/s) 3757 3940 3849
Liner mass (g) 32.6 25.1 40.2
Jet mass (g) 4.0 3.1 4.5
Jet % from the liner mass 12.31 12.53 11.18
Jet K.E. (kJ) 36.0 37.2 35.7
Jet tip velocity (m/s) 5476 6075 5320
Cut-off velocity (m/s) 1610 1720 1747
Time (t
o
at Z
o
) (s) 18.50 16.52 19.30
Initial Z
o
(cm) 10.7 11.5 12.2
8.5.2 Jet density distribution
The density of the jet along its length was calculated from the jet formation model for three
liners where Mie-Gruneisen EOS based on the shock Hugoniot was used (Section 4.5.3,
page 110). The density of the collapsed liner material is directly related to the liner
compressibility and the pressure generated from the explosive load. Distributions of
density, compressibility and velocity over the entire jet length are depicted in Figure 8-12
for the copper jet. This figure shows that jet density decrease from slug to tip along the jet.
Besides, the density contours also shows a radial density distribution on the circular cross-
section of the jet (i.e. the density on the tip premises of the jet is 0.6% larger than that at its
centreline). Figure 8-13 shows the velocity and the density histories of the copper jet
recorded at the fixed gauge point. The distributions of jet density and velocity along the jet
axis for the copper liner at a given time are shown in Figure 8-14. Both Figure 8-13 and
Figure 8-14 demonstrate the increase of density reduction with the corresponding jet
velocity.

Figure 8-12 (a): Density, (b): compressibility and (c): velocity of the copper jet just before
the jet tip impacts the target; (d): a picture of the recovered copper slug.
216

Figure 8-13 Jet velocity and density histories recorded at a fixed gauge point.

Figure 8-14 Jet density and velocity distributions along the jet axis for copper liner.
8.5.3 The penetration depth calculations
The projected effective jet lengths for the three shaped charge perforators with different
liners were calculated by the back projection of the relation between time and effective jet
length from the moment when the jet reaches the first steel layer. An example of the
relation between time and effective jet length established by the data obtained from
Autodyn jetting analysis is shown in Figure 8-15. However, this value cannot be used
directly with Eq.(8-18) because the effective jet length has to be modified taking account
of the thicknesses of the laminated steel and water layers.
217

Figure 8-15 The fan plot of the copper jet showing the original and the modified effective
jet length due to the presence of the laminated test layers.
The correction of the jet tip velocity is based on the uniform density formula Eq.(8-12).
The exit jet tip velocity perforating a finite thickness target [103] is given by
I
cx
= I
n
[
z
i
z
i
+1
i
i
,
8-20
where V
jex
and V
jin
are the exit and the input jet tip velocities, respectively; Z
i
is the
effective jet length at the front of the target surface, T
i
is the target thickness and i refers to
the number of the target layer to be perforated.
However, Eq.(8-20) is not suitable for the penetration formula, Eq.(8-18), where non-
uniform density effect is considered. Thus Eq. (8-20) was modified based on Eq.(8-18) to
determine the exit jet velocity with considering the density reduction effect,
cx
bv
]in
(u+b)_
Z
i
+
i
Z
i
]
by
i
-u
. 8-21
Eq.(8-21) together with Eq.(8-18) can be used to predict the penetration depth of a
continuous shaped charge jet into a multi-layered target when the non-uniform density
distribution of the jet is considered. The values of the exit jet tip velocity and the relevant
effective jet length for the test layers are presented in Table 8-5 while Table 8-6 gives the
218
penetration depth calculated using various methods including the modified virtual origin
model [Eq.(8-18)] in Section 8.2 and the penetration reduction due to the density gradient
and reduction.
Table 8-5 The effective jet length and the jet exit velocities of the three test layers.
Property Zirconium Copper Powder
V
tip
(m/s) 6075 5476 5320
V
cutoff
(m/s) 1620 1610 1747
Z
o
(cm) 11.5 10.7 12.2
Solid jet density
j
(g/cm
3
) 6.51 8.93 11
Target density
T
(g/cm
3
) 2.75 2.75 2.75
=

p
t
p
j

0.65 0.55 0.50
Vp
1
(m/s); Eq.(8-21) 5891.8 5324.2 5198.9
Vp
2
(m/s); Eq.(8-21) 5564.5 5053.7 4979.3
Vp
3
(m/s); Eq.(8-21) 5153.9 4714.4 4698.1
Vp
3
(m/s); Eq.(8-20) 5219.5 4769.7 4762.1
Z
final
(cm) 14.4 13.6 15.1
Vp
1
, Vp
2
and Vp
3
are the exit jet velocities as it perforate the steel, the water and the steel
layers of the API laminated target layers, respectively.
The calculated penetration depths and the reduction percentages in penetration due to the
density gradient along the jet length indicate that the reduction term has considerable
influence on the predicted penetration depth of a shaped charge jet. Data in Table 8-6 are
presented in Figure 8-16. It clearly shows that the modified virtual origin model largely
improve the predictions of penetration depth by virtual origin model for all three liners.
Table 8-6 Comparison among experimental result, numerical simulation and the virtual
origin model predictions for the penetration of jets with three different liners.
Liner
V
j

(m/s)
Eq.
(8-20)
V
c

(m/s)
a
Value
Penetration depth (cm)
Difference
Between
Eqns. (8-12)
and (8-18)
VO
Eq.
(8-12)
Mod.VO
Eq.
(8-18)
Exp. Sim.
(cm) (%)
Zirconium 5153.9 1620 0.077 72.73 63.57 68.0 59.0 9.2 12.5
Copper 4714.4 1610 0.089 82.67 72.04 64.0 65.0 10.6 12.8
Powder 4698.1 1747 0.049 97.10 78.18 80.0 75.0 18.9 19.4
219

Figure 8-16 Comparison among experimental result, numerical simulation and the virtual
origin model predictions for the penetration of jets with three different liners.
8.6 Summary
The density reduction of a shaped charge jet is developed during the jet formation, which
has been shown experimentally and numerically in this chapter. This leads to the non-
uniform distribution of the jet density and the original virtual origin penetration model is
incapable of dealing with penetration of jet with non-uniform density distribution. A
correlation between jet density reduction and jet velocity is proposed in this chapter, based
on which an analytical solution of the modified virtual origin model is obtained. The
validity of the modified virtual origin model is demonstrated by its largely improved
predictions in comparison with experimental and numerical results.
Chapter 9: Zirconium Shaped Charge Jet Breakup Time
220
CHAPTER.9 ZIRCONIUM SHAPED
CHARGE JET BREAKUP TIME
9.1 Introduction:
The velocity gradient of a shaped charge jet causes the stretching of the jet after it has been
formed. This leads to the axial breakup of the jet into small fragments, which can
significantly decrease its penetration capability. Thus, it is necessary to predict the breakup
time and the characteristics of the jet fragments.
Three approaches including hydrocode simulations, one-dimensional models and empirical
formulae have been employed to predict the jet breakup time. These approaches have been
summarised in Ref. [67] and discussed in Ref. [85]. Among these three approaches, the
semi-empirical formula presented by Hirsch [4-5] has demonstrated its reliability and
efficiency for the prediction of the breakup time of a shaped charge jet. Hirsch [5]
estimated the breakup time t
b
of a jet element according to:
t
b
=
2r
v
PL
9-1
where r is the initial radius of the jet element when the jet forms, which can be measured
from flash x-ray or estimated from:
r = 2RI
L
sin_
[
2
]
9-2
in which R is the initial inner radius of the liner element and is the elemental collapse
angle of the liner element calculated from jetting analysis (Pugh-Eichelberger-Rostoker
model [50]). V
PL
is a characteristic plastic velocity representing the average velocity
difference between the neighbouring jet segments [5]. The physical meaning of the
proposed breakup time formula in Ref.[5] is the same as the breakup time of a one-
dimensional homogeneous ductile metal that undergoes a very high strain-rate
deformation.
The reciprocal of V
PL
(i.e. 1/ V
PL
) represents the specific breakup time of a certain liner
material [5]. Experimentally, V
PL
can be measured using multiple flash x-ray units, where
the position, the length, the radius and the velocity of each jet segment are determined,
based on which the velocity difference between each pair of neighbouring jet segments can
221
be determined. In addition to the determination of breakup time, V
PL
can also be used to
calculate the total number of jet fragments (n) according to Hirsch [4]:
n =
I
p
-I
cu
I
PL

9-3
where V
tip
and V
rear
are the jet tip and the rear velocities, respectively.
In the first part of this chapter, the parameters in Johnson-Cook (J-C) constitutive equation
for the zirconium material will be calculated based on the data obtained from the tensile
testing of zirconium specimens at different strain-rates and different temperatures. The
second part includes the calculations of the V
PL
for some zirconium liners based on J-C
constitutive equation. The calculated V
PL
values are implemented in a simple breakup time
formula as a function of the scaled liner thickness to the charge diameter and the jet radius.
The V
PL
is calculated for shaped charge with conical zirconium liners of an apex angle of
46
o
and an outer diameter of 36mm with different liner wall thickness values of 0.7, 1, 1.3,
1.7, 2, 2.3, 2.7, 3, 3.3, 3.7 and 4mm. The used explosive is HMX with a loading density of
1.891g/cm
3
and a total mass of 30.75gm. All the shaped charge jet output data were
calculated using Autodyn jetting analysis, in which the elemental jet velocity and axial
distances were used to calculate the jet strain-rate that was used to calculate V
PL
using the
Johnson-Cook constitutive equation.
9.2 Determination of V
PL

Various methods have been suggested by researchers for the calculation of V
PL
in order to
estimate the shaped charge breakup times for different liners [163].
Haugstad [86] presented a method to determine V
PL
based on dimensional analysis of the
parameters governing the jet breakup time model. The model consists of a rod of length L
o

clamped at one end and moving with velocity V
o
at the other end. Assuming a constant
density
o
and constant flow stress
o
and according to the -theorem, only one non-
dimensional number exists, i.e.
__
o
o
p
o
]
0.5
_
1
o
]_ = Const. 9-4
where
o
is the strain-rate that can be calculated by
o
=V
o
/L
o
,
o
= nr
o
is the perturbation
wave length that causes the initial necking [164], r
o
is the initial radius of the jet element.
Thus,
222
I
PL
=
o
o
= o
o
o
p
o
9-5
where a=0.87 is a constant determined numerically using HEMP code [86]. Therefore, the
average velocity difference between two neighbouring successive fragments, V
PL
, is
I
PL
= .87
o
o
p
. 9-6
Pfeffer [163] obtained a similar relation for V
PL
using 2-D numerical hydrocode, i.e.
I
PL
= .9
_
c
p
. 9-7
According to Hirsch [4], the velocity difference between the broken fragments in a copper
jet is
I
PL
=
o
p
9-8
where
y
is

the dynamic yield stress of the copper material.
An alternative method to calculate V
PL
was presented by Walters and Summers [163],
where they used Kolskys plastic velocity model to calculate the velocity difference
between the particulated copper jet fragments uing different dynamic constitutive
equations, which were validated against previous field measurements. The advantage of
the Kolsky model is that V
PL
is completely determined by the given constitutive equation.
In addition, the V
PL
values obtained from Kolsky model with different constitutive
equations (i.e. Johnson-Cook, the modified Johnson-Cook and the Zerilli-Armstrong (Z-A)
equations) are very close to the measured V
PL
values for copper jets. Moreover, these
constitutive equations contain the effect of temperature, strain-rate and grain size, which
have direct effects on V
PL
values, and hence, the breakup time. Therefore, the approach
based on Kolsky model will be implemented in this study for zirconium liner using
Johnson-Cook constitutive model (J-C).
Kolsky model [165] considers the equation of the plastic deformation along a wire during
the wire drawing fabrication process. The plastic velocity is defined as the velocity, at
which the wire would break [165]. Based on the relationships between engineering stress
(strain) and true stress (strain), Walters and Summers [163] obtained the average velocity
difference between two neighboring fragments, i.e.
I
PL
=
1
p

o
c

e
c

s
cN
0
e
c

=
1
p

o
e
-o
s
N
0
e

9-9
223
where
N
is the true necking strain, is the true plastic stress, is the true strain and is
the initial liner density; subscript e refers to engineering terms.
9.3 Calculation of the J-C constitutive equation parameters for
zirconium
The Johnson-Cook constitutive equation (J-C) has been established in 1983 to study the
effects of strain, strain-rate and temperature on the flow stresses for some metals and alloys
[121]. The tensile test specimens of zirconium rods were machined from a cylinder of
6mm diameter and 20mm length. The strain-rates that were applied to the samples were
810
-5
, 1.610
-3
and 1100 s
-1
, while the temperatures that were tested were 300 and 400K.
The data published by Ramachandran et al. [166] for the zirconium material were
combined with above measurements to determine the J-C parameters for zirconium. These
combined experimental data extend the extrapolation ranges of temperature and strain-rate,
which, however, was found to have a minor effect on the flow stresses at relatively low
temperatures [166].
The temperature of 400K was chosen as the upper limit of the elevated temperature test
because this temperature rise due to the severe plastic work was predicted from the
Autodyn jet formation model for the zirconium liner driven by HMX explosive. The
temperature calculation will be discussed in details in Section 9.4.
The general form of the J-C constitutive equation is:
o = (A +Be

n
)(1 +Cne
-

)(1 -I
H
m
) 9-10
where is the dynamic flow stress, is the effective plastic strain, A is the yield strength,
B is the hardening constant, n is the hardening exponent, C is the strain-rate constant and
m is the thermal exponent constant.
-

is the normalized effective plastic strain-rate (i.e.
the applied true strain-rate divided by the reference strain-rate). T
H
is the homologous
temperature that can be calculated by:
H
oom
mcI oom
9-11
where T
room
is the room temperature and T
melt
is the melting temperature of the material.
The J-C constants are calculated using the experimental data in this study together with the
data reported in [166]. The general deduced J-C equation for the zirconium material is:
224
o (HPo) = (17 +e

0.6
) [1 +.1n
s

s
c
(1 -I
H
0.5
) 9-12
where e is the true dynamic strain-rate and can be calculated by:
c
-s
9-13
in which e
c
is the engineering strain-rate and e
o
is the reference true initial strain-rate
corresponding to e

co
=810
-5
s
-1
and can be calculated by:
e
o
81
-5 -e
9-14
Thus
s

s
cc
e
c
810
-S
, and Eq. (9-12) can be expressed as
o (HPo) = (17 +e

0.6
) [1 +.1n
e
c

81
-
(1 -I
H
0.5
) . 9-15
Eq. (9-15) is used to calculate the plastic stress-strain curves for the three test specimens as
depicted in Figure 9-1 with their measured curves. It can be concluded that the presented J-
C constitutive equation can reasonably predict the plastic behavior of the zirconium
material up to the true strain of 0.40. The flow stresses calculated using J-C Eq. (9-15) are
quite similar to those calculated by Ref. [166], apart from the grain size effect that is
included only in Z-A model.

Figure 9-1 The measured and the calculated stress-strain curves for the four zirconium test
specimens.

225
9.4 The jet temperature estimation
Von Holle and Trimble [167] measured the temperatures of shaped charge copper jets with
Composition B explosive charges. The charge diameter was 81.3mm. The jet temperatures
were measured at travelling distance of eight times this diameter. The measured average
temperature was 432
o
C with a standard deviation of 76
o
C. Racah [14] investigated
analytically the three different mechanisms of the liner heating when the liner collapsed to
form a jet and when this jet is stretched. The first mechanism is the liner heating by the
detonation wave as it grazes the liner. The calculated temperature for this mechanism is
30K for the copper liner. The second mechanism is the liner heating during the liner
collapse process while the third one is the jet heating during its elongation. In order to find
the zirconium jet temperature, Autodyn was used in this study to simulate shaped charge
jet formation, in which the Mei-Gruneisen thermodynamic model based on shock equation
of state was employed. Since the first mechanism only causes a small increase of
temperature, it focuses on the jet heating due to severe plastic work of the jet rather than
the heat transfer from the detonated explosive products to the jet material [164]. To
validate and verify the hydrocode results for the jet temperature calculations, the shaped
charge BRL-81.3 [168] was modelled using the jet formation simulation algorithm, during
which the jet temperature was recorded at different times. It was assumed that the plastic
deformation is continuous and smooth along the jet until its breakage, while the total jet
heating due to these mechanisms was considered. The jet temperature was found to vary in
the hoop direction along the crossection area of the jet. However, the absolute value of 780
K (i.e 482
o
C+25
o
C) is observed to be common along the jet profile, while the experimental
measurement showed a temperature increase of 428
o
C (i.e. 701K) using Composition B
explosive. This means that the zirconium jet temperature can be reasonably predicted using
hydrocode.
The OWP shaped charges with zirconium liners of different thicknesses were modelled in
the same way and their jets were allowed to elongate up to 3 times its diameter. The jet
temperatures were investigated over the entire jet length. From the produced jet
temperature contours depicted in Figure 9-2 for the zirconium liner with thickness of
1.7mm, it can concluded that the temperature is not constant over the extended zirconium
jet and its slug, but ranges between 425K and 402K at the rear and the tip of the jet,
respectively. The jet elements were found to have the same temperature gradient when the
jet is allowed to elongate to a distance of seven times the charge diameter. This means that
the jet temperature can be assumed to be approximately constant from the time of jet
226
collapse on the liner axis to the moment when the jet breaks up. Walters and Summers
[163] assumed an isothermal deformation of copper shaped charge jets in a similar attempt
to calculate the V
PL
for copper jet. The assumption gave very accurate results of V
PL
using
different constitutive equations when compared with the experimentally measured values
for a range of shaped charges. Thus, it is reasonable to assume a homogeneous distribution
of jet temperature and an isothermal process during the jet stretching.

Figure 9-2 The zirconium jet temperature contours at the moment of jet formation for the
zirconium liner with a wall thickness of 1.7mm.
9.5 Results
9.5.1 Calculations of the necking strain and V
PL
using J-C constitutive equation
The J-C constitutive equation is used to calculate V
PL
using the true stress and strain in
Eq.(9-12). Assuming that velocity variables of both the jet tip and its slug remain
unchanged during the stretching process, the engineering strain-rate will be independent of
strain, and therefore d
c c

equals zero. Besides, it has been shown that the jet
stretching deformation process is isothermal (i.e.
d1
H
m
ds
= ). Therefore, differentiation of J-
C Eq. (9-10) gives:
o
e
= (Bne

n-1
) _1 +Cn
e
c

e
co
_(1 -I
H
m
).
9-16
Since
s

s
cc
e
c
810
-S
, thus
d
ds
= (Bne
P
n-1
) [1 +Cn
e
c
81
-
(1 -I
H
m
). 9-17
Applying the stability condition in order to calculate the necking strain (
N
), i.e.
d
ds

which leads to
227
(Bne
N
n-1
) _1 +Cn

c

eo
_(1 -I
H
m
) = (A +Be
N
n
) _1 +Cn

c

eo
_(1 -I
H
m
) 9-18
or (Bne
N
n-1
) = (A +Be
N
n
) 9-19
according to Eqns.(9-10) and (9-17).
The solution of this equation gives the values of the maximum necking strain
N
, at which
the V
PL
will be estimated. Substitute the values of the J-C constants into Eq. (9-19), we
have .6
N
-0.4
-
N
0.6
= .S78 .
The necking strain is
N
=0.351, which is independent of strain-rate and temperature of the
jet. The velocity difference between the jet fragments can be calculated using Eq.(9-9)
based on true stress and true strain
o
e
-o = ((Bne
P
n-1
) -(A +Be

n
)) _1 +Cn

c

eo
_(1 -I
H
m
) ; 9-20
I
PL
=
1
p
((Bne

n-1
) - (A + Be

n
)) _1 + Cn

c

eo
_(1 - I
E
m
)
s
N
0
u ;
9-21
I
PL
=
_
B(1 - I
E
m
) _1 + Cn

c

eo
_
p
(ne

n-1
) - _e

n
+
A
B
]
0.351
0
u .
9-22
The integration is done using the area under the curve over the strain range from 0 to
0.351. The area was found to be 0.265.
Substituting the values of the J-C parameters, the velocity difference between two
neghibouring zirconium fragments is calculated by
I
PL
= .26
_
1
6
(1 -I
H
0.5
) _1 +.1n
e
c
8 1
-
]
6

9-23
Table 9-1 lists the shaped charge parameters related to the jet breakup using the output data
from the jetting analysis. The engineering strain rate
c
can be calculated as follow:
c
=
I
tip
-I
rcor
]ct
,
9-24
where L
jet
is the jet length; V
tip
and V
rear
are the jet tip and slug velocities, respectively.
228
The calculated V
PL
values and the number of fragments are also listed in Table 9-1. The
melting temperature of the zirconium material is 1852
o
C [169], while the reference
temperature was taken to be 300K.
Table 9-1 Shaped charge parameters related to the jet breakup for the studied zirconium
liners with different liner wall thicknesses
T
L
(mm) 0.7 1 1.3 1.7 2 2.3 2.7 3 3.3 3.7 4
CD (mm) 36 36 36 36 36 36 36 36 36 36 36
T
L
/CD 0.019 0.027 0.036 0.047 0.055 0.063 0.075 0.083 0.091 0.102 0.111
V
tip
(m/s) 7307 6879 6634 6262 5928 5647 5352 5131 4905 4661 4496
V
slug
(m/s) 4872 2721 1477 1338 1239 1155 1073 1226 441 414 395
Initial jet length
L
jet
(mm)
44.0 61.8 67.1 56.4 45.2 41.5 35.0 32.4 29.8 28.7 26.9
Liner Mass (g) 9.89 15.8 20.9 26.9 31.4 35.6 40.5 44.2 47.6 52.2 54.9
Jet mass (g) 2.01 3.57 4.69 5.49 6.09 6.63 7.32 7.17 9.31 9.88 9.93
Jet KE (kJ) 47.2 53.4 52.2 49.1 46.7 44.3 40.7 38.3 34.5 31.5 29.3
Jet temp. (K) 450 430 416 402 390 380 365 355 347 340 330
V
PL
(m/s) 64.6 65.5 66.2 66.9 67.6 68.2 69.1 69.7 70.3 70.8 71.7
1/ V
PL
(s/mm) 15.5 15.3 15.1 14.9 14.8 14.7 14.5 14.3 14.2 14.1 13.9
No. of fragments 38 63 78 73 69 66 62 56 63 60 57
Note: T
L
is the liner thickness and CD is the charge diameter.
The calculated velocity difference between the zirconium jet fragments is compared with
that of the copper material measured by Hirsh [5], as shown in Figure 9-3. It shows that, at
a certain value of T
L
/CD, the zirconium material has lower V
PL
values, which means that
the zirconium jet has a longer time of elongation before its breakup. Therefore the
zirconium liner showed a remarkable increase in its ductility when compared with copper
liner. Two experimental measurements of the velocity difference between the zirconium jet
fragments (V
PL
) were 64.3 and 74.9m/s, respectively, according to Bourne et al.[13], which
are close to the calculated values as shown in Figure 9-3. Bourne et al experiments to
measure the V
PL
values were carried out using 2 x-ray radiography pictures at two times in
the range of 1000s and about 50 s apart.
229

Figure 9-3 The velocity difference between the jet fragments for different T
L
/CD values for
both copper and zirconium jets.
Since the specific breakup time is considered as a characteristic property for a given
shaped charge with a certain liner thickness, the obtained V
PL
values and the scaled
(T
L
/CD) values are correlated as depicted in Figure 9-4 for zirconium material with copper
material as a baseline. The relation for zirconium is almost linear and can be described by
1
v
PL
1
L
C
9-25
which is a similar to that proposed by Hirsch [5] for the copper material, which is
1
I
PL
= 1S.886 -11.19 _
I
L
C
].
9-26
Eqns. (9-25) and (9-26) have the same linear form, but the slope of Eq. (9-25) is much
smaller than that of Eq. (9-26). Therefore, for a given value of T
L
/CD, zirconium can
achieve higher ductility than copper, which agrees with the experimental observations in
Bourne et al. [13].
230

Figure 9-4 The specific breakup time (1/V
PL
) as a function of the scaled value (T
L
/CD) for
zirconium and copper [5].
The importance of Eq.(9-25) is that it gives a direct method to estimate the specific
elemental breakup time if T
L
and CD are known for a zirconium shaped charge. The
reciprocal of V
PL
(i.e. 1/ V
PL
) represents the liner specific breakup time per 1mm of jet
length (i.e. s/mm) and can be used directly to calculate the breakup time of a shaped
charge jet according to Eq.(9-1), i.e.
t
b
= 2r _ _

__ 9-27
where the initial jet element radius r can be measured from flash x-ray or calculated using
Eq.(9-2), in which the collapse angle is obtained from the jetting analysis. If 1/V
PL
is in
s/mm and the radius r is in mm, then the breakup time will be in s. The calculated jet tip
radii and their breakup times for the four OWPs are shown in Table 9-2. This table shows
the clear difference among the four liners in their jet tip radius and their breakup times
although they have the same charge design except the liner thickness. Generally the higher
the jet velocity of the liner element, the lower the breakup time of this element.
Table 9-2 The jet tip radius and jet breakup time for zirconium OWPs.
T
L
(mm) 1 1.3 1.7 2
V
tip
(m/s) 7879 7634 7262 6928
Jet tip radius (mm) 1.135 1.25 1.765 1.84
Breakup time (s) 34.5 38.15 52.59 54.5
231
9.6 Summary
- The J-C constitutive equation constants are determined for the zirconium material
and used to estimate the characteristic V
PL
for some zirconium OWPs.
- The specific breakup time of the liner material, which is the reciprocal of the
velocity difference between the particulated jet fragments, is calculated for
zirconium shaped charge liners and found to be in the range of 64.6 to 71.7 m/s for
the liner wall thickness ranges between 0.7 and 4mm, respectively.
- The breakup time of zirconium shaped charges is presented, by which the breakup
time can simply be calculated by the jet radius and the scaled T
L
/CD values.

Chapter 10: Conclusions and Future Work
232
CHAPTER.10 CONCLUSIONS AND
RECOMMENDATIONS FOR FUTURE STUDIES
10.1 Introduction
This research covers a wide range of parametric studies based on Autodyn hydrocode and
experiments for a range of OWPs with different liner materials and different shapes in
order to investigate the characteristics of the produced jets and their penetration
capabilities.
10.2 Conclusions
The following findings have been observed:
- The jetting analysis validation showed that the difference between the measured tip
velocity and the numerical simulation is only 1.34%, while the difference between
the measured and the calculated penetration depth was only 1.6%, which means that
Autodyn can be used effectively in the jetting analysis and jet penetration.
- The produced jet tip velocity of an OWP has nearly the same value as the detonation
velocity of the used explosive and is about 2.5 times the Gurney velocity of the
explosive.
- The optimum design of a shaped charge with liner thickness of 1.5mm, cone apex
angle of 56
o
and total RDX mass of 26g can achieve 100cm depth of penetration into
35MPa concrete.
- Behind 20mm water stand-off distance, the penetration depth of OWP begins to
decrease dramatically due to water resistance and jet erosion.
- Increasing the OWP steel charge casing from 1 to 8mm can increase the jet tip
velocity by 800m/s.
- A simple empeical relation was presented to estimate the Gurney velocity of an
explosive material in terms of its C-J pressure and its impulse. The maximum
difference between the Gurney velocity calculated by this relation and that found in
the survey for a range of explosive materials was 5.48%.
- Allison-Vitalli formula (i.e. VO model) of jet penetration depth is modified to
include the target strength effect using Johnsons damage number and the
confinement pressure effect using the Drucker-Prager model. The penetration
reduction correction terms due to the consideration of concrete strength were found
233
to be 0.14 and 0.29 for the concrete targets with unconfined compressive strength of
26 and 55 MPa, respectively.
- The critical collapse angle as a function of flow velocity was calculated numerically
and analytically for the zirconium liner material to determine the coherency of
shaped charge jet. Four zirconium OWP liners with different shapes were studied and
found to produce coherent jets. The bi-conical liner exhibited the largest penetration
depth into target, which is 22% greater than that of the baseline conical zirconium
liner.
- It was demonstrated that jet density reduction should be considered in the study of jet
penetration. A modified VO penetration model based on non-uniform jet density is
presented in this research, where the penetration reduction due to the jet density
reduction was found to be 12.5, 12.8 and 19.4% for the zirconium, copper and
copper-tungsten jets, respectively.
- The characteristic V
PL
for some zirconium OWPs with different liner wall
thicknesses ranging between 0.7 and 4mm, respectively, was calculated. It was found
to be in the range of 64.6 to 71.7 m/s for the liner wall thickness ranges between 0.7
and 4mm, respectively. The V
PL
was found to increase linearly with the increase in
the liner wall thickness. For a given value of scaled liner thickness to the charge
diameter (T
L
/C
D
), zirconium can achieve higher ductility than copper.
- A simple breakup time relation for zirconium OWPs is presented based on the
characteristic V
PL
, and the jet radius.
10.3 Future work
The following future research studies may be pursued:
- The proposed modified VO model has ignored target compressibility. Further study
should be performed to assess whether the target compressibility during the jet
penetration has a significant effect on the jet penetration.
- The homogeneity of the powder liner density may be largely improved using
alternative pressing methods (e.g. cold/hot isostatic pressing), which is very
important for the formation of a high quality jet.
- Other manufacturing processes could be further improved. For example, spinning
techniques could be used to replace the current deep drawing method. Automatic
filling and testing facilities may be used to replace the current manual filling,
pressing and detonating procedures to reduce the variation and uncertainty.
234
- Jet density reduction and non-uniform distribution should be further investigated
using more accurate instantaneous jet density measurement and numerical tools.
- Dynamic behavior of zirconium in a wider range of temperature and strain-rate
should be tested to increase the accuracy and reliability of the zirconium J-C model.
- The zirconium jet temperature was estimated numerically. Reliable measurements of
jet temperature are needed to validate the hydrocode.

References
235
REFERENCES
[1] L. Zernow, The density deficit in stretching shaped charge jets. International
Journal of Impact Engineering, 1997. 20, 610, p. 849-859.
[2] F. E. Allison and R. Vitali, A New method of computing penetration variables for
shaped charge jets, Ballistic Research Laboratory Report No. 1184, 1963.
[3] G. R. Cowan and A. H. Holtzman, Flow configurations in colliding plates:
explosive bonding, Journal of Applied Physics, 1963. 34, 4, p. 928-939.
[4] E. Hirsch, A Formula for the shaped charge jet breakup-time. Propellants,
Explosives, Pyrotechnics, 1979. 4, 5, p. 89-94.
[5] E. Hirsch, Scaling of the shaped charge jet breakup time. Propellants, Explosives,
Pyrotechnics, 2006. 31, 3, p. 230-233.
[6] W. P. Walters and J. Zukas, Fundamentals of shaped charge. Wiley Interscience
Publication, New York, 1989.
[7] A. Schwartz, M. Kumar, and D. Lassila, Analysis of intergranular impurity
concentration and the effects on the ductility of copper-shaped charge jets. Journal
of Metallurgical and Materials Transactions A, 2004. 35, 9, p. 2567-2573.
[8] E. Tamer, The factors affecting the performance of explosive reactive Armours.
2004, Master of Science Thesis, Military Technical College: Cairo, Egypt.
[9] C. E. Joachim, Rapid runway cutting with shaped charges, Technical Report No.
ADP001712, 1983.
[10] N. J. Schoeneberg, Generation of transient antennas using cylindrical-shaped
charges. Master Thesis, Texas Tech. University. 2003.
[11] W. H. Lee, Oil well perforator design using 2D Eulerian code. International
Journal of Impact Engineering, 2002. 27, 5, p. 535-559.
[12] S. L. Renfre, Liner and improved shaped charge especially for use in a well pipe
perforating gun. Patent. US No. 5,509,356, 1996.
[13] B. Bourne, K. G. Cowan, J. P. Curtis, Shaped charge warheads containing low
melt energy metal liners, Proceedings of the 19
th
International Symposium on
Ballistics. Switzerland, 2001.
[14] E. Racah, Shaped charge jet heating, Propellants, Explosives, Pyrotechnics, 1988.
13, p. 178-182.
[15] Z. Bogdan and W. Zenon, Formation of jets by shaped charges with metal powder
liners, Propellants, Explosives, Pyrotechnics, 2008, 33, 6, p. 482-487.
[16] L. A. Glenn, J. B. Chase, J. Barker, D. J. Leidel, Experiments in support of
pressure enhanced penetration with shaped charge perforators. 18
th
International
Symposium & Exhibition on Ballistics, San Antonio, TX, 1999.
References
236
[17] D. Davinson and D. Pratt, A hydrocode-designed well perforator with exceptional
performance, 17
th
International Symposium on Ballistics. 1998. Midrand, South
Africa, p.1-9.
[18] M. Held, Determination of the material quality of copper shaped charge liners.
Propellants, Explosives, Pyrotechnics, 1985. 10, 5, p. 125-128.
[19] W. P. Walters and D. R. Scheffler. A Method to increase the tip velocity of a
shaped charge jet. 23
rd
International Symposium on Ballistics, 2007. Tarragona,
Spain, p. 135-144.
[20] M. Held, Hydrodynamic theory of shaped charge jet penetration. Journal of
Explosives and Propellants. Taiwan, 1991. 7.
[21] E. Tamer and L. Q.M., Parametric analysis of the penetration performance of
shaped charges used as oil well perforators, 41
st
International. ICT-Conference.
2010, Karlsruhe, Germany.
[22] M. H. Keshavarz and S. Abolfazl, The simplest method for calculating energy
output and Gurney velocity of explosives, Journal of Hazardous materials, 2006,
131, 1-3, p.1-5.
[23] T. Urbanski, M. Jureki and S. Laverton, Chemistry and Technology of Explosives.
Vol. 1. Pergamon Press Oxford, UK, 1964.
[24] D. J. Leidel and J. P. Lawson, High performance powder metal mixtures for shaped
charge liners. 2001, US Patent No. US7,811,354,B2.
[25] H. Elwany, The use of polymer bonded explosives in improved linear shaped
charge designs, PhD Thesis, Cranfield University, U.K. 1994.
[26] P.R.C. Green. Ammunition for the Land Battle. UK: BRASSEYS, 4
th
Edition. Vol.
4. 1990.
[27] P. Y. Chanteret and A. Lichtenberger, About varying shaped charge liner thickness.
17
th
International symposium on ballistics. 1998, Midrand, South Africa.
[28] M. Defourneaux, Theorie hydrodynamique des charges creuses. Mememorial
dArtillerie Francaise, 1970. 44, 2.
[29] J. Brown, P. J. Edwards and P. R. Lee, Studies of shaped charges with built-in
asymmetries. Part II: Modelling. Propellants, Explosives, Pyrotechnics, 1996. 21,
2, p. 59-63.
[30] M. Moyses, Penetration by shaped charge jets with varying off-axis velocity
distributions. 17
th
International Symposium on Ballistics, 1998. Midrand, South
Africa.
[31] M. Held, Liners for shaped charges, Journal of Battlefield Technology, 2001, 4, 3,
p. 1.
[32] W. Guo, S. K. Li, F. C. Wang and W. M. Wang, Dynamic recrystallization of
tungsten in a shaped charge liner. Scripta Materialia, 2009. 60, 5, p. 329-332.
References
237
[33] T. F. Wang, and H. R. Zhu, Copper-Tungsten shaped charge liner and its jet.
[34] B. Bourne, P. Jones and K. G. Cowan, Grain crystallographic texture effects on the
performance of shaped charges. International Symposium on Ballistics, 1993.
Quebec, Canada.
[35] V. Y. Gertsman, M. Hoffmann, H. Gleiter and R. Birringer, The study of grain size
dependence of yield stress of copper for a wide grain size range. Acta Metallurgica
et Materialia, 1994. 42, 10, p. 3539-3544.
[36] M. Meyers, U. Andrade, and A. Chokshi, The effect of grain size on the high-strain,
high-strain-rate behavior of copper. Metallurgical and Materials Transactions A,
1995. 26, 11, p. 2881-2893.
[37] S. Fujiwara and K. Abiko, Ductility of ultra high purity copper, Journal de
physique 1995. 5, (Colloque C7), p. 111.
[38] F. J. Zerilli and R. W. Armstrong, Constitutive relations for the plastic deformation
of metals. American Institute of Physics, API Conference proceeding, 1994, 309, p.
989-992.
[39] J. P. Curtis, A Breakup model for shaped charge jets. 16
th
International Symposium
on Ballistics. 1996. San Francisco.
[40] W. H. Tian, A. L. Fan, H. Y. Gao, J. Luo and Z. Wang, Comparison of
microstructures in electroformed copper liners of shaped charges before and after
plastic deformation at different strain-rates. Materials Science and Engineering A,
2003. 350, (1-2), p. 160-167.
[41] L. Lu, S. X. Li, and K. Lu, An abnormal strain-rate effect on tensile behavior in
nanocrystalline copper. Scripta Materialia, 2001. 45, 10, p. 1163-1169.
[42] J. W. Reese and A. Hetz, Coated metal particles to enhance oil field shaped charge
performance. 2001, US Patent No. US 2002/0178962 A1.
[43] Y. Liu and Z. Shen, Numerical simulation on formation and penetration target of
powder metal shaped charge jet, International Conference on Computer
Application and System Modelling (ICCASM), 2010, 9, p. 518-521.
[44] W. P. Walters, P. Peregino, R. Summers and D. Leidel, A Study of jets from un-
sintered powder metal lined non-precision small calibre shaped charge, Army
Research Lab. Report No. ARL-TR-2391, 2001.
[45] E. Bakker, K. Veeken, L. Behrmann, P. Milton, G. Shirton, A. Salsman and I.
Walton, The new dynamics in underbalanced perforating. Oilfield Review, 15, 4,
2003.
[46] B. Grove, J. Heiland and I. Walton, Geologic materials' response to shaped charge
penetration. International Journal of Impact Engineering, 2008. 35, 12, p. 1563-
1566.
References
238
[47] C. . Karacan, and P.M. Halleck, Comparison of shaped-charge perforating
induced formation damage to gas- and liquid-saturated sandstone samples. Journal
of Petroleum Science and Engineering, 2003. 40, 1-2, p. 61-75.
[48] Y. Luo, and Z. Shen, Study on orientation fracture blasting with shaped charge in
rock. Journal of University of Science and Technology Beijing, Mineral, Metallurgy
and Material, 2006. 13, 3, p. 193-198.
[49] G. Birkhoff, D. P. MacDougall, E. M. Pugh, and G. Taylor, Explosives with lined
cavities, Journal of Applied Physics, 1948, 19, 6, p. 563-582.
[50] E. M. Pugh, R.J. Eichelberger and N. Rostoker, Theory of jet formation by charges
with lined conical cavities, Journal of Applied Physics, 1952. 23, 5, p. 532-536.
[51] S. K. Godunov, A. A. Deribas and V. I. Mali, Influence of material viscosity on the
jet formation process during collisions of metal plates. Combustion, Explosion and
Shock Waves, 1975. 11, 1, p. 1-13.
[52] W. P. Walters, Influence of material viscosity on the theory of shaped-charge jet
formation. Memorandum Report No. ARBRL-MR-02941, 1979.
[53] E. Hirsch, A simple representation of the Pugh, Eichelberger, and Rostoker solution
to the shaped charge jet formation problem," Journal of Applied Physics, 1979. 50,
p. 4667-4670.
[54] G. X. Liu, The simplified model for predicting shaped charge jet parameters,
[55] E. F. Allison and R. Vitalli, An application of the jet-formation theory to a 105mm
shaped charge. BRL Technical Report No. MD AD0277458, 1962.
[56] R. J. Eichelberger, Re-examination of the nonsteady theory of jet formation by
lined cavity charges. Journal of Applied Physics, 1955. 26, 4, p. 398-402.
[57] J. Carleone, P. C. Chou and C. A. Tanzio, One-dimensional computer code to
model shaped charge liner collapse, jet formation and jet properties. 1975, Dyna
East corporation.
[58] G. Randers-Pehrson, An improved equation for calculating fragment projection
angle. International Symposium on Ballistics. 1976. Daytona Beach, FL.
[59] L. Behrmann and N. Birnabum, Calculation of shaped charge jets using
Engineering approximations and finite difference computer codes. Technical
Report No. AFATL-TR-73-160, 1973.
[60] J. Carleone, Mechanics of shaped charges, M. Baltimore, Editor. 1987,
Computational Mechanics Associates.
[61] R. Gurney, The initial velocities of fragments from Bombs, shell and grenades.
Aberdeen Proving Ground, Ballistic Research Lab. Report No. Md. BRL-405,
1943.
[62] L. H. Thomas, Theory of the explosion of cased charges of simple shape. 1944.
References
239
[63] T. Sterne, A Note on the initial velocities of fragments from warheads. Technical
Report No. BRL-648, September, 1947.
[64] P. Y. Chanteret, An analytical model for metal acceleration by grazing detonation.
7
th
International Symposium on Ballistics. 1983. Hague, Netherlands.
[65] J. E. Kennedy, Gurney energy of explosives: estimation of the velocity and impulse
imparted to driven metal, Technical Report No. SC-RR-70-790. 1970.
[66] G. E. Jones, J. E. Kennedy, and L.D. Bertholf, Ballistics calculations of R.W.
Gurney. American Journal of Physics, 1980. 48: p. 264-269.
[67] P. C. Chou and W.J. Flis, Recent developments in shaped charge technology.
[68] G. E. Duvall and J. O. Erkman and C. M. Ablow, Explosive acceleration of
projectiles. Israel Journal of Technology, 1969, 7, 6, p. 469.
[69] L. A. Shushko, B. I. Shekhter and S. L. Krys'kov, Bending of a metal strip by a
sliding detonation wave. Combustion, Explosion, and Shock Waves, 1975. 11, 2, p.
229-236.
[70] E. Hirsch, Improved Gurney formulas for exploding cylinders and spheres using
Hard Core Approximation. Propellants, Explosives, Pyrotechnics, 1986. 11, 3, p.
81-84.
[71] P. C. Chou, Improved formulas for velocity, acceleration, and projection angle of
explosively driven liners. Propellants, Explosives, Pyrotechnics, 1983. 8, 6, p. 175-
183.
[72] A. Koch, N. Arnold, and M. Estermann, A simple relation between the detonation
velocity of an explosive and its Gurney energy. Propellants, Explosives,
Pyrotechnics, 2002. 27, 6, p. 365-368.
[73] I. G. Henry, The Gurney formula and related approximations for the high-explosive
deployment of fragments. Report No. PUB-189, Aerospace Group, Hughes Aircraft
roration, California, 1967.
[74] M. Held, Plate velocities for asymmetric sandwiches. Propellants, Explosives,
Pyrotechnics, 1997. 22, 4, p. 218-220.
[75] A. Deribas, Physics of explosive work hardening and welding, N. Nauka, Editor.
1972.
[76] H. R. Kleinhanss, Experimentelle untersuchungen zum kol- iapsprozeb bei
hohlladungen. Proceedings of The 3
rd
International Symposium on Military
Applications. 1971. FRG.
[77] V. F. Lobanov, Numerical simulation of flow during compression of cylindrical
samples by a glancing detonation wave. Journal of Applied Mechanics and
Technical Physics, 1975. 16, 5, p. 795-798.
[78] Y. A. Gordopolov, A. N. Dremin and A. N. Mikhailov, Experimental determination
of the dependence of the wavelength on the angle of collision in the process of the
References
240
explosive welding of metals. Combustion, Explosion, and Shock Waves, 1976. 12,
4, p. 545-549.
[79] A. Deribas and I. Zakharenko, Surface effects with oblique collisions between
metallic plates. Combustion, Explosion, and Shock Waves, 1974. 10, 3, p. 358-367.
[80] D. P. Bauer and S. J. Bless, Strain-rate effects on ultimate strain of copper.
Technical Report No. AFML-TR-79-4021, 1979.
[81] P. C. Chou and Carleone, Calculation of shaped charge jet strain, radius and
breakup time. Technical Report No. BRL-CR-246, 1975.
[82] J. M. Walsh, R. G. Shreffler and F. J. Willig, Limiting conditions for jet formation
in high velocity collisions, Journal of Applied Physics, 1953. 24, 3, p. 349-359.
[83] P. C. Chou, J. Carleone and R. R. Karpp, Criteria for jet formation from impinging
shells and plates, Journal of Applied Physics, 1976. 47, 7, p. 2975-2981.
[84] A. Cowan, Hydrocode and analytical code modeling of the effect of liner material
grain size on shaped charge jet breakup parameters. 17
th
on Ballistics. 1998. Midrand, South Africa.
[85] E. Hennequin, Modelling of the shaped charge jet breakup. Propellants, Explosives,
Pyrotechnics, 1996. 21, 4, p. 181-185.
[86] B. Haugstad, On the breakup of shaped charge jets. Propellants, Explosives,
Pyrotechnics, 1983. 8, 4, p. 119-120.
[87] G. Pfeffer, Determination par simulations numtriques de L'etat et des lois de
fragmentation des jets de charges Creusses " 5
th
International Symposium on
Ballistics, 1980. Toulouse, France.
[88] P. C. Chou and J. Carleone, A One dimensional theory to predict the strain and
radius of shaped charge jets. Proceeding of the first International Symposium on
Ballistics. 1974. Orlando, FL.
[89] P. C. Chou and J. Carleone, Breakup of shaped charge jet. 2
nd
International
Symposium on Ballistics. 1976. Daytona Beach, USA.
[90] R. Karpp and J. Simon, An estimate of the strength of a copper shaped charge jet
and the effect of strength on the breakup of a stretching jet. US Army Ballistic
Research Lab., BRL Report, 1976.
[91] C. Miller, A New approach to the numerical analysis of shaped charge jets.
Proceedings of the 6
th
International Symposium on Ballistics. 1981. Orlando, FL.
[92] C. Miller, Generation of necks in an elongated shaped charge jet. Proceedings of
The BRL Symposium on Ballistics, ADPA. 1982.
[93] J. Carleone and P. C. Chou, Shaped charge jet stability and penetration
calculations. Technical Report No. ADA050117, 1977.
[94] J. M. Walsh, Plastic instability and particulation in stretching metal jets. Journal of
Applied Physics, 1984. 56, 7, p. 1997-2006.
References
241
[95] J. E. Backofen, The use of analytical computer models in shaped charge design.
[96] D. C. Pack and W. M. Evans, Penetration by high-velocity (Munroe) jets: I,
Proccedings of the Physical Society, London, 1950. B64, p. 298.
[97] R. Eichelberger, Experimental test of the theory of penetration by metallic jets.
Journal of Applied Physics, 1956. 27, 1, p. 63-68.
[98] E. M. Pugh, A Theory of target penetration of jets, National Defense Research
Committee Armour and Ordnance Report No. A-274 (OSRD No. 3752), 1944.
[99] O. A. Klamer, Shaped charge scaling. Technical Memorandum No. AD0600273,
1964, Dover, NG, USA.
[100] G. R. Abrahamson and J. N. Goodier, Penetration by shaped charge jets of
nonuniform velocity. Journal of Applied Physics, 1963. 34, 1, p. 195-199.
[101] R. Dipersio, J. Simon, B. Merendino, Penetration of shaped-charge jets into
metallic targets. Ballistic Research Laboratory Memorandum Report No. 1296.
1965.
[102] W. Schwartz, Modified SDM model for the calculation of shaped charge hole
profiles. Propellants, Explosives, Pyrotechnics, 1994. 19, 4, p. 192-201.
[103] W. P. Walters, W. J. Flis, P. C. Chou, A survey of shaped-charge jet penetration
models, International Journal of Impact Engineering, 1988, 7, 8, p. 307-325.
[104] V. P. Alekseevskii, Penetration of a rod into a target at high velocity. Combustion,
Explosion, and Shock Waves, 1966. 2, 2, p. 63-66.
[105] N. S. Sanasaryan, Penetration of a cumulative jet into a barrier. Fluid Dynamics,
1975. 10, 6, p. 997.
[106] D. R. Christman and J. W. Gehring, Analysis of high-velocity projectile penetration
mechanics, Journal of Applied Physics, 1966. 37, 4, p. 1579.
[107] J. R. Doyle and R. L. Buchholz, Design, development, fabrication and testing
program to demonstrate feasibility of the mass focus fragmentation warhead. 1973.
[108] A. Tate, A theory for the deceleration of long rods after impact, Journal of the
Mechanics and Physics of Solids, 1967. 15, p. 387-399.
[109] A. Tate, Further results in the theory of long rod penetration. Journal of the
Mechanics and Physics of Solids, 1969. 17, 3, p. 141-150.
[110] A. Tate, Long rod penetration models-Part I. A flow field model for high speed
long rod penetration. International Journal of Mechanical Sciences, 1986. 28, 8, p.
535-548.
[111] M. Held and A. A. Kozhushko, Radial crater growing process in different materials
with shaped charge jets. Propellants, Explosives, Pyrotechnics, 1999. 24, 6, p. 339-
342.
References
242
[112] A. Tate, Long rod penetration models-Part II. Extensions to the hydrodynamic
theory of penetration. International Journal of Mechanical Sciences, 1986. 28, 9 p.
599-612.
[113] D. A. Matuska, A model for high velocity penetrating. Proc. of ARO Workshop on
Computational Aspects of Penetration Mechanics. 1982. Aberdeen Proving Ground
MD.
[114] M. Held, N. S. Huang, D. Jiang and C. Chang, Determination of the crater radius as
a function of time of a shaped charge jet that penetrates water. Propellants,
[115] M.S. Cowler, Autodyn jetting Analysis tutorial, 2001.
[116] S. Malcolm, Autodyn theory manual, R. 3.0, Editor. 1997, Century Dynamics:
USA.
[117] V. S. Berg and D. S. Preece, Shaped charge induced concrete damage predictions
using RHT constitutive modeling, Journal of International Society of Explosives
Engineers, 2004, 2. p.261-272.
[118] G. Baudin and R. Serradeill, Review of Jones-Wilkins-Lee equation of state. New
Models and Hydrocodes for Shock Wave Processes in Condensed Matter, Paris,
France, Edited by Laurent Soulard; EPJ Web of Conferences, 2010, 10, p. 21.
[119] E. L. Lee, H. C. Hornig and J. W. Kury, Adiabatic expansion of high explosive
detonation products, California University, Livermore. Lawrence Radiation Lab
Report No. UCRL-50422, 1968.
[120] C. M. Tarver, W. C. Tao and C. G. Lee, Sideways plate push test for detonating
solid explosives, Propellants, Explosives, Pyrotechnics, 1996. 21, 5, p. 238-246.
[121] G. R. Johnson and W. H. Cook, A Constitutive model and data for metals subjected
to large strains, high strain-rates and high temperatures. Proceedings of the 7
th
Int.
Symposium on Ballistics. Netherlands, 1983. p. 541-547.
[122] W. Herrmann, Constitutive equation for the dynamic compaction of ductile porous
materials, Journal of Applied Physics, 1969, 40, 6, p. 2490-2499.
[123] N. Heider and S. Hiermaier, Numerical simulation of the performance of tandem
warheads, 19
th
International Symposium of Ballistics, Interlaken, Switzerland,
2001, p.1493.
[124] W. Riedel, K. Thoma, S. Hiermaier and E. Schmolinske, Penetration of reinforced
concrete by beta-b-500. numerical analysis using a new macroscopic concrete
model for hydrocodes, 9
th
International Symposium on the Interaction of the Effects
of Munitions with Structures, 1999. p. 1.
[125] Z. Tu and Y. Lu, Evaluation of typical concrete material models used in hydrocodes
for high dynamic response simulations. International Journal of Impact
Engineering, 2009. 36, 1, p. 132-146.
[126] J. Leppnen, Concrete structures subjected to fragment impacts, PhD thesis,
Chalmers University of Technology, Gteborg, Sweden, 2004, p. 40.
References
243
[127] C. J. Hayhurst, R. A. Clegg, I. H. Livingstone and N. J. Francis, The application of
SPH techniques in Autodyn 2-D to ballistic impact problems, 16
th
International
Symposium on Ballistics, San Fransisco, USA, 1996.
[128] J. Brookes and W. Yang, Target for testing perforating system, 2001. US Patent
No. US6,238,595 B1.
[129] R. A. Clegg, C. J. Hayhurst and I. Robertson, Development and application of a
Rankine plasticity model for improved prediction of tensile cracking in ceramic and
concrete materials under impact, 14
th
DYMAT Technical Meeting. 2002: Sevilla,
Spain.
[130] B. Adel, Numerical approach in predicting the penetration of limestone target by an
ogive-nosed projectile using Autodyn 3-D, 12
th
International Colloquium on
Structural and Geotechnical Engineering (ICSGE), 2007, Egypt.
[131] B. Lusk, W. Schonberg, J. Baird, R. Woodley and W. Noll, Using coupled Eulerian
and Lagrangian grids to model explosive Interactions with buildings. Technical
report No. ADA481951, 2006.
[132] J. Leppnen, Concrete subjected to projectile and fragment impacts: Modelling of
crack softening and strain-rate dependency in tension. International Journal of
Impact Engineering, 2006. 32, 11, p. 1828-1841.
[133] Marinko Ugri, Duan Ugri, FEM techniques in shaped charge simulation,
Scientific Technical Report Review,Vol. LVIX, No.1, 2009, p. 26-34.
[134] J. E. Backofen, Air cratering by eroding shaped charge jets. 23
rd
International
Symposium on Ballsitics. 2007. Spain. p. 689-696.
[135] D. Hasenberg, Consequences of coaxial jet penetration performance and shaped
charge design criteria, Technical Report No. NPS-PH-10-001, 2010.
[136] J. Curtis, F. Smith and A. White, The formation and stretching of bi-material
shaped charge jets. American Institute of Physics Conference proceedings, 2011,
1426. p.116-119.
[137] M. H. Keshavarz, Prediction method for specific impulse used as performance
quantity for explosives. Propellants, Explosives, Pyrotechnics, 2008. 33, 5, p. 360-
364.
[138] D. R. Hardesty and J. E. Kennedy, Thermochemical estimation of explosive energy
output. Combustion and Flame Journal, 1977. 28, p. 45-59.
[139] M. G. Vigil, Optimized conical shaped charge design using the SCAP code, Sandia
National Laboratories Report No. SAND88-1790. 1988.
[140] J. A. Zukas and W. Walters, Explosive effects and applications. 1997, Springer.
[141] W. Johnson, Impact strength of materials. London, Edward Arnold, 1972, p. 303.
[142] Q.M. Li, Johnsons damage number in impact dynamics, International Journal of
Impact Engineering, 2011. 38, 5, p. 273-274.
References
244
[143] API, Evaluation of well perforators, recommended practice RP 43, American
Petroleum Institute (API), January, 1991.
[144] API, Recommended practice 19B for evaluation of well perforators, First Edition,
2001
[145] Book of ASTM standard, ASTM C 192/C 192M-02 Standard practice for making
and curing concrete test specimens in the laboratory, 2002, 04, 02. p.2.
[146] Q.M. Li and H. Meng, About the dynamic strength enhancement of concrete-like
materials in a split Hopkinson pressure bar test. International Journal of Solids and
Structures, 2003. 40, 2, p. 343-360.
[147] A. N. Mikhailov and A. N. Dremin, Flight speed of a plate propelled by products
from a sliding detonation. Combustion, Explosion, and Shock Waves, 1974. 10, 6,
p. 787-792.
[148] J. T. Harrison, Improved analytical shaped charge code: Basc, Technical Report
No. ARBRL-TR-02300, 1981.
[149] J. D. Walker, Incoherence of shaped charge jets, American Institute of Physics
Conference. Proceeding, 1994, 309, p. 1869-1872.
[150] M. Held, Diagnostic of superfast jets with 25 km/s tip velocities, Propellants,
[151] R. H. Nielsen, G. Wilfing, ULLMANNS Encyclopedia of industrial chemistry,
Zirconium and Zirconium Compounds, 2002, Wiley-VCH Verlag GmbH & Co.
KGAa, Vol. 39, p. 753.
[152] Zirconium-Machining, http://www.espimetals.com/index.php/technical-data/337-
zirconium-machining, accessed on 1/10/ 2012.
[153] M. C. Chick and L. A. Learmonth, Determination of shock initiation and
detonation characteristics of PE4 in proof test geometries, Defence science and
technology Organisation Materials research Laboratories, Report No. MRL-R-979,
1986.
[154] R. K. Whartona, S. A. Formby and R. Merrifield, Airblast TNT equivalence for a
range of commercial blasting explosives, Journal of Hazardous Materials, 2000,
79, 12, p. 31-39.
[155] W. L. Porter, B. Satterwhite, Evaluation of well perforator performance, Journal of
Petrolleum Technology, 1976. 28, 12, p. 1466-1472.
[156] A. R. Kiwan and H. Wisniewski, Theory and computations of collapse and jet
velocities of metallic shaped charge liners, Ballistic Research Labs Aberdeen
Proving Ground MD, Report No. AD0907161. 1972.
[157] F. E. Allison and G. M. Bryan, Cratering by a train of hypervelocity fragments.
Proceedings of the 2
nd
Hypervelocity Impact Effects Symposium. 1957.
[158] F. Jamet, Measurements of densities in shaped charge jets and detonation waves,
American Society for Non-Destructive Testing, 1976.
References
245
[159] K. D. Werneyer and F. J. Mostert, Analytical model predicting the penetration
behaviour of a jet with a time-varying density profile, 21
st
on Ballistics, 2004, South Africa, p. 390-397.
[160] F. M. Milton, D. W. Klaus and F. J. Mostert, An analytical penetration model for
jets with varying mass density profiles, 22
nd
International symposium on Ballistics,
2005,Canada. p. 622-629.
[161] B. Grove and I. Walton, Shaped charge jet velocity and density profiles, 23
rd

International Symposium on Ballistics, 2007, Spain. p.103-110.
[162] AccuPyc-II-1340-Pycnometer,,,http://www.micromeritics.com/Product-
Showcase/AccuPyc-II-1340.aspx, accessed on 1/10/2012.
[163] W. P. Walters and R. L. Summers, The velocity difference between particulated
shaped charge jet particles for face-centered-cubic liner materials, Army Research
Lab Aberdeen Proving Ground MD Report No. ADA257348, 1992.
[164] K. Cowan and B. Bourne. Further analytical modelling of shaped charge jet
breakup phenomena. 19
th
International Symposium on Ballistics, Switzerland,
2001, p. 803-810.
[165] H. Kolsky, Stress waves in solids, Dover Phoenix Editions, 1963.
[166] V. Ramachandran, A.T. Santhanam and R. E. Reed-Hill, Dislocation-solute
interactions and mechanical behaviour of zirconium and titanium. Indian Journal
Technology, 11, 10, p. 485-492.
[167] W. Von Holle and J. Trimble, Temperature measurement of shocked copper plates
and shaped charge jets by two color ir radiometry, Journal of Applied Physics,
1976, 47, p. 2391-2394.
[168] C. L. Aseltine, Analytical predictions of the effect of warhead asymmetries on
shaped charge jets. Technical Report No. ARBRL-TR-02214, 1980.
[169] Matweb,,,http:/www.matweb.com/search/Datasheet.aspx?MatGuid=6e8936bad994
f13bfb29923cc1506a9&ckck=1, accessed on 1/10/2012.

Appendices
246
APPENDICES
Appendix A: The numerical calculation of the critical angles
c
at different V
2

The analytical model used to calculate the critical collapse angles at different flow
velocities is obtained on the basis of the momentum balance according to:
2
Pjp
c
v
2
2
[

+1
-P

[
(p
c
v
2
2
-P)
2
(A1)
where P is the pressure,
o
is the initial liner density, is the compressibility (i.e. =/
o
-
1).
The critical angles can be determined for a given impinging velocity V
2
from the
condition [ p / = . Thus, for =
c,

dP
d
P|P-p
c
v
2
2
]
(+1)|p
c
v
2
2
-P(+2)]

(A2)
So, EOS of the liner material is used with Eq. (A2) to calculate the critical angles
c
at
different values of V
2
. Shock EOS take the form of:
pC
0
2
(+1)
(1-(-1))
2
(A3)

where C
o
is the sound speed of the liner material and S is the slope of the shock speed-
particle velocity line. For the zirconium material, S=1.018.
From Eq. (A3),
dP
d
(1-(-1))pC
0
2
((1-(-1))(2+1)+2(
2
+)(-1)
(1-(-1))
4
(A4)
For simplicity, assume S1, therefore Eq. (A4) becomes:
dP
d
o
2
(A5)
Equating Eqs. (A2), (A5), and solving for the critical compressibility (i.e.
c
). The
obtained values for
c
were

substituted into EOS, Eq. (A3) to get the corresponding critical
pressure; P
c
. The Values of
c
and P
c
are used with Eq. (A1) to get
c
.

P|P-p
c
v
2
2
]
(+1)|p
c
v
2
2
-P(+2)]

pC
2
2p + 1 . (A6)
Thus,
P
2
-(C
0
2
(p +2)(2p +1)(p +1) -I
2
2
)pP -(pC
0
2
(2p +1)(p +1)(ppI
2
2
)) = (A7)
Appendices
247
P
c

=
(C
0
2
(+2)(2+1)(+1)-v
2
2
)p _(C
0
2
(+2)(2+1)(+1)-v
2
2
) p)
2
+4(pC
0
2
(2+1)(+1)(pv
2
2
))
2
. (A8)
Let Y=+1 and Z= 2+1
c

(C
0
2
z(+1)-v
2
2
)p _(C
0
2
z(+1)-v
2
2
) )
2
p
2
+4(p
2
C
0
2
zv
2
2
)
2
. (A9)
From EOS, s=1
P
c
= pC
0
2
p =
(C
0
2
z(+1)-v
2
2
)p _(C
0
2
z(+1)-v
2
2
)
2
p
2
+4(p
2
C
0
2
zv
2
2
)
2
(A10)

0
2
(C
0
2
z(+1)-v
2
2
) _(C
0
2
z(+1)-v
2
2
)
2
+4(C
0
2
zv
2
2
)
2
(A11)

2C
0
2
p -C
0
2
Z( +1) +I
2
2
= (C
0
2
Z( +1) -I
2
2
)
2
+(C
0
2
ZpI
2
2
), (A12)

(2C
0
2
p -C
0
2
Z( +1) +I
2
2
)
2
= (C
0
2
Z( +1) -I
2
2
)
2
+(C
0
2
ZpI
2
2
), (A13)
C
0
4
p
2
2
+C
0
2
p(-C
0
2
Z( +1) +I
2
2
) = (C
0
2
ZpI
2
2
) , (A14)
C
0
2
p

-C
0
2
Z( +1) +I
2
2

= ZI
2
2
. (A15)
Substitute the values of Y and Z,
p

-Z( +1) +
v
2
2
C
0
2

= Z
v
2
2
C
0
2
, (A16)
p

(p +1)

-(p +1)(2p +1)(p +2) -(2p +1)
v
2
2
C
0
2

= -
v
2
2
C
0
2
, (A17)
2p
3
+6p
2
+_7 +2 [
v
2
C
0
2
] p +1 =

, (A18)

For the Zirconium material, =6510 g/cm
3
and K=98.5G Pa.
Calculating the critical pressure P
c
and the critical compressibility
c
, the critical collapse
angle is estimated from Eq. (A1) at different flow velocities V
2
.
Appendices
248
Appendix B: Proof of Inequality Eq.(8-6)

According to Eq.(8-5),
dP()
d
= I
(t) +t
dv
]
()
d
. (B1)
For uniform density distribution, Eq.(B1) together with Eqs.(8-9) and (8-10) lead to
dP()
d
= v
o
[
t
o
t

+1

+t
d
d
_v
o
[
t
o
t

+1
_ = v
o
[
t
o
t

+1

+t [
-

+1
v
o
(t
o
)

+1
(t)
-

+1
-1
. (B2)
Therefore, the inequality Eq.(8-6) is equivalent to v
o
[
t
o
t

+1

[

+1
v
o
[
t
o
t

+1
or 1

[

+1
, which is automatically satisfied.
For non-uniform density distribution, the velocity is determined by Eq.(8-16), which can
be rewritten as
_
v
0
v
]
]
b
0
_b _
v
0
v
]
] +o] = (o +b) [
b
0
. (B3)
For the shaped charge liners studied in this paper, a0.1, b1 and I
o
> I
, thus b _
v
0
v
]
] +
o b _
v
0
v
]
] and o +b b with a maximum error less than 10%. Therefore, Eq.(B3) can
be approximated by:
_
v
0
v
]
]
b
0
+1
[
b
0
. (B4)
The inequality Eq.(8-6) can be proved easily by Eqs.(8-5), (8-14) and (B4) using the same
procedure for uniform density distribution situation, which will not be presented here.
Meanwhile, we also calculated the time histories of I
(t) and
dv
]
()
d
numerically. Based on
Eq.(B1), it has been shown that inequality Eq.(8-6) can be satisfied for the studied cases in
Chapter 8.

Appendices
249
Appendix C: Derivation of Eq. (8-18)
Assuming the hydrodynamic theory and starting from Eq. (8-8) for non-uniform density
jet,
-dv
v
j
()

v()
v
c
1
c
dv
v
j
()_
a
V
]
o
v
j
(t)+b_
v()
v
c
d
c
(C1)
v
]
c
v()
1
c
dv
v
j
()_
a
V
]
o
v
j
(t)+b_
v()
v
c

c
(C2)
o
V
]
o
v
j

by
c
+1
by
c
_
a
V
]
o
v
j
+b_
_
a
V
]
o
V
]
o
+b_
1
by
c

(C3)
The total penetration will be achieved within time duration of t and V
c
; the cutoff element
velocity or the velocity of the last penetrating element. Thus, Eq. (C3) will be.
o
V
]
o
v
c
by
c
+1
by
c
Cuc]]
i
1
by
c

(C4)
where
cutoff
is the square root of the ratio of the target to the jet densities at the cutoff
element, while
tip
is the square root of the ratio of the target to the jet densities at the tip
of the jet. When
cutoff
=
tip=

o,
and

b=1, this

equation reduces to the constant density
equation. The second term can be defined as the density reduction term.
From Eq.(C4), the impact velocity of the jet is determined by an algebraic equation of
b _
v
0
v
]
]
b
0
+1
+o _
v
0
v
]
]
b
0
= (o +b) [
b
0
.
(C5)
Eq.(C5) reduces to Eq.(8-9) for constant jet density when a=0 and b=1.
This equation also reduces to the constant density equation when
cutoff
=
tip
=constant and
b=1.
To calculate the penetration of jet with non-uniform density distribution, the derived
definitions of the total penetration time and the jet velocity at this time are used in the
penetration equation (i.e. Eq. (8-5)).

o
V
]
o
V
]c
1
by
c

Cuc]]
i
1
by
c

(C6)
Appendices
250
Appendix D: The detailed drawing of a conical liner OWP Preliminary Design

D-1: OFHC Copper Liner

Units in mm

Appendices
251
D-2: Steel casing

Units in mm

Appendices
252
D-3: PE4 Explosive charge

Units in mm.
End of Thesis
253

(This page is intentionally left blank)

Criteria of Design Improvement of Shaped Charges

Uploaded by

Copyright:

Available Formats

Criteria of Design Improvement of Shaped Charges

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Criteria of Design Improvement of Shaped Charges

Uploaded by

Copyright:

Available Formats

What is the main topic of research discussed in the document?

What factors are discussed that can affect shaped charges used as oil well perforators?

CRITERIA OF DESIGN IMPROVEMENT OF

cos (x) where U

sin([ -o) -sin([ -o -)

and Eq.(6-2) can be reduced to the hydrodynamic equation, which is

is the velocity of a continuous jet; u is the velocity of the jet-target interface or

/ (t) are introduced. Thus,

You might also like