EVALUATION OF LIGHT CERAMICS FOR BALLISTIC
PROTECTION
EVALUATION DES CERAMIQUES LEGERES POUR PROTECTION
BALISTIQUE
A Thesis Submitted
To the Faculty of the Royal Military College of Canada
by
Cezar Constantin Craciun
In Partial Fulfillment of the Requirements for the Degree of
Master of Applied Science in Mechanical Engineering
May, 2009
© This thesis may be used within the Department of National Defence,
but copyright for open publication remains the property of the author Library and Archives Bibliotheque et 1*1 Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition 395 Wellington Street 395, rue Wellington OttawaONK1A0N4 OttawaONK1A0N4 Canada Canada
Your file Votre reference ISBN: 978-0-494-66042-3 Our file Notre reference ISBN: 978-0-494-66042-3
NOTICE: AVIS:
The author has granted a non L'auteur a accorde une licence non exclusive exclusive license allowing Library and permettant a la Bibliotheque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par I'lnternet, preter, telecommunication or on the Internet, distribuer et vendre des theses partout dans le loan, distribute and sell theses monde, a des fins commerciales ou autres, sur worldwide, for commercial or non support microforme, papier, electronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats.
The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in this et des droits moraux qui protege cette these. Ni thesis. Neither the thesis nor la these ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent etre imprimes ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author's permission.
In compliance with the Canadian Conformement a la loi canadienne sur la Privacy Act some supporting forms protection de la vie privee, quelques may have been removed from this formulaires secondaires ont ete enleves de thesis. cette these.
While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n'y aura aucun contenu removal does not represent any loss manquant. of content from the thesis.
••I Canada Ill
ACKNOWLEDGEMENTS
I would like to thank my thesis supervisor, Dr. I. E. Boros for his support along each step of the project and his confidence in my work and abilities to reach the final destination of this extended journey.
I would also like to recognize the contribution of Dr. Petru Lucuta and his entire
family, who made possible the production of most of the materials tested in the present
research.
Specific thanks must go to Guy Bergeron from Defence Research and
Development Canada - Valcartier as well as to Dominique Proulx and Pierre Vieillette
from Munitions Experimental Test Centre in Valcartier for their invaluable contribution
to the experimental testing program.
Many thanks are dedicated to Mechanical Engineering staff, which made my
work at RMC both fruitful and pleasant.
And last, but not least, my thoughts of gratitude go to the three emotional pillars
of my life; my mother Georgeta, my wife Floriana and my daughter Natalie-Maria. IV
ABSTRACT
Craciun Cezar C. M.A.Sc. (Mech. Eng.). Royal Military College of Canada. May 2009. Evaluation of Light Ceramics for Ballistic Protection Supervisor: Dr. I. E. Boros.
Newly developed ceramic materials have become the material of choice in many modern applications. About half a century ago, ceramics started to be included in armour systems that provide ballistic protection to personnel and equipment on the modern battlefield.
Due to the continued development of various threats, particularly small arms
ammunition with increased penetrating capabilities, the modern ceramic based armour
must improve its stopping characteristics in order to remain efficient.
Presently, the interaction between the projectile and the ceramic is not completely
understood; therefore no theory can fully describe or simulate all the complex processes
taking place at the moment of impact.
Research aimed at improving the quality of armour grade ceramic relies mostly on
experimental study, with depth of penetration (DOP) being one of the most common
testing methods used to characterize the effectiveness of such ceramics.
In the present study, three modern ceramics, namely silicon carbide, boron
carbide and CERAMOR® with various areal densities and with or without front covering
materials have been compared with the one already in use on the CC-130 Hercules V aircraft. The comparison was based on the length of the penetration of an armour piercing projectile in a polycarbonate block, after impacting a ceramic tile.
All the three modern ceramic materials outperformed the alumina, at lower areal densities compared with the alumina.
The results confirmed the need to replace the obsolete ceramic currently in use on the CC-130 Hercules aircraft. Based on this study, further ballistic testing of complete armour systems would be required to finalize the selection process. VI
RESUME
Craciun Cezar C. M.A.Sc. (Genie Mecanique). College Militaire Royal du Canada. May 2009. Evaluation des Ceramiques Legeres pour Protection Balistique Superviseur: Dr. I. E. Boros.
Les materiaux a base de ceramiques recemment developpes constituent des materiaux de choix pour des applications modernes. Depuis presque un demi-siecle, on a commence a inclure des ceramiques dans des systemes de blindage offrant une protection balistique pour les militaires et les equipements sur les champs de bataille.
Parce que les menaces ont evolues, surtout en ce qui concerne les projectiles perforants de petits calibres, il s'avere necessaire d'augmenter le pouvoir d'arret de projectiles des blindages contenant ces ceramiques afin de pouvoir continuer a les utiliser.
Le processus d'interaction entre le projectile et la ceramique n'est pas entierement compris. C'est pourquoi aucune theorie ne peut decrire ou simuler exactement les evenements complexes qui ont lieu au moment de l'impact.
La recherche visant 1'amelioration de la qualite des ceramiques utilisees dans des blindages est, consequemment, strictement experimentale, la methode de profondeur de penetration (PDP) etant le plus souvent utilisee pour faire la caracterisation de telles ceramiques. vii
Dans la presente etude, la performance de trois ceramiques modemes (carbure de silice, carbure de bor et CERAMOR®) avec des densites sur surface differentes a ete comparee a celle de la ceramique couramment utilisee dans les avions CC-130 Hercules.
La comparaison a ete basee sur la distance parcourue par un projectile perforant dans du polycarbonate, apres avoir impacte une piece de ceramique.
Tous les ceramiques modernes ont eu une meilleure performance, meme a une densite sur surface plus bas que celle d'alumina.
Les resultats ont confirme la necessite de remplacer la ceramique devenue obsolete encore utilisee a bord des avions CC-130 Hercules. II s'averera cependant necessaire d'effectuer des essais balistiques plus pousses sur des systemes de blindage complets, afin de confirmer la selection des ceramiques sur la base des donnees obtenues
au cours de la presente etude. viii
TABLE OF CONTENTS
ABSTRACT iv RESUME vi TABLE OF CONTENTS viii LIST OF FIGURES x LIST OF TABLES xii LIST OF ACRONYMS (ABBREVIATIONS) xiii LIST OF SYMBOLS xv CHAPTER 1: INTRODUCTION 1 1.1 HISTORICAL OVERVIEW 1 1.2 OBJECTIVE 4 1.3 SCOPE 5 1.4 THESIS OUTLINE 6 CHAPTER 2: LITERATURE REVIEW 7 2.1 SMALL PROJECTILES 8 2.2 ARMOUR MATERIALS 13 2.2.1 Metals 14 2.2.2 Continuous Fibre Reinforced Composites (CFRC) 15 2.2.3 Ceramics 15 2.3 CERAMIC MANUFACTURING 22 2.3.1 Ceramic Forming 22 2.3.2 Ceramic Hardening 23 2.3.3 Ceramic Hot Pressing 23 2.3.4 Reaction Bonded Ceramic 24 2.4 ARMOUR USE OF CERAMICS 25 2.5 COMBINED ARMOUR 29 CHAPTER 3: THEORETICAL CONSIDERATIONS ON BALLISTIC IMPACT 32 3.1 ELASTIC IMPACT 32 3.2 IMPACT DAMAGE 41 3.3 ARMOUR CHARACTERISTICS 49 3.3.1 Impact on Metals 52 3.3.2 Impact on Continuous Fibre Reinforced Composites 54 3.3.3 Impact on Ceramics 55 3.3.4 Armour Ceramic Improvements 77 3.4 TYPES OF BALLISTIC TESTS 89 3.5 DOP TEST 94 IX
CHAPTER 4: EXPERIMENTAL TESTING 100 4.1 PRELIMINARY WORK- DROP TOWER IMPACT APPARATUS.... 100 4.2 DOP EXPERIMENTAL PROGRAM 104 4.2.1 Protection Level /Projectile Selection 105 4.2.2 Experimental Set up Apparatus 108 4.2.2.1 Backing 108 4.2.2.2 Clamping System-the lig 110 4.2.2.3 Ceramic - Polycarbonate Adhesive 112 4.2.3 Ceramic Samples 114 4.2.3.1 Tiles Minimum Dimensions 116 4.2.3.2 Alumina Tiles 117 4.2.3.3 Boron Carbide Tiles 119 4.2.3.4 CERAMOR® Tiles 119 4.2.3.5 Silicon Carbide Tiles 121 4.2.4 The Firing Range 122 4.2.5 Targets Tested. 125 4.2.6 DOP Measuring Method 132 CHAPTER 5: RESULTS 133 CHAPTER 6: INTERPRETATION OF RESULTS 143 6.1 FLAT VERSUS MAP FOR POROUS CERAMOR® 143 6.2 POROUS VERSUS DENSE FOR CERAMOR-MAP® 144 6.3 SPALL COVER INFLUENCE ON CERAMOR-MAP® 145 6.4 FLAT VERSUS MAP FOR SILICON CARBIDE 146 6.5 THICKNESS INFLUENCE ON BORON CARBIDE 146 6.6 MATERIAL INFLUENCE FOR HIGH AREAL DENSITY TILES 147 6.7 MATERIAL INFLUENCE FOR LOW AREAL DENSITY TILES 148 CHAPTER 7: DISCUSSION 150 7.1 LIMITATIONS 150 7.2 TEST VALIDITY 151 7.3 SPALL COVER EFFECT 152 7.4 INFLUENCE OF LATERAL CONFINEMENT 153 7.5 SHATTERING PATTERN FOR A1203, SiC AND B4C 154 7.6 DOP AND EFFECT ON PROJECTILE FOR A1203, SiC AND B4C 160 7.7 CERAMOR® POROSITY AND PERFORMANCE 162 7.8 RESULTS INCONSISTENCIES 166 7.9 CALCULATIONS BASED ON DOP RESULTS 170 7.10 ADDITIONAL TESTS 171 7.11 CC-130 HERCULES ALUMINA RESULTS 172 7.12 CONCLUDING REMARKS 173 CHAPTER 8: CONCLUSIONS 175 CHAPTER 9: RECOMMENDATIONS 178 REFERENCES 182 VITA 189 APPENDIX: Table containing DOP test results and efficiency calculated values 190 X
LIST OF FIGURES
Figure 1 - The effective armour thickness 2 Figure 2 - Section through the frontal armour of T-80 and T-64B Soviet MBT 3 Figure 3: Examples of projectiles1 10 Figure 4: Reference drawing for FSP from STANAG 45692 10 Figure 5: Example of a phase diagram for projectile impact1 13 Figure 6: A Qualitative Representation of a Ceramic Stress-Strain Curve6 17 Figure 7: Elements present in ceramic materials , 18 Figure 8: Carbide compositions and their melting temperatures7 19 Figure 9: Nitride compositions and their melting temperatures7 19 Figure 10: Oxide compositions and their melting temperatures7 20 Figure 11: Example of a modern combined armour system 30 Figure 12: Waves generated at the impact on solids19 36 Figure 13: Plastic wave attenuation during wave propagation in solids 39 Figure 14: Classic illustration of the modes of applied stress for crack propagation 42 Figure 15: Projectile armour interaction19 51 Figure 16: Evolution of damage produced by a spherical particle impact54 66 Figure 17: Sectional view of surface crack pattern in Si3N4 impacted by 2.4 mm WC sphere at 231 m/s55 67 Figure 18: Cracks generated in a confined ceramic target 68 Figure 19: A generic graph exhibiting the shatter-gap phenomenon2 76 Figure 20: Edge cracking of a ceramic tile due to wave reflections66 81 Figure 21: Configurations used in edge optimization evaluation by Bryn James45 83 Figure 22: Impactor manufactured in house 102 Figure 23: Frame for drop impact testing 103 Figure 24: 7.62 x 51 Bofors - Carl Gustaf WC core, AP round 107 Figure 25: The 5" polycarbonate billet being machined in the lathe 110 Figure 26: Clamping system for the DOP targets Ill Figure 27: The removal of excess adhesive 114 Figure 28: Alumina ceramic tiles glued for DOP testing 118 Figure 29: ACERAM's high temperature electric furnace for ceramic sintering 120 Figure 30: DOP target setup 124 Figure 31: Experimental arrangement used for ballistic testing 125 Figure 32: DOP versus Areal Density 137 Figure 33: DOP versus areal density for heavy, inefficient samples 139 Figure 34: DOP versus areal density for heavy, efficient samples 140 Figure 35: DOP versus areal density for light, efficient samples 141 Figure 36: Uncovered alumina targets, after impact 154 XI
Figure 37: Kevlar® covered alumina target after impact 155 Figure 38: Boron carbide targets; the thick samples (CEK) shown in top row, the thin . samples (CEN) in bottom row 156 Figure 39: Uncovered SiC MAP tile completely shattered after impact (MB 33) 157 Figure 40: SiC MAP tile sandwiched within PC films and diamond coated, after the impact (MBDP 30) 158 Figure 41: Fragmentation of a SiC tile confined within PC films and radial cracks in the PC cylindrical backing (MBDP 28) 158 Figure 42: Comminuted (pulverized) ceramic and the AP core contained in cracked PC backing cylinder for flat SiC tiles sandwiched between PC films (MFDP 20) 159 Figure 43: Radial and lateral cracks in PC -thick B4C target (CEK 2) 160 Figure 44: Radial and lateral cracks in PC - thin B4C target (CEN 11) 161 Figure 45: Flat CERAMOR® low density samples after impact. The projectiles cores were left intact 163 Figure 46: WC cores penetrated in the PC backing after impacting: E-CERAMOR®-MAP sandwiched between PC films, low density tiles. The cores have curved trajectories. 164 Figure 47: WC fragmented cores penetrated the PC cylinders after impacting the CERAMOR®-MAP low density tiles 164 Figure 48: Low density CERAMOR®-MAP tile fissured, before testing 167 Figure 49: Low density CERAMOR®-MAP tile (B-2) after the impact (V = 935.3 m/s, DOPof74mm) 168 Figure 50: Low density CERAMOR®-MAP tile (B-4) after the impact (V = 935.4 m/s, DOPof92mm) 169 Figure 51: CF-130 Hercules armour panel perforated by two 7.62 AP WC core projectiles at muzzle velocity (930 m/s), exit face 173 xii
LIST OF TABLES
Table 1: Material properties for various ceramics compared to RHA 26 Table 2: Property Comparison of Typical Armour Ceramics. ' 27 Table 3: 2001 Ceramic Armour Producers16: 29 Table 4: Wave velocities in isotropic solids1 37 Table 5: Ceramic Material Evaluation Summary of Ballistic Test Methods41: 93 Table 6: The STANAG 4569 accepted ammunition for level III threat2 105 Table 7: Mechanical properties of CERAMOR® ceramics compared to boron carbide69. 115 Table 8: Tests performed in the experimental program 126 Table 9: Tiles tested in the first set of trials 128 Table 10: Tiles tested in the second set of trials 130 Table 11: Average results for the ceramic samples tested 134 Table 12: Flat versus MAP results for porous CERAMOR® 143 Table 13: The influence of the porosity for CERAMOR-MAP® 144 Table 14: Spall cover influence on CERAMOR-MAP® 145 Table 15: MAP and spall cover influence on silicon carbide results 146 Table 16: Thickness influence on boron carbide results 147 Table 17: High areal density tiles, material's influence 147 Table 18: Material influence for tiles with low areal density 148 Table 19: Appendix - DOP test results and efficiency calculated values 191 xm
LIST OF ACRONYMS (ABBREVIATIONS)
3D Three Dimensional
AP Armour Piercing
APC Armoured Personnel Carrier
APFSDS Armour Piercing Fin Stabilised Discarding Sabot
API Armour Piercing Incendiary
ASTM American Society for Testing and Materials
BAD Behind Armour Debris
Cermet Ceramic - Metallic Composition
CFRC Continuous Fibre Reinforced Composite
DOP (Modified) Depth of Penetration or Depth of Penetration of a Projectile into Backing Material after Perforating the Ceramic
DPT Dwell / Penetration Transition
DRDC Defence Research and Development Canada
DREV Defence Research Establishment Valcartier
DU Depleted Uranium
DWE Dwell Tests
FSP Fragment Simulating Projectile
FTG Fixed Target Geometry
FEE High Explosive XIV
HEL Hugoniot Elastic Limit
LAV Light Armoured Vehicle
LOP Length of Penetration
MAP Modular Advanced Protection
MBT Main Battle Tank
METC Munitions Experimental Test Centre
NATO North Atlantic Treaty Organization
NDP Non-Deforming Penetration
PEN Penetration Depth Direct or Reverse Impact
RHA Rolled Homogenous Armour
RMC The Royal Military College of Canada
SLAP Saboted Light Armour Piercing
STANAG NATO Standardisation Agreement
TAD Target Areal Density Performance Maps
TCA Tandem Composite Armour
V50 Ballistic Limit Velocity (50% of the projectiles at that velocity did not penetrate the armour) or VBL
VBL Ballistic Limit Velocity or V50 or Residual Data
WWH The Second World War LIST OF SYMBOLS
Area
Length of a Surface Crack or Half Length of an Internal Crack
Alumina or Aluminium Oxide
Brittleness Factor
Boron Carbide
Carbon
Sonic Velocity or Elastic Wave Velocity
Velocity of Bulk Elastic Waves
Longitudinal Pulse Velocity
Elastic Compressive Longitudinal Wave Velocity
Plastic Wave Velocity
Cobalt
Lransversal Pulse Velocity
Computer Lomography
Ballistic Energy Dissipation Characteristic
Target Width
Projectile Diameter
Length unit xvi dt Time unit
E Elastic Modulus or Young Modulus
Ec Kinetic Energy
Eeq Equivalent Thickness or Thickness Efficiency
Em Mass Efficiency
F Force
Fe Iron
FL Longitudinal Force
G Shear Modulus
H Knoop Hardness
Hb Backup Layer Thickness
Hc Ceramic Thickness between Projectile and Back Surface
Hy Vickers Hardness
ID Inside Diameter
K Stress Intensity Factor at the Tip of a Crack
Kt Stress Concentration Factor at the Tip of a Crack
Kc Fracture Toughness
Kic Fracture Toughness for Mode I of Fracture
m Mass
m Bonding Quality Factor
ME Relative Mass Efficiency
Meq Equivalent Mass or Mass Efficiency
OD Outside Diameter XV11
Pbac Penetration Depth into Unprotected Backing Material
PC Polycarbonate
Pres Residual Penetration of Projectile in Backing behind Ceramic q2 Quality Factor or Ballistic Efficiency Factor
S Structural Factor
SisN4 Silicon Nitride
SiC Silicon Carbide
Tcer, tCer Ceramic Target Thickness t^t Critical Thickness
U Wave Velocity in Solids
Up Particle Velocity v Velocity
vO Impact Velocity
VL Particle Longitudinal Velocity
Vth Threshold Velocity or Critical Velocity or Ballistic Limit
W Tungsten
WC Tungsten Carbide
Y Dimensionless Parameter or Function Associated with Cracks
ZrC>2 Zirconia or Zirconium Dioxide
zp Acoustic Impedance of a Projectile
zt Acoustic Impedance of a Target
Avs Change in the Particle Velocity in Shear
s Strain (longitudinal or transversal) ys Specific Surface Energy yp Plastic Deformation Energy Associated with Crack Extension r\ Ballistic Efficiency
X,\i. Lame Parameters p Density
Pbac Density of Backing Material pCer Density of Ceramic pt Curvature Radius of a Crack Tip a Stress, Uniaxial
Go Magnitude of the Nominal Applied Tensile Stress a'oc Effective Strength of Fragmented Ceramic
Goc In-plane Strength of Ceramic Layer
cob In-plane Strength of Backup Layer
CTC Critical Stress
CTHEL Maximum Stress for Elastic Wave Propagation in Solids
cim Maximum Stress at a Tip of a Crack
u Poisson Ratio
T Shear Stress 1
CHAPTER 1: INTRODUCTION
1.1 HISTORICAL OVERVIEW
The human history is one of wars. The fastest and most remarkable technical achievements had been accomplished for military purposes. And in this bellicose world the most basic race has been fought between the projectile and the shield. In the beginning, the spear, the sword and the arrow were used to penetrate the personal protection or the shelter of the enemy with the purpose of killing or injuring him.
The Middle Age saw improvements in fortification, as well as in the protection of ships and soldiers. The projectiles, on the other hand, became faster and more powerful, particularly after the formula for gunpowder manufacturing had been brought to Europe.
For centuries, just as the thickness of the protection cover, metal or stone, increased, so did the kinetic energy of the projectile, to augment its penetration.
It was not until WWII that an excellent idea was conceived: to use inclined armours, first appearing on Soviet tanks. The concept of defeating a projectile was not based solely on stopping it anymore, but deviating it by ricochet. For the impacting projectile it is imperative to be as close to the perpendicular as possible on the target in order to expend most of its kinetic energy in the perforation process. A lower angle of 2 impact will lead to a longer penetration through the armour or even a ricochet, and eventually to more protection from the same armour thickness (as seen in Figure 1).
Figure 1 - The effective armour thickness.
The latest major enhancement of the passive armour systems came on the market
after WWII, in the form of sandwich type armour, including not only metal, but also
advanced composite materials and ceramics. One example of the advanced armour
system currently used is illustrated in Figure 2, from various open sources. It features
hard steel on the impact surface, followed by a more ductile steel layer. A glass fibre
layer and more steel increase the protection, probably against shape charges. The lead
liner might protect against harmful radiation. t"-"t Line (lead imprenated plastic foam uneven thickness) LJ—U Mine plough attachment points
Figure 2 - Section through the frontal armour of T-80 and T-64B Soviet MBT.
The development of the 3D battlefield raised new protection issues, this time related to aircrafts / helicopters and their crew against more and more vicious ammunition.
On the ground or on water, an increment in the mass of armour has a lower effect on the vehicle or boat's fighting and movement capabilities compared to the air, where the effect is more drastic.
Therefore, the research and development of lighter armours with better protection capabilities is an ongoing activity, fuelled by the demands and menaces of the modern battlefield. 4
1.2 OBJECTIVE
The objective of this work is to identify and recommend improvements for the ceramic based ballistic protection panels already in use with the Canadian Air Force against small arms ammunition.
The study benchmarks the current protection solution against the small arms
ammunition and researches the possibilities for improvement through the replacement of
the ceramic materials used and / or changes in the configuration of the ceramic in the
protective systems.
The benefits of these potential changes are likely an armour with the same
stopping capabilities against the same threat with a lower areal mass or a better ballistic
behaviour with no mass penalty. 5
1.3 SCOPE
The aim of this research project is to examine ways for improving the armour panels used by the Canadian Air Force on its CC-130 Hercules transport aircraft. The panels are used as ballistic protection against small arms ammunition, a threat encountered during take-off and landing in operation zones. Since the main component of these panels is represented by the ceramic, the present research focuses on the improvement of this material. As a consequence, the effect of any upgrading of the ceramic has a greater effect on the entire armour system, than any other component of the panel.
This is to be accomplished by comparing the ceramic in use with the ceramics
currently available on the ballistic protection market, through a representative test. 6
1.4 THESIS OUTLINE
The first chapter of this study contains the introductory information pertaining to the present research.
In its second chapter, the study reviews the small arm projectiles and the armour
materials developed to protect against them, as presented in the studied literature.
The third chapter presents some theoretical considerations on ballistic impact,
focused mainly on ceramics, and a few possible ways of improvement, derived from the
studied literature as well.
The forth chapter includes the experimental research performed by the author. It
contains an earlier research performed at RMC focused on manufacturing an impact
apparatus for testing various elements of an armour panel. This research was substituted
with ballistic testing of ceramic components in a representative DRDC facility, work
presented also in the forth chapter of the present study.
The results obtained during the ballistic testing, their interpretation and discussion
are presented in chapter five, six and seven, respectively; the conclusions are outlined in
chapter eight.
Suggestions for improvements, outlined in chapter nine, based on the literature
survey and testing performed conclude the present work. 7
CHAPTER 2: LITERATURE REVIEW
Many materials are used for ballistic protection today. Although different types of materials used in applications involving protection against projectiles with high kinetic energy behave in dissimilar ways under the same impact conditions, the particularities of the target to be protected establishes, to a certain extent, the type of materials to be used.
For instance, in personal protection, the flexibility of the garment is an important feature; therefore the use of fabrics makes the most sense.
The mass, velocity and material of the expected projectile also have a great influence on the choice of protective material. For example, the impact velocity of projectiles could lead to a totally different response of the target material, such as in the case of a steel plate impacted at 800 m/s compared with the same steel plate impacted at
1600 m/s by a projectile with the same kinetic energy. This might happen due to the mechanical properties of the metal, which could be affected by penetration velocity.
In light of the research topic discussed here, the current section first tackles the types of small arms projectiles that can be encountered by the armour used on the battlefield, after which it outlines the materials that are commonly used in ballistic protection against small arms ammunition. 8
2.1 SMALL PROJECTILES
It is generally accepted that the main danger on the battlefield is created by kinetic energy threats impacting the targets. The obvious example of such a threat is the projectile shot from a weapon. Although one can argue that an explosion can produce more damage, the blast shock is effective only in the vicinity of the explosion, but the fragments generated by either the charge shell, the target or the environment (i.e. rocks near the explosion) could do damage over a much wider area.
The projectiles with regular shape are typically fired from weapons directly against a target. There are many types of projectiles, some being designed to defeat a particular type of target. The most relevant projectiles used against targets protected by
light armour are shot by small arms, up to a 20 mm calibre.
The most representative small arms projectiles are the following:
Regular - The regular projectile has a heavy core made of lead or mild steel,
covered by a jacket made of copper or copper alloy. Its kinetic energy at impact does the
damage to the target.
High Explosive (HE) - The high explosive projectile has a metallic shell filled
with an explosive composition. In the vicinity of the target, upon impact or after
penetrating the target, the explosive charge is initiated, generating a shock wave and
fragments that affect the target. This type of projectile has an extremely limited use in
small calibre weapons.
Armour Piercing (AP) - The armour piercing projectile is similar in construction
to the regular projectile, except that its core is made of hard steel or other heavy metals 9 such as tungsten or depleted uranium (DU), capable of penetrating thicker armour than a regular projectile.
Armour Piercing Incendiary (API) - The armour piercing incendiary projectile has enhanced penetration capabilities similar to the AP projectile, combined with the ability of igniting flammable materials behind the armour, due to a pyrotechnic composition contained in the projectile.
Armour Piercing (Fin Stabilised) Discarding Sabot (APFSDS) or Saboted
Light Armour Piercing (SLAP) - It is a projectile with a diameter smaller than the
calibre of the barrel. It consists of a long narrow rod pointed at the front. During flight, it
is stabilised by fins located in the rear. It is propelled through the barrel with the aid of
light sabots that are discarded after the projectile leaves the muzzle. The high velocity of
the projectile combined with its narrow cross section gives it excellent perforation
capabilities. It is, however not usually encountered in the calibres discussed here.
Another major threat to an unprotected or lightly protected target is represented
by metallic fragments, generated by exploding munitions. These fragments are irregular
in shape and flying patterns and their mass follows a Gaussian distribution, depending on
the original munitions that generated them. Metallic fragments with regular shapes,
ejected from munitions are another threat. Images of the projectiles described above are
presented in Figure 3'.
In order to simulate the effect of these irregular fragments against targets, a
fragment simulating projectile (FSP) was designed and standardised for the international
research community2. The projectile is fired from a regular small arm at different
velocities in order to obtain certain impact energies on the target. This FSP is usually 10 cylindrical with its length comparable to its diameter. The front is chisel shape and its rear might have a rim for engaging in the grooves of the barrel. A reference drawing of the FSP is given in Figure 4.
A8M0R PIERCING .50-CALIBER CAPPED WITH FIN STABILIZED BURSTING CHARGE DISCARDABLE SA80T
ARMOR P1ERCIHG CAP / CHROME MOLYBDENUM SABOT PENET8A10R BODY fir STEEL CORE
K£9
j*"^-—rnTtunrxonzj..-^ i m IL~—___— —M 2) FRAGMENT SIMULATORS REGULAR SOLIDS IRREGULAR FRAGMENTS (FSP)
Figure 3: Examples of projectiles1.
SHARP EDGES
=w —i m VALUE TZSH-
1—0
Figure 4: Reference drawing for FSP from STANAG 45692 11
Znkas provides an excellent overview of the situations encountered in regards to projectile characteristics such as geometry, material density, their flight and impact behaviour as well as their possible final state, as follows:
Geometry: Basic Shape: Solid Rod Sphere Hollow Shell Irregular Solid Nose Configuration: Cone Ogive Hemisphere Right Circular Cylinder
Material Density: Lightweight: Wood, Plastics Ceramics, Aluminium Intermediate: Steel, Copper Heavy: Lead, Tungsten
Flight Characteristics: Trajectory: Straight (stable) Curved (stable) Tumbling (unstable) Impact Conditions: Normal Oblique
Final Condition: Shape: Undeformed Plastically Deformed Fractured Shattered 12
Location: Rebound- Partial Penetration Perforation
A typical phase diagram for projectile impact is provided in the same bibliographic material (Zukas1), as seen in Figure 5. In this instance, the diagram portrays the behaviour of a 6.35 mm ogival-nosed projectile striking a 6.35 mm thick aluminium alloy target. Every projectile - target pair diagram has unique numerical values but is usually comparable in shape.
As one can see, depending on the impact velocity and the impact obliquity, the interaction between a given projectile/target pair can lead to different results. The projectile can ricochet, embed or perforate the given target, in various states, namely intact, broken or shattered. 1400 PERFORATES SHATTERED
120O
1000
£
9. 800
^600 I- PERFORATES UMfTACT
(BALLISTIC LIMIT CURVE) 400
RICOCHETS INTACT
20O EMBEDS INTACT
{RICOCHET CURVE)
JL i I L. O 20 40 60 80 OBLIQUITY {degrees)
Figure 5: Example of a phase diagram for projectile impact1,
2.2 ARMOUR MATERIALS
Properly designed armours can protect against a large range of ballistic threats but the armour becomes heavier and bulkier for increased threat levels. As a consequence, the threat assessment to define an appropriate level of protection is very important, as stated by Iremonger and Gotts3. 14
Armour is the last line of defence of a target. Over time, this line has been transformed and improved in order to keep up with the developments in ammunitions designed to destroy such targets. The science behind the design of armour has to take into consideration a multitude of factors starting with the threat that has to be defeated and ending with the comfort of the personnel protected by that armour.
Anderson considers that nowadays, the armour is a synergy of mechanisms and materials, the latest being used to amplify the performance of the mechanics. He outlines the importance of weight in this application, and the demands to use less material and push the components to their limit, namely failure.
Thus, the process of creating new armour involves a fair amount of theoretical research and experimental data to back up the proposed design.
Since the material properties are utilized in the region of material failure, i.e., if the armour doesn't fail for a given ballistic threat, it could be made slightly lighter4, a certain dose of statistical uncertainty will always be present.
The subsequent pages will review the materials used most in armour design, with
emphasis on ceramics, since they are the materials tested and evaluated in the present
research.
2.2.1 Metals
Metals represent the oldest choice for effective armour, particularly for mobile
targets, capable of absorbing the kinetic energy that would otherwise kill or injure the 15 defended personnel. Although in the early days copper and brass alloys were the metals of choice, projectile evolution demanded a harder metal. Today steel and steel alloys are encountered as metal armour materials.
2.2.2 Continuous Fibre Reinforced Composites (CFRC)
Modem composite materials with continuous (or long) fibres as the main element
are increasingly used in applications where their behaviour can be tailored to obtain the
maximum output with minimum use of material. When flexibility is required, only fibres
are used - mostly woven. When flexibility is undesirable, hard composites containing a
matrix that embeds the fibres to provide the needed stiffness are used. The woven
composite armour found one of its first applications in personal protection systems, such
as bullet-proof vests called "flak-jacket" in the early days. Helmets and add-on stiff
inserts for added protection are manufactured using the same continuous fibres reinforced
composites (CFRC) in epoxy matrices.
2.2.3 Ceramics
In his book "A concise introduction to ceramics", Phillips5 explains the origin of
the term ceramics as being derived from the Greek word "keramos" which means "burn
stuff'. He emphasises the fact that today ceramics are more popular than ever before due 16 to the fact that they are often harder, lighter, stiffer and more resistant when compared to metals. Their brittleness, on the other hand is the penalty to be paid.
At room temperature both crystalline and amorphous ceramics almost always fracture before any visible plastic deformation can occur following an applied tensile stress. To make the matter worse, there is a considerable variation and scatter in fracture strength of these brittle materials, depending on their porosity and the existence of flaws capable of initiating a crack.
Under compressive stress, ceramics display much higher strengths than in tension.
Although still brittle, some authors suggest that some plasticity might be present.
Matsalla is one of the few authors who believe that ceramic materials do permit
some plastic flow before rupture, particularly when under compressive stress. He
illustrates his belief with the following graph (Figure 6), in which the area under the
stress-strain plastic curve is representative of the energy that can be absorbed by the
material. He characterises the plastic energy, as being absorbed permanently in the form
of destructive work, a feature desired in ballistic applications of ceramic materials, in
order to defeat an impacting projectile. 17
Tensile Stress, a
Tensile Strain, e Compressive Strain, -e
Compressive Yield Strength, -Y Compressive Stress, -cr
Figure 6: A Qualitative Representation of a Ceramic Stress-Strain Curve6
Due to the relatively low cost and abundance of their raw materials, ceramics have experienced an increase in their use for various applications in many fields. This variety of applications requires customization of the ceramic material; therefore new formulations are tested and accepted into production on regular basis. Figure 7 offers a better understanding of the large percentage of elements from the periodic table of
elements that can be found in ceramics . 18
I 2 3 4 5 6 7 8 9 IB 11 12 13 14 15 IS i; IS IA V1IA
DA U1A IVA VA VIA V11A
3 4 5 6 7 S Li Be B C N 0
12 14 14 15 11 1ITB IVS VE VIB V1IB vin IB IIB 13 Na Al Si Si P
19 20 21 -j*2 23 24 25 26 27 28 29 30 31 32 Ca Sc Ti V Cr Mn Ft Co Nfi Cu Zn Ga Ge
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Rb Sr Y Zr Nb Mo Tc Ru Rh Pel Ag Cd In Sn Sb
55 56 72 73 74 75 76 77 78 79 80 81 82 83 a Bil Hf Ta W Re Os Ir Ft An n Pb Bi
S7 88 Fr Ka
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 La Ce Pr Nd Pin Sm Eo Gd Tb Dy Ho . Er Tm Yb Lu
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 Ac Th Po U Np Pn Am Cm Bk Cf Es Fm Md No Lw
Figure 7: Elements present in ceramic materials
There are three main groups of ceramics widely employed in armour manufacturing as well as other more common applications. These groups are carbon based (carbides), nitrogen based (nitrides) and oxygen based (oxides), ceramics. The ceramic formulations based on the above groups together with their melting temperatures
(in degrees Kelvin) are presented in Figures 8 through 10 : 19
Carbides
Compositions and Tmp (JC)
1 2 4. 5 f, 7 Si 9 10 II 13 14 15 16 n 18 VILA
VIA HA 1VA VA VHA
ss 2720
NuO, A1.C, SiC 973 BIB 1VB VB VIB V'Iffi vm IB UB 2970
TiC VC G-jC, 3433 3600 2168
NbC Mo-C >I97C 3533 3770 1MB . ToC 3813 2900
TbC UC MS* 2863
Figure 8: Carbide compositions and their melting temperatures7.
Nitrides Compositions and T™ (K.)
1 2 4 5 6 7 8 9 10 11 12 13 14 1.1 16 17 : IS LA . VIIA
IU MA 1VA VA VIA VUA. BN 1118 2513 3000
AIM Sf IIIB IVB VB VIB VIIB vm IB HB >2475 2715
Ca3N, TiN YN CrN ClhN 1468 3200 2593 1770 573
2rN NbN 3250 2323
Ts,N 3360
TliN UN 21K13 3113
Figure 9: Nitride compositions and their melting temperatures7. 20
Oxides
Compositions and Tmp (K.)
1 > 3 4 5 7 H 9 10 11 12 13 14 15 16 !7 IS VA. VHA
nu 1VA VA VIA V1IA HA Li,0 BcO B-O, >J97S 1725 7b
MgO Al3Oj SiO; 3098 IVB VB VIB V11B IB ms 2322 1978
00 TO, V.05 Cr-,0, MnO Fe-Qs NiO 0*0 ZnO 703 3183 2113 M7 >26Q3" 1840 2257 15(18 2248
SiO •ZsO, Nb20s MoO, CdO SnO Sb205 Tea 2933 3123 1764 1068 3? 1773 2183 1353 92S 1006
BaO TO PbO 2283 2100 1744 ll59 1098
CeO, >2873
ThO, UO, H9i 3151
Figure 10: Oxide compositions and their melting temperatures7
From all of the above ceramics, some of the most widely used include:
• Alumina, (A1203) which up to date is still the dominant structural ceramic
for a wide variety of wear, erosion and impact resistant applications
including ceramic armour, due to its specific physical properties. Although
alumina based ceramics have a higher density (up to 4 g/cm3), their costs
for both the material and the manufacturing process, make them attractive
to the end users since normally they do not require the use of expensive
fabrication equipment. 21
• Zirconia (ZrCb) is an important wear, erosion and impact resistant
material. In combination with alumina, it is used for ballistic protection
when the weight of the armour is not critical given that the density of
alumina-zirconia ceramics ranges from 4.2 to 4.5 g/cm3 depending on the
actual composition .
Non-oxide armour ceramics such as boron carbide, silicon carbide, silicon nitride, aluminium nitride, titanium diboride, have better physical properties associated with
relatively low density (except titanium diboride-based ceramics) that are more beneficial
for ballistic applications than alumina based ceramics. However, the manufacturing of
these ceramics is relatively expensive and less productive.
• Silicon carbide (SiC) is the clear leader in terms of tonnage produced
primarily because there is a large market for abrasive grit. Silicon carbide
ceramics are in the form of reaction bonded or reaction sintered silicon
carbide, where the grains are bonded with a second phase or phases.
• Boron carbide (B4C) is more expensive than SiC mainly because of the
smaller volumes produced. If similar volumes were produced, the prices
would be comparable according to Tresser9.
The manufacturing process for the ceramic materials consists of the forming of
the ceramics into the shape desired and then the hardening (sintering) or the densification
process, at high temperature. 22
2.3 CERAMIC MANUFACTURING
The ceramic manufacturing consists generally of a forming process in which the shape of the ceramic part is obtained, followed by a hardening process to achieve the desired mechanical properties.
2.3.1 Ceramic Forming
There are several methods for forming ceramics. For simple shapes and mass production, uniaxial dry pressing is the procedure of choice. The dry granules of the ceramic powder with plasticizer, lubricants and binder are pressed automatically into a mould, generating parts with the same dimensions and similar density. Thus the quality of the parts is kept relatively constant.
The extrusion process can also be used in mass production, for parts with constant cross section. For more complicated parts, pressure can be applied isostatically. This procedure is performed by increasing the pressure of a liquid in which the ceramic material is introduced, in a sealed envelope.
The casting method for complex parts requires the ceramic material to be mixed with a liquid, usually water, forming a slurry (slip). This slurry is injected under pressure into a porous mould that absorbs or drains out the liquid. This procedure can be fully automated. 23
2.3.2 Ceramic Hardening
In order to obtain the final mechanical properties of the ceramic material, the powder conglomeration has to coalesce by establishing molecular links between the granules. The easiest way to initiate the reaction process between the ceramic grains is to increase their temperature over a well established limit for several hours in a controlled cycle. Usually done in natural gas or electric heated furnaces, at atmospheric pressure
(pressureless sintering process), the process is relatively simple for oxide based ceramics such as alumina.
In the case of non-oxide ceramics such as SiC and B4C, the oxygen present in the air during the heating will react with the ceramic molecules, degrading the carbon based finite composition. As a consequence, the heating process has to be performed in a controlled environment, where the oxygen has been replaced with inert gases. Developed only in the last 20 years, this controlled process is more expensive than the air firing, but highly effective, routinely producing SiC and silicon nitride (Si3N4) based ceramics to
99+% of theoretical density9.
2.3.3 Ceramic Hot Pressing
A transitional process for forming and hardening the ceramic pieces at the same time is the hot-pressing process, in which the material is pressed at high temperature to trigger the densification reaction. 24
When the pressure is increased during this hot pressing process, the porosity of the finite material is practically eliminated. The uncertainty factor introduced by the porosity in armour applications of the hot pressed ceramics is therefore greatly reduced, so the products perform more consistently and usually better than their pressureless sintered counterparts.
2.3.4 Reaction Bonded Ceramic
The procedure of linking the ceramic granules with the aid of a second phase has been applied for several ceramics. At high temperatures, the molten material will fill the voids and act as a binder. When part of the metal reacts with the ceramic matrix (usually containing free carbon) generating more ceramic material that fills up the voids, the process is called reaction bonding. These finite products are usually not well suited for ballistic impact applications.
In one publication, Hazell10 talks about the reaction bonded process for the manufacture of ceramics, such as SiC that has been around since the 1950's. He considers that its main advantage is the low cost for manufacturing ceramic simple forms to controlled tolerances and relatively low processing temperatures, with no pressure required; reducing the cost of the final product. The main disadvantages of these ceramics making them less attractive for armour applications are the presence of porosity and the relatively large amount (10-20 %) of unreacted material such as silicon, which are sources of failure during impact and penetration.10 25
The reaction bonded process is still promising from another perspective, namely the possibility to incorporate particles or fibres with superior characteristics in the starting mixture, essentially generating ceramic based composite materials. According to
Tresser , these ceramic based composites, although potentially better suited for certain applications, have not found widespread commercial use at this time, probably because of the difficulty to obtain homogenous mixtures during the manufacturing process.
2.4 ARMOUR USE OF CERAMICS
The armour application for ceramic materials is a relatively new development.
Only in about the last 30 years has this type of armour proved its efficiency in large scale deployments. Its main advantage is its reduced mass; the ceramic armour has a lower areal mass compared to the steel armour needed to defeat the same kinetic energy based threat.
According to Robertson11, the highest performance armour ceramics are related in physical characteristics to diamond. Nevertheless these are difficult to mould into pieces of ceramic that are of use in an armour system. As a consequence, it is necessary to either contaminate the system with additives to aid sintering and sacrifice much of the improved armour performance or apply simultaneously heat and pressure (hot pressing) with the associated cost implication.11 The ceramics identified by Robertson11 for being of principal interest for armour manufacturing are boron carbide, silicon carbide and aluminium oxide (alumina). 26
A comparison of the characteristics of the rolled homogenous armour (RHA) steel and some of the most representative ceramic materials used in armour applications is presented in Table l6.
Table 1: Material properties for various ceramics compared to RHA6.
RHA AL203 SIC BtC GLASS
Density, p (g/cm3) 7.85 3.4-4.0 3.0-3.2 2.4-2.5 2.2-2.5
Elastic Modulus, E (GPa) 205 220-400 280-476 440-470 46-94
Compressive Strength (MPa) 794-1170 1930-2600 1725-2500 1400-3400 1000
Tensile Strength (MPa) 794-1170 250-400 310 155-350 50
Poisson's Ratio, v 0.3 0.21-0.27 0.19 0.16-0.21 0.17-0.24
Sonic Velocity, c (km/s) 5.1 9-10 9-11 11-13 5.7
0 5 Fracture Toughness, Ktc (MPam ' ) 10-20 3-6 3 to 5 2.5 to 4 0.8
Knoop Hardness, H (kg/mm2) 150-350 2000-2100 2700-2800 2800-3500 500
Relative Mass Efficiency, ME 1.0 1.68-2.78 2-3.5 3.45-4.38 1.06
Raw Material Relative cost 1.0 1.0-2.4 2.0 10-19 1
In comparing chemically similar ceramics, the manufacturing process plays a major role in their final properties. Starting from essentially the same chemical ingredients, the hot pressing process can lead to ceramic armour with different mechanical properties compared to a reaction bonded process. As an example, Table 2 compares few ceramic materials manufactured in different ways and the influence of the manufacturing process on their final characteristics. 27
Table 2: Property Comparison of Typical Armour Ceramics.
Ceramics Density Vickers Fracture Young's Sonic Flexural g/cm Hardness Toughness Modulus Velocity, Strength GPa Kic, MPa.m05 , GPa km/s MPa Alumina, 3.60 - 12-18 3.0-4.5 275 - 9.5 200 - Sintered 3.95 450 11.6 400 Alumina-Zirconia 4.05 - 15 - 20 3.8-4.5 300 - 9.8 350 - Sintered 4.40 340 10.2 550 Silicon Carbide 3.10 - 22-23 3.0-4.0 400 - 11.0 - 300 - Sintered 3.20 420 11.4 340 Silicon Carbide 3.20 - 20-23 5.0-5.5 440 - 11.2 - 500 - Hot-pressed 3.28 450 12.0 730 Silicon Carbide 3.06 22 3.9 384 - 284 Reaction Bonded Silicon Nitride 3.20 - 16-19 6.3-9.0 - - 690 - Hot-pressed 3.45 830 Boron Carbide 2.45 - 29-35 2.0-4.7 440 - 13.0 - 200 - Hot-pressed 2.52 460 13.7 500 Boron Carbide 2.57 28 5.0 382 - 278 Reaction Bonded Titanium Diboride 4.55 21-23 8.0 550 - 350 Sintered Titanium Diboride 4.48 - 22-25 6.7-6.95 550 11.0 - 270 - Hot-pressed 4.51 11.3 700 Aluminium Nitride 3.20 - 12 2.5 280 - 300 Hot-pressed 3.26 330 400
In Chu's opinion, boron carbide tends to be the most desirable ceramic for
armour applications due its combination of high hardness (superior to all other ceramics
except diamond) and low density (2.52 g/cm3). The main disadvantage in using this
material is the processing cost associated with ballistic applications. He points to silicon
carbide as a generally accepted economical alternative to boron carbide. Major
drawbacks present for all the ceramic materials used in the ballistic field include, in his
opinion, difficult fabrication methods, limited armour tile geometries and a high cost for 28 the finished tile. The relative lack of multi-hit capability due to the shattering is mentioned as well.
Buchar14 considers that if one takes into account the cost of different ceramics, one can conclude that alumina is still the most effective ceramic for providing reasonable protection in view of the fact that the use of higher performing ceramics such as silicon carbide and boron carbide is relatively costly for large volume applications due to processing difficulties14.
In fact, all the ceramics with projectile stopping capabilities are currently utilized in ballistic applications. The particularity of each armour application needs a specific ceramic formulation whose qualities can be tailored for that particular application.
Referring to their manufacturing, it is worth mentioning what Barnett15 stated, that perhaps one of the most serious shortcomings of ceramic technology is the difficulty
of consistently producing the same material with identical and uniformly distributed
micro and macro structure.
Overall, the ceramic armour production is increasing due to growing demand
driven by the new conflicts involving developed nations all over the world. At the same
time, the quality of the final products is of paramount importance since the lives of the
military personnel depend on the functionality and reliability of their protection.
Therefore, not too many companies are sharing this particular market. According to
Gooch16, in 2001 there were only a few major producers of ceramic armour, as shown in
Table 3. Most of these companies can still be found on the market. 29
Table 3: 2001 Ceramic Armour Producers :
Ceramic Type Producer / Type Sintered Coors CAP3 99.5% Alumina Morgan Matroc (UK) Alumina ETEC Alumina Ceradyne Sintered SiC Pure Carbon SiC Reaction-Bonded M-Cubed (Simula) SiC 2 MC (Australia) SiC, B4C Metal Matrix Composite Lanxide Dimox AS 109 Lanxide Dimox - HT Hot Pressed Cercom B4C, TiB2, WC Ceradyne B4C, TiBj, SiC Saint-Gobain B4C
2.5 COMBINED ARMOUR
It is rare for armour to be manufactured from a single material on the modern
battlefield anymore. In general, modern and effective protection can be provided by a
combination of steel or other metal, composite materials and ceramic. For lighter armour
systems, when the weight is of concern or the threat is not particularly high, modern
armour can consist of ceramic materials backed by metals or composites or metal inserts
in front of composite materials. The general rule is that the impact face material has to be
hard enough to shatter or mushroom the projectile while the back material deforms and
stops the fragments from penetrating any further.
As an example, a ceramic armour system consists, in general, of a ceramic layer
bonded with a combination of high tensile strength backing materials such as long fibre 30 composite (e.g. Kevlar®, Spectra® or fibreglass) and, sometimes with soft metals (e.g. aluminium). A layer of ductile material in front of the ceramic to retain the back ejected fragments (the spall cover) can also be part of the armour system. Upon impact of the bullet with high kinetic energy, the hard-faced ceramic is cracked and broken, and the residual energy is absorbed by the backing material. The backing material stops the fragmented ceramic and the bullet, previously deformed, eroded or shattered by the ceramic. As an example of such armour system, a chest plate used in personal protection
(body armour) is presented in Figure 11.
Figure 11: Example of a modern combined armour system .
The efficient use of these materials can significantly reduce the armour weight,
1 Q while providing the same level of protection as the RHA. Gotts and Pippa stated that the 31 baseline component of the armour requirement is the threat to be defeated. In their opinion, this should be a threat that is of direct relevance to the end user. In other words, the fulfilment of this condition will provide the level of protection desired, while eliminating the excess mass of armour not needed in the particular circumstance in which it is used.
The NATO Threat Assessment to the Dismounted Combat Soldier states that personnel and equipment require effective protection against conventional fragmentation and ballistic munitions in a mobile environment. The report outlines the importance of
weight versus the efficiency of ballistic plates used in various situations, either as
individual equipment or as vehicle protection. Hence, the obsolete use of massive steel
plating as a standard for protection, has been replaced by multi-layer covers for vehicles
such as tanks, light armoured vehicles (LAV) or armoured personnel carriers (APC)
containing a smaller amount of steel together with other materials, such as rigid ceramic
plates or composite materials.
Despite the fact that studies related to ceramic-based composite material
protective armour have led to the continuous improvement in their performance, these
systems are far from being fully optimized, including the selection of the materials, their
layering sequence and fibre orientation, the goal of the optimisation process being to
obtain a lighter, thinner armour with better ballistic characteristics. 32
CHAPTER 3: THEORETICAL CONSIDERATIONS ON
BALLISTIC IMPACT
3.1 ELASTIC IMPACT
During and after World War II, most of the major research laboratories throughout the world carried out intensive studies of dynamic processes in materials,
focusing on high velocity ballistic impact. Thousands of research papers have been
published in the last 60 years and countless conferences debated every aspect of the
phenomena involved in the ballistic impact of materials.
The essential physical phenomenon on impact is the time-dependent pressure
generated in both the impactor and the target. The pressure causes stresses that can cause
failure, above a certain pressure limit.
No matter the part of this subject investigated, the simple equation relating the
kinetic energy with the velocity of a mass is always considered the starting point:
2 Ec=—mv (Equation 1)
The kinetic energy of a projectile increases proportionally with its mass (m), and
with the square of its velocity (v). The basic formulation for the energy delivered by an
object of mass mtoa target can be expressed as: 33
dE = F dl (Equation 2)
where F is the force that will act over the length dl. This kinetic energy is translated into damage on the projectile and the target.
A concept of paramount importance for the present study is represented by the fundamental differences between the static (or quasi-static) and the dynamic deformation of a material. With static and quasi-static deformation we have a state of equilibrium at
any time; explicitly the sum of forces acting on any element in the body is close to zero.
At a high rate of deformation, one portion of the body is stressed while the other portion
has not experienced this stress yet. In other words, in the second case the transformation
has to take into account the velocity of the stress travelling through the material. The
deformation associated with the stress (the strain) also travels through the body of the
material. These unique traveling velocities of the stress as well of the strain in each
material, easily calculated with a good degree of confidence, are called stress waves
respectively strain waves.
The most common waves encountered in impact-type applications are elastic
waves. Their velocities through continuous materials depend on the characteristics of that
particular material as well as the type of wave.
When the amplitude of the stress wave exceeds the elastic limit of the material at
a given strain rate (Hugoniot Elastic Limit (HEL)), plastic deformation takes place in the
ductile material. This plastic deformation propagates in the impacted material in a similar
manner to elastic deformation, namely through waves, called plastic waves. An important 34 observation is that these plastic waves propagate at lower velocities compared with the elastic waves.
According to Meyers19, when a medium is bound (no lateral flow of material is allowed), the plastic wave can become extremely steep and as narrow as a fraction of a micrometer. He states that these waves can travel through the material without changing its macroscopic dimensions (since the material is bound), and therefore a state of compressive stress close to hydrostatic compression establishes itself. These waves are called shock waves.
On the same topic, Zukas1 considers that the shock waves form in media whose characteristics are that the velocity of propagation of large disturbances is greater than the propagation velocity of smaller ones. Thus, the stress pulse develops a steeper front on passing through the medium, and the thickness of this front is ultimately determined by the molecular constitution of that medium.
It is worth mentioning for our study that during an impact event, similar waves are generated in both bodies involved in the event. For the case of a projectile impacting a target, the elastic, plastic or shock waves, depending on the impacts particular conditions, are produced in the target material, as well as in the impactor.
In his book on dynamic behaviour of materials, Meyers19 illustrates and describes the elastic waves produced in solids, usually generated by an impact event.
According to him, the most common types of elastic waves in solids and their descriptions are:
1. Longitudinal Waves. These waves correspond to the motion of the particles back and forth along the direction of wave propagation such that the particle velocity (Up) is 35 parallel to the wave velocity (U). If the wave is compressive they have the same sense; if it is tensile, the sense of the U is opposite compared to the sense of Up.
2. Distortional (or Shear) Waves, are generated when the motions of particles conveying the wave are perpendicular to the direction of propagation of the wave itself.
There is no resulting change in density and all longitudinal strains Su, £22 and £33 are zero. In other words, all the particle displacement is perpendicular to the wave propagation direction (Up is perpendicular to U).
3. Surfaces Waves are analogous to waves on the surface of water. The particles in
these waves describe elliptical trajectories. The surface waves, also called Rayleigh
waves in solids, are described as a particular case of interfacial waves when one of the
materials has negligible density and elastic wave velocity.
4. Interfacial (Stoneley) Waves are special waves that form at the interface of two
semi-infinite media with different properties in contact.
5. Waves in Layered Media (or Love waves) are mostly generated during
earthquakes, since the earth is composed of layers, or strata, with different properties.
This layering is responsible for the emergence of special wave patterns. It is known that
earthquakes produce waves in which the horizontal component of displacement can be
significantly larger than the vertical component, behaviour not consistent with Rayleigh
waves.
6. Bending (or flexural) Waves involve propagation of flexure in a one-dimensional
(bar) or a two-dimensional configuration.1
When the fronts of some waves are combined, the result can be called a Head
Wave. 36
Most of the waves described above are illustrated in Figure 12.
Wavelets Rayletgh wove
Longitudinal wave
Figure 12: Waves generated at the impact on solids 19
Zukas1, in his elastic theory for isotropic solids indicates that impact generates mostly two types of waves. He defined them as dilatational (or longitudinal) waves, in which the particle motion induced by the disturbance is normal to the wave front and distortional (or transverse or shear) waves as those wherein material particles move in a plane at right angles to that in which the wave front propagates. According to him, under the proper loading conditions, torsional and flexural waves may also be generated.1
The velocities of the longitudinal and shear waves generated in isotropic solids are calculated with the formulae presented in Table 4. 37
Table 4: Wave velocities in isotropic solids1.
Extended Media Bounded Media
Longitudinal: _ A + 2/i _ E(l-o) nl c , E L i= — ° " P p{\ + v){\-2v) P
Shear: ^2 JJ. G E , G c~s= — P P 2p(l + y) P
Where: E = Young's modulus
u = Poisson's ratio
p = density
X,[i = Lame parameters
G = shear modulus
In the same work, Zukas1 mentions the elastic waves that may be propagated along the surface of a solid. The most studied types according to him are the Rayleigh waves and the Love waves. In describing them, Zukas states that Rayleigh surface waves decay exponentially with depth from the surface to the medium interior. The Love wave, confined as well to a relatively shallow surface zone, is a shear wave. The Love wave is mostly generated in a layer of material that possesses one set of physical constants, overlying a layer that possesses different physical constants.1
The waves traveling through solids generate stresses in their path. In order to
obtain the intensity of the propagated stress due to a longitudinal wave, we start with the 38 relationship between the longitudinal stress and the longitudinal particle velocity at any point in a body, from Newton's second law, as suggested by Zukas1.
FLdt = (mvL) (Equation 3) where: Fi = longitudinal force acting on a given cross section
dt = the time the force acts
m = the mass the force acts upon
vi - the particle velocity imparted to the mass by the force
F Since a = —- and m = pAdl where dl is the distance the pulse has moved in time dt, the above equation can be written as:
CiAdt = pAdldvL (Equation 4) or
dl . <7 = pcL(AvL) (Equation 6) Similarly, Zukas proposed the equation for the shear stress generated by a traverse disturbance as: T = pcs(Avs) (Equation 7) where: r= the shear stress cs = the velocity of propagation of the transverse disturbance Avs - the change in the particle velocity in shear 39 For impact generating stress pulses above the maximum stress for elastic wave propagation (JHEL, there will be two waves traveling through the solid as mentioned before, namely an elastic wave followed by a slower, plastic wave. A typical illustration of a generic wave propagation and plastic wave attenuation in solids following an impact is given by Matsalla . The image, presented below can be found slightly modified in one of the Zukas credited publications1. Stress Distance P __. CT J el- L^_£ J^yjf cL MA CI! CL' 1 „ ^ * Figure 13: Plastic wave attenuation during wave propagation in solids6. 40 The elastic compressive longitudinal wave travels through the solid with the ci velocity, followed by a plastic wave with the velocity c£, lower than the elastic one. A release wave, generated at a later moment traveling with the same velocity as the first elastic wave ci attenuates the plastic wave by unloading it from the rear. The distance at which this unloading occurs is called the catch-up distance and depends on the incident pulse. For most of the commonly used armour ceramics, which are the materials of interest in this thesis, the longitudinal elastic wave velocities are presented in Table 2 as sonic velocities. Knowing the values of the longitudinal elastic wave velocities in the target and the projectile materials, one can calculate the shock stress generated upon the impact. This shock stress will generate a shock wave travelling through both the target and the projectile, usually with devastating consequences for both materials. Useful information can be obtained if the stress generated by this shock wave in the material is to be determined. Robertson , assuming a one-dimensional state of strain proposed the following equation for the stress generated by the shock wave: AClP2C2 a = v0 (Equation 8) plcl+p2c2 where: a = one-dimensional stress generated at impact Pi = the bulk density of the projectile p2 = the bulk density of the target Ci = the elastic longitudinal wave speed for the projectile C2 = the elastic longitudinal wave speed for the target material 41 vo = the impact velocity of the projectile. Gotts21, who used the same formula to obtain the unidirectional stress in the target, went a step further and concluded that the pressure in the target is one third of the generated stress. The next equation for evaluating the shock stress, suggested by Robertson and Gotts11 uses the acoustic impedance of the materials that can be easily determined for well known homogenous materials. zt x z a = V0 — (Equation 9) Z Z ,+ P where: a = the general shock stress vo = the projectile impact velocity zt = the acoustic impedance of the target zp = the acoustic impedance of the projectile 3.2 IMPACT DAMAGE For the vast majority of the impact events of interest, the stress generated in the target is higher than the elastic limit of the material. For ductile materials such as most metals, plastic deformation will occur. Brittle materials such as ceramics, on the other hand, will fracture with no significant plastic deformation when stressed above their elastic limit. 42 In general, fracture is classified according to 3 modes of loading: Mode I (tension or opening); Mode II (shear or sliding); and Mode III (transverse shear or tearing). For all modes illustrated in Figure 14, tension at the tip of the crack is needed for the crack to grow. MODE I MODE II MODE III Opening Sliding Tearing Figure 14: Classic illustration of the modes of applied stress for crack propagation In understanding the physical phenomena and the relationships among different parameters involved in the failure of materials, the work done by Callister22 has been found to be extremely useful. In assuming Mode I of crack propagation, namely a crack oriented perpendicular to the applied stress, the maximum stress at the tip of the crack am is: 43 f _ ~\ 1/2 1 + 2 (Equation 10) U t; where: am = the maximum stress at the tip of the crack pt = the radius of curvature of the crack tip a = the length of a surface crack or half length of an internal crack For brittle materials, the crack tip curvature has an extremely small radius relative to the length of the crack, therefore the fractional part of the above equation may be very large and thus, by neglecting the other unit term, the above equation could be approximate by cr = 2crr (Equation 11) A J where: crm = the maximum stress at the tip of the crack (To = the magnitude of the nominal applied tensile stress Pt = the radius of curvature of the crack tip a = the length of a surface crack or half length of an internal crack From equation 11, the stress concentration factor at the tip of the crack Kt could be obtained as ' a ^ K ,= ^ = 2! (Equation 12) \Ptj where: Kt = the stress concentration factor at the tip of the crack 44 am = the maximum stress at the tip of the crack ao = the magnitude of the nominal applied tensile stress pt = the radius of curvature of the crack tip a = the length of a surface crack or half length of an internal crack This factor presents the degree of amplification of the stress at the tip of the crack relative to the nominal stress applied to the body. 22 Callister , referring other authors, published the formula of the critical stress /2 (VRV V o-, = (Equation 13) V na J where: crc = the critical stress required for a crack to grow in a brittle material E = the modulus of elasticity ys = specific surface energy a = the length of a surface crack or half length of an internal crack The crack propagation releases the energy stored in the elastically deformed material, called elastic strain energy. The specific surface energy ys is a term related to the energy stored on the new surfaces created at the faces of a crack. For materials that are not perfectly brittle, the ys factor in the above equation has to be replaced with (ys + yp) where yp represents a plastic deformation energy associated with crack extension. 45 Besides the stress concentration factor Kt, there is a similar parameter K, which defines the stress intensity at the tip of a crack, or in other words, it provides for a convenient specification of the stress distribution around a flaw" . It is defined as: K = Yo^lTTa (Equation 14) where: K = the stress intensity factor at the tip of the crack a = the length of a surface crack or half length of an internal crack Y = dimensionless parameter or function that depends on both the crack and the specimen sizes and geometries as well as the manner of load application"" From this relation (equation 14), a very important parameter can be obtained, namely the fracture toughness Kc of a material, whose equation is Kc = Y(TC ^Tia (Equation 15) where: Kc — the fracture toughness of a material crc = the critical stress for crack propagation a = the length of a surface crack or half length of an internal crack For "Opening" or Mode I of fracture and relatively thick specimens, the fracture toughness Kc is known as the plane strain fracture toughness KJC, defined as K!C = Ycr^Jna (Equation 16) where: Kic = the plane strain fracture toughness for mode I of fracture a= the magnitude of the nominal applied tensile stress 46 a = the length of a surface crack or half length of an internal crack Y = dimensionless parameter or function that depends on both the crack and the specimen sizes and geometries as well as the manner of load application In some particular cases the Y factor can approach unity. It is worth mentioning that fracture toughness Kic values are given for quasi static conditions. In dynamic events, the material behaviour might be different compared to the quasi-static behaviour, resulting in different values for some mechanical proprieties of the material, including the above mentioned fracture toughness. For ballistic ceramics, the fracture toughness KJC is an important parameter in terms of material properties. Fracture toughness alone does not give a value for ballistic resistance, but when used in conjunction with elastic modulus, strength and hardness of the material, the fracture toughness KJC defines the ballistic properties of the ceramic material. 9T Medvedovski suggested a semi-phenomenological criterion of evaluation of ballistic energy dissipation ability based on the fracture toughness, by using the formula: „. ...^HvxcxE D = 0.36 ; (Equation 17) Kic where: D = ballistic energy dissipation characteristic Hv = Vickers hardness of the material c = sonic velocity in the material E = modulus of elasticity Kic = plane strain fracture toughness 47 By generalizing the above formula, Medvedovski reasoned that, since ballistic energy dissipation characteristic also depends on the phase composition and structure of ceramics, a "structural" factor S taking into account the micro-structure of the ceramic material has to be inserted in the formula, resulting in: „. _, HvxcxE D = Sx (Equation 18) where: D = ballistic energy dissipation characteristic S = "structural" factor Hv = Vickers hardness of the material c = sonic velocity in the material E = modulus of elasticity Kic = plane strain fracture toughness Another formula suggested by the same author relates the brittleness B of a ceramic material to its fracture toughness by the relation n HvxE B = — (Equation 19) Kic where: B = brittleness factor Hv = Vickers hardness of the material E = modulus of elasticity Kic = plane strain fracture toughness 48 Medvedovski indicates that these formulae may be successfully applicable for dense homogeneous armour ceramics. The limitation for heterogeneous armour ceramics is that the ballistic energy dissipation ability and ballistic performance depend significantly on phase composition and structure of the ceramics and related crack propagation.23 In armour systems consisting of ceramic strike face and a backing, after the ceramic is fractured, it still resists the penetration due to inertial effect and backup support. Although difficult to quantify, the effective strength of the fragmented ceramic a'oc can be estimated by an empirical equation suggested by Moshe , namely: - mcrob a'0c = 0.2or0c 1 - exp (Equation 20) H 0.2cr0c \ cJ where: a'oc = the effective strength of the fragmented ceramic Goc = the in-plane strength of the ceramic layer aob = the in-plane strength of the backup layer Hb = the thickness of the backup layer Hc = the ceramic thickness between the current projectile front and the interface, m = bonding quality factor, unity for 100% optimum bonding This equation might provide some benefit for simulations, but it is not particularly useful during experimental testing, due to the difficulty in measuring the remaining thickness of the ceramic Hc at intermediate moments during the event. 49 In summary, many authors have tried to quantify the ballistic behaviour of ceramic targets through various theoretical or more empirical equations. However, the ballistic performance of a ceramic is system dependent (no one design is optimal for all ceramics or ceramic variations ) and the failure mechanism of these brittle materials Oft under dynamic loading conditions still is not fully understood . As a consequence, the available relations do not fully cover the dynamic response of a ceramic target subjected to ballistic impact. 3.3 ARMOUR CHARACTERISTICS According to Kaufmann27, there are several beneficial properties of ceramics undergoing ballistic impact. He considers that the initial resistance to penetration is provided by the compressive strength of the ceramic. High hardness value is beneficial in blunting or destroying the tip of an armour piercing round. High bulk, shear and Young's moduli increase the ceramics resistance to deformation. High shear strength allows the material to resist failure due to the large shear stresses generated near the impact site. A high Hugoniot Elastic Limit (HEL), which represents the dynamic yield strength of a material, is beneficial as high transient pressures can be reached near the point of impact.27 Robertson and Gorts believe that in order for the armour to be efficient in defeating the relevant high velocity round, the strike face material must transmit a stress pulse of sufficient magnitude and for a sufficient duration back into the structure of the 50 bullet to disrupt or destroy the penetrator element of the bullet either by yield processes, which leads to mushrooming of soft core rounds, or by fracture that results in fragmentation of the hard materials such as AP penetrator. As a consequence, the destruction of the integrity of the projectile reduces the stress concentration on the armour in a function relating to the diameter of the dispersing projectile. Alternative processes such as erosion can also be important in defeating high velocity projectiles, although these processes do not necessarily involve an increase in the area of the interaction between armour and projectile during the ballistic process and as a result have a different response in relation to projectile velocity.11 Each type of material reacts differently under ballistic impact. The dynamic phenomenon together with the intrinsic properties of that particular material yields a unique kind of response and, implicitly, defines the best usage of that material for ballistic protection. According to Meyers , there are several modes in which a projectile interacts with the armour and, therefore different ways in which the armour can be penetrated under impact. These mechanisms depend on the armour composition, projectile characteristics and impact velocity. These interaction modes are illustrated in Figure 15 and include: a. spalling; b. plugging; c. ductile hole growth; d. conoid formation; e. melting / vaporization and plume formation; 51 f. concurrent erosion of projectile and target; g. petaling; h. comminution (cracking and crushing). Plume Spoil •OHH (a) (e) lb) (C) (d) Figure 15: Projectile armour interaction Spalling (a), plugging (b), ductile hole growth (c) and petaling (g) are encountered in the impact of ductile targets such as steel. The spalling encountered in stiff CFRC 52 based armours is called delamination. The conoid formation (d) and the comminution (cracking and crushing) (h) are characteristic of impact with brittle materials, such as ceramics and glass. The concurrent erosion of the projectile and the target (f), as well as the melting / vaporization and plume formation (e) can happen for both brittle and ductile materials in the target, when the impact velocity exceeds a certain limit. Iremonger and Gotts3 made some predictions on where the development of projectiles and armour systems will head in the near future. They expect pre-formed fragments to be smaller than present ones and delivered with greater accuracy. The percentage of bullets with hardened penetrators (AP rounds) encountered by a target will be greatly increased. The ability to place two high velocity shots on an armour plate at ever decreasing separations will also be improved. Unfortunately, Iremonger and Gotts do not expect a matching increase in armour material performance. They conclude that the industry is now considered to be in the business of fme-tuning and, although huge advancements would make armour designers lives easier, they are not expected in the foreseeable future.3 3.3.1 Impact on Metals Metals are the cheapest and oldest choices for ballistic protection. They have been used for the longest time against projectiles, and their behaviour under impact is well documented. 53 However, the continuous improvements in projectiles especially during and after WWII led to more ways of defeating a metal shield. The increase in projectile velocity beyond the elastic limit of the metal, the use of new materials for the projectile's perforating core (harder or pyrophoric), the improved shapes of projectiles and the explosive driven projectiles are some of the challenges that a steel armour designer has to take into consideration. One of the main drawbacks in employing metals in armour is the spalling phenomenon. This phenomenon is present when the projectile does not perforate the entire armour, but metal fragments from the back face of the armour are projected, sometimes doing more damage than the projectile itself. The explanation for this behaviour resides in impact dynamics. The initial stage of impact is characterised by shock waves developed in the armour frontal layer and in the projectile. When the shock wave in the armour reaches the back surface, reflected waves called rarefaction waves are generated which propagate backwards towards the impact surface. The axial stress, now in tension usually exceeds the tensile strength of the metallic armour material, leading to breakage and ejection of one or several "slabs" from the back of the armour. Projected with high velocities, these metal fragments produce a tremendous amount of damage, especially in enclosed spaces such as tanks and bunkers. On the modern battlefield, the rolled homogenous steel armour (RHA) is being either integrated in composite armours or entirely replaced by other materials for most armour applications. 54 3.3.2 Impact on Continuous Fibre Reinforced Composites The impact behaviour of the CFRC has some peculiarities compared to the classic materials. The flexible material elongates during the projectile stopping process, absorbing a large amount of the entire kinetic energy of the projectile. This deformation slows down and stops the projectile over a larger period of time and over a longer distance when compared to harder or more brittle armours. If the projectile's kinetic energy is too high, after the elongation process there is fibre breakage and the armour is defeated. In the case of projectiles stopped by the armour, the deformation of the composite still takes place and can lead to blunt trauma if used in close vicinity of the personnel's body, such as in protective vests. If the continuous fibres are used in conjunction with a rigid matrix, the fibres are added in plies with specific orientation until the desired resistance is achieved. During the impact, the first damage to occur is delamination between the plies. Then, the matrix holding the fibres yields and the fibres start acting similarly to flexible (no matrix) designs. Although the damage of the matrix absorbs some energy, the fibres in stiff armour designs are essentially deformed over a shorter length compared to unbounded fibres, and only over the area where the matrix has been destroyed. The short length that has been deformed in the case of these fibres will absorb only a fraction of the energy absorbed by the flexible armour before breakage. This occurrence is desirable where a deeper deformation of the armour will affect the protected person or equipment. A common example of this situation is the protective 55 helmet where a blunt trauma due to the elongation of the composite fibres is unacceptable. A common use of the continuous fibre reinforced composites in armour design is where the composite is placed behind a harder material, such as ceramics. The incoming projectile will be shattered and its fragments spread over a larger area by the frontal ceramic layer before being caught by the composite material. 3.3.3 Impact on Ceramics The brittleness of the ceramic material imposes some particularities to the impact process compared to more ductile armour materials such as steel or aluminium alloys. Metals absorb the energy of projectile by a plastic deformation mechanism. In the case of ceramics, the kinetic energy of projectile is absorbed by a fracture energy mechanism, as mentioned by Medvedovski. The elastic deformation of the ceramic armour can absorb some of the projectile's impact energy, but typically, not all of the total impact energy is absorbed. Above the elastic limit of ceramics, the kinetic energy of the incoming projectile is dissipated through a fracture generating mechanism, identified by Matsalla as "plastic deformation". Therefore, the initial moment of the impact is of critical importance in defeating the projectile since the shattered ceramic will be less resistant to penetration than the intact material. As the penetration increases, more and more ceramic will exceed the elastic limit and will fracture, until the projectile is either stopped or the ceramic is 56 perforated. As a consequence, the resulting behaviour of the ceramic is a complicated combination of the integrated responses of the damaged and undamaged regions, as suggested by Normandia.2 Yadav and Ravichandran"0 admit that the ballistic penetration of ceramics is a complex field of endeavour and sometimes researchers have proposed alternative explanations for nominally the same phenomenon. These authors acknowledge that for brittle materials such as ceramics, the target resistance may depend on the initial impact velocity and the velocity at which damage propagates within the material. On the other hand, they consider that there is not enough information about damage propagation velocity within the brittle material . It is agreed upon that the ballistic performance of ceramic materials depends on a number of properties. According to Medvedovski , these properties include density and porosity, hardness, fracture toughness, Young's modulus, sonic velocity, mechanical strength and some others. He considers that one cannot find any single property that can be directly correlated with ballistic performance because the fracture mechanism during the bullet impact is very complicated since the crack formation is caused by different stress factors, and has an extremely short duration23. Medvedovski23 concludes that crack propagation and energy dissipation mechanisms and ultimately ballistic performance are strongly influenced by the microstructural features of the ceramic material. Pickup agrees that it is difficult to correlate a single material property of a ceramic to its ballistic behaviour. He mentions that researchers have considered hardness, fracture toughness, Weibull modulus and combinations of these to generate ceramic 57 failure models. The correlation difficulty seems to increase for ceramics where there is a dwell transition, which involves many material and geometrical factors. As a result, the vast majority of researchers concede the challenges faced in predicting or simulating the behaviour of ceramic armours under impact. Renahan admits that because of the complexity of ceramic fracture mechanism, the researchers working on ceramic armour development have been reliant on testing methods to evaluate their ceramics, as predictive methods have yet to fully explain the ceramics' behaviour. He states that although the fracture of ceramics has been studied in depth, an accurate model of their failure has not yet been developed0 . Renahan considers also that the predictions for ceramic fracture are erratic because of the high velocity of the impact, hence the dynamic loading, which results in little strain before failure, along with the brittleness of the ceramic. In addition to this, he admits that the mechanisms involved in this impact are also not completely understood and it is not clear how to relate a ceramic's mechanical properties to its ballistic efficiency. "3 0 LaSalvia as well acknowledges that despite the efforts and expense of the previous 30 years, no much visible progress was apparently made in establishing a generalized understanding of the connection between a ceramic's physical and mechanical characteristics and its consequential ballistic performance. As the reasons for this lack of correlation he includes ceramic variability, non-standardized material characterisation and ballistic test methodologies. In addition to that, a general lack of understanding of the fundamental response of non-monolithic targets subject to ballistic impact and penetration is also responsible for the lack of useful information. 58 In describing the possible phases in ceramic perforation by a projectile. JangJ identified three phases that can possibly occur during the defeat of a ceramic protector by a ballistic impact, namely: 1. Shock, in which a high-pressure shock wave followed by a release wave propagates through the ceramic; 2. Penetration, in which the projectile penetrates the target by compressive yielding of the target materials in its path; 3. Fracture, when a fracture front propagates into the ceramic ahead of the projectile.34 Some other authors such as Hazell and Iremonger35 consider that under dynamic loading in compression, brittle materials generally deform inelastically. The inelastic response of these brittle materials is attributed to the nucleation, growth and coalescence of micro-cracks which can cause extensive stiffness loss and strength degradation within the material. Lu Jin and Zhai Jun , in their studies, concluded that for brittle materials such as glasses, ceramics and hard composites, the primary and most important failure mechanism is crack and micro-crack development and interaction with their neighbouring micro-cracks. Lu Jin considers that the micro structural interaction can absorb irreversibly the energy through a range of dissipative mechanisms associated with microstructural changes such as nucleation, growth, and coalescence of micro-cracks, internal friction, irreversible phase transformation, and the chemical reactions in elastic-brittle materials.36 Zhai Jun concludes that analyses at the micro and nano-size scales require two important considerations, namely the explicit representation and account of micro and 59 nano-material structures and the explicit tracking of failure processes in the form of crack / micro-crack initiation, growth and coalescence. These analyses have yet to be performed in a comprehensive manner. In accordance with the failure mechanism based on micro-crack development, in Medvedovski hints that in general, the energy dissipation through microcracking consists of a formation of numerous micro-cracks, characterized by the appearance of numerous stress micro concentrators. It appears therefore that the energy is spent on the formation of numerous surfaces. In order for the energy dissipation to be effective he considers that the microcracking should occur faster than general macro-crack propagation. LaSalvia noticed the influence of the micro-cracks that started the shattering process of the ceramic under impact loading. He thinks that these micro-cracks are perhaps initiated by "sharp" pores, which then coalesce together. According to most of the publications consulted, the authors such as McCuiston40, Normandia29'41, Lucuta26, Medvedovski23, Moshe24'42, Lopez-Puente43, St-Denis44 and Bryn James45 had similar descriptions on the ceramic behaviour during a ballistic impact event. They agree that during a ballistic impact event between a ceramic target and a projectile, a shock wave develops in the ceramic frontal layer and in the projectile. For the ceramic target, this shock wave precedes the projectile's penetration and propagates through the target. The Shockwave, though initially compressive in nature, can introduce and grow strength limiting defects in the target, in the form of micro-cracks, thus weakening it. Upon arriving at free boundaries or discrete material interfaces, the shock wave is reflected backwards towards the impact surface and becomes tensile. This propagating tensile shock wave will further weaken the ceramic target substantially as 60 ceramics typically perform poorly under tensile loads. After a short delay, the rarefaction wave is followed by break-up of the ceramic whose front can be considered to be a "shatter wave" as mentioned by Moshe- . Moshe predicts that the velocity of the break up wave front or shatter wave is estimated to be CQ/2 where Co is the velocity of bulk elastic waves and is approximately equivalent to the velocity of the Rayleigh surface wave for maximum crack propagation. In some unfortunate circumstances the weakening of the target by the shock wave can happen before the projectile even begins to penetrate the target, essentially decreasing the expected performance of that target, according to McCuiston40. The waves generated in the ceramic target following an impact event are also generated in the projectile as mentioned earlier. Due to the complexity of the stress pulse effect resulting from the combined waves and the shape and materials of the projectile, the final outcome for the impacting solid might not be easily predicted11. Projectiles, including the hard cored type can be shattered into small fragments, under certain conditions, as observed by Gooch16, Kaufmann27 , St Denis44, Robertson11, Woolmore46 andGotts21. Madhu47 found that as the velocity of the projectile is increased, both the projectile and the target break into a larger number of finer pieces, leading to more efficient energy absorption by spatially spreading the impact energy. This complex phenomenon has been observed by nearly everyone who tests armour but according to Skaggs48 it has never been adequately explained. Flinders49 underlines the fact that one cannot fully understand the damage phenomenon occurring in the projectile also in his study. He found that ballistic testing 61 with a WC-Co core projectile at 907 m/s showed that higher projectile hardness resulted in lower depth of penetration in ceramic targets. There are some arguable hypotheses trying to explain the above anomaly. As observed in a research document mentioned earlier , there are instances when the alumina present in ceramic armour might undergo a phase transformation when impacted by AP with WC core projectiles. The phase transformed alumina might have better stopping capabilities, decreasing the penetration of the projectiles. Another explanation given by Robertson and Gotts11 is that the ceramic exhibits a pressure hardening behaviour and its penetration resistance generally increases with increasing applied stress unlike metals and polymers. However, in the same document, Robertson and Gotts presented examples that contradicted their previous assumption by affirming "some ceramic materials at high shock stress undergo phase transformation and become amorphous. As a result a considerable effective reduction in yield stress is observed, i.e., effectively becoming softer. " In order to sustain this affirmation, they gave the example of hot pressed B4C that, at high bullet velocities, appears to reduce its effective hardness to something not much greater than high grade alumina, a fact confirmed by Robertson in another study. He observed that above a threshold shock stress, the ballistic performance of the boron carbide decreases considerably from one similar to silicon carbide down to something similar to a considerably less hard alumina . The residual chemical element left in the reaction bonded ceramics such as the silicon appears to undergo a similar transformation at shock levels about 11 GPa, lowering the performance of the SiC compared with its hot pressed counterpart when dealing with WC-cored AP. 62 As mentioned earlier, during a ballistic type impact between a projectile and the ceramic faced armour, deformation occurs in the projectile as well. For soft core projectiles the mushrooming effect caused by the ceramic is extremely effective in preventing the penetration of the armour. The AP projectiles, having a hard core made of hard steel or WC, are more efficient in perforating the ceramics. Few of these AP projectiles have their core relatively ductile. During the impact the hard type of core might still deform, but the deformation is not desired since it is relatively detrimental to performance. The deformation increases the contact area with the target and thus the penetration can decrease. In most of the instances though, the AP core is hard and brittle; therefore their cross section remains relatively constant during the perforation process. As a consequence, there are particular circumstances in which the core can be shattered into pieces upon impact. In examining recovered pieces of fragmented core of an impacting projectile, Kaufmann observed that the core was shattered into small fragments. By analyzing the fragmentation pattern he concluded that the fracture mechanisms involved in a ballistic impact are difficult to reproduce numerically due to the random nature of the crack propagation within the core . This random nature of the core fragmenting process can be influenced by a multitude of factors. Gotts18 admits that one of these factors is the consistency of the test bullets. He believes that features such as jacket thickness and hardness need to be consistent, otherwise the impact properties will also vary.18 The link between the projectile core fragmenting and the consistency of the softer parts of the same projectile might not be obvious, but the influence is notable. The first 63 occurrence during the impact of an AP projectile at the contact with armour is the damaging and erosion of the soft tip of the projectile. In some instances, as noted by the same author as above (Gotts, Philip L.) in another article21, the soft tip can damage some ceramic materials enough to significantly lower their performance before the core of the projectile engages them. For that reason, the time of contact loading between the core and the intact ceramic is of paramount importance in initiating failure in the core and decelerating the core and ultimately determining the amount of damage done in the core and the projectile's defeat. There are several authors that observed and reported the interaction between a hard core projectile and ceramic armours. Lucuta26, Medvedovski23, Zhai Jun37, LaSalvia50, Grady31, Pickup31, Moshe42, Hazell52 and Gooch16 described the interaction and the damage caused by such impact. Essentially, the simple mechanism of the ballistic failure process as explained by the above authors can be summarised as: 1. The projectile tip is destroyed at contact, followed by a shattering phase, when the penetrator can fracture and break on the surface of the ceramic plate. The shattering appears if the high compressive strength of the ceramic overmatches the loading produced by the penetrator impact, and the penetrator material can flow and shatter. At velocities above a certain limit, the material from the projectile tip can be ejected laterally with high speed. 2. There is a damage accumulation in the ceramic material initiated by compressive waves followed by tensile and shock waves and their reflections, and 64 bending of the ceramic tile and backing plate. During this stage, the ceramic material undergoes cracking. 3. The development and intersection of the above cracks produce fragmentation of the ceramic through a relatively large area of impact into particles ranging from a very fine powder (comminuted ceramic) to relatively large chunks. 4. During the ceramic fragmentation process, the penetrator or the rest of it might be left intact, might be eroded or shattered more, depending on several conditions such as impact velocity, geometry and mechanical characteristics of the ceramic and the projectile. 5. Complete break-up of the ceramic is presumed to occur when the front of the shatter wave reaches the forward extent of the "inelastic zone" of comminuted (or pulverised) ceramic surrounding the imbedded projectile, as enounced by Moshe. 6. Subsequent to break-up, continued penetration is resisted by fragmented ceramic pieces held in place and constrained by the surrounding material and by inertial effects. 7. During the entire process, the ceramic material contributes to defeat of the penetrator core through erosion mechanisms. The stages of ceramic cracks development, as suggested by the several authors mentioned above consist of the following: 1. The cracking starts close to the penetrator periphery as rig cracks concentric about the impact site due to the stress field (initially elastic) with largest tensile stresses in the radial direction. 65 2. A fracture conoid is then initiated close to the interface between the projectile and the target, following the path normal to the direction of the principal tensile stress. Thus, several large Hertzian cone cracks extend through the target, assuming trajectories 25-75° from the initial normal-to-the-surface direction, as explained by Hazell and Iremonger.52 3. Radial tensile cracks might be initiated at the back surface close to the axis of impact and star cracks might be formed at the side of the conoid. 4. The backup material yields at the ceramic interface and tensile spall planes are generated by tensile waves reflected from free surfaces interacting with inhomogeneities to nucleate micro-cracks.33 5. The tension that results in the ceramic as it bends following the motion of the projectile initiates axial cracks. 6. As the projectile continues to advance the principal tensile stresses are now in the circumferential direction that invokes six to twelve large radial cracks propagating outward from the impact centre . Blazynski considers that only hard angular penetrators generate these radial cracks. 7. Lateral cracks develop beneath the impact surface parallel to it in unloaded ceramic areas from the crater wall, as the penetrator of the projectiles enters into the ceramic. 8. The material in front of the penetrator is (micro) cracked and forms a comminuted or crushed region directly beneath the impact center which appeared to be wholly contained between the set of cone cracks which formed first. The comminuted 66 region consisted of a high density micro-cracked ceramic3 , compacted by the advancement of the penetrator into the target. An illustration of the evolution of the cracks during the impact of a ceramic material with a metallic sphere is presented in Figure 16, from RirterD . Side View Top View Surface /- Ring *r~ (f5)) Cracks Ring ~^ Cracks .s~J> ^ Figure 16: Evolution of damage produced by a spherical particle impact . 67 It is worth mentioning that in Figure 16 the target is considered to be semi- infinite, and thus there are no cracks due to target bending or reflections from the free surfaces. Other authors such as Shockey and Marchand55 or Wells56 impacted different ceramic target configurations then performed a comprehensive examination of each target, in an attempt to determine the cracks generated by the impact. Shockey, was able to identify several types of cracks, as presented in Figure 17 by sectioning the impacted ceramic. Figure 17: Sectional view of surface crack pattern in Si3N4 impacted by 2.4 mm WC sphere at 231 m/s . 68 Wells3 observed the ballistic damage in confined ceramic targets with the aid of X-ray Computed Tomography. The cracks observed included conventional conical, radial and laminar cracking, as well as spiral or circular beach-mark cracking and outer edge radial and periodic laminar cracking, which seem to be less-reported by other authors. An illustration of his findings is presented in Figure 18. Confining Resistance of Bulk Surrounding Elastic Ceramic Figure 18: Cracks generated in a confined ceramic target56. Three other observations about failure mechanism worth mentioning include: 1. The ceramic target not being penetrated at all upon an impact. 2. The uncertainty with regards to crack formation. 3. The presence of "shatter-gap". 69 These observations are developed in the following paragraphs. 1. The ceramic target might not be penetrated at all following an impact. During the initial phase, complete prevention of penetration can be accomplished by containing and constraining the ceramic target in such a way as to cause the projectile to be destroyed on the impact surface of the target ceramic, due to stresses at the projectile- target interface exceeding the elastic limit of the projectile material. The plastic deformation and radial flow is proportional with the projectile's velocity. This phenomenon has been called "interface defeat" or "dwell" and involves the total deceleration, erosion and radial flow of the projectile up to a threshold projectile velocity, Vth, as mentioned by Pickup , with no significant penetration in the ceramic. Despite the avoidance of penetration with interface defeat, considerable cracking damage still occurs, nevertheless, within the ceramic target. This threshold velocity of the projectile, Vth is commonly referred to as a critical impact velocity or ballistic limit. According to Zukas1, some of the parameters affecting the above mentioned ballistic limit are: Material hardness of the projectile Impact angle of the projectile (striking yaw) Projectile's nose shape, especially for lower velocities Projectile's material density, mainly at higher velocities Length-to-diameter ratio of the projectile, a higher value having a higher ballistic limit. 70 Anderson and Walker57 as well as Lundberg P and Lundberg B38 studied in depth the phenomenon related to the interface defeat and the transition to ceramic penetration. Anderson and Walker found in their research that the time during impact when the penetration starts is largely independent of impact velocity for thin ceramic targets impacted by small arms projectiles with a velocity range of several hundred meters per second. However, the time duration of dwell is dependent upon tile thickness (as well as the geometric and material properties of the substrate material).37 By analyzing the failed material of the projectile after impact they concluded that it consists of extremely small particulates and, as a result, they referred to the dwell failure mechanism as erosion. CO Lundberg P and Lundberg B studied the influence of ceramic characteristics in increasing the threshold velocity for interface defeat. They found that the fracture toughness of the ceramic has more influence on the above velocity than the ceramic hardness. That led them to conclude that by increasing the confining pressure of the ceramic target, the initiation and propagation of cracks is suppressed, increasing the dwell * co . , to penetration transition velocities. They included in their conclusions the possibility that "'there exists a unique transition velocity for each combination of projectile, target CD material and target configuration". Holmquist and Johnson5 have done a more in-depth research on the confining effect of the ceramic impacted by a projectile. They tested thin and thick ceramic targets at two prestressed levels (small and large) and two prestressed states (radial and isostatic). Their results demonstrated that apart of the type of pre-stress, the ballistic performance of both targets was improved by pre-stressing. In prestressed thick targets, the velocity at which ceramic perforation occurs (ballistic limit) can be increased and the 71 penetration velocity can be reduced.^ For thin targets, the pre-stressing delays tensile failure at the rear face of the ceramic, therefore the projectile interacts with intact (stronger) ceramic for a longer time, resulting in improved ballistic performance.39 Their experimental results are yet to be backed by a complete understanding of the stress state that occurs from pre-stressing during impact, understanding needed for a more scientific approach towards a comprehensive ballistic improvement. The effect of hydrostatic pressure on the ballistic performance has been described by Pickup60 as well. At the end of his study, he concluded that if the hydrostatic pressure is relieved, possibly by insufficient constraint or release waves reflected from low impedance boundaries, the ceramic strength is greatly reduced which may terminate or even prevent dwell.60 2. There is a certain dose of uncertainty with regards to the crack formation in ceramics, due to the peculiarities of the material and the dynamic interaction with the projectile. As most authors agree, most brittle solids contain inhomogeneities such as small holes, cracks or phases which have different moduli or strength from those of the matrix. When a brittle material is subjected to a large confining stress, any of these inhomogeneities can act as nuclei for new cracks. These micro-cracks eventually coalesce to cause axial splitting. In other words, for elastic-brittle materials subjected to impact loading, some micro-cracks can nucleate and grow under tensile stress even though the loading is compressive.36 72 Meyers1 , in his book on dynamic behaviour of materials enumerates the mechanisms through which the microstructural inhomogeneities create conditions for tensile stresses under compressive loading, namely: 1. Spherical voids subjected to compression generate tensile stresses. 2. The shear stress due to the applied compressive load generates tensile stresses at the extremities of an elliptical flaw, eventually cracking it. 3. Elastic anisotropy of a polycrystalline ceramic aggregate leads to incompatibility stresses at the boundary. Under compressive loading plastic deformation can occur. When the stress is released, localised tension regions created by the residual stresses due to plastic deformation can open up cracks.19 Another peculiarity of a dynamic fracture is that, as mentioned earlier, there is a limiting velocity of propagating cracks, usually accepted as the Rayleigh wave velocity19. As a consequence, at a certain critical velocity, cracks tend to branch out (bifurcate) in order to decrease the overall energy of the system. Thus, a quasi-static failure will result in the propagation of a single crack, whereas dynamic crack propagation can produce fragmentation. As Meyers19 also remarked, the fracture toughness of materials is often dependent on the rate of crack propagation. Due to the complexity in establishing the stress intensity factor at the tip of a running crack, the fracture toughness determination is not as straight forward as under quasi-static conditions. However, in the comminuted zone in front of the penetrator, also referred to as Mescall zone by Stepp l, there is no clear tensile stress, but a rather a large hydrostatic stress. He considers that the effect of 'crushing' or 73 'pulverizing' of the material ahead of the penetrator might be explained and evaluated when there is a degree of plasticity within the material. According to Hazeir5, the plasticity in ceramics has been noticed by several authors and is generally revealed by the process of twinning and / or slip within the crystalline structure. He suggested that twinning and slip within individual grains and plasticity within interfacial materials give rise to the nucleation of intergranular micro-cracks, which eventually coalesce causing failure of the material. 3 Hazell concluded that, due to the complexity of the deformation at the granular level, this physical process is practically impossible to be accurately simulated33. In agreement with him, St-Denis44 observed that the fracture of the ceramic in front of the projectile is variable, indicating the statistical character of the comminution process. Hazell and Iremonger5 add that the crushing process can be augmented by the arrival of a tensile failure wave from the lateral surfaces of the target, if the target is thick enough. Lundberg62 studied the transition from interface defeat to penetration in thick ceramic targets under normal impact. He relates this transition to "the collapse of a highly constrained domain in the ceramic material in front of the projectile". Similar to previously nominated authors, Lundberg considers that in this domain, the pressure and shear stresses are so large that massive micro-mechanical damage and possibly plastic yield occur. It is understood that the size and shape of the micro-mechanical damage depend on the surface load and its distribution and as a consequence, on the density, strength and velocity of the projectile. As the surface load also depends on the shape of the projectile, it can be expected that the transition velocities will be different for projectiles with different geometries or materials. 74 Although the resistance to comminution is an important factor in the ceramic penetration, the ability of the penetrator to move through the comminuted ceramic particles is maybe more important for the ballistic performance of confined ceramics, as underlined by Stepp . For the projectile to advance, it is required that the fine ceramic fragments in front of the penetrator move away from its path. As a consequence, material properties, such as dynamic compressive failure energy, friction and flow and abrasive properties of the finely fragmented ceramic material govern the resistance of ceramic targets to penetration, as Sarva remarked. This movement has to be done through an ejection process that clears the pulverised ceramic away to accommodate the penetration of the projectile. According to the same author, initially the ceramic powder is ejected from the impact face and during the later stages of the penetration event it is predominantly ejected from the rear face. Implicitly, a considerable amount of the energy transferred to the target is converted to the kinetic energy of the ejecta. During the ejection process, in the presence of the high pressure around the projectile-target interaction area, the small unevenly shaped particles can lead to erosion of the projectile and resistance to penetration as observed by Hazell and Iremonger52. The erosion is increased by the large pressure generated by the impact and advancement of the projectile which creates "stiffness in the comminuted material"44. This erosion of the projectile and resistance to penetration is strongly affected by the confining pressure in the zone around the penetrator, the initial strength of the ceramic and the rate of damage accumulation of the ceramic according to Pickup31 as well as by the loading rate. Zhai Jun suggests that the energy release rate increases with loading rate due to the inertial effect37. 75 3. The "shatter-gap" is one important abnormality of a ceramic target subjected to ballistic loading. Moynihan discovered that for B4C ceramic targets impacted by WC core projectiles it is possible to obtain complete penetrations at low velocities followed by partial penetrations at higher velocities. At still higher velocities, complete penetrations can be obtained. He recovered the projectile fragments as well as the B4C fragments after impact. After post-mortem examination Moynihan concluded that the shatter gap is apparently not caused by a change in the physical state of the core during penetration. On the contrary, it appears to be related to a change in B4C failure at high impact velocities. Moynihan labelled as "shattered" the impacts resulting in a greater mass of small fragments which appear to occur at higher impact velocities. The impact with a lesser mass of small fragments, labelled "normal" appear to coincide with lower impact velocities. 4 Moynihan described the term shatter gap as a situation where there are two or more impact velocities where there is a transition from partial penetration to complete penetration with increasing impact velocity . Moynihan's findings have been confirmed by Robertson , who, in researching the same projectile-ceramic pair, concluded that in the impact velocity interval 800 to 880 m/s the boron carbide penetration behaviour is extremely inconsistent from shot to shot. Following previous authors' research on the same subject, Robertson and Gotts1 admitted that the shatter-gap effect could easily surprise an unwary armour designer. The existence of this phenomenon of paramount importance is taken into account during the design of new armours. It is generally acknowledged that the shatter-gap phenomenon is exhibited by a few projectile/armour material interactions, being most common with ceramic armour systems. The shatter-gap phenomenon is defined as being the event when 76 the projectile core is shattered and defeated by the armour when impacted at relatively high velocities. At lower velocities, the projectile could defeat the armour because of insufficient impact energy to break the projectile core. As a direct result, some projectile/armour combinations can have multiple ballistic limit values, also called V50, namely the velocity at which 50% of the impacting projectiles will defeat the armour. The danger associated with it is that one armour manufacturer can design and test a new product concluding that it can withstand a certain velocity of a test projectile, while, at lower velocity, the same armour can be defeated by the same projectile. The STANAG recommends that when a shatter gap is suspected, the test procedure should cover all the possibilities of low velocity penetration. A generic graph that can be obtained during the testing of a pair projectile- ceramic exhibiting a shatter-gap phenomenon is presented in Figure 19. 100 I* 75 !5 re £2 I 50 c .2 I 25 c 0 650 700 750 800 850 900 950 Projectile Velocity (m/s) Figure 19: A generic graph exhibiting the shatter-gap phenomenon2. 77 3.3.4 Armour Ceramic Improvements More effective projectiles with higher penetration capabilities appearing on the market demand improvements for target protection. The armour has to be improved and this is the main goal that currently fuels much of the research in this field. Some authors consider that the improvement of armours reached its limit and only minimum fine- tuning is achievable, even for the newer types of armours such as ceramic. LaSalvia , for instance considers that there are not clear and substantiated ideas on how to improve ballistic performance of ceramic materials. In our research, the armour improvement suggestions are focused solely on improving the ceramic component of the system. The empirical rules established by Mark Wilkins in 1986 at Lawrence Livermore National Laboratory as enounced by Skaggs are still referred to by the armour designers nowadays. These rules are: 1. the armour needs to be at least as hard as the bullet to stop it and the bullet has to be harder than the armour to defeat it, thus the reason for tungsten carbide core bullets; 2. the armour layer needs to be at least Vi the bullet diameter in thickness to successfully defeat the bullet; 3. the front ceramic needs to be about 1/3 of the thickness of the total plate and the backing needs to be about 2/3 of the total; 4. the basic characteristics of the composite plate are that the front armour material needs to break and rotate the bullet and the backing material needs to be ductile to catch the penetrating products without failing; 78 5. understanding the fracture / cracking / break-up of the ceramic is only important for thick ceramics. The ceramic materials are prized for having a lower areal density compared to other armour materials for the same level of protection, which makes them the first choice when the weight of the armour might be a concern. For this reason, the main improvement is focused on decreasing its mass or increasing its performance at the same mass. To achieve this goal, there are several directions pursued by the major manufacturers. The energy dissipation is being improved through mixes of two or more ceramic phases or by new methods of manufacturing. The size distribution of these phases is being optimised either at the beginning of the manufacturing process or in the final product. Medvedovski2" considers that ceramics with multiple phases of the same component may demonstrate increased ballistic energy dissipation. Some more ductile materials are added to the ceramic mixtures in order to increase their "plasticity" and decrease the shattering. Various ceramic grains are embedded in other armour materials to make use of their hardness. Ceramic materials can be reinforced by whiskers or nanotubes, resulting in toughened ceramics, capable to withstand more hits on a smaller area. Another major objective is the cost reduction for the finite product. The raw materials can add up the costs substantially, therefore various by-products or recycled ceramics from other products along with non-traditional materials are included in future research. The manufacturing process might be extremely costly as well; therefore any way to reduce it is being taken into consideration. The development of new manufacturing 79 processes or the use of inexpensive processes as well as an increase in productivity can reduce that cost. The change in manufacturing process is not an easy task. After each change, the new products have to be extensively tested in order to confirm their expected performance. There have been situations when the change in the manufacturing process resulted in a reduction in performance for the newer ceramics, as mentioned by Robertson \ Another important part in the manufacturing process is the struggle to keep the geometric tolerances of the finished product within certain limits. A less expensive process might not be that accurate, therefore an additional step to remove the excess surface material is usually needed. The opinions on either this polishing process will affect the performance of the ceramic are contradictory. Some authors such as Baraett13 consider that the surface finish may change the behaviour of the surface material so that it is effectively a different material, since he argues that the response of a ceramic specimen may depend very sensitively on the "micro-geometry" of its surface. Bryn James66, on the same note considers that the firing process for ceramic tiles often has an advantageous effect on the surface, by leaving a relatively flaw free layer with a residual compressive stress which tends to increase the ballistic performance of the tile (up to approx 5%). The benefit of keeping the strict geometrical tolerances for further assembling by machining away this external layer may be offset by replacing it with a surface microscopically damaged by the machining process. Other authors such as St-Denis affirm that the investigation in the polished versus unpolished surfaces, indicates that the surfaces flaws are not the controlling mode of fracture for ballistic impact, therefore not affecting the ballistic properties of the ceramic. 80 Luckily, most of the other physical characteristics affected by the manufacturing changes are well studied and do not generate any conflicting affirmations. A good example is the reduction of porosity, which improves the ceramic's ballistic efficiency. Krell and Strassburger 7 added the minimization of residual porosity on the list of ways to design better ceramic armour, along with decreasing the grain size (significantly below 1 urn) and increasing the purity of the raw materials. Arguably, one of the less expensive modifications that can be done to a ceramic in order to improve its ballistic stopping capabilities is its geometry. By changing the shape, the size or some surfaces of the tile, its ballistic behaviour might be improved. Testing will always be needed to confirm that those modifications will actually lead to the expected optimization. A trivial example of why a geometric variation needs to be studied in order to provide an optimum result is the tile thickness variation. St-Denis acknowledges that the tile thickness must be increased to an optimal value for the given threat, beyond which any additional thickness benefit is offset by the relative weight increase. In other words, the fine-tuning of the ceramic thickness has to be done in accordance with the threat to be expected, and the lowest thickness is recommended to be used. The lesson to be learned from these tests is that armour should be customized to protect against the most likely actual threat, though this may be difficult to define in some instances. The further reduction of casualties by higher levels of protection may not be significant and, at longer ranges and after ricochets, higher than intended levels of threat may be stopped, according to Iremonger and Gotts . 81 Another example given by Bryn James66 is related to the surface covered by a single tile. He suggests that a larger tile performs worse than a small tile due to residual stress in the material following its firing at high temperatures. On the other hand, he also admits that the shock wave, propagating in a ceramic tile, created by the impact of the projectile can induce tensile cracking at the edge of the tile as illustrated in Figure 20. Figure 20: Edge cracking of a ceramic tile due to wave reflections . Hazell and Iremonger confirmed this finding by observing in their research that by reducing the lateral dimensions of a ceramic tile, the time of dwell was shortened, indicating that greater fragmentation had occurred ahead of the penetrator. 82 St-Denis suggested that in order to avoid the tile edge damage due to the reflected shock waves, a ratio of tile width (D) to projectile diameter (d) should be greater than 20 (D/d>20). This ratio or higher will lead to a so called "self confinement condition ',44. Other out of ordinary studies concentrated on various other geometric modifications than just the tile thickness. One of the already mentioned authors, namely Bryn James, in another study , investigated a number of edge profiles when assembling the ceramic tiles into panels. He concluded that the most successful profile in reducing the damaging reflected waves was shown to be a 45° chamfer of the tile edge. That particular chamfered edge apparently provided a weight saving of 30% of the ceramic mass, for a specified protection level. It is worth mentioning that the edges had the same thickness as the rest of the tile. The configurations tested in this study are presented in Figure 21, the most successful profile being the third from the top. 83 LAT< LAT SAW S LAT 45 LAT 30< LAT ID Figure 21: Configurations used in edge optimization evaluation by Bryn James . Parameshwaran68 performed a fundamental study on the addition of geometric features to armour and the way these modifications alter the nature and extent of damage compared to conventional armour. In the experimental setup he replaced the ceramics with a polycarbonate type polymer, which is easier to machine than ceramics, glued to a back plate. The geometric features were either cylindrical voids or impact surface variations. At the end of the study he concluded that the incorporation of geometric features into the front plate of a composite armour is a "viable concept for improving the ballistic performance of composite armor without any additional weight penalties" . As a collateral finding, he mentioned that the load intensity on the back plate is dictated 84 mainly by the fracture conoid formed at the impact, the inside of the conoid being confined and fragmented / crushed by the impact and displaced towards the back plate. Probably the only ceramic armour available on the market that incorporates geometric features on the impact surface is the CERAMOR-MAP® (CERAmic arMOR Modular Advanced Protection69) produced by ACERAM - Materials and Technologies Inc., a subsidiary of Morganite Canada Corporation, Canadian Operations of Morgan Crucible PLC. Based on the fact that the ballistic performance of a ceramic tile is increased with the angle of incidence of the incoming projectile to the surface of the ceramic, the MAP design is based on an array of nodes (bumps) on the strike face that create deflective surfaces improving the dual role for the ceramic to deflect and defeat the projectiles, according to Lucuta . It is believed that the system acts in two ways. First the projectile is deflected at the impact due to the high angle of incidence, increasing its impacting area and then the MAP design "reduces the conoid propagation as interferences from the nodes in the propagation are assumed" . Apparently, the array of hemispherical nodes (or bumps) on the ceramic surface has been documented to increase the mass efficiency of the ceramic by as much as 30%"70. Another method of improving the ceramic's ballistic performance is by adding in the raw material of other elements, essentially creating a composite, ceramic based material. One of the most appealing ceramic composite contains fibres that act as reinforcements. Meyers estimates that by adding the fibres, the fracture toughness of the material can be increased by two to three folds. His theory is that the fibres can act like crack arresters or, by sliding and detaching themselves from the matrix, as an additional 85 fracture barrier. He believes that when a ductile fibre is added, its plastic deformation requires work and enhances the toughness of the ceramic.19 In commenting on reinforcing the ceramics with fibres and whiskers, Robertson and Gotts11 consider that the practice "has been generally found to be detrimental to armour performance"11. They are still optimistic that in the future, the use of nano-sized reinforcements and other nanotechnology applications might improve the mechanical reliability and behaviour of ceramics, this approach offering the best prospect for performance improvement. However, they admit that due to the high relating costs, the process is yet to be a commercially viable solution . The adding to the ceramic of a second phase that undergoes a dilatational / shear phase transformation when subjected to stress is considered by Meyers19 to be another feasible fracture toughness enhancement. When the ceramic cracks, this addition acts at the crack tip decreasing the stresses and creating a "process" zone, increasing the toughness of the material. As an example, he points out that the addition of partially stabilised zirconia to alumina doubles its fracture toughness. The addition of the secondary material in the ceramic, if done in layers can create so called "stop bands" in which wave propagation is forbidden or, at least attenuated, thus slowing down the velocity of damage propagation in the ceramic under high velocity impact and increasing the target resistance, according to McCuiston40, Yadav and Ravichandran30. Yadav and Ravichandran studied laminated structures composed of ceramics and polymers that yielded promising results following ballistic impacts. They concluded that the enhanced penetration resistance of the laminated ceramic / polymer structure is given by three mechanisms, namely: 86 1. attenuation of the stress waves due to the periodicity of the laminated structure, 2. reduced damage velocity in the laminated structure, and 3. the crack arresting feature of the polymer layer. Strassburger71 studied the impact of AP projectiles against laminated ceramics as well, and compared the results against monolithic ceramics with the same thicknesses. He used the same type of ceramic for the two layers that comprised the laminated target, as well as for the monolithic target, but varied the thicknesses of each layer for the first case. Strassburger concluded that the lamination process can increase significantly the ballistic resistance of the target against both steel and tungsten core projectiles, if the front ceramic layer is thinner in relation to the rear one. He explained that, when the projectile first hits a thin front layer, this layer will be fragmented very rapidly and will exhibit only a low ballistic resistance. However, pre-damage to the second plate will be reduced so that the already damaged projectile has to penetrate a material with higher ballistic resistance compared to the case of a monolithic target. When the thick plate is at the front, the penetration of this plate is like that in a monolithic target. However, the thin plate at the back will be shattered by wave reflections before the penetration can occur.71 A variation of laminated ceramics, already produced by several companies is the covering of the front face of a ceramic with another material such as powders, polymers or polymer matrix composites, or fibre reinforced polymers. Nunn , Sarva and Nemat-Nasser studied the use of the fibre reinforced polymers on the front of ceramic armours and the ballistic behaviour of the resulting laminated structure. They observed a significant improvement that ranged from 25% to 87 40%, against AP projectiles, with an increase in areal density of only about 3%63. Sarva63 concluded that the front face restraint accounts for most of the improvement, while the back face restraint, if exists, has only a modest contribution. He found that there is a critical thickness of the front cover, in his research being three composite layers, a further increase in its thickness not generating any significant performance enhancement. Nemat-Nasser73 used flash radiography to see the interaction between the projectile and the target during the impact. He came to the well accepted conclusion that constraining the ceramic tile will result in a much greater mushrooming and erosion of the projectile. The greater erosion and reduced velocity of the projectile are also manifested in the form of a significant delay in the back-face displacement of the ceramic tile.73 By using high speed photography Nemat-Nasser discovered that the front face confinement of the ceramic tiles studied alters significantly the flow of the pulverised ceramic that is ejected. He found that the front-face ejecta from a bare tile are radially dispersed and conical. For a constrained tile the ejecta flow is more acute and cylindrical. Additionally, the ejecta velocity for a constrained tile is nearly 40% higher.73 Although impressive, the improvements in ballistic impact performance of the ceramics covered with fibre reinforced composites cannot be fully explained according to most authors. Nunn72 speculates that the composite layers act to delay the onset of fracture and fragmentation of the ceramic material, but "the actual mechanisms are not clear. The composite layers may provide a lateral constraint on the ceramic tile, which could slow the spread of cracks and the separation of tile fragments".72 He also suggested the possibility that the cover may provide a form of acoustical damping that affects the propagation of stress waves in the ceramic tile resulting in delayed fracture. 88 A similar increase in performance seems to be obtained by using only a non reinforced polymer cover according to Lucuta7 . The system containing the polymeric protection of the ceramic strike face is patented and manufactured by ACERAM - Materials and Technologies Inc. TX 7Q 77 As Nemat-Nasser , Normadia , Kaufmann and other authors remarked, the added compression on ceramic during impact and penetration is a good way to increase the time it erodes the projectile and, implicitly to increase its efficiency, even in the 7Q 77 highly damaged or comminuted state . Kaufmann considers that ceramics still maintain significant strength after fracture if in compression. The strength of both the intact and fragmented ceramic materials is dependent upon pressure, with strength increasing as the confining pressure increases. As damage accumulates, pressure increases due to bulking27 or compacting. He admits though that their response to high pressure and strain rate is complicated, this being one of the main difficulties in understanding the process. Surprisingly at first sight, in his study on ballistic impact of ceramics, St-Denis44 relays some information from other sources, namely DRDC Valcartier that modelling of the ballistic impact test using a Johnson-Holmquist constitutive model for brittle material led the researchers to the conclusion that "ceramics are sensitive to hydrostatic tensile pressure which develops during impact and radial confinement can be detrimental to the performance" . This theory contradicts the findings of all of the authors of similar studies consulted for this thesis. St-Denis too does not agree with DRDC findings and further in his study declares that the use of lateral confinement is beneficial to the ballistic potential of the ceramic by delaying the damage induced by the lateral tension waves44. Later in his 89 report, St-Denis dials down the importance of the confinement, concluding that the benefit or insignificance of confinement is yet unclear due to evidence presented by various authors in support of either condition . As a general conclusion on armour improvement, one has to admit that the conflicts around the world involving western democracies as well as other states or non- state entities are fuelling the innovations on this topic. In addition, the media coverage of attacks and their detrimental effects on light armoured targets and military personnel add more weight on the importance of research in this field. 3.4 TYPES OF BALLISTIC TESTS Tremendous amounts of funds and time are invested in studying the response of ceramic materials under ballistic impact. The challenge is to a better understand their response and, as a consequence, to make them more attractive and efficient as armour components through future improvements. An interesting approach in this field of study is the "reverse engineering" suggested by Gotts . He thinks that by careful deconstruction of successful armour systems it may be possible to identify the factors that play the main role in the beneficial ballistic properties. The difficulties with this approach reside not only in the carefulness of the deconstruction in order to isolate the important factors, but also in the costs involved by studying entire systems. 90 The most utilized and classic approach in this research, based on the correlation of static properties with ballistic performance proved to be not entirely accurate, as found by many authors such as Flinders49, Adams74, Robertson and Gotts11. Nevertheless, the static properties are the properties that can be easily obtained for each material. A correlation between these static properties and the final impact response of the ceramic would be the desired finding for the armour designer. The effective resistance to penetration (hardness) and dynamic fracture processes (toughness) are the properties that are used the most in studying the armour ceramics according to Robertson and Gotts . They consider though that, whereas ceramic hardness values can be used in most cases as a guide, quasi-static fracture toughness measurements (Kic) are an extremely poor guide to any aspect of ceramic armour performance. They also emphasise the fact that there is a need to measure the dynamic fracture resistance and that the absence of any recognised low cost measurement technique for such a property is according to them, probably the single biggest impediment to the development of improved ceramic armour materials.11 Another important property of the ceramic, as suggested by Horsfall75 is its resistance to shear failure. This has been quantified by him as "the energy dissipated by frictional loses during shear failure under compression." He believes that the magnitude of the energy dissipated in shear failure is a function of the friction coefficient between fracture surfaces which, apparently has been shown to be equal to the ratio of compressive strength to fracture toughness. Compressive, shear and tensile strengths of the ceramics are considered to be the most important factors affecting their behaviour under impact, according to Matsalla . 91 Having obtained the values for the main static properties of ceramics, several authors tried to use them for armour design purposes since the dynamic properties for the same materials are difficult to obtain accurately. As an intermediate step in their research, some scientists proceeded to simulate the behaviour of the same ceramics under impact. The simulation process, if correlated correctly with reality can offer valuable information at lower costs compared to manufacturing and testing the physical products. Some computer applications for high velocity interactions had been used for gaining insight into the physical phenomena which occur and affect the performance of ceramic armours during ballistic events, but to date the results do not seem impressive. One explanation, given by LaSalvia for the limitations of the computational modelling is that the models are simulating the phenomenon and may not incorporate the correct physics. As a consequence, the validity of the results obtained can be subject to question. This scepticism in the results offered by simulation is shared by most of the authors. Matsalla6 admits that there has been little success in modelling the dynamic failure of ceramic materials and therefore most of the knowledge in this field is based on empirical observation. Adams74 considers that the computer simulations are not capable of accurately predicting all the conditions under which a projectile will completely penetrate the armour. He gives credit to the efforts to use discretized, finite element / finite difference modelling or "hydrocode" modelling but underlines the limitation that these analytical tools are not adequate to design armour systems or to adequately predict material performance. He believes that they simply don't describe, with sufficient accuracy, all of the important phenomena that affect performance. Adams concluded that the basic dynamic material behaviour, under the conditions of projectile impact and 92 penetration, is still imperfectly understood and therefore, as Hazell confirms, "accurate modelling of the failure of these brittle materials under dynamic loading conditions still remains an illusive goal. " " As a consequence, the testing of ceramic materials under conditions similar to their use remains the main method of evaluating and improving their performance. These tests have their limitations as well. Some results measured in a certain test configuration might not necessarily be extrapolated to the real impact in the field. In the attempts to save some money during the testing phase, an unsuspecting researcher, testing only parts of the entire armour system, might fail to take into account some variables that prove to be important in the behaviour of the final product. As an example, Normandia mentions the attempts to use data generated from sphere impacts to rank candidate armour materials. He concluded that in very light systems that rely on the fracturing or shattering of the penetrator, the results from testing do not correlate well with the final systems' behaviour, while for confined ceramics systems relying more on the erosion of the penetrator, the results are more encouraging. As a lesson learned by many researchers on ceramic armours, it seems that the relationship between target design, target areal density, impact velocity and defeat of the projectile are complex. Therefore, it is best to measure, directly, the principal variable of interest and hold all other variables as constant as possible, as suggested by Adams74. During the last several years, a series of tests were used and eventually standardised for evaluating the ballistic behaviour of the ceramics and other armour components. Normandia and Gooch41, in one of their published article summarised these tests in Table 5. 93 Table 5: Ceramic Material Evaluation Summary of Ballistic Test Methods41: TEST TEST TYPE INFORMA TION OBTAINED NDP Non-Deforming Typically used for soft metals and hard targets, Penetration this applies for concrete, limestone and other geological materials. Various researchers attempt to isolate target resistance in this penetration mode. PEN Penetration Depth Penetration-velocity curves, penetration Direct or Reverse resistance, penetration rate, penetrator Impact consumption rate. DO? Modified Depth-of Relevant for determination of performance goals Penetration as a function of ceramic thickness — similar to TAD, but in a semi-infinite configuration. DWE Dwell Tests Total interface defeat conditions. DPT Dwell / Penetration Velocity defines a load that is characteristic of a Transition failure shear strength of the ceramic, or of a transition strain. FTG Fixed Target Geometry Generic material comparison experiment in armor-like configurations, particularly at obliquity. TCA Tandem Composite Configuration to minimize the use of damaged Armour material. VBL Ballistic Limit Velocity Typical requirement for acceptable armor, (V50) or Residual Data individual tests measure residual penetrator characteristics. BAD Behind Armour Debris Used to measure the lethality of the penetrator or the vulnerability of the target to an overmatched threat. Data quantification utilized in lethality assessment tools. TAD Target Areal Density Helps determine near-optimal armor Performance Maps configurations Theories permit extrapolation to different threats. Most of the tests in Table 5, such as DWE, DPT, FTG, BAD and TAD are pertaining to entire armour systems rather than individual elements. For example, the test 94 preferred during the final stages of the armour development is the Ballistic Limit Velocity (V50) in which the final armour is impacted with a standard projectile at different velocities until the limit velocity is found (50% of the projectiles at that velocity did not penetrate the armour). By varying the thickness of the ceramic or other parameters, a ballistic limit curve can be obtained for that particular projectile-armour pair, but testing costs are high. In order to reduce the high costs, some tests in the early stages of a ceramic based armour development are required. 3.5 DOP TEST Arguably one of the preferred tests in the early stages of armour manufacturing is the Depth of Penetration (DOP) in which the ceramic or other armour materials are backed by a semi-infinite material. The impacting projectile is designed to overmatch the ceramic material and its residual forward momentum is to be stopped in the semi-infinite backing. A DOP value is obtained by measuring the depth at which the projectile embedded itself in the backing material with and without protection. By comparing the penetration of a projectile into a ceramic protected target and an unprotected target, the ceramic's ballistic performance may be determined. According to Darin Ray76, another way to compare this ballistic data is to measure the initial diameter of the hole in the backing material. As an alternative to measuring the depth in the backing, especially when X-ray or other expensive equipment might be needed, he considers that when the depth of penetration is shallow, the harder ceramic 95 has spread the bullet fragments and ceramic debris that penetrates into the backing compared to the smaller initial diameter when a softer ceramic is used. Several authors published materials relating to DOP tests performed on armour ceramics. They include: Flinders49, Kaufmann27, Gotts21, Horsfall75, Gilde77, Adams77'74, Hazell10'78'79, Robertson20'78'79, Madhu47, Lopez-Puente43, Yadav and Ravichandran30, St-Denis44, Peron80, Gooch81, Murat82, Bryn James66, Krell and Strassburger67. It is generally agreed that the DOP testing doesn't offer the final answer on formulating the best armour design; it is still an excellent screening tool for the possible ceramic materials, in which a large amount of data can be collected quickly and relatively inexpensively from a minimal amount of samples. The ranking obtained in the DOP by the tested materials might not be entirely mirrored in the final design; however Robertson78 considers that materials failing to impress in terms of performance against the particular test projectile in DOP regime will most of the time perform poorly in the practical armour system. The data collected during the DOP tests does not usually include stress, strain or damage accumulation to be used in future simulations. However, the DOP testing is a practical and highly effective method of assessing relative ceramic performance. This type of testing is highly efficient in terms of materials use and time. In other words, the DOP testing is an expedient screening test to determine which materials are worthy of a more extensive evaluation. One must be cautious in comparing different sets of data from DOP, because they may vary with the specific method and materials used in the assembly of samples for 96 ballistic testing. The data must be taken into account only as a comparative ranking rather than combined for a more global evaluation of the armour ceramics on the market today. In order to compare the results obtained in DOP tests, several formulae were proposed by several authors. Flinders4 and Darin Ray76 suggested a mass efficiency (Em) based comparison using the following equation: E = Pbac boa (Equation 21) PcJcer + PbacDOP where: Em = the mass efficiency Pbac= the density of the backing material pcer = the density of the ceramic Pbac= the penetration depth into the unprotected backing material tcer = the thickness of the ceramic target DOP = depth of penetration into backing after striking the ceramic target. In the above equation, the mass efficiency increases with decreasing DOP for a given set of materials and ceramic thickness, until the ceramic completely defeats the projectile. At the same time, increasing the thickness of the ceramic target increases the areal density and decreases the efficiency. A variation of the equation 21 for calculating the ballistic performance of a ceramic in following DOP tests is suggested by Peron80, Madhu47, Gilde and Adams77. They use a formula to determine an equivalent thickness or thickness efficiency (Eeq), another formula to obtain an equivalent mass or mass efficiency (Meq) and then, by 97 combining them, they determine a quality factor or ballistic efficiency factor (q ). These formulae are: £ = -J™ £2- (Equation 22) A/f = boc res ., r-^bac _ r ., A^o 1 1eq ~ rp eq (Equation 23) cer r"cer rce 2 q = Meq x Eeq (Equation 24) where: Eeq = the equivalent thickness or thickness efficiency Pbac= the projectile penetration in the semi infinite backing Pres = the residual penetration of the projectile in the backing after penetrating the ceramic tile = Tcer the ceramic thickness Meq = the equivalent mass or mass efficiency pbac= the density of the backing pcer = the density of the ceramic q2 = the quality factor or ballistic efficiency factor The equivalent thickness or the thickness efficiency is defined as the thickness of the backing material, i.e., the PC replaced by one thickness unit of the ceramic material and the equivalent mass or the mass efficiency represents the mass of the backing material, replaced by one mass unit of the ceramic material to obtain the same effect on the projectile. 98 It can be seen that Eeq and Meq are dimensionless factors that compare the ballistic performance of the ceramic to the backing material. The values of Eeq and Meq for the backing material are unity. A number higher than unity denotes better ballistic performance of a material when compared to the reference backing material. 77 7 According to Gilde and Adams' opinion, the armour quality factor, q , is important to armour designers because it relates both the mass and thickness or space efficiencies, since both the weight of the armour and the space it takes up are critical factors in designing armours. Another name for the mass efficiency Meq equation used in DOP testing has been R7 suggested by Murat . He names it ballistic efficiency (r\) of a ceramic tile: P JJ = Pbac( bac ~ Pres) (Equation 25) r cer cer where: n = the ballistic efficiency of ceramic tiles pbac= the density of the backing Peer= the density of the ceramic tcer= the ceramic thickness (Pbac - Pres) = the reduction in backing thickness penetrated due to ceramic tile in place, i.e. the difference between the reference depth and the residual depth. According to Krell and Strassburger67, usually, a linear decrease of the residual penetration is observed when the ceramic thickness increases resulting in a linear increase of the mass efficiency Em with increased values of Tcer. From such plots, a linear extrapolation of Em to a ceramic thickness which would stop the projectile just at the 99 ceramic-backing interface (named critical thickness) defines the maximum mass efficiency Em,max as a characteristic material parameter. In one of his articles, Horsfall75 recommended a direct formula for calculating the critical thickness (tent) of a ceramic to just defeat the projectile, in a DOP test setting as: er tcrit = *' ^ (Equation 26) bac res where: tcrit = the critical thickness of the ceramic Pbac - the DOP in the backing material only tcer = the thickness of the ceramic tile tested Pres = the residual DOP when the ceramic tile is used. The above equations (equations 21 to 26) constitute the basic set of tools for evaluating the ceramics during DOP tests. They provide enough information to the researcher in order to choose the ceramics worth being kept in the competition for a more in-depth research. The DOP test is considered the test of choice in screening ceramics due to its relevance, low costs, simplicity of the test and convenient data analysis. 100 CHAPTER 4: EXPERIMENTAL TESTING The experimental testing program consisted of the manufacturing of a drop weight impact apparatus with the purpose of testing armour components o RMC premises. Subsequently, a depth of penetration (DOP) testing program has been finalized in a research facility. The details of these parts of the experimental work are presented in this chapter. 4.1 PRELIMINARY WORK - DROP TOWER IMPACT APPARATUS Due to cost constrains and the availability of materials, it was initially decided to build an impacting system for armour materials that can be used on RMC premises. A gas gun was the first option for impact testing, since there is already one available in the Mechanical Engineering Department. This gas gun could have been modified for the present purpose. It uses compressed helium fed into the barrel by a valve controlled by a solenoid. The regular projectile, approximately 10 mm in diameter is made of aluminium, with flat front and hollow back. The gun is muzzle loaded; hence the 101 tolerances of the projectile diameter have to be extremely tight. The maximum projectile velocity is about 100 m/s83. After calculating the maximum kinetic energy of the projectile and the necessary modifications to increase it, it became evident that the use of the Mechanical Engineering gas gun is not a viable method for studying the impact on armour materials. A second option taken into account was to modify a drop tower used for testing body armour to stab resistance. It is installed in the testing area of the Mechanical Engineering Department. The tower consists of a tube about 5 metres high, with an inner square section, roughly 90 mm x 90 mm. It can accommodate a missile with detachable blade tips, held by an electro-magnet. The electro-magnet is raised and lowered inside the tube with the aid of a surveyor's tape and a pulley. At the base of the tube there are two magnetic velocity detectors, 50 mm apart. When the missile crosses in front of each detector, a signal is sent to an oscilloscope and a time is determined from the delay between the first and the second signal. Under the tube there is a table on which the testing samples are usually installed. For blunt impact testing, the stab resistance impacting missile was not suitable anymore, being too light and having a sharp tip. Another missile had to be designed and manufactured, in accordance with international standards suitable for this kind of tests. The most suitable standard for measuring the damage resistance of a material following a drop impact event has been proven to be ASTM D 7136/D 7136M - 0584. It defines a standard impactor with a blunt, hemispherical striker tip. This tip generates more damage compared with a sharp striker tip. The blunt shape is arguably in accordance with the idea that in a high velocity impact, the tip of the projectile is deformed before starting to 102 penetrate the armour. The mass of the impactor was required to be 5.5 ± 0.25 kg and the smooth hemispherical striker tip had to have a diameter of 16 ± 0.1 mm and a hardness of 60 to 62 HRC, as specified in the standard. The impactor, manufactured in house in accordance with the ASTM indications is presented in Figure 22. Figure 22: Impactor manufactured in house. The lateral metallic rim has the role to trigger the velocity sensors and the back disk is to be in contact with the electro-magnet for holding and raising the impactor. The impactor's tip is attached to the mass by a threaded cylinder, for easy replacement. A major change has been decided for the holding mechanism and the sample size, compared to the above ASTM specifications. Instead of simply supporting the sample in place, the use of an all contour vice type clamping with bolted frames was considered more appropriate. The size of the sample holder has been modified as well, in order to accommodate samples with the testing area of 300 mm x 300 mm, compared with 75 mm 103 x 75 mm as suggested in the standard. The manufactured frame with a part of a model sample is presented in Figure 23. Figure 23: Frame for drop impact testing. During some preliminary tests and after consulting various bibliographic materials it has been concluded that a drop test does not accurately describes the impact of a projectile, although the kinetic energy might be similar. The damage dynamics are completely different and, as a consequence, the use of this apparatus in simulating a ballistic impact event has been cancelled. 104 Nevertheless, the test would be very useful in evaluating the possible damage of the ceramic based armour during everyday use. The tear and wear of these armours, especially the ones used in the personnel protection equipment might decrease their effectiveness significantly. It is well known that the dropping or rough use, generating small impacts, can result in micro cracks in the ceramic strike face. Micro cracking can or result in a partial or complete ballistic failure under lower level of threats. A study on the resistance of the ceramic armour material to small impacts in the theatre of operations and everyday use could take full advantage of the drop impact apparatus existing at RMC, but this research is beyond the scope of this project. 4.2 DOP EXPERIMENTAL PROGRAM The goal of the experimental program was to screen the potential ceramics for replacing the alumina in use on the CC-130 Hercules aircraft. The program consisted of 86 impacts on ceramic specimens, focusing on the effects of the ceramic material, the thickness of the ceramic (areal density), porosity, and use of spall cover, diamond coating and the shape of the impact surface. The DOP experiments have been performed at one of the most suitable facility in Canada for this work, namely the Munitions Experimental Test Centre (METC) in Valcartier Quebec. Before the testing could begin, a test plan for the DOP assessment has been put in place. Some of the decisions that shaped the test plan are presented in this chapter. 105 4.2.1 Protection Level / Projectile Selection From STANAG 45692, a level III kinetic energy threat has been selected. This corresponds to a 7.62 mm calibre rifle with AP projectiles, the most probable threat an aircraft may encounter in the theatre of operations at landing and take off. The chosen threat level must defeat the ceramic material. As a consequence, there should be a residual penetration to permit the evaluation of the ceramic materials tested. In the available revision of the STANAG 45692, the ammunition accepted for the level III threat is 7.62 x 51 AP and 7.62 x 54R B32 API, ammunition presented in Table 6. Since the latest is manufactured by Barnaul Machine Tool of CIS Russia and used by the former Soviet countries with Dragunov type sniper rifles, it is more difficult to be purchased and tested; therefore the first option has been selected. Table 6: The STANAG 4569 accepted ammunition for level III threat2. Ammunition Supplier Specific test Test Projectile ammunitions Velocity Mass (m/s) (grams) Bofors Carl FFV AP M993 WC 930 8.2 7,62 mm x 51 Gustaf core AP Nammo AP8, 8.4 g W alloy 870 10.5 Lapua core Barnaul 7,62 mmx Machine B32API7N13 854 54R Tool in CIS Russia The chosen 7.62 mm x 51 AP is the one manufactured by Bofors Carl Gustaf. The mass of the FFV projectile is 8.2 g compared with the mass of the AP8 projectile 106 which is 10.5 g as found in Jane's Ammunition Handbook . The mass of the projectiles' core stated in the STANAG 4569 is 8.4 g for both ammunitions, which seems to be inaccurate. Nevertheless, they are well established munitions widely fielded by several Western armies. This fact increases the chances of reproducing the present tests. The bullet core, described by Moynihan64 , Robertson and Hazell78 has a mass of 5.96 grams and consists of tungsten carbide (composition by percentage weight C 5.2, W 82.6, Co 10.5, Fe0.41) of hardness 1200 Hv, mounted in a low carbon steel jacket with gilding metal, on an aluminium cup. An excellent description of the projectile's core material is given by Gooch \ He describes the two ceramics of interest formed by tungsten and carbon, namely tungsten monocarbide (WC) and ditungsten carbide (W2C). WC has a very high elastic modulus but the WC materials currently in production are, in fact, cermets. The cermets, by definition are alloys of ceramics and metal binders that are sintered to form a hard dense material. The WC cermet apparently contains 8-10% cobalt by weight, added as a liquid- phase sintering aid to eliminate the porosity in the material and allow it to be fully densified at lower temperatures and pressures as compared to binderless WC. Gooch suggests that the cobalt addition reduces yield strength and hardness, but increases toughness.81 The image of this ammunition and its elements, from the DRDC Valcartier is shown in Figure 24. 107 iMHSkaWUMMU fo 2 4 6 8 10 12 7.ft2mm Natu AP BOIWH Pokfe :127 ^raimiS.iju Nov ;iu Utint^taii' Figure 24: 7.62 x 51 Bofors - Carl Gustaf WC core, AP round. As an observation, in comparing the penetration data of a steel core with a WC core projectile, the differences are significant. While against SiC based ceramics Darin Ray found a difference of 13% in the favour of the WC core bullet, the difference could increase to 60% against steel armour according to Robertson and Hazell78, causing the weight of armour required to defeat such rounds to be significantly increased relative to that necessary to defeat the traditional steel cored AP ammunition. 108 4.2.2 Experimental Set up Apparatus For DOP testing on ceramic tiles, an experimental design has been outlined. The experimental design is explained in the following sections. 4.2.2.1 Backing Probably the most important variable in a DOP design is the backing material. In the literature it has been found that the most widely used materials for DOP serm-infinite backing are steel, aluminium and polycarbonate (PC). There are some advantages and disadvantages for all of them. The steel, usually rolled homogenous armour (RHA) grade is largely used for .50cal and higher AP rounds, since it has a higher stopping capability. It is used in sections, mostly 1 inch thick and clamped in blocks in position. The distance travelled by the core of the AP projectile is measured by counting the totally perforated sections and approximating the depth of the penetration in the section with the core embedded in it. Some authors such as Hazell , and Robertson seem to prefer aluminium rods. The variability due to the "crossing" of gaps between the backing sections is therefore eliminated. The aluminium is softer than the RHA steel allowing a longer path of the projectile core, and consequently, a better discrimination of the results. The downside is that the assessment of the depth of penetration in the large aluminium block is more difficult that in the case of thinner segments. For a precise result, a computerised X-ray or a CT-scan image is required. In addition to this, this backing design might lead to some 109 material being unused or wasted, since it is impossible to prepare the length of the rod to match the penetration for each future shot. Nevertheless, the design is suitable for all small calibre AP rounds. Gotts21 and other authors prefer to use polycarbonate for their DOP backing. Its density is lower than its metal counterparts, permitting a better resolution in comparing the depths of penetration. Among the positive aspects in using this material one can specify that the measurements can be done visually, since the material is transparent, as well as the attractive price and the machining convenience. Considering all of the above aspects, it was decided that the best option for the backing material to be used in the present testing program is the polycarbonate. Relatively long rods were chosen to permit the stopping of the penetrating round in single blocks. Five inches nominal diameter, 8 feet long polycarbonate rods were purchased for these tests. They were cut in 6 inches billets for the simplicity of use. The ends on each billet were straight cut by the author, as seen in Figure 25. 110 PC billet 7 Figure 25: The 5" polycarbonate billet being machined in the lathe. After machining, the billets that were designed to have the ceramic tiles glued on one face were further machined on a numerically controlled milling machine with a hole made in order to accommodate the geometry of the tiles. 4.2.2.2 Clamping System - the Jig The billets had to be held firmly in place during the impact, but also to be replaced in a timely manner between the shots. For that purpose, a jig was designed. It Ill accommodated the use of three billets, each 6 inches (150 mm) long at the same time, to simulate the "semi-infinite" length, although only the first billet was expected to be penetrated during each shot. They were held in place by two C channels (C 4 @ 5.4), one on the top and one on the bottom. The channels were connected between them via twelve double extension bars with threaded rods and nuts. In the back, a welded metallic part acted s a stopper, impeding billets' axial movements during impact. An image of this jig is presented in Figure 26. Figure 26: Clamping system for the DOP targets. The jig was clamped front and back in a horizontal position on a test table that could be raised or lowered to bring the center of the ceramic tile in line with the trajectory of the projectile. 112 4.2.2.3 Ceramic - Polycarbonate Adhesive The adhesive used to glue the ceramic tiles to the backing material could be a major source of variability in DOP testing. This variability comes from two aspects. The first aspect is the impedance of the bonding material, which can be substantially lower compared to the impedance of the ceramic and the backing material. During the impact, this mismatch generates a strong tensile reflection into the ceramic, which acts to shatter the ceramic layer and to provide energy for ejection of the shattered material. In addition to this, a strong shear wave is set up at the interface, which serves to "unzip" the adhesive interface, as Bryn James45'65 observed. He recommended that ideally, the backing should be in contact with the ceramic through a soft material with similar acoustic impedance to both the ceramic and the backing. In the present study, the consistency and some other properties of an epoxy type adhesive have been considered to be similar with the properties of the polycarbonate material. Therefore the mismatch in their impedance might not be significant. The other aspect of the testing variability given by the adhesive is its thickness, as observed by Gooch81 and other authors. Lopez-Puente43 studied the effect of the adhesive layer thickness on the ballistic limit of ceramics, impacted by an AP projectile similar with the one chosen for the present study. He obtained contradicting results and concluded that the efficacy of the armour is affected by three different effects related to the adhesive thickness. The shear stress decreases with the increase in thickness of the adhesive, reducing its possibility of failure. The ceramic spalling is reduced with thin adhesive layers due to low compression distance of the adhesive that prevent bending of 113 the brittle tile. The energy absorption by the backing plate is greater when a thick layer of adhesive is used because of the cushion effect during the load transfer from ceramic to the backing. Gooch , Hazell and Chu Henry recommend that the adhesive thickness should be kept to a minimum in order to eliminate or reduce variability in ballistic performance. For the present tests, a WEST SYSTEM® brand 2-part marine grade epoxy was chosen. It consists of a liquid epoxy resin as its base (105 Resin) and a medium viscosity epoxy curing agent (205 Fast Hardener). The working time for the mixture is 9 to 12 minutes, while the time for curing to a solid state is between 6 and 8 hours, at room temperature. Its maximum strength is achieved in 1 to 4 days, as stated on the WEST SYSTEM® brand website87. The epoxy adhesive chosen was applied to the matting surfaces and then the ceramic was pushed by hand against the polycarbonate block and oscillated by hand until an even thin adhesive surface has been achieved with no gaps or obvious air inclusions. More adhesive material was added if necessary to completely fill the lateral gaps between the side of the ceramic and the polycarbonate block. Since the curing time of the thin layer of the adhesive was several hours at room temperature, the mass of the tile during this time acted in thinning the adhesive layer even further. The excess hardened adhesive was subsequently removed, as shown in Figure 27. 114 Figure 27: The removal of excess adhesive. 4.2.3 Ceramic Samples The ceramic materials to be tested included most of the ones already in use for various armour applications. Our selection included the original alumina from the CC- 130 Hercules aircraft armour panels, boron carbide (B4C) material manufactured by Ceradyne Inc, silicon carbide (SiC) manufactured by Morgan Crucible PLC under the trade name Pure-bite® and CERAMOR® material manufactured by ACERAM - Materials and Technologies Inc. Other potential armour ceramics such as titanium diboride were deemed to have a density that was too high to be included in our research, since the weight of the armour has to be kept to a minimum. 115 All the ceramic materials tested are well known and their mechanical properties are presented in Table 1. The only exception is the CERAMOR® ceramic material, described in a document published by Lucuta2 . Lucuta states that the CERAMOR® material is a new functional oxide based composite which is designed specifically for ballistic applications. According to him, the CERAMOR® ceramic material is a two-phase composite that exhibits a hardness of 17.2 GPa, strength of 492 MPa, and a fracture toughness of 6.9 MPa m1/2; about three times higher that any other ceramic materials. Lucuta considers that the CERAMOR® destroys the armour piercing penetrator in two ways, depending on the material of the penetrator. If the penetrator is made of hard steel, it is eroded during the impact, starting from the tip. The WC penetrator is broken into small fragments by the ceramic. In another document, Lucuta69 presented the mechanical properties of the CERAMOR® material in comparison with the published properties of boron carbide. The material properties, namely hardness, flexural and compressive strength, and fracture toughness were measured on specimens cut from slip cast standard chest plates made of the CERAMOR® ceramic material. Table 7: Mechanical properties of CERAMOR® ceramics compared to boron carbide . Property Units CERAMOR E-CERAMOR B4C Hardness N/mm2 22,000 32,500 31,680 Bending Strength N/mm2 390 390 350-575 Compressive Strength MPa 2420 2730 3050 E-modulus GPa 387 402 327 Fracture Toughness MPaJm1'2 7.2 7.1 2.9 116 The E-CERAMOR® as stated by Lucuta in the same document is essentially the CERAMOR® material with a diamond coating on the strike surface in order to improve the performance of the CERAMOR® systems against steel core armour-piercing projectiles. This solution, at the publishing date of the document (2006) was subject to worldwide patent pending protection. 4.2.3.1 Tiles Minimum Dimensions In terms of the actual tiles tested, most of the tiles were off the shelf products of the manufacturing companies. Since the tiles differed in dimensions from one producer to another, an optimum dimension of the target tile or tiles was decided, to accommodate the supplied materials and also eliminate the lateral edge effects. Bryn James 6 considers that in order for the lateral dimensions to have no influence upon the DOP result, the ceramic tile's smallest lateral dimension should be greater than 30 times the calibre of the projectile for velocity impacts below 1600 m/s. He suggests that if no tiles of satisfactory size are available, the tested tile may be clad in a supportive frame to rnimic a larger tile. In order to fulfil the above condition in testing with 7.62 mm diameter projectiles, a tile with the smallest lateral dimension of over 200 mm would have had to be used, a practically unfeasible solution. St-Denis44 recommends that in order to avoid the edge effect in DOP testing, a tile with a diameter to thickness ratio of minimum 5 is required. This latest requirement could be fulfilled by all the samples experimentally tested. 117 Another condition that had to be taken into consideration during our testing, as suggested by Chu Henry1" in his DOP tests, was that each target to have sufficient impact surface area to maintain a minimum of 38 mm between point of impact and the free edge of the material. The tiles tested during this research project in Valcartier had a minimum lateral dimension that fulfilled this requirement. Some confidence was added by the fact that these tiles were embedded in the polycarbonate, their lateral sides being not free, but relatively constrained. With the exception of the alumina tiles, the distance between the point of impact and the edge for the smallest tiles used in the present tests was 39 mm, enough to proceed with only one tile per test. 4.2.3.2 Alumina Tiles The square alumina tiles were recovered from an old CC-130 Hercules armour panel provided by CFB Trenton. The tiles, 2" x 2" x 0.35" (51 x 51 x 9 mm) have been visually inspected for defects before being glued with the epoxy adhesive to form a larger target, by using halves and corners of the same material, with the purpose of providing sufficient impact surface. The total surface of the alumina target was 14 sq. inches (about 90 sq cm) and its final shape was octagonal, as presented in Figure 28. The epoxy adhesive, although not identical with the original adhesive used in the original armour panels is considered to have similar bonding properties. 118 Figure 28: Alumina ceramic tiles glued for DOP testing The CC-130 Hercules alumina tiles have been tested in two configurations, namely as bare tiles or covered on the strike face with a single ply of woven Kevlar®/epoxy, the latest configuration being used in the actual armour panels on the strike face. 119 4.2.3.3 Boron Carbide Tiles Ceradyne Inc. supplied its boron carbide square tiles in two thicknesses, namely 0.443 in (11.2 mm) and 0.333 in (8.5 mm). The tiles were standard 51x51 mm (2"x2") tiles used by the company in their ballistic protection products. In order to be accommodated by the designed holding system and to have a similar shape with the alumina tiles, the corners of the square tiles have been removed. The final impact area of the boron carbide was similar to the area of the alumina targets. All the cutting of the ceramics, both alumina and boron carbide was performed by the author at the ACERAM - Materials and Technologies Inc. manufacturing facility in Kingston ON, using a diamond circular blade saw. A visual inspection for cracks followed the cutting. 4.2.3.4 CERAMOR® Tiles The typical CERAMOR® tiles used during the tests had a regular hexagonal shape with a 1.85" (47mm) side and a total area of about 54.5 sq. cm. The tiles tested were either flat or nodded (MAP design with deflective surface shapes), as well as pressed or slip cast tiles. Due to unforeseen circumstances, some of the pressed tiles had a lower density than expected. The slip cast tiles and the second batch of pressed tiles had the desired optimum density values. The manufacturing process used raw powders with proprietary information regarding the composition, particle size and surface area. The powders were manually =Tonne Hydraulic'. The >R® tiles were then s tgfa temperature in on ACE: terials and Tect >, as shown in Figure 29. After an overnight cooling shrinkage was approximately 20% along the uncomprs L ••£: f? ; I I :E: 121 It is worth mentioning that all geometric dimensions of the CERAMOR® tiles were considered identical, due to the use of the same mould during the automatic pressing. The shrinkage of the tiles, although theoretically dependent on heterogeneities in the actual mixture composition was also considered identical. A precise measurement of each dimension for each tile in order to find the possible geometric differences was not achievable with the available tools. Other factors whose influence was not expected to play any significant role in the final ballistic test result might have been the specific humidity of the powder mixture and the position in the furnace of each tile, parameters that cannot be fully controlled. 4.2.3.5 Silicon Carbide Tiles The Pure-bite® silicon carbide (SiC) raw material for the specimens had been supplied by Morgan Crucible Inc. to ACERAM - Materials and Technologies Inc. in order to be pressed into shape with the same die. As a result, both silicon carbide and the CERAMOR® tiles have the same regular hexagonal shape and roughly the same area, both in flat or MAP configuration. The "green" Pure-bite SiC tiles were shipped to a Morgan Crucible facility to be sintered in a controlled atmosphere furnace. After sintering, the density of each tile was measured and recorded using the water immersion method based on Archimedes' principle. The mass of the tile in both air 122 and water was measured, the volume of the tile being obtained, which allowed the calculation of the density. The SiC and CERAMOR® tiles were tested in bare a state, as well as covered on the impact face either with the ACERAM - Materials and Technologies Inc. patent pending diamond coating, or encapsulated in PC films, a feature also patented by the same company. 28 tiles were covered with both the diamond coating and the polycarbonate film. The final variety of CERAMOR® and SiC tiles included tiles with flat impact surfaces as well as tiles with nodded impact surfaces, (MAP design) with areal densities comparable either with the boron carbide (B4C) tiles or the alumina (AI2 O3) tiles. In general, the number of ballistically tested samples of each type was four. If the samples performed consistently, only three were tested and in rare occasions five samples were tested to reduce the uncertainty in the results. The budget and testing time constrains also played a role in the decision to diminish the number of tests performed. 4.2.4 The Firing Range The live-fire testing was conducted in two stages at the Munitions Experimental Test Centre in Valcartier, in the 100-metre Small Calibre Terminal Ballistics Laboratory. For firing the Bofors-Carl Gustaf AP bullets, the laboratory utilised a universal receiver with a NATO test barrel. An electric switch behind an armoured window actuated the firing device. The impact velocity of each projectile onto the target, situated at 5 m from the muzzle was measured by two means, namely: 123 1. A first system was based on a Doppler radar, with a precision of ± 1%. The velocities were measured at a distance of 1.5 meters from the target and projected to the impact point using the retardation curve of each projectile . 2. A second system used two Oehler light screens with an electronic counter, the first screen located at 1 m from the muzzle of the gun. To protect them from ceramic ejecta, a transparent polycarbonate panel with a small opening for the projectile was clamped at the target end of the last screen. A laser beam from behind the breach travelled through the barrel, aligning it with the impact point on the target surface. Another important issue to address during ballistic testing as required by the STANAG 45692 is the measurement of projectile impact yaw angle, since its value will certainly influence whether a partial penetration or a complete penetration result is obtained for a given threat / target geometry. In the above mentioned STANAG it is suggested the use of a yaw "card" as a simple and effective method of assessing projectile yaw. It is recommended that the card material utilised should produce a clean hole matching authentically the presented area of the projectile. A critical requirement for the material is to not disturb in any way the onward flight characteristic of the projectile. A good example of such material is paper. When the hole in the yaw card is a perfect circle it is understood that no yaw is present in the projectile at the point of measurement. During several tests, a mobile holder with a sheet of ordinary paper was intercalated between the target and the last velocity screen, after the laser aiming procedure. All of the projectile perforations in the recovered papers were considered to be perfect circles upon visual inspection. However, due to the ceramic ejecta, sometimes the card was badly damaged, making the inspection of the perforation impossible. 124 The test jig was firmly clamped to a mobile table adjustable for height and axially aligned with the direction of the shot. The test jig position was accurately adjusted to ensure that the centre of the target block corresponded with the centre of the shot-line. The upper C channel used as clamping device was easily removable for replacing the samples. Behind the billet holding the ceramic sample, other two flat faced billets, each 150 mm in length, were ensuring enough DOP distance to mimic a semi-infinite material, and to supply enough length for the residual trajectory of the penetrating projectile. An image of the final target setup with the yaw card is presented in Figure 30. Figure 30: DOP target setup. The entire experimental arrangement used for ballistic testing consisted of the following parts: - Testing barrel mounted on a universal receiver Doppler radar for velocity measurement - Velocity measurement light screens - Yaw card on a holder 125 Target Bullet trap The drawing of the ballistic testing arrangement is presented in Figure 31. 5m Screens Trap Radar Target Card L r* lm H • Figure 31: Experimental arrangement used for ballistic testing. 4.2.5 Targets Tested A total number of 94 tests have been performed during the entire experimental program, including 4 tests on entire armour panels and 2 tests on bare polycarbonate for baseline values and 2 tests on aluminium faced polycarbonate. The aim was to test 4 samples in each configuration. When the samples in one configuration performed consistently, due to time constraints, the number of tests for that configuration was reduced to 3. These tests are synthesized Table 8. Table 8: Tests performed in the experimental program. Target Shape No of Impact Manufact. Spall Diamond No. Material Tests Surface Method Cover Coat 1. Alumina Octagon 3 flat pressed no no -> Octagon J Kevlar® 2. Alumina flat pressed no epoxy Hexagon 4 dry 3. CERAMOR® flat no no pressed Hexagon 4 dry 4. CERAMOR® flat no yes pressed Hexagon 4 dry 5. CERAMOR® flat PC film no pressed Hexagon 4 dry 6. CERAMOR® flat PC film yes pressed Hexagon 4 dry 7. CERAMOR® MAP no no pressed Hexagon 3 dry 8. CERAMOR® MAP no yes pressed Hexagon 4 dry 9. CERAMOR® MAP PC film no - pressed Hexagon 4 dry 10. CERAMOR® MAP PC film yes pressed 11. CERAMOR® Hexagon 5 MAP slip cast no no 12. CERAMOR® Hexagon 4 MAP slip cast PC film yes Boron Octagon hot 13. Carbide 11.3 4 flat no no pressed mm thick Boron Octagon hot 14. Carbide 8.5 4 flat no no pressed mm thick 127 Target Shape No of Impact Manufact. Spall Diamond No. Material Tests Surface Method Cover Coat Silicon Hexagon dry 15. 4 flat no no Carbide pressed Silicon Hexagon dry 16. 4 flat PC film yes Carbide pressed Silicon Hexagon dry 17. 4 MAP no no Carbide pressed Silicon Hexagon dry 18. 4 MAP PC film yes Carbide pressed CERAMOR® Hexagon dry 19. 4 MAP no no thick pressed CERAMOR® Hexagon dry 20. 4 MAP PC film yes thick pressed CERAMOR® Hexagon dry 21. 4 MAP no no thin pressed CERAMOR® Hexagon dry 22. 4 MAP PC film yes thin pressed 2x8 mm Al Round 23. 2 flat - - - discs 24. Polycarbonate Round 2 flat extruded - - CC-130 25. Square 2 flat - - - armour panel CERAMOR® dry 26. Square 2 map PC film yes armour panel pressed The purpose of the first series of tests was to compare the alumina (AI2O3) tiles currently in use on the CC-130 Hercules transport aircraft with several configurations 128 proposed by ACERAM - Materials and Technologies Inc, made from CERAMOR® ceramic material. The targets used in this first series of tests, included the following tiles: Table 9: Tiles tested in the first set of trials. Impact Manufact. Spall Diamond No. Tile material Notation Surface Method Cover coat 1. Alumina flat pressed no no H-l to H-3 Kevlar® HK-lto 2. Alumina flat pressed no epoxy HK-3 dry 3. CERAMOR® flat no no F-l to F-4 pressed dry FC-1 to 4. CERAMOR® flat no yes pressed FC-4 dry FP-ltoFP- 5. CERAMOR® flat PC film no pressed 4 dry FCP-1 to 6. CERAMOR® flat PC film yes pressed FCP-4 dry 7. CERAMOR® MAP no no B-l to B-4 pressed dry BC-1 to 8. CERAMOR® MAP no yes pressed BC-3 dry BPlto 9. CERAMOR® MAP PC film no pressed BP4 dry BCP1 to 10. CERAMOR® MAP PC film yes pressed BCP4 DB5to 11. CERAMOR® MAP slip cast no no DB9 DBCP1 to 12. CERAMOR® MAP slip cast PC film yes DBCP4 129 A few observations about the tiles tested in this first set are worth mentioning, namely: 1. All the CERAMOR® tiles were hexagonal, compared to alumina tiles which were octagonal, as declared before. 2. The Kevlar® / epoxy spall cover was identical with the spall cover used on CC- 130 Hercules. It was recovered from the same old armour panel as the alumina ceramic, and glued to the octagonal tile with the same epoxy adhesive used during this research for assembling the targets. 3. The PC film that formed the spall cover for the CERAMOR® tiles covered not only the front surface, but the back and the sides of the tile as well. It was glued to the ceramic material with a polyurethane adhesive. 4. Due to a manufacturing error, the porosity of the dry pressed CERAMOR® material was higher than expected, the density of the tiles being constantly 1% lower than expected, affecting the DOP results. Following the results obtained, the requirement of a second set of tests was deemed to be needed, in order to validate the previous results, as well as to expand the investigation on other available ceramic materials. This second test series was performed at the same research facility, namely Munitions Experimental Test Centre in Valcartier, with the same ammunition. The aim of the second set of tests was to compare the three ceramic materials widely used for ballistic protection, namely the boron carbide, the silicon carbide and the 130 dry pressed CERAMOR® ceramic material, in optimum configurations resulted from the first set of tests. The following tiles were tested during the second scheduled live testing: Table 10: Tiles tested in the second set of trials. Impact Manufact Spall Diamond No. Tile material Notation Surface Method Cover coat Boron Carbide hot 1. flat no no CEK1 to CEK4 11.3 mm thick pressed Boron Carbide hot 2. flat no no CENlltoCENH 8.5 mm thick pressed Silicon dry MF16,MF17,MF23, 3. flat no no Carbide pressed MF15 Silicon dry PC MFDP19,MFDP20, 4. flat yes Carbide pressed film MFDP22, MFDP24 Silicon dry MB31.MB32, 5. MAP no no Carbide pressed MB33,MB35 Silicon dry PC MBDP27 to 6. MAP yes Carbide pressed film MBDP30 CERAMOR® dry ABK36, ABK38, 7. MAP no no thick pressed ABK39, ABK43 CERAMOR® dry PC ABDPK37,ABDPK41, 8. MAP yes thick pressed film ABDPK42, ABDPK45 CERAMOR® dry 9. MAP no no ABN46 to ABN49 thin pressed CERAMOR® dry PC ABDPN50 to 10. MAP yes thin pressed film ABDPN53 131 A few observations about the tiles tested in this second set are worth mentioning: 1. All the CERAMOR® and silicon carbide tiles were hexagonal, compared to boron carbide tiles which were octagonal, due to the removal of their comers. 2. The PC film that formed the spall cover for the CERAMOR® and silicon carbide tiles covered the entire surface of the tile, compared to the diamond coating that covered only the strike surface. The PC film was glued to the ceramic material same as in the first set of tiles, namely with a polyurethane adhesive. One might question the necessity of testing two thicknesses of the same product. The boron carbide tiles have been supplied in two thicknesses, both of them having been manufactured to be integrated in protective armour against small arms ammunition. The thick CERAMOR® tiles were the manufacturer's recommended choice against 7.62 mm AP projectiles. The thin CERAMOR® tiles have a similar areal density with the thick (0.443") Boron Carbide tiles, thus adding another point of comparison between the two materials. No flat CERAMOR® tiles have been tested this time, since the previous set of tests demonstrated the superiority of the nodded ceramic against the threat. It is worth mentioning again that, compared with the previous tests; the density of this batch of pressed CERAMOR® tiles has been adjusted to its optimum value, 1% higher than the porous material. 132 4.2.6 DOP Measuring Method Due to the transparency of the PC backing, the DOP values could be measured right after the impact, on the jig, using a ruler. These values were recorded on paper and on the impacted PC billet. If the projectile penetrated more than one billet, all of the affected billets were removed and preserved as a unit. A second, more precise measurement was performed on the RMC premises, when all targets were also photographed. In some instances, when the curvature of the trajectory of the projectile core was excessive, attempts have been made to calculate the actual distance traveled by the projectile's core, named length of penetration (LOP), instead of using the value of the penetrating depth (DOP) in the PC billet. If the LOP was higher with more than 5 % of the total value compared with the DOP value, the first value vas used in calculations. 133 CHAPTER 5: RESULTS Following the two sets of DOP testing, the results were consolidated in a table format, presented in the Appendix at the end of this report. The table contains the mass and the surface of the samples, from which their areal densities were obtained. The recorded impact velocities of the projectiles, as well as the measured depth of penetration for each target are also presented. If the trajectory of the damaged core of the impacting projectile in the PC was significantly deflected, a second value, of a length of penetration (LOP) was measured. Since the measurements were performed visually from outside of the PC, a small degree of visual distortion and, therefore some error is expected. For most of the important values (such as the DOP or LOP, the impact velocities, the masses, the densities and the areal densities of each group of targets) their average value with the standard deviation were included. Due to proprietary issues, the average thickness values of the CERAMOR® and E-CERAMOR® tiles are not included as they can be used to obtain their density. Because of this limitation, the calculated values of the equivalent thickness or 2 thickness efficiency (Eeq) and the quality factor or ballistic efficiency factor (q ) (using the formulae provided earlier in this document) were not possible for the specified CERAMOR® materials in either configuration. The critical thickness values, in other 134 words the minimum ceramic thicknesses needed to just defeat the projectile, were also not available for the CERAMOR® materials. It is worth noting that the velocities displayed in the table of results represent the actual velocities at the point of impact, obtained with the retardation curve of each projectile from the measured point situated 1.5 m in front of the target; therefore no correction for the air drag effect was required. Some additional information for describing the results obtained is also presented in the "Remarks" column of the same table of results in the Appendix. The table from the Appendix can be consolidated in a shorter table containing the average values for each type of ceramics tested. This information is presented in Table 11. Table 11: Average results for the ceramic samples tested. Tile Spall Area dens DOP StDev Equiv Equiv Quality Critic thick Cover [g/cm2] [mm] thick mass factor [mm] A1203 no 3.3 194 1.73 1.8 0.6 1.1 118.8 A1203 Kevlar 3.38 190.3 1.53 2.1 0.7 1.5 102.5 B4C thick no 2.67 12 15.32 17.8 8.6 152.5 11.9 B4Cthin no 2.15 25.8 5.91 21.8 10.5 228.6 9.7 SiC no 2.66 9.8 12.82 23.6 9.1 216.1 9.0 SiC "E"+ PC 2.57 15 7.8 23.7 9.2 218.7 8.9 SiC MAP no 2.59 6 12 24.6 9.6 236.7 8.6 SiC MAP "E"+ PC 2.83 14.5 12.18 21.6 8.4 182.3 9.8 CERAMOR no 3.29 150.8 15.59 - 2.2 - - porous 135 Tile Spall Area dens DOP StDev Equiv Equiv Quality Critic thick Cover [g/cm2] [mm] thick mass factor [mm] CERAMOR 3.26 128 52.57 - ^ - - "E" j porous CERAMOR PC 3.48 157.3 18.98 - 1.9 - - porous CERAMOR "E"+ PC 3.51 153.5 12.77 - 2 - - porous CERAMOR no 3.36 127.8 52.39 - 2.9 - - MAP porous CERAMOR "E" 3.4 173 7.21 - 1.4 - - MAP porous CERAMOR PC 3.61 107.5 38.27 - 3.5 - - MAP porous CERAMOR "E"+ PC 3.61 124.3 43.12 - 2.9 - - MAP porous CERAMOR no 3.4 53.2 32.06 - 5.6 - - MAP-cast CERAMOR "E"+ PC 3.69 33.3 39.53 - 5.8 - - MAP-cast CERAMOR no 3.19 22.8 8.5 - 7.1 - - MAP dense CERAMOR "E"+ PC 3.39 12.3 15.11 - 7.1 - - MAP dense CERAMOR no 2.42 51.5 14.46 - 8 - - MAP dense CERAMOR "E"+ PC 2.64 61.3 63.16 - 6.9 - - 1 MAP dense Another quick reference guide of the results obtained is the graph containing the DOP versus the areal density of each sample, presented in Figure 32. In addition to the above tiles, two DOP tests were performed against the bare polycarbonate material in order to obtain the value for the depth of penetration with no added ceramic, a value needed for further calculations. Compared with the ceramic covered PC, the core of the projectile has not been stripped from its jacket by the plain 136 PC target, the entire projectile penetrating the PC material in intact shape. An attempt to strip the core of the projectile by adding 2 discs, 8 mm thick each was successful, although the result obtained was not used in further calculations. Two shots were also performed against a complete original CC-130 Hercules armour panel and two others against an ACERAM - Materials and Technologies Inc. designed armour panel. Both panels failed each test. DOP versus Areal Density 3.80 • Alumina bare 3.60 • Alumina with Kevlar/epoxy A CERAMOR porous 3.40 •- OE-CERAMOR porous -+- X CERAMOR porous, in PC + 0 A A <* A fa + A E-CERAMOR porous, in PC „ 320 + CERAMOR-MAP porous < • E-CERAMOR-MAP porous E u o E-CERAMOR-MAP porous, in PC a 3.00 A CERAMOR-MAP dense cast (A • E-CERAMOR-MAP dense cast, in PC C A A 2.80 • B4C thick (0 • B4C thin 9> I O. OSiC 2.60 f- ^fer • "E"-SiC in PC « SiC-MAP A"E"-SiC-MAPinPC 2.40 A CERAMOR-MAP dense pressed • E-CERAMOR-MAP dense pressed, in PC A CERAMOR-MAP dense pressed, thin 2.20 • E-CERAMOR-MAP dense pressed, in PC, thin 2.00 20 40 60 80 100 120 140 160 180 200 DOP in PC [mm] 138 In comparing the effectiveness of different ceramic materials, the main parameter to which the DOP measurement could be related is the areal density of the ceramic tile. For materials with similar areal densities, the ones with lower DOP values are preferred. The materials with similar DOP values can be easily sorted by their areal densities, since a lighter material with the same ballistic performance is desired. The graph of the DOP measurements versus the areal densities for all the tests performed, shown in Figure 32, can be divided in three distinct groups of results. The groups can be arbitrarily divided by a depth of penetration value of 100 mm, and by their areal densities, namely lower or higher values than 3 g/cm . The resulting groups, presented in the following three figures could be defined as follows: 9 9 1. A first group with a high areal density (3.10 g/cm to 3.70 g/cm ) and high DOP values (above 100 mm), Figure 33. 9 9 2. A second group with a high areal density (3.20 g/cm to 3.80 g/cm ) and low DOP values (less than 100 mm), Figure 34. 9 9 3. A third group with a low areal density (2.10 g/cm to 2.90 g/cm ) and low DOP values (less than 100 mm), Figure 35. Some ceramic tiles might have results crossing in two groups on the graph; the group with the majority of the results being considered as appropriate. DOP versus Areal Density for heavy, inefficient samples 3.70 3.60 X D Alumina bare 350 ^r ~ • Alumina with Kevlar/Epoxy < A CERAMOR porous E X A X .o OE-CERAMOR porous jo> X CERAMOR porous, in PC •f 3.40 c A E-CERAMOR porous, in PC « Q + CERAMOR-MAP porous "<5 + 9> • E-CERAMOR-MAP porous o E-CERAMOR-MAP porous, in PC 3.30 ^Q- A A O/ • B4C thin 3.20 3.10 40 60 80 100 120 140 160 180 200 DOP in PC [mm] »*1 TO S DOP versus Areal Density for heavy, efficient samples re"< W *. o 3.80 • o II *s II • CERAMOR-MAP dense cast < 3.70 • E-CERAMOR-MAP dense cast, in PC 'A09 • s09 • CERAMOR-MAP dense pressed to 1 • E-CERAMOR-MAP dense pressed, in PC 5 3.60 a re • 2s 3.50 5 • © 1 «5 le i > 3.30 si t re c S a> **- n Si 3.20 A 3 ea l • A A 5- < 3.10 3.00 2.90 • 2.80 10 20 30 40 50 60 70 80 90 10C DOP in PC [mm] o DOP versus Areal Density for light, efficient samples 2.90 A * \ 2.80 • B4C thick • B4C thin 2.70 0 OSiC ir • • • • • "E"-SiC in PC • 2.60 41 * SiC-MAP < A"E"-SiC-MAPinPC E 11 • CERAMOR-MAP dense pressed, thin 2.50 • E-CERAMOR-MAP dense pressed, in PC, thin • w c 2.40 Q a> 2.30 2.20 a • • 2.10 • 2.00 4 'i —r i 1 1 1 i 0 20 40 60 80 100 120 140 160 180 200 DOP in PC fmml 142 The first group (Figure 33) is the least desirable since it will generate relatively heavy ceramic armour with poor performance. It includes the tested alumina tiles, either in bare state or Kevlar®/epoxy spall covered, as well as most of the porous CERAMOR® material. As exceptions, few porous tiles of CERAMOR®-MAP material performed slightly better than average, having DOP values less than 100 mm. The second group (Figure 34) with a similar average areal density compared to the first one performed better, having lower DOP values. This category includes the cast CERAMOR® tiles, the few results from the porous CERAMOR® tiles and the dense pressed CERAMOR®-MAP tiles. The third and most attractive group (Figure 35) contains the targets that should generate the primary ceramic choices for future investigation and use. They contain relatively light tiles with acceptable behaviour under impact. From this group it is worth mentioning the SiC tiles, the B4C tiles and the thin CERAMOR®-MAP tiles. By taking into consideration only the results from the most attractive group from Figure 35, the materials that are worth further investigations are: B4C in both thicknesses, SiC flat as well as in MAP configuration and the CERAMOR®-MAP materials. If possible, it might be worth clarifying the influence of the PC films and the diamond coating through further testing. 143 CHAPTER 6: INTERPRETATION OF RESULTS The interpretation of the results focuses on the assessment of the importance of each factor that might have an influence on the performance of each group of samples. The DOP average results and standard deviations along with the mass efficiency, quality factor and critical thickness, where applicable, are compared in this chapter. 6.1 FLAT VERSUS MAP FOR POROUS CERAMOR® The influence on the MAP design versus the flat tiles was compared using the DOP and equivalent mass results in Table 12. Table 12: Flat versus MAP results for porous CERAMOR®. POROUS CERAMOR-MAP POROUS CERAMOR FLAT Spall Equiv Equiv DOP Std Dev DOP Std Dev Cover Mass Mass Bare 127.8 52.39 2.9 150.8 15.59 2.2 "E" 173 7.21 1.4 128 52.57 3 PC 107.5 38.27 3.5 157.3 18.98 1.9 "E" + PC 124.3 43.12 2.9 153.5 12.77 2 144 From the results obtained, the MAP design is in general superior in defeating the tested projectile. The only anomaly is the presence of the diamond coating which seems to worsen the performance of the MAP tiles, while improving the performance of their flat counterparts. As a result, for the second set of tests, only the MAP design has been selected. 6.2 POROUS VERSUS DENSE FOR CERAMOR-MAP® The influence of the porosity on the performance of the CERAMOR-MAP® tiles was compared in Table 13. The comparison includes only the bare tiles and the tiles covered with diamond coating and polycarbonate, at similar area densities. Table 13: The influence of the porosity for CERAMOR-MAP®. POROUS CERAMOR-MAP DENSE CERAMOR-MAP Spall Equiv Equiv DOP Std Dev DOP Std Dev Cover Mass Mass Bare 127.8 52.39 2.9 22.8 8.5 7.1 "E" + PC 124.3 43.12 2.9 12.3 15.11 7.1 The results lead to the conclusion that the influence of the porosity on the DOP result is extremely significant. 145 6.3 SPALL COVER INFLUENCE ON CERAMOR-MAP® The influence of the spall cover on the performance of the CERAMOR-MAP® tiles was compared. The comparison includes only the bare tiles and the tiles covered with diamond coating and polycarbonate, for low and high area densities. The cast CERAMOR-MAP® tiles are included in this comparison as well. Table 14: Spall cover influence on CERAMOR-MAP® POROUS CERAMOR- CAST CERAMOR- DENSE CERAMOR- MAP MAP MAP Spall Std Equiv Std Equiv Std Equiv DOP DOP DOP Cover Dev Mass Dev Mass Dev Mass Bare 127.8 52.39 2.9 53.2 32.06 5.6 22.8 8.5 7.1 "E" + PC 124.3 43.12 2.9 33.3 39.53 5.8 12.3 15.11 7.1 For the porous MAP tiles, the adding of spall cover did not improve their performance significantly. The performance of the cast MAP tiles was improved by the addition of the spall cover, although their general performance was still lower than their pressed counterparts with low porosity. From Table 14 it can be concluded that the best performing CERAMOR® tile is the pressed MAP with spall cover. 146 6.4 FLAT VERSUS MAP FOR SILICON CARBIDE The effect of the MAP design for the silicon carbide has been evaluated for tiles with or without spall cover. When present, the spall cover included diamond coating and the polycarbonate film. Besides the DOP values with their standard deviations, the comparison includes the quality factor and the critical thickness for each group. Table 15: MAP and spall cover influence on silicon carbide results. SILICON CARBIDE MAP SILICON CARBIDE FLAT Spall Std Quality Critical Std Quality Critical DOP DOP Cover Dev Factor Thick. Dev Factor Thick. no 6 12 236.7 8.6 9.8 12.82 216.1 9.0 yes 14.5 12.18 182.3 9.8 15 7.8 218.7 8.9 From Table 15 it seems that the best performing design for silicon carbide material is the MAP configuration, with no spall cover. 6.5 THICKNESS INFLUENCE ON BORON CARBIDE The effect of the thickness on the performance of boron carbide tiles was evaluated. Besides the DOP values with their standard deviations, the comparison includes the quality factor and the critical thickness for each group. 147 Table 16: Thickness influence on boron carbide results. BORON CARBIDE THIN BORON CARBIDE THICK Std Quality Critical Std Quality Critical DOP DOP Dev Factor Thick. Dev Factor Thick. 25.8 5.91 228.6 9.7 12 15.32 152.5 11.9 From the results in Table 16 it appears that although the DOP performance of the thin boron carbide tiles is lower, their quality factor is higher, therefore their calculated efficiency is higher than for their thick counterparts. 6.6 MATERIAL INFLUENCE FOR HIGH AREAL DENSITY TILES The effect of the material on the DOP results was compared for tiles with an areal density over 3 g/cm . The materials mclude CERAMOR® and alumina. From the CERAMOR®, the comparison includes only the dense MAP tiles. When present, the CERAMOR® spall cover in this comparison includes both diamond coating and polycarbonate, while the alumina spall cover consists of a Kevlar® epoxy sheet. Table 17: High areal density tiles, material's influence. ALUMINA CERAMOR-MAP Spall Equiv Equiv DOP Std Dev DOP Std Dev Cover Mass Mass no 194 1.73 0.6 22.8 8.5 7.1 yes 190.3 1.53 0.7 12.3 15.11 7.1 148 From Table 17 it is obvious that the CERAMOR® material vastly outperforms the alumina. 6.7 MATERIAL INFLUENCE FOR LOW AREAL DENSITY TILES The effect of the material on the DOP results was compared for tiles with an areal density under 3 g/cm . The materials include CERAMOR®, silicon carbide and boron carbide. From the CERAMOR® samples, the comparison includes only the dense MAP tiles with low areal density. When present, the CERAMOR® or silicon carbide spall cover in this comparison includes both diamond coating and polycarbonate. Table 18: Material influence for tiles with low areal density. BORON BORON CARBIDE SILICON CERAMOR-MAP CARBIDE THIN THICK CARBIDE MAP Spall Std Equiv Std Equiv Std Equiv Std Equiv DOP DOP DOP DOP Cover Dev Mass Dev Mass Dev Mass Dev Mass no 25.8 5.91 10.5 12 15.32 8.6 6 12 9.6 51.5 14.46 8 yes ------14.5 12.18 8.4 61.3 63.16 6.9 Based on DOP results, among the ceramic materials tested in tiles with low areal densities, the group with the best performance appears to be the MAP silicon carbide with no spall cover, followed by the thick boron carbide. Based on the values of mass 149 equivalency, the thin boron carbide and the MAP silicon carbide with no spall cover seem to be the best choices. For the parameters investigated in this research project, the ceramics of choice for light armour are the bare silicon carbide in MAP configuration and the boron carbide at the higher thickness. The CERAMOR-MAP® is the third best option, especially when a slight increment in the mass of the armour is not critical. 150 CHAPTER 7: DISCUSSION Three ceramic materials, namely boron carbide, silicon carbide and CERAMOR® have been tested as tiles for depth of penetration (DOP), using polycarbonate billets as semi-infmite backing. These results have been compared with the results of alumina tiles from the armour panels in use on the CC-130 Hercules aircraft. The testing ammunition was 7.62x51 mm AP round with WC core, corresponding to level III threat according to STANAG 45692. 7.1 LIMITATIONS The small number of tiles and the performance range did not justify a comprehensive statistical analysis. An analysis based on such a small number of samples from each category could not produce any generalisation that can be accepted without reasonable doubt, as observed by Gilde and Adams77. They consider that the small number of tests corroborated with the large spread in DOP data for ceramics makes it difficult to draw viable conclusions although this is done often, due to the costs and complexity of the ballistic testing. Therefore, the subsequent discussion on the results 151 obtained, pertains only to the specific ones obtained in this study, and only for the range of the parameters investigated. The fact that some of the target configurations exhibited no penetration in polycarbonate in 75% of the shots (MAP SiC) or 50% of the shots (E-CERAMOR®- MAP, B4C-thick, flat SiC), although impressive at first sight, diminished even more the data base that could have been used for a statistical analysis. The conditions and constraints of the current test program did not allow the author to study the behaviour under various impact velocities of the boron carbide tiles in more depth. The armour designer must be aware that, according to some researchers such as Moynihan64, the B4C impacted with the same ammunition used in the present tests (7.62 x 51 mm, AP, WC core) exhibits a shatter gap at projectile velocities around 870 m/s. As a result, ways to mitigate this effect have to be taken into consideration when using B4C ceramic in armour design. 7.2 TEST VALIDITY The highest impact velocity during this test series was 940.8 m/s and the lowest impact velocity was 930.2 m/s. Due to this low variation of the velocities (935.5 ± 5.3 m/s), all the tests are in within the tolerance required by the STANAG 45692, namely ± 20 m/s, therefore considered valid hits. As mentioned before, attempts were made to measure the yaw angle of each shot with the aid of yaw cards mounted roughly 30 cm in front of the target. Unfortunately, in 152 most situations, the ceramic fragments ejected during the impact damaged the cards. In the few cases when the cards could be recovered, the orifice made by the projectile was perfectly round. Therefore, for the further interpretation of the results it was assumed that all projectiles impacting the targets had a zero yaw angle, 7.3 SPALL COVER EFFECT The ejection of ceramic fragments was more predominant for the tiles with no PC film or Kevlar®/epoxy spall covers. Apparently, the addition of spall covers has a beneficial effect in containing the fragmented ceramic. The DOP has not been reduced significantly with the Kevlar®/epoxy spall cover addition to the alumina tiles (only about 2% difference between average values), and the PC film spall covers generally seems to have worsened the performance of the SiC tiles. The use of spall cover for the CERAMOR® tiles generated mixed results. However, the spall cover is worth considering for decreasing the collateral damage done by the ejecta (or secondary fragments) generated during the impact. St-Denis44 affirms that following DOP tests there are some indications that the frontal confinement effect is more pronounced at lower velocities than at higher velocities. As a consequence, in particular impact situations the confinement added by either the PC films or the Kevlar®/epoxy spall covers might help. On average, the bare SiC tiles outperformed the diamond coated and PC film covered SiC counterparts, for similar areal densities, in both flat (9.8 mm versus 15 mm 153 DOP) and MAP (6 mm versus 14.5 mm DOP) configurations. Although the MAP design appeared to be more effective for bare SiC tiles, the adding of the diamond and the PC film seemed to "neutralize" the effect of the nodes, the average performance of both configurations being similar (14.5 mm versus 15 mm). The addition of diamond coating seemed to have worsened the ballistic behaviour of the CERAMOR®-MAP tiles. On average, they decreased the performance of both the bare tiles as well as the PC film sandwiched configurations. 7.4 INFLUENCE OF LATERAL CONFINEMENT One might inquire about the contribution of the lateral confinement provided by PC and its cracking, and ways to take it into account and somehow quantify it. It is generally understood that the lateral confinement allows for the constraint of broken ceramic fragments; thereby enhancing the erosion phase. Some research has been done in measuring this influence by Yadav and Ravichandran30 without much success. They concluded that for DOP tests, it is very difficult to quantify the effects of confinement due to lateral support or due to a front cover plate. Due to these previous unsuccessful attempts and the limited amount of data available after the current research, no attempts had been made to quantify the above mentioned confinement related to the present testing. 154 7.5 SHATTERING PATTERN FOR A1203, SiC AND B4C The shattering pattern of each type of target had particularities that are worth mentioning in our discussion. The uncovered alumina tiles shattered in relatively large pieces and few cracks extended to adjacent tiles, compared to the Kevlar®/epoxy covered ones, where the fragments were smaller in size and the cracking was more contained, as seen in Figure 36 and Figure 37. Figure 36: Uncovered alumina targets, after impact. 155 Figure 37: Kevlar® covered alumina target after impact. The boron carbide targets also generated large fragments upon impact, with radial cracks extending to the edges of the tiles. Apparently, the thicker B4C tiles (11.2 mm) underwent more damage over a larger area, compared with the thinner ones (8.5 mm), as seen in Figure 38. 156 Figure 38: Boron carbide targets; the thick samples (CEK) shown in top row, the thin samples (CEN) in bottom row. The thick B4C tiles had the average DOP value less than half compared to the thin B4C tiles. However, because the damage of the thick B4C tiles was over a larger area compared to their thin counterparts, one might question the effectiveness of these tiles against subsequent impacts in the vicinity of the first hit. Same as B4C, the SiC tiles exhibited explosive shattering at impact, as remarked also by Medvedovski in his study. He observed that most of the SiC material is transformed into a powder at the impact area with a minimum amount of large ceramic fragments, fragments that are peeled off the backing material in spite the strong bonding. The uncovered SiC tiles embedded in the PC billets were completely shattered and, in 6 157 out of 8 tests no material remained attached to the epoxy adhesive and the PC backing. The tiles covered with the PC film spall covers although completely destroyed, usually remained attached to either the front film or the backing. Some results are shown in Figure 39 to Figure 42. Cracks in PC cylindrical backing and the size of the shattered grains of SiC in the confined tiles are also visible. Figure 39: Uncovered SiC MAP tile completely shattered after impact (MB 33). 158 Figure 40: SiC MAP tile sandwiched within PC films and diamond coated, after the impact (MBDP 30). Figure 41: Fragmentation of a SiC tile confined within PC films and radial cracks in the PC cylindrical backing (MBDP 28). 159 Ceramic and projectile material Figure 42: Comminuted (pulverized) ceramic and the AP core contained in cracked PC backing cylinder for flat SiC tiles sandwiched between PC films (MFDP 20). It is worth noting the shape of the large fractures in the PC cylinder in Figure 42. The majority of the cracked PC billets presented similar patterns for both octagonal and hexagonal tiles. 160 7.6 DOP AND EFFECT ON PROJECTILE FOR A1203, SiC AND B4C The effect of both covered and uncovered alumina targets on the cores of the impacting projectiles was relatively insignificant, both their shape and trajectory not being notably affected. Both thicknesses of the B4C targets shattered consistently the AP core. The polycarbonate material around the trajectory of the projectile's core in the PC billet was plastically deformed, and several large radial and lateral cracks developed in the unaltered portions of the PC billet, extending large distances from the penetrated region, as seen in Figure 43 and Figure 44. Figure 43: Radial and lateral cracks in PC - thick B4C target (CEK 2). 161 Figure 44: Radial and lateral cracks in PC - thin B4C target (CEN 11). Seven out of sixteen silicon carbide tiles exhibited no penetration in the PC backing. The percentage of no penetration in each group of four varied from 25% to 75%. As mentioned earlier, the bare SiC tiles outperformed the diamond coated and PC film covered SiC counterparts, for similar areal densities, in both flat (9.8 mm versus 15 mm DOP) and MAP (6 mm versus 14.5 mm DOP) configurations. The SiC MAP configuration seems to be more effective in stopping the AP rounds compared to the flat configuration, but only for the bare tiles (6 mm versus 9.8 mm). The adding of the diamond and the PC film seemed to "neutralize" the effect of the nodes, the average performance of both configurations being similar (14.5 mm versus 15 mm). 162 7.7 CERAMOR® POROSITY AND PERFORMANCE Most of the CERAMOR® material tested in the first test series was pressed. Due to a manufacturing error, the density of the tiles obtained was lower than expected. As a consequence, the DOP results were poor compared to expected values for this material due to the increased porosity, and consequently they have been ignored in the final recommendations. Nevertheless, the DOP results of the CERAMOR® porous tiles are a reasonable tool in assessing the influence of the MAP design compared to the flat tiles, as well as the influence of adding the diamond coating and / or polycarbonate spall covers. From this perspective the flat CERAMOR® porous tiles performed similarly. Apparently the diamond coating or / and the PC film did not improve significantly the DOP results. On average, the DOP results were above 150 mm for three out of four groups. One singular exception was the diamond coated group (average DOP 128 mm) which contained both the highest and the lowest DOP values, namely 180 mm and 57 mm respectively, with no apparent explanation except for a normal variability of the ceramic material. The low density CERAMOR®-MAP in bare state, had also an interesting behaviour. On average, the DOP for this group was 127.8 mm, but a standard deviation of 52.4 mm does not give much confidence in these results. The tile with the lowest density happened to be hit by the projectile with the highest velocity in the group, leading to the highest DOP of 178 mm. Two other tiles with similar densities (their masses differing by less than 0.1%) were hit by projectiles with identical velocities but performed in dissimilar manner; one registered a DOP of 74 mm while the other had a 163 value more than double, at 167 mm. The fourth target although having the highest density was hit by a bullet with similar velocity to the others and had a 92 mm DOP. It is believed that a major improvement exhibited by the low density MAP configuration CERAMOR® tiles (compared to their flat counterparts), is their capability to visibly erode and fragment the WC core of the projectile. Compared to the flat diamond coated and PC film covered low density CERAMOR® tiles, the similarly configured CERAMOR®-MAP tiles managed to erode one side of the tip of the core, leading to a shorter, curved penetration trajectory in the PC cylindrical backing. Even more remarkable, some bare porous CERAMOR®-MAP tiles were able to fragment the projectile's core, a fact not observed with any of their flat counterparts, proving the superiority of the MAP design. A few pictures of impacted CERAMOR® tiles are presented below. Figure 45: Flat CERAMOR® low density samples after impact. The projectiles cores were left intact. 164 Figure 46: WC cores penetrated in the PC backing after impacting E-CERAMOR®-MAP sandwiched between PC films, low density tiles. The cores have curved trajectories. Figure 47: WC fragmented cores penetrated the PC cylinders after impacting the CERAMOR®-MAP low density tiles. 165 It should be noted that the dimensions for DOP values written on the billets in the above figures represent the rough initial measurements taken from the top face of the PC, the depth of the socket in which the ceramic tile was mounted being subtracted subsequently, during a second, more accurate set of measurements. The high density CERAMOR®-MAP cast tiles (having similar areal densities with the low density CERAMORJD-MAP pressed tiles) performed significantly better, with most DOP values lower than half of that of the low density pressed tiles. These results were convincing enough to consider the manufacturing of the second batch of the CERAMOR®-MAP pressed tiles, under a better quality control of the manufacturing process in order to further eliminate most of their porosity and obtain the desired density. Because of the MAP design superiority, only CERAMOR®-MAP had been selected for the second set of tests. The tiles were tested either in their bare state or diamond coated (E-CERAMOR®) and sandwiched between PC film spall covers. The DOP test results for these dense CERAMOR® tiles were somehow mixed. For the tiles with a higher areal density, the addition of the diamond coating and the PC film covers improved the average DOP results (22.8 mm versus 12.3 mm). The diamond and PC film additions on the lighter CERAMOR®-MAP tiles had an opposing effect, increasing the average DOP values from 51.5 mm to 61.3 mm. 166 7.8 RESULTS INCONSISTENCIES The causes that generated the significant discrepancies in the DOP values for targets from the same groups are difficult to highlight. One reason might be that, for the MAP configuration, the exact point of impact on a node plays a significant role in defeating the projectile. This source of inconsistency is not available for flat tiles. However, even for the flat tiles the difference in DOP values is significant, the SiC tiles being a good example. For the uncovered tiles, 50% of the shots exhibited zero penetration in PC, while the worst performing target yielded a DOP of 27 mm. As a result, the average DOP was 9.8 mm and the standard deviation was 12.8 mm. Radial and lateral cracks were present in the PC cylinders for most of the SiC targets (similarly to the B4C targets). The CERAMOR® material had the uppermost variation of the DOP results compared with the other ceramics. The variation was higher for MAP tiles, but significant for their flat counterparts, as well. The calculated standard deviation for the CERAMOR® high density groups tested in the second set increases the uncertainty of the results. With such a small number of tests - about four for each group - each result has a significant weight on the average value. For instance, in the case of the thicker CERAMOR®-MAP tiles, the highest penetration for both bare and covered tiles was similar, namely 31 mm. Only the fact that there were two zero DOP values in the case of PC film and diamond covered tiles improved their overall results. 167 An explanation of the relatively high standard deviation of the DOP results, beside the normal variation of the ceramic material, might reside in the inconsistency of the quality of the material and / or the manufacturing technique. The inspection of some low density CERAMOR®-MAP tiles found a number of them containing fissures, most probably due to the manufacturing process. One example is illustrated in Figure 48. Figure 48: Low density CERAMOR®-MAP tile fissured, before testing. The different patterns of these fissures, corroborated with the inherent variability in ceramic DOP test results might explain some of the unexpected results for this particular material. This conclusion is supported by the fact that for the same set of tiles the fragmentation of the ceramic under the impact has been different, as illustrated in Figure 49 and Figure 50. Figure 50 shows evidence of considerable spalling of the ceramic tile. 168 Figure 49: Low density CERAMOR®-MAP tile (B-2) after the impact (V = 935.3 m/s, DOP of 74mm). 169 Figure 50: Low density CERAMOR®-MAP tile (B-4) after the impact (V = 935.4 m/s, DOP of 92mm). Another cause of the large discrepancies in the test results might be due to the fact that the tests were performed at the limit "capabilities" of the tiles, where a slight variation of a parameter might have made the difference between a complete defeat of the projectile and the penetration of a broken projectile core. Probably a much larger number of identical tiles tested under similar conditions would lower the uncertainty of the results. 170 In some instances, the DOP for some sets of samples did not correlated well with either the areal density of those samples and/or the impact velocity of the projectile. From this perspective it seems that there might be other factors not taken into account whose inputs might be important during a DOP type testing. One explanation for the differing DOP results as well as the end state of the ceramic tile might be that the adhesive did not bond similarly to each of the bare ceramic materials or to the PC film used in the back of the ceramic. 7.9 CALCULATIONS BASED ON DOP RESULTS A more mathematical approach in comparing the performance of the ceramic tiles under ballistic impact should make use of formulae to determine the equivalent thickness or thickness efficiency Eeq (Equation 22), the equivalent mass or mass efficiency Meq (Equation 23) and the quality factor or ballistic efficiency factor q2 (Equation 24), previously defined in this study. The results of these formulae are also presented in the table Table 11 and in the appendix, with the comment that due to proprietary information, the CERAMOR® material could not be included in some of the calculations. One may choose the material for future armour based solely on the equivalent thickness if the space where the armour has to be mounted is limited, or on the equivalent mass, if the weight is an issue for the protected target. Otherwise both results are combined to generate a quality factor in order to characterize the overall efficiency of the tested ceramic. 171 The critical thickness, another theoretical value derived from the equivalent thickness results represents the theoretical thickness of the ceramic capable of just defeating the projectile, in the tested configuration. A higher performance ceramic will require less thickness to stop the projectile, therefore a lower value is preferred. 7.10 ADDITIONAL TESTS In order to be able to use the formulae, a value for the DOP in PC with no ceramic tile was needed. That value was consistent in two successive trials, namely 211 mm. The fact that the projectile travelled the entire trajectory in PC intact, (with the copper jacket) added some concern that further calculations might be slightly inexact, since after impacting a ceramic tile, only the core will travel in the PC. An attempt to strip the core of the projectile with two 8 mm thick aluminium discs was successful, 198 mm and 201 mm DOP values being obtained respectively. However, these values were about 6% lower than the previous ones, thus probably quantifying the energy that was consumed to plastically deform the jacket and penetrate the discs. After consulting the work of other authors, such as Gotts21, who used the same backing material (namely PC), it has been determined that the more appropriate DOP value to be used should be the one obtained with no aluminium discs, namely 211 mm. 172 7.11 CC-130 HERCULES ALUMINA RESULTS At first, the DOP values obtained for the alumina ceramics from the CC-130 Hercules aircraft appear to be too disproportionate in comparison with the values obtained for other ceramics, and difficult to explain from an engineering perspective. The DOP data obtained during the testing of this ceramic lead to the apparent conclusion that, in order to stop the AP projectile used, a thickness of more than 100 mm of this alumina was needed. All the other ceramics tested exhibited values less than 10% of the alumina, leading to the conclusion that unseen conditions might have affected the behaviour of the old alumina targets. An analysis of the factors that could have influenced these results highlighted a few possible explanations for the poor performance of the CC-130 Hercules alumina tiles. A first possibility is that the tiles manufactured probably at about the same time as the aircraft (more or less 40 years ago), were not designed for this type of projectile, which came on the market more than 25 years later. The size, the materials and the manufacturing technique of these tiles might not have been suited to withstand an impact by a WC core AP projectile at muzzle velocity. Another possibility is that the relatively small area of the tiles tested generated a large amount of reflected waves from the lateral walls, thus leading to a premature fragmentation of the ceramic and implicitly, a poor ballistic behaviour of these samples. One might argue that the adhesive used in mounting the alumina tiles was not identical with the one used in the actual armour and the manufacturing of the target increased the gaps between the tiles and, implicitly, the energy of the reflected waves. 173 Besides, the armour is designed to work best in combination with the appropriate backing material, not on the testing bench. Although not contesting the above opinions, it is worth mentioning that the testing of the entire system against the same projectile failed to provide the expected protection, as seen in Figure 51. Figure 51: CF-130 Hercules armour panel perforated by two 7.62 AP WC core projectiles at muzzle velocity (930 m/s), exit face. 7.12 CONCLUDING REMARKS In comparing the results obtained for the samples from the modern ceramics with no proprietary limitations, namely the boron carbide (B4C) and the silicon carbide (SiC) 174 tiles in all available thicknesses and configurations, several observations are worth mentioning. All ballistic efficiency factor results lay within a band between a value of 122 and 260. The best average ballistic efficiency was given by the SiC MAP tiles with the value of 236.7. The thick B4C tiles performed the worst, with an average value of only 152.5. The best theoretical critical thickness was found for the SiC in MAP configuration, with a value of 8.6 mm. The flat SiC had similar results, between 8.6 mm and 9.5 mm. The MAP SiC diamond coated and sandwiched in PC films had the highest, hence the worst critical thickness value for SiC, at 10.3 mm. The thick (11.2 mm) B4C yielded the least desirable theoretical critical thickness result of all samples tested, with an average of 11.9 mm, worse than the thin B4C tiles (with a resultant average of 9.7 mm), The average theoretical critical thickness of the thin B4C is slightly better than the worst SiC sample of 9.8 mm. This leads to the apparently challenging conclusion that the use of a thinner ceramic is recommended over a thicker one. Although not necessarily technically correct, the conclusion might be accurate if we take into consideration the lower protection increment by increasing the thickness (and implicitly the areal density) of the ceramic. In considering only the results obtained from the efficiency factor and the critical thickness formulae for the ceramic samples tested, we can conclude that the best choice for a proposed armour panel would incorporate the MAP SiC in bare configuration, as its ceramic component. 175 CHAPTER 8: CONCLUSIONS Four major ceramic materials in use for ballistic protection against small arms ammunition, namely alumina (AI2O3), silicon carbide (SiC), boron carbide (B4C) and CERAMOR® have been tested in various configurations for the depth of penetration (DOP) in polycarbonate (PC), at level III, STANAG 45692. The projectile used was 7.62x51 AP Bofors - Carl Gustaf with WC core. The purpose of the test was to compare the performance of the alumina material (currently used in protective panels on CC-130 Hercules aircraft) with these other materials available on the market today. The conclusions are based solely on the results from the tested samples in the present study, for the range of parameters investigated. It has been demonstrated that the alumina material currently in use with the CC- 130 Hercules aircraft had the worst ballistic performance in comparison with all the other materials tested. Its areal density was also high compared to most of the other ceramics tested. Therefore its replacement would certainly give a much better performance in a lighter configuration. For the CERAMOR® and E-CERAMOR® materials, the influence of the density (or porosity) has been confirmed; the dense materials outperformed their porous counterparts as expected. 176 The MAP design, containing shaped nodes on the strike face of the CERAMOR®, E-CERAMOR® and SiC ceramics have improved the ballistic performance of the above materials, compared with their flat counterparts for the same areal densities. The addition of the PC spall cover film around the ceramic tiles and the diamond containing coating of the impact face did not always improve the ballistic behaviour of either the CERAMOR® or the SiC targets against the projectile used, their influence being not entirely understood. Some of the ceramic materials namely the B4C and the SiC exhibited explosive shattering upon impact, an effect that has to be taken into account in designing the entire protective armour panel for multiple hits protection. The performance of the thicker B4C tiles sometimes failed to prove their superiority compared to the thinner B4C tiles. An optimum ceramic thickness to balance its performance with its areal density should be chosen for any given threat. The addition of a PC film or a Kevlar® sheet on the impact face reduced the ejecta that could act as secondary fragments during an impact, a subject worth investigating if the threat posed by such ejecta is considered to be substantial. Some of the materials exhibited significant variations in the DOP results from one sample to another. The cause could be the inherent variability of the ceramics as well as other unforeseen factors. Further testing should be conducted to pinpoint some of these presently unknown factors. Based on the results of the testing performed in this research, it is suggested that entire panels should be tested against the same threat, in order to recommend 177 replacements for the obsolete, alumina based armour panels, currently in use on CC-130 Hercules aircraft. The most promising ceramics that should be inserted in the panels for future testing are the bare MAP SiC, the thin (8.5 mm) B4C and the CERAMOR®-MAP at areal densities lower than previously used alumina on the CC-130 Hercules aircraft. 178 CHAPTER 9: RECOMMENDATIONS Based on the results obtained in this project, the testing on complete armour panels made with the best performing ceramic materials and configurations in the present research is highly recommended. As a consequence of this research program, the materials proposed are the bare MAP SiC, the B4C ceramic, more desirable at the lower thickness (8.5 mm) and the CERAMOR®-MAP ceramics at areal densities lower than the alumina used on CC-130 Hercules aircraft. Since the mass of the ceramic tiles in the ballistic panels designed to defeat AP projectiles according to some authors represents between 75% (Robertson and Gotts11) and 80% (Lucuta69) by weight, the significant weight contribution of the ceramic indicates that by using the better performing ceramics with lower areal densities compared to the previously used alumina, the decrease in areal mass would be transferred in a high proportion to the entire panel. The other materials included in a stand-alone panel configuration beside the ceramic will usually form the front spall cover and the backing material. These materials may range from a simple ballistic fabric wrap to a complex layering system incorporating other ceramics, metal and organic composite materials, whose importance should not be underestimated, although their mass addition to the total system remains relatively low. 179 Due to the technology behind the manufacturing and the integration of the spall cover and the backing material with the ceramic layer, the panels to be tested have to be procured from the same manufacturer that supplied the selected ceramic material or the manufacturer that integrated this ceramic in the complex armour system, in the hope that - every ceramic-composite system that forms a particular type of panel is customised and optimised by its manufacturing company for the best performance. The testing of the proposed stand-alone panels has to be performed with the same ammunition used to test the alumina based reference panel. If no penetrations are obtained in the new configurations, the replacement of the obsolete CC-130 Hercules aircraft used panels would be highly advantageous. From a practical perspective, more DOP tests should be performed on other ceramics in existence on the market, in order to increase the data base for this type of tests. Further results on the already tested materials should also be performed to provide additional data for a complete statistical analysis. The test conditions have to be closely replicated in the subsequent testing for the results to be comparable. Some novel materials and shapes that are currently in various research stages of development could also be added to this data base, as soon as they become available. A number of heterogeneous materials with optimal compositions and structures might have remarkable ballistic performance, as suggested by Medvedovski . These materials might contain ceramic-metal compositions (cermets). The proportion of each part may vary in the thickness of the tile, with increased percentage of ceramic at the impact face for damaging the projectile, replaced by metal towards the back of the tile, for increased plasticity. 180 Dual ceramic tiles manufactured from high performance more expensive materials cladded on alumina based ceramic tiles should be studied in future testing. The projectile used for ballistic testing could be changed since for the same protection level, more than one projectile is recommended. Most probably, other projectile configurations and materials will interact differently with the ceramic samples. From this interaction, one might obtain a new classification of the best performing ceramics. Another important factor worth investigating by an armour manufacturer is the impact at lower than muzzle velocities, in order to eliminate the possibility of the shatter gap that can lead to armour penetration. This possibility might appear when the threat projectile is fired from a significant distance from the target to be protected, such as during the approach for landing or just after the take-off of an aircraft or helicopter. For entire, stand-alone panels, the multi hit capabilities are worth investigating given that the same area might be hit by more than one projectile or fragment. The enhanced edge shooting capabilities is an added subject of interest. The problem is still in existence and the research could be particularly rewarding. The improved edges of the armour panels should protect efficiently the entire area covered. Another area of interest for future research would be the better understanding of the projectile / panel impact phenomena. High velocity video recordings as well as various sensors embedded in different parts of the target (presumably not affecting its behaviour) could provide additional information to be used for future computer simulations, the inexpensive method for optimising next generation armour panels. 181 Ultimately, the actual manufacturing of armour systems may be designed to address a large number of factors not directly related to their ballistic performance such as the ease of fabrication, the comfort of wear and reduction of day-to-day "fatigue" damage, as suggested by Gotts21. The everyday manipulation and use of the ceramic based panels could slightly damage the system. Most probably not visible, the damage could accumulate over time decreasing the stopping capabilities of the armour. The potential research might compare new armour panels or entire systems with similar systems subjected to "fatigue" loading as well as with armours that have been in use in theatres of operations. Referring to theatres of operations where troops and equipment might be deployed, the missions tend to become more diversified rather than just war fighting; therefore the threats against the supplied armour would be more difficult to anticipate. As a consequence, it might be advantageous to manufacture in the future modular or multipurpose armours. This approach would add importance to the design and manufacturing aspects of the armour compared to just the fine tuning of the ballistic 2 n materials , as suggested by Iremonger and Gotts . As a consequence, a complete research program would include the ballistic testing of the final product in deployed situations on the actual equipment that has to be protected. 182 REFERENCES 1. Zukas J.A. et al."Impact Dynamics" John Wiley & Sons, Inc. 1982. 2. STANAG 4569 "Procedure for Evaluating the Protection Levels of Logistic and Light Armoured Vehicles for KE and Artillery Threats". 3. Iremonger M.J and Gotts P. L. "Practical Personal Ballistic Protection: Past, Present and Future", PASS 2002. 4. Anderson Ch. E. Jr. "Developing an Ultra-Lightweight Armor Concept", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 5. Phillips G. C. "A Concise Introduction to Ceramics", Van Nostrand Reinhold, 1991. 6. Matsalla D. Capt, "Modular Defence Panel System - Land Forces Technical Staff Programme Student Paper" The Royal Military College of Canada, 2004. 7. Shackelford J.F., Alexander W. - "The CRC Materials Science and Engineering Handbook", CRC Press, 1992. 8. Medvedovski E. "Silicone Carbide-Based Ceramics for Ballistic Protection" Ceramic Armor and Armor Systems Symposium, Tennessee, USA, April 2003. 9. Tresser R.E. "An Assessment of Low Cost Manufacturing Technology for Advanced Structural Ceramics and Its Impact on Ceramic Armor", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 10. Hazell P.J. et al. "The Penetration of Armour Piercing Projectiles through Reaction Bonded Ceramics" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 11. Robertson Colin, Gotts Philip "Review of Hard Materials and Their Use in Future Personal Armour Applications" PASS 2004. 183 12. Norwood J.D. "Interceptor Body Armor" Presentation at the EDGA 2nd Annual Conference on Lightweight Materials for Defence, Washington, D.C. July 28 2004. 13. Chu Henry et al. "Ballistic Properties of Pressureless Sintered SiC/TiB2 Composites" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 14. Buchar J. Et al."Numerical Simulations of the Ballistic Efficiency of Some Add on Armors", 21st International Symposium n Ballistics, Adelaide, South Australia, 2004. 15. Barnett R.L. "The Design - Testing Interface" Structural Ceramics and Testing of Brittle Materials, Illinois Institute of Technology Research, 1967. 16. Gooch W.A. Jr. "An Overview of Ceramic Armor Applications", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 17. Aceram - Materials and Technologies Inc. presentation brochure. 18. Gotts Phil and Pippa Kelly "Ballistic Testing: What the Results Mean?" PASS 2002. 19. Meyers M.A. "Dynamic Behavior of Materials" John Wiley & Sons Inc. 1994. 20. Robertson C.J. et al. "The Effective Hardness of Hot Pressed Boron Carbide with Increasing Shock Stress" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 21. Gotts P.L. et al "Determination of the Critical Components of a Complex Personal Armour Rigid Plate" PASS 2006. 22. Callister William D. Jr. "Materials Science and Engineering, An Introduction" Fourth Edition, John Wiley & Sons Inc. 1997. 23. Medvedovski Eugene "Ceramics for Personnel Armour Systems: Current Status and Future" PASS 2004. 24. Moshe Ravid et al. "Penetration Analysis of Ceramic Armor with Composite Material Backing" 19th International Symposium of Ballistics, 7-11 May 2001, Interlaken, Switzerland. 25. Normandia M.J. et al "A Comparison of Ceramic Materials Dynamically Impacted by Tungsten Carbide Spheres" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 184 26. Lucuta P.G. et al "Ultra-light Advanced Ceramic Armour System for Personal Ballistic Protection" PASS 2002. 27. Kaufmann, C et al. "Numerical Modeling of the Ballistic Response of Ceramic Tiles Supported by Serm-infinite metallic backings" PASS 2002. 28. Medvedovski E. "Armor Alumina Ceramics", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 29. Normandia M. J. "Ceramic Defense: Improving Ceramic Armor Performance" Ceradyne Inc. report, Costa Mesa, California, 2006. 30. Yadav S., Ravichandran G. "Penetration Resistance of Laminated Ceramic / Polymer Structures", International Journal of Impact Engineering 28 (2003) 557- 572. 31. Pickup Ian "The Correlation of Microstructural and Mechanical Characteristics of Silicone Carbide with Ballistic Performance" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 32. Renahan C.C. "Modelling the Ballistic Performance of Ceramic Armour Using Artificial Neuronal Networks" Master Thesis, The Royal Military College of Canada, 2006. 33. LaSalvia J.C. et al." A Simple Pre-ballistic Evaluation Methodology for Ceramics", 21st International Symposium n Ballistics, Adelaide, South Australia, 2004. 34. Jang J. et co "Failure of ceramic / fibre-reinforced plastic composites under hypervelocity impact loading", Journal of Materials Science 32,1997. 35. Hazell Paul J. and Iremonger Michael J. "The Numerical Analysis of Dynamically Loaded Ceramic: a Crack Softening Approach" Int. J. for Numerical Methods in Engineering 2000; 48 pg. 1037-1053. 36. Lu Jin et al "A Preliminary Study on Damage Wave in Elastic-brittle Materials" Int. J. of Damage Mechanics, vol. 14, April 2005. 37. Zhai Jun et al "Micromechanical Simulation of Dynamic Fracture Using the Cohesive Finite Element Method" J. of Engineering Materials and Technology, April 2004, vol. 126, pg 179-191. 38. Medvedovski E. "Armor Silicon Carbide-Based Ceramics" 27th International Conference on Advanced Ceramics and Composites, Florida, USA, 2003. 185 39. LaSalvia J.C. et al "Sphere Impact Induced Damage in Ceramics: II. Armor-grade B4C and WC" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 40. McCuiston R. Et al. "Fabrication and Simulation of Random and Periodic Macrostructures" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 41. Normandia M.J., Gooch W.A. "An Overview of Ballistic Testing Methods of Ceramic Materials", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 42. Moshe Ravid et al. "Penetration Analysis of Ceramic Armor Backed by Composite Materials" Ceramic Armor and Armor Systems Symposium, Tennessee, USA, April 2003. 43. Lopez-Puente J. at al "The Effect of the Thickness of the Adhesive Layer on the Ballistic Limit of Ceramic / Metal Armours. An Experimental and Numerical Study", International Journal of Impact Engineering 32 (2005) 321-336. 44. St-Denis CD., "The Fabrication and Ballistic Performance of Pressureless Sintered Silicon Carbide Ceramic Armour Tiles", Master's Thesis, The Royal Military College of Canada, 2001. 45. Bryn James "Practical Issues in Ceramic Armor Design", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 46. Woolmore N.J. et al. "An investigation into Fragmenting the 14.5 mm BS41 Armour Piercing Round by Varying a Confined Ceramic Target Setup" Ceramic Armor and Armor Systems Symposium, Tennessee, USA, April 2003. 47. Madhu V. at al "An Experimental Study of Penetration Resistance of Ceramic Armour Subjected to projectile Impact", International Journal of Impact Engineering 32 (2005) 337-350. 48. Skaggs S.R. "A Brief History of Ceramic Armor Development" 27th International Conference on Advanced Ceramics and Composites, Florida, USA, 2003. 49. Flinders, Marc et co. "High-Thoughness Silicon Carbide as Armor", Journal of American Ceramics Society, 88, 2217-2226, 2005. 50. LaSalvia J.C. et al "Sphere Impact Induced Damage in Ceramics: I. Armor-grade SiC and TiB2" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 51. Grady D.E. "The Spall Strength of Condensed Matter", J. Mech. Phys. Solids, Vol. 36 No. 3, 353-384,1988. 186 52. Hazell P.J., Iremonger M.J. "Ceramic Armour Penetration", Lightweight Armour Systems Symposium, 1999. 53. Blazynski T.Z. - "Materials at High Strain Rates", Elsevier Applied Science, 1987. 54. Ritter J.E. - "Erosion of Ceramic Materials", Trans Tech Publications, 1992. 55. Shockey D.A., Marchand A.H. - "Failure Phenomenology of Confined Ceramic Targets and Impacting Rods", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 56. Wells J.M. et al. "Visualisation of Ballistic Damage in Encapsulated Ceramic Armor Targets" 20th International Symposium on Ballistics, Florida, USA, 2002. 57. Anderson C. E., Walker J.D. "An Analytical Model for Dwell and Interface Defeat", International Journal of Impact Engineering 31 (2005) 1119-1132. 58. Lundberg P., Lundberg B. "Transition between Interface Defeat and Penetration for Tungsten Projectiles and Four Silicon Carbide Materials", International Journal of Impact Engineering 31 (2005) 781-792. 59. Holmquist T.J., Johnson G.R. "Modeling Prestressed Ceramic and Its Effect on Ballistic Performance", International Journal of Impact Engineering 31 (2005) 113-127. 60. Pickup I.M. et al. "The Effect of Coverplates on the Dwell Characteristics of Silicon Carbide Subject to Impact", 21st International Symposium n Ballistics, Adelaide, South Australia, 2004. 61. Stepp D.M. "Damage Mitigation in Ceramics: Historical Developments and Future Directions in Army Research", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 62. Lundberg P. et al "Impact of Conical Tungsten Projectiles on Flat Silicon Carbide Targets: Transition from Interface Defeat to Penetration", International Journal of Impact Engineering 32 (2006) 1842-1856. 63. Sarva S. at al "The Effect of Thin Membrane Restraint on the Ballistic Performance of Armor Grade Ceramic Tiles", International Journal of Impact Engineering 34 (2007) 277-302. 64. Moynihan T.J. et al. "Analysis of the Shatter Gap Phenomenon in a Boron Carbide / Composite Laminate Armor System" 20th International Symposium on Ballistics, Florida, USA, 2002. 187 65. Robertson C. et al. "Update on Practical Non Destructive Testing Methods for In- Service QA of Ceramic Body Armour Plates" PASS 2006. 66. Bryn James "Depth of Penetration Testing", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 67. Krell A., Strassburger E. "High-purity Submicron a-A1203 Armor Ceramics Design, Manufacture, and Ballistic Performance", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 68. Parameswaran V. et al. "A New Approach for Improving Ballistic Performance of Composite Armor" Experimental Mechanics, vol. 39, no. 2, June 1999. 69. Lucuta P.G. et al "Present and Future Trends in the Development of CERAMOR Systems for Personal Protection" PASS 2006. 70. Lucuta P.G. - Private Communications. 71. Strassburger E. et al. "Ceramic Armor with Submicron Alumina Against Armor Piercing Projectiles", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 72. Nunn S.D. et al. "Improved Ballistic Performance by Using a Polymer Matrix Composite Facing on Boron Carbide Armor Tiles" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 73. Nemat-Nasser S. et al. "Novel Ideas in Multi-Functional Ceramic Armor Design", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 74. Adams M.A. "Theory and Experimental Test Methods for Evaluating Ceramic Armor Components", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. 75. Horsfall I. "Glass Ceramic Armour Systems for Light Armour Applications" 19th International Symposium of Ballistics, Interlaken, Switzerland, 7-11 May 2001. 76. Darin Ray et al "Effect of Room-Temperature Hardness and Toughness on the Ballistic Performance of SiC-Based Ceramics" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 77. Gilde G.A. and Adams J.W. "Processing and Ballistic Performance of A1203 / TiB2 Composites" International Conference on Advanced Ceramics and Composites, The American Ceramic Society, 2005. 188 Robertson, C. and Hazell PJ. "Resistance of Different Ceramic Materials to Penetration by a Tungsten Carbide Cored Projectile" Ceramic Armor and Armor Systems Symposium, Tennessee, USA, April 2003. Robertson, C. and Hazell P.J. "Resistance of Silicon Carbide to Penetration by a Tungsten Carbide Cored Projectile" Ceramic Armor and Armor Systems Symposium, Tennessee, USA, April 2003. Peron P.F. "Ballistic Development of Tungsten Carbide Ceramics for Armor Applications", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. Gooch W.A. et al. "Ballistic Development of U.S. High Density Tungsten Carbide Ceramics", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. Murat Vural et al. "Ballistic Performance of Alumina Ceramic Armors", Proceedings of the Ceramic Armor Materials by Design Symposium, Hawaii, 2001. Underhill Ross - Private Communications. D 7136/D 7136M - 05 - Standard Test Method for Measuring the Damage Resistance of a Fiber-Reinforced Polymer Matrix Composite to a Drop-Weight Impact Event, April 2005. Coffelt R.A "Rifle Level Body Armor Capabilities, Use and Care", PASS 2006. Jane's Ammunition Handbook 2005 - 2006. www.westsystem.com Munitions Experimental Test Centre - Ballistic Test Report number CEEM 06/07- 074, August 31 2006. 189 VITA Name: Cezar Constantin Craciun Place / date of Birth: Romania, 1 March 1970 Education: Military High School, Romania, 1989 Military Technical Academy, Bachelor in Engineering, Specialisation in Obstacles, Landmines, Demolitions and Camouflage, 1996 Military Technical Academy, Advanced Studies in Explosives, Special Combustibles and Pyrotechnics, 1998 Military Postings: Combat Engineer's Research Centre 1996-1997 Combat Engineer's Special Equipment Testing Centre 1997-1999 Combat Engineer's Explosives, Demolition, Demining and EOD Training Centre 1999-2002 EOD Office, Land Forces Headquarters 2002-2003 Missions Abroad: Demining, EOD and Booby traps Course - Turkey, 1998 Romanian Engineering Battalion, SFOR, 2000 EOD Training Course, the Netherlands, 2001 EOD Officer, Dutch Battle Group, SFOR, 2001 Canadian Experience: Mining Resources Engineering Ltd, Kingston, 2003-2004 The Royal Military College of Canada, Department of Mechanical Engineering, 2004-2006 Canadian Explosives Research Laboratory, 2007-present 190 APPENDIX: Table containing DOP test results and efficiency calculated values. 191 Table 19: Appendix-DOP test results and efficiency calculated values. No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt & no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] CC-130 Hercules bare tiles - Alumina 1 H-1 297 3.47 90 3.30 9.50 931.7 193 1.89 0.66 1.25 111.4 few PC cracks 2 H-2 297 3.47 90 3.30 9.50 934.9 196 1.58 0.55 0.87 133.6 few PC cracks 3 H-3 297 3.47 90 3.30 9.50 936.1 193 1.89 0.66 1.25 111.4 more PC cracks Average= 297.0 3.47 3.30 934.2 194.0 1.79 0.62 1.12 118.8 Stddev= 0.00 0.00 0.00 2.27 1.73 0.18 0.06 0.22 12.9 CC-130 Hercules tiles with Kevlar - Alumina 4 HK-1 305.6 3.47 90 3.40 10.00 932.3 190 2.10 0.73 1.54 100.5 5 HK-2 308.1 3.47 90 3.42 10.00 930.5 189 2.20 0.77 1.69 95.9 6 HK-3 299.2 3.47 90 3.32 10.00 933 192 1.90 0.66 1.26 111.1 Average= 304.3 3.47 3.38 931.9 190.3 2.07 0.72 1.49 102.5 Stddev= 4.59 0.00 0.05 1.29 1.53 0.15 0.05 0.22 7.8 Porous CERAMOR® flat, bare 7 F-1 179.2 54.5 3.29 934.5 135 2.80 AP core cracked app 3mm gap 8 F-2 179 54.5 3.28 930.3 143 2.51 between PC blocks app 2mm gap 9 F-3 178.9 54.5 3.28 934.7 154 2.10 between PC blocks 10 F-4 180.3 54.5 3.31 936.4 171 1.46 Average= 179.4 3.29 934.0 150.8 2.2 Stddev= 0.65 0.01 2.59 15.59 0.58 192 No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt &no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] Porous E-CERAMOR® flat (diamond coated) 11 FC-1 169.9 54.5 3.12 935.7 180 1.20 12 FC-2 180.9 54.5 3.32 937 57 5.61 3mm gap between 13 FC-3 179.2 54.5 3.29 930.5 151 2.21 PC blocks 14 FC-4 181.3 54.5 3.33 935 124 3.16 Average= 177.8 3.26 934.6 128.0 3.0 Stddev= 5.36 0.10 2.82 52.57 1.89 Avg = 151.7 without FC-2 value Porous CERAMOR® flat, sandwiched in PC 15 FP-1 188.2 54.5 3.45 935.3 165 1.61 16 FP-2 192.1 54.5 3.52 936.1 135 2.61 2-3mm gap 17 FP-3 190.6 54.5 3.50 930.2 150 2.11 between PC blocks 18 FP-4 188.1 54.5 3.45 939.9 179 1.12 Average= 189.8 3.48 935.4 157.3 1.9 Stddev= 1.95 0.04 3.99 18.98 0.64 Porous E-CERAMOR® flat, sandwiched in PC core between PC 19 FCP-1 191.1 54.5 3.51 940.8 144 2.31 blocks; gap AP broken 20 FCP-2 188.7 54.5 3.46 937 171 1.40 21 FCP-3 193.2 54.5 3.54 936.3 155 1.91 core stop between 22 FCP-4 191.5 54.5 3.51 938.3 144 2.31 PC blocks =>gap Average= 191.1 3.51 938.1 153.5 2.0 Stddev= 1.86 0.03 1.98 12.77 0.43 193 No Tile type Tile Areal Impact Equiy Equiv Quality Crit Density Area Thick DOP LOP Remarks crt &no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] Porous CERAMOR-MAP®, bare layered cracks in 23 B-1 182.7 54.5 3.35 935.3 167 1.59 ceramic AP frag; molten PC; 24 B-2 182.8 54.5 3.35 935.3 74 4.94 low ceram dispers 25 B-3 177.6 54.5 3.26 938.9 178 1.23 AP frag in 2; molten 26 B-4 189.1 54.5 3.47 935.4 92 96 4.01 PC Average= 183.1 3.36 936.2 127.8 2.9 Stddev= 4.71 0.09 1.78 52.39 1.81 Porous E-CERAMOR-MAP® 27 BC-1 184.9 54.5 3.39 934.6 171 1.43 28 BC-2 182.1 54.5 3.34 936.2 167 1.59 29 BC-3 189.7 54.5 3.48 932.2 181 1.04 Average= 185.6 3.40 934.3 173.0 1.4 Stddev= 3.84 0.07 2.01 7.21 0.28 Porous CERAMOR-MAP®, sandwiched in PC AP frag; curved 30 BP-1 194.9 54.5 3.58 933.6 92 105 3.59 traject; molten PC PC cracks; AP 31 BP-2 194.9 54.5 3.58 932.3 89 102 3.69 curved trajectory 32 BP-3 199.1 54.5 3.65 934.8 158 158 1.76 PC cracks AP frag; curved 33 BP-4 197 54.5 3.61 937.2 52 65 4.89 trajectory; Cu in molten PC Average= 196.5 3.61 934.5 97.8 107.5 3.5 Stddev= 2.01 0.04 2.08 44.09 38.27 1.29 194 No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt &no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [mfe] [mm] [mm] Eq22 Eq23 Eq24 [mm] Porous E-CERAMOR-MAP®, sandwiched in PC low adherence front 34 BCP-1 193.5 54.5 933 180 180 1.06 3.55 PC film 35 BCP-2 195.2 54.5 3.58 936.2 71 77 4.53 curved trajectory 36 BCP-3 198 54.5 3.63 934.5 96 110 3.36 curved trajectory AP core fragmented 37 BCP-4 200.2 54.5 3.67 97 130 2.67 932.8 curved trajectory Average= 196.7 3.61 934.1 111.0 124.3 2.9 Stddev= 2.97 0.05 1.58 47.55 43.12 1.45 Porous CERAMOR-MAP®, high density, cast, bare 38 DB-5 192.5 54.5 3.53 935.5 27 6.3 Cu in PC AP core fragment; 39 DB-6 189.7 54.5 3.48 939.6 53 5.49 Cu in molten PC 40 DB-7 187 54.5 3.43 938.9 82 4.55 curved trajectory curved trajectory; 41 DB-8 156.9 54.5 2.88 937.9 88 5.17 Cu in molten PC Cu+ceram+AP in 42 DB-9 201.1 54.5 3.69 933.9 16 6.39 PC; cracked PC Average= 185.4 3.40 937.2 53.2 5.6 Stddev= 16.81 0.31 2.39 32.06 0.78 E-CERAMOR-MAP®, high density, cast, sandwiched in PC 43 DBCP-1 206.4 54.5 3.79 933.5 55 4.98 curved trajectory PC cracked; NO AP 44 DBCP-2 205.3 54.5 3.77 940.7 0 6.78 core in PC 45 DBCP-3 188 54.5 3.45 935.3 78 4.67 46 DBCP-4 204.3 54.5 3.75 934.3 0 6.81 tile shattered Average= 201.0 3.69 936.0 33.3 5.8 Stddev= 8.71 0.16 3.25 39.53 1.14 195 No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt & no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] Boron carbide (Ceradyne), thick (0.443 in -11.2 mm) dent only, cracks in 47 CEK1 240.1 2.52 90 2.67 11.20 928.9 0 18.84 9.06 170.62 11.2 PC 48 CEK2 239 2.52 90 2.66 11.20 930.8 32 15.98 7.69 122.84 13.2 cracks in PC dent only, cracks in 49 CEK3 242.4 2.52 90 2.69 11.20 929.0 0 18.84 9.06 170.60 11.2 PC 50 CEK4 238.5 2.51 90 2.65 11.20 924.2 16 17.41 8.39 146.09 12.1 cracks in PC Average= 240.0 2.52 2.67 11.20 928.2 12.0 17.77 8.55 152.54 11.9 Stddev= 1.73 0.00 0.02 0.00 2.82 15.32 1.37 0.65 22.93 1.0 Boron carbide (Ceradyne), thin (0.333 in - 8.5 mm) cracks from impact 51 CEN11 196.1 2.52 90 2.18 8.50 935.0 20. 22.47 10.79 242.56 9.4 point cracks from impact 52 CEN12 198.2 2.51 90 2.20 8.50 934.4 25 21.88 10.53 230.52 9.6 point cracks from impact 53 CEN13 187 2.51 90 2.08 8.50 925.4 34 20.82 10.03 208.89 10.1 point cracks from impact 54 CEN14 194.4 2.52 90 2.16 8.50 930.2 24 22.00 10.57 232.43 9.6 point Average= 193.9 2.52 2.15 8.50 931.3 25.8 21.79 10.48 228.60 9.7 Stddev= 4.87 0.00 0.05 0.00 4.45 5.91 0.70 0.32 14.17 0.3 Silicone carbide (Morgan), flat 55 MF16 139.6 3.12 52 2.68 8.61 935.8 0 24.50 9.51 232.97 8.6 dent only 56 MF17 139.8 3.11 52 2.69 8.63 934.1 12 23.05 8.96 206.46 9.2 57 MF23 134.7 3.12 52 2.59 8.31 934.4 27 22.16 8.59 190.42 9.5 58 MF25 139.2 3.12 52 2.68 8.57 932.3 0 24.61 9.54 234.69 8,6 dent only Average= 138.3 3.12 2.66 8.53 934.2 9.8 23.58 9.15 216.13 9.0 Stddev= 2.43 0.00 0.05 0.15 1.43 12.82 1.18 0.46 21.47 0.5 196 No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt &no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] Silicone carbide (Morgan), flat diamond coated and sandwiched in PC 59 MFDP 19 132.1 3.10 52 2.54 8.19 932.4 12 24.30 9.48 230.32 8.7 60 MFDP 20 135.2 3.11 52 2.60 8.36 933.0 17 23.20 9.03 209.44 9.1 61 MFDP 22 139.6 3.12 52 2.68 8.59 929.9 0 24.55 9.51 233.49 8.6 dent only 62 MFDP 24 134.1 3.13 52 2.58 8.25 932.6 16 23.64 9.15 216.33 8.9 Average= 135.3 3.12 2.57 8.27 932.7 15.0 23.71 9.22 218.70 8.9 Std dev = 3.17 0.01 0.06 0.18 1.39 7.80 0.62 0.24 11.41 0.2 MAP Silicone carbide Morgan) 63 MB 31 135.8 3.12 52 2.61 8.38 930.8 24 22.31 8.66 193.29 9.5 64 MB 32 132 3.11 52 2.54 8.15 933.5 0 25.89 10.06 260.36 8.2 dent only 65 MB 33 135.7 3.11 52 2.61 8.38 927.8 0 25.17 9.78 246.27 8.4 dent only 66 MB 35 135.7 3.12 52 2.61 8.36 932.1 0 25.23 9.78 246.84 8.4 dent only Average= 134.8 3.12 2.59 8.32 931.1 6.0 24.65 9.57 236.69 8.6 Std dev = 1.87 0.00 0.04 0.11 2.43 12.00 1.59 0.62 29.66 0.6 MAP Silicone carbide Morgan), diamond coated, sandwiched in PC 67 MBDP 27 135.7 3.11 52 2.83 9.11 930.4 9 22.17 8.63 191.38 9.5 68 MBDP 28 136.2 3.11 52 2.84 9.13 935.4 23 20.59 8.01 164.88 10.2 69 MBDP 29 135 3.12 52 2.82 9.02 931.4 26 20.51 7.95 163.06 10.3 70 MBDP 30 136.1 3.12 52 2.83 9.07 930.5 0 23.27 9.02 209.91 9.1 dent only Average= 135.8 3.12 2.83 9.08 931.9 14.5 21.63 8.40 182.31 9.8 Std dev = 0.54 0.01 0.01 0.05 2.39 12.18 1.33 0.51 22.50 0.6 197 No Tile type Tile Areal impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt & no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] CERAMOR-MAP®, thick, high density 71 ABK36 173.8 54.5 3.19 930.3 29 6.91 72 ABK38 173.4 54.5 3.18 925.9 31 6.85 73 ABK39 173.9 54.5 3.19 928.3 17 7.36 74 ABK43 174.2 54.5 3.20 927.8 14 7.46 Average= 173.8 3.19 928.1 22.8 7.1 Stddev= 0.33 0.01 1.83 8.5.0 0.31 E-CERAMOR-MAP®, thick, sandwiched in PC, high density 75 ABDPK 37 173.3 54.5 3.39 930.7 31 6.42 76 ABDPK41 174.5 54.5 3.41 927.4 0 7.48 dent only 77 ABDPK42 173 54.5 3.39 928.8 18 6.89 78 ABDPK45 174.1 54.5 3.41 926.6 0 7.50 dent only Average= 173.7 3.39 929.7 12.3 7.1 Stddev= 0.69 0.01 1.78 15.11 0.52 CERAMOR-MAP®, thin, high density 79 ABN46 132.2 54.5 2.43 930.6 40 8.53 80 ABN47 132.9 54.5 2.43 930.6 56 7.72 81 ABN48 131 54.5 2.40 928.5 40 8.61 82 ABN49 132.9 54.5 2.44 927.0 70 7.00 Average= 132.3 2.42 929.2 51.5 8.0 Stddev= 0.90 0.01 1.75 14.46 0.76 198 No Tile type Tile Areal Impact Equiv Equiv Quality Crit Density Area Thick DOP LOP Remarks crt &no mass density velocity thick mass factor thick [grams] [g/cc] [cmA2] [g/cmA2] [mm] [m/s] [mm] [mm] Eq22 Eq23 Eq24 [mm] E-CERAMOR-MAP®, thin, sandwiched in PC, high density 83 ABDPN 50 130.7 54.5 2.61 933.7 70 6.53 84 ABDPN 51 132.3 54.5 2.64 928.3 67 6.59 85 ABDPN 52 121.9 54.5 2.45 930.7 186 1.23 86 ABDPN 53 132.2 54.5 2.65 933.1 47 7.49 Average= 129.3 2.64 931.7 61.3 6.9 Stddev= 4.97 0.09 2.47 63.16 2.85 PC backing with 2x8mm Al discs in front 87 DOP1 1.21 918.7 198 88 DOP 2 1.21 933.8 201 Average= 926.2 199.5 Stddev= 10.6 2.1 Bare backing polycarbonate Cu jacket remained 89 Poly-1 935.9 211 on the core Cu jacket remained 90 Poly-2 937.2 211 on the core Average= 936.6 211.0 Std dev = 0.9 0.0 CC-130 Hercules, armour panel 91 HP-1 935.3 defeated 92 HP-2 941.7 defeated ACE RAM - porous CERMOR-MAP® panel 93 NP-1 936.5 defeated 94 NP-2 934 defeated