Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Long term soft tissue fixation and mechanical reliability of a ceramic housing for a new radio frequency… Hori, Bryan David 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2006-0505.pdf [ 44.28MB ]
JSON: 831-1.0058936.json
JSON-LD: 831-1.0058936-ld.json
RDF/XML (Pretty): 831-1.0058936-rdf.xml
RDF/JSON: 831-1.0058936-rdf.json
Turtle: 831-1.0058936-turtle.txt
N-Triples: 831-1.0058936-rdf-ntriples.txt
Original Record: 831-1.0058936-source.json
Full Text

Full Text

Long Term Soft Tissue Fixation and Mechanical Reliability of a Ceramic Housing for a new Radio Frequency Transmitter by Bryan David Hori B.Eng, Carleton University, 2003 A THESIS SUBMITTED IN PARTIAL FULLFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Chemical and Biological Engineering) THE UNIVERSITY OF BRITISH COLUMBIA July 2006 © Bryan David Hori, 2006 ABSTRACT This project was focused on the design and suitability of the housing component of a new telemetry device to be implanted into young Steller sea lions. The housing's suitability is assessed on its long term performance for stable implantation for lifetimes of up to 30 years. An aluminum oxide ceramic material is selected as the housing material as it meets radio frequency, biocompatibility and strength requirements. The housing design consists of a solid base and porous top surface with an inner cavity for electronics potted in epoxy. Aptness of the design for implantation involved investigating the response of the housing to biological and mechanical factors. Biological response was examined by assessing tissue fixation of porous aluminum oxide. Disc implants (36), with a top porous surface of pore size 32 pm and thicknesses of 0.5 mm and 1.0 mm, were sub-dermally implanted into the backs of young rabbits. Due to surgical complications, 33 tags were inserted under the cutaneous trunci muscle, while the remaining were inserted above it. A favourable tissue reaction was assessed in all cases. All implants migrated with the skin growth a distance of 4.69 + 1.48cm. Half of the implants moved an additional 1.74 ± 1.93cm caused by a combination of externally applied forces and loose tissue attachment. Loose tissue attachment was a result of implantation into subcutaneous fat tissue and the inability of implant encapsulated tissue in integrating with the fat layers. The response of the housing to mechanical factors was examined by applying loading conditions (cyclic fatigue, compression, puncture and impact) that simulate what is expected in-service. Implants were able to resist fracture due to compression and puncture while impact suitability is achieved when considering energy absorption by the surrounding tissue. The derived housing design has good potential for future implantation into Steller sea-lions. Further research is required to examine implant fixation and migration in dermal tissue compared to subcutaneous tissue. As the implants will move from the insertion location in growing skin, cranial skin growth patterns should be considered prior to implantation into Steller sea lions. ii T A B L E O F C O N T E N T S Abstract 11 Table of Contents 111 List of Tables v List of Figures v l Nomenclature X l Acknowledgements x " CHAPTER I: Introduction 1 CHAPTER II: Background 3 2.1 Introduction 3 2.2 Biomaterials 3 2.2.1 Polymers 5 2.2.2 Metals 10 2.2.3 Ceramics 13 2.2.4 Composites 15 2.3 Tissue Reaction 16 2.3.1 Inflammation and the Foreign Body Response 16 2.3.2 Porous materials and implant migration 25 2.4 Mechanical Properties of Ceramics 28 2.4.1 Fatigue 31 2.4.2 Impact 33 2.4.3 Hertzian Contact 38 2.4.4 Effect of Porosity on Mechanical Properties 39 2.4.5 Effect of Biological Environment on Material Strength 41 CHAPTER III: Design 47 CHAPTER IV: Tissue Attachment and Implant Migration 5 2 4.1 Introduction 52 4.2 Materials and Methods 54 4.2.1 Manufacturing of Alumina samples 54 4.2.2 Animal Surgery Procedures 56 4.2.3 Tissue Adhesion Procedures 60 4.2.4 Histology 63 4.2.5 Implant movement tracking 65 4.3 Results and Discussion 65 4.3.1 Manufacturing of Alumina Samples Results and Discussion 65 4.3.2 Experimental Results and Discussion 72 4.4 Conclusions 93 4.5 Future Work 9 4 CHAPTER V : Mechanical Durability 95 5.1 Introduction 95 iii 5.2 Materials and Methods 97 5.2.1 Alumina Housing Fabrication 97 5.2.2 Test Apparatus Fabrication 98 5.2.3 Test Procedures 100 5.3 Results and Discussion m 5.3.1 Fatigue Results 112 5.3.2 Puncture Results 119 5.3.3 Impact Results 124 5.4 Conclusions 132 5.5 Future Work 134 CHAPTER VI: Overall Conclusions and Recommendations for Future Work 136 References ^38 Appendix I: Engineering Drawings Appendix II: Animal Care Forms 156 Appendix III: Histology Report 1 6 2 Appendix IV: Tissue Attachment Loading Curves ....169 Appendix V: Housing Fracture Pictures " ° iv LIST OF TABLES 2-1: Mechanical Properties of Selected Polymers [5, 12] 6 2-2: Mechanical Properties of Selected Metals: Titanium alloys [5], 316L Stainless Steel [7] and CoCr alloys [7] 11 2-3: Mechanical Properties of Selected Bioceramics [27] 13 2- 4: Reported strength reductions of alumina placed in different environments 42 3- 1: Material Selection Design Table 50 4- 1: Reported tissue attachment forces 54 4-2: General info on Rabbits used in Experiments (SQ - Subcutaneous, ID - Intradermal) 57 4-3: Skin growth movement tracked using a tattoo marker on the rabbit's skin. The mean migration distance is used to adjust implant migration distance in Table 4-4. CV -Caudal Ventral, C - Caudal, V - Ventral 73 4-4: Implant Migration Table. The distance that each implant moved was adjusted using the determined skin growth from Table 4-3. For pore layer face, up refers to a pore layer facing the skin and down refers to a pore layer facing muscle tissue. Listed next to the implant location is the thickness of the pore layer 74 4-5: Mean adjusted migration distances with standard deviation. The sample size is shown in brackets 75 4-6: Recorded peak pull-out force for each alumina disc. Results for rabbit 9 and 12 are omitted due to different implantation position ..79 4-7: Summary of peak forces for different implant orientations and pore layer thickness. The sample size is shown in brackets for each case 80 4-8: Mean peak pull-out forces for the different implantation locations. No statistical difference was observed (p < 0.5). The sample size is shown in brackets for each case 81 4- 9: Tissue tensile break loads from the literature 85 5- 1: Bite force for various species 105 5-2: Reported values for human skull fracture 109 5-3: Recorded forces and stresses from compression tests on six specimens 112 5-4: Pass/fail results from cyclic fatigue tests run until 70 000 cycles 117 5-5: Peak force data for initiation of the first crack and for the maximum load at total fracture ..122 5-6: Drop-weight Impact Test Results 124 v LIST OF FIGURES 2-1: Material Stress (s) - Strain (e) Curve. A - Brittle Material, B - Ductile Material 29 2-2: Typical S-N curve 32 2-3: Stress state during contact loading. During the loading phase, material is compressed beneath the indenter and material is 'pulled' towards the indenter creating tensile . stresses at the surface. During the unloading phase, residual stresses build up as the material attempts to return to its initial state. This creates a tensile stress below the area of indentation and compressive stress on the surface [105, 106] 38 2-4: Introduction of a pore or hole distorts the stress lines of a component. Force lines become thicker at the apex of the pore and larger stress fields result [32] 41 2-5: Stress corrosion crack growth diagram. No crack growth occurs until a specific stress intensity level is reached. Region A represents the crack growth rate.governed by reactions with the environment. Region B shows a steady crack growth velocity limited by the rate of diffusion of environmental molecules to the crack tips. Region C is the stage where sub critical cracks have reached an unstable size and the failure toughness (for an inert environment) is reached 44 2-6: Bond rupture at crack tip from environmental species 45 2- 7: Bond rupture of silica via water. The water provides a lower energy path to bond breakage during stretching of the silica bonds [124] 45 3- 1: 3D Solid View of Electronics Enclosure : 47 3-2: Outline of the housing's longitudinal cross section 48 3-3 Side view of young Steller sea-lion skull with a mock implant 49 3- 4 Top view of young Steller sea-lion skull with a mock implant 50 4- 1: Alumina disc die and mold for axial pressing 55 4-2: Design matrix showing the locations of all implants in the back of the rabbits. The top of the blocks is where the head is and the bottom is where the tail is. The bold names indicate those that were sent for histological examination 58 4-3: Template aligned to the dorsal spinous process of the 2nd thoracic vertebrae. All implants were inserted according to this template 59 4-4: Tissue block extraction from rabbits post mortem 60 4-5: Experimental setup for implant tissue adhesion tests 61 4-6: Force gauge attached to implanted alumina discs via force grips 62 4-7: Extracted alumina disc with surrounding fat tissue 62 4-8: Aluminum Oxide fabricated discs. Diameter of 17.5mm and 4mm thickness 66 4-9: Fabrication of porous and solid layered disc produced a non-uniform thickness with a higher densification of the solid layer (bottom) 67 vi 4-10: SEM view of the porous surface 67 4-11: SEM view of the solid/smooth surface 68 4-12: Graph of pore size variation with applied load during axial pressing of the powder. Only small differences in pore size were observed 69 4-13: Surface pore viewed under SEM (800x) 71 4-14: Surface pores viewed under SEM (250x) 7 1 4-15: High magnification (4500x) view of the surface. Alumina particles of many different sizes are observed with small gaps in between. Higher temperatures at longer sintering periods can reduce the sizes of these gaps 72 4-16: Cross section view of the tissue around the implants. The implant pouch is at the bottom of the image as indicated. Magnification is 100x 76 4-17: Relation between migration and tissue attachment for migrated and non-migrated implants. The dashed line is for migrated implants and the solid line is for non-migrated implants. (SE equals standard error of relation) 77 4-18: Relation between migration and tissue attachment showing results for different pore layer thicknesses. The dashed line is for 0.5mm data points and the solid is for 1mm data points 77 4-19: Migration vs. Peak pull-out force graph. For the case of migrated samples, peak pull-out force increased with migration distance (p < 0.05). The solid line represents the trend line for migrated implants only 84 4-20: Crystalline debris is shown adjacent to where the implant was located. Debris is both intracellular and extracellular. Magnification is 100x 67 4-21: Relation between implant migration and the encapsulation thickness (p < 0.8). The dashed line represents the trend line for the migrated discs while the solid line represents the non-migrated discs 88 4-22: Relation between implant migration and the amount of crystalline debris present around implants (p < 0.18). The dashed line represents the trend line for the migrated discs while the solid line represents the non-migrated discs 89 4- 23: Bite marks on the backs of rabbits 9 u 5- 1: Alumina housings for mechanical reliability tests 98 5-2: Test apparatus. A - Shield, B - Pressure plate, C - Housing holder, D - Puncture head 99 5-3: Actual in-service layout of implanted housing. The soft tissue surrounding the implant provides some dampening of the applied loads. The performed mechanical experiments do not simulate this, thus the results are conservative estimates 100 5-4: Cyclic fatigue test setup 103 5-5: Dynamic fatigue test setup 104 vii 5-6: Dynamic fatigue test setup. The alumina housing is shown in compression .104 5-7: Puncture test setup 107 5-8: Drop-weight apparatus 111 5-9: Weibuli diagram of compression strength tests 113 5-10: Probability of failure of the housing versus the applied stress. Applied stresses that are 40% of the characteristic strength will give <5% probability of failure .113 5-11: Failure of a sample housing under compression. Fracture occurred in one of the side walls and correlates well with what is predicted in Finite Element models 115 5-12: Load-time diagram for Puncture Experiments for all 6 tests .120 5-13: Weibuli diagram for puncture stresses 120 5-14: Top surface fracture of a specimen. A ring crack can be seen around the area of puncture with lateral cracks propagating outwards 121 5-15: Simplified representation of the housing under impact when implanted. The underlying bone and soft tissue act as springs that can absorb some of the energy from impact. The amount absorbed is related to the stiffness of the materials 127 5-16: Stress-strain curves for ductile (a) and brittle materials (b). The amount of energy a material can absorb before fracture is related to the area under the curves [206] 128 5-17: Fracture of housing #8 130 A1-1: Alumina disc mold die 151 A1-2: Alumina disc mold press top 151 A1-3: Alumina disc mold press bottom 152 A1-4: Housing Bottom three view drawing 153 A1-5: Housing Top three view drawing 153 A1-6: Pressure Plate .154 A1-7: Puncture Head 154 A1-8: Housing Holder 155 A1-9: Shield 155 A3-1: Cross section view of the tissue around the implants. The implant pouch is at the bottom of the image as indicated. Magnification is 20x 167 A3-2: Cross section view of a control tissue block. No inflammation can be observed Magnification is 20x 167 A3-3: Crystalline debris at 200x magnification. Some debris is in the cytoplasm of the macrophage while some is outside 168 viii A4-1: Tissue Attachment Force for 0.5mm pore layer implants 169 A4-2: Tissue Attachment Force for 1 mm pore layer implants 169 A4-3: Tissue Attachment Force for Front Left Location implants 170 A4-4: Tissue Attachment Force for Back Left Location implants 170 A4-5: Tissue Attachment Force for Front Right Location implants 171 A4-6: Tissue Attachment Force for Back Right Location implants 171 A4-7: Tissue Attachment Force for Pore Up implants 172 A4-8: Tissue Attachment Force for Pore Down implants 172 A5-1: Compression Fracture Picture 1 173 A5-2: Compression Fracture Picture 2 173 A5-3: Compression Fracture Picture 3 174 A5-4: Compression Fracture Picture 4 174 A5-5: Compression Fracture Picture 5 175 A5-6: Compression Fracture Picture 6 175 A5-7: Compression Fracture Picture 7 176 A5-8: Compression Fracture Picture 8 176 A5-9: Compression Fracture Picture 9 177 A5-10: Compression Fracture Picture 10 177 A5-11: Compression Fracture Picture 11 178 A5-12: Compression Fracture Picture 12 178 A5-13: Impact Fracture Picture 1 179 A5-14: Impact Fracture Picture 2 179 A5-15: Impact Fracture Picture 3 180 A5-16: Impact Fracture Picture 4 180 A5-17: Impact Fracture Picture 5 181 A5-18: Impact Fracture Picture 6 181 A5-19: Puncture Fracture Picture 1 182 A5-20: Puncture Fracture Picture 2 182 ix A5-21: Puncture Fracture Picture 3 183 A5-22: Puncture Fracture Picture 4 183 A5-23: Puncture Fracture Picture 5 184 A5-24: Puncture Fracture Picture 6 184 A5-25: Puncture Fracture Picture 7 185 A5-26: Puncture Fracture Picture 8 185 A5-27: Puncture Fracture Picture 9 186 A5-28: Puncture Fracture Picture 10 186 A5-29: Puncture Fracture Picture 11 187 A5-30: Puncture Fracture Picture 12 187 A5-31: Puncture Fracture Picture 13 188 x NOMENCLATURE Stress Characteristic Stress Strain Toughness Fracture toughness or Stress Intensity Mode I fracture toughness Dynamic fracture toughness Stress intensity during dynamic crack growth Stress intensity during dynamic crack arrest Stress intensity for sub critical crack initiation Crack size Surface Energy Surface Energy without porosity Young's Modulus/Stiffness Young's Modulus of indenter Young's Modulus without porosity Total Energy Mechanical Work Energy Energy for surface formation Kinetic Energy Temperature Time Poisson ratio Displacement Density Plane wave speed Longitudinal wave speed Shear wave speed Surface wave speed Force Area Mass Velocity Pressure during contact loading Area of indenter during contact loading Radius of indenter Porous material property Porous material property Strength Strength without porosity Crack growth velocity Probability of failure Weibuli modulus Gravity height xi ACKNOWLEDGEMENTS I would like to thank my supervisor, Dr. Royann Petrell, for her support and guidance throughout the entire project. I would also like to thank my thesis committee, Dr. Goran Fernlund and Dr. James Feng, for providing advice and for taking the time to read my thesis. I am also grateful to Dr. Nemy Banthia, Dr. Warren Poole and Dr. Aleksey Tsvetkov for their assistance in helping me organize the mechanical experiments and material fabrication. I thank Dr. Tamara Godbey for her assistance in completing experiments with the rabbits. I thank Mrs. Pamela Rosenbaum for helping in all purchase orders. And finally, I am indebted to the Marine Mammal Research Consortium for their financial support to make this project possible. xii CHAPTER I: Introduction Since the late 1970's, the population of Steller sea-lions (Eumetopias jubatus) have been declining in the Aleutian Islands and Gulf of Alaska [1]. Steller sea-lions are now listed as an endangered species in the United States and the trend of their drastic decline starts at 250 000 and ends at 50 000. Many scientists are investigating the cause, but difficulty arises in monitoring the young (the most vulnerable) at their remotely located rookeries. Telemetry tags could help to determine the cause of the decline by monitoring the movements and survivorship of young Steller sea-lions but existing products are problematic for young animals. Tags that glue onto the fur of the animal fall off when animals molt or hairs break. Also, many implantable types of telemetry devices are of limited range and longevity, too large or too surgerically invasive for use in a young animal. A small radio frequency (RF) tag placed under the skin would be a possible means of electronically monitoring an immature individual sea-lion. This project described herein is part of a larger project aimed at the development of a RF tag (with a 3-year working life) for placement under cranial skin of a sea-lion pup. The electronics housing is the focus of this project to determine its suitability for implantation into Steller sea-lions. The housing is assessed on its ability to not impede RF transmission, its biocompatible, its mechanical reliability, its longevity (which equals the lifetime of Steller sea-lions, 20-30 years) and its capacity to fixate the implant near the incision area. Fixation of the implant within the tissue is an issue because telemetry tags have been known to move within the body [2, 3]. Migration of implants may lead to re-location of the implant into positions that may be harmful to the animal and that may prevent proper transmission of an antenna signal. The implant must stay positioned relative to where it is implanted. Implant longevity is required as implants are inserted into young sea-lions and must endure the life of the animals. Unfortunately while human implants can be removed if defective or if any problems arise, animal implants must generally be left inside the animal for its lifetime as constant 1 examination and removal is not possible. Biocompatibility and mechanical reliability must be ensured for the entire life time of the Steller sea-lion. The objectives of this research project are the design, and biological and mechanical testing of a suitable housing for the radio-frequency tag under development. The hope is that the methodology and knowledge generated will lead to improved industrial implant design for the health and safety of wildlife. This thesis is organized to address the research objectives by beginning with the necessary background material to comprehend all aspects of the project. The second chapter describes the background needed to design the housing by addressing the size, shape and selected material for the RF tag and background relating to implant fixation and mechanical testing. Chapter 3 relates to the housing. Chapter 4 relates to the interactions of the housing with the biological surroundings. Chapter 5 is used to address the mechanical durability of the housing when subjected to applied forces that represent in-service conditions that the housing will experience. Overall conclusions on the suitability of the housing for this application are presented last in Chapter 6. The thesis diverges from traditional thesis style in that chapters 4 and 5 are written in paper style as they have separate introduction, materials and methods and results sections. This has been done for clarity between the two main issues of tissue fixation and structural mechanics. Reference style follows the style recommended by the Journal of Biomaterials. 2 CHAPTER II: Background 2.1 Introduction The design of the tag involved consideration of its biocompatibility, mechanical reliability (its resistance to in-service loading conditions of deep ocean diving, biting and impact), fixation in tissue and its compatibility with the radio frequency antenna. The following literature review was performed to select a material for the housing, to design its shape and size, and to plan out longevity experiments that address mechanical reliability and tissue fixation. Topics include: • Biomaterial properties to select a material for the housing that is biocompatible and has good strength • Tissue in growth to porous surfaces for material fixation • Mechanical properties of ceramics in fatigue, impact and Hertzian contact These topics provided the necessary background to design the housing, plan the longevity experiments and understand the results. The long term reliability or durability was evaluated by examining the material, biological and mechanical properties of the electronics housings. 2.2 Biomaterials The material properties are important for selection purposes. Strength, biocompatibility and longevity are important parameters that are considered in selection. Strength refers to a materials resistance to failure, biocompatibility refers to the tolerance a biological environment has to a material (see section 2.3 for an expanded explanation) and longevity indicates the duration a material can last in vivo before failure. There are several types of materials that can be chosen and these are briefly described to highlight the advantages that ceramics have in this application and thus, why they were selected. 3 T h e r e a r e m a n y m a t e r i a l s a v a i l a b l e f o r u s e in i m p l a n t d e v i c e s a n d e a c h o f t h e m is c o n s i d e r e d b i o c o m p a t i b l e h o w e v e r t h e i r i n t e r f a c i a l r e s p o n s e w i t h t i s s u e m a y v a r y [4]. F o u r d i f f e r e n t t y p e s o f m a t e r i a l s c a n b e d i s c e r n e d b a s e d u p o n t h i s r e s p o n s e : • T y p e 1: Inert m a t e r i a l w i t h a s m o o t h s u r f a c e • T y p e 2: Inert m a t e r i a l w i t h a m i c r o p o r o u s s u r f a c e • T y p e 3 : A m a t e r i a l w i t h a c o n t r o l l e d r e a c t i v e s u r f a c e • T y p e 4: M a t e r i a l s t h a t a r e r e s o r b a b l e T y p e 1 m a t e r i a l s p r o d u c e m i n i m a l t o x i c i t y to t h e s u r r o u n d i n g b i o l o g i c a l e n v i r o n m e n t h o w e v e r a t h i n f i b r o u s c a p s u l e f o r m s a r o u n d t h e i m p l a n t s e p a r a t i n g t h e i m p l a n t f r o m t h e h o s t t i s s u e . T h i s f i b r o u s c a p s u l e p r e v e n t s a d h e r e n c e o f t h e i m p l a n t w i t h t h e c o n t i g u o u s t i s s u e t h u s a l l o w i n g t h e i m p l a n t to m o v e u n d e r a n y s t r e s s o r e x t e r n a l f o r c e . T y p e 2 m a t e r i a l s c o n t a i n a m i c r o - p o r o u s s t r u c t u r e t h a t a l l o w s in g r o w t h o f t i s s u e . T h e f i b r o u s c a p s u l e sti l l f o r m s a n d i n t r o d u c i n g p o r e s in to a m a t e r i a l l o w e r s its e l a s t i c m o d u l u s [5]. P o r o u s m a t e r i a l s c a n b e in b u l k f o r m o r a s a c o a t i n g . T y p e 3 m a t e r i a l s a r e d e s i g n e d to c o n t r o l a n d e l ic i t a s p e c i f i c t y p e o f p r o t e i n o r m o l e c u l a r r e a c t i o n a t t h e i n t e r f a c e . S t r o n g c h e m i c a l b o n d s c a n r e s u l t f r o m t h i s s u r f a c e m o d i f i c a t i o n . A n d t y p e 4 m a t e r i a l s a r e d e s i g n e d to b e e v e n t u a l l y r e p l a c e d b y t h e s u r r o u n d i n g t i s s u e a s it r e g e n e r a t e s itself . T h e t y p e o f r e a c t i v i t y a n d a t t a c h m e n t c a n i n f l u e n c e t h e b i o l o g i c a l r e s p o n s e v i a t h i c k n e s s o f t h e i n t e r f a c i a l l a y e r b e t w e e n t h e i m p l a n t a n d s u r r o u n d i n g t i s s u e [6]. T h e a p p l i c a t i o n d i c t a t e s w h a t m a t e r i a l t y p e to u s e a s a l l h a v e u s e s in s p e c i f i c a p p l i c a t i o n s . T y p e s o f m a t e r i a l s ( i .e . p o l y m e r s , m e t a l s ) a r e n o t c o n f i n e d to a s p e c i f i c m a t e r i a l t y p e a l t h o u g h s o m e c a n n o t b e u s e d in a l l c a s e s . M i n i m u m r e q u i r e m e n t s ( a s s p e c i f i e d ) f o r m a t e r i a l s f o r s o f t t i s s u e i m p l a n t a t i o n a r e (1) e x h i b i t p h y s i c a l p r o p e r t i e s t h a t m e t t h e d e s i g n s p e c i f i c a t i o n s ; (2) m a i n t a i n p h y s i c a l p r o p e r t i e s in vivo t h a t a r e m e a s u r e d in vitro; (3) e l ic i t n o n e g a t i v e t i s s u e r e s p o n s e ; (4) d i s p l a y n o c a r c i n o g e n i c , t o x i c , a l l e r g e n i c , a n d / o r i m m u n o g e n i c e f f e c t ; (5) s t e r i l i z a b i l i t y w i t h o u t c o m p r o m i s e o f p h y s i c o c h e m i c a l p r o p e r t i e s [7]. U n f o r t u n a t e l y , t h e r e e x i s t s n o m a t e r i a l s t h a t m e e t 4 all these requirements as every material elicits at least a small reaction and material properties tend to degrade over time when immersed in biological environments. If a material is not mostly inert, then the foreign object may be rejected and ejected from the host tissue [8]. The main types of materials that are used for implants include polymers, ceramics, metals and composites. Their advantages and disadvantages are discussed next. 2.2.1 Polymers Polymers such as epoxy, silicone rubbers, polyurethanes and polyparaxylylene are widely used as biomaterials for implants in both wildlife and human applications and will be used as examples. Attributes that affect a polymer's properties include molecular weight, polymer chain orientation, the degree of crosslinkage, and chemical composition [3]. These attributes affect properties such as strength, resistance to stress cracking, viscosity, hydrophobicity and hydrophilicity, water absorption, temperature resistance, solubility, and electrical properties. Strength The mechanical strength properties of epoxy, silicone, polyurethane and polyparaxylylene are shown in Table 2-1. Polymers in general have very low strength relative to ceramics and metals. Their low stiffness and, thus, high ductility is the main reason for their use in many non-load bearing applications such as catheters or cardiovascular systems [9]. Epoxy has one of the highest strengths of all polymers and has uniform properties with temperature changes. However this strength is very low compared to other materials as is seen in subsequent chapters. Adding multifunctional crosslinking agents such as polyamines, acid anhydrides, polysulfides and polyamides can be mixed with epoxy to yield better strength, hardness, and high abrasion resistance [10]. 5 Silicones are also well known for their uniform mechanical properties from very cold to very hot temperatures, resistance to physical aging, hydrophobicity and chemical and physical inertness [11]. Small amounts of silica can be added to improve strength if needed. Polyurethanes are thermosetting polymers that are used to coat implants [7]. They can be formulated to yield products ranging from rigid to highly compliant and flexible [5]. They possess greater tensile, tear and flexural strength and fatigue resistance than silicones. Polyparaxylylene (also known as Parylene) provides little physical protection to the implant as it has poor adhesive properties (i.e. it can be removed quite easily by applying a shear force). Table 2-1: Mechanical Properties of Selected Polymers [5,12]. Properties Silicone Elastomer Polyurethane Epoxy Parylene C Tensile Strength (MPa) 5.2 31-62 30-90 68.9 Compressive Strength (MPa) - 65-90 70-130 -Young's Modulus (GPa) 0.001-0.01 1.3-2 13.8-26.7 2.8 Biocompatibility Some polymers (epoxy, silicone, polyurethane, parylene) have been shown to be biocompatible [5, 6, 8, 11, 13, 14-21] with applications in cardiovascular systems, hip implants, catheters and sutures. Epoxy is biocompatible and is versatile in that different fillers or additives can be mixed with it to create desired properties. For example fillers such as calcium carbonate, clay, silica, barium sulfate, mica, and various metallic oxides are used to tailor the physical behaviour of the epoxy such as glass transition temperature, viscosity, shrinkage, physical and electrical strength, abrasion resistance and thermal expansion [10, 13]. These can aid in increasing an epoxy's mechanical strength but fillers can be toxic to biological tissue. Medical grade epoxies exist to handle this problem as they contain minimal amounts of potential toxic additives that can leach out [11]. Silicone Rubbers, developed by Dow Corning Company, are also widely used as materials for biomedical implants in humans and animals [13, 22] and are the only non-carbon chain polymer used in implants. Medical grade silicone is the type used in implants and is biologically inert and 6 blood compatible. However, if the silicone is not fully cured, toxic products such as acetic acid and alcohol can leach out. Polyurethanes are also biocompatible and similar to silicone and epoxy; medical grades exist to improve biocompatibility [13]. Parylene C is used in medical applications as it possesses the best combination of electrical, physical and moisture permeation properties when compared with the others [11]. It is sterilizable and nontoxic. Inconsistent polymer performance can be a result of processing variations and can affect biocompatibility. For example, partially reacted or incompletely cured plastics/rubbers can evoke a biological response instead of being inert. In addition, the use of any additives or fillers must be shown to be biocompatible before inclusion in the polymer as they may leach out during use which can be toxic to the host. Longevity The longevity of polymers is directly related to their degradation in biological tissue and their long term inability to block moisture. Degradation can affect strength, resistance to stress cracking, viscosity, hydrophobicity and hydrophilicity, water absorption, temperature resistance, solubility, and electrical properties. The in vivo breakdowns of these properties are due to changes in cross-linking, oxidative or hydrolytic degradation, phase changes and absorption of bodily fluids [13, 14]. Pesch et al. [15] showed that the hardness of some polymers (polyethylene and polyacetal) decreased after a few months in vivo due to the absorption of bodily fluids. Moisture permeation is very important when polymers are used to encapsulate electronics. Any water permeation can cause corrosion of electrical components thus leading to a leaching of metallic ions into the host tissue. Moisture permeation leads to an inability to provide a hermetic enclosure. 7 When dealing with long-term implants, epoxies should only be used with rigid hermetic housings [23]. Numerous studies have shown that epoxies are poor ion barriers and absorb moisture easily [24]. This water absorption can result in penetration of the interior of the package if no hermetic housing is used and can change the mechanical and electrical properties of the epoxy. Studies on various different epoxies have shown weight changes from a minimum of 1% in two weeks to as much as 11% after eight weeks due to water absorption [10]. The water permeability is increased when curing agents or fillers are added. To attempt to alter epoxy's poor water absorption, hydrophobic diluents can be added. However as mentioned, some fillers or additives are toxic and can leach out of the coating into the neighboring tissue [14]. Similar to epoxy, silicone should be used in conjunction with a hermetically sealed container to protect the implant. Although more resistant to water than epoxy, because of fewer impurities and polar groups, silicones still swell which alters their mechanical properties [12]. Polyurethane provides good resistance to oil and chemicals however is susceptible to hydrolysis and is less resistant to water permeation than silicone. Polyester-type polyurethanes have been shown to degrade in the body and thus are not viable for long-term implantation [5, 13]. However poly-ether segmented polyurethanes have shown excellent stability over long implant periods along with high elastic modulus, biocompatibility, and resistance to bending [13]. Polyparaxylylene (also known as Parylene) is much more resistant to moisture than silicone and epoxy however it provides little physical protection. This makes it adequate for use as a secondary barrier to moisture. Thus, the suitability of polymers, in general, for long-term implantation is negated by problems with moisture permeation and degradation. This is further elaborated in the next few paragraphs. One of the most frustrating obstacles in implant design, are that many implants fail in the testing stages because the selected materials do not survive long term exposure to saline solutions or they do not adequately create a hermetic seal to prevent moisture from reaching the substrate 8 [11]. Polymers have been proven to allow moisture to permeate and thus true hermetic sealing can only be attained through metal, glass or ceramic containers [10, 11, 13]. A summary in the early 1990's [11] was made after many different results from the literature were reviewed by the authors. Their results showed that 1) the primary mode of electronic device failure occurred from the intrusion of water and ions to the substrate that lead to the corrosion of circuitry and shorting of conducting surfaces; 2) voids, cracks and poor adhesion greatly accelerated water permeation; 3) polymer encapsulants only retarded the water and ion permeation; 4) hard shells of glasses, ceramics and self-passivating metals provided the best corrosion resistance, chemical stability and water impermeability; 5) the most widely used polymers, medical grade's of epoxy and silicone, are susceptible to environmental stresses; 6) and that in general, no optimum encapsulation material exists. The last point is well illustrated when examining polymer encapsulants. Polytetrafluoroethylene is very moisture resistant and biocompatible, however its mechanical properties are poor especially its adhesion to substrates. Polyurethane is the opposite in that it has fairly good mechanical properties however is poorly resistant to chemicals in the body. The physiological environment exposes implants to various electrolyte fluids containing water, water-soluble proteins, lipids, sugars, and ions [1]. All polymeric encapsulants absorb some water and transmit water vapor or ions (primarily Na + and CI") to the substrate [10, 11, 13]. Water intrusion into the polymer will result in hydrolytic breakdown and delamination of the polymer and corrosion and shorting of any conducting circuitry on the substrate. Transmission of water vapor through the polymer occurs through interstitials (voids or cracks that form during curing). Reichert et al. [11], derived an analysis to prove that polymer encapsulated devices are non-hermetic and it was determined that only days are required for the internal humidity to reach that of the external. The same analysis showed that metals, glasses and ceramics can be considered hermetic for periods of months to years depending on the thickness. 9 Long term implant fixation is also difficult with polymers as surface re-arrangement is possible [16]. Surface atoms or molecules can rotate or translate in response to the biological environment shifting the surface structures. Thus, porous structures for fixation can be shifted and altered. The reasons for these changes in surface structure are minimization of surface energy and entropy increase. Stability can be obtained through high levels of crosslinking (inhibiting the motion of the chains) or bonding the surface layer to a stiff intermediate layer between the surface and substrate. In vivo, alterations of the surface can also occur by way of cellular interaction. Pesch et al. [15] showed that polymers such as polyethylene can have their surfaces worn away after only two months in vivo via the enzymatic actions of macrophages. A final disadvantage of polymers as long-term implant materials is their susceptibility to viscoelastic forces [5]. Stress relaxation is the increase of strain or deformation with constant applied stress. Polymers undergoing a constant stress over time will provide less protection for any substrate electronics as the material deforms in a time-dependent manner. Polymers have been shown to be biocompatible, cheap and compatible with the antenna (as indicated by their low dielectric constants, e.g. -3.5 for epoxy). However, the low strength and life time of polymers along with their inability to hermetically seal the housing indicates their unsuitability as a material for the intended housing. 2 . 2 . 2 Metals The three main metallic materials used for biomedical applications are stainless steels, cobalt-chromium alloys and titanium alloys. These materials have been used in medical applications (e.g. hip implants) and were potential candidates for the enclosure. 10 Strength Metals have a good mixture of mechanical properties. They exhibit reasonable strength while also maintaining a ductility that provides good resistance to fracture. Properties for titanium, stainless steel and cobalt chromium alloys are listed in Table 2-2. Table 2-2: Mechanical Properties of Selected Metals: Titanium alloys [5], 316L Stainless Steel [7] and CoCr alloys [7]. Material Tensile Strength Compressive Young's (MPa) Strength (MPa) Modulus (GPa) Titanium 620 620 100 Ti-6AI-4V 900 900 110 Annealed 316L 485 550 200 Cold-Worked 316L 860 965 200 CoCrMo 655 700 240 CoNiCrMo - Cold worked 1793 - 240 The elastic modulus for most metals is greater than 100 GPa which indicates a good resistance to deformation. Stainless steels are approximately 200 GPa, cobalt-chromium alloys range from 220-234 GPa and titanium is around 100-110 GPa. The strength of metals is reasonable in both compression and tension. Also, the fracture toughness of metals is quite high which is an indication of resistance to both crack formation and crack growth. Biocompatibility The good biocompatibility of metals has been well documented [5, 7, 15, 25]. Applications of which metals have been applied include joint replacements, screws, bone plates, heart valves and pacemakers. Problems with metals tend to arise from corrosion which is discussed in the next section. Longevity Metals are impervious to moisture and exhibit excellent electrical and thermal conductivity and mechanical properties. The main concerns with using metallic materials are corrosion. Corrosion can weaken the metal and corrosion products from the metal can leach into the nearby tissue. Results of metallic corrosion from implant retrieval have shown tissue discolorations of black 11 (titanium), blue-green (cobalt) and brown (iron based alloys) [5]. Austenitic stainless steels provide the best corrosion resistance of all stainless steels. 17-20% chromium content imparts the corrosion resistance and the addition of 2-4% molybdenum (which also reduces the grain size and increases strength) enhances resistance to pitting corrosion. Although the stainless steels contain alloying elements that improve the corrosion resistance, it is still recommended that stainless steels be used for short-term use only since they are still vulnerable to corrosion. Cobalt-chromium alloys demonstrate the most balanced properties of corrosion resistance; fatigue resistance and strength of all implant metallic alloys [5]. Two types of cobalt-chromium alloys exist, the castable CoCrMo alloy and the wrought CoNiCrMo alloy. The compositions of the two differ by alloying element weight percents of molybdenum, nickel, iron, tungsten and chromium. As in stainless steels, chromium is present in high percentages (20 up to 30%) for corrosion resistance. The wrought CoNiCrMo is highly resistant to corrosion especially to solutions containing chloride ions. These make the alloy appropriate for long-term applications. However wear, any corrosion, and fretting may cause metallic ions to be released into the surroundings with cobalt and nickel ions being highly toxic. Titanium has a good inherent corrosion resistance to accompany its good biocompatibility. It is commercially used in fabricating hermetic implantable containers (i.e. pacemaker cases with typical lifetimes from 5-10 years) and implants such as joint replacement parts thus it has a good record of long term use. Titanium has been manufactured with porosity for implant fixation and shown positive results [5, 26]. In addition to problems with corrosion, metals also are electrical conductive which will interfere with any signal produced by the antenna. Electromagnetic waves will be absorbed by the metals which would prevent any signal from being sent to the base station. The use of metals would require an exterior antenna, which would increase the complexity of the housing design and electronics systems. 12 Metals have high strength and good biocompatibility but their use as a material for the housing is limited by their susceptibility to corrosion (which limits their lifetime) and their conductivity. Another limiting factor could be the cost as titanium (~$34/kg) can be quite expensive however stainless steel is not (~$2.20/kg). 2.2.3 Ceramics Ceramics are generally known as very high strength materials that are brittle. They have good biocompatibility with applications in orthopedics. However, there are difficulties in manufacturing ceramics which can lead to high costs. Strength Mechanical properties that make ceramics useful are its good resistance to corrosion; its non-ductile behaviour yields good creep resistance, its high hardness and its high compressive strength. Ceramics have extremely high strength in compression but low strength in tension. Mechanical properties for different types of bioceramics are shown in Table 2-3. Table 2-3: Mechanical Properties of Selected Bioceramics [27] Material Tensile Compressive Young's Dielectric Strength (MPa) Strength (MPa) Modulus (GPa) Constant Aluminum Oxide (A l 2 0 3 ) 300 3000-4500 410 9.9 Silicon Carbide (SiC) 200 1200-2200 480 10 Zirconium Oxide (Zr0 2 ) 640 1800-2200 160-240 26 To achieve the high mechanical properties that ceramics are capable of, small grain size is required (<5 microns for alumina) [28] as well as high purity (>99.5% for alumina) [29]. Hip joint implants make use of ceramics for their high wear resistance and mechanical properties. The Wagner hip join cap has a wall thickness of 1-2.5mm and can withstand external loads greater than 30000 N and internal pressures of up to 100 bars (close to 100 atm) [30]. Thus, despite the ceramics brittleness, it is still used successfully. Calculations and tests show that the mechanical 13 lifetime of these hip joints and thus ceramics in general is approximately 30 years however this will vary with loading conditions, implant geometry and manufacturing [30]. Biocompatibility The biocompatibility of ceramics has been reported in the literature and they are regarded as the most inert materials in biological tissue [5-8, 15, 28, 29, 31-37]. Most applications of bioceramics are in the dentistry area although some implant applications use ceramics for blood interfaces, hip replacements and drug delivery systems [38, 39]. Alumina has been used in orthopedic surgery for over 20 years [40]. Histology results of subcutaneous and intramuscular implantation of alumina have shown no presence of foreign body cells such as macrophages or giant cells [31, 35, 36]. Thin fibrous capsules of 50-100pm thicknesses have been observed [31, 36]. When alumina has been placed intramuscularly, a connective tissue membrane of 24.4pm thickness formed around the ceramic to separate it from the muscle [24]. In vitro and in vivo assessment of alumina particles have been performed also showing no cell damage, adverse tissue reaction or immunological response [24]. The favorable reaction to particles reflects the good tolerance the body has for wear debris from a solid implant. Reactions to porous ceramics are discussed in section 2.3.2. The blood compatibility of alumina and other ceramics have also been proven to have good thromboresistance [32, 41]. Although, this material property is usually only important in the design of cardiovascular components. Good compatibility with osseous tissue is also reported along with the ability of certain ceramics to increase the rate of bone growth [9, 35, 42]. Implants removed after a year showed no ill effects in the implantation area with a good vascular system in the bone implant area. The surrounding soft tissue was also examined and no adverse biological reaction was viewed. In general, the response to an implanted ceramic is characteristic of the normal wound healing process [31]. 14 Longevity Similarly to glass, once a crack has been initiated the ceramic material will fail upon any load due to the high stress concentration that forms at the crack tip. This brittle nature is due to the ionic arrangement that renders dislocation motion difficult. High values for young's modulus are also a result of the impaired dislocation movement. This brittle nature can be problematic for ceramics in load bearing applications. Ceramics used as implant encapsulations provide high resistance to moisture (i.e. hermeticity) and have good corrosion resistance [7, 11]. In applications of joint replacements and dental implants, alumina (Al 20 3) ceramic has shown good long-term viability (up to 14 years and more). They are also non-conductive which translates to ease of permeation of electromagnetic radiation from the telemetry systems part of the implant (i.e. aluminum oxide has a dielectric constant -10) [10]. Ceramics are advantageous due to their high strength, biocompatibility and longevity however their inherent brittleness can pose problems under shock loadings. The brittleness of ceramics also leads to high manufacturing costs as much care must be taken to produce high quality components. 2.2.4 Composites In composite biomaterials, it is important that both constituents (fiber and matrix) are biocompatible [7]. Materials that are lightweight, strong and compliant can be formulated with composites. Polymers can be reinforced with fibers to increase stiffness, strength, fatigue resistance, and creep resistance. However it is important that the interface between the constituents not be degraded by the body otherwise many properties are altered. Problems dealing with long-term reliability are the same as with the other materials on account of composites being manufactured from polymers and metals or ceramics. The use of composites as biomaterials is still being explored with few existing applications. 15 Of all the available biomaterials, ceramics appear to be the most suitable as they show a good combination of high strength, good biocompatibility and longevity. Ceramics are also compatible with the radio frequency antenna (with medium level dielectric constants from 10-30 and low conductivity [5, 12]). Main problems will arise from the brittleness of ceramics and manufacturing costs. 2.3 Tissue Reaction In this section, the tissue reaction to porous surfaces is discussed. First, the general reaction to materials is presented to understand how foreign objects are treated when implanted in host tissue. Overviews of different attempts to alter the foreign body response are also included. Then it is discussed how porous surfaces affect the reaction that materials undergo with a focus on the results from research with porous ceramics. Also discussed is the usefulness of porous materials in reducing implant migration. 2.3.11nflammation and the Foreign Body Response Inflammation is a reaction of tissues to injury and is part of the natural wound healing process [8, 43, 44]. Bacteria and other foreign objects are destroyed or neutralized by the inflammation process, and debris is transported from the wound. Factors that can affect the wound healing process are [43]: 1. position or location of the wound in the body 2. interference of blood supply to the wound area 3. placement and immobilization of any cut surfaces 4. the general health of the host Surgical procedures incite the inflammatory response as there is severe vascular and cellular damage to expose the underlying tissue. Platelets aggregate, begin the coagulation process and release various proteases, growth factors and chemicals that recruit other cells and begin the wound healing process. Within the first few days, the implant is only seen as part of the initial 16 injury. A layer of proteins is absorbed onto the surface that can act as a signal for chemotactic stimulus for neutrophils (the method in which neutrophils move to the inflammation area) [45]. The absorbed proteins can also promote more platelet activation and their adhesion to the biomaterial is dependent on several biomaterial surface properties such as hydrophobicity. The inflammatory response can either be short lived (acute) or persistent over a long period of time (chronic) [45]. Acute inflammation involves rapid healing and is the initial response in all implantation cases where surgical procedures are necessary. The process involves the infiltration of neutrophils (polymorphonuclear leukocytes) into the injury area to remove any pathogenic, organism and cellular debris. When the site has been cleared and cleaned, neutrophils are extruded with the eschar (or scab) or they are phagocytosed by macrophages or fibroblasts [46]. Their presence can be prolonged through the persistent presence of foreign particles such as debris from a biodegradable material [47]. Their prolonged presence can result in the formation of an abscess or a localized buildup of liquid pus composed of necrotic neutrophils [32]. Fibroblasts appear afterwards to repair the injured area [43]. If the initial inflammatory response persists for a longer period of time, then it becomes chronic; however tissue damage is minimal or non-existent in contrast to acute inflammation [43]. Chronic inflammation is characterized mainly by the presence of macrophages and lymphocytes with possibly a few neutrophils present [11]. The macrophage initially attempts to phagocytize the material or if the material is too large, then macrophages can coalesce their cytoplasm's to become a multinucleate giant cell (also named polykaryocytes) to attempt to engulf the object. If the object is still too large then granulation tissue, consisting of fibroblasts and angioblasts in a matrix of collagen, forms around the masses of macrophages and encapsulates the material in a dense membrane of connective tissue. Collagen (a protein) is synthesized by the fibroblasts (i.e. fibroplasias [4]), which are stimulated to do so by the macrophages secreting lactate [44]. Granulation tissue formation is also facilitated by the macrophages releasing growth factors interleukin-1 (IL-1), interleukin-6 (IL-6), TNF-a, fibroblast growth factor (FGF), platelet-derived 17 growth factor (PDGF) and transforming growth factor 6 (TGF-fs) [25, 46, 47]. These growth factors have effects on the fibroblast proliferation, connective tissue matrix production and the recruitment of inflammatory cells and monocytes. The fibrous capsule or granuloma separates the foreign object from the surrounding tissue [43]. Thus the granuloma consists of the foreign material surrounded by macrophages (possible in the form of a polykaryocyte) all encapsulated in a connective tissue layer. It is connected to the stroma of the surrounding tissue through a network of collagen. The inner surface or interface between the foreign object and the fibrous capsule is an "unfinished" surface [8]. If the object is removed then the capsule will collapse and resume the tissue regeneration until the space is completely filled. As time progresses, the granuloma can undergo caseation necrosis (degeneration of the tissue into a soft cheese-like substance), calcification (tissue hardening by deposition of calcium), liquefaction (tissue becoming a liquid) or hyperplasia (the increase of cells in the tissue). Macrophages have been recognized as a main component in the cellular response during inflammation and experiments have been performed on the activity of macrophages following the implantation of small samples [44]. During the first week after implantation, macrophages were observed engulfing debris which then became incorporated into the lysosomes (i.e. phagolysosomes). Tissue debris was easily degraded by the lysosomal enzymes while the non-degradable materials were left in the cytoplasm. The foreign body response is a specific type of chronic inflammation as other sources or items can induce inflammation. The granuloma can affect the performance of implants as it has been found to impair the function of biosensors [48, 49]. The fibrous capsule blocks the analyte from reaching the sensor. In this project, the fibrous capsule could affect the implant but in a different 18 manner. A thick fibrous capsule may absorb the transmission signal, preventing the signal from reaching the receiving base station. The response to a biomaterial can be either local or systemic [45]. Inflammation, necrosis and neoplastic transformation are local responses while end organ (part of the sensory nerve) destruction, distant carcinogenicity and immunological reactions are systemic. Several implant variables can affect the foreign body response in addition to the above-mentioned factors that can influence wound healing. These can be chemical, mechanical, geometrical and surface factors. Chemical and material factors Any chemical or material debris that finds its way into the surrounding tissue from the implant can result in chemical toxicity. This can be caused by metallic ions leaking into the environment from corrosion or from polymers leaching additives. Any impurities or filler materials that leak into the tissue can also lead to an adverse response (i.e. carcinogenic) [4, 14]. Autian [14] cites many different medical cases where this has been found. Non-toxic effects can also result from the implant. These include surface interactions such as protein adsorption. Denaturation of proteins can possibly evoke an immunologic response contributing to the inflammatory response [43]. Studies have shown that different types of materials or material chemistries elicit a different tissue response. Parker et al. [50] performed a test comparing silicone elastomer to poly-L-lactic acid. The initial response to poly-L-lactic acid showed a thicker capsule than that of silicone however after a steady-state cell activity was reached, the capsule was thinner. Wood et al. [51] performed tests on silicone, vinyl, polyurethane and Teflon (PTFE) and measured the fibrous capsule thickness of each after 4 months. Inflammation was more pronounced with the 19 polyurethane and the vinyl capsules. Teflon showed the thinnest capsule at 87 pm while vinyl had the thickest at 252 pm. Matlaga et al. [52] performed tests on six medical grade polymers; polypropylene, polyethylene, polyurethane, silicone rubber, polyvinyl chloride), and Teflon. Results varied but, in general, it was observed that polypropylene exhibited the least reaction while polyurethane demonstrated the highest response. Werkmeister et al. [53] also corroborate this conclusion of polyurethane inducing a high inflammatory response leading to a thicker capsule. Experiments have also been done to examine the differences between biodegradable and non-biodegradable materials in vivo [54]. Due to the degradation of the materials tested, inflammation increased with time as fragments or debris from the material entered the surrounding tissue. Foreign body cells would entrap these fragments as the inflammatory response persisted. The non-biodegradable material (e-PTFE) showed limited foreign body response with an adjacent capsule. These findings are also mentioned by Woodward et al. [55]. The effect of different materials on tissue response is explained as the result of different surface hydrophobic and hydrophilic characteristics [46]. Surfaces that are more hydrophilic promote cell adhesion and hence a better mechanical interlocking [50] leading to a reduction in the inflammatory response. The surface chemical composition can also be changed to affect the inflammatory response. Ward et al. [17] mixed in polyethylene oxide groups into the synthesis of polyurethane to investigate its effects. It was suspected that the hydrophilic nature of the polyethylene oxide group would improve the cellular adhesion to the polyurethane. Experiments confirmed these suspicions and showed a significantly thicker fibrous capsule for polyurethane without these side groups (104.9 pm vs. 193.2 pm). Negative effects can also occur from chemical composition. Toxic impurities in the material can leach out and cause an inflammation effect [55]. Not only does the surface characteristic change (i.e. brittleness), but foreign body cells will respond with phagocytosis inciting further capsule reaction. 20 Mechanical factors Implant movement with respect to the adjacent tissue will lead to irritation and an increase in inflammatory response [43]. Movement can be induced by interior forces such as muscular contraction, joint movement or inner fluid movement, or exterior forces such as impact forces from a source hitting the implant area. This 'mechanical irritation' has been studied by one group [56]. They implanted both smooth and micro-grooved membranes in two different locations where one location would produce implant motions (subcutaneous layer in the back of an animal) and in one location where the implant would be securely placed (sub-periosteally on the frontal bone of the skull). Their results showed a similar capsule composition of fibroblasts and collagen (although in early weeks it was different, at the final test period the content was the same) for both locations and for both surfaced materials. The capsule thickness increased for all subcutaneous implants and it decreased for all sub-periosteal implants. The interface was the same with all implants containing a layer of macrophages and foreign body cells. These results confirm the notion of 'mechanical irritation' or frictional force effect on the foreign body response. Geometrical factors It has been found that the size and shape of the implant can have an effect on the inflammatory response [43, 53]. Implants of the same surface composition and material but with different shape, exert different mechanical stresses on the surrounding tissue affecting capsule formation and contraction. Tubular or rounded implant designs are considered optimal [18, 55]. Ward et al. [17] looked at varying thicknesses of 1.2 cm square samples. Thicknesses were 300 pm and 2000 pm. Results showed that the thinner samples had a fibrous capsule thickness of 84.3 pm. This was approximately 20 pm less than the thicker samples. This result is corroborated by Woodward et al. [55], who state that smaller implants generate more uniform and lesser responses than larger implants. The shape of implants and their effect on tissue response has been documented [52]. Matlaga et al. [52] examined the foreign body reaction to circular-, triangular-, and pentagonal-shaped cross 21 sectioned rods. After an implantation period of 14 days, the samples were excised and it was found that triangular rods showed the highest degree of macrophage activity with pentagonal rods second and the circular rods showing the least. The proposed explanation for this was that the acute angles of the triangular cross section caused movement of the implant within the host initiating a greater tissue response. The size of animal being tested on also seems to affect the response. Data obtained from large animals is more comparable to human wound healing processes then from small animals like rats or mice [57] whose response is somewhat different. This was noted by Janssens et al. [58], who tested implants on animals of different weights and noticed more inflammation in the lower weight animals. Surface factors Smooth surfaced implants will always be surrounded by a fibrous capsule with the thickness of the sheath varying with implant shape, movement and reactivity [11]. Surface micro-texturing through grooves and ridges has been speculated to influence the number of inflammatory cells, capsule thickness, capsule organization, and the number of blood vessels present around the implant [18, 20]. The hypothesis is that a mechanical interlocking between the implant interface and the surrounding tissue will reduce the capsule size. This would be due to a reduction in the stress and movement of the implant to the surrounding tissue. Otherwise the results are a 'mechanical irritation', which induce tissue damage, fibrosis, and increased inflammatory response (hence leading to a thicker capsule). Taylor et al. [19] speculate that the increased numbers of macrophages observed at textured surfaces are responsible for a thinner fibrous capsule. This is explained by the possibilities that 1) macrophages reduce the proliferation of fibroblasts and hence collagen production; 2) the increased number of macrophages may produce a substance that causes fibroblasts to reduce their collagen output; 3) the textured surface stimulates collagenase and elastase secretion in the interfacial macrophages which leads 22 to degradation of the capsule; or 4) the collagen produced is of an altered structure that is more soluble and susceptible to collagenase than normal collagen. Many different groups have attempted to examine the effects of textured surfaces on the formation of the fibrous capsule. Varying results have been seen. Some [19-21, 59-61] have performed studies demonstrating differences in tissue reactions and attachment while others [17, 18, 50, 56, 57] have shown no change. In general, studies have been performed with several different implant membranes consisting of micro-grooved surfaces. For example, den Braber et al. [20], used surface micro-grooves of widths 2.0, 5.0, and 10 pm with a constant depth of 0.5 pm. Walboomers et al. [18] used implants with micro-grooves of 2 microns with varying depths between 0.5-6 microns. Thus studies have been performed by examining not only the effect of micro-textures, but also the variability in micro-texture geometry with respect to ridge width and depth. The layer of inflammatory cells has been seen in several different studies [18, 19, 56, 57] and is explained as a phenomenon called rugophilia [44, 56]. This phenomenon describes the behaviour of macrophages in vitro and in vivo and how they are attracted to roughened surfaces. This layer of macrophages is used as a possible source of differentiation between in vivo and in vitro results in terms of effects of surface micro-machining. It is suggested [18, 56, 57] that this layer of inflammatory cells 'hides' the surface texture and any cells in contact with the layer of macrophages does not detect the actual structure. Another suggestion given is described by Walboomers et al. [57] in that the difference in Young's modulus of elasticity between the implant and surrounding tissue is the cause. This mechanical compatibility has been mentioned in other literature as well [4, 62]. The mismatch of stiffness can create stress and strains in the interface that can cause the surrounding tissue to detach and become separated. The implant is then free to move against the host tissue creating 'mechanical irritation' which leads to an increase in inflammatory response. 23 Trends over different implantation periods have shown similar results for different micro-textures [20]. The different bars represent the same material but with different groove sizes (with one control that was without grooves). The capsule comparable at each time period indicating the grooves had no effect. This does not indicate that all surface textures are ineffective as some have had success with different patterns such as pillars [19, 61]. Taylor et al. [19] performed experiments on micro-pillars that were patterned randomly. Their results showed a 30% reduction in fibrous capsule thickness with textured surfaces compared to smooth surfaces after 8 weeks. Different surface treatments have been applied to materials in hope to manipulate the tissue response. Parker et al. [50] have looked at surface pre-treatments of fibronectin pre-coatings and radio-frequency glow discharge (plasma treatment). The plasma treatment is an attempt to increase the surface free energy to improve tissue attachment. The fibronectin coating is an adhesion protein (integrin) that has a hydrophilic chain that has been shown to promote cell adhesion in vitro [63]. Results showed that an increased inflammatory response occurred around the fibronectin coated materials while the radio-frequency glow discharge treated materials showed no difference when compared to controls. Sterilization effects on material surfaces have also been examined to determine in vivo responses [64]. It is known that sterilization can have significant effects on polymers. Heat or steam sterilization can cause hydrolysis, melting or degradation of the polymer while radiation sterilization can degrade the polymers through chain scission and de-polymerization [13, 65]. Zhang et al [64] examined the effects of ethylene oxide, beta, and steam sterilization on silicone and polyurethane with respect to tissue response. The results showed a similarity between both materials and sterilization techniques after a long implantation time. Initial findings demonstrated different reactions however these converged to similar levels as time increased. It was concluded that the sterilization technique used showed no difference in in vivo responses. 24 2.3.2 Porous materials and implant migration The surface characteristics of implants in soft tissue reflect its ability to interact with its contiguous environment. Improperly secured or loose implants can result in migration or even expellation through the skin [13]. Ceramics (i.e Bioglass, Ceravital, alumina or hydroxyapatites) can be made to be surface reactive or porous to form strong bonds with adjacent tissue [7]. Various sources have reported varying pore sizes for optimal soft tissue integration. 5-15 and 20-40 microns has been reported for soft tissue [33, 35, 66] and 150 and 200 microns for bone tissue [6, 31, 33, 35]. The surface pores and inner interconnected pores must be of these minimum dimensions. A few experiments have been performed to examine the effect of pore size on the inflammatory response [17, 35, 67-69]. Klinge et al. [67] performed experiments on two different pore sizes for a polypropylene material. Mean pore size was 460 pm and 2800 pm. In the small pore sized material, the inflammatory response (in the form of granulocytes and macrophages) increased over the 90day test period. This contrasted to the larger pored material where inflammation decreased over time. Ward et al. [17] performed a similar test with pore sizes of 1 pm and 60 pm. Although, the materials used were different for the different pore sizes thus comparison between the two pore sizes could not be done without taking into consideration the material effect. The final capsule size was similar despite this difference and was also close in thickness to a non-porous polyurethane also tested. These results would indicate that pore size has no effect on capsule thickness. Their results did indicate that capsule density significantly differed with the non-porous material showing a high density of close packed collagen and fibroblasts. The porous capsule density was much less dense similar to the wound repair in normal subcutaneous tissue without the presence of a foreign object thus indicating a less severe inflammatory response. Sharkawy et al. [68, 69] confirm this conclusion with porous PTFE. Ceramics have been fabricated to have surface porosity and their ability to promote tissue in growth and fixation has been investigated [31, 35, 37, 70, 71]. Tissue reactions have been 25 observed to improve with the introduction of porosity. In one study, the thickness of the fibrous capsule for dense ceramics varied in between 15-30 cells thick while porous implants showed encapsulations of <6 cells thick [31]. Tissue, both fibrous and muscle grew into the pores along with blood vessels. For implants with larger pore sizes, almost no fibrous encapsulation was observed and similarly to dense ceramic implants, no inflammatory cells were visible. Post-surgical inflammation was also shorter for porous implants. The larger encapsulation of dense ceramics was attributed to the lack of attachment to surrounding tissue that enabled the implants to move freely and abrade the tissue [31, 71]. Porous implants were firmly held in place and could not move. A study on the use of porous alumina for tracheal applications showed the implant firmly in place with no movement [71]. Within the porous structure, connective fibrous tissue and blood vessels are observed indicating good biocompatibility [35, 71]. The movements of implants within a host have been observed in many different applications such as in monitoring of live stock [58, 72], fixation screws in hip fractures [73, 74], dental implants [75], cochlear implants [76] and acetabular cups [77, 78]. For some applications, the movement of the implant may affect its performance such as in the case of our RF tag application or in biosensors [79]. It has been suggested that material parameters such as size, shape, surface texture, and biocompatibility can affect foreign body migration [79]. Repetitive body motions (i.e. from joint or muscles), capillary action and gravity are also plausible causes that have been suggested [80]. One research group attempted to measure migration of glass rods using a staining technique [79]. Results from implantation times of less than 7 weeks in adult cats showed small movements of less than 5mm. Migration of lag screws in hip fracture applications lead to 8-17% failure rates and migration distance has been studied [73, 74]. Screw migration is thought to occur from the cyclic dynamic loading of the screws which leads to rotation of the screws. Rotations of up to 30 degrees have been observed with migration distances not being reported. Implant recovery from research with live stock have showed 10-20% losses of implants as implants migrated from the 26 implantation location or were expelled from the animal [58, 72]. The measured migration distance of implants showed only small movements of less than 1cm [72]. The failure rate in cochlear implants is much lower with reported failures (migration of electrodes out of the cochlea) to be between 1 and 1.3% [76]. Dental implants have been observed to show significant movement [75, 81]. Radiographs have shown movements of dental implants into the maxillary sinus which has been reported on several occasions due to weak bone interfaces or infection. A similar case of implant migration in facial tissue was observed with an orbital floor implant across the nasal septum [82]. The implant was not fixated and caused breathing problems in the patient. Hip implants are also subject to implant migration. The acetabular cup can migrate and rotate [77, 78, 83, 84]. Measurements of the migration and rotations have yielded results of 130pm movements and rotations of less than 1 degree [77], maximums of 1-1.6 mm movements with an average mean of 0.04mm [78] and 1mm movements and 1 degree of rotations [83, 84]. One very severe case of implant migration was reported for the movement of an acetabulum from a hip implant into the bladder [85]. It is reported that the implant was able to move so far because of radiotherapy in that area and thus the irradiated tissue was easier for the implant to migrate through. Migration of fixation devices has been seen to be a common occurrence [80]. Pin migration from intrathoracic, intraperitoneal, intrapelvic, intracranial and intraspinal implants has been reported with some cases requiring medical attention. Some fatal cases were reported for pins used in shoulder operations [80, 86]. Other types of implants where migration has been observed are central venous catheters, ventriculoperitoneal shunts and ocular prostheses [86]. The bond strength of an implant and surrounding tissue can be enhanced via porous surfaces. Micro-motions of approximately 25 microns has been calculated for bone and metallic implants 27 [66]. These micro-motions can cause irritation to the adjacent tissue further inciting the inflammatory response and can either deteriorate the material or the surrounding tissue [6, 31]. Mechanical interlocking of porous materials with adjacent tissue can reduce the movements. 2.4 Mechanical Properties of Ceramics Fatigue, impact and Hertzian contact of ceramics are presented in this chapter as each are loading conditions that the electronics enclosure will encounter while implanted. Fatigue occurs from repeated deep ocean diving, impact can occur from collisions with other sea-lions or rocks and forces similar to Hertzian contact loads can occur from biting. Sub critical crack growth is discussed in the fatigue chapter and the effect of dynamic loads on brittle solids is given in the impact chapter. An introduction to fracture mechanics is given first. Knowing a material's strength is important when designing a component or part that is under stress. The strength of a material determines its resistance against deformation from applied stresses. A material's stress-strain curve (Figure 2-2) indicates important strength parameters of the material such as the yield stress (YS) and the ultimate tensile stress (UTS). The yield strength indicates the point at which the material response transitions between elastic to plastic (where permanent deformation begins) and the UTS is the maximum stress the material can withstand. Where the stress-strain curve ends, is where fracture occurs. Brittle materials deform very little before fracture as shown in Figure 2-1. The yield stress, along with a safety factor, is usually the design point for structural design [87]. 28 s Ultimate Tensile Strength Fracture Fracture e Figure 2-1: Material Stress (s) - Strain (e) Curve. A - Brittle Material, B - Ductile Material The area under the stress-strain curve yields the amount of energy absorbed before fracture or failure [87]. This is also called the toughness and is defined as: Failure can occur by means of a brittle or ductile nature. Brittle failure is a low energy failure mechanism characterized by rapid crack propagation with little plastic deformation. Only small amounts of energy are required to initiate crack growth with failure occurring close to the yield stress. Cleavage is a characteristic type of low energy brittle failure which occurs along specific crystallographic planes or grain boundaries (i.e. intergranular fracture). Coarse grain structures are easily susceptible to this type of failure as the crack failure path is not impeded. More grain boundaries or a fine grain structures increase the energy required for crack growth to continue as the crack must re-orientate itself as it encounters a new grain to continue along the growth plane. High energy or ductile failure occurs with extensive plastic deformation via shear [87]. The fracture toughness is a method of determining when a material will undergo fracture [87]. The applied stress and crack size before fracture are related to this material property. As a stress Te =}o-ds ( 2 . 1 ) 2 9 is being exerted on a material, cracks within the material grow and once the crack reaches a critical size, it grows rapidly to cause failure in the material. The relation is expressed below: The fracture toughness is also related to the amount of energy required to form a crack. This energy is expressed as twice the surface energy with the relation below: In general, the fracture toughness for ceramics and polymers are low with composites exhibiting moderate levels and metals exhibiting the highest resistance to fracture. The differences in fracture toughness values are mostly due to the ductility of the materials and the materials ability to absorb externally applied energy through plastic deformation. The range of values for K ! c is <1 MPa m 1 / 2 (brittle) to >100 MPa m 1 / 2 (very tough) [87]. The mechanical properties of ceramics show various advantages and disadvantages with respect to fracture. It is well known that ceramics are brittle but they also have very high compressive strengths. Inherent flaws coupled with a ceramics inability to undergo plastic deformation, lead to a low fracture toughness and large susceptibility to mechanical shock loadings [88]. Inherent flaws also lead to a large statistical spread in measured mechanical properties as each component has different amounts and sizes of flaws. Flaws are introduced to parts or components through the manufacturing process (i.e. machining or sintering) and by microstructural arrangement or properties. Examination of material properties is usually performed by constructing Weibull diagrams. Weibull diagrams are plots that show the probability of failure of a component for a given failure mode. Strength of ceramics is controlled by temperature and microstructure (and can then vary from flaws). Porosity, grain size and high temperatures can significantly affect the elastic constants. ( 2 . 2 ) Kc=j2yE ( 2 . 3 ) 30 Pores in the material can act like flaws or cracks and lead to early failure or failure at lower stresses; a smaller grain size can improve the strength of the material but makes the material more stiff and thus the fracture energy is lower; and high temperature can create new flaws by causing limited plasticity [88]. Cracks or flaws are the dominant factor in the failure of any material and the focus of fracture mechanics. How different loading conditions affect crack growth is thus important to understand how failure of a material occurs. 2.4.1 Fatigue Fatigue is a type of failure caused by repetitive or cyclic loadings [87, 89]. It is usually brittle in nature with very little plastic deformation associated with it. The cyclic stress is less than the yield stress and sources of fatigue can be thermal, rotational or vibrational. Thus, failure from fatigue occurs below the expected failure point. Four different approaches can be utilized for the design process of parts under fatigue. The first is the 'infinite life' method where the stresses are low enough that fatigue cracks never initiate and thus fatigue failure never occurs; the second is the safe life approach whereby a finite cycle limit is specified and then the part is removed from service when the limit is reached; the third is the fail safe scheme which accepts that failure is possible but failure of the given component will not be catastrophic to the entire system and thus failure is acceptable; and the fourth method is the damage tolerance approach where the notion that cracks or defects are present at the start of component service, that those cracks or defects will continue to grow over time, and that accurate prediction of failure can be achieved by using fracture mechanics [87]. The fourth method is used most often as cracks are always expected to be present after component fabrication. Prediction of a components life using fracture mechanics then can allow criteria from the other methods to be incorporated. Through experiment, fatigue stress-cycle (S-N) or strain-cycle (e-N) curves can be plotted and then used to extrapolate the lifetime for a part given an applied stress or strain. Problems 31 associated with developing and reading S-N curves are that many samples are required to produce the curves and the plots do not indicate when crack initiation has begun or the rate of crack growth. These curves are useful when predicting the lifetime of components or parts. A typical curve is shown in Figure 2-2. Failure by fatigue occurs first through crack initiation (usually instigated at the surface) followed by crack propagation until fracture [87, 89]. Separation of slip bands initiate the crack and are a result of sliding or shearing of atomic planes within the crystal structure. Further propagation of the crack is a result of a stress concentration forming at the tip and its rate of movement through other grains can be accelerated with the presence of defects [89]. Stress Number of Cycles Figure 2-2: Typical S-N curve Failure from fatigue occurs because the strength of materials, including ceramics, is time-dependent. This is different from the time-dependent failure from viscoelastic effects as strength degrades at lower temperature ranges where grain boundary slip is not possible. Failure from fatigue, which can occur at much lower stresses than the yield strength, is attributed to the slow and continuous growth of subcritical flaws until critical size is reached [88]. The environment and temperature can aid in the growth of cracks by accelerating their growth at an applied stress. Internal residual stresses or stress concentrations from geometric sources are also involved in subcritical crack growth [90]. There are two types of fatigue, static and cyclic, with each producing different failure limits [88, 90, 91]. Results have shown that failure from cyclic fatigue occurs at lower stress levels than 32 failure from static fatigue (static fatigue is the application of a single stress until failure opposed to increasing and decreasing the applied load between a maximum and minimum). The main reason for the difference in fatigue lives is the crack bridging mechanism [27, 92, 93]. This mechanism helps to stabilize crack growth and to reduce the stress intensity at the crack tip to below the levels of applied stress in the rest of the material. Crack bridging occurs in the wake of a crack and involves the bridging or re-connection of grains behind the tip of the crack. The bridging ligaments create closure forces that reduce the crack stress field in a material subjected to a static load. The bridging occurs via frictional forces of asperities. Cracks that are subjected to cyclic loads initially undergo the crack bridging mechanism but the strength of the frictional forces joining the grains in the crack wake are degraded by the repeated opening and closing forces that follow the cyclic load path. 2.4.2 Impact In a fracture mechanics approach, cracks are expected to be formed and are allowed as long as they do not exceed a critical size. How cracks grow depend on how the material is loaded. For a dynamic load, the crack is subjected to a mechanical force rapidly. For low loading rates, this can be analyzed using static equations with varying material characteristics. But at higher speeds, inertia effects or stress wave effects occur and must be considered [94]. This is the biggest difference between static and dynamic loading; the rate of loading is much faster in impact so much so than static that the component under impact is unable to deflect in a quasi-static manner and mechanisms that dissipate strain energy in static crack growth do not occur [95]. The Griffith energy balance concept describes the growth of a crack via an energy balance in the system [96]. The energy in the system is related to the strain potential energy of the elastic medium, the potential energy from the externally applied load and the surface energy required in forming new crack surfaces. U = UU+US (2.4) 33 This idea has the equilibrium concept that the change in energy with crack growth is zero as crack extension will lower the mechanical potential energy while increasing the surface energy. Or, crack extension is furthered by a release in mechanical/internal strain energy and the release in energy is used to create new crack surfaces. Thus, crack growth occurs when U M > U s [32]. d U = 0 ( 2 . 5 ) dc This concept forms the basis for formulation of solutions for static crack growth. However, when an unbalanced dynamic force is applied, the system acquires kinetic energy and is considered a dynamic system. Dynamic crack growth is achieved when the crack reaches the critical failure length (as described by the stress intensity factor) and its velocity accelerates until material failure. This can arise from a static loading. Rapid or time-varying loading such as impact can also cause dynamic crack growth but the added input of kinetic energy accelerates the time for the crack to reach critical levels. The Griffith concept can be modified to incorporate inertial effects and such has been done by Mott so that [96]: U = UM+US+UK ( 2 . 6 ) The added input of energy leads to a lower strain energy needed for fracture and thus fracture is attained in materials for lower levels of applied stress. Crack growth from dynamically applied loads occurs from interactions with stress waves if the input of kinetic energy does not lead to material failure first. Cracks that interact with stress waves are subject to loading related to the amplitude of the stress wave. Each pass of the stress wave will incrementally increase the crack length until the crack reaches its critical length or the stress waves dissipate [96]. Under dynamic loads, the resistance or toughness is different than 34 the static equivalents. Three different stages (initiation, propagation and arrest) of the growth of a crack under dynamic loading are used to describe the evolution of dynamic crack propagation. A crack growing after a dynamic load is applied is growing with a stress intensity that is different than the stress intensity of a static load. The dynamic stress intensity factor for crack initiation is related to the dynamic stress intensity through the relation [97]: The left hand side is the dynamic stress intensity factor at a time tf when the crack propagation begins. The right hand side is the crack initiation stress intensity for mode I loading and shows a dependence on the temperature and the rate of loading. The temperature dependence occurs from the increase in material ductility with increasing temperature or from internal energy dissipation associated with inelastic/plastic deformation near the crack tip zone. The rate dependence occurs due to the rate dependence of the inelastic material response in the crack tip zone and the inertial nature of the stress field in the crack tip zone [97]. Once the crack has been initiated in dynamic loading, the stress intensity factor becomes related to the velocity of the crack or the kinetic energy in the crack tip zone [97]. Thus: As the crack velocity increases towards a terminal velocity, the crack may bifurcate or branch to dissipate some of the kinetic energy [96]. This is not seen in quasi-static cases. Crack branching can occur because of interactions with secondary micro-fractures or defects in the crack path that link up with the advancing crack. Defects ahead of the crack interact with the advancing stress field and can grow until met by the propagating crack (micro crack coalescence). Crack Kfn{tf) = Kld{T,dKdrldt) (2.7) K?"(t,v) = Kld(v,T,dK?"/dt) (2.8) 35 branching can also occur by interactions with stress waves. As the crack grows, the associated stress field also moves and the stress waves can reflect off of free surfaces or micro-structural in homogeneities. For strong interactions of reflected stress waves and stress waves from the propagating crack, the crack may be deflected onto adjacent cleavage planes or possible even arrested. Deflection can lead to crack branching and is a function of the fast moving crack [96]. While the crack is growing, it is possible for it to cease growth or for the dynamic crack to arrest. This is described by another stress intensity value called the dynamic crack growth arrest toughness and crack growth cannot be maintained when [97]: K?"(t)<KIa(T) ( 2 . 9 ) Upon impact, stress waves are generated from the area of impact and spread throughout the material. The wave-propagation equation can be used to describe the movement of a plane wave through a material (a plane wave is one where the displacement is constant over all points of a plane perpendicular to its direction of travel): 32« 1 d2u , E —T = ~7—T" where c = J — ( 2 . 1 0 ) dx2 c2 dt2 11 The resulting stress and strain are: F l d(mv) a = ^  = ^ ^T- =pvc° (2-n) A A dt s = — ( 2 . 1 2 ) c„ 36 The initial wave moves towards the material extremity and will reflect as a stress wave with the same sign and amplitude but as either compressive (if initial was tensile) or tensile (if initial was compressive). The wave then reaches the impact end and reflects once again until the case of a static loading is reached. If an interface is reached in which the material properties or cross section are changed, the wave is split into components with part of the wave continuing on while another part of the wave is reflected [98]. Experiments have shown that values of KM for alumina can range between 3.5-5.7 MPa m 1 / 2 . It has also been shown that these values can vary with temperature (decreasing with increasing temperature) [99]. As the crack is propagating, it is possible that the crack may curve or branch into multiple crack paths. The dynamic crack curving criterion uses the assumption that the stress field ahead of the crack will determine whether branching or curving occurs. A critical value for the circumferential stress at a distance r0 from the crack will dictate the curving at a certain angle ?„. The circumferential stresses cause the off-axis microcracks to enlargen and connect to the main crack and thus cause the crack to either curve or diverge [99]. A few researchers have looked at the impact of ceramics, primarily in applications to ballistics as armour materials [100-103]. These experiments use ceramics along with a substrate material to provide protection against high velocity projectiles. The ceramic material is used for its high strength and the substrate is used to absorb the kinetic energy of the projectile. A few have looked at the impact strength of ceramic hip implants as the acetabular head and cup are predicted to be loaded dynamically for 1 to 2 million cycles per year [88, 104]. Drop-weight and pendulum impact tests were conducted to measure the impact energy and the resistance of alumina to specified applied loads. Recorded impact energy at fracture ranged from 14.4 J [104] to 22 J [88]. An impact fatigue limit was observed for <9 J [88]. 37 2.4.3 Hertzian Contact Hertzian contact involves analyzing the stress field from the contact of a blunt indenter with a solid surface [27, 105-109]. This is a special type of fracture situation that exhibits uncommonly high tensile gradients near the edge and corners of an indenter. For ceramics, this can be problematic as ceramics are generally much weaker in tension than compression. Below the indentation load, compressive forces occur. During unloading, the stress fields switch as residual stresses in the material form when the force is removed. Figure 2-3 shows this schematically. The contact pressure can be expressed as either a function of the applied load or the strain. Applied Load Load Removed Compressive Stresses Tensile Stresses Figure 2-3: Stress state during contact loading. During the loading phase, material is compressed beneath.the indenter and material is 'pulled' towards the indenter creating tensile stresses at the surface. During the unloading phase, residual stresses build up as the material attempts to return to its initial state. This creates a tensile stress below the area of indentation and compressive stress on the surface [105,106]. p0 = Flm2 = (3E/4nk)(ac/r) where k =9/16[(l-v 2) + (1 - v 2 ) E / E s ] (2.13) Contact loading from blunt indenters generates cone cracking [105, 106]. Cracks grow in the following manner: 1) Pre-existing cracks or flaws are subjected to the tensile stresses just outside of the contact zone. Initiation of crack growth depends on the existence of pre-existing flaws and well fabricated specimens will yield high resistance to contact loading. 38 2) As the loading increases, a flaw grows and expands around the ring of the indenter forming a 'ring' crack at the surface. Tensile stresses are at a maximum on the surface where the edge of the indenter meets the material. 3) The crack grows downward into the material as loading continues to increase. Crack growth occurs in the tensile stress field where the ceramic is weakest. 4) During unloading, the crack closes as the tensile stress field becomes a compressive stress field. The load necessary for failure depends on the geometry of the indenter and can vary significantly [27, 107-109]. 2.4.4 Effect of Porosity on Mechanical Properties Porosity can be introduced into ceramics, metals and polymers as a fully porous material or as a coating [5]. Despite some disadvantages, porous materials are widely used due to their ability to form a mechanical interlocking with surrounding bone and soft tissue. This attribute is highly used in implant fixation applications. Porosity also can affect mechanical strength of the material and a summary of results are discussed next. Material strength can be affected by material imperfections or defects that can create internal stress concentrations [6, 87, 89]. The effect of porosity on material strength has been described by many researchers and an exponential relation is generally well accepted up to about 40-50% porosity [5, 6, 47, 88, 90, 105, 110]. Tensile and compressive strength are examples of properties that depend on the porosity and the following relations have been determined empirically to describe strength (usually in terms of strength at 0% porosity): 39 E = E0(\-apP-bpP2) [ 3 2 , 1 0 5 ] ( 2 . 1 4 ) E = E0(\-P)2 [ 1 0 5 ] ( 2 . 1 5 ) E = EoexV(-bpP) [ 8 8 , 9 0 ] ( 2 . 1 6 ) E = Eo ( 1 - P ) 2 [ 9 0 ] ( 2 . 1 7 ) l + (— -\)P PN For oxides, bp is typically 4 (3.5 for aluminum oxide) and ? N is assigned a value of 0.4 or less. Of the above equations, the relation derived by Sprigg (2.16) [88, 90] has been shown to give the best fit. Similar forms of the relation can also be applied for the influence of porosity on fracture energy and strength but with different material constant bp [32, 88]. The strength of porous materials is described as decreasing with an increase in porosity in all the above relations (a 5% porosity leads to a 20% reduction in stiffness). There are two main reasons for this; the first is that the load bearing area decreases as pores are introduced into the material and second is that stress becomes concentrated at pore locations [105]. Figure 2-4 illustrates the second point of stress concentrations. An introduction of a hole or pore can disrupt the stress field in the material and stress is able to concentrate at apex locations. For the example shown in Figure 2-4, the stress concentration at the ends of the major axis of the ellipse can be described by the equation [105]: 7 = Yo e x p ( - 6 p P ) ( 2 . 1 8 ) S = S0 e x p ( - ^ P ) ( 2 . 1 9 ) 40 °max = °1 1 + 2. P (2.20) The stress for an elliptical pore will depend on the size of the pore and the radius of curvature at the end of the major axis. For a circular pore, c = ?, and thus the stress concentration is approximately three times the nominal stress in the material. The equation also shows that for pores that are very sharp (small radius of curvature), the stresses can become very high. Unfortunately, the above equation can only describe a single hole in a plate and in practice many pores of various sizes and shapes exist, each interacting with each other creating a very complex stress state. Solutions for these situations are complex and beyond the scope of this project. t t Figure 2-4: Introduction of a pore or hole distorts the stress lines of a component. Force lines become thicker at the apex of the pore and larger stress fields result [32]. 2.4.5. Effect of Biological Environment on Material Strength Several researchers have studied the influence that non-standard environments can have on the material strength of alumina ceramics [6, 111-119]. Test environments include water, Ringer's solution (a mixture that is used to represent the physiological environment) and in vivo. Each of these environments demonstrates that strength degradation of alumina is possible in 41 physiological environments. A summary of results from the literature for aluminum oxide are presented in Table 2-4. Table 2-4: Reported strength reductions of Alumina placed in different environments. Strength Reduction Media Appl ied Stress? Reference - 3 0 % Ringer's Solution During aging [111] 9 to 10% De-ionized Water After aging [111] 35% (Fatigue Strength at Isotonic NaCI Solution During aging [112] 10 million cycles) - 0 % (High purity A l 2 0 3 ) Ringer's Solution During aging [116] 40% (with Silica and media permeation) - 0 % (>99.7% A l 2 0 3 ) In vivo After aging [117] Results from experiments vary and tend to be conflicting as some researchers report very little strength reduction while others report large strength reductions. One research group [111] even reported an increase in strength of alumina after 6 months of subcutaneous implantation. The flexural strength and Weibuli modulus were measured with the flexural strength showing an increase of up to 10% and no change in Weibuli modulus. The Weibuli modulus indicates the density of defects that can reach critical crack size with low values indicate a higher. No change in Weibuli modulus means there was no crack growth during the implantation. The increase in strength was believed to be due to cracks in the material being filled with tissue that helped to increase the flexural resistance. . Reported strength reductions may be caused by impurities in the materials. Silica has been shown to decrease the strength of the material via its reactions with the environment [113-115]. Silica present at the tips of stress concentrations or cracks can enhance the stress state through chemical reactions between the environment and the chemical bonds at the tips. The silica in the glass phase undergoes a hydration reaction which leads to further crack separation. Silica can also be present along the grain boundaries and intergranular failure has been observed as the primary mode of crack growth in aluminas with silica impurities [113-115]. This is easily concluded as crack growth in air for alumina is a mixture of intergranular and transgranular 42 failure. High purity alumina implants have shown very little reduction in strength over time [116, 117, 119]. The effect of the environment must be accompanied by a stress state to cause corrosion [136]. Alumina, like many ceramics, has a high resistance to corrosion due to the strong ionic and/or covalent bonds, but can still be susceptible to stress corrosion cracking [7, 11, 88, 120]. Strength tests that occur after the material has been placed in a corrosive media and then removed and tested in an inert environment should show little difference in results [120]. For corrosion to take place there is a dependence of environment on an applied stress as the two components acting separately do not produce the same result of both working together. In ceramics, corrosion refers to simple dissolution of the material or chemical reaction. Reactions occur at defects and can lead to bond breakage at the crack tips similarly to the reaction of silica with the environment. It has been reported that A l 3 + can be leached from bars of alumina after aging in saline solution or water [121]. A certain stress level or stress intensity is required to initiate crack growth which is the reason why corrosion only occurs if combined with an applied stress. Once the initiation stress is reached, environmental reactions increase the crack growth rate until failure. Figure 2-5 shows the evolution of crack growth from stress corrosion using the following relation of crack velocity during sub-critical crack growth and stress intensity factor at the crack tip. The diagram is the same as for normal sub-critical crack growth but the presence of water or physiological media increases the crack growth velocity for a given stress intensity factor and also allows crack initiation at a lower stress level [120, 122]. v = v f K V \KIcJ (2.21) 43 LN v B / c A K|C LN K, Figure 2-5: Stress corrosion crack growth diagram. No crack growth occurs until a specific stress intensity level is reached. Region A represents the crack growth rate governed by reactions with the environment. Region B shows a steady crack growth velocity limited by the rate of diffusion of environmental molecules to the crack tips. Region C is the stage where sub critical cracks have reached an unstable size and the failure toughness (for an inert environment) is reached. If insufficient levels of stress are applied to induce the stress corrosion mechanism, then no effect of environment may be observed. For the one reported case of 0% strength reduction for a fatigue test on samples in vitro, the applied fatigue stress was 25-30% that of the failure stress for the samples [116]. This applied stress is much lower than what others report as stress levels causing failure from environmental effects thus it is probably that the applied stress was insufficient to cause sub critical crack growth. The mechanisms of crack growth from environmental elements have been explained by a few by considering the reactions at a molecular level [123, 124] and by the presence of silica at interfaces and grain boundaries [113-115]. The bond rupture at the crack tip can be envisioned as shown in Figure 2-6. The bond rupture process begins as a force is applied to stretch the bond between the material molecules BB. As the bond is stretched, it reaches a critical separation point where the state becomes more energetically favorable for the bond to break. The energy required to break the bonds lowers when a chemically reactive species is present such as AA [124]. Species AA provides an alternative lower energy path to bond rupture and thus, allows crack growth at lower energies or applied stresses. For ceramics, this has been seen when silica impurities is present in the material and subject to degradation from water [124]. 44 The structure of the water allows it to easily break the bonds of silica. Figure 2-7 illustrates the process. Figure 2-6: Bond rupture at crack tip from environmental species [123]. Figure 2-7: Bond rupture of silica via water. The water provides a lower energy path to bond breakage during stretching of the silica bonds [124]. The effect of biological environment on porous materials is more detrimental when compared to dense materials. The porous surface creates more area exposed to the environment and thus 45 more area where corrosion can take place. In vivo tests on 35% porous alumina have shown strength reductions of up to 35-40% after 3 months [6, 121]. Fatigue tests in de-ionized water have shown strength reductions of 10% and tests in bovine blood have shown strength reductions of 15% [121]. The above literature review revealed that ceramics are an excellent choice for the housing material. Ceramics have good biocompatibility, high strength and can be used for long term applications. Their disadvantages were their brittleness and cost of manufacturing. Mechanical tests were needed to determine how the brittle nature of ceramics would limit the suitability of ceramics as the housing material. Fatigue, impact and Hertzian contact forces were needed to be tested to assess the housing strength. The susceptibility of the housing to these forces can increase as porosity is introduced into the material and when placing the material in a potentially corrosive medium. The use of porous ceramics can be used to provide fixation for the housing via tissue in growth. The foreign body response can be reduced by altering the surface chemistry, the use of a porous structure and using rounded corners instead of sharp corners. This information was used in the design of the housing and is presented next. 46 CHAPTER III: Design Design of the housing precedes any reliability testing. The overall dimensions, shape and material must be chosen to meet biological and mechanical criteria. Dimensions and shape must be chosen to suit the intended area of implantation, the electronics and to reduce the foreign body response while material is selected to provide good compatibility with biological tissue and resistance to any mechanical loading. The designed housing is shown in Figure 3-1 as a 3D solid modeled using a CAD program named PRO-Engineer (PTC). FEMLAB (Comsol Inc.), a finite element program, was used to help determine suitable dimensions. Detailed drawings of the enclosure are shown in appendix 1. The overall dimensions are 58mm x 30mm x 6.4mm. The length and width dimensions are at the maximum allowed criteria set by the Marine Mammal Research Consortium to provide maximum space for the electronics. The height is restricted such that it does not show a large skin protrusion or overly stretches the skin. It is also desired to have as small an implant as possible as literature has shown that larger implants produce a larger foreign body reaction. Figure 3-1: 3D Solid view of Electronics Enclosure 47 The inside shape of the housing was requested by the electronics group in the Steller sea-lion project to accommodate the electrical hardware, antenna and batteries (Figure 3-2). The inside shape is designed with different heights to properly fit all the electrical systems. The type of antenna selected for the application is a loop antenna which fits in the rounded space at the front of the housing. The solid section that the loop antenna encircles is used to provide reinforcement for the top lid against any applied loadings. Once the electronics have been placed in the housing, an epoxy is used to fill in any of the remaining spaces to reduce movement of the electronics and to provide mechanical reinforcement of the housing lid. The epoxy provides a strong reinforcement for the top lid as it prevents it from bending under applied pressures. Bending of the top lid can lead to high tensile stresses in the bottom side of the cover which will lead to early fracture as alumina has low tensile strength. This is especially important for any localized forces over the center of the cover plate such as in the case of puncture. Batteries Electrical Hardware Antenna t A \ , "N f v Figure 3-2: Outline of the housing's longitudinal cross section. The base and lid design was chosen to allow easy installation of the electrical equipment. All corners are rounded to reduce the foreign body response attributed to sharp corners or edges. The outer shape was designed with a taper to conform to the shape of the Steller sea-lions. Figures 3-3 and 3-4 shows the skull of a young Steller sea-lion with a mock implant housing to demonstrate the conformation of the housing to the Steller sea-lion head. The implant is placed on one side of the skull as a crest forms in the middle of the skull as the sea-lion ages. Although, the implant seems large by comparison to the skull of a young sea-lion, consultations with a veterinarian (Dr. Tamara Godbey, DVM) concluded that the size would not pose any 48 problems. The implant weight was also determined to be negligible as the weight of the implant must not exceed 3-5% the body weight of the animal [125]. Young Steller sea-lions are reported to weigh 23 kg which would give a maximum implant weight of 0.69 - 1.15kg [126]. The measured weight of the alumina housing without electronics is approximately 0.0245 kg. Figure 3-3: Side view of young Steller sea-lion skull with a mock implant. Ceramic materials were deemed optimal for this long-term application with aluminum oxide being the selected ceramic. Material selection was based on biocompatibility, strength, longevity and compatibility with the antenna and was discussed in the literature review. A design table (Table 3-1) shows how ceramics compared to other materials. Polymers were the worst choice due to their low strength and long-term life. Metals were also a poor choice as metals will absorb the transmission signal from the antenna thus blocking the signal. The longevity of metals was also questionable due to their susceptibility to corrosion. Ceramic materials are very strong but are also brittle and this brittleness can limit the use of ceramics in load bearing applications. Mechanical tests were needed to demonstrate the durability of ceramic materials in conditions simulating the in-service environment. 49 Figure 3-4: Top view of young Steller sea-lion skull with a mock implant. Table 3-1: Material Selection Design Table Biocompatibility Strength Longevity Antenna Compatibility Ceramics Metals Polymers v V V V X X V V V X Aluminum oxide is also very biologically inert and will be well tolerated by the body. Soft tissue reactions should be minimal however the fixation of the housing requires an additional characteristic. Porous materials have been shown to promote soft tissue in growth that can possible lead to material fixation. The lid of the housing was designed to have a layer of porous alumina on top of a dense alumina layer. The porous layer will be in contact with dermal tissue and should provide a path for tissue in growth while the solid layer maintains the hermetic sealing. A layer of porous alumina covers only the top as the overall dimensions of the material will increase with a layer of porous alumina covering the entire surface area. The overall strength of the material would also degrade as the amount of porous structures increases. The literature review shows that material strength significantly degrades with the introduction of pores and thus, the amount of pore structures should be as low as possible. The pore size must meet the 50 minimum requirements for tissue in growth which has been discussed in the literature review to be 5-15 um. The thickness of the lid with the addition of the porous layer must be as thin as possible without limiting the strength of the lid. Pore layer thickness has not been investigated in the literature and the thickness of a pore layer for proper tissue fixation needs to be determined to finalize the thickness of the housing lid. Fixation of materials in growing animals has only been sparsely studied with only qualitative measurements available. Quantitative measurements were required to properly justify the use of a pore layer to fixate the material and to demonstrate its anchorage. Mechanical strength was tested to evaluate the strength of the brittle ceramic housing under forces that simulate in-service conditions. Fixation experiments were also conducted to determine a final pore layer thickness and to quantify the amount of fixation that a porous layer will provide to an implant in a growing animal. The results from both showed the reliability of the housing design to be suitable for implantation into Steller sea-lions. 51 CHAPTER IV: Tissue Attachment and Implant Migration 4.1 Introduction Reports from the literature indicate that fixation of implants is currently still a problem [2, 3] with only one general qualitative study showing no gross migration on a limited number of samples [127], No reports have been found on the topic of implant migration in growing animals. The intended location of implantation for the radio frequency implant is sub-dermal at the back of the Steller sea-lion's head. Insertion into the location will occur in young (three month old) animals thus the implant will be placed in growing tissue. Movement within the growing tissue was a concern and fixation of the implant is desired so that it maintains its position at the back of the head. The overall objective of these experiments was to examine the fixation and biocompatibility of porous alumina discs that are representative of the material and thickness of the proposed housing. Previous literature was used to design the housing for biocompatibility and fixation in the area of implantation however it was not known what the response would be in growing animals. This experiment was used to clarify what the response will be to implants that represent the housing design. This experiment consisted of the implantation of aluminum oxide discs into the subcutaneous tissue of young rabbits (Oryctolagus Cuniculus) for three months. The implants that were fabricated for this experiment were manufactured to show similar characteristics to the housing design. Aluminum oxide material was used with the implant consisting of a layer of solid alumina, representing the base, and a layer of porous alumina representing the lid. The effect of thickness of the pore layer on implant fixation was not known and two thicknesses were used, 0.5mm and 1mm. A thinner pore layer was desired to reduce the overall dimensions of the housing to allow more inner space for the electronics. The hypothesis of the experiment was that the pore structures would yield implant fixation. Fixation was analyzed by histological assessment, relative 52 movement of implants from the location of implantation and by quantitatively measuring the attachment of implants to their surroundings. The force of attachment of implants to soft tissue was used to determine the amount of fixation in the surrounding tissue. Attachment force was examined by measuring the pull-out force of removing the implants from the surrounding tissue. The attachment force has only been investigated by a few researchers [130-134]. Results from this have varied as the type of material and pore size and structure has varied. This variability in results is expected as characteristics can vary between materials (i.e. biocompatibility and material friction) and pore structures that allow more tissue in growth (i.e. more inner pore volume) will undoubtedly provide more resistance to tissue removal. Biological variability has also been observed as some porous samples have shown some foreign body response, fibrous tissue in growth instead of surrounding tissue in growth or insufficient tissue in growth for mechanical tests [133]. A related conclusion that was made amongst the different research groups is that tissue tends to break or rip when tissue is torn from the implant's porous structure [133]. This conclusion is due to observed forces that are higher then researchers expected from a failure mode involving tissue overcoming frictional forces to be removed from the material. This also makes sense as tissue present in an interconnected pore structure would be wrapped around biomaterial and a tearing of tissue would be necessary to pull the tissue away. Thus, the attachment force is mostly dependent on the amount of tissue in growth (minimum pore size to allow tissue in growth and inner volume for tissue anchorage) and strength of in grown tissue with only some small dependence on material type. Reported results for different materials are shown in Table 4-1. Observations of similar tissue attachment indicate tissue in growth and fixation of the implants in their surroundings. Results from this experiment were compared to the recorded attachment forces listed in Table 4-1. 53 Table 4-1 Reported tissue attachment forces Material Pore Size (pm) Porosity (%) Test Duration Attachment Force Ref. Cobalt Alloy 5-25 35 4 weeks 1.0 ± 0.6 g/mm [133] 25-45 39 1.5 ± 0.9 g/mm 45-150 38 11.4 ±3.6 g/mm 8 weeks 4.3 ± 2.8 g/mm 6.7 ± 3.5 g/mm 13.0 ± 4.9 g/mm 12 weeks 4.6 ± 5.0 g/mm 6.4 ± 6.6 g/mm 8.3 ± 4.2 g/mm 16 weeks 6.8 ±5.1 g/mm 13.2 ± 4.0 g/mm 27.5 ± 10.3 g/mm e-PTFE 22 - 3 to 52 weeks 20.6 N [130] Tantalum 400-600 75-80 4 weeks 60.7 ± 37.4 g/mm [132] 8 weeks 70.9 ± 37.5 g/mm 16 weeks 89.4 ± 43.1 g/mm Gore-Tex 30-60 - 14 days 10N [134] 30 days 12.5 N 60 days 17N 90 days 21 N 4.2 Materials and Methods In general, alumina discs were fabricated and inserted into the sub-dermal region of young rabbits for three months. One-third of the implanted discs were removed and sent to a pathologist to perform histological analysis on the tissue reaction and fixation while the other two-thirds were used to mechanically test the fixation of discs in their surrounding tissue. Details of the procedures are listed below. 4.2.1 Manufacturing of Alumina samples In this experiment, the actual full housing design was not used because the porous layer was the main item to be tested with regards to implant fixation. The size of the housings was also a problem as they would be too large to be implanted in rabbits and manufacturing costs were also reduced by not employing the full housing design. The type of alumina powder used was 99.5% A l 2 0 3 and was supplied by the Materials Engineering Department at the University of British Columbia (UBC). An organic binder (a 6.7 wt% polyethylene glycol and polyvinyl alcohol mixture) was used to help in improving the adhesiveness (increasing the particle friction) of the alumina particles and burns out of the material at approximately 200-300 °C during sintering. 54 Fabrication of the implants was achieved using a cold axial pressing method (Figure 4-1). A simple hand press was used to compress the powder in a mold. The mold was manufactured using a high strength 4140 steel and is shown in Figure 4-1. The mold was designed to produce a disc implant diameter of 17.5mm. Detailed engineering drawings of the mold are contained in appendix 1. Die Powder pressed inside mold V • • V M o l d Figure 4-1: Alumina disc die and mold for axial pressing. To make the porous structure, 99.99% graphite (496588, Sigma-Aldrich, Oakville, ON) was mixed into the alumina powder at a specific weight percent. This method has been used by several others [136-138]. The particle size of the graphite used was 100 mesh to try and produce a micron sized open porosity. Two sets of alumina powders were prepared to create the two layers of solid and porous alumina. The first set was pure alumina mixed in with the organic binder and the second set was alumina powder mixed with graphite and the organic binder. To produce the two layered structure; first the alumina-graphite powder mixture was placed in the mold and flattened using the mold die. Then the pure alumina powder was placed on top of the alumina-graphite powder mixture in the mold. The overall thickness of the implants was 4 mm with half the discs having a 0.5 mm thick porous layer and others having a 1mm thick porous layer. 55 The materials were then pressed and finally sintered to 1400 °C (the maximum limit for the available ovens) for 2-3 hours. After sintering the material, a scanning electron microscope (SEM) was used to examine the surface structure. A pore structure is created when the graphite particles burn out of the alumina leaving a small opening in its path. The average pore size was measured by taking the measurements of randomly selected pores at different locations on the surface. Attempts were made to alter the pore size by varying the press load pressure from 4 to 7 tons/area2 and varying the wt% graphite used to make the pore structure. 4.2.2 Animal Surgery Procedures For implantation, 12 three month old female New Zealand White rabbits (Oryctolagus Cuniculus, Animal Care Center, UBC) were used with an average weight of 2.45 kg ± 0.22 kg. Steller sea-lions cannot be used as they are endangered animals and it is not ethical practice to use wild animals [139] thus rabbits were selected as rabbits have been used by many other research groups for similar experiments [20, 21, 134]. Rabbits are generally used because of their ease of handling, inexpensive purchase and housing costs and size [43]. The dermal tissue of rabbits is the intended implantation location. Differences exist between rabbits and Steller sea-lions in that the subcutaneous tissue in rabbits is more loosely attached to the dermal tissue. This was not a concern as the intended location of implantation was the dermal tissue and not the subcutaneous tissue. The animal care form to acquire permission from the Canadian Council of Animal Care (CCAC) to experiment is in appendix 2. Surgery dates were staggered to accommodate the schedule of the veterinarian (Dr. Tamara Godbey, DVM) and to ensure no severe complications occurred (Table 4-2). If one set of rabbits showed signs of rejection or severe inflammation, then the remaining rabbits would not be operated on until the problem was determined. Four rabbits were operated on at each surgery date and the duration of implantations was 98, 105 and 111 days. The duration of the study was 56 chosen as a minimum of 90 days implantation as required for attaining a cellular steady state around the implants [140]. Table 4-2: General Info on Rabbits used in Experiements (SQ - Subcutaneous, ID - Intradermal) Rabbit # Surgery Date Weight (kg) Suture Type 1 2 3 4 5 6 7 8 9 10 11 12 09/28/2005 09/28/2005 09/28/2005 09/28/2005 10/04/2005 10/04/2005 10/04/2005 10/04/2005 10/11/2005 10/11/2005 10/11/2005 10/11/2005 2.47 2.34 2.61 2.76 2.63 2.38 2.52 2.6 2.5 2.62 2 2:03 3-0 Monocryl SQ, 4-0 Nylon skin 3-0 Monocryl SQ, 4-0 Nylon skin 3-0 Monocryl SQ, 4-0 Nylon skin 3-0 Monocryl SQ, 4-0 Nylon skin 3-0 Monocryl SQ, 4-0 Vicryl ID 3-0 Monocryl SQ, 4-0 Vicryl ID 3-0 Monocryl SQ, 4-0 Vicryl ID 3-0 Monocryl SQ, 4-0 Vicryl ID 4-0 Monocryl SQ, 4-0 Monocryl ID 4-0 Monocryl SQ, 4-0 Monocryl ID 4-0 Monocryl SQ, 4-0 Monocryl ID 4-0 Monocryl SQ, 4-0 Monocryl ID In each rabbit, 4 locations of implantation were identified on the back of the rabbits. One location was a control consisting of only an incision and formation of a pocket under the skin (i.e. no implant in that location). The control was used to differentiate any abnormal response from being caused by the implants or by the surgery. The other three locations were implanted each with an alumina disc. In each rabbit, one alumina disc with porous layer of 0.5 mm, one disc with porous layer of 1 mm and a second disc of either 1 mm or 0.5 mm porous layer were implanted. The second disc was randomly selected and its purpose was for histological examination. The other two discs were intended for testing mechanical pull-out force. The control pocket was also sent for histological examination. Discs were implanted with the porous layer facing in a random direction (towards the skin or the body). In total, 36 discs were implanted with 18 having a porous layer of 0.5 mm and 18 having a porous layer of 1 mm. Selected locations for placing each type of implant were done using a randomization block design. The locations of each implant for each rabbit are shown in Figure 4-2. 57 Rabbit #1 0.5mm Control 1mm 1mm Rabbit m Control 1mm 0.5mm 1mm Rabbit #7 0.5 mm 1mm 1mm Control Rabbit #10 0.5 mm Control 1mm 1mm Rabbit #2 0.5mm 1mm Control 0.5 Rabbit #5 Control 1mm 0.5 mm 1mm Rabbit m Control 0.5 mm 0.5mm 1mm Rabbit #11 1mm Control 1mm 0.5mm Rabbit #3 0.5 mm 0.5 mm 1mm Control Rabbit #6 0.5mm Control 0.5mm 1mm Rabbit #9 1mm 0.5mm Control 0.5 mm Rabbit #12 1mm 0.5mm 0.5mm Control Figure 4-2:- Design matrix showing the locations of all implants in the back of the rabbits. The top of the blocks is where the head is and the bottom is where the tail is. The bold names indicate those that were sent for histological examination. Before surgery, the back of the rabbit was shaved, washed and disinfected. Implants were sterilized using an autoclave. Rabbits were administered a general anesthesia of Ketamine (2 mg/kg) induced intramuscularly. During surgery, anesthesia was maintained with Isoflurane (3-4%) induced via facemask. To maintain a consistent implant location within each rabbit, a template was used and referenced to the dorsal spinous process of the 2 n d thoracic vertebrae (Figure 4-3). Thus all implants were inserted into the same approximate position in each rabbit. Incisions were made caudal to the template locations and the sterile implants inserted into subcutaneous tissue pockets. The original aim was to implant the alumina discs right beneath the dermal layer however there was extreme difficulty in separating the cutaneous trunci muscle (right beneath the dermal layer and is the muscle responsible for raising hair) from the skin layer. Thus, it was only possible to insert implants into the subcutaneous tissue with the exception of implants in rabbit's number 9 and 12 where three implants were inserted into the dermal layer. Direction of porous layer was not controlled in the subcutaneous implants as the same type of tissue would be present around 58 the entire implant (opposed to the original intention of having dermal on the porous side with subcutaneous tissue on the non-porous side). Implants inserted into the dermal layer were oriented with the pore layer facing the epidermis. Figure 4-3: Template aligned to the dorsal spinous process of the 2nd thoracic vertebrae. All implants were inserted according to this template. Implants were not fixed with sutures. The wounds were closed using sutures described in Table 4-2. Suture types were changed from skin to subcutaneous due to rabbits in the first set attempting to remove the skin sutures by biting at them. Postoperative, rabbits were given an analgesic of Butorphanol (0.2 mg/kg) and an antibiotic of Baytril/Enrofloxacin (5 mg/kg) intramuscularly once daily for 5 days. After surgery, rabbits were placed in a cage and allowed to move unrestrictedly. After the three month implantation period, the animals were euthanized by administration of Xylazine (10 mg/kg) induced intramuscularly followed by pentobarbital (120 mg/kg) via ear vein. 59 The backs of the rabbits were then shaved and areas for extraction were marked. Two inch square blocks of tissue were extracted and placed in 10% buffered formalin for fixation (Figure 4-4). This was only performed for those samples intended for histology. Figure 4-4: Tissue block extraction from rabbits post mortem. 4.2.3 Tissue adhesion procedures A digital force gauge (DPS-110, Imada Inc., Northbrook II.) was used to determine the pull-out force required to remove an implant from its tissue surroundings following a method described in [130]. Before performing the tissue adhesion tests, a ventral incision to the implant location was made which allowed access to the implant. The rabbit was then placed on a test stand (LV-220, Imada Inc., Northbrook II.) below the force gauge (Figure 4-5). Grips (SC-8, Imada Inc, Northbrook II.), attached to the force gauge, were used to grab onto the implant (Figure 4-6). 60 Figure 4-5: Experimental setup for implant tissue adhesion tests. Connected to the force gauge was a laptop computer using a special RS-232 cable (CB-203, Imada Inc. Northbrook II.). Data was recorded using a terminal program called Simpleterm Gold (Ptronix). Data from the force gauge was queried 20 times per second to acquire enough data for load-time curves. Right before the test, the force gauge was set to zero and then the implant was pulled from the animal using a vertical force. The veterinarian held the animal down as the implant was pulled out. Figure 4-7 shows an implant removed from a rabbit with surrounding tissue. 61 Figure 4-6: Force gauge attached to implanted alumina discs via force grips. Figure 4-7: Extracted alumina disc with surrounding fat tissue. 62 4.2.4 Histology Histological examination was performed by Dr. Nick Nation (Animal Pathology Services APS" LTD., Edmonton, Alberta). Blocks of tissue containing alumina discs with 0.5mm and 1mm thick pore layers were sent for examination as well as control blocks (tissue that had been initially cut open without insertion of an implant). A series of skin samples obtained from twelve rabbits were sent and received fixed in formalin. There were two sections of skin per rabbit: one with an implanted alumina disc in the subcutaneous connective tissue, and one control. Fixed skin samples were examined grossly, the implanted transmitters were removed, and cross sections were made of the skin through the area in which the transmitter was located, or, in control samples, through the middle of the submitted tissue sample. All trimmed samples were placed in tissue cassettes, processed into paraffin blocks, sectioned at 5 microns, stained with hematoxylin and eosin stain and mounted on glass slides using standard techniques. The amount of pigment was assessed on a five point scale and also recorded for each sample in a tabular form. Slides were examined by a board certified veterinary pathologist and each section of tissue was either recorded as normal, or a description was made of any abnormalities. The following measurements were recorded: 1) The number of inflammatory cells present on both smooth and porous surfaces (i.e. macrophages, foreign body giant cells). This measurement gives an indication of the type of foreign body response and material compatibility with the host tissue. 2) The presence of blood vessels or blood cells was measured as the proximity or presence of blood vessels near the implant surfaces. Distance was measured in microns and was defined as the distance of blood vessels to the surface of the implants. The presence of blood vessels indicates a normal wound healing process and a remodeling of a new connective tissue matrix in the wound area. Chronic inflammation from an implant would prevent the completion of the wound healing response thus the formation and presence of blood vessels is an indication of no 63 chronic inflammation [20, 68, 69, 141]. Close proximity to blood vessels also leads to thinner fibrous encapsulations as nutrient supplies are more available [142]. 3) The amount of vascularization refers to the density of capillaries and vessels near the implant. A high density of blood vessels is desired for the same reason as mentioned above for the presence of blood vessels. Avascular fibrous capsules are indications of chronic inflammation and are thus not desired [141]. 4) The extent of fibrous tissue encapsulation per area covered was measured to give an indication of the degree of encapsulation. The area of fibrous tissue was measured and related to the amount of implant area in contact with the host tissue [143]. 5) The degree of tissue in growth in the porous surface was measured as the depth, in microns, of tissue growth from the surface. Tissue penetration was measured by examination of implant cross sections. Several measurements were made per cross section. Tissue in growth is an important parameter to measure to analyze the differences in pull-out forces to 0.5 and 1 mm pore layer thickness. 6) The cell thickness of encapsulation for both smooth and porous surfaces was measured. Thickness was calculated as the distance from the surface to the outer limit of the capsule adjacent to the surrounding tissue. The encapsulation thickness was measured at four representative sites, using a micrometer eyepiece, and the average thickness recorded in a tabular format. The thickness of encapsulation will be an indicator of biocompatibility and implant integration into host tissue [61]. A thinner encapsulation can also indicate tissue fixation as less irritation occurs from the movement of the implant within the encapsulation [71]. 7) The type of tissue around the implant encapsulation to determine what tissue the implant is fixated in. 64 4.2.5 Implant movement tracking < During implantation of the alumina samples, a template was used to ensure the same placement of all samples in the rabbits. The template is shown in Figure 4-3. Implant movement was checked on by the veterinarian on three different dates. Before tissue adhesion tests and implant extraction for histology, the distance from the original location of implantation to the current implant location was measured. General skin growth was tracked in the first four rabbits. Tattoos were made where the implant was after insertion. As the animal grew, the tattoo locations moved as the body mass of the rabbit increased and skin tissue moved. The tattoo locations were identical to where the implant template marked the initial implant locations. Thus, skin tissue growth was measured as the distance from the original location of implantation (as indicated by the template aligned to the spinal position) to the final location of the tattoos. 4.3 Results and Discussion Herein are the results from manufacturing the alumina discs and from experiments concerned with the fixation of implants with porous surfaces. 4.3.1 Manufacturing of Alumina Samples Results and Discussion Aluminum oxide discs were made as shown in Figure 4-8. Eighteen discs were manufactured with a porous layer thickness of approximately 0.5mm and eighteen discs were manufactured with a porous layer thickness of approximately 1mm. The two different layers, joined together to form a non-homogeneous material, produced a non-uniform cross-sectional thickness (Figure 4-9). This effect is called Warpage and was expected as the sintering densification mechanisms are different for powders with and without pore forming agents [144, 145]. The alumina powder without graphite was able to contract and shrink to produce a highly dense layer as material is not hindered to fill any spaces. This is how the normal sintering process works. 65 The alumina powder mixed with graphite was not able to sinter to the same degree as large pores formed by the graphite powder burning out. The larger pores created by the carbon removal prevents the powder from reaching the same density as the pure alumina layer (i.e. coarsening). Some material still fills in the pores or spaces however the process is not complete (and this explains why pore sizes are not the same as the graphite particle size used). The result reflected the two different shrinkages as the porous layer bulged outwards from the solid layer. To achieve a more uniform section, the edges were sanded using silica paper. The surfaces of the porous and solid layer are shown in Figures 4-10 and 4-11. Figure 4-8: Aluminum Oxide fabricated discs. Diameter of 17.5mm and 4mm thickness 66 Figure 4-9: Fabrication of porous and solid layered disc produced a non-uniform thickness with a higher densification of the solid layer (bottom). Figure 4-11: SEM view of the solid/smooth surface. To achieve the desired pore structure, several different techniques were attempted to get a reasonable pore size. The first technique came from exerting different load pressures when pressing the green body compact as observed in [146]. It has been reported and shown [146] that as the forming pressure is increased, both the pore size and porosity decrease. An increase in the pore size was examined similarly by reducing the forming pressure and the results obtained are summarized in Figure 4-12. No observable difference is seen with the increase in forming pressure. The large standard deviation leads there to being no statistical difference between the different press pressures even though the means vary slightly. Thus the 7 tons press force was selected to get the highest degree of powder packing density and for higher strength [144, 146]. Press force was varied using 20 wt% graphite in the porous alumina-carbon layer. 68 -\ 1 1 1 1 4 5 6 7 8 Applied press load (tons) Figure 4-12: Graph of pore size variation with applied load during axial pressing of the powder. Only small differences in pore size were observed. Weight percent of graphite was also varied from 10 wt% to 50 wt%. Negligible differences in pore size were observed however porosity did change as expected. However, it was very easily observed that graphite weight percentages exceeding 30% yielded a porous layer that was very weak and readable broke during handling. The manufacturing process is most likely the reason for the weak structure as will be discussed. Therefore, due to limited available equipment, using a weight percent of graphite that exceeds -30% is not possible. The difficulty in achieving high porosities using ceramic powdered particles has been reported [138]. Ceramics with porosities above 40-50% lack much strength because of the thin connections made between alumina particles between pores. There is also a minimum volume amount of graphite required to have completely open porosity in the pore structure [138]. Volume percents of less than -10-12% will yield a closed porous structure and volume percents exceeding that of 20% will yield a completely open structure. Graphite weight percents 10 and 20% give graphite volume percents of 17.5 and 35% respectively. Therefore, a 20 wt% graphite alumina mixture was selected to give a complete open porosity and solid connected porous structure. _ 50 | 40 V 30 N w 20 2> o 10 0 69 The final average pore size attained was 32.8 ± 15.4 microns with an associated 35% porosity [137]. The types of pores observed are shown in Figures 4-13 and 4-14. The solid surface at higher magnifications is shown in Figure 4-15. Figure 4-15 clearly shows the final sintering state of the alumina powder particles. Spaces or gaps still exist between particles indicating the sintering process could be improved by increasing sintering temperature and time or better control of the sintering temperature to account for burn-off of binders [137, 144]. The former is something that was difficult to achieve using the given equipment however the latter could have been modified. Proper control of binder burn-off can significantly improve the densification of the product as too-rapid removal of the binder can result in the formation of cracks or spaces between particles. Doing so will increase the density of the final product yielding a much stronger material. The final product was selected because it has a few advantages. First, the smaller pore size is better for mechanical strength [135]. At a given porosity, the stress concentrations are higher for surfaces with a lower density of defects (large pores) than for surfaces with a higher density of defects (small pores). The interactions of pores cause a reduction in stress concentration level and more of the applied load is supported by the material in between pores. The reduced applied load from this can be illustrated by comparing stress failure from stress concentrations and a reduced load bearing cross section. Stress concentrations can drastically increase the applied stress and a simple circular hole in a plate can produce local stresses three times that of the stress field in the rest of the material. The effect of having a reduced load bearing cross section is not as drastic. For example, having a porosity of 35% means a load bearing area of 65% which translates to an increase of stress by only 1.5 times. The smaller pores are also advantageous as the volume of porous interconnections increases as pore size decreases (for a given porosity). These smaller but numerous interconnections provides a good framework for tissue in growth and fixation. Inner pore connectivity volume is one of the most important factors for achieving biological fixation. 70 Figure 4-14: Surface pores viewed under SEM (250x). 71 Figure 4-15: High magnification (4500x) view of the surface. Alumina particles of many different sizes are observed with small gaps in between. Higher temperatures at longer sintering periods can reduce the sizes of these gaps. 4.3.2 Experimental Results and Discussion During the experimental period the skin of the young rabbits, where discs were implanted, grew uniformly approximately 4.69 cm towards the rear. At the same time the rabbits gained approximately 1.85 kg equal to a 75% weight increase. According to histological examination, implants were surrounded by mature fibrous connective tissue with encapsulating tissue growing into the porous layer. The encapsulation thickness and tissue type were typical for porous implants [29, 31]. Eighteen out of the 36 discs moved with the incision point on the skin, while the other 18 migrated further than 1 cm away from the incision point. The average migration distance was 1.74 ± 1.94 cm and only two implants showed gross migration (approximately 7.5cm). Pore layer thickness did not effect force measurements or tissue fixation (as viewed from histology). This means that the smaller pore layer thickness can be used in the tag application to reduce the size of the housing. Contradictory to what is implied in the literature; tissue in growth does not lead to implant fixation in the surrounding tissue. 72 The histology report is in appendix 3 and will be referred to throughout the section as notes in quotations. More detailed results follow. General characteristics of implant migration The results from the movement of the incision points are shown in Table 4-3. The incision points were only measured for the first four rabbits and the table lists the movement of each of the incision points. All incision points showed movement in the caudal-ventral direction. The implants were expected to follow this same movement and should have also moved that same distance and direction. However, approximately 50% of the implants moved further than the incision points. To account for this additional movement, an adjusted migration distance was calculated as the migration distance minus 4.69cm (Table 4-3). Distances that were less than 1 cm were assumed to have not migrated and are shown in the table to have a 0 cm adjusted migration. Disc migration results are summarized in Table 4-4. The direction that the pore layer faced is also listed as most implants were inserted with pore layer facing a random direction. Migration was not affected by pore layer thickness and pore layer orientation. Individual animal effects are not evident as patterns for implant migration within animals were not observed. Table 4-3: Skin growth movement tracked using a tattoo marker on the rabbit's skin. The mean migration distance is used to adjust implant migration distance in Table 4-4. CV -Caudal Ventral, C - Caudal, V - Ventral. Rabbit 1 Rabbit 2 Rabbit 3 Rabbit 4 Upper-Right 6.6 CV 4.5 CV 2.5 V 4.5 CV Lower-Right 7.4 CV 4 CV 4.5 V 4 C Lower-Left 6.9 C 4.2 V 4.5 V 4 C Upper-Left 7.2 C 3.8 C 2.5 V 4 C Total Mean Movement - 4.69cm Standard Deviation - 1.48cm 73 Table 4 -4 : Implant Migration Table. The distance that each implant moved was adjusted using the determined skin growth from Table 4 -3 . For pore layer face, up refers to a pore layer facing the skin and down refers to a pore layer facing muscle tissue. Listed next to the implant location is the thickness of the pore layer. Rabbit # Implant Location Migration (cm) Adj. Migration (cm) Pore layer face 1 Front left,0.5mm 7.2 2.51 down Back right, 1mm 11.9 7.21 up Back left, 1mm 8.6 3.91 up 2 Back Right,0.5mm 4 0 up Front right, 1mm 5.66 0.97 up Front left, 0.5mm 4.3 0 down 3 Back Left, 1mm 6.5 1.81 down Front Left, 0.5mm 3.9 0 down Front right, 0.5mm 5.5 0.81 down 4 Front Right, 1mm 4.5 0 up Back Left, 0.5mm 4 0 up Back Right, 1mm 4 0 up 5 Back left,0.5mm 7.6 2.91 up Back right, 1 mm 8.6 3.91 up Front right, 1mm 8.2 .3.51 down 6 Back Left, 0.5mm 6.7 2.01 up Back Right, 1mm 8.2 3.51 down Front left, 0.5mm 6.4 1.71 up 7 Front Right, 1mm 6 1.31 up Front Left, 0.5mm 3.8 0 down Back Left, 1mm 6.5 1.81 up 8 Front Right,0.5mm 12.5 7.81 down back right, 1 mm 7.5 2.81 up Back left, 0.5mm 6.8 2.11 up 9 Front Right,0.5mm 3.2 0 up Front left, 1mm 3.5 0 up Back right, 0.5mm 6.2 1.51 up 10 Front left,0.5mm 3.4 0 up Back left, 1 mm 6 1.31 down Back right, 1mm 5.3 0.61 up 11 Back right,0.5mm 5.6 0.91 up Front left, 1mm 3.5 0 up Back left, 1mm 7 2.31 up 12 Front right,0.5mm 2.3 0 up Front left, 1 mm 2.5 0 up Back left, 0.5mm 4.3 0 up Table 4-5 shows the calculated mean distances. ANOVA showed no significant difference in tag migration between pore layers of 0.5mm and 1mm thickness (p < 0.97). The total average migration distance for implants is approximately 1.74cm. 74 Table 4-5: Mean adjusted migration distances with standard deviation. The sample size is shown in brackets. Type Pore layer up (cm) Pore layer down (cm) Total Average (cm) 0.5mm thick layers 1.12 ± 1.02 (10) 1.85 ±2.81 (6) 1.39 ± 1.93 (16) 1mm thick layers 1.91 ± 2.05 (13) 2.54 ± 0.99 (4) 2.06 ± 1.87 (17) Total average 1.56 ± 1.76 (23) 2.13 ± 2.29 (10) 1.74 ± 1.93 (33) The results show that implant fixation did not occur in all cases as some implants showed movement while others did not. There are three possible causes that could explain the migration of implants in the host tissue; these are 1) tissue adhesions and the area of implantation, 2) the foreign body response and 3) the presence of externallyapplied forces. All three possible causes are mechanisms that work to move implants in the body [86]. To determine if these mechanisms played a role in the movement of the implanted alumina discs, the characteristics of discs that moved were compared to discs that did not. Characteristics that appeared in both immobile and migrated discs are not likely to be responsible for implant migration. Any discrepancies in available data might explain the difference in migration. Tissue Adhesion and the Area of Implantation Tissue adhesion was investigated using histology and pull-out force measurements. Histology was used to verify tissue in growth and the attachment of encapsulating tissue to the implant while pull-out force measurements were used to quantify the amount of implant attachment to its surrounding tissue. Histology results Firmly attached implants would not move unless subjected to extreme forces that would rip or break the adjacent tissue. Histological results could reveal tissue damage as more recent wound healing activity would be present similar to what would be expected from the initial implantation. However histological examination showed "no gross evidence of inflammation or a foreign body reaction in any of the samples" and that "each of the implants was located in a well vascularised pouch of connective tissue". Thus no implants showed any gross differences. The histology report 75 also states that "the connective tissue was growing into the porous, but not the smooth, surface of some implants." Figure 4-16 shows the fibrous tissue encapsulation and surrounding tissue. Implant location Figure 4-16: Cross section view of the tissue around the implants. The implant pouch is at the bottom of the image as indicated. Magnification is 100x. Implants that are not firmly attached may be easily moved from external forces. The attachment of the fibrous encapsulation to the implant surface was described using a ranking scale (see appendix 3). The interaction of the encapsulation with the implant shows that some discs had signs of tissue in growth while others did not. A correlation between tissue attachment and migration was calculated to see if the lack of tissue attachment in some implants leads to migration. Figure 4-17 shows that a direct relation did not exist between tissue attachment and implant migration (p < 0.78 using a Student t-test). ANOVA showed no significant difference between tissue attachment when comparing discs that had migrated and discs that had not migrated (p < 0.41). Tissue attachment was also compared for different pore layer thicknesses and no statistical differences for tissue attachment were observed (p < 0.29 using ANOVA). Figure 4-18 shows the scatter in results for the relation between tissue attachment and migration for different pore layer thicknesses. 76 o 5 4 3 c o 1 2 f 1 0 SE= 0.98 • Migrated • Not Migrated SE= 0.45 0 1 2 3 4 5 Tissue Attachment Rating (0 - none, 4 - Close Attachment) Figure 4-17: Relation between migration and tissue attachment for migrated and non-migrated implants. The dashed line is for migrated implants and the solid line is for non-migrated implants. (SE equals standard error of relation) 0 • 1mm • 0.5mm SE = 1.52 * SE = 1.47 0 1 2 3 4 5 Tissue Attachment Rating (0 - none, 4 - Close Attachment) Figure 4-18: Relation between migration and tissue attachment showing results for different pore layer thicknesses. The dashed line is for 0.5mm data points and the solid is for 1mm data points. The lack of tissue attachment (5 out of 12 implants showed no attachment) viewed during histological examination may be attributed to porosity or pore size. It has been shown that a minimum of 5-15 microns is required for tissue in growth into alumina ceramic [35] however the 77 rate of in growth can vary greatly as pore size increases [129]. A doubling of the pore size can lead to an in growth rate increase of 5 times [129]. The pore size used in this experiment was approximately 32 microns which is large enough to accommodate tissue in growth according to other sources [33, 35, 66]. The results from histology (see appendix 3) revealed that not all of the implants were tightly attached to their encapsulated tissue. This can mean that either the pore size used was insufficient for tissue in growth or that movement of implants prevented them from anchoring in place. A longer term study may have revealed different results as the'three month study may not have been long enough for tissue to firmly grow into pores. The use of larger pore sizes would show if any differences in implant movement can be achieved with a higher rate of tissue in growth and larger area for tissue penetration. Also, using implants with all porous surfaces instead of just one porous surface may have improved attachment. However, in this case the degree of tissue attachment or lack of tissue attachment did not correlate to implant movement. Low levels of tissue attachment can lead to implant movement when in combination with external forces such as what would occur during rabbit fights. These concepts are discussed next. Pull-out force results The tissue attachment of the encapsulation to the surrounding tissue of implants was also measured using the force of extraction procedure mentioned in section 4.2.3. Quantitative measurements were collected to describe the integration of implants with their biological surroundings by measuring the removal force. The failure mode was also noted. Several key observations were made during the tests: 1) Some encapsulated implants were embedded in pouches of fat layers and removal of the implants required a stretching and tearing of the fat tissue. Results from these samples yielded very high forces however these forces were more indicative of the failure force of the fat tissue. 78 2) The surfaces of some implants clearly showed small blood vessels that had grown into the porous network indicating some tissue in growth and attachment. This supports the inconsistent tissue attachment observations made in the histology report. Not all implants showed visible tissue in growth which agrees with the histology report indicating tissue attachment in only a few implants. 3) Some encapsulated implants were highly embedded in their surrounding tissue and could not be grabbed onto by our test apparatus. Additional incisions had to be made to be able to firmly grab onto the implant to extract it from the surrounding tissue. These incisions led to lower recorded extraction forces. Rabbit's 9 and 12 particularly showed this characteristic of highly embedded encapsulated implants. Results from three implants from rabbit's 9 and 12 were omitted from the upcoming analysis because those implants were tightly embedded into the dermal layer and extraction was impossible without additional incisions. Removal of those implants required more cutting of the surrounding tissue than the others. These three implants were from rabbit #9 - front right and rabbit #12 - front right and front left. The recorded peak pull-out forces for all other discs (those that were inserted into the subcutaneous tissue) are shown in Table 4-6. Peak values were recorded and comparisons were made between discs with 0.5 and 1mm thick pore layers and with discs that had pore layers facing dermal tissue or fat tissue. No statistical differences were observed (p < 0.29 for orientation and p < 0.7 for pore layer thickness using ANOVA). A summary of those peak forces is shown in Table 4-7. Force loading graphs of each implant is shown in appendix 4 with comparisons between locations, pore layer thickness and implant orientation. Table 4-6: Recorded peak pull-out force for each alumina disc. Results for rabbit 9 and 12 are omitted due to different implantation position. Rabbit # Implant Location and Pore layer Thickness Peak Force (N) 1 Front left,0.5mm 20.2 Back right, 1mm 129.2 2 Back Right,0.5mm 104 79 Front right, 1 mm 60.3 3 Back Left, 1mm 68.3 Front Left, 0.5mm 54.3 4 Front Right, 1mm 56.8 Back Left, 0.5mm 73 5 Back left,0.5mm 68.8 Back right, 1 mm 66.4 6 Back Left, 0.5mm 75.7 Back Right, 1mm 98.5 7 Front Right, 1mm 48.7 Front Left, 0.5mm 43 8 Front Right,0.5mm 94.2 back right, 1mm 70 9 Front left, 1mm 81 10 Front left,0.5mm 118.8 Back left, 1mm 68 11 Back right,0.5mm 48.2 Front left, 1mm 72 Table 4-7: Summary of peak forces for different implant orientations and pore layer thicknesses. The sample size is shown in brackets for each case. Type Pore layer up (cm) Pore layer down (cm) Total Average (cm) 0.5mm thick layers 81.42 ± 25.59 (6) 52.93 ± 30.96 (4) 70.02 ± 29.99 (10) 1mm thick layers 73.05 ± 24.76 (8) 78.27 ± 17.52 (3) 74.47 ± 22.28 (11) Total average 76.64 ± 24.5 (14) 63.79 ± 27.66 (7) 72.35 + 25.04(21) The effect of location on recorded pull-out force was also examined and results are summarized in Table 4-8. Although the front locations show a mean peak force lower than the rear locations, there is a large variability which leads to no statistical difference (p < 0.26 using ANOVA). Thus, the location of the initial implantation did not affect the amount of force required to extract implants from tissue. This is expected as the type of connective tissue surrounding the implants is expected to be the same regardless of the location in the back. 80 Table 4-8: Mean peak pull-out forces for the different implantation locations. No statistical difference was observed (p < 0.5). The sample size is shown in brackets for each case. Implant Location Mean Peak Force Front Left 60.61 + 30.64 (7) Back Left 70.76 + 3.06(5) Front Right 64.74 ± 15.55 (5) Back Right 86.05 ± 27.13 (6) The location on the back may not affect implant fixation however the area of implantation within the back tissue may be related to migration. Rabbits possess skin that is more loosely attached to the underlying tissue than many other animals or mammals and can easily be pushed or pulled away from the body [155]. The dermal layers are attached to the underlying muscle through a loose fibrous connective tissue called areolar tissue which consists of various elastic fibers. Also, part of the rabbit's subcutaneous tissue is adipose tissue which is fat storing tissue. The looseness of the subcutaneous tissue provides little resistance to the movement of any object within the tissue. The looseness of adipose tissue has also been described in detail in rats [147]. The layer between the skin and underlying muscle is called the sliding layer and allows the contractions of muscle relative to the dermal layers. Very weak forces are required to move the sliding layer and the layer allows relative movement of up to 10-15 mm. The sliding layer itself is composed of thin collagen sheets of elastic fibers with each sheet connected to each other through very sparse interconnections. During implantation, there was difficulty for the veterinarian to separate the dermal layer and the cutaneous trunci (the muscle underneath the skin that causes hair to rise) and thus most of the alumina discs were inserted into the loose subcutaneous tissue or the sliding layer. Even at the time of extraction, it was observed that some implants could be easily moved by applying an external force. Implants positioned in the sliding layer would exhibit this behaviour as the implant encapsulation forms very little interconnectivity with the collagen sheets [43, 156]. Histological results show that the implant encapsulations were connected only to dermal connective tissue and not to subcutaneous tissue. Implant movement can thus be related to the amount of encapsulation attachment to dermal connective tissue. Implants that are inserted into deeper 81 pockets of fat will experience more movement as little fixation can occur; and implants that are placed adjacent to the dermal layers are more likely to fixate with the skin and move with the skin as it grows. General body movement (from muscle groups) and external forces (from other animals or objects) would also easily lead to more implant movement of discs placed deeper inside the fat tissue. Another consideration is the differential growth of the dermal tissue and subcutaneous tissue. Little connectivity exists between the layers and therefore growth of tissue in the subcutaneous layer may not follow the growth pattern of dermal tissue. This would lead to movement of implants inserted into the subcutaneous fat region that cannot be accurately described by the use of skin tattoos. The exact nature of how tissue grows needs to be addressed to more accurately describe this movement however the movement of implants in the dermal layer may be able to be described by the change in tattoo location. There were three implants that the veterinarian could insert into the dermal layer (in between the cutaneous trunci and epidermis) and these implants were observed to be more securely embedded in tissue at the time of extraction. As mentioned previously, those three implants were omitted from peak force and migration analysis due to different positioning however measurements were still made. As observed in Table 4-4, the highlighted implants (from rabbits 9 and 12) showed no relative movement to the tattoo markers. A larger sample size is required to further prove that insertion into this layer or layers that do not slide provide more stability. One of the key observations made during the tests was that the higher recorded pull-out forces were for implants that were embedded in pouches of fat tissue. These values were observed to be more reflective of the amount of force required to tear or break the fat tissue surrounding the implants rather than the force required to tear the tissue away from the surfaces of the alumina discs. Figure 4-8 shows an extracted implant with tissue still attached indicating that the encapsulation tissue surrounding the implant was not separated. Thus, the mode of removal was 82 observed to be the failure of the tissue surrounding the encapsulated implant instead of the failure of the tissue-alumina interface. This indicates a weak connection between the implants fibrous capsule and the surrounding tissue as fibrous tissue was removed along with the implant. The highest pull-out forces were, therefore, due to the stretching and tearing of the surrounding pouches of fat tissue. Histology showed that the adjacent tissue to implants was connected to only the dermal tissue on one side and not to the adipose tissue surrounding the rest of the implant. The collagen sheets in the sliding layer show little connectivity with the above dermal layers and it can be expected that the elastic sheets would connect very little with the encapsulation tissue around the implants [146]. This explains the easy removal of the implant with encapsulated tissue from the fat layers. The recorded forces are then more representative of the fat tissue breaking force as opposed to the extraction force. They were high as the collagen sheets and elastic fibers provide the fat tissue with a high tensile strength [148]. The actual force to extract the implants from their surrounding tissue would have been much lower as the implant's encapsulation is only loosely attached to adjacent fat tissue. Figure 4-19 shows the trend of pull-out force versus migration. A correlation exists between implants that have migrated and peak pull-out force (p < 0.05 using a Student t-test). Similarly, no statistical difference is observed (p < 0.84 using ANOVA) between peak pull-out force of implants that migrated versus implants that did not migrate. A likely explanation for these results relates to the type of tissue the implant is embedded in. Implants that are more embedded in the fat tissue are expected to show more movement as little attachment exists between the encapsulated tissue and the adjacent tissue. The peak pull-out force, however, would be high as mentioned above due to the collagen and elastin content in the fat tissue. These implants migrated with the growth of the fat layers while implants that showed little or no movement must have integrated with the dermal tissue and migrated with the skin growth. Implants that were integrated into the dermal tissue required some extra incisions to attach the apparatus to the implant for measuring the pull-out force. The results showed lower removal forces although the 83 implants were more firmly fixed than implants extracted from fat tissue. The differing characteristic that lead to some implants moving while others did not was their connection of encapsulation tissue with fat or dermal tissue. Figure 4-19 shows how the implants that experienced more migration (due to positioning in fat tissue) also showed a higher peak pull-out force. 0 2 4 6 8 10 Migration (cm) Figure 4-19: Migration vs. Peak pull-out force graph. For the case of migrated samples, peak pull-out force increased with migration distance (p < 0.05). The solid line represents the trend line for migrated implants only. A comparison between the tissue tensile breaking loads from the literature [149-154] and the observed results show similarities that confirm the measured fat tissue breaking force. Although there is no data available for rabbit subcutaneous tissue, it can be expected to have a tensile force less than many other types of tissue as it is a loose connective tissue. Results for the tensile properties of soft tissue are shown in Table 4-9. The average peak force was 72 N which is less than the tensile breaking loads for most other types of tissue but still in the same approximate magnitude. Thus, the recorded peak forces that were highest seem to correspond to the breaking strength of fat tissue as the observed forces were much higher than what has been previously reported. 84 Table 4-9: Tissue tensile break loads from the literature. Type of Tissue Tensile Breaking Load Reference Rabbit Stomach Canine Aorta Human Bladder Rabbit Muscle Sheep Trachea Rabbit Ligament Rabbit ACL 200 N [154] 38.55 N (Medial) [154] 117.6 N (Lateral) 120 g/mm [151] 400-485 N [152] 80 g/mm [150] 50 g/mm [149] 198 N [153] The lower recorded forces showed in Figure 4-19 corresponded to implants that were more firmly attached to dermal tissue as they were well embedded. The implants that were in fat tissue were embedded in the tissue but could be moved around easily to grab onto using our test apparatus. To measure the extraction force of the implants attached to the dermal layer, extra incisions had to be made so that the grips could attach to the alumina discs. The results showed lower removal forces although the implants were more firmly fixed than the implants extracted from fat tissue. Variability in results may be attributed to the degree of attachment of the fibrous capsule to the above dermal tissue. This would indicate that fixation is more related to the connection between the fibrous encapsulation and surrounding tissue then between the porous surface and surrounding tissue. No correlations were observed between peak force and any implant characteristic (pore orientation and pore layer thickness) and similarly no relation existed for peak force and implant migration. Large variability in tissue attachment experiments has been seen by others and is expected when working with biological systems [132, 133]. Thus it is not unexpected to see the same with these research results. Several reasons can be used to explain the spread of results. Tissue collagen content or tissue maturity and organization can affect the attachment strength [133]. Results can also vary between animals as the wound healing response and tissue content can differ between test subjects of different health and age [43]. More samples per animal would be needed to 85 discount any statistical differences between rabbits. In general, more samples are needed to get more precise results with lower p-values. Foreign Body Response Another source causing implant migration is the foreign body response. Surgical incisions are required for the alumina discs to be implanted and this causes injury to the animal. Thus, the implants are subject to the wound healing response as soon as they are implanted [43, 47, 53, 156, 157]. This means the implant will be subjected to a foreign body response regardless of biocompatibility. The wound healing process begins with inflammation followed by the formation of granulation and fibrous tissue that forms a base tissue matrix for remodeling of the injury site. The final stage of the wound healing process involves the replacement of the fibrous tissue with a more functional and mature collagen matrix that is specific to the area of injury [53]. The presence of the implant slows the wound healing process and thus the implant is subjected to a longer period of inflammation [157]. The initial stage of wound healing involves the presence of inflammatory cells such as macrophages and lymphocytes. These inflammatory cells interact with the alumina discs and despite the good biocompatibility of aluminum oxide, there is still some reaction as is the case with all materials. Therefore, a small degree of fibrous encapsulation surrounds the implants preventing a good integration between the implant material and host tissue. Inflammatory cells were also observed under histological examination due to the presence of debris from the porous surface of the implant (Figure 4-20) and are reported as: "In most implants, it appeared that there was some minor erosion of the porous disc surface, but not of the hard surface... Surrounding the fibrous capsule there were irregular clusters of macrophages and foreign body giant cells which contained intracytoplasmic clusters of the same granular crystalline material." 86 - v V Figure 4-20: Crystalline debris is shown adjacent to where the implant was located. Debris is both intracellular and extracellular. Magnification is 100x. Appendix 3 shows a ranking for the amount of particle debris present. The debris can further contribute to the inflammatory process and slow the formation of the fibrous encapsulation [86]. The implant can then be moved by external forces until the fibrous tissue matures into more organized connective tissue that interacts with the surrounding tissue to fixate it. However, histology has shown no differences between implants that have moved or not with similar vascular connective tissue surrounding all implants. Differences would appear in encapsulation thickness as a more severe foreign body response leads to thicker encapsulations [18, 20]. No correlation was found between encapsulation thickness (listed in appendix 3) and migration (p < 0.8 using a Student t-test) or between encapsulation thicknesses of implants that migrated versus implants that did no migrate (p < 0.65 using ANOVA). Figure 4-21 shows the relation for encapsulation thickness and migration. No relation, using a Student t-test, was observed for implants that migrated and their encapsulation thickness (p < 0.1) and for implants that did not migrate and their encapsulation thickness (p < 0.79). Tissue encapsulation thickness varied from 20 to 140 microns and similar sizes of 10 to 120 microns have been reported elsewhere for 87 porous materials [31, 158]. This is thinner than the reported encapsulation thicknesses of 100 to 200 microns for solid alumina indicating a more favorable tissue reaction [31]. The presence of inflammatory cells is observed as related to the amount of crystalline debris in the surrounding tissue as they are not present at the implant surface or in the fibrous capsule. This is mentioned in the histology report "Surrounding the fibrous capsule there were irregular clusters of macrophages and foreign body giant cells which contained intracytoplasmic clusters of the same granular crystalline material." Once again, no correlation (p < 0.18 using a Student t-test) can be observed when plotting implant migration versus the amount of crystalline debris present (Figure 4-22). Therefore, it can be concluded that foreign body reaction does not cause the movement of the implants as no relation exists for the encapsulation thickness and present inflammatory cells with implant movement. 5 E 4 c 3 o - f—J 2 CD 1 0 Migration no Migration SE = 0.68 SE = 0.45 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 Encapsulation Thickness (cm) Figure 4-21: Relation between implant migration and the encapsulation thickness (p < 0.8). The dashed line represents the trend line for the migrated discs while the solid line represents the non-migrated discs. 88 E c o "•4—• 5 4 3 2 1 0 Migrated non-migrated • • • _ SE = 0.99 , — — SE = 0.18 Crystalline Debris (0 - none, 5 - lots of debris) Figure 4-22: Relation between implant migration and the amount of crystalline debris present around implants (p < 0.18). The dashed line represents the trend line for the migrated discs while the solid line represents the non-migrated discs. Analysis of variance (ANOVA) showed no statistical differences (F value < F critical for a=0.05) for the effect of different pore layer thicknesses on encapsulation wall thickness (p < 0.2) or amount of crystalline debris (p < 0.19). The manufacturing procedure of the porous and solid layered alumina disc showed some problems as noted above. The result was that the porous layer appeared to not be completely sintered as the porous surface was powdery. This is almost certainly the crystalline material that was observed in the macrophages and foreign body giant cells present near the implant. Erosion of the surface via biological degradation would be much easier on these porous surfaces. No gross inflammatory response was observed as it has been shown that alumina powder is biocompatible [24]. Surface erosion of alumina materials from implantation has been mentioned by a few others [158-160]. Corrosion pitting is observed where interactions with the biological environments are concentrated on parts of the surface (where defects may be located) that lead to the formation of 89 pits [159]. However, this was observed with alumina femoral heads subjected to wear. The surface of dense alumina after implantation has been observed to have no surface changes [158, 160]. Histological observations of our implants confirm this "In most implants, it appeared that there was some minor erosion of the porous disc surface, but not of the hard surface." No account of erosion of porous ceramics has been reported leading to the conclusion of manufacturing technique causing the weakening of the porous layer and hence the easy removal of material. Externally Applied Forces Externally applied forces such as interactions with other rabbits or the rubbing of the implant area against a wall can lead to potential movement of the alumina discs [86]. Unfortunately we cannot accurately analyze this scenario as constant monitoring of the rabbits was not performed to examine their behaviour and activities to list all possible sources. It was, however, observed on a few separate occasions that rabbits had bite marks on their backs. These bite marks indicate that rabbits fought each other and that the back was a target area (Figure 4-23). During these fights, the implants could have been dislodged as rabbits moved around and attacked one another. Figure 4-23: Bite marks on the backs of rabbits 90 Other sources may also exist as well although the only evidence that is observed is bite marks. These extrinsic forces can be thought as only a factor to implants that are able to move because of the lack of implant fixation. This mechanism can easily occur in combination with lack of tissue attachment or lack of integration of the fibrous encapsulation with the surrounding tissue. An implant that is very loosely attached to its surrounding tissue will be more easily moved around by external forces. The lack of attachment of the encapsulating tissue to the surrounding tissue provides no resistance against any externally applied forces. This lack of attachment is expected when the implant is inserted into the subcutaneous tissue which was observed during tissue attachment force experiments. The placement of the implant in the sliding layer also subjects the implant to muscle movements from the underlying dorsal muscle. The sliding layer moves as the muscle contracts or extends and implants within the sliding layer would also move. Applications to the Steller Sea-lion Housing The results from this experiment showed that implant movement occurred in 50% of the cases when the location of implantation was the subcutaneous fat tissue. Insertion into dermal tissue yielded better integration of the encapsulation tissue with the surrounding tissue (although a limited number of samples were available) however implants still migrated with the growing skin. In the application of implantation into the Steller Sea-lion, the same results might be expected. If possible, implantation should be in the dermal tissue to reduce migration of the implant from the area of incision. The dermal tissue is more tightly bound then the loose subcutaneous tissue and can provide better implant resistance to migration. The subcutaneous tissue is also poorly vascularized which can lead to slower healing and encapsulation times. Implantation into the dermal layer could increase the wound healing rate around the implant and the formation of a connection between the encapsulated tissue and surrounding tissue. The growth of the Steller sea-lion skin will determine what the final position of the implant will be. It should be considered that the final position of the implant may not coincide with the incision location. Before implantation Jnto the back of the head of the Steller Sea-lion the skin growth 91 should be researched to understand how the implant may migrate and to what location. This research is especially important as Steller sea-lions have a tighter interconnection between the dermal layer and subcutaneous tissue compared to the loose connection observed in rabbits. The tighter connection may provide added implant resistance to movement. However, the main conclusions from this experiment showed that implantation of the housing should be in the dermal layer and not the subcutaneous layer. The role of the loosely connected subcutaneous tissue, specific to rabbits and not observed in Steller sea lions, will only need to be considered if the housing is to be placed in the subcutaneous tissue of the sea lions otherwise the results will not differ between the two species. The tissue reaction to the aluminum oxide material in young animals was reported to be minimal in the histology report. Thus the use of alumina in the Steller Sea-lion is safe. The presence of the debris particles can be eliminated via improved manufacturing technique. All observations showed that no statistical significance existed for implants with different pore layer thicknesses. This indicates that the pore layer used in the housing design can be thinner resulting in an overall thinner housing. 92 4.4 Conclusions Results from experiments with implanted alumina discs have lead to several key conclusions about the migration of porous materials and the effect of pore layer on implant fixation. i. Implant discs that were encapsulated in well-vascularized tissue do not lead to fixation into surrounding tissue as previous literature results suggest. ii. Migration occurred in 50% of the discs with two showing gross migration (~7cm) from the incision points. iii. A comparison of discs that had migrated and had not migrated (in terms of pull-out force, encapsulation thickness and tissue in growth) did not show significant differences. iv. The most likely explanation for the migration movement was the weak interconnection between the implant encapsulation tissue and the surrounding subcutaneous fat tissue that can lead to easy movement of the implant from external or internal forces. v. The strongest fixation without migration occurred when the discs were implanted in the dermal layer between the cutaneous trunci and dermis. vi. No adverse tissue reaction was observed for aluminum oxide in growing animals. vii. Implants will move as the tissue they are embedded in or attached to grows. Implants move with the growth of tissue away from the incision location. Care must be taken before implanting into Steller Sea-lions for long term. 93 4.5 Future Work Topics for future work include: i. Investigation of the increase in tissue attachment force for increases in porosity and pore size. Determine the variation in tissue attachment force and see if there is a maximum limit of pore size for maximum tissue attachment force. ii. Comparison of the movement of implants with tissue/skin growth as functions of time. Measure the distance of implant movement relative to location of implantation as the animal grows to see if implants follow the same growth path as the tissue they are embedded in. Determine if porosity or pore size affects this movement and relate results to tissue attachment forces. iii. Measuring the in vivo biological forces exerted on implants and comparing those forces to tissue attachment forces and tissue in growth rates. Measurement of the amount of in grown tissue needed to fix an implant of certain dimensions can be used to derive a model that predicts times for biological fixation for different pore sizes and porosities. This is examined only if an effect of pore size exists. iv. Examine implant migration in different types of tissue to see any difference compared to the loose subcutaneous adipose tissue. This will help determine to what degree the looseness of the tissue plays a role in the migration of implants. Also, compare the tissue growth of subcutaneous tissue with dermal tissue to determine if any relation exists and if dermal tissue growth tracking can adequately model sub-dermal tissue growth. v. Study of the growth of tissue in different animals to determine a model that predicts how implants may move during implantation in growing animals. Correlate this data to research performed on how implants move with tissue growth. 94 CHAPTER V: Mechanical Durability 5.11ntroduction Throughout the lifetime of the Steller sea-lion, several different situations exist where the implant may be subjected to an external force that could break the implant. Research into the durability of the implant to resist external forces needs to be performed to assess the enclosure's reliability. Reliability testing involves ensuring a mechanically robust enclosure that will last the specified life time of the animal. An enclosure that is not mechanically robust will crack or break when subjected to applied loading cases and could cause failure of the internal electronics. Biological harm to the host can also take place as the electronics are corroded and release metallic ions into the adjacent tissue or from the leakage of chemicals from the batteries. The aluminum oxide housing must be durable enough to prevent these and this will require meeting certain criteria. The ultimate goal of product development is to design and manufacture items that meet the needs or requirements of customers. Components have to conform to a certain level of reliability that includes customer requirements as a minimum throughout the intended period of use. Thus, the testing of component reliability is a very important stage when developing a new product with reliability testing being the best method to verify a products functionality and conformity to specifications. Reliability performance testing involves taking the final product and identifying any problems with its use. If problems arise, then corrective actions must be determined to redesign the product to not fail. Tests are undertaken to establish whether or not the product meets the established minimum reliability requirements or failure criteria. In the case of examining mechanical strength, the RF tag housing must be able to endure various loading conditions without cracking or breaking. As is the case with brittle materials such as aluminum oxide, cracking very quickly leads to material failure and thus, cracking and failure indicate the same end. 95 Tests that are used to predict service performance must reflect actual loading conditions and intended service life. For example, hip implants are expected to last 40 years, and endure many types of loadings. Accelerated testing methods have been developed to predict the 40 year service time in 6 months [161]. Cost of implants and test and time commitments restraint the test number and types actually conducted. For wildlife implants, service performance data is generally lacking. For long living and endangered animals such as some sea-lions, service data should be a necessary pre-requirement before tags are implanted into them. This thesis is concerned with the development of service performance tests associated with how well the housing would protect the animal over its 20-30 year lifetime from its potential toxic interior. The loads that the RF tag housing must endure should correspond to the same expected loading conditions when implanted in the Steller Sea-lions (i.e. conditions must be as close to real as possible). These loading conditions consist of (with more details in the methods section): 1) Deep ocean pressures associated with sea-lion diving. Sea-lion diving patterns have been recorded and mean depths and dives have been measured. These measurements form the basis of cyclic fatigue tests with a stress range from very low (when at the surface) to stresses associated with depths of twice the mean recorded depth. The expected type of stress is a mixture of uniaxial and hydrostatic pressure. 2) Puncture forces as a result of Steller Sea-lion bites. The fighting behaviour of Steller Sea-lions has been observed and biting is their main conduit for harming each other. The possibility of the implant being in an area subject to biting has been discussed and although the occurrence may be rare, it is still a potential problem. Therefore, the housing must be able to endure the Steller Sea-lion bite force. 3) Impact dynamic forces from falling or interactions with other Steller Sea-lions. Impact forces comprise very complex loading states coupled with a variety of loading conditions. Impact loading conditions can vary depending on the area of impact and the type of surfaces in contact 96 with the implant upon loading. A sea-lion falling onto smooth blunt surfaces will experience a different impact load then from rough surfaces. Impact loads from interactions with other animals occur via other sea-lions jumping or running over other animals (larger adults running over younger sea-lions). These impact forces need to be considered for the housing. The following performance reliability tests demonstrate the product suitability to resist mechanical loading that will be experienced during implantation in Steller Sea-lions. To obtain this information, tests and models were developed. In particular, cyclic fatigue tests are conducted to model the mechanical response of the implant from pressure forces; puncture fracture loads were recorded and compared; and the impact response of the implant was analyzed. All three of these results are used to describe the suitability of the housing to meet the specified design criteria. Herein, results are discussed along with comments about the added effects of alumina aging in biological environments and the effects of porosity when a thin layer of porosity is included in the top surface. It is hoped that manufacturers of other wildlife implants will follow similar methodology to provide service performance data for other animals. 5.2 Materials and Methods Tests methods are designed to accurately test the mechanical response of the housing to similar in-service conditions. Failure criteria and analysis is derived to assess the suitability of the housing during different loading conditions. Methodology and failure criteria are discussed herein. 5.2.1 Alumina Housing Fabrication The mechanical reliability tests are used to assess the service performance of the enclosure when implanted sub-dermally in Steller Sea-lions. Thus, to prove the reliability of the final design performance tests are used to evaluate the mechanical robustness of the final design. Performance tests are used to observe the operation of the final product. Only the final product was used so that adequate conclusions can be made. 9 7 Alumina housings manufactured to specification must be fabricated to carry out the performance tests. A sufficient number of alumina (99.5% Al 2 0 3 ) housings were fabricated (International Ceramic Engineering, Worcester, MA) for testing the housings under the three different loading cases (Figure 5-1). The housings were identical to the final product with the exception of the porous layer on the top side of the lid (see appendix 1 for detailed dimensions). Porosity was not included in these manufactured enclosures due to high costs of fabrication and because it was expected that the strength of the housing is determined by the solid alumina layer and not the porous. The effect of porosity is discussed in the results section. The interior of the housings were filled with epoxy (EPO-TEK 301, Epoxy Technology, INC., Billerica, MA) to simulate the final product with the absence of the electronics. Figure 5-1: Alumina housings for mechanical reliability tests 5.2.2 Test Apparatus Fabrication To perform mechanical tests on the alumina housings in test frames, apparatus had to be designed to accommodate the enclosures for specific loadings. The test apparatus for the 98 pressure and puncture tests are shown in Figure 5-2. The puncture head area reflects the area of mature Steller Sea-lion teeth (as determined by measurements on Steller Sea-lion skulls). Detailed drawings for all the components are shown in appendix 1. Each component has a threaded end to screw into the load frame. Figure 5-2: Test apparatus. A - Shie ld , B - Pressure plate, C - Hous ing holder, D - Puncture head For all tests the enclosure holder was designed to align the applied loads normal to the top surface. This represents the loading conditions in vivo as the enclosure is tapered to fit the skull of the sea-lion and also assumes the tag is firmly fixated within the tissue. The components were manufactured from 4140 high strength steel. The rigidness of the components leads to very conservative experiments as in practice the housing will be sitting on top of soft tissue and bone (Figure 5-3). Bone tissue is much more compliant than high strength steel and will absorb some of the loadings especially in impact. This dampening of the load is less prevalent in compression and puncture tests as the rate of loading is much slower than impact. A more accurate representation of the impact tests can be achieved by considering the energy absorption of the surrounding tissue. This will be considered in the discussion. 99 Dermal Layer Bone tissue Figure 5-3: Actual in-service layout of implanted housing. The soft t issue surrounding the implant provides some dampening of the applied loads. The performed mechanical experiments do not simulate this, thus the results are conservative estimates. 5.2.3 Test Procedures Fatigue Tests To determine the life of the product under cyclic loading, the housing was subjected to pressure forces expected in practice (from deep ocean diving). The compressive forces used were applied using a uniform displacement as application of a uniform pressure was not possible with the available equipment. However it is expected that the use of uniform displacement is valid because of the small size of the implant and that any applied pressure will act over the entire housing. The applied compressive forces are conservative in nature comparing to the actual case where some pressure is also applied on the sides. The pressure on the sides of the housing in practice will reduce the stresses in the load bearing walls and thus the actual conditions are better than what is represented in the tests. The applied forces were calculated using Steller Sea-lion dive data [162]. In the research, 25 Steller sea-lions were monitored for dive behavior. Depth, duration of dive and frequency of dives was recorded. The mean dive depth was 135.36 meters with a mean dive frequency of approximately 6.3 times a day. The associated pressure with a dive depth of 135.36m is 1.33 MPa and the total number of cycles for a Steller sea-lion living a lifespan of 30 years [163] is 68 985 cycles. Therefore, the failure criteria for the RF tag housing is a mean fatigue pressure of 1.33 MPa for 68 985 cycles. To account for strength degradations from tissue implantation and the inclusion of a porous layer, a higher load was applied. A maximum load of 26 kN and 100 minimum load of 1 kN were selected and housing samples were cycled 70 000 times. The number of specimens used was limited by the costs of manufacturing the housings thus only 10 samples were available for cyclic fatigue tests. From such a small sample size, S-N curves can not be constructed to predict the fatigue life [164]. A minimum of 16 samples is needed to generate an S-N curve for four stress levels to adequately test the reliability of the housing. Instead a binomial statistical test method was used to characterize the acceptability of the housing to the specified test conditions by classifying test trials as successes or failures [164]. A successful trial was when the housing did not fail after 70 000 cycles was reached. The problem of having a small sample size also shows up using the binomial test method as the low sample size will lead to a low reliability for a high confidence level (90%) or a high reliability (90%) with low confidence level. The frequency of cyclic loading was chosen to be 10 Hz. This frequency is different than the real loading conditions of 1 cycle per 66 seconds (0.015Hz) [162], so that the testing can be completed within a reasonable time frame. It has been reported that the effect of cycle frequency is negligible (for metals or ceramics) up to 20 [165], 100 Hz [166] and 170 Hz [167] after which temperature effects start to occur. It has also been reported that effect of frequency occurs for all ranges with experimental data showing such [27]. The frequency effect described in [27] states that lower frequencies shift the stress-life curve towards the static fatigue case and thus the life time is decreased for increasing frequency. Due to the costs of producing more samples for a static fatigue test (at least six more), the dependence or effect of fatigue strength on cycle frequency was not tested. Plotting such results in a stress vs. time to failure graph would, however, show any frequency effects as fatigue strength plots would show variations at different frequencies. with R" '< 1 — C for the case of no failures (5.1) 101 In addition to static and cyclic tests, dynamic fatigue can provide some important information when considering deep ocean pressures. Depths of over 400m have been reported for Steller Sea-lion dives. A dynamic test to determine the maximum compressive forces the housing can withstand is also of interest as the results can show the maximum depth the sea-lion can dive to before the housing ruptures from compression. The applied loading rate is the speed at which a Steller Sea-lion dives and is 6.79 m/s or approximately 70 kPa/s [162]. The number of samples used was 6. Experiments were carried out by using the appropriate test apparatus and loading the housings cyclically and dynamically until failure in air at room temperature. The temperature that the housing will be subjected to will vary as the Steller Sea-lion dives deep into the ocean or rests on a beach however temperature effects such as viscoelastic creep do not occur until very high temperatures (i.e. 1100 °C) [89]. The failure load was recorded as well as the time to failure. If no failure occurred by 70 000 cycles, the experiment was stopped and the housing deemed a non-failure. The experimental setup is shown in Figure 5-4. The test setup for the dynamic fatigue experiments are shown in Figures 5-5 and 5-6 using a Forney FX600 compression machine (Forney Inc.). Results from the dynamic fatigue tests are analyzed by constructing a Weibuli diagram [27, 168] and the failure mode is recorded to know how the housing may fail in service. The failure probability, P, is related to the strength measurements using the function: P = 1 - exp (5.2) 102 Figure 5-4: Cyclic fatigue test setup. Where m w is the Weibull modulus, s is the strength and s 0 is a reference strength. The reference strength represents the stress level where the probability of failure is 63.3% for a stress level equal to or less than the reference strength and is used for normalizing data. The probability function is plotted, using the strength measurements, by taking the double-logarithm of both sides such that: The graph allows the determination of the Weibull modulus (slope) and the reference strength (y-intercept). The y-axis is typically plotted with a scale of -4 to 1 with 0 representing the probability (63.3%) when the reference strength and failure strength are equal. In In l-P 1 (5.3) 103 Figure 5-5: Dynamic Fatigue Setup Figure 5-6: Dynamic Fatigue Setup. The alumina housing is shown in compression. 104 Puncture tests The puncture tests were used to assess the strength of the housing under bite loads from Steller Sea-lions. The housing was placed in a load frame and a vertical force applied through an indenter. The biting force that is applied by Steller Sea-lions has not been recorded in the literature. Results for other animals are summarized in Table 5-1. Table 5-1: Bite force for various species Species Recorded Force (N) Reference Dusky Shark -590 [169] Humans 440-670 [170] -800 [171],[172] Horn Shark 338 [173] African Lion -4200 [174] Alligator -9500 [174] Dogs 800 [175] Although no value for bite force for Steller Sea-lions is known, it could be guessed that the bite force lies somewhere near that of an African lion. What should be noted about the above recorded forces is that the types of teeth used to generate those forces are not mentioned. Flat or blunt teeth can exert much higher forces than pointed teeth since high stresses accumulate in pointed teeth at high loads which could lead to tooth fracture. This is described in [172] as the ratio offerees is 4:2:1 for molars, premolars and incisors. The highest stresses occur when biting on hard surfaces. Consequently, the maximum force for African lions may be 4000 N but the maximum force for incisors or teeth with the function of tearing is approximately 1000 N. All the teeth in a Steller sea-lion are pointed and thus they are not meant for biting down on hard surfaces. Thus, the failure load or stress of Steller sea-lion teeth should also be considered. Unfortunately, this is also not known but values for humans have been recorded that are of the magnitudes 888 N [175] and 611-1807 N [176]. The stress field in the tooth will be related to the applied load, geometry of the tooth and the area of contact. However, it could be estimated, using the results for human teeth that the failure load could be as high as twice the maximum applied load. 105 Therefore, two sets of criteria can be set to ensure mechanical reliability under bite puncture force. The first is that the enclosure must endure the maximum applied load that a Steller sea-lion can exert which we are assuming to be approximately 1000 N. The second is supplementary to the first criteria and is that the implant must endure the load necessary to fracture a sea-lion's tooth. The first criteria is the most important to pass however, the second will clearly show that the housing can survive any possible animal bite. Tests were conducted in air at room temperature. The variation in temperature that the housing could experience (approximately ±5-10 °C) that would contribute to the mechanical response is negligible considering the good resistance of ceramics to thermal stress [90]. The loading rate used was that of the bite force rate for a human which has been recorded to be around 300 N/s [177]. The number of samples used was 6. Loads were applied until failure of the housing. The failure load was recorded as a load vs. time curve. The experimental setup is shown in Figure 5-7. A load frame (Instron, Inc.) was used to apply the puncture loads. A Weibuli probability diagram was constructed to analyze the results. 106 Figure 5-7: Puncture test setup Impact Forces Impact forces occur when a shock force is applied to the housing. Shock forces are high loading rates that are applied for a short duration. These can occur in two different situations during the lifetime of a Steller sea-lion. The first is from the environment. The habitat of Steller sea-lions are haul outs and rookeries. These rocky islands consist of many small cliffs that sea-lions may fall off of. Even a drop of a few meters could result in large impact forces being applied to the implant if the housing is in the area of contact. The second situation involves interactions with other sea-lions. Fights between sea-lions occur and it is possible that a sea-lion may be pushed over onto a rocky or hard surface that will impact the housing. Or it is also possible that a sea-lion will step or run over another sea-lion's head. Thus, there are several different impact loading cases however only the worst case will be examined. Unfortunately, only estimates can be made regarding the severity of each potential situation. Simple estimates can be made using the impulse equation and energy balances. 107 These two equations can be used to determine an approximate impact force. The impulse equation is derived from Newton's second law of motion: F = mAv (5.4) The force is related to the change in velocity, v, of a mass, m, over a period of time t. The mass is the mass of a Steller sea-lion, the time is the duration of contact and the velocity is calculated using an energy balance relating potential and kinetic energy. The energy balance yields an equation for velocity: Where g is the gravitational constant and h is the height of the fall. From these two equations, simple cases can be put together to determine an appropriate failure criteria for the housing. The first case to look at is the case of a sea-lion falling several meters onto a rocky surface. Assuming a fall of 5 meters, sea-lion mass of 680 kg [163] and a contact time of 0.5 seconds, the calculated impact force is approximately 13.5 kN. The impact force can be deceiving as it is the stress or added energy put into the material that causes fracture. However, most research groups report impact force and thus it can be used 'loosely' for comparison. The energy associated with the above situation is 33.3 kJ. The calculated energy is an extremely conservative value as it assumes the sea-lion falls directly on the implant with all related energy being used converted to strain energy in the housing. In actuality, the body of the sea-lion would absorb most of the kinetic energy and is analogous to research conducted on ballistic impacts for alumina with various different backings. The alumina is used for its high strength and a more ductile backing is used to absorb the kinetic energy (which can exceed a kJ) [100-103]. (5.5) 108 The next case is that of a Steller sea-lion running over the head, and thus the housing, of another sea-lion, particularly young sea-lion pups. Conditions are estimated as in the previous case with a height of 25cm (corresponding to the height of a sea-lion flipper), a sea-lion mass of 170 kg (assuming each flipper supports a quarter of the sea-lion mass) and a contact time of 0.5 seconds. The resulting force and energy are 753 N and 417J. The final case is the impact of a sea-lion falling over. Conditions for this situation are estimated as a height of 1m (corresponding to the height of a sea-lion), sea-lion head mass of 10 kg and a similar contact time of 0.5 seconds. The calculated force and stress are 88.6 N and 98J. In each case, the most important parameter is the resulting energy which is a function of the velocity at time of impact or initial height. The worst case scenario comes from a Steller sea-lion falling from a rookery from a height of 5m. However, this is only an estimate of a possible scenario thus instead of specifying a pass/fail criteria for impact testing, tests shall first be conducted to determine the fracture characteristics of the housing. These results will be applied to the above mentioned cases to calculate conditions for which the housing will not fail. The results will also be compared to the fracture strength of skulls. The implant may fail from a specific loading case but if the applied loading also surpasses the skull fracture strength, then implant failure may not be such a concern. Typical values for skull fracture in humans are shown in Table 5-2 as no results for Steller sea-lions have been recorded. Table 5-2: Reported values for Human Skull Fracture, Reported values for skull fracture References 73N - 1020N (deDendina on situation) H 781 33 ft-lbs of energy (~ 44J) [179] 7 3 - 8 7 3 N [180] 4.43 m/s (min. velocity) or 27.6J [181] 82.7 MPa [182] 1000 Nor16.6J [183] Tests were conducted in air at room temperature using a drop-weight impact tester. The number of samples used was 11 and the height at failure was recorded for each sample. The test 109 apparatus was constructed to be similar to specifications mentioned in ASTM standards [184, 185]. The test apparatus is shown in Figure 5-8. The weight dropped onto the samples has a mass of 4.89 kg. As in the compression tests, a uniform displacement force is applied to the housing instead of pressures as an impact is expected to strike the entire housing at once instead of just one section. Reports from the literature [88, 104] report an impact energy of 14 to 22J for alumina used in hip implants. Others have reported the impact energy for alumina to be 1.96J [186], 1 -3 kJ/m 2 [187] (which is equivalent to 1.74 - 5.2J for our geometry) and 5.9J [12]. A starting point that corresponds to energies between 1 and 22J is used as the fracture energy for the alumina housing will most likely fall within this range. A starting height corresponding to approximately 6J was chosen arbitrarily. The experimental procedure followed the Bruceton Staircase Method or the Up-and-Down method [184]. The first specimen was tested at a height corresponding to 6J. If the specimen did not fracture, the height of the dropped weight is increased by 10mm otherwise the height is reduced by 10mm. This is continued until all specimens have been tested. Failure and non-failures are recorded and the mean failure height and energy are calculated as in [184]. 110 Figure 5-8: Drop-weight apparatus. 5.3 Results and Discussion Results from the mechanical experiments show that the durability of the aluminum oxide housing is sufficient for cases of cyclic fatigue, high compressive forces and puncture. Cyclic fatigue results showed no housing failure after 70 000 cycles for stresses of approximately 15 MPa; dynamic fatigue results show failure of the housing at a minimum stress level of 236 MPa; and puncture forces at failure are at a minimum level of 15 kN. The response of the enclosure to impact was failure at an energy level of 6.7 J . This conservative result coupled with the energy 111 absorption of the surrounding tissue leads to the possibility that the housing is sufficient under impact loading. More detailed results are discussed in the following sections along with remarks about the potential effects of environmental ageing and the addition of the porous layer. Recommendations for improvement are also made where applicable for re-design. 5.3.1 Fatigue Results Dynamic compression force tests were run on samples until failure. The results are shown in Table 5-3 and as a Weibuli diagram shown in Figure 5-9. The characteristic strength or normalizing strength was calculated to 570.6 MPa and a plot of the probability of failure of the housing as a function of applied stress is shown in Figure 5-10. The large variation in results was expected and is a result of the scatter in the inherent flaw sizes of ceramics [168]. Processing, machining and finishing of ceramics can give rise to defects of various sizes which leads to a large scatter in results. Plotting the probability of failure of the material takes this variation in processing into consideration (Figure 5-10). Table 5-3: Recorded forces and stresses from compression tests on six specimens. Specimen # Force to Failure (kN) Stress (MPa) 1 840.1 482.8 2 1060.5 609.5 3 538.5 309.5 4 1117.6 642.3 5 925.1 531.7 6 837.1 481.1 112 Figure 5-10: Probability of failure of the housing versus the applied stress. Applied stresses that are 40% of the characteristic strength will give <5% probability of failure. Results from Figure 5-10 show a low probability of failure for applied stresses that are two orders of magnitude higher than what is expected during the actual application. Sea-lion's have been observed to reach depths as low as 400m however the resulting stress from those depths is only approximately 4 MPa. A 95% reliability of success of the housing corresponds to a strength of 113 236 MPa which is 59 times stronger than is required. This strength gives rise to failure at depths of 24000m; depths that exceed the maximum capabilities of sea-lions. High strength results of the housing were expected as aluminum oxide has been proven to have very high compressive strengths. Failure below the theoretical values of 2.5-3 GPa occurs due to the general variability of inherent flaws contained in each fabricated specimen and from the geometry of the sample. Under uniform pressure, the presence of epoxy helps to reduce the stresses in the top plate. Without epoxy, the cover will deflect in the middle putting a tensile stress on the bottom side of the plate. This can be extremely important as the strength of ceramics is lower in tension and premature failure of the housing can occur originating from the area under tension. However, in our assumption of uniform displacement the epoxy does not play a role as the load is supported by the walls of the housing. Therefore, with most of the stresses in the walls of the housing that is the first place that one would expect to see failure or fracture. Samples from the compression tests verify this as cracks formed in the walls and failure subsequently occurred. All six samples showed failure in the walls and are a result of shear forces from the uniaxial compression. Figure 5-11 shows an example of one of the failures and pictures of all six samples are shown in appendix 5. Fracture patterns for all six specimens appear to be similar with cracks originating in the interior of the housing wall and propagating outwards. The many fragments are a result of crack branching as the crack reaches a critical size or velocity [188, 190]. . , 114 Figure 5-11: Failure of a sample housing under compression. Fracture occurred in one of the side walls and correlates well with what is predicted in Finite Element models. Although the housing has surpassed the required criteria, the effects of having a porous layer and of the implant in a biological environment need to be considered. Porosity in a material has been well studied however only a few have examined the strength of a multi-layered material with a dense layer and porous layer [189-191]. Strength of the material decreases with the introduction of a porous layer. The porous structure lowers the stiffness of the layer and provides locations for cracks to form [190]. Cracks from neighboring pores can coalesce to form a larger crack that can reach critical levels. One group [190] ran numerical simulations and showed that an inclusion of 7.5% porosity can reduce the strength of the material by 28% (this correlates well with predictions of strength decreases when introducing porosity into the entire bulk of a material). Two other groups showed a two-thirds [192] and 75% [193] drop in strength of a porous coated titanium material. 115 Although, the strength of the material is lowered, the layers can also create a barrier to crack propagation and thus toughness may be slightly increased from the separate layers [191]. Therefore, the critical crack length may be larger as more energy is required for cracks to propagate from the porous layer to the solid layer and vice versa. Using the Sprigg relation for alumina with a porosity and strength, the introduction of a 35% porosity yields a strength reduction of 70%. At a failure probability of 5%, our failure strength is predicted to be approximately 236 MPa. Adjusting this value for the introduction of a porous layer gives a new strength of 71 MPa. The housing strength and related calculations should use this value as it may more accurately represent the final product. However, one problem is associated with this assumption. The failure mode seen in the compressive strength tests clearly show a failure in the walls of the housings as a result of stress applied from the plate and transferred from the epoxy (thus giving rise to stress levels above those applied). The above calculations use the assumption that now failure will occur from the porous layer on top of the housing cover plate as cracks form from the pores. The interactions of stress fields from neighboring pores create very complex stress states which can be difficult to predict and model. Therefore, the actual failure stress may be different than the above calculation. Compression tests should be conducted with the porous layer to verify the analysis. Environmental effects on material strength are important in static fatigue as water can lead to sub-critical crack growth and failure below the fracture strength. Although this may not be directly related to dynamic tests, during the life of the housings they will be subjected to this form of corrosion and it is acceptable to expect the presence of larger cracks to exist in the material as time goes by. These cracks can easily propagate under applied stresses and cause premature failure. Failures at stress levels that are 30-40% lower than the fracture strength have been reported. Applied to the alumina housing, the strength to failure can be reduced from 71 MPa to 28 MPa after long term implantation. This final adjusted value for porosity and environmental effects is still 7 times that of the required 4 MPa corresponding to a sea-lion dive depth of 400m. 116 Cyclic fatigue tests were run until failure or 70 000 cycles was reached. The high compressive strength, determined from the dynamic tests indicates that the housing will also have relatively high fatigue strength. Results from the cyclic fatigue tests are shown in Table 5-4. Table 5-4: Pass/Fail results from cyclic fatigue tests run until 70 000 cycles. Sample # Pass/Fail? 1 Pass 2 Pass 3 Pass 4 Pass 5 Pass 6 Pass 7 Pass 8 Pass 9 Pass 10 Pass All samples passed the applied loading conditions of 0.57 MPa to 14.94 MPa for 70 000 cycles. Using the Binomial statistical test method for no failures, these results correspond to a reliability of 90% with a 65% confidence level or a confidence level of 90% with a 79% reliability for 10 samples. The distinction between the reliability and the confidence level is that the reliability is the probability that the housing will function at the specified loading conditions and the confidence level is the indication that the results from the experiment are the true results. In using the Binomial test method, a high reliability and confidence level means that it is most probable the component will pass the test whenever subjected to the same loading condition. With respect to the observed results, we can either say that: 1) There is a high probability that the component will pass similar loading conditions however the result may not be completely indicative of the true response or 2) The result is the true response but we do not expect the housing to have a high probability of surviving similar loading conditions. 117 The first case shall be used as we can prove, using results from the literature, that the fatigue strength of alumina far surpasses the testing conditions. Evidence for this would give good indications that we can be confident about the results. Several researchers have tested the cyclic fatigue strength of alumina and have calculated the fatigue limit for the material [194-196]. Results can vary between experiments as material strength can vary according to different geometries and material processing (of which can generate different flaws ranging from the inclusion of impurities or micro-cracks from residual thermal stresses) [197]. Fatigue strength has been reported as 86 MPa [194], 130 ± 10 MPa [195] and 120 MPa [27, 196]. These calculated fatigue strengths are much higher than the 15 MPa used for our experiments. Therefore it is reasonable to be confident that the results obtained are accurate and precise and that the housing will be able to withstand cyclic stresses up to 15 MPa for 70 000 cycles. The reduction in strength of a porous layer is similar to the above calculations. A 70% reduction in strength would mean a new fatigue strength of 4.5 MPa. This is still above the mean expected pressure of 1.33 MPa from deep ocean diving of Steller sea-lions. This fatigue strength is only based off the experimental results and results from the literature indicate that the actual fatigue strength is several times larger. Constraints of machine capabilities restricted the use of larger stresses however the levels tested can still confirm the minimum capabilities of the housing's performance under cyclic fatigue. Thus, although experimentally we do not have results that show the much higher fatigue capabilities of the housing, we can still make the statement that the results obtained are very conservative as compared to the literature. And so the fatigue strength of a porous layer may be calculated, in our instance, to be only three times the expected service conditions, in actuality it is much more. The effect of the environment is that of sub-critical crack growth as micro cracks slowly grow at stress levels below the fracture strength. Reductions of our conservative fatigue limit of 4.5 MPa 118 yield a fatigue strength of 1.8 MPa which takes into consideration both porosity and corrosion from a biological environment. This calculated value still exceeds the expected service loads of 1.33 MPa. The safety factor is only 1.35 however the conservative nature of our experiment must still be considered. 5.3.2 Puncture Results Load-time diagrams were recorded for each housing test. Results for the six tests are shown in Figures 5-12 and 5-13. The bite of an animal can be considered a type of Hertzian contact load (as the tooth isrounded on the end) arid thus, it is expected to show similar types of cone cracking failure. The load-time diagrams for the puncture experiments show several important features related to the formation and development of cone cracking. The first characteristic is the point at which initial cracks begin to form. On the load-time plot, this corresponds to the drop-off of force that occurs around 7 seconds. Force is applied to the housing until a crack forms which reduces the resistance to the indenter. As a result, the force applied to the specimen decreases. The initial cracking can comprise the formation of the ring around the indenter or any sub-surface cracks however catastrophic failure of the housing does not occur. Sub-surface cracks are characteristic of the fracture response of aluminas to contact forces. These are formed by a shear process that leads to the formation of micro-cracks intragranularly and originate from grains that have shear faults or defects [27, 198]. The cracks then grow outwards to weak grain boundaries and a stress intensity is applied along the interface. Once a critical stress is attained, failure along the grain boundaries occurs. 119 0 10 20 30 40 Time (seconds) Figure 5-12: Load-time diagram for Puncture Experiments for all 6 tests. 1 _ 0.5 6T o i & -0.51 I -1 -1.5 -2 19 s = 65.7 MPA ln(s) 18.3 Figure 5-13: Weibuli diagram for puncture stresses Further load is applied and any additional cracking appears as more drop offs in the load-time plots. This continues until total fracture of the specimen which is marked by the final drop off in 120 force. A typical specimen fracture is shown in Figure 5-14. Pictures of all specimens are shown in the appendix 5. The importance of the epoxy filler is clearly shown in the case of puncture forces. With the case that no epoxy is used to fill the inner cavity, the top plate acts similarly to a beam under three point loading. The puncture force causes a deflection in the top plate which creates a tensile stress on the bottom side of the plate. Due to the low tensile strength of alumina, failure will occur at low stress levels. The presence of the epoxy filler provides stiffness to prevent the bending of the top plate. This improves the resistance of the housing to puncture or any other localized load. Figure 5-14: Top surface fracture of a specimen. A ring crack can be seen around the area of puncture with lateral cracks propagating outwards. 121 As can be seen from Figure 5-14, a ring crack is visible around the puncture location. Lateral cracks can be seen propagating from the ring crack outwards. Cracks on the bottom of the housing indicate cone cracking that penetrated the depth of the housing. Variations in load-time plots of each housing are due to the degree of initial cracking. The loading rates were the same for each housing and this is confirmed by the slopes of each loading curve. The first sample has a different loading plot as the applied loading rate was faster than intended and the experiment was halted after the first cracking was heard. The maximum load before total fracture was lower for samples that experienced more cracking during that initial stage. The Weibuli plot uses the stress associated with the first stage of cracking as any sign of cracking in the housing is considered a failed specimen. The initial cracking force for each housing was very similar while the maximum force varied much more. The normalizing strength is calculated as 76.5 MPa. Table 5-5 shows a summary of peak force results from the experiment. Table 5-5: Peak force data for initiation of the first crack and for the maximum load at total fracture. Sample # Load at First Crack (N) Maximum Load (N) 1 2006.06 N/A 2 2065.15 . 6265.3 3 2118.86 6326.17 4 1686.62 6852.53 5 1966.93 5811.06 6 1792.5 5355.8 For a reliability of 95%, the applied force to initiate cracks is 1515.6 N. Comparing this to the expected applied force for Steller sea-lion's using the African lion as a model; the housing will fail at approximately 1.5 times the assumed bite force of 1000N. Failure loads for teeth for humans have varied greatly and observations as high as 1800N have been made. At a load of 1800N, the probability of failure of the housing is approximately 26% meaning that most likely an animal biting with that force will fracture its tooth before the housing cracks. 122 The load causing failure correlates well with what has been observed for alumina under indentation loading as 2000 N has been seen as initiating cracking by another research group [198]. Porous ceramics has also been examined under contact damage [198, 200]. A porous silicon nitride was tested and results showed damage at forces as low as 250 N [200]. Porous alumina has also been tested and showed higher resistance to contact force than silicon nitride. Loads of 1000 and 1500 N were applied to porous blocks. At 1000 N, very little surface damage could be seen and no sub-surface damage was observed. However at 1500 N both surface and sub-surface damage was seen as multiple cracks were visible [199]. Introducing a porous layer to the housing will reduce the resistance of the housing to puncture forces. Although the solid layer will not break, the porous layer will undergo some cracking or fracture. Cracks may propagate into the solid layer causing failure from other loading sources such as impact or pressure. Fracture of the porous layer could lead to alumina pieces being distributed into the surrounding tissue which can lead to local inflammation. The results from [199] show that the failure of porous alumina at loads of 1000 N did not take place however no statistical analysis was performed to determine the probability that failure will not occur at this level. The Sprigg equation used previously showed that the alumina porous layer is 70% weaker than the solid alumina layer. Using that relation in this situation would lead to a force of only around 500 N for a 95% reliability, thus it is possible that the porous layer will fail from a sea-lion bite. There are several possible methods that researchers are currently examining to improve the strength of porous materials which can be achieved through different processing techniques [201-203]. The first fabrication technique uses a method called pulse electric current sintering [201]. This technique is a pressurized sintering method that has been shown to improve the densification of alumina by increasing interactions between particles during the formation process. The denser microstructure leads to a higher strength of the solid volume in the porous material. Hot isostatic pressing (HIP) has also been shown to increase the strength of porous 123 materials in a similar manner as it is also a pressurized sintering technique [202]. Flexure strength was tested and observed to be twice that of porous alumina fabricated by other methods at 30% porosity. An overall strength decrease of 20% was observed which is a significant improvement than what has been observed in the past (i.e. 50-60% decreases in strength). This strength increase was also achieved by the addition of second phase particles which has also been described by another research group as a method to increase porous ceramic strength [203]. The addition of A l 2 0 3 nanoparticles to the powder before sintering was examined to determine any strength increase. The presence of nanoparticles can increase the contact between particles during the sintering phase and thus leading to a better connection between neighboring particles. No explanation was given as to why this occurs however a stronger bond between particles leads to a higher strength. Experiments showed fracture toughness increased from 3 to 5 MPa m 1 ' 2 and flexural strength increased from 175 to 240 MPa [203]. Environmental factors mostly do not affect the outcome of an animal bite. Cracks initiate sub-surface and are not in contact with the external environment. Cracks that have formed from contact loads and moved to the surface can be subject to sub-critical crack growth. And cracks that have formed at the surface and propagated downwards can grow unstably if then subjected to a contact load. However, no direct relation is likely to occur. 5.3.3 Impact Results Drop weight impact tests were conducted on eleven specimens and results are summarized in Table 5-6 similarly to [184], • Table 5-6: Drop-weight Impact Test Results Drop Test Number (x = failure, o = pass) Height, cm 1 2 3 4 5 6 7 8 9 10 11 17 x 16 x o 15 o o 14 o 13 x o 12 x o 11 o 124 The mean failure height (using a 4.89 kg mass) was calculated to be 14 ± 6.9 cm with associated mean failure energy of 6.7 J . This shows similar results to what has been reported elsewhere for alumina [12, 88, 104, 186, 187]. More samples are needed to be tested to achieve a smaller variance and more precise mean failure height. Unfortunately, the low failure energy would indicate that the housing could only survive falls from a few centimeters when implanted in the head of a sea-lion. Comparing the mean failure energy of 6.7 J to the cases presented in the methods section (33.3 kJ, 417 J and 98 J) show that the housing would fracture in all three of the cases. The more limiting scenario is the potential for fracture of the sea-lion's skull. Fracture energy for human skulls are listed as also being quite low (~20-40 J) and a sea-lion skull subjected to loads described in the cases present in the methods section would most likely fracture and possibly kill the animal. Thus, the real limiting factor is the sea-lion skull strength. The actual fracture strength of the sea-lion skull is not known however several reports have been made about the fracture strength of human skulls [178-183]. These reports are observations based on information from cadavers and from reported hospital cases. This indicates that information is gathered from results on whole bodies and that reported fracture energies may be higher than the actual skull fracture energy. During a fall, the skull is loaded as it impacts a surface however the entire body absorbs some of the impact energy. Thus, the actual impact fracture energy of the skull is probably lower than reported. Steller sea-lion and human skulls are quite different. General morphology is easily observed and thus the response to impact should be different. The thickness of the skull, layer of skin surrounding the skull and subcutaneous fat and tissue beneath the skin can all affect the fracture properties of the skull. Similarly, an implanted housing will also yield different fracture properties as it is placed on a softer substrate that can help reduce the force of impact. Tissue has been predicted to reduce impact force by 15% [204]. 125 A more representative case of the housing while implanted and subjected to impact loading can be expressed as an impacted mass supported on springs (Figure 5-15). The springs represent the underlying tissue and reduce the overall energy of the impact through deformation and deflection. An energy balance shows that the overall system energy, in Figure 5-15, is equal to the kinetic energy of the impacting mass minus the energy absorbed by the springs (eq. 5.6). The spring stiffness for soft tissue can be calculated using the Young's modulus of the material. Energy Balance = E = 1/ 2 mv2 - Vi kx2 (5.6) Spring stiffness = k = (AE) / x (5.7) Where m = mass of impactor, v = the velocity of the impactor, k = spring stiffness, x = spring deflection, A = area of impact and E = Young's modulus of bone. Thus, the actual mean failure energy of the housing is the failure energy of 6.7J plus the amount of energy dissipated by the underlying tissue. Estimates using equation 5.6 and assuming only bone is beneath the housing (and that bone deflects 1mm) yields a potential of over 500J that can be absorbed. The spring stiffness was calculated using the Young's modulus of bone which is 18 GPa [5]. This indicates that the housing can be sufficiently strong to resist possible in-service .impact loads. To show that the housing can survive larger impacts with more absorbent underlying materials, four impact tests were conducted using expanded polystyrene (EPS) beneath the housings. Results showed fracture at heights of 60-70cm. Fracture at these heights was expected as EPS can absorb approximately 30J of energy [205]. It is, therefore, reasonable to assume that when the housing is implanted it will tolerate higher impact loads. 126 Impacting mass Spring stiffness, k '7 7 / / / / / / / 7 7 Figure 5-15: Simplified representation of the housing under impact when implanted. The underlying bone and soft tissue act as springs that can absorb some of the energy from impact. The amount absorbed is related to the stiffness of the materials. The model in Figure 5-15 can also be used to represent the case with steel below the ceramic housing. In this case, the stiffness of the steel is so high that no energy is absorbed by the holder as the deformation of the steel is almost zero. All of the energy is put into the ceramic housing and the alumina housing fractures. The brittle nature of ceramics causes the ease of fracture under impact [187]. Ceramics can resist high stress levels but they have low elastic energies that can easily be exceeded when placed under impact leading to failure. The failure of alumina from impact also occurs differently than for sources of lower strain rates as inertia effects must be considered [206, 207]. The poor energy absorption of brittle materials can be seen in Figure 5-16 where the stress-strain curves for brittle and ductile materials are displayed. 127 (a) e 'tens \r tb) * Figure 5-16: Stress-strain curves for ductile (a) and brittle materials (b). The amount of energy a material can absorb before fracture is related to the area under the curves [206]. The impact strength of the material can thus be increased by improving the ductility of the ceramic. Currently, one method exists to improve the ductility of ceramics without compromising their strength. This involves fabricating alumina using nano-sized alumina particles [207-210]. It has been shown that ceramics that have nanostructures can undergo some plastic deformation before fracture. This is on account of the increased amount of volume occupied by grain boundaries. Grain boundaries are the interfaces between grains and have amorphous structures where the atoms are more loosely bonded than in grains. The loose bonding leads to the possibility of elastic or plastic deformation under loading. However, grain boundaries are only a few nanometers in width and occupy a very small space in coarse or micron sized grain structures. For nanostructured materials, the grains are only 10-20 nm in size and the volume of grain boundaries increases leading to an overall increase in ductility of the material. As a result of this ductility, the strength may decrease however the toughness of the material would increase as the stress-strain curve develops into a similar shape as seen in Figure 5-16a. Impact strength will increase as the material is able to absorb more energy before fracture. Under quasi-static loading, crack growth and failure is mostly determined by the critical crack size however under dynamic loading, the velocity of the crack governs the failure [206]. The fast 128 moving crack can propagate via a transgranular mode (as well as intergranular) that provides less resistance for growth. Intergranular crack growth, although requires less energy for growth as impurities and small pores congregate at grain boundaries, provides more mechanisms for crack energy dissipation by way of crack deflection. Thus, the crack loses less kinetic energy and can continue to expand until failure. It has also been shown that dynamic crack arrest does not occur in ceramics as it does in metals or polymers and so once the crack has reached the critical dynamic stress intensity level, it will not stop and the material will fracture [206]. Of the 11 housings tested, only 4 fractured and fracture was localized (i.e. catastrophic failure of the housing did not occur). Figure 5-17 shows the fracture of one of the four housings and pictures of all the housings that fractured are shown in appendix 5. In all the four fractures, only chipping was observed as no discernable cracks could be seen and correlates with similar observations for impact failure reported in literature [211]. The impact causes loading that produces subsurface lateral cracks from stress wave propagation. These lateral cracks are responsible for the chipping mechanism. Although not visible, other subsurface cracks may have formed beneath the surface that have not reached critical and thus, even the samples that do not show fracture may still fracture after repeated impact loads. The effects of the environment also have to be considered. Unfortunately, no research has been performed on the impact strength of materials subjected to corrosive environments. As the failure mechanism is different for dynamic loads due to kinetic effects, it is impossible to predict the behavior of the material based on results for static cases. However, as the mechanisms are separate (i.e. slow crack growth vs. high strain rate crack propagation), it can be intuitively predicted that there is no effect of the environment on the impact strength of the material. 129 Figure 5-17: Fracture of housing #8. The presence of a porous layer also has to be considered. This has been investigated by one group that examined the dynamic fracture behavior of layered alumina [212, 213]. Results from their tests show an increase in fracture energy as porosity increases and is explained as the outcome of increased locations for crack deflection. Under impact, the stress-strain curve resembles that of a more ductile material. As with all porous materials, the strength is decreased however the shape of the stress-strain curves show that more energy can be absorbed before failure. The shape of the curve is most likely a result of the mode of failure that has the porous layer being crushed and absorbing the kinetic energy as the impact load first occurs. The force of impact is also reduced as the time that the weight works on the ceramic is increased. The longer contact time reduces the impact force. This corresponds to the peak of the stress-strain curves however material failure continues as the cracks propagate into the solid layer leading to total fracture. This behavior causes the material to be more impact resistant as more energy can be absorbed before failure. 130 It is also observed that the stress-strain curves differ as impact velocity changes however explanation is given for the cause of this. 131 5.4 Conclusions Mechanical reliability experiments were conducted to assess the resistance of the housing to in-service loading conditions. The following conclusions were made based on the results from those experiments. i. Compression, cyclic fatigue and puncture tests all showed results that passed the set criteria. Impact tests showed failure during conservative experiments but consideration of the surrounding tissue leads to the housing passing the set criteria. ii. Compression tests were conducted to simulate deep ocean diving and experiments showed the housing possesses high strength with a strength of approximately 236 MPa for a 5% probability of failure. Failure was observed in the walls of the housing. iii. Cyclic fatigue tests were conducted to simulate repeated ocean diving and experiments showed no failure at peak loads of 15 MPa for 70 000 cycles with a 90% reliability. iv. Puncture tests, to simulate sea-lion biting, showed good resistance to failure with fracture occurring at 15 kN for a 5% probability of failure. Failure occurred similarly to what is expected from Hertzian contact loads with ring cracks forming around the indenter. v. Impact experiments, to simulate sea-lions falling or from being stepped on, showed fracture occurred at a mean energy level of 6.7 J and failure was in the form of chipping near edges or corners. Calculations that consider the energy absorption of the underlying tissue show that failure of the housing will occur at levels greater than 500J which may be sufficient under many impact conditions. vi. In all cases except impact loading, porosity has a negative effect as it is a source of crack initiation. In impact, the weak porous structure can deform to absorb some of the impact energy improving the impact resistance of the housing. 132 vii. The environment can also have a negative impact on the performance of the housing especially when impurities are present in the material. Higher purity aluminas (>99.9%) should be used to avoid this. 133 5.5 Future Work Topics for future work include: i. Measuring the bite force of Steller sea-lions for more accurate results. Also, modeling the response of the Steller sea-lion tooth to biting on an alumina substrate to determine the tooth stresses. This information can be used to see if the implant fails before the tooth does. ii. Studying the frequency effect on cyclic fatigue results. There have been mixed results showing both no effect and some effect. To examine how the housing reacts to variations in frequency would be of interest as the actual loading rate is much slower. Performing a study on the frequency effect from 0 to 10 Hz (0 being the static fatigue state) will give a more accurate lifetime prediction. iii. Performing experiments in vivo or in vitro and with a porous layer. These two missing components in the mechanical reliability tests can drastically affect the results. Estimates can be made using results from others in the literature, but a more accurate result could be obtained by examining the strength degradation of RF tag housings after a certain amount of time in vivo. iv. Performing mechanical experiments on housings with the electronics to' investigate any differences in the operations of the antenna when under any of the applied loadings. Although the housing has high failure limits, it is possible that the inner electronics may fail at lower conditions. These should be examined as the performance of the radio-frequency tag may degrade before specified. v. Performing non-destructive evaluation of sample housings that do not fail to see if any damage on the inside has occurred. This is most likely or possible under impact loading as stress waves move through the housing. Damage on the inside would mean potential harm to the inner electronics. 134 vi. Investigating the possibility of strengthening the porous layer through different processing techniques. Doing so may increase the reliability of the material under puncture and cyclic fatigue loading. vii. Researching methods to increase the ductility of the material to improve impact resistance. The use of nanoparticles to produce a grain structure of nanometer size is one method. The strength of nanostructure alumina under impact loading should be tested. viii. Modeling of the response of the housing using finite element programs. Simulation of different loading conditions can help better represent the expected in-service loading conditions and the effect of the porous layer. 135 C H A P T E R VI: Overall Conclus ions and Recommendations Research has been conducted to assess the biomechanical reliability of an alumina housing for long term sub-dermal implantation of a radio frequency transmitter in a Steller sea-lion. Biological and mechanical attributes of the housing were tested and results have been compiled in previous chapters. Biological reliability of the implant refers to assessment of the housings ability to be compatible with a soft tissue environment and fixation at the location of implantation. Histological observations have shown good tissue compatibility of aluminum oxide with the presence of a thin fibrous connective tissue capsule surrounding implants. The growth of animals leads to skin growth. Implants move along with the tissue as it grows and can migrate even further away from the location of implantation. Similar attributes of migrated and non-migrated discs indicated that externally applied forces to the objects contained in loosely bound tissue are the cause of the movement. No migration relative to the tissue growth was observed in only half of the implants and porosity was observed to have no influence on fixation. Further research is required to study the behaviour of growing tissue and tissue attachment to porous materials to ensure long term anchorage. Mechanical reliability of the implant was determined by its resistance to applied loading conditions that are expected during service. Cyclic fatigue, compression, puncture and impact tests were conducted to establish the suitability of the radio frequency tag's housing. The alumina case passed all the required conditions with the possible exception of impact. Ceramics are known to be strong materials but brittle and thus they have low resistance to impact forces. Improving the ductility of alumina will improve its impact strength. Further research is required to make this improvement but assessment of the housing's risk to such loading condition should be made first. Other research that needs to be conducted is the response of the full 136 configuration of the housing (i.e. with porous layer) to applied forces while implanted in soft tissue. Results may vary from what has been discussed and more accurate predictions can be made. While the mechanical and biological reliability were examined during this research, both were treated separately however in practice, each can affect the other. Questions need to be answered that address both topics as one subject. For example, 'how will deep ocean pressures affect the fixation of the implant?' or 'how does impact loading change when applied to the implant in the body?'. More sophisticated tests need to be designed to answer such questions. As is the case in most medical device testing, customized tests and apparatus are needed to simulate the correct conditions. Cadavers may also be needed to represent the loadings absorbed by the body. Along with these tests, better defined criteria need to be set to establish the performance of the implant for long term application. Much progress has been made on predicting the performance of the housing however more work must be done to ensure the proper biological and mechanical stability for long term implantation. Additional research must be accomplished before using the device for endangered animals. 137 REFERENCES [I] Trites, A. and Larkin, P., Changes in the Abundance of Stellar Sea-lions (Eumetopias jubatus) in Alaska from 1956 to 1992: how many were there?, Aquatic Mammals, 22: 153-166. [2] Whitfield-Gibbons, J . and Andrews, K., PIT Tagging: Simple Technology at Its Best, Bioscience, 54: 447-454, 2004. [3] Brockway, K., Database Support for Microchip Identification in Companion Animal Identification, M. Sc. Thesis, 2002. [4] Hench, L. and Ethridge, E., Biomaterials: An Interfacial Approach, Academic Press, New York, 1982. [5] Von Recum, A., Handbook of Biomaterials Evaluation, Macmillan, New York, 1986. [6] Hench, L., Bioceramics: From Concept to Clinic, Journal of American Ceramics Society, 74: 1487-1510, 1991. [7] Park, J . and Bronzino, J. , Biomaterials: Principles and Applications, CRC Press, Boca Raton, 2003. [8] Davilla, J. , Lautsch, E. and Palmer, T., Some physical factors affecting the acceptance of synthetic materials as tissue implants, Annals of the New York Academy of Sciences, 146: 138-147, 1968. [9] Hulbert, S., Cooke, F., Klaiwitter, J . , Leonard, R., Sauer, B. and Moyle, D., Attachment of prostheses to the musculoskeletal system by tissue ingrowth and mechanical interlocking, Journal of Biomedical Materials Research Symposium, 4: 1-23, 1973. [10] Ko, W., Implantable Sensors for Closed-loop Prosthetic Systems, Futura Pub. Co., New York, 1985. [II] Reichert, W. and Saavedra, S., Materials consideration in the selection, performance, and adhesion of polymeric encapsulants for implantable sensors, In: Williams, D., editor, Medical and dental materials, Wiley, Weinheim, 1992. [12] [13] Boretos, J . , Concise Guide to Biomedical Polymers, Thomas Books, Illinois, 1973. [14] Autian, J . , Problems in the use of polymeric materials in medical and biomedical applications, Annals of the New York Academy of Sciences, 146: 251 -261, 1968. [15] Winter, G., Leray, J . and de Groot, K., Evaluation of Biomaterials, Wiley, New York, 1980. [16] Ratner, B. and Castner, D., Surface Modification of Polymeric Biomaterials, Plenum Press, New York, 1997. [17] Ward, K., Slobodzian, E., Tiekotter, K., and Wood, M., The effect of microgeometry, implant thickness and polyurethane chemistry on the foreign body response to subcutaneous implants, Biomaterials, 23: 4185-4192, 2002. [18] Walboomers, X. and Jansen, J . , Microgrooved silicone subcutaneous implants in guinea pigs, Biomaterials, 21: 629-636, 2000. 138 [19] Taylor, S. and Gibbons, D., Effect of surface texture on the soft tissue response to polymer implants, Journal of Biomedical Materials Research, 17: 205-227, 1983. [20] den Braber, E., Ruijter, J . , and Jansen, J . , The effect of a subcutaneous silicone rubber implant with shallow surface microgrooves on the surrounding tissues in rabbits, Journal of Biomedical Materials Research, 37: 539-547, 1997. [21] Bern, S., Burd, A. and May, J. , The biophysical and histologic properties of capsules formed by smooth and textured silicone implants in the rabbit, Plastic and Reconstructive Surgery, 89: 1037-1044, 1992. [22] [23] Mayr, W., Bijak, M., Rafolt, D., Sauermann, S., Unger, E. and Lanmuller, H., Basic design and construction of the Vienna FES implants: existing solutions and prospects for new generations of implants, Medical Engineering and Physics, 23: 53-60, 2001. [24] Harms, J . and Mausle, E., Tissue Reaction to Ceramic Implant Material, Journal of Biomedical Materials Research, 13: 67-87, 1979. [25] Petillo, O., Barbarisi, A., Margarucci, S., De Rosa, A., and Peluso, G., Materials/Biological Environment Interaction. In: Barbucci, R., editor. Integrated Biomaterials Science, Kluwer Academic, New York, 2002. [26] Pacquay, Y., Ruijter, J . , van der Waerden, J . and Jansen, J . , Wound healing phenomena in titanium fiber mesh: the influence of the length of implantation, Biomaterials, 18: 161-166, 1997 [27] Munz, D. and Fett, T., Ceramics: Mechanical Properties, Failure Behaviour, Materials Selection, Springer, Berlin, 1999. [28] Bonfield, W., Lacefield, W., Clark, T., Kelly, S., Kellman, D., Lee, R., Page, D., Renz, E., Soltesz, U., Tolbert, J . and Wolfarth, D., Biomechanical Stability and Design: Strength, In: Ducheyne, P. and Lemons, J . , editors, Bioceramics: Material Characteristics versus in vivo behaviour, Annals of The New York Academy of Sciences, New York, 1988. [29] Lemons, J. , Aluminum Oxide Ceramics as Biomaterials, In: Vincenzini, P. editor, Ceramics in Surgery, Elsevier, Amsterdam, 1983. [30] Zeibig, A. and Luber, H., Bioceramic Hip Joint Components - Industrial Production and Testing Procedures to Ensure High Functional Reliability, In: Vincenzini, P., editor, Ceramics in Surgery, Elsevier, Amsterdam, 1983. [31] Hulbert, S., Morrison, S. and Klawitter, J. , Tissue Reaction to Three Ceramics of Porous and Non-Porous Structures, Journal of Biomedical Materials Research, 6: 347-374, 1972. [32] Ravaglioli, A. and Krajewski, A., Bioceramics, Chapman and Hall, London, 1992. [33] Klaiwaitter, J . , Weinstein, A., Cooke, F., Peterson, L., Pennel, B. and McKinney, R., An Evaluation of Porous Alumina Ceramic Dental Implants, Journal of Dental Research, 56: 768-776, 1977. [34] Hulbert, S., Morrison, S. and Klawitter, J. , Compatibility of Porous Ceramics with Soft Tissue; Application to Tracheal Prostheses, Journal of Biomedical Materials Research Symposium, 2: 269-279, 1971. 139 [35] Klawitter, J . and Hulbert, S., Application of Porous Ceramics for the Attachment of Load Bearing Internal Orthopedic Application, Journal of Biomedical Materials Research Symposium, 2: 161-229, 1971. [36] Davis, S., Gibbons, D., Martin, R., Levitt, S., Smith, J . and Harrington, R., Biocompatibility of Ceramic Implants in Soft Tissue, Journal of Biomedical Materials Research, 6: 425-449, 1972. [37] Hulbert, S., Matthews, J . , Klawitter, J . , Sauer, B. and Leonard, R., Effect of Stress on Tissue Ingrowth into Porous Aluminum Oxide, Journal of Biomedical Materials Research Symposium, 5: 85-97, 1974. [38] [ [39] Zhang, S., Emerging biological materials through molecular self-assembly, Biotechnology Advances, 20: 321-339, 2002. [40] [Bumgardner, J . , Wiser, R., Gerard, P., Bergin, P., Chestnutt, B., Marini, M., Ramsey, V., Elder, S. and Gilbert, J . , Chitosan: potential use as a bioactive coating for orthopaedic and craniofacial/dental implants, Journal of Biomaterial Science, 14: 423-438, 2003. [41] Lee, W. and Hairston, P., Cardiovascular Implants State of the Art and Potential For Bioceramics, In: Hulbert, S. and Young, F., editors, Use of Ceramics in Surgical Implants, Gordon and Breach, New York, 1969. [42] Hentrich, R., Graves, G., Stein, H. and Bajpai, P., An evaluation of inert and resorbable ceramics for future clinical orthopedic applications, Journal of Biomedical Materials Research, 5: 25-51, 1971. [43] Coleman, D., King, R. and Andrade, J . , The foreign body reaction: a chronic inflammatory response, Journal of Biomedical Materials Research, 8: 199-211, 1974. [44] Salthouse, T., Some aspects of macrophage behaviour at the implant interface, Journal of Biomedical Materials Research, 18: 395-401, 1984. [45] Shankar, R. and Greisler, H., Inflammation and Biomaterials, In: Greco, R. editor, Implantation Biology: The Host Response and Biomedical Devices, CRC Press, London, 1994. [46] Magnani, A., Gianfranco, P., Margarucci, S. and Chittur, K., Protein Adsorption and Cellular/Tissue Interactions. In: Barbucci, R., editor. Integrated Biomaterials Science, Kluwer Academic, New York, 2002. [47] Thomsen, P. and Gretzer, C , Macrophage interactions with modified material surfaces, Solid State and Materials Science, 5: 163-176, 2001. [48] Fraser, D., An introduction to in vivo biosensing: progress and problems, In: Fraser, D., editor, Biosensors in the body: continuous in vivo monitoring, Wiley, England, 1997. [49] Sharkawy, A., Klitzman, B., Truskey, G. and Reichert, W., Engineering the tissue which encapsulates subcutaneous implants. I. Diffusion properties, Journal of Biomedical Materials Research, 37: 401-412, 1997. [50] Parker, J . , Walboomers, X., Von den Hoff, J. , Maltha, J . and Jansen, J . , Soft tissue response to microtextured silicone and poly-L-lactic acid implants: fibronectin pre-coating vs. radio-frequency glow discharge treatment, Biomaterials, 23: 3545-3553, 2002. 140 [51] Wood, R., LeCluyse, E. and Fix, J. , Assessment of a model for measuring drug diffusion through implant-generated fibrous capsule membranes, Biomaterials, 18: 957-959, 1995. [52] Matlaga, B., Yasenchak, L. and Salthouse, T., Tissue response to implanted polymers: the significance of sample shape, Journal of Biomedical Materials Research, 10: 391-397, 1976. [53] Werkmeister, J . , Tebb, T., White, J . and Ramshaw, J . , Collagenous tissue formation in association with medical implants, Solid State and Materials Science, 5: 185-191, 2001. [54] Zhao, S., Pinholt, E., Madsen, J . , and Donath, K., Histological evaluation of different biodegradable and non-biodegradable membranes implanted subcutaneously in rats, Journal of Cranio-Maxillofacial Surgery, 28: 116-122, 2000. [55] Woodward, S. and Salthouse, T., The tissue response to implants and its evaluation by light microscopy. In: von Recum, A., editor. Handbook of Biomaterials Evaluation, Macmillan, New York, 1986. [56] Parker, J . , Walboomers, X., Von den Hoff, J . , Maltha, J. , and Jansen, J . , The effect of bone anchoring and micro-grooves on the soft tissue reaction to implants, Biomaterials 23: 3887-3896, 2002. [57] Walboomers, X., Croes, H., Ginsel, L, and Jansen, J . , Microgrooved subcutaneous implants in the goat, Journal of Biomedical Materials Research, 42: 634-641, 1998. [58] Janssens, S., Rocha, L, Bosschaerts, L, Barbosa, M., Puers, R., Ville, H., and Geers, R., Implant recovery and tissue reaction in growing pigs following implantation of packaging materials for injectable electronic identification and monitoring of devices, Preventative Veterinary Medicine, 25: 249-258, 1996. [59] Chehroudi, B. and Brunette, D., Subcutaneous microfabricated surfaces inhibit epithelial recession and promote long-term survival of percutaneous implants, Biomaterials, 23: 229-237, 2002. [60] Chehroudi, B., Gould, T. and Brunette, D., Titanium-coated micro-machined grooves of different dimensions affect epithelial and connective-tissue cells differently in vivo, Journal of Biomedical Materials Research, 24: 1203-1219, 1990. [61] Picha, G. and Drake, R., Pillared-surface microstructure and soft-tissue implants: effect of implant site and fixation, Journal of Biomedical Materials Research, 30: 305-312, 1996. [62] Hayashi, K., Mechanical properties of biomaterials: relationship to clinical applications. In: Boretos, J . and Eden, M., Contemporary Biomaterials, Noyes Publications, New Jersey, 1984. [63] Carey, D., The surface stability of polymers with absorbed fibronectin, Journal of Biomedical Materials Research, 49: 12-6, 2000. [64] Zhang, Y., Bjursten, L, Freij-Larsson, C , Kober, M., and Wesslen, B., Tissue response to commercial silicone and polyurethane elastomers after different sterilization procedures, Biomaterials, 17: 2265-2272, 1996. [65] Giardino, R. and Aldini, N., Infection and Sterilization. In: Barbucci, R., editor. Integrated Biomaterials Science, Kluwer Academic, New York, 2002. [66] Rose, R. and Litsky, A., Biomechanical Considerations in the Loosening of Hip Replacement Prostheses, In: Williams, D., editor, Current Perspectives on Implantable Devices, Jai Press, London, 1989. 141 [67] Klinge, U., Klosterhalfen, B., Birkenhauer, V., Junge, K., Conze, J . , and Schumpelick, V., Impact of Polymer Pore Size on the Interface Scar Formation in a Rat Model, Journal of Surgical Research, 103: 208-214, 2002. [68] Sharkawy, A., Klitzman, B., Truskey, G. and Reichert, W., Engineering the tissue which encapsulates subcutaneous implants. I. Diffusion properties, Journal of Biomedical Materials Research, 37: 401-412, 1997. [69] Sharkawy, A., Klitzman, B., Truskey, G. and Reichert, W., Engineering the tissue which encapsulates subcutaneous implants. II. Plasma-tissue exchange properties, Journal of Biomedical Materials Research, 40: 586-597, 1998. [70] Helmer, J . and Driskell, T., Research on Bioceramics, In: Hulbert, S. and Young, F., editors, Use of Ceramics in Surgical Implants, Gordon and Breach, New York, 1969. [71] Hulbert, S., Morrison, S. and Klawitter, J. , Compatibility of Porous Ceramics with Soft Tissue: Application to Tracheal Prostheses, Journal of Biomedical Materials Research Symposium, 2: 269-279, 1971/ [72] Stark, K., Morris, R. and Pfeiffer, D., Comparison of electronic and visual identification systems in pigs, Livestock Production Science, 53: 143-152,' 1998. [73] Baumgaertner, M., Curtin, S., Lindskog, D. and Keggi, J. , The value of the Tip-Apex Distance in Predicting Failure of Fixation of Pertrochanteric Fractures of the Hip, The Journal of Bone and Joint Surgery (American Volume), 77:'1058-1064, 1995. [74] Ehmke, L., Fitzpatrick, D., Krieg, J . , Madey, S. and Bottlang, M., Lag screws for hip fracture fixation: Evaluation of migration resistance under simulated walking, Journal of Orthopaedic Research, 23: 1329-1335, 2005. [75] Kitamura, A., Removal of a migrated dental implant from maxillary sinus by transnasal endoscopy, British Journal of Oral and Maxillofacial Surgery, (Article in Press), 2006. [76] Hoffman, R., Cochlear implant in the child under two years of age: skull growth, otitis media and selection, Otolaryngology, Head and Neck Surgery, 117: 217-219, 1997. [77] Maher, S. And Prendergast, P., Discriminating the loosening behaviour of cemented hip prostheses using measurements of migration and inducible displacement, Journal of Biomechanics, 35: 257-265, 2002. [78] Hendrich, C , Bahlmann, J . and Eulert, J . , Migration of the Uncemented Harris-Galante Acetabular Cup, The Journal of Arthroplasty, 12: 889-895, 1997. [79] Fitzpatrick, T., Liinamaa, T., Brown, I., Cameron, T. and Richmond, F., A Novel Method to Identify Migration of Small Implantable Devices, Journal of Long-term Effects of Medical Implants, 6: 157-168, 1996. [80] Jou, I. and.Lai, K., Acute median nerve injury due to migration of a kirschner wire, Journal of Hand Surgery (British and European Volume), 23: 112-113, 1998. [81] lida, S., Tanaka, N., Kogo, M. and Matsuya, T., Migration of a dental implant into the maxillary sinus, International Journal of Oral and Maxillofacial Surgery, 29: 358-359, 2000. [82] Massaro-Giordano, M., Kirschner, R. and Wulc, A., Orbital Floor Implant Migration Across the Ethmoidal Sinuses and Nasal Septum, American Journal of Ophthalmology, 126: 848-850, 1998. 142 [83] Kneif, D., Downing, M., Ashcroft, G., Knight, D., Ledingham, W., Gibson, P. and Hutchison, J . , The Correlation Between Immediate Radiolucent Lines and Early Implant Migration in Cemented Acetabular Components, The Journal of Arthroplasty, 21: 215-220, 2006. [84] Wilkinson, J . , Eagleton, A., Stockley, I., Peel, N., Hamer, A. and Eastell, R., Effect of pamidronate on bone turnover and implant migration after total hip arthroplasty: a randomized trial, Journal of Orthopaedic Research, 23: 1-8, 2005. [85] O'Sullivan, A., O'Malley, K., Fitzpatrick, J . and Smith, J. , Migration of a Prosthetic Acetabulum into the Bladder, The Journal of Urology, 166: 617, 2001. [86] Ng, W., Foreign materials in tissues, pathogenesis and clinical implications, Current Diagnostic Pathology, 9: 85-95, 2003. [87] Beddoes, J . , Lecture Notes: Materials: Strength and Fracture, Carleton University, 2002. [88] Dorre, E. and Hubner, H., Alumina, Springer-Verlag, Berlin, 1984. [89] Benham, P., Crawford, R. and Armstrong, C , Mechanics of engineering materials, Longman Ltd., England, 1996. [90] Wachtman, J . , Mechanical Properties of Ceramics, Wiley-lnterscience, New York, 1996. [91] Guiu, F. and Li, M., Role of Crack-Bridging Ligaments in the Cyclic Fatigue Behaviour of Alumina, Journal of the American Ceramic Society, 75: 2976-2984, 1992. [92] Hu, X. and Mai, Y., Crack-Bridging Analysis for Alumina Ceramics under Monotonic and Cyclic Loading, Journal of the American Ceramic Society, 75: 848-853, 1992. [93] Lathabai, S., Rodel, J . and Lawn, B., Cyclic Fatigue from Frictional Degradation at Bridging Grains in Alumina, Journal of the American Ceramic Society, 74: 1340-1348, 1991. [94] Miannay, D., Time-Dependent Fracture Mechanics, Springer, New York, 2001. [95] Poursartip, A. and Street, K., Proceedings of the International Conference on Composite Materials Volume V: Structures, Woodhead Publishing Limited, Vancouver, 1995. [96] Lawn, B., Fracture of brittle solids, Cambridge University Press, Cambridge, 1993. [97] Ravi-Chandar, K., Dynamic Fracture, Elsevier, Oxford, 2004. [98] Vollmann, J . , Profunser, D., Goossens, J . and Dual, J. , Mechanical Wave Propagation at phase- and material interfaces of microstructures, Proc. of SPIE, 5392: 203-210, 2004. [99] Kobayashi, A., Dynamic and Impact Fractures of Ceramic Composites at Elevated Temperatures, In: Nair, S. and Jakus, K., editors, High Temperature Mechanical Behaviour of Ceramic Composites, Elsevier, Amsterdam, 1995. [100] Sherman, D. and Ben-Shushan, T., Quasi-static impact damage in confined ceramic tiles, International Journal of Impact Engineering, 21: 245-265, 1998. [101] Madhu, V., Ramanjaneyulu, K., Bhat, T. and Gupta, N., An experimental study of penetration resistance of ceramic armour subjected to projectile impact, International Journal of Impact Engineering, 32: 337-350, 2005. 143 [102] Rajendran, A. and Dandekar, D., Inelastic response of alumina, International Journal of Impact Engineering, 17: 649-660, 1995. [103] Lopez-Puente, J . , Arias, A., Zaera, R. and Navarro, C , 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: 321-336, 2005. [104] Maher, S., Lipman, J . , Curley, L, Gilchrist, M. and Wright, T., Mechanical Performance of Ceramic Acetabular Liners Under Impact Conditions, The Journal of Arthroplasty, 18: 936-941, 2003. [105] Green, D., An introduction to the mechanical properties of ceramics, Cambridge University Press, Cambridge, 1998. [106] Lawn, B., Fracture of Brittle Solids, Cambridge University Press, New York, 1993. [107] Ciavarella, M. and Macina, G., New Results for the fretting-induced stress concentration on Hertzian and flat rounded contacts, International Journal of Mechanical Sciences, 45: 449-467, 2003. [108] Ciavarella, M., Indentation by nominally flat or conical indenters with rounded corners, International Journal of Solids and Structures, 36: 4149-4181, 1999. [109] Ciavarella, M. and Hills, D., The influence of the indenter tip-radius on indentation testing of brittle materials, Journal of the European Ceramic Society, 19: 239-245, 1999. [110] Gitzen, W., Alumina as a Ceramic Material, The American Ceramic Society, Ohio, 1970. [111] Cook, S. and Weinberg, L., The effect of aging and surface modification on the mechanical properties of dense alumina oxide, Biomaterials, Medical Devices arid Artificial Organs, 12: 67-74, 1984. [112] Kimura, Y. and Takubo, S., Corrosion fatigue of bio-ceramic sapphire in isotonic sodium chloride solution, International Journal of Fatigue, 22: 899-904, 2000. [113] Barinov, S., Ivanov, N., Orlov, S. and Shevchenko, V., Dynamic Fatigue of Alumina Ceramics in Water-Containing Environment, Ceramics International, 24: 421-425, 1998. [114] Barinov, S., Ivanov, N., Orlov, S. and Shevchenko, V., Influence of Environment on Delayed Failure of Alumina Ceramics, Journal of the European Ceramic Society, 18: 2057-2063, 1998. [115] Barinov, S., Fateeva, L., Ivanov, N., Orlov, S. and Schevchenko, V., Effect of Acidic Environment on Subcritical Crack Growth in Alumina Ceramics, Scripta Materialia, 38: 975-980, 1998. [116] Krainess, F. and Knapp, W., Strength of a Dense Alumina Ceramic after Aging in vitro, Journal of Biomedical Materials Research, 12: 241-246, 1978. [117] Fahr, A., Brown, R. and Day, D., In vivo and in vitro aging of orthopedic aluminas, Journal of Biomedical Materials Research, 17: 395-410, 1983. [118] Zhang, Y. and Lawn, B., Long-Term strength of ceramics for Biomedical Applications, Journal of Biomedical Materials Research Part B: Applied Biomaterials, 69: 166-172, 2004. [119] Krell, A., Pippel, E., Woltersdorf, J . and Burger, W., Subcritical crack growth in Al203 with submicron grain size, Journal of the European Ceramic Society, 23: 81-89, 2003. 144 [120] Jones, R., Stress-Corrosion Cracking: Materials Performance and Evaluation, ASM International, Ohio, 1992. [121] Drummond, J . , Simon, M. and Brown, S., Degradation of Ceramic Materials in Physiological Media, In: Syrett, B. and Acharya, A., editors, Corrosion and Degradation of Implant Materials, ASTM, Baltimore, 1979. [122] Zhu, P., Lin, Z., Chen, G. and Kiyohiko, I., The predictions and applications of fatigue lifetime in alumina and zirconia ceramics, International Journal of Fatigue, 26: 1109-1114, 2004. [123] Lawn, B., Physics of Fracture, Journal of the American Ceramic Society, 66: 83-91, 1983. [124] [123] Michalske, T. and Freiman, S., A molecular interpretation of stress corrosion in silica, Nature, 295: 511-512, 1982. [125] Cochran, W., Wildlife telemetry. In S. D. Schemnitz, editor, Wildlife management techniques manual, The Wildlife Society, Washington, D.C., 1980. [126] [127] Drobeck, H., Rothstein, S., Gumaer, K., Sherer, A. and Slighter, R., Histologic Observation of Soft Tissue Responses to Implanted, Multifaceted Particles and Discs of Hydroxylapatite, Journal of Oral and Maxillofacial Surgery, 42: 143-149, 1984. [128] Wang, X. and Agrawal, C , Interfacial fracture toughness of Tissue-Biomaterial Systems, Journal of Biomedical Materials Research, 38: 1-10, 1997. [129] Hulbert, S., Cooke, F., Klawitter, J. , Leonard, R., Sauer, B. and Moyle, D., Attachment of Prostheses to the Musculoskeletal System by Tissue Ingrowth and Mechanical Interlocking, Journal of Biomedical Materials Research Symposium, 4: 1-23, 1973. [130] Greene, D., Pruitt, L. and Maas, C , Biomechanical Effects of e-PTFE Implant Structure on Soft Tissue Implantation Stability: A study in the Porcine Model, Laryngoscope, 107: 957-962, 1997. [131] Bundy, K., Schlegel, U., Rahn, B., Geret, V. and Perren, S., An improved peel test method for measurement of adhesion to biomaterials, Journal of Materials Science: Materials in Medicine, 11:517-521,2000. [132] Hacking, S., Bobyn, J . , Toh, K., Tanzer, M. and Krygier, J . , Fibrous tissue ingrowth and attachment to porous tantalum, Journal of Biomedical Materials Research, 52: 631-638, 2000. [133] Bobyn, J . , Wilson, G., MacGregor, D., Pilliar, R. and Weatherly, G., Effect of pore size on the peel strength of attachment of fibrous tissue to porous-surfaced implants, Journal of Biomedical Materials Research, 16: 571-584, 1982. [134] Bellon, J . , Contreras, L., Bujan, J . and Carrera-San Martin, A., Experimental assay of a Dual Mesh polytetrafluoroethylene prosthesis (non-porous on one side) in the repair of abdominal wall defects, Biomaterials, 17: 2367-2372, 1996. [135] Rice, R., The Porosity Dependence of Physical Properties of Materials: A Summary Review, Key Engineering Materials, 115: 1-20, 1996. [136] Maca, K., Dobsak, P. and Boccaccini, A., Fabrication of graded porous ceramics using alumina-carbon powder mixtures, Ceramics International, 27: 577-584, 2001. 145 [137] Corbin, S., Lee, J . and Qiao, X., Influence of Green Formulation and Pyrolyzable Particulates on the Porous Microstructure and Sintering Characteristics of Tape Cast Ceramics, Journal of the American Ceramic Society, 84: 41-47, 2001. [138] Corbin, S. and Apte, P., Engineered Porosity via Tape Casting, Lamination and the Percolation of Pyrolyzable Particulates, Journal of the American Ceramic Society, 82: 1693-1701, 1999. [139] Jonker, R. and Treites, A., Morphometric measurements and body condition of healthy and starveling steller sea lion pupus (eumetopias jubatus), Aquatic Mammals', 26: 151 -157, 2000. [140] ISO Standards 10993 series for Biological Evaluation of Medical Devices [141] Ravin, A., Olbrich, K., Levin, L., Usala, A. and Klitzman B., Long- and Short-term effects of biological hydrogels on capsule microvascular density around implants in rats, Journal of Biomedical Materials Research, 58: 313-318, 2001. [142] Sanders, J . , Stiles, C. and Hayes, O , Tissue response to single-polymer fibers of varying diameters: evaluation of fibrous encapsulation and macrophage density, Journal of Biomedical Materials Research, 52: 231-237, 2000. [143] Bilsen, P., Popa, E., Brouwer, L., Vincent, J : , Taylor, C , de Leij, L., Hendriks, M. and van Luyn, M., Ongoing foreign body reaction to subcutaneous implanted (heparin) modified Dacron in rats, Journal of Biomedical Materials Research Part A, 68: 423-427, 2004. [144] Richerson, D., Modern Ceramic Engineering, Marcel Dekker Inc, New York, 1992. [145] Ring, T., Fundamentals of Ceramic Powder Processing and Synthesis, Academic Press, San Diego, 1996. [146] Chao, W. and Chou, K., Studies on the Control of Porous Properties in the Fabrication of Porous Supports, Key Engineering Materials, 115: 93-108, 1996. [147] Kawamata, S., Ozawa, J. , Hashimoto, M., Kurose, T. and Shinohara, H., Structure of the rat subcutaneous connective tissue in relation to its sliding mechanism, Archives of Histology and Cytology, 66: 273-279, 2003. [148] Roberts, C , Walker, D. and Schellenberg, R., Extracellular Matrix, In: Zweiman, B. and Schwartz, L., editors, Inflammatory Mechanisms in Allergic Diseases, Marcel Dekker, New York, 2002. [149] Ohara, T., On the comparison of strength of various organ tissues, Journal of Kyoto Prefectural University of Medicine, 53: 577-597, 1953. [150] Mochizuki, T., On the tension test made upon the wall of the thoracic aorta in animals, Journal of Kyoto Prefectural University of Medicine, 52: 103-106, 1952. [151] Asami, I., Study on the strength of human urinary organs, Journal of Kyoto Prefectural University of Medicine, 70: 767-788, 1961. [152] Tsuang, Y., Sun, J . , Chen, I., Hsu, S., Tsao, K., Wei, K. and Hang, Y., The effects of cyclic stretching on tensile properties of the rabbit's skeletal muscle, Clinical Biomechanics, 13: 48-53, 1998. 146 [153] Behrend, M., Kluge, E., Schuttler, W. and Klempnauer, J. , Breaking strength of native and sutured trachea: an experimental study on sheep trachea, European Surgical Research, 33: 255-263,2001. [154] Abe, H., Hayashi, K. and Sato, M., Data book on mechanical properties of living cells, tissues and organs, Springer, New York, 1996. [155] Craigie, E., Bensley's Practical Anatomy of the Rabbit, University of Toronto Press, Toronto, 1948. [156] Andrade, J. , Interfacial phenomena and biomaterials, Medical Instrumentation, 7: 110-120, 1973. [157] Shanklin, D. and Smalley, D., Dynamics of Wound Healing after Silicone Device Implantation, Experimental and Molecular Pathology, 67: 26-39, 1999. [158] Zhu, C , Ohsaki, K., Li, K., Ye, Q., Tran, Y., Ohba, Y. and Moriyama, K., Long-term observation of subcutaneous tissue reaction to synthetic auditory ossicle (Bioceram) in rats, The Journal of Medical Investigation, 46: 97-103, 1999. [159] Catledge, S., Cook, M., Vohra, Y., Santos, E., McClenny, M. and Moore, D., Surface crystalline phases and nanoindentation hardness of explanted zirconia femoral heads, Journal of Materials Science: Materials in Medicine, 14: 863-867, 2003. [160] Milleding, P. and Karlsson, S., On the surface elemental composition of non-corroded and corroded dental ceramic materials in vitro, Journal of Materials Science: Materials in Medicine, 14: 557-566, 2003. [161] Batchelor, A. and Chandrasekaran, M., Service Characteristics of Biomedical Materials and Implants, World Scientific, London, 2004. [162] Loughlin, T., Sterling, J . , Merrick, R., Sease, J . and York, A., Diving behavior of immature Steller sea lions (Eumetopias Jubatus), Fisheries Bulletin, 101: 566-582, 2003. [163] [164] Lee, Y., Pan, J. , Hathaway, R. and Barkey, M., Fatigue Testing and Analysis: Theory and Practice, Elsevier, Oxford, 2005. [165] Kaplan, A., Effect of Loading Cycle Parameters on Fatigue-Crack Growth (Review), Moscow Aviation Institute, 14: 58-68, 1978. [166] Matthewson, J . and Padiyar, V., Cyc//c fatigue of high strength optical fibers in bending, Proceedings of the International Society for Optical Engineering, 4215: 53-59, 2001 [167] ASTM, Standard practice for Conducting Amplitude Axial Fatigue Tests of Metallic Materials, ASTM Standard E 466, ASTM Designation E 466 - 82, pp. 465-469, 1982. [168] Suresh, S., Fatigue of Materials, Cambridge University Press, New York, 1998. [169] [170] [171] 147 [172] Homewood, C , Cracked Tooth syndrome - Incidence, clinical findings and treatment, Australian Dental Journal, 43, 1998. [173] Huber, D., Eason, T., Hueter, R. and Motta, P., Analysis of the bite force and mechanical design of the feeding mechanism of the durophagous horn shark Heterodontus francisci, The Journal of Experimental Biology, 208: 3553-3571, 2005. [174] [175] [176] Rosentritt, M., Plein, T., Kolbeck, C , Behr, M. and Handel, G., In vitro fracture force and marginal adaptation of ceramic crowns fixed on natural and artificial teeth, The International Journal of Prosthodontics, 13: 387-391, 2000. [177] Kohyama, K., Hatakeyama, E., Sasaki, T., Dan, H., Azuma, T. and Karita, K., Effects of sample hardness on human chewing force: a model study using silicone rubber, Archives of Oral Biology, 49: 805-816, 2004. [178] [179] [180] [181] Johnson, K., Fischer, T., Chapman, S. and Wilson, B., Accidental head injuries in children under 5 years of age, Clinical Radiology, 60: 464-468, 2005. [182] Wood, J . , Dynamic Response of Human Cranial Bone, Journal of Biomechanics, 4: 1-12, 1971. [183] Hoick, P., What can a baby's skull withstand? Testing the skull's resistance on an anatomical preparation, Forensic Science International, 151: 187-191, 2005. [184] ASTM, Standard test method for Impact Resistance of flat, rigid plastic specimen by means of a striker impacted by a falling weight (Gardner Impact), ASTM Standard D 5420, ASTM Designation D 5420 - 93, pp. 457-463, 1993. [185] ASTM, Standard test method for Resistance of organic coatings to the effects of rapid deformation (Impact), ASTM Standard D 2794, ASTM Designation D 2794 - 93, pp. 378-379, 1993. [186] Huang, X. and Tu, M., Effect of single impact damage on strength of alumina and zirconia-toughened alumina, Journal of Materials Science Letters, 15: 1925-1926, 1996. [187] Davidge, R. and Phillips, D., The significance of impact data for brittle non-metallic materials, Journal of Materials Science, 7:1308-1314, 1972. [188] Varner, J . and Frechette, V., Fractography of glasses and ceramics, The American Ceramic Society, New York, 1988. [189] Chen, C , Takita, K., Honda, S. and Awaji, H., Fracture Behavior of Cylindrical Porous Alumina with Pore Gradient, Journal of the European Ceramic Society, 25: 385-391, 2005. [190] Nakamura, T. and Wang, Z., Simulations of Crack Propagation in Porous Materials, Transactions of the American Society of Mechanical Engineers, 68: 242-251, 2001. 148 [191] Leguillon, D., Tariolle, S., Martin, E., Chartier, T. and Besson, J . , Prediction of crack deflection in porous/dense ceramic laminates, Journal of the European Ceramic Society, 26: 343-349, 2006. [192] Wolfarth, D., Ducheyne, P., Effect of a change in interfacial geometry on the fatigue strength of porous coated Ti-6AI-4V, Journal of Biomedical Materials Research, 4: 417-425, 1994. [193] Kohn, D. and Ducheyne, P., A parametric study of the factors affecting the fatigue strength of porous coated Ti-6AI-4V implant alloy, Journal of Biomedical Materials Research, 24: 1483-1501,1990. [194] Healy, J . , Bushby, A., Mai, Y. and Mukhopadhyay, A., Cyclic fatigue of long and short cracks in alumina, Journal of Materials Science, 32: 741-747, 1997. [195] Lathabai, S., Mai, Y. and Lawn, B., Cyclic fatigue behavior of an alumina ceramic with crack-resistance characteristics, Journal of the American Ceramic Society, 72: 1760-1763, 1989. [196] Zheng, Q., Zheng, X. and Wang, F., On the expressions of fatigue life of ceramics with given survivability, Engineering Fracture Mechanics, 53: 49-55, 1996. [197] Reece, M., Guiu, F. and Sammur, M., Cyclic fatigue crack propagation in alumina under direct tension-compression loading, Journal of the American Ceramic Society, 72: 348-352, 1989. [198] Guiberteau, F., Padture, N. and Lawn, B., Effect of Grain Size on Hertzian Contact Damage in Alumina, Journal of the American Ceramic Society, 77: 1825-1831, 1994. [199] Latella, B., O'Conner, B., Padture, N. and Lawn, B., Hertzian Contact Damage in Porous Alumina Ceramics, Journal of the American Ceramic Society, 80: 1027-1031, 1997. [200] She, J . , Yang, J. , Beppu, Y. and Ohji, T., Hertzian contact damage in a highly porous silicon nitride ceramic, Journal of the European Ceramic Society, 23: 1193-1197, 2003. [201] Jayaseelan, D., Kondo, N., Brito, M. and Ohji, T., High-strength porous alumina ceramics by the pulse electric current sintering technique, Journal of American Ceramic Society, 85: 267-269, 2002. [202] Biasini, V., Parasporo, M. and Bellosi, A., Fabrication and characterization of AI203 porous bodies by hot isostatic pressing, Thin Solid Films, 297: 207-211, 1997. [203] Li, G., Jiang, Z., Jiang, A. and Zhang, L., Strengthening of Porous Al203 Ceramics through Nanoparticle Addition, Nanostructured Materials, 8: 749-754, 1997. [204] Kannus, P., Parkkari, J . and Poutala, J . , Comparison of Force attenuation Properties of Four Different Hip Protectors Under Simulated Falling Conditions in the Elderly: An In Vitro Biomechanical Study, Bone, 25: 229-235, 1999. [205] Mills, N. and Moosa, A., Impacts of Hemispherical Strikers on Polystyrene Bead Form, Journal of Cellular Plastics, 35: 289-310, 1999. [206] Yang, K. and Kobayashi, A., Dynamic Fracture Responses of Alumina and Two Ceramic Composites, Journal of the American Ceramic Society, 73: 2309-2315, 1990. [207] McMillan, P. and Tesh, J. , Dynamic effects in the impact testing of brittle materials, Journal of Materials Science, 10: 621-632, 1975. 149 [208] Jiang, B. and Weng, G., A theory of compressive yield strength of nano-grained ceramics, International Journal of Plasticity, 20: 2007-2026, 2004. [209] Webster, T., Siegel, R. and Bizios, R., Design and evaluation of nanophase alumina for orthopaedic/dental applications, Nanostructures materials, 12: 983-986, 1999. [210] Bhaduri, S. and Bhaduri, S., Enhanced low temperature toughness of Al203-Zr02 Nano/Nano Composites, Nanostructured Materials, 8: 755-763, 1997. [211] Ghadiri, M. and Zhang, Z., Impact attrition of particulate solids. Part 1: A theoretical model of chipping, Chemical Engineering Science, 57: 3659-3669, 2002. [212] He, Z., Ma, J . , Wang, H., Tan, G., Shu, D. and Zheng, J. , Dynamic fracture behavior of layered alumina ceramics characterized by a split Hopkinson bar, Materials Letters, 59: 901-904, 2005. [213] He, Z., Ma, J. , Wang, H., Tan, G., Shu, D. and Zheng, J . , Static and Dynamic Fracture Behavior of Layered Alumina Ceramics, Ceramic Engineering and Science Proceedings, 26: 251-255, 2005. 150 APPENDIX I: Engineering Drawings Drawings of the alumina disc mold Figure A1-1: Alumina disc mold die Figure A1-2: Alumina disc mold press top 151 Figure A1-3: Alumina disc mold press bottom 152 Drawings of the Housing 153 Drawings of parts for mechanical tests Figure A1-8: Housing holder APPENDIX II: Animal Care Forms fa rAaRiWsr i - , . - • -APPLICATION TO USE ANIMALS FOR RESEARCH Thet&e of ankndssor re&earctt Isa prtvaege S€*»* a jKSccof iousear t i r ias t i a research pnfsd Uap pra*2d.»» STPJOOWtrauit«>ow tfat i t s ase <S3HlTC3t5 &p£325d,ItSS B3= pQj=Ci I S HISS!, 1 M t i l ! tf» procedures 53 WKOlSEE antral* tie SWi B2 C2fft9rJ O M T h e acterBAr;rcsraraqiBamanl*lo De mil i* U H ainefng agerxy ccreiiEls f»?rre^ew. TnescKnl£cRiarilrec|Eirenemt<rKtart^£indsd artohssy sponsored research vma be conducted ey tfte anmi care c w w m t far Mature calsaamss A & 8. and By ssecta) peer rata* comutston **tmasMe categories C . D S E . faHawtts guidelines tscosifstfeirig Wsrarm. H a a s s s u t a i ' f i f e j f e f s f g e e d o r t g l h i p t a ISptoScaftes) of the complete app0c3Toncortsrtng ataXacnmer&lalhsCceTtmi ls oo Mta ta lCare .OsaGea lS^s tc t i Services andAotii&f&raian, 1Q2,SJ9aAgrar3as>f.Rd2d,., 3 C . .VST. .iza. TUs ot Project Be re topmas t a f a Sabder tna l S F tag f o r Ste l ler Sea l i o n s Mpx B J M M H * QnUFKHct BlOtmamtaion Botur.-(If renewal, provide pes$a>£ protocol ntg iAaaod 33 a * a a»p-ogress reparSfj ccACCa^c^«imasMei«is^ercrto€Uda«RS): B J A | | B H , C a s m* B S ftrtrapa iwessgasor Sumsm»: p S p 5 GWaiNanvafs): Kjgaa8 Acadamfc SanSc . t : : n a t i ?iof»: .3r U3C facuy/ o=faEMaa: J^ t ' i j a i ca l ' and B i o l o g i c a l Enfra-eeriiig CsSoe Address: MacMOlan 2U 23*7 M a i n M a l l Lan Address (ifOSarenn: S3 Tet^aneNw»isrfl3s^):822'447» Tsf-aanonaUansiBrjLaEt, J f a x » O T ^ § § j i § § § &^aJA**=sis:rp»s3gcwjLtesc .ca •eaarason: i .s ie unaersgrea, asst resra SJ annas-used in irti p a p a s * wtB Se cared £ r fei accordance wSi Sis prtnetotes prdrmlgasd Be C anaotaci Coonca on Animal Care and t i t Urt tenty of BrSsh CoRnttJa. I agree to emargsiqr veSarenary o r e Of 5ft= U r t M S f / V M M 0 M i s i r M r « i » « M M C i arpas erRMEK. fcwss.atpsopis tang arsrsas urWar ir,yoweroniwttttrraniatam* jjprapTjtt I S M H M K ana rave read std agreed io oomjfjt wttB M S praaooL DectaraKn: i . t ie roderst-grseo, assure «B> EieselacraHsmasiCCAC s&id3KS and nm be avataae tar t ie procedures dssarlBed tee-Mae: Ffts> s is Oepar*rsert Mead IS 8ie Pr»3|;al Kweaegaior, Bis sgnaltre a! Ifte o san is raqared. U8C Ospsfteert Head p e q u t e d for All -Satsi fetes): MEidjBl imjsstgalar: SEgjiGsxe Hams: Dr.2Kifia"sSSS B3S; Base: A n a t e s w r . j ^ Mams Maram* etssaareii urtt, FisBsrtss c e * s „ Room i s . » J 8-s, Head. FfehartesCsnSs Sweat ..%ns»sJsr Rstvm tmtpjetisd agpsktsiam io Research Ssruses 1JS U B C C c s ^ a ' t cs As^sai Cafe Scco. 102. 6S9D A^ax^'jH^4 -V^£ics^,3C 5'6T IZi 156 IV. Artsus sttmudtafl Attentate s to anmamBa. n w stares-vistaae «&s * • » nature mwe SMncsMtMW? vmsrware cay91 nwe Jar rsisaiiig wan* a qptctfc *MBalMf Mast asst.w»ww»o et statsd ana jusse-3 appro pussy, wate iwssi a reses/w ay sews can M aOTareitusi/^ raasoHay •sss'.as*nw-vmttmows 1=..?-mstamBz#vcmpsv} cr«rawuocwjasmcatw$.ssMwOjiSr«MHKrcJHB#8K a n There a t BO alternatives to lw* aninml feting for this type of research. Smiitahan of the type of hostile biological environment where the implant will be placed cannot be reproduced accurately in a laboratory setting. Use use of in Tiiro models would produce misleading and maccuraie results. oupilcstn>n: o o ? ; «t*nagr araa parasole»stKyaujaaaprswjsmpctiMnaJ K j * * , a p B s w a s opewwtsescessary. fl m m n Ho, in* subdermal implant, which is in development is tailored to ihaStellar sea Bees » A it has not been implanted, iiai any animal. There fe no diipScation of previous research. A S T * yiMor { S-3yrcs (e.g.! L-xsCDa Rabbit 24 nl Cv-Ce' tie Ar.' Case C=i ire c 1 '.I ..... ' i B — intmaij arsiooe wrt , g«K 1*12 RHH mmss aiti « t agency p m c n g t a c S?a a_*jft as as j r s x s i e . M eeac* aspect» ana c « m wt>« MNMMst iw* M M esoaMws? Rabbits have been chosen as many other research groups around the world hare used rabbits for these sorts of experiments Correlation with their results E important to infer conclusionsas to the tissue response of the test materials. Figs were considered, however there is no advantage to using pigs and it is much, easier to care for smaller animate. The Association for the Advancement of Medical Instrumentation (AASH) also recommends the use of rabbits for these types of expeiimenti-1 for e ^ s r i i i s n a An wfi&i smmsis <SQ,ISS S U B irsm cem-msx sn&i! an 35s oxnsgs *are oaamaosd, spatting vwwtnesenunsofjs arairasasa. ! f i J say »!= woac= S3 rem'Ba's*! se uses aacaS>3 B-3a' Sis *>|oo«pwiocjuu«!«n>3tt»»v »*HS?f ? Til;aasifiotSRS^dojisarotf *«to«u?ijnaaa. Twenty-four rabbits are needed for fee proposed experiments. In each rabbit, 3 small test samples will be implanted along with a control sample Each implant will ran Sans a different surface top-ographk size §. e. p«re s&e) with the control being a smooth surface. Three samples of each implant and control are needed for statistical analysis;.. Position in the rabbit is important and results could vary with location. Thui &e four different tmplanfe will have to be tested m each of the different locations. Therefore 12: rabbits wilt be needed for each set of implants. In addition to USE, too different materials mill be PTanmied "Tich flsat the final total required is 24. 158 il. O S S O S B arocedurss BwaMnq artrr»3ts Was stespe Bsgosae, ana da not exssrpl pages »ST» gram appcaskMis. OGQE anaad 02 g wan as to wrs i a s rappsiv.o t ie anma irom ssan-ioiiwsn, M M n g M w M n N n w B M m o f l B a n d . iatflfy^ovApaffbnnca^aKKaawt. s a d i t a i r M y 3reqsa3£2dioaaio. Kapartinema drugs or cm M s w a t t v s t a . g u a a c SWDRCMR *SK&. AM BisrvSa! surgerymusst aoMu i i ngasep i f e t ecM l i pMa . AnacR aaaassiai pasj-es Ifnecessary. Implantation mill be perforated by Dr. Tamam ^^.Jjj&usiBg gjflfJIuKKB anaesthesia.. Th« area of implantation, fee back of the anintaL will be shaved, preppedaBd draped. IncfcioES must be imde m.rangls iaefuH thickness of theikhs so that subcutaneous implantation can be obtained. Subcutaneous packets mil be made through blinf dissection with iciisen and implants will be inserted into the pockets. Each animal will receive 4, 15 cm square, implants Kith a thickness of 3mm. Tee subcutaneous pockets will then be dosed with 3-0 nylon sutures. Before waking, the animals Trill be grren a dose of aaffciefta andbn4»fflbjB,oJi for pain relief. For the postsp erative monitoring,, animate mil be monitored for infection and wiB be given antibiotics to reduce any possible infection. Stomtarteg wilt be performed twice daily. After 12 weeks, animals will be sacrificed after a lethal dose of general anaesthesia aad implants wilt be retrieved by pptoWWBi o f feme blocks airronndmg ineiBtplauts. Half of the tissue from each sample wfflbe removed by apeel test apparatus to determine adhesive bonding strength. Tbe other half of each san$)k will then be feed in a 40/O buffered formaMehyde for 4 hours, then dehydrated is a graded series of etaaiiol and embedded completely in paraffin. Sections out be cut uifflg a microtome to obtain cross sections of the tissue blocks for histological analysis. Ptease sacra wrfca oftlie SSi:»6>_ CCAC Keywords aca><: nu w n * _ x attMtar QMmatoe B U B M O T uaatamm twotcmommmasm stagaa s a w n Bcaoiaer* Pamsrf cm C u w i Tissue OOBscton assdsig Tra^ngfflfasnjj i i awsg/Taggtng tnjes&m Sood Samps:! a s s i l S a r a a Cs#S«BWl Pfts'ScaSReiSaSrt sifecSm lrouc»on ft»a> gody RaSaftxi P W s t e i arftaaaaa FoodOe|r t raSM____Wa5soq*f ! (aSon SpecialMet Altares SMKUsraeol Exposure ttaSmotofi AtonsSraSsfi Chesses S tpxure M I S H N B Ags-sts Iromanageiic or liSBnuraSaiy Ag eats .-JWSTO C o r e d * * ftiJ»3S tfajsr Sagoy M m SMgay _jt .SJ t^fjjajS Sugary SflWhal Surgery Mutate surgeries post praseauremoiicnag: w n a i s f tesea tMKytCjMaMa sfflsesjwaafcaurawnjnurcoer or days, ana ?s?pe isjseasc pa?arrs=3"s org.f?SP3S-#?S?;5, manm^awcrviisR # ^ a a c ^ £ e . m ^ 3 * 5 g e t M i m b M H pertxroed. t tpostarosetkn m o m « n a » « t t M e « i ^ r « N t t a a « , i i i i st tx ld&estaadand e s p a t e d a s (lesassaiy.. 3. »5t 24 SHUTS: T U B M M «M sue assewed or . Tamara tJ&djSgfc carfcat vaennaian a r sssss-resaled sjsisptonss. fiaqoagcf n a s= «ar w a y 2t»w«t!tte s « t d f l - / saa San. a. wconS24 raur^: E 3 3 5 w . T s a « • te a s s a e a w K t c fHsaEar : fasti ar(T»3l « • te chKXm tisfce {Sa%> c a s s i a ! (sscs as sroour ase a sn>s fa»«S»g « H t a « a | . D s t t K o T i M a n i m a l I m o i a a acocptaH* sratf-pelret The ejgjeriment reaches its end-point at 12 weeks after implantation. ossGKoe r . a ffiorsaagje se K » = = S w a «ci r s K s : 1.1 zjuarasca j o a n d a o n « ne a a r ! 3 . s x - y s n o3ir , .a = S O H O ; in e e ^ a t w i K f . .'rife:.!'-', n u i_ i . n l n« 55 |S 1 ^S m bemorveKJKBJ_S_ i i:>i.isdi?it e v i l . i i j i . i l 159 VLCO»rr5NT )0US ISSUES Tins lonoissng tyses « sqoarrasTs ategeiwaiyccrss^KJio oa a! Jconareecsn nut . ?Hi»*»KSca*Ksi»|f aii>r5r«retK!re apsy »yours*»j>. NdKinaw^dsasaiaiiarapclntlsflafparrMiBd. 1. wrrilca! restrain! (mauidsrarae^eaa}. (tryes.gfceduaocn} T « i does SOT Indod: JISOR f a t s o ; s u a a ; taifej ssood samj&s 2. Food an aw waiar depttam. gryct , g t e duaswi) 3. Extant*vartrtoreIJi •MkORrwRR peat, c<M,Igtit, n*se, taRMy.e&.}. 4. Sicgfcaljxocsdueswlere ajjpraprtata a » wattes { B M O j ) a ^ « a f l a % « l a l t c « a ^ u u t t & 5. Us * of reurarmKOlar ttccnng agsas. a. Ssdro^rtodc or nsgatte MKarcuR t f l i 7. a r e s , asisrsavarasraunia. 8. Ppeda Wj tey a SgMog sxpslBssnla. a tescsjalfc pan £n c a r * * * . * artmSs. tQ. S r « « U o r I m a t f r a p r o c e d u r e s » « i r e K w a y 11. Exercise to exrausaon stidss. 12. Mt tSSasi rasKs. . 11 ^ e r a S i t B of sensory systems aSeti cause changes ia me. BthRfcur cs ws ls -a of l i e asf mats. 14. LettialraaaSdn. 15. Fisunsrs eansj*ae a<j!MaHL 15. rraurxf 5 t fcon^ ia? a i ^ u v a m 17. T ran tgMark rocM« la r *na ) t &^ited.{Byes.canisf8etaquesllai i G S S K S J H ¥55 • ' Y2» H * • Y=S s » S ^ s • Y e ; • r a ; Hi wo • YSS Hi* 0 • Yes Bj«o BY** US Mo • . VrS • YS5 • ;y35 • , Y s » 18**° • j Y e s Hi*** O .Ye; @!*» tryauanswered 'yes 53 any « r t JOO« is r rs (son. wig iSana is j . prowia: A} Asstcatoalcrarecontoom wag BRpouft B} ftsdsadetaSsregardlEK!siesecandraons, ajg. daatoB cttattraM, ttaiaw^ art W t e ^ o t a l a a w i i w a g , sfe; C} A B a n t w m w i a t W a n j a o a t w * 8 » c t t w i « « aflknattart <f aa? joMBttewnpKaJw*; D) Ttaraeanret a O M t o a M M B a a r ••Bantu any pain, dSooman, or stess andtne aMo«i«*cI i»«setatentr i r t»goe> Deyono sial predteied ey fx txjssWsSifii assign: E) T ie nairs&s of animals that « • 6= suspected *D 8ie consfSans BidScated; r ) T»a pom s w i f t * antes! sassssig Miss i f tsnmisao j -sutBr tzad; G} Sfsakgsrtcsir*«x*3<aiananas are sang used, titanlcQMtdtidt ajajGK. •se 160 Appendix III: Histology Report RABBIT SKIN SAMPLES HISTOPATHOLOGIC FINDINGS PREPARED FOR: DR PETRELL/BRIAN HORI CHEMICAL AND BIOLOGICAL ENGINEERING UNIVERSITY OF BRITISH COLUMBIA BY: ANIMAL PATHOLOGY SERVICES APS LTD. 18208 ELLERSLIE ROAD, EDMONTON, ALBERTA T6W 1A5 Draft Report #2 APRIL 13, 2006 162 SUMMARY: A s e r i e s o f s u b c u t a n e o u s i m p l a n t s o b t a i n e d f r o m t w e l v e r a b b i t s , e a c h w i t h a c o n t r o l s a m p l e from t h e s a m e a n i m a l , w e r e e x a m i n e d g r o s s l y a n d m i c r o s c o p i c a l l y . A t h i n w a l l e d p o u c h o f f i b r o u s c o n n e c t i v e t i s s u e h a d f o r m e d a r o u n d e a c h i m p l a n t . G r o s s l y , n o i n f l a m m a t i o n w a s e v i d e n t , b u t m i c r o s c o p i c a l l y t h e r e w a s a m i l d m u l t i f o c a l f o r e i g n b o d y r e a c t i o n n o t t o t h e i m p l a n t s t h e m s e l v e s , b u t t o a d a r k c r y s t a l l i n e m a t e r i a l t h a t w a s c o m i n g from t h e s u r f a c e o f t h e i m p l a n t s . T h e a d h e s i o n o f s u b c u t a n e o u s c o n n e c t i v e t i s s u e t o t h e s u r f a c e o f a f e w i m p l a n t s s t r o n g l y s u g g e s t e d t h a t t h e c o n n e c t i v e t i s s u e w a s g r o w i n g i n t o t h e p o r o u s , b u t n o t t h e s m o o t h , s u r f a c e o f s o m e i m p l a n t s . T h e h i s t o p a t h o l o g i c s t u d y r e p o r t e d h e r e i n w a s c o n d u c t e d b y m y s e l f . T e c h n i c a l p r o c e d u r e s o n f i x e d t i s s u e s w e r e c o n d u c t e d b y a r e g i s t e r e d l a b o r a t o r y t e c h n o l o g i s t . P. N. Nation, DVM, MVSc, PhD, Diplomate ACVP. Date per Animal Pathology Services APS Ltd. 163 PROCEDURE: A series of skin samples obtained from a total of twelve rabbits were received fixed in formalin from UBC. There were two sections of skin per rabbit: one with a implant in the subcutaneous connective tissue, and one control. Fixed skin samples were examined grossly, the implants were removed, and cross sections were made of the skin through the area in which the implant was located, or, in control samples, through the middle of the submitted tissue sample. All trimmed samples were placed in tissue cassettes, processed into paraffin blocks, sectioned at 5 microns, stained with hematoxylin and eosin stain and mounted on glass slides using standard techniques. Slides were examined by a board certified veterinary pathologist and each section of tissue was either recorded as normal, or a description was made of any abnormalities. The thickness of the pouch wall was measured at four representative sites, using a micrometer eyepiece, and the average thickness recorded in a tabular format. The amount of pigment was assessed on a five point scale and also recorded for each sample in a tabular form. RESULTS: A: GROSS TISSUE EXAMINATION Gross examination of the tissue samples revealed similar findings in each. All control samples were grossly normal pieces of skin and subcutaneous connective tissue, many with the cutaneous trunci muscle attached. Each of the implants was located in a well vascularised pouch of connective tissue. The connective tissue and supportive blood vessels were tightly adherent to the surface of the implants. All implants has a hard, smooth disc surface on one side, and a chalky, porous surface on the other. In most implants, it appeared that there was some minor erosion of the porous disc surface, but not of the hard surface. There was no gross evidence of inflammation or a foreign body reaction in any of the samples. B: HISTOPATHOLOGIC EXAMINATION Microscopically, the control sections were all anatomically normal. Microscopic findings from the implant sites were all similar, but there was some variation in the degree of the findings. There was a well defined wall of mature fibrous connective tissue around each implant, forming a collagenous capsule for the implant. This wall generally varied between five and ten fibrocytes in thickness, and occasionally was as much as fifteen 164 cells thick_There was a black amorphous granular crystalline material in the collagenous interstitium between fibrocytes. Surrounding the fibrous capsule there were irregular clusters of macrophages and foreign body giant cells which contained intracytoplasmic clusters of the same granular crystalline material. Occasionally, in the centre of the largest accumulations of such cells, there was crystalline material free in the interstitium. It appeared that in these areas the cells originally phagocytosing the material may have broken down, releasing it. T A B L E I OBSERVATIONS ON POUCH WALL RABBIT ID POUCH WALL ATTACHMENT TO AMOUNT OF THICKNESS 1 IMPLANT CRYSTALLINE DEBRIS 1 0.10 None 4 2 0.14 None 2 3 0.07 Closely attached 5 4 0.02 Very close attachment 2 5 0.08 Very close attachment 4 6 0.07 Attachment 4 7 0.05 Slight attachment 4 8 0.08 None • 4 9 0.02 None 3 10 0.07 Slight attachment 3 11 0.08 None 2 12 0.09 Slight attachment 2 KEY TO TABLE I CRYSTALLINE DEBRIS 0 = No debris 1 = less than one crystal per 200X field 2 = crystalline material arranged linearly but not continuously adjacent to the wall of the pouch 3 = continuous crystalline material on or adjacent to the pouch wall 4 = crystalline material in focal masses of macrophages or giant cells scattered in the subcutis 5 = large, continuous masses of cells and crystals 1 Average thickness in mm, of four measurements at points typical of the fibrous lining of the wall 165 DISCUSSION: I n t h i s s t u d y , t h e i m p l a n t s d i d n o t h a v e a n y d e l e t e r i o u s e f f e c t u p o n t h e s k i n , a n d i n f a c t , g r o s s e x a m i n a t i o n s u g g e s t e d t h a t t h e y h a d n o e f f e c t . M i c r o s c o p i c a l l y , e a c h i m p l a n t w a s l o c a t e d w i t h i n a w e l l d e f i n e d p o u c h o f m a t u r e f i b r o u s c o n n e c t i v e t i s s u e . T h e o n l y i n f l a m m a t o r y c h a n g e s w e r e m i n o r , a n d c o n s i s t e d o f m a c r o p h a g e s a n d g i a n t c e l l s w i t h p h a g o c y t o s e d c r y s t a l l i n e g r a n u l a r m a t e r i a l i n t h e c y t o p l a s m . T h i s m a t e r i a l i s o r i g i n a t i n g f r o m t h e s u r f a c e o f t h e i m p l a n t s a n d a p p e a r s t o o r i g i n a t e b y d e g e n e r a t i o n o f t h e s u r f a c e m a t e r i a l w h i c h i s b e i n g p i c k e d u p a n d r e m o v e d b y t h e p h a g o c y t i c s y s t e m . T h e c r y s t a l l i n e m a t e r i a l i s o n l y f o u n d t o b e c o m i n g f r o m t h e p o r o u s s i d e o f t h e i m p l a n t s , n o t t h e s o l i d s i d e . T h i s f i n d i n g s u g g e s t s t h a t t h e i m p l a n t s m i g h t c o n t i n u e t o b r e a k d o w n w i t h t i m e , a n d t h u s t h a t t h e r e m a y b e a t i m e l i m i t t o t h e i r u s e f u l n e s s . T h e r e a c t i o n t h a t i s p r e s e n t i s m i l d , a n d w o u l d n o t b e e x p e c t e d t o c a u s e a n y d i s c o m f o r t . 166 Figure A3-1:Cross section view of the tissue around the implants. The implant pouch is at the bottom of the image as indicated. Magnification is 20x. Figure A3-2: Cross section view of a control tissue block. No inflammation can be observed. Magnification is 20x. 167 Figure A3-3: Crystalline debris at 200x magnification. Some debris is in the cytoplasm of the macrophage while some is outside. 168 APPENDIX IV: Tissue Attachment Loading Curves 140 0 1 2 3 4 5 Time (sec) Figure A4-1: Tissue Attachment Force for 0.5mm pore layer implants 140 0 1 2 3 4 5 Time (sec) Figure A4-2: Tissue Attachment Force for 1mm pore layer implants 169 CD O 140 120 100 80 60 40 20 0 R3 (1, pd) R4 (0.5, pu) R5 (0.5, pu) R6 (0.5, pu) 1 v / LV A R10(1,pd) \ i 2 3 Time (sec) Figure A4-4: Tissue Attachment Force for Back Left Location implants 170 140 120 100 80 60 40 20 0 0 1 R2 (1, pu) R4(1, pu) R7(1, pu) R8 (0.5, pd) 4 Time (sec) Figure A4-5: Tissue Attachment Force for Front Right Location implants 171 CD o •R9(1, pu) R10 (0.5, pu) R11 (1,pu) R4 (0.5, pu) R5 (0.5, pu) R6 (0.5, pu) R2 (1, pu) •R4(1, pu) R7(1, pu) R1 (1, pu) R2 (0.5, pu) R5(1, pu) R8(1, pu) R11 (0.5, pu) 0 1 2 3 4 Time (sec) Figure A4-7: T issue Attachment Force for Pore Up implants CD O 120 100 80 60 40 20 0 R1 (0.5, pd) R3 (0.5, pd) R7 (0.5, pd) R3(1, pd) I R10(1,pd) R8(0.5, pd) A R6(1, pd) I i / \ //A //A / \. T \ \ \\ V L i L 1 V, 1 0 1 2 3 4 Time (sec) Figure A4-8: T issue Attachment Force for Pore Down implants 172 APPENDIX V: Housing Fracture Pictures Compression Test Specimens Specimen #1 Figure A5-2: Compression Fracture Picture 2 173 S p e c i m e n #2 Figure A5-3: Compress ion Fracture Picture 3 Figure A5-4: Compress ion Fracture Picture 4 174 Specimen #3 Figure A5-5: Compression Fracture Picture 5 Figure A5-6: Compression Fracture Picture 6 175 Specimen #4 176 Specimen #5 Figure A5-9: Compression Fracture Picture 9 Figure A5-10: Compression Fracture Picture 10 177 Specimen #6 Figure A5-11: Compression Fracture Picture 11 Figure A5-12: Compression Fracture Picture 12 178 Impact Test Fractures Specimen #1 Figure A5-13: Impact Fracture Picture 1 Figure A5-14: Impact Fracture Picture 2 179 Specimen #2 Figure A5-15: Impact Fracture Picture 3 Specimen #8 Figure A5-16: Impact Fracture Picture 4 180 Specimen #11 Figure A5-17: Impact Fracture Picture 5 Figure A5-18: Impact Fracture Picture 6 181 Puncture Test Fractures Specimen #2 Figure A5-19: Puncture Fracture Picture 1 Figure A5-20: Puncture Fracture Picture 2 182 Figure A5-21: Puncture Fracture Picture 3 Figure A5-22: Puncture Fracture Picture 4 183 Figure A5-25: Puncture Fracture Picture 7 Figure A5-26: Puncture Fracture Picture 8 185 Specimen #5 • Figure A5-29: Puncture Fracture Picture 11 Specimen #6 Figure A5-30: Puncture Fracture Picture 12 187 Figure A5-31: Puncture Fracture Picture 13 188 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items