<%BANNER%>

Foramen is Tougher than drilled hole in equine third metacarpus

University of Florida Institutional Repository

PAGE 1

FORAMEN IS TOUGHER THAN DRILLED HOLE IN EQUINE THIRD METACARPUS By BARBARA GARITA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2002

PAGE 2

Copyright 2002 by Barbara Garita

PAGE 3

I dedicate this work to my family, Donato, Noemi, Esteban, Pablo, Ifigenia, Ignacio, Natalia and my grandmother Margarita.

PAGE 4

ACKNOWLEDGMENTS I would like to acknowledge that this work would have been much more difficult without the advice, company, commitment, and friendliness of the people that coexist with me everyday. All of them in one way or the other helped this work enormously, probably much more than they know or I can express. I would particularly like to let Dr. Rapoff know that his counseling, patience and support were crucial, obviously in the completion of my thesis, but also in my out of school education. Similarly, I would like to thanks Wes Johnson for sharing so many ideas, for coaching me through graduate school, and simply for caring for me. I feel that Dr. Satchi Venkatarama got me closer to the reality of graduate school, and offered immense strengthening and understanding, and Stephanie Buskirk offered valuable help and company. Also, I would like to thank to Dr. Hafka and Dr. Jing Huang, Olivier Fontanel, Renaud Rinaldi, Ron Brown, Matt Gabriel and Nicole Gasparina. Outside school many friends significantly participated in my research by showing interest in my work like Natalia P. and Ryanne C. and my family in Costa Rica. Finally I must acknowledge the energy, tools and means behind my research: the charm of nature and the methods of science, the available technology and the monetary sources made available by the Biomedical Engineering Department, NASA and Aerochem. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.........................................................................................................................x CHAPTER 1 INTRODUCTION...........................................................................................................1 2 METHODS AND MATERIALS.....................................................................................2 Anatomy...........................................................................................................................2 Loading Environment and Fracture Incidence.................................................................3 Bone Procurement and Identification..............................................................................5 Beam Dimensions............................................................................................................5 Machining of Beams........................................................................................................7 Drilling Holes...................................................................................................................8 Mechanical Testing Setup..............................................................................................11 Mechanical Testing Plan................................................................................................13 Monotonic Tests...............................................................................................17 Fatigue Tests....................................................................................................17 Staining..........................................................................................................................18 Sectioning of Beams for Microscopy............................................................................18 Microscopy Section Analyses Setup..............................................................................20 Morphometric Parameters..............................................................................................21 Damage Parameters.......................................................................................................23 Specific Examples of Damage Types and Quantification.............................................24 Statistical Analyses........................................................................................................32 3 RESULTS......................................................................................................................35 Morphometric Parameters..............................................................................................38 Damage Parameters.......................................................................................................39 4 DISCUSSION................................................................................................................51 v

PAGE 6

5 OSTEON TRAJECTORIES NEAR THE EQUINE METACARPUS NUTRIENT FORAMEN...................................................................................................................56 Introduction....................................................................................................................56 Materials and Methods...................................................................................................56 Results............................................................................................................................57 Discussion......................................................................................................................58 6 CONCLUSION..............................................................................................................61 APPENDIX: QUANTIFICATION PROTOCOL..............................................................62 LIST OF REFERENCES...................................................................................................65 BIOGRAPHICAL SKETCH.............................................................................................69 vi

PAGE 7

LIST OF TABLES Table page 2-1: Testing plan. .............................................................................................................17 2-2: Staining procedure......................................................................................................19 2-3: Crack types and descriptions......................................................................................25 3-1: Results for beam specimen tested monotonically to failure.......................................36 3-2: Results for the remainder of the beams......................................................................37 3-3: Results for porosity and spatial locations...................................................................40 3-4: Results for osteon density and spatial locations.........................................................40 3-5: Results for average osteon diameter and spatial locations.........................................41 3-6: Results for density of BM and spatial location..........................................................43 3-7: Results on crack length per area for black wispy mesocracks (BWM)......................46 3-8: Result on crack length per area for diffused stained mesocracks (DSM)..................47 3-9: Results for intra-osteonal microcracks (IOM)............................................................48 3-10: Descriptive statistics for cement line cracks............................................................49 3-11: Angles summary.......................................................................................................50 vii

PAGE 8

LIST OF FIGURES Figure page 2-1: Skeleton of a horse.......................................................................................................4 2-2: The arrow points to the location of the nutrient foramen (right)..................................4 2-3: Pictures identifies between right and left cannon bones. ............................................6 2-4: Finite element model of the beam with the holes. .......................................................7 2-5: An equine MC3 (upper left), dashed lines indicate proximal and distal cutting locations. Remnant proximal and distal cuts (top right)..........................................9 2-6: Mediolateral cuts being made with a diamond blade band saw (left)........................10 2-7: Drilling of the holes with nitrogen powered drill. ....................................................10 2-8: The toothpick with 320 Silicon Carbide powder adhesively bonded. ......................11 2-9: Oblique view of free body diagram for a beam in four point bending (A)...............14 2-10: Actual tested beam. Locations where the Teflon tape has worn out down by the friction of the loading noses during cyclic testing (A). .......................................15 2-11: Testing setup (top)....................................................................................................16 2-12: Stained bone beam....................................................................................................19 2-13: Plan view of bone beam, shown with holes near each end as they exist in the actual beam. ....................................................................................................................19 2-14: Three sections quantified for compositional and microcrack parameters................23 2-15: The three figures show the M-image. .....................................................................26 2-16: Example of the damage type: bundle of microcracks (BM).....................................27 2-17: Example of black wispy mesocracks (BWM), denoted by the arrows. .................28 2-18: Example of diffused stained mesocracks (DSM), indicated by the arrows. ...........29 viii

PAGE 9

2-19: Example of an intra-osteonal crack (IOM), specifically a case of a stained wedge. 30 2-20: Example of an intra-osteonal crack (IOM), specifically a case of lamellar debonding. ............................................................................................................31 2-21: Example of a cement line crack (CLC), it is the black wispy line surrounding the osteon. ..................................................................................................................32 3-1: Peak strain level versus number of cycles to failure (S-N curve)...........................36 3-2: A groove present on an 8,000 specimen (left). A permanently deformed beam (right).....................................................................................................................37 3-3: Graph for porosity and spatial location. ...................................................................40 3-4: Graph for osteon density and spatial location............................................................41 3-6: Graph of density of BM and hole type. ....................................................................43 3-7: Graph of density of BM and proximity to the hole. .................................................44 3-8: Number of BM split by proximity to the hole and hole type.....................................44 3-9: Graph for crack length for black wispy mesocracks (BWM) and hole type..............45 3-10: Graph for diffused stained mesocracks (DSM) hole type and proximity of the hole...................................................................................................................46 3-11: Interaction plot for intra-osteonal microcracks and the effects of proximity to the hole.........................................................................................................................48 3-12: Graph for intra-osteonal microcracks (IOM) and the effects of hole type...............48 5-1: Palmar view of the equine metacarpus (left).............................................................59 5-2: Lateral view of the equine metacarpus (left).............................................................60 ix

PAGE 10

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FORAMEN IS TOUGHER THAN DRILLED HOLE IN EQUINE THIRD METACARPUS By Barbara Garita December 2002 Chair: Dr. Andrew Rapoff Department: Biomedical Engineering The purpose of this work was to compare the damage resistance and tolerance of bone in the vicinity of a natural hole, the nutrient foramen and a drilled hole in the equine third metacarpus. The nutrient foramen hole starts to form in early fetal stages, and the microstructure around it has been adapting to its presence since then. The indefinite adaptability is a reason why related studies have investigated the foramen and the microstructure around it. These studies have revealed that the microstructure reduces the local stress concentrations near the foramen, and that it increases structural strength. Also, elastic and strength manifestations of this microstructure were mimicked in the design and fabrication of a plate with a central hole. This plate demonstrated superior performance over a uniform plate with a hole. The present research deepens even more into the microstructure around the foramen and verifies that it also increases damage resistance and tolerance (toughness) in response to cyclic loading. Also, osteon trajectories near the foramen were uncovered; they exist in regions of transverse tension x

PAGE 11

stress in response to global loading. Their arrangement locally increases toughness, as well. xi

PAGE 12

CHAPTER 1 INTRODUCTION It is of interest to quantify and analyze the occurrence of physical damage in the form of cracks in the vicinity of a natural hole, in the equine third metacarpus (MC3). The equine MC3 is the major load bearing bone in the forelimb of the horse, and nutrient foramen is a hole that exists in its back or palmar aspect. Most holes in structures present themselves as stress concentrations, but the foramen is a region of stress reduction [22]. The microstructural variations near the foramen are responsible for making this hole a site of stress reduction [22,40]. The purpose of this work is to continue investigating the microstructure near the foramen; specifically, to evaluate its damage resistance and tolerance to cyclic loading. The following chapters will describe how the damage resistance and toughness of the microstructure in the vicinity of the foramen was compared to that of a drilled hole. Beams were procured from the palmar aspect of the equine MC3, all containing the foramen. A drilled hole was placed a distance away from the foramen. The beams were cyclically loaded at high strain levels effecting the same loading both holes. The cyclic loading induced damage around the holes, and later that damage was quantify and analyzed. Chapter 5 deals with the latest findings on osteon trajectories near the foramen. The location and arrangement of these osteonal trajectories reveal further localized adaptation and damage tolerance around the foramen. 1

PAGE 13

CHAPTER 2 METHODS AND MATERIALS The equine third metacarpus (MC3) is a captivating bone. At first glance, it is distinctly recognized because the remnants of two vestigial bones are attached to its sides. Its shape reflects evolutionary adaptation that allows for efficient and high speed ambulation [31]. Still, it has been repeatedly studied because of its incidence as a stress fracture site [6,18,35,33,34,37,38]. Like many other long bones, it has a nutrient foramen located on its palmar aspect, around a third of its length away from proximal metaphysis (Figure 2-2). The following pages intend to give details on specimen preparation: harvesting bone beams, drilling holes, staining with basic fuchsin; mechanical testing, quantifying compositional and damage parameters, and performing statistical analyses. Anatomy The third metacarpus bone (MC3) stretches from the elbow to the fetlock joint (Figure 2-1). The second and fourth metacarpals are two vestigial metacarpals attached through syndesmoses to the medial and lateral aspects of the third metacarpus; together they are known as the cannon bone. There is no trace of the first and fifth metacarpals; consequently, the horse walks only on the middle digit. The nutrient foramen is a natural hole, an intrusion through the bones cortex, on the palmar side of the third metacarpus, located approximately 60 mm away from the proximal end, and usually centered, equidistant from the lateral and medial sides of the bone. Typically it has an elliptical shape and its mayor axes measure about 3 mm and 2 2

PAGE 14

3 mm. A nutrient vessel crosses the cortex, through the foramen hole, to supply the bone with necessary nutrients. At a macroscopic level, cortical or cancellous bone appears solid. Cancellous bone is highly porous and appears at the ends of long bones. Cortical or osteonal bone, is found in the diaphyses (middle region) of long bones (like the MC3), and it is of concern in this study. The porosity of cortical bone is usually between 5% and 10%; its pores consist of canals (Haversian and Volkmanns canals) or temporary cavities (resorption spaces). Cortical bone is composed of lamellar layers, either in a parallel order (plexiform bone) or in a circumferential arrangement (osteonal bone). Osteonal bone is composed of osteons, which are concentric layers of lamellar wrap around a central canal (Haversian canal). In addition, osteons are cylindrical in shape, about 200 m in diameter and usually oriented parallel to the longitudinal axis of long bones. Osteons are surrounded by a weak interfacial layer, called the cement line, which arrests and deflects cracks (physical form of damage). Cortical bone undergoes remodeling (removal and replacement of bone), in part, to repair cracks. Loading Environment and Fracture Incidence The MC3 is one of the most common sites for stress fractures in racing horses, particularly, in young thoroughbreds [38,34,35]. Peak periosteal strains, in other animal species can reach peaks strains of 2,000 to 3,000 [34]; however, peak compressive strains of nearly 5,000 have been measured on dorsal surfaces of the MC3 during galloping [38]. Shear strains approximately measuring 44% less in magnitude were measured during trotting, indicating a torsional loading component. Still, the equine third metacarpus is primarily axially loaded in compression due to gravity and in vivo tests using strain gages suggest dorsopalmar bending [34]

PAGE 15

4 Horse Skeleton Fetlock Joint Dorsal direction Palmar direction Distal direction Proximal direction Equine third metacarpus Figure 2-1: Skeleton of a horse. The ellipse surrounds the equine third metacarpus. An expanded view elaborates on the terms describing a two dimensional orientation of the bone. In the case of a three dimensional orientation, the lateral direction will be out of the page, and the medial direction will be into the page. Figure 2-2: The arrow points to the location of the nutrient foramen (right). Location of the principal nutrient vessels, as it crosses the palmar cortex (left). Proximal direction Palmar View Distal direction Medial direction Image from [16] Lateral direction

PAGE 16

5 Bone Procurement and Identification Six pairs of metacarpi (12 bones total) were obtained from a commercial processor (Animal Technologies, Tyler, Texas). The animals were between three and five years of age who died from reasons unrelated to this work. The bones were stored at C. Every bone was handled following the guidelines of our Institutional Animal Care and Use Committee. After thawing, the limbs were cleared of the remaining soft tissue and identified as right and left cannon bones. The identification of the right and left metacarpal bones was complicated, since the providers of the limbs included the carpal bones with the third metacarpals. In order to anatomically differentiate between right and left metacarpal, the carpal joint (the joint between the carpal bones and the MC3) had to be separated. A depression on the top view of the metaphysis identified the lateral aspect [39] (Figure 2-3). Beam Dimensions When preparing the beam specimens, the ASTM Standards D 790M-82 were followed as strictly as possible. However, due to the anatomical difficulties on the palmar aspect, a support span to depth ratio of 40 to 1 was used instead of a 16 to 1 ratio, as has been used by others [35]. The ASTM Standards recommends 25 mm width beams for a depth of 2 mm [4]; however, the palmar aspect has a limited width (due to the presence of the second and third vestigial metacarpals); and only 15 mm width was possible. Considering the overhanging length, outside the supports, the beams for experimentation measured 2 x 15 x 110 mm.

PAGE 17

6 metaphysis cuts on MC3s carpal bones top view of metaphysis of MC3 carpal bones Figure 2-3: Pictures identifies between right and left cannon bones. The black arrows point the carpal bones, and the white arrows point on a depression on the metaphysis of the third metacarpal bones that identify the lateral aspects. The length of 110 mm in the bone beams was required, because of the necessity of positioning the two holes (the foramen and a drilled hole) between the loading noses. A finite element model of a beam with mechanical properties representative of the MC3 was constructed by a colleague (Dr. Jing Huang, Figure 2-4). This model was used to confirm that stress fields of the drilled hole and the foramen would not interact. The model was subjected to loading and boundary conditions representing those of the experiments. The results demonstrated that a distance of 12 mm between the hole centers was sufficient for the stress fields not to interact.

PAGE 18

7 D C A B Figure 2-4: Finite element model of the beam with the holes. The model confirms the distance between the holes is enough for the stress fields from the holes to act independently. A and B show the intricate detail of the mesh, with more nodes in the proximity of the holes. C and D show the stress fields next to the holes. In figure D the arrows point to a close up of the stress fields; evidently, there is no interaction between them (they do not touch each other). Machining of Beams Beams were machined from the palmar aspect using a combination of rough and precision guided cuts. After proper anatomical identification, a normal band saw was used to make the initial rough cuts on the cannon bones; they were cut 60 mm proximal from the foramen and 80 mm distal from the foramen. The remaining bulks were cut longitudinally, separating the palmar from the dorsal aspects of the bone (Figure 2-5). The finer longitudinal cuts (slabs) were made using a precision diamond blade band saw. The specimens were grinded and polished to the final thickness of approximately 2 mm

PAGE 19

8 to eliminate all the possible surface scratches introduced by the cutting blades. The beams were kept hydrated throughout these and all other preparations. Mediolateral cuts were made with the aid of a plexiglas fixture (Figure 2-6). The width of the beams were cut with a diamond band saw (Gryphon C-40, Sylmar, California). The same plexiglas assured the same width for all specimens (15 mm) (Figure 2-6, right). Drilling Holes The drilled hole and the control hole were machined with a nitrogen powered drill at a speed of 75 RPM. A powered drill was used to reduce wobble [36]. Slow drilling will significantly decrease the maximal bone temperature on the contact surfaces [43]. Application of a moderate axial force and drilling in as short of time as possible through the bone will reduce the damaging effects of the drilling process [3]. The plexiglas fixture also served to create a pilot hole when using the drill and guaranteed the same location for the drill holes in each specimen (Figure 2-7). The drilled holes were made as similar as possible to the foramen. After drilling the holes, the foramen was carefully measured and compared in dimensions to the drilled holes. If any major differences were present, a milling machine was used to modify the drilled holes. Finally, the drilled holes were meticulously ground and polished to corresponding elliptical measures, using a toothpick to which Silicon Carbide 320 grit powder (Buehler; Lake Buff, IL) was glued. Before any testing, the beam specimens were polished again with 600 sand papers and polishing cloths (Buehler Texmet Cloth; Lake Buff, IL) and water [20]. All beams were marked with a chamfer on the lateral and distal corner to discern the correct anatomic orientations throughout the experiments.

PAGE 20

9 Figure 2-5: An equine MC3 (upper left), dashed lines indicate proximal and distal cutting locations. Remnant proximal and distal cuts (top right). A longitudinal cut being made in the Exkact saw (Exkact; Germany) (center), notice the foramen is visible. Resulting machined sections from the six equine third metarpi (bottom left) and a close up view of one pair (bottom right).

PAGE 21

10 Figure 2-6: Mediolateral cuts being made with a diamond blade band saw (left). Close up fixture used to guide blade (right). Figure 2-7: Drilling of the holes with nitrogen powered drill. The plexiglas fixture is used to drill the pilot hole, to diminishes wobble, and locate repeatedly the site of the drilled holes.

PAGE 22

11 Fixture Drilled hole Foramen Control hole Figure 2-8: The toothpick with 320 Silicon Carbide powder adhesively bonded. Used for meticulously sculpting the drilled hole like the foramen (top left). Plexiglas fixture used to hold the beam and even forces while grinding and polishing beams manually (top right). Example of a beam as it looked before testing (bottom). Mechanical Testing Setup Two strain levels where chosen: a supraphysiological strain of 8,000 very high physiological strain of 5,000 Others [35] have tested in three point bending lateral, dorsal, and medial bone beam from the MC3 for 10,000 cycles at 5,000 and found no reduction in elastic modulus or yield strength. Others [21] have fatigue tested the same aspects of the MC3, in four point bending at 10,000 until failure. These beams endured between 300 and 21,000 cycles. Neither group tested the palmar aspect. Considering that the beams in this experiment have two holes in the loaded area, the strain levels chosen seem reasonable. Bending was accomplished by attaching a custom fixture to a servohydraulic testing machine (MTS 858 MiniBionix with a 407 controller)

PAGE 23

12 and a load cell (Entran, NJ) (Figure 2-11). Each beam was placed in the fixture so that the periosteal surface experienced compression. Various fixture for holding the beams where investigated. Initially the beams moved off the supports. Possibly the anisotropy of the bone created a nonlinear response to the stresses induced while fatigue testing. The problem was solved by restraining the beams to the support rollers using rubber bands (Figure2-10). A stainless steel four point bending fixture was designed and constructed. The fixture has a pivot axis to assure that the support noses and the loading noses remain parallel to each other when loading a specimen (Figure 2-11). Rollers where held by rubber bands, to assured movement at the contact points, and to reduce the friction forces against the beams surfaces, while cyclic tests took place (Figure 2-10). The support span was 80mm and the loading span is 40mm, as recommended by the by the ASTM 790-82 standard [4] (Figure 2-9). Studies on cyclic testing on bone specimen using four and three point bending fixtures [23] have noted that stainless steel fixed rollers can create grooves on the bone beams[24]. These grooves become the primary cause of an alteration in the stiffness of the beam. Guided by finite element results [23] the smallest diameter rollers suggested in the ASTM standards are recommended for use (3 mm diameter rollers). Considering their recommendation, 3.2 mm rollers were used in this experiment. Additionally, Teflon tape was used around the rollers as well as on the bone to diminish the wear [24]. In other cases, especially when testing bone beams at 8,000 the Teflon tape was not enough to eliminate the formation of grooves. For the higher strain levels, a sacrificial piece of equine bone from the dorsal aspect, having the same thickness and width as the bone

PAGE 24

13 beam but measuring only 10 mm in length, was glued with cyanoacrylate to the surface of the bone. The bone sacrificial piece and the beam were then covered with Teflon tape (Figure 2-9). All tests took place in a temperature controlled aqueous solution, in a cylindrical plexiglas container. Temperature plays an important role in fatigue testing of bone [7]. Fatigue life has been found to depend significantly upon temperature of bone [15]. These data suggest that a linear relationship exists between the number of cycles to failure and range of temperature. For example, the fatigue life at room temperature (22 C) is one half that at physiologic (37 C) for bone. Additionally, it seems reasonable that if any advantageous structural adaptation exists near the foramen regarding toughness, then such adaptation may give the best results at physiological temperature. All beams were tested at a physiological temperature of 37C in a calcium buffer solution to prevent leakage of calcium ions [26]; otherwise, leakage of ions can cause a reduction in modulus of nearly 2.5% [26]. A heater (Fisher Scientific Automerse Immersion Heater Model 199) was used to control the solution temperature at 37C. The testing solution consisted of 0.9% saline solution with a concentration of 57.5 mg/L of CaCl 2 [26]. The specimens were allow to thaw for 2 hours and acclimate at least 45 minutes in the bath prior to testing. Mechanical Testing Plan From the five pairs of metacarpi (12 bones) procured, a total of fourteen beams were mechanically tested. Four were dorsal specimens used for protocol development only. The remainders were palmar beams.

PAGE 25

14 D C A B a a 1/2 (L + a) 1/2 (L a) Pa Bending Moment Control Hole Drilled Hole ForamenHole Loading Span Support Span (L) P P P P -P +P Transverse Shear Figure 2-9: Oblique view of free body diagram for a beam in four point bending (A). The loading span is 80mm and the support span is 40mm (B). A transverse shear diagram shows a region of zero shear between the loading forces and a bending moment diagram shows a region of constant moment (D).

PAGE 26

15 C B A Sacrificial bone pieces Figure 2-10: Actual tested beam. Locations where the Teflon tape has worn out down by the friction of the loading noses during cyclic testing (A). Cartoon demonstrating the manner in which the beams were held by the rubber bands to the support noses (B), and the location of the sacrificial bone pieces to prevent grooves from forming. Figure B shows a top view, and figure C shows

PAGE 27

16 Figure 2-11: Testing setup (top). Custom made four point bending fixture, note the pivot point, and the way in which the rubbers bands hold the loading noses to allow for rotation (bottom).

PAGE 28

17 Table 2-1: Testing plan. Repeating letters under the horse column signify the bones come from the same horse. Four beams were procured from the palmar aspect of horse S (two from the right leg and two from the left leg.) Of importance is the specimen ID, for it will be mentioned throughout this study. Number Horse Side Specimen ID Test Cyclic Strain Level 1 S Right SRA Fatigue 5000 2 S Left SLA Fatigue 8000 3 S Right SRB Monotonic -4 S Left SLB none control 5 A Right AR Fatigue 8000 6 A Left AL Fatigue 5000 7 G Right GR Fatigue 5000 8 G Left GL Fatigue 8000 9 L Left LL Monotonic -10 L Right LR Fatigue 5000 11 I Left ILB Fatigue 8000 12 I Left ILA Lost --Monotonic Tests The monotonic tests were done to failure, in displacement control at a rate of 1 mm/sec. The elastic moduli were calculated from the load-deflection data and used to estimate the required deflections for cyclic strain levels. Fatigue Tests The fatigue tests were performed at a frequency of 2 Hz using sinusoidal waveform. The nominal cyclic strain level was calculated form the beam theory and the elastic modulus estimated from monotonic tests. Then the beams were loaded to approximately 10 N for the purpose of holding them. The test machine controller was set for the displacement corresponding to the nominal strain level desired. The data acquisition rate was 40 Hz. The data were recorded following a burst acquisition method, which saved 6 cycles every 100 cycles. Tests were terminated at 60,000 cycles (8hours and 20 minutes) unless the beam failed before that.

PAGE 29

18 Staining Basic fuchsin staining is widely used to separate induced from artefactual microcracks in bone. An established protocol [9] served as a guide to stain the loaded beams. Immediately after the specimens were tested, they were placed in 1% basic fuchsin (J.T. Baker, Basic Fuchsin) and 70% ethanol. The ethanol concentrations were increased in each successive step (Table 2-2). Each step lasted 24 hours, and at least the first 2 hours were spent under a 20 mm Hg vacuum to enhance penetration of the stain (Figure 2-12). The complete process takes 11 days. Also, it is important to note that the solution can not be reused; since the bones become dehydrated by releasing water and reducing the concentration of the ethanol, this can reduce the uptake of stain. Sectioning of Beams for Microscopy After the bones beams were stained they were sent to a commercial processing laboratory (Pathology Associates Advance, North Carolina), where eight 100 um thick sections were prepared from relevant regions. Three sections were prepared in the vicinity of both loaded holes, and two sections near or through the control hole. In this his way, sections 1, 4, and 7 were suitable for comparison (Figure 2-13). Sections 3, 5, 6, and 8 did not contain a hole, but are important when analyzing the microdamage in the proximity of these holes. All the beams were chamfered on the overhanging unloaded lateral endosteal corner, so that the commercial processing laboratory could identify the correct orientation of the specimens. Regardless of the precautions and explanations given to the laboratory, they were unable to provide with certainty the orientation of the slides. The correct orientation of the slides was determined after laborious observation of the microstructure and by looking at the reconstructed embedded beams.

PAGE 30

19 Table 2-2: Staining procedure. Figure 2-12: Stained bone beam. 3 sections at each of these holes 1. 24 hours in1% basic fuchsin in 70% ethanol 2. 24 hours in1% basic fuchsin in 80% ethanol. 3. 24 hours in 1% basic fuchsin in 90% ethanol. 4. 24 hours in 1% basic fuchsin in 100% ethanol. 5. 24hours in 1% basic fuchsin in 100% ethanol. 6. Remove beams and allow to air dry for 2 days. 7. Rehydrate in deionized water for 4 days. chamfer 125 mm 5 4 3 8 7 6 2 1 5 mm S(RA) 1 Lat Perios 1 x 2 glass microscope slide glass coverslip section Figure 2-13: Plan view of bone beam, shown with holes near each end as they exist in the actual beam. The transverse lines represent the location of desired sections (top). Bone section mounted on slide.

PAGE 31

20 Microscopy Section Analyses Setup Two of the fatigued beams were analyzed with respect to morphometric and damage parameters describe subsequently. A total of three sections per beam were analyzed: one through the foramen and one through the drilled hole, and one through the control hole. The morphometric parameters included microstructural characteristics intrinsic to the bone. The damage parameters included quantitative measures of damage. Images of the entire bone sections were taken, and put together in a sort of mosaic, denoted as an M-image. The M-image is a paper banner composed of three printouts and used as a map to locate important features and for organization. The images that comprised the M-image were taken using a brightfield microscope (Olympus BX-60 Melville, N.Y.), a digital color camera (Cohu CCD RGB), and a frame (Integral Technologies, Flashpoint128). The microscope was used with a 4X objective, due to the magnification of the camera the total magnification was 142.5X. Using a square grid as a guide, images of each bone section were taken every 1 mm. A complete M-image was comprised of 30 images, or 2 rows of 15 images. These images were cropped to square millimeters and put together using Matlab. The M-image was used extensively: morphometric parameters were measured with it and it served as a fixed frame of reference where the location of every single crack was noted. Initially, each section was divided in the mediolateral direction into 10 regions; The 10 regions were identified considering their proximity to the hole. (medial D, medial C, medial B, medial A, medial hole, lateral hole, lateral A, lateral B, lateral C, and lateral D). Each region measured 1 mm in length and the height of the beam (approximately 2mm), except for the two regions next to the holes (the near regions) which measure 0.50 mm in length (Figure 2-15). The reason why the near regions are shorter in length is

PAGE 32

21 because next to the foramen a more involved microstructure expected and subsequently noted (Figure 2-15). The 10 regions were aggregated into far (medial B, C, D and lateral B, C, D) and near (medial hole, medial A, lateral hole, and lateral A) regions based on preliminary statistical comparisons. Each section was divided in the endoperiosteal direction (through the thickness) into regions according to the mode of strain experienced by the beam; tension, compression or central (low tension and low compression). The tensile and compressive regions measured 0.5 mm in thickness each, and the central region measured 1 mm in thickness. Morphometric Parameters Morphometric parameters are intrinsic to the bone and should reflect adaptation to its loading environment in vivo. The morphometric parameters quantified on the M-images were porosity and osteon density and osteon diameters Detailed explanations on the method for acquiring the images and eventually their quantification can be found in the Appendix; herein a brief overview of the process is provided. Transparencies, on which a 20 m x 20 m grid was printed, were taped on top of the M-images to count and mark the parameters. The M-images were scanned for the number of Haversian canals, the length Volkmanns canals, and the area of resorption spaces. All these values were put into the following expression to determine porosity. area of voids% Porosity = total area

PAGE 33

22 22250m# of Haversians canals 2+ length of Volkmann's Canals (m) 50m+ area of resorption spaces (m) =total area (m) It was assumed that the diameter of all Haversian canals and the Volkmanns canals were 50 m wide. The areas of the resorption spaces were counted, explicitly if they were larger than 0.250 m 2 The osteon density was calculated form the number of Haversian canals in a region. Finally, the osteon diameters were the last of the morphological parameters quantified. Using Matlab, scan lines were placed on the M-images in the middle of each region, in the endoperiosteal direction (Figure 2-14). Using the brightfield microscope and 10X objective images where successively taken, following down the scan lines. These images where not put together or cropped, they were simply printed out and osteons were measured on them with a ruler (following some scaling). Not all osteon were measurable, only those that followed a specified criterion [33] described below. The reason for this is that the cement line must be clearly visible to measure osteon diameters. The criteria for choosing an osteon and measuring its diameter: 1. secondary or primary osteons with presence of a clear cement line 2. circular or elliptical shape 3. refilling completed or almost completed 4. all osteons were measured through the Haversian canal in perpendicular and transverse directions only

PAGE 34

23 5. two diameters were measured if their shape was elliptical Mediolateral or (proximity to the hole ) Perioendosteal ( com p ression-tension ) Proximodistal or (hole type) Control Section Drilled section Foramen section Figure 2-14: Three sections quantified for compositional and microcrack parameters. Damage Parameters One group [48] defined micro cracks damage in three sizes: microcracks, mesocracks and macrocracks. Microcracks are smaller than 50 m, mesocracks are between 50 m and 200 m, and macrocracks are larger than 200 m. The same group noted that the magnification used to observed or quantify damage is a defining factor, because the origin of the damage could be as small as 3m long cracks. Before settling on a method for quantifying cracks, the bone sections are observed and different types of microcracks were inspected, and a protocol was prepared that included the relevant and repeating damage characteristics. In the six sections observed, there were five types of repeatedly, identifiable cracks. These were bundles of microcracks (BM), black wispy mesocracks (BWM), diffused stained mesocracks (DSM), intra-osteonal microcracks (IOM), and cracks along

PAGE 35

24 cement lines (CLC) (Table 2-3, Figures 2-16 through 2-21). The BM, BWM, and DSM are all linear cracks, and IOM and CLC are parallel to the lamellar layer of the osteons, so these were measure in arc lengths via the radius and angle (Figures 2-19 through 2-20). The black wispy cracks were seen in either two ways: as small groups of cracks less than or equal to 20 m in length (BM) (Figure 2-16), or as single cracks usually longer than 20 m (BWM) (Figure 2-17). Diffuse stained mesocracks were thicker, and very much stained (Figure 2-18). These last cracks were carefully quantified, paying special attention to lamellar orientations, so not to be confused with stained Volkmanns canals. The cement line microcracks (CLC) were either deeply stained red, or looked scalloped and black (Figure 2-21). The intra-osteonal microcracks (IOM) composed of stained wedges (Figure 2-19) and debonding lamellar layers (Figure 2-20) were clear. There were areas of increased stain uptake especially observed in interstitial bone, where the population of osteocyte lacunae is high. These were not quantified, since no linear cracks were present. Often the BM were present in these areas, and those were quantified. Specific Examples of Damage Types and Quantification A transparency with scaling marked on it (Figures 2-16 to Figure 2-21), worked as an outside eyepiece that measure length and angle of the cracks. For quantifying microdamage three essential tools used were a microscope, the frame grabber software, a CCD camera, two computers and a transparency placed on the monitor of one computer that served to measure the length and the crack orientation. On one computer the live images were displayed at a magnification of 356.25X, and at this magnification the cracks were measured. Another computer had the M-image of the section and in this one, the coordinate location of every crack was recorded. The data recorded was length of

PAGE 36

25 each crack, the angle, the type, and the x-y coordinate location for each crack. The cracks were quantified per region by a blind observer, and all the data was put into spread sheets, which later differentiate endosteal, periosteal, or center located cracks. Later a custom code was used to process the data into total crack density, length per area, and angle. Table 2-3: Crack types and descriptions. Crack Type Acronym Description Bundles of microcracks BM Damage in interstitial bone, comprised of five, 20 m long, wispy black cracks (Figure 2-16). Black wispy mesocracks BWM Individual cracks in interstitial bone, from interstitial space to osteons, or osteons to osteons or into Volkmanns Canals; longer than 20m wispy black cracks (Figure 2-17). Diffused stained mesocracks DSM Cracks from interstitial bone to osteons, or osteons to osteons, or into Volkmanns Canals; longer than 20m wispy black cracks; longer than 40m and stained red (Figure 2-18). Intra-osteonal microcracks IOM Stained wedges (Figure 2-19)and lamellar debonding (Figure 2-20). Cement lines microcracks CLC Loopy black or stained cement line cracks (Figure 2-21).

PAGE 37

Region 1mm Tension Compression Fa r N ea r Lateral Medial Periosteal Endosteal 26 Figure 2-15: The three figures show the M-image. Notice the region, and how they are identified by a letter, the scan line (orange) when counting osteons sizes. Finally, notice the transparencies and how they were used.

PAGE 38

27 20 m 30 60 90 N A, 0 Figure 2-16: Example of the damage type: bundle of microcracks (BM). The five microcracks represent one bundle. Notice they are all about the same size, and they appear perpendicular to the side of the osteons

PAGE 39

28 -30 -60 90 30 60 N A, 0 20 m Figure 2-17: Example of black wispy mesocracks (BWM), denoted by the arrows. The cracks measures 60 m and it is oriented -60 with respect to the neutral axis. Notice that the crack expands between two osteons.

PAGE 40

29 20 m -30 -60 90 30 60 N A, 0 N A Figure 2-18: Example of diffused stained mesocracks (DSM), indicated by the arrows. It measures about 60 m and is oriented -30 with respect to neutral axis.

PAGE 41

30 20 m -30 -60 90 30 60 N A, 0 Figure 2-19: Example of an intra-osteonal crack (IOM), specifically a case of a stained wedge. The wedge is the crack; it is identified for an increase in stain intake compared to its surrounding matrix. The angle of the wedge () is approximately 60 and the radius (r) of the osteon is measured to be about 80 m. The arc length (s) is then calculated from (s=r).

PAGE 42

31 20 m -30 -60 90 30 60 N A, 0 Figure 2-20: Example of an intra-osteonal crack (IOM), specifically a case of lamellar debonding. The angle measure () is 360 and the radius (r) of the debonded lamellae is about 40 m. The arc length (s) is then calculated, (s=r).

PAGE 43

32 20 m -30 -60 90 30 60 N A, 0 Figure 2-21: Example of a cement line crack (CLC), it is the black wispy line surrounding the osteon. The angle measure () is 360 and the radius (r) of the osteon is about 40 m. The arc length (s) is then calculated, (s=r). Statistical Analyses Due the amount of details involved in analyzing these sections, an organization scheme was developed, in which independent and dependent variables were explicitly identified. The independent variables are those that describe the source or origin of the bone section. They include side (right or left metacarpus), the number of cycles endured, the strain level (5,000 or 8,000), hole type (foramen, drilled or control), proximity to the hole (far or near), and endoperiosteal location (tension, compression or central). The dependent variables divide into morphometric and damage parameters and describe the microstructure of the bone and changes introduced in the testing process, respectively.

PAGE 44

33 Morphometric parameters are porosity, osteon density and osteon diameters. The damage parameters crack density (#/mm 2 ), crack length (mm/mm 2 ), and crack angle. The primary goal of the statistical analyses of the damage measured per region was to compare specific damage amongst to the three holes types (foramen, drilled hole, or control hole). All statistical analyses were performed with commercially available software (StatView; SAS Institute; Cary, North Carolina). The confounding effects were proximity to each hole (near or far) and applied strain level (5,000 or 8,000 ). The types of physical damage, the hole type, and the two confounding effects were accounted for by performing a full interaction analysis of variance (ANOVA). If no significance resulted in the interaction, the strain level factor was eliminated from the model and the ANOVA repeated with only the hole type-proximity interaction active. Consequently, if no significance resulted, then the proximity factor was eliminated, leaving only the hole type factor. At any stage of the analysis, if a significant ANOVA resulted, Fishers protected least significant difference tests were performed to determine if significant individual comparisons existed. For all analyses, significance was assumed to exist for P < 0.05. To perform statistical analyses on the cracks orientation, another database was constructed in which the experimental unit was each individual crack. Angles were measured for black wispy mesocracks (BWM), diffused stained mesocracks (DSM), and bundles of microcracks (BM). All angles were measured with respect to the neutral axis (NA) in each section; four levels were chosen, 0, 30, 60 and 90. By limiting the alternative of angles to only four possible outcomes, the response variable (angle), becomes a nominal response variable. Statistical analyses on the orientation of the linear

PAGE 45

34 microcrack and spatial occurrence (both nominal variables) were done using Contingency Tables Analyses (determines the relationship between nominal variables). Chi square tests for independence are reported, which test the likelihood that response variables are independent of each other. A small chi square P-value (significance if P<0.05) would show a significant relationship between certain response variables exists.

PAGE 46

35 CHAPTER 3 RESULTS The beams refer to in the testing plan (Figure 2-1) were either monotonically or cyclically tested. In a meaningful way they were arranged for a strain level. For example, four beam were harvested from horse S, and on was assigned to monotonic tests (SRB), two were assigned to undergo cyclic tests, 5000 and 8000 respectively (SRA and SLA), and finally one was left as a control specimen (SLB). Table 3-1 shows the results of the two monotonic tests performed. Table 3-2 shows the results of the fatigue tests performed. For two specimens relaxation tests were attempted, and the resulting data were lost. Evidently, fatigue tests at the strain level of 5,000 did not affect much the modulus of the specimen. The specimen GL (Table 3-2 and Figure 3-1) failed unexpectedly at only 3,254 cycles through the foramen. Specimen GR (Table 3-2 and Figure 3-1) endured almost half a million cycles and did not fail. AR specimen (Table 3-2 and Figure 3-1) also failed due to the contact forces between rollers and the bone beam, which created grooves. A strain-number of cycles curve (Figure 3-1), shows the strain levels at which 10 beams were loaded, and the number of cycles each tolerate. Two beams at 8000 reached failure (AL and GR), while none of the specimens tested at 5000 failed. Though Teflon tape was placed at the contact areas, grooves were created on the bone beams. The grooves were much evident on the high strained specimens (8000) (Figure 3-2), and not so dramatic on the lower strained specimens (5000). As

PAGE 47

36 mentioned before, AR failed through the site of one of the grooves before the groove problem was solved with the addition of the sacrificial bone piece. 080001600024000110000100000000Number of fatigue cyclesStrain Level (microstrain) LL LR ILA SRB SLA AR AL GR GL SRA Figure 3-1: Peak strain level versus number of cycles to failure (S-N curve). Arrows pointing to right indicate test stopped prior to failure. Table 3-1: Results for beam specimen tested monotonically to failure. Specimen ID yield yield ultimate ultimate Beam Apparent Elastic Modulus LL 88 MPa 14,000 90 MPa 20,000 7.4 GPa SRA 142 MPa 11,000 147 MPa 20,000 16.6 GPa

PAGE 48

37 Table 3-2: Results for the remainder of the beams. Specimen ID Strain level Number of Cycles Failure Modulus Reduction Initial Modulus Final Modulus SLA 8,000 60,000 yes 4.7% 12.7 GPa 12.1 GPa SRB 5,000 60,000 no none ----SLB -------------------Control beam----------------ARt 8,000 59,900 yes 11.8% 14.3 GPa 12.6 GPa AL 5,000 60,000 no none GL 8,000 3,254 yes 19.6% 14.4 GPa 11.5 GPa GR 5,000 452,836 no none ----ILA 8,000 6,500 yes Lost data ---Relaxation tests--ILB 8,000 26,000 no Lost data ---Load control--LR 5,000 21,000 no Lost data ---Relaxation tests--beam groove Figure 3-2: A groove present on an 8,000 specimen (left). A permanently deformed beam (right).

PAGE 49

38 Morphometric Parameters The morphometric parameters are porosity, osteonal density, and osteonal diameters per region. There is no significant relation between right and left beams and morphometric parameters, except near the drilled hole, which is significantly different in porosity and osteonal diameter (P = 0.0059 and P = 0.0353, respectively). The porosity (mean 6.1% standard deviation 0.9%) near the drilled hole in the right MC3s is lower than in the opposing MC3 (7.3% 1.4%) (Table 3-3). Also near the drilled hole, in the right MC3, osteon diameter (13.8%) is larger than in the left MC3 (Table 3-5). There is a significant difference between hole types for all morphometric parameters (porosity, osteon density, and osteon diameter) (Figures, 3-3. 3-4, 3-6). Porosity is significantly related to the hole type (P<0.0001) and proximity to the hole (P=0.0005). Fisher PLSD demonstrated that significance existed between foramen and drilled (P<0.0001), and foramen and control (P<0.0001), but not between control and drilled (P=0.1078) (Figure 3-3). The foramen has the highest porosity, while the drill and the control holes porosities are similar. Near the foramen, the highest porosity is found near it, while farther away the porosity is significantly lower (Table 3-3). There exists a significant difference between osteon density, hole type (P<0.0001) and the proximity to the hole (P<0.0163). Again, the significance relies on the structural differences found in the vicinity of the foramen hole. Fisher PLSD demonstrated that significance existed between foramen and drilled (P<0.0001), and foramen and control (P<0.0001), but not between control and drilled (P=0.3578) (Figure 3-4). The osteon density is higher in the foramen section, and highest near the foramen hole. For the

PAGE 50

39 drilled hole and the control hole sections the osteon density remains similar, and there is no marked difference between their far and near regions (Table 3-4). A significant ANOVA resulted from the interaction between hole type (P=0.001) and average osteon diameter, and no significant for proximity to the holes (P=0.0798) and osteon diameter. The average osteon diameter is significantly smaller in the foramen and drill sections, when compared to the average osteon in the control section. These values correspond to a significant difference in average osteon diameters between the control hole and the foramen (P = 0.008), and between the control and the drilled holes (P = 0.0295), but not between the foramen and the drill holes (P = 0.1948) (Figure 3-5). Investigating the near and far regions in each hole type, there is no significant ANOVA for average osteon diameter in proximity to the hole; however, the region near the foramen has the smallest osteon diameters (near 0.152 mm 0.010 vs. far 0.171 mm 0.012), followed by the drill section (near 0.165 mm 0.009 vs. far 0.173 mm 0.023), while the near and far regions in the control hole are very similar (near 0.184 mm vs. far 0.180 mm 0.019) (Table 3-5). Notice that there could be a structural trend for small osteons in the region close to the foramen, and it could overlap longitudinally into the region where the drill hole was placed. Damage Parameters As a reminder, damage parameters in the form of five different types of cracks were measured near the foramen, drilled holes, and control holes: 1) bundles of microcracks (BM), 2) black wispy mesocracks (BWM), 3) diffused stained mesocracks (DSM), 4) intra-osteonal microcracks (IO), and 5) cement line cracks (CLC). Black, wispy mesocracks, diffused stained mesocracks and bundles of microcracks are all linear

PAGE 51

40 Table 3-3: Results for porosity and spatial locations. Porosity Far Near Hole Type Mean Std. Dev Mean Std. Dev Control 8.00% 1.40% 8.00% 1.20% Drilled 6.60% 1.10% 6.90% 1.60% Foramen 9.10% 1.70% 15.90% 5.50% 0 .03 .05 .08 .1 .13 .15 .17 .2 .23Porosity (mm^2/mm^2) Control Drilled Foramen near far Er r o r Ba r s : 95 % Co n f i d e n c e In t e r v a l c b a abc Figure 3-3: Graph for porosity and spatial location. The repeating letters symbolize significant Pvalues between spatial locations. The bars represent the standard deviation (Std. Dev). Table 3-4: Results for osteon density and spatial locations. Osteon Density (#/mm^2) Far Near Hole Type Mean Std. Dev Mean Std. Dev Control 18.5 4.1 18.2 4.6 Drilled 16.4 1.9 18.4 3.1 Foramen 23.2 4.5 29.4 5.6

PAGE 52

41 0 5 10 15 20 25 30 35Number of Osteons/mm^ 2 Control Drilled ForamenCell near far E rro r Ba r s : 95% C onf i d e n c e I n t e r val a,b, a,b, b a Figure 3-4: Graph for osteon density and spatial location. 0 .03 .05 .08 .1 .13 .15 .17 .2 .23Osteon diameter (mm) Control Drilled ForamenCell near far E rro r Ba r s : 95% Co n f i d e n c e In t e r v a l b a a,b Figure 3-5: Graph for average osteon diameter and spatial location Table 3-5: Results for average osteon diameter and spatial locations.. Osteon Diameter (mm) Far Near Hole Type Mean Std. Dev Mean Std. Dev Control 0.18 0.015 0.184 0.023 Drilled 0.173 0.023 0.165 0.009 Foramen 0.171 0.012 0.152 0.01

PAGE 53

42 cracks that occur predominantly in interstitial spaces, between osteons or from interstitial areas into osteons. Intra-osteonal microcracks are made up of diffused stained areas inside of osteons, either lamellar debonding or stained wedges. There were 1822 cracks accounted for in the six bone sections. A total of 884 (48.51%) were found in the 5000 beam, and 938 (51.48%) were found in the 8000 beam. 679 (37.26%) were found in the foramen sections, 654 (35.89%) in the drill section, and 489 (26.83%) in the control section. Near the holes (foramen, drilled, and control) 470 (25.80%) cracks were quantified, and far from the holes 1352 (74.20%) were quantified. In the same way, 1045 (57.35%) appeared in the region defined as the center region, 422 (23.16%) in the closer to the endosteal surface (tension region), and 355 (19.48%) closer to the periosteal surface (compression region). Of the 1822 cracks 475 (26.07%) were identified as BM, 132 (7.24%) identified as BWM, 150 (8.23%) as DSM, 65 (3.58%) as IOM, and 1000 (54.88%) as CLM. Bundles of Microcracks Due to the characteristics describing this type of damage; that is, groups of approximately five 20 m long cracks (which add to 100 m), the analysis is done on the number of bundles of microcracks in each region, rather than on the length of this microcrack per region. No significance resulted from the ANOVA between strain level and the BM (P=0.3205), as well as for the interaction between hole type and proximity to the hole (P=0.1160). However, there is a significant relationship between BM and hole type (P= 0.0002) (Figure 3-6), and between BM and proximity to the hole (P= 0.0002) (Figure 3-7).

PAGE 54

43 Fishers PLSD tests demonstrated that significance difference exist between drilled and foramen holes ( P= 0.0003) and between the drill and control hole ( P= 0.0002). No significant differences exist between the foramen and control holes ( P= 0.6172) (Figure 3-6). The drilled holes contain many more cracks of this type (273) in its vicinity compared to the foramen (121) and the drilled hole (81) (Figure 3-8). Concerning the proximity to the hole, there are significantly more BM far (382) than near (93) the hole ( P= 0.0001) (Figure 3-8). Table 3-6 accounts for the mean and standard deviation (Std. Dev.) of the length per area for the BM in far and near regions for each hole type. Notice that there are over twice the amount of BM in the drilled when compared to the foramen or the control hole. Table 3-6: Results for density of BM and spatial location. Bundle of microcracks (#/mm2) Far Near Hole Type Mean Std. Dev. Mean Std. Dev. Control 6.8 5.3 5 2.2 Drilled 17.6 5.1 7.8 2.8 Foramen 9.2 6.6 2.2 1.1 Figure 3-6: Graph of density of bundles of mi crocracks (BM) and hole type. Significance exists between hole type and density of BM. 0 2 4 6 8 10 12 14 16 18 Control Drilled Foramen Cell a b a,b BM density (#/mm2)

PAGE 55

44 Figure 3-7: Graph of density of BM and proximity to the hole. A significance ANOVA results form BM and proximity to the hole. Figure 3-8: Number of BM split by proximity to the hole and hole type. 0 2 4 6 8 10 12 14 16 far near Cell y Error Bars: 95% Confidence Interval a a BM density (#/mm2) 0 25 50 75 100 125 150 175 200 225Bundles of microcracks Control Drilled Foramen near far Split By: Proximity to the hole 81 273 121

PAGE 56

45 Black Wispy Mesocracks The full interaction ANOVA for hole type, proximity to the hole and strain level resulted not insignificant for BWM ( P= 0.6812). The same is true for the interaction between hole type and proximity to the hole ( P= 0.3898). However, a significance was seen for the hole type along with and BWM ( P= 0.0126) (Figure 3-9). Fishers PLSD test demonstrated a clear significance between hole types and BWM. Significance between drill and foramen sections ( P= 0.0151), the drill section has more damage than the foramen section. Also, there is a significant difference between drill and control ( P= 0.0268), and the drill section also leads in damage. The drilled hole has over 50% more crack length per area over the foramen, while the foramen only has 33% more crack length per area when compared to the control section (Table 3-7). In addition, there no significance between the foramen and control sections, the mean difference between them being scarcely 0.054 mm/mm2 (Table 3-7). Figure 3-9: Graph for crack length for black wispy mesocracks (BWM) and hole type. 0 .1 .2 .3 .4 .5 .6mm/mm^2 Control Drilled Foramen Cell a a,b b

PAGE 57

46 Table 3-7: Results on crack length per area for black wispy mesocracks (BWM). Black wispy mesocracks (mm/mm 2 ) Hole Type Mean Std. Dev Control 0.1 0.07 Drilled 0.34 0.32 Foramen 0.15 0.13 Diffused Stained Mesocracks The data do not reveal any significance relating this type of damage and strain level, proximity to the hole, or hole type. A significant ANOVA results from the interaction between hole types and proximity to the hole (P= 0.0188). This significant interaction rejects the null hypothesis that the effect of hole types is indifferent to proximity of the hole concerning diffused stained mesocracks (DSM) (Figure 3-10). For this type of damage there is a 50% increase in crack length per area in the vicinity of the foramen when compare to the drilled hole, while the mean crack length for the rest of the hole type remain similar (Table 3-8). 0 .05 .1 .15 .2 .25 .3 .35 .4 .45mm/mm^2 Control Drilled ForamenCell near far Er r o r Ba r s : 95 % Co n f id e n c e In t e r v al Figure 3-10: Graph for diffused stained mesocracks (DSM) hole type and proximity of the hole.

PAGE 58

47 Table 3-8: Result on crack length per area for diffused stained mesocracks (DSM). Diffused stained mesocracks (mm/mm 2 ) Far Near Hole Type Mean Std. Dev Mean Std. Dev Control 0.19 0.11 0.12 0.08 Drilled 0.11 0.07 0.1 0.06 Foramen 0.11 0.07 0.26 0.19 Intra-Osteonal microcracks Arc length per area is the measure corresponding to this type of damage. For the case of intra-osteonal microcrack and the interaction of hole type, proximity to the hole and strain level, there is no significant ANOVA (P=0.4896). After the strain level factor is eliminated from the analyses, significance results from the interaction of hole type and proximity to the hole (P=0.0165). The single ANOVA for hole type is low but not significant (P=0.0671), as expected, the single significance analysis for intra-osteonal cracks and proximity to the hole is relevant (P=0.0007). As a consequence Fishers PLSD test for proximity to the hole uncovers a significant value between near and far regions (P=0.0002) (Figure 3-11), where more cracks of this type are found in the near regions, or the regions closest to the holes. The low non-significant value in concerning the hole type, further suggests that additional analysis is necessary. Effectively, Fishers PLSD uncovers a significance between foramen and control sections (P= 0.0067) and between foramen and drilled sections (P= 0.0192) (Figure 3-12). In both cases the foramen section shows more intra-osteonal microcracks (IOM), or arc length per area than the drilled and the control sections. There is no significant discrepancy between drill and control sections, though the drill section exposes more length per area of this type of damage. Far away from the

PAGE 59

48 hole the arc length per area is highest in the drilled section, while near the holes, the foramen has the highest arc length per area for this type of damage (Table 3-9). 0 .1 .2 .3 .4 .5 .6 .7 .8 .9Arc length/mm^2 far nearCell Er r o r Ba r s : 95 % Co n f i d e n c e In t e r v al a a Figure 3-11: Interaction plot for intra-osteonal microcracks and the effects of proximity to the hole. 0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1Arc length/mm^2 Control Drilled ForamenCell E rror B ars: 95% C on fi d ence I n t erva l b a a, Figure 3-12. Graph for intra-osteonal microcracks (IOM) and the effects of hole type Table 3-9: Results for intra-osteonal microcracks (IOM). Intra-osteonal microcracks (arc length/mm 2 ) Far Near Hole Type Mean Std. Dev Mean Std. Dev Control 0.06 0.04 0.31 0.32 Drilled 0.21 0.19 0.34 0.31 Foramen 0.1 0.05 0.88 0.39

PAGE 60

49 Cement Line Cracks As in the case of intra-osteonal cracks, arc length per area is the measure analyzed in cement line cracks. There is no significance in the interaction of hole type, proximity to the hole, and strain level (P=0.733). No significance results when the strain level is eliminated from the model (P=0.184). Significance results in relation to the hole type (P=<0.0001). Further analysis reveals a marked difference between foramen and drilled sections, and between foramen and control regions (for both cases P= <0.0001), where the foramen CLC lead considerably (Figure 3-13). No significance is expressed between the control and the drill regions; unexpectedly, the control hole shows a slightly higher value in arc length per area than the drill hole, while the foramen has the highest amount of cracks of this type (Table 3-10). 0 1 2 3 4 5 6Arc length/mm^2 Control Drilled ForamenCell a,b b a Figure 3-13: Interactive bar plot for cement line cracks and the effect of hole type. Table 3-10: Descriptive statistics for cement line cracks. Cement line cracks Hole Type Mean Std. Dev Control 2.54 0.82 Drilled 2.5 0.74 Foramen 4.3 1.8

PAGE 61

50 Linear Cracks and Angle Frequencies Angles were measured for linear cracks only. These include BWM, DSM, and BM. As a reminder, all angles were measured with respect to the neutral axis (NA) in each section; and four levels were chosen, 0, 30, 60, and 90. By limiting the alternative of angles to only four possible outcomes, the response variable angle becomes a nominal response variable. Contingency table analyses for angle and type of damage reveal a significantly low Chi-square P value (P=<0.0001). The significance predominately relies on the bundles of microcracks damage type. Out of 475 BM quantified, 469 were oriented in the range on 60-90 from the neutral axis. The remaining linear cracks (BWM and DSM) are more evenly distributed, with higher amounts found 0 from the NA axis (38% and 35%, respectively). Table 3-11: Angles summary. Observed Frequencies for An g le, T y pe of Dama g e0306090TotalBundles of Microcracks24178291475Black Wispy Mesocracks51173628132Diffuse Stain Mesocracks53264031150

PAGE 62

CHAPTER 4 DISCUSSION The differences in measured morphometric parameters (porosity, osteon density, and average osteon diameter) in the vicinity of the natural and drilled holes represent mechanical adaptation experienced while the animal was still alive. The parameters near the foramen were unmistakably different than near the drilled and control holes. Near the foramen existed the highest porosity, the highest osteon density, and the smallest osteons. Furthermore, the regions nearest the foramen were regions of yet the highest porosity, the highest osteon density, and the smallest osteons. It was expected that such morphometric differences would exist and was further expected that they would play the role that they did in terms of evolution of damage in each region. The porosities reported in the current work in the vicinity of the foramen, the drilled hole, and the control hole (approximately 12%, 7%, and 8%, respectively) are higher than those reported by others [35] in specimens from the dorsal, medial, and lateral aspects of the MC3 (approximately 4% to 5%). The highest porosity near the foramen is primarily due to resorption spaces, and indication of current remodeling for a mature animal not yet in senescence [16]. Similarly local regions of relatively high porosity were found in earlier work on the foramen [22], and played a major role in routing the higher stresses away from the foramen into regions of lower porosity and higher strength. The highest osteon density and the smallest osteons were found nearest the foramen, and these parameters go hand in hand. Higher osteon density and smaller 51

PAGE 63

52 osteons directly imply more cement lines per region of bone. A significant literature exists that promotes cement lines as feature that arrest and deflect microcracks in osteonal bone [2,13,15,23,28,45,47]. Mechanical testing was performed to generate damage so that correlations could be drawn between it and the morphometric parameters. To that end, methods of test preparation were developed that exposed the foramen and a drilled hole to the same mechanical duress, that minimized damage due to processing, and provided a control for existing damage. Specifically, the foramen and the drilled hole resided in the constant moment region of a four point bending beam. The drilled holes (as well as the control hole) were meticulously sculpted to provide the same topology as the foramen and to remove surface damage due to their drilling. Finally, the control holes were located in an overhanging unloaded region of each beam. In this way, the damage states near the foramen and the drilled hole could be normalized with respect to the control hole. Mechanical testing was performed on more beams than were analyzed morphometrically and for damage. The reason for this was due to the laborious and time consuming nature of the analyses. Nevertheless, interesting conclusions can be drawn from the mechanical test results themselves. The apparent initial elastic modulus of the beams containing the foramen in the current work was 14.2 1.4 GPa (mean standard deviation). The apparent initial modulus of four point beams from the palmar aspect of the equine MC3 (but not containing the foramen) has previously been reported to be 14 2 GPa [6]. The similarity between the current and previous results is encouraging and reflects the local nature of the special microstructure near the foramen.

PAGE 64

53 Another result from the mechanical tests is that of modulus reduction after cyclic loading. Modulus reduction has been shown to be a measure of the mechanical manifestation of damage in bone. The beams tested at the 5,000 strain level exhibited limited reduction in modulus. The beams tested at 8,000 exhibited a slight but clear reduction in modulus. Others have observed similar results [1]. Three point bending beams from the dorsal, lateral, and medial aspects of the MC3 strained at 5,000 were reported to exhibit no reduction in modulus [34]. Conversely, four point bending beams from the dorsal, lateral, and medial aspects of the MC3 strained at 10,000 exhibited dramatic reduction in modulus, with failure imminent after just 10,000 cycles of loading [21]. Another important aspect of this work was the implementation and improvement upon existing protocols for staining and quantifying microcrack damage in bone. The observations damage of this work agree with those of other researchers interested in fatigue damage in equine bone [35], fractures in human metatarsals [19], and the effect of damage on the mechanical properties in bone [8,12,43]. These other researchers presented images and comments describing the appearance of damage that were of great assistance in the current work [19, 30, 41]. Two types of damage bundles of microcracks and black wispy mesocracks clearly relate to the special microstructure near the foramen. Others [35] have reported a type of damage very similar in character to what was defined as bundles of microcracks in the current work. In the previous work, they were called unstained cracks and were described as small cracks on the order of 25 m. These unstained cracks were perpendicular to the periosteal surface (90 from the neutral axis of bending in parlance

PAGE 65

54 of the current work), and near the periosteal facing sides of osteons. They were also reported to be longer (by over 17%) in mechanically tested specimens compared to untested controls. Bundles of microcrack in the current work appeared black. Nearly all were in proximity to osteons and directed 90 from the neutral axis. Significantly more cracks of this type were found near the drilled hole compared to the foramen. Individually, the foramen and the drilled hole each had a decreased number of microcrack bundles further away in regions of decreased osteon density. It was expected that the black wispy mesocrack type of damage would reflect the local microstructure near each of the hole types. Given that the region nearest the foramen exhibited the highest porosity, it would be expected that more of these cracks would be found here. However, the opposite was found: it seems as if the high porosity is mitigated by the combination of higher osteon density and smaller osteons. The drilled hole does not posses the specialized osteon arrangement; near it, the crack length per area was over 2 times greater than that near the foramen. In contrast, the crack length per area near the foramen was only approximately 1.5 times greater than near the control hole. The other damage types (diffuse stained microcracks, intra osteonal microcracks, and cement line cracks) did not correlate with the expectations of decreased damage nearest the foramen. These damage types are characterized by their ability to uptake stain in higher proportions. It is also very possible that these damage types were in actuality stained Volkmanns canals. Other researchers have noted this and have thus given little credence as to these type of damage as reliable indicators of the true damage state. Finally, regions of increased remodeling (and decreased mineralization)

PAGE 66

55 necessarily uptake more stain [42], thereby increasing the perceived level of these damage types in regions nearer the foramen.

PAGE 67

CHAPTER 5 OSTEON TRAJECTORIES NEAR THE EQUINE METACARPUS NUTRIENT FORAMEN Introduction The purpose of this study is to describe osteon trajectories near a natural hole the primary nutrient foramen in the equine third metacarpus (MC3). Materials and Methods Four equine third metacarpi were obtained from an animal tissue service for this and a related study. The bones came from skeletally mature animals less than five years of age and without skeletal abnormalities. All procedures involving animal tissue use was conducted under the approval and auspices of our Institutional Animal Care and Use Committee. The bones were cleaned of soft tissue and stored at C until further preparation. Rough cuts were made first with a diamond blade band saw (Gryphon) to separate the palmar from the dorsal aspects. Three longitudinal sections, 2 mm thick in the endoperiosteal direction, were cut with a precision diamond blade band saw (Exakt) from the palmar aspect with the foramen centrally located. One parasagittal section (approximately 5 mm thick in the mediolateral direction) was cut from the palmar aspect with the foramen centrally located. Each section was mounted on a petrographic glass slide, ground (Buehler Minimet) with 600 grit silicon carbide waterproof sand paper in combination with a 3 m diamond slurry (Buehler METADI Supreme) and polished manually (Buehler TEXMET 1000) until no scratches were visible under the microscope. 56

PAGE 68

57 Successive polishings were performed on the parasagittal section so as to observe osteon trajectories while approaching the foramen mediolaterally. The bones and sections were kept hydrated with purified water during all the processing and storage. All sections were observed under a reflected light microscope (Olympus BX-60). The most periosteal surface of each longitudinal section and the most lateral surface of the parasagittal section was observed under 40X magnification. Digital images of regions around each foramen were captured with a CCD camera (Cohu), installed on the microscope and connected to a frame grabbing card (Integral Technologies FlashPoint128). The images were cropped and merged to form mosaics of larger regions (Figures 5-1 & 5-2, center images). Results Osteons with endoperiosteal trajectories were evident on the most periosteal surface, specifically on the distal and proximal apexes of the nutrient foramen (Figure 5-1). These osteons did not disrupt the lamellar layer lining the foramen evident in the images and reported previously by us [22]. These osteons were bundled together within a triangular shaped region, and were similar in size and elliptical shape. Osteons with tangential trajectories were observed elsewhere and enclosed the foramen mediolaterally. Observations on the parasagittal section yielded diverging osteon trajectories (Figure 5-2). Osteons become apparent in the mediolateral direction, outlining the foramen edges. Longitudinal osteons did not reach the edge of the hole, and some turned to become endoperiosteal osteons observed in the longitudinal sections. Many of the osteons were elliptical in shape.

PAGE 69

58 Discussion The gross shape and heterogeneous mechanical properties of the MC3 reflect its function in response to evolutionary adaptation for efficient and safe high speed ambulation [31]. Its length and large cortical cross sectional area give it an inherent resistance to both buckling and static compression failure: safety factors (SFs) exceed 5 on buckling and approach 2 on compressive strength for horses in a trot [5,16], but only slightly exceed 1 on fatigue strength [16]. The low fatigue SF reflects the high incidence of stress fractures in racing horses, up to 70% in young thoroughbreds [37]. These fractures occur on the dorsal aspect of the MC3, an area of predominant tension [37], diametrically opposite the location of the foramen. Thus, the foramen exists in a region of predominate longitudinal compression. It is well known that a transverse tension stress field exists in the up- and downstream regions near an elliptical hole in a plate subjected to far field longitudinal compression. Such regions exist near the proximodistal apexes of the MC3 foramen. It is in these regions that we observed endoperiosteal osteon trajectories, especially evident nearest the periosteal surface of the MC3. The periosteal surface represents the surface of greatest compressive bending normal stress, and, thus, the greatest transverse tension in the foramen apexes. Osteons perpendicular to this tension present their cement lines as possible crack arrestors and, along with diverging osteons which may deflect cracks, may increase toughness in these regions.

PAGE 70

59 HC HC 100 m 1 mm 10 mm D D P P Figure 5-1. Palmar view of the equine metacarpus (left). A longitudinal section containing the foramen (center). Magnified views of proximal (P) and distal (D) apices (right). Osteons with endoperiosteal trajectories in the apexes are evident by the presence of Haversian canals (HC) in these views, while osteons tangent to the foramen edge are visible elsewhere. A lamellar layer lines the foramen.

PAGE 71

60 L EP HC ML L L ML Figure 5-2. Lateral view of the equine metacarpus (left). Parasagittal section through a portion of the foramen (center). The lightly striped regions are where osteons with mediolateral (ML) trajectories and, thus, tangential to the foramen were found. Osteons with longitudinal (L) trajectories were found in the apexes of the foramen. Osteons with longitudinal and tangential trajectories are found in these regions as well. Magnified views of the ML region in the center image (top right). Mediolateral osteons are evident by the presence of Haversian canals (HC) in this view. Osteons diverge in this view (bottom right). Labeled are Haversian canals traversing in a general longitudinal direction (L) and those in an endoperiosteal (EP) direction.

PAGE 72

CHAPTER 6 CONCLUSION The ultimate outcome of the current work was to complete the description of the foramen as a structurally optimized feature in bone. The previous work of colleagues clearly demonstrated the static strength advantages of the microstructure near the foramen [22,46]. These features were then mimicked in the design of an inhomogeneous plate containing a hole [27]. The fabrication and mechanical testing of this plate design with twice the strength of a homogeneous panel demonstrated the obvious advantage of biomimickry [14]. The current work describes further the microstructure near the foramen specifically with respect to its implication on fatigue behavior and local toughening. The demonstration of reduced damage near the foramen compared to a drilled hole under the same mechanical duress, then, hopefully completes the description of the foramen as an optimized structural feature in bone. 61

PAGE 73

APPENDIX QUANTIFICATION PROTOCOL 1) Placement of the Slide a) Place bone slide on Merz grid, separated by the thickness (1.4mm) of the double sided sticky tape, and orient the bone section parallel to one of the Merz grid lines. b) Use the double-sided sticky tape to secure the bone slide in a firm position. c) Record the orientation of the bone slide, if different from the usual; lateral to the right, and periosteal toward the observer. 2) Images are observed on the microscope with a 4X objective. An image grabber program (Flashpoint 3.1), and a CCD camera connected to the microscope are use to look at the bone sections magnified 142.5X on the computer screen. Images are saved with the frame grabber program, and with the aid of the Merz grid, each image is taken 1mm apart. a) Focus on the bone and take image. b) Scan the entire length taking images, the bone section fits in two rows of 15 images each. c) Reaching final quadrant (a quadrant boxes in the Merz grid) take an image then focus on grid and take an image. The image of the grid will be useful defining the appropriate area to crop off the images when putting together. 3) Highlight placement On printout, representing entire Mertz grid, highlight the quadrants in which the bone was placed, and its medial and endosteal corner. 4) Producing mother image a) Images entered into Matlab to be cropped b) Run program to convert to larger image c) Name this image the mother image 5) Placement of grid over mother image a) Print out mother image, made up by three 8x11 printouts. Each printout is made up by ten images, which were cropped to a square millimeter and put together with Matlab. b) Align grid along both sides of the hole or center of the bone c) Secure transparencies with tape d) Note the orientation of the bone 6) The grid is composed of boxes measuring 400 m 2 From the center, the grid is 62

PAGE 74

63 sectioned into two columns, on both sides of the hole, measuring 50mm each. From there, the rest of the columns are 100mm in width covering the rest of the bone. The grid also has three rows separating periosteal, center, and endosteal sections. 7) Area per section a) Number of squares (measuring 10,000 m^2) covering bone are counted b) This number is multiplied by 10,000 giving the area of bone for each quadrant 8) Osteon count a) Haversian canals are marked red with marker b) Count number of red dots in each quadrant 9) Resorption space area a) Resorption spaces are marked blue b) In each quadrant, the number of small boxes (400 m^2) colored blue are counted c) The number is then multiplied by 400 giving the total resorption space area for that quadrant 10) Volkmanns canal length a) Volkmanns canals are marked yellow b) Number of small boxes (400 m^2) colored yellow are counted c) This number is then multiplied by 20 to give the total length of the canals for the individual quadrant 11) The data collected for steps 8-10 is then entered into an Excel spreadsheet. These are entered into the appropriate column and then divided by the area of bone found in that particular quadrant on the grid. 12) Preparing mother image for osteon diameter measurements a) Open mother image (6165x822) in Paint Shop a. Locate where the lamellar layer ends, or the drill hole start, for the medial and lateral sides b) Record these numbers, in pixels, in orange cells on specific chart. The yellow cells are the location, in pixels, of the scan lines for osteon diameter measurements. The two cells under the yellow cells are the locations of the boundary lines. c) The numbers in the yellow and white cells must be entered in Matlab d) Run the program 13) Labeling the sections a) Open mother image, now containing scan lines and boundary lines, in Paint Shop. b) Label each section with a letter following the below lettering system:

PAGE 75

64 Medial hole Lateral .. 200 100 50 50 100 200 F G H I J K MM M L LL O P Q R S T 14) Print out mother image with appropriate lines and labels and images of individual sections. Preview printouts of sections, circling possible osteons that might be measured later. 15) Getting images of osteons a) Looking at the slide at 10X magnification and under circular polarized light b) Each section is scanned following scan lines as guides, for osteons present in the boundaries. c) Take images of osteons, scanning down each section, using printouts as references 16) Measuring osteon diameters a) Open images of sections in Paint Shop b) Label secondary osteons with an s c) Measure osteons according to guidelines stated below: Rules for Measuring Osteons Circular osteons 1 diameter (x-direction) Elliptical osteons 2 (x,y) diameters For Secondary Osteons Clearly see the cement line (on x axis) or (x, y axis) For Primary Osteons Clearly see the cement line (if not on x and y directions), then only on 45, 135, 225, 315, 90, 180, 270 or 360. All measurements must go thru the Haversian Canals 17) Print out images of the osteons with labels and measurement lines with a ruler. a) Enter the numbers in the appropriate excel sheet.

PAGE 76

LIST OF REFERENCES 1. Akkus O, Knott DF, Davy DT, Rimnac CM. Linear microcracks may contribute little to degradation of Youngs modulus in cortical bone tissue. Bioengineering Conference; ASME 2001; 50: 79-80. 2. Akkus O, Rimnac CM. Corical bone tissue resists fatigue fracture by deceleration and arrest of microcrack growth. J of Biomechanics; 2001; 34:757-764. 3. Anderson MA, Mann FA, Kinden DA, Wagner-Mann CC. Evaluation of cortical bone damage and axial holding power of nonthreaded and enhanced threaded pins placed with and without drilling of a pilot hole in femurs form canine cadavers. JAVMA; 1996; 208(6): 883-887. 4. ASTM 790M-82. Flexural properties of unreinforced and reinforced plastics and electrical insulating materials. 1982. 5. Bartel DL, Schryver HF, Lowe JE, Parker RA. Locomotion in the horse: a procedure for computing the internal forces in the digit. American Journal of Veterinary Research 1978:39(11):1721-7. 6. Bigot G, Bouzidi A, Rumelhart C, Martin-Roset W. Evolution during growth of the mechanical properties of the cortical bone in equine cannon-bones. Med. Eng. Phys.; 1996; 18(1):79-87. 7. Bonfield W, Li CH. The temperature dependence of the deformation of bone. J of Biomechanics; 1968; 1:323-329. 8. Boyce MT, Fyhrie DP, Glotkowski MC, Radin EL, Schaffler MB. Damage type and strain mode associations in human compact bone bending fatigue. Journal of Orthopaedic Research: 1998; 16:322-329. 9. Burr DB, Hooser M. Alteration to the en bloc basic fuchsin staining protocol for the demonstration of microdamage produced in vivo. Bone; 1995; 17(4): 431-433. 10. Burr DB, Schaffler MB, Frederickson RG. Composition of the cement line and its possible mechanical role as a local interface in human compact bone. J of Biomechanics; 1998: 21(11):939-945. 11. Burr DB, Stafford T. Validity of the bulk-staining technique to separate artifactual from in vivo bone microdamage. Clinical Orthopaedics and Related Research; 1990; 260:305-308. 65

PAGE 77

66 12. Burr DB, Turner CH, Naick P, Forwood MR, Ambrosius W, Hasan MS, Pidaparti R. Does microdamage accumulation affect the mechanical properties of bone? J of Biomechanics; 1998: 31: 337-345. 13. Burr DB, Turner CH, Naick P, Forwood MR, Pidaparti R. En Bloc staining of bone under load does not improve dye diffusion into microcracks. J of Biomechanics; 1998; 31: 285-288. 14. Buskirk SR, Venkataraman S, Ifju PG, Rapoff AJ. Functionally graded biomimetic plate with hole. Proceedings of Society for Experimental Mechanics Annual Conference & Expositions, Milwaukee, WI, 2002. 15. Carter DR, Hayes WC. Fatigue life of compact bone I. Effects of stress amplitude, temperature and density. J Biomechanics: 1976; 9: 27-34. 16. Cheney JA, Liou SY, Shen CK, Wheat JD. Cannon-bone fracture in the throughbred racehorse. Med Biol Eng; 1973: Sep 11(5): 613-20. 17. Cowin, Stephen C. Bone Mechanics Handbook. Boca Raton, CRC Press LLC, 2001. 18. Dallap BL, Bramlage LR, Embertoson RM. Results of screw fixation combined with cortical drilling for treatment of dorsal cortical stress fractures of the third metacarpal bone in 56 Thoroughbred racehorses. Equine Veterinary Journal; 1999; 31(3): 252-257. 19. Donahue SW, Sharkey NA, Modanlou KA, Sequeira LN, Martin RB. Bone strain and microcracks at stress fracture sites in human metatarsals. Bone; 2000; 27(6): 827-833. 20. Frost HM. Preparation of thin undecalcified bone sections by rapid manual method. Stain Technology; 1958; 33:273-277. 21. Gibson VA, Stover SM, Martin RB, Gibeling JC, Willits NH, Gustafson MB, Griffin LV. Fatigue behavior of the equine third metacarpus: mechanical property analysis. Journal of Orthopaedic Research; 1995; 13(6): 861-868. 22. Gotzen N, Cross AR, Ifgu PG, Rapoff AJ. Understanding stress concentration around a nutrient foramen. Journal of Biomechanics; 2002; invited paper accepted for publication. 23. Gou XE, Liang LC, Goldstein SA. Micromechanics of osteonal cortical bone fracture. J of Biomechanical Engineering; 1998; 120: 112-117. 24. Griffin LV, Gibeling JC, Gibson VA, Martin RB, Stover SM. Artifactual nonlinearity due to wear grooves and friction in four-point bending experiments of cortical bone. J Biomechanics; 1997; 30: 185-188.

PAGE 78

67 25. Gross TS, McLeod KJ, Rubin CT. Characterizing bone strain distributions in vivo using three triple rosette stain gages. J of Biomechanics; 1992; 25(9): 1081-1087. 26. Gustafson MB, Martin RB, Gibson V, Storms DH, Gibeling J, Griffin L. Calcium buffering is required to maintain bone stiffness in saline solution. J of Biomechanics; 1996; 29(9): 1191-1194. 27. Huang J, Venkataraman S, Rapoff AJ, Haftka RT. Optimization of axisymmetric elastic modulus distributions around a hole for increase strength. Structural and Multidisciplinary Optimization 2002; accepted for publication. 28. Jepsen KJ, Davy, DT. Comparison of damage accumulation measures in human cortical bone. J of Biomechanics; 1997; 30: 891-899. 29. Jepsen KJ, Davy DT, Krzypow DJ. The role of the lamellar interface during torsional yielding of human cortical bone. J of Biomechanics; 1999; 32: 303-310. 30. Lee TC, OBrien JO, Taylor D. The nature of fatigue damage in bone. International Journal of Fatigue; 2000; 22: 847-853. 31. Les CM, Stover SM, Keyak JH, Taylor KT, Willist NH. The distribution of material properties in the equine third metacarpal bone serves to enhance sagittal bending. J of Biomechanics; 1997; 30: 355-361. 32. Martin RB, Burr DB, Sharkey NA. Skeletal Tissue Mechanics. New York, Springer-Verlag, 1998. 33. Martin RB, Gibson VA, Stover SM, Gibeling JC, Griffin LV. Osteonal structure in the equine third metacarpus. Bone; 1996; 19 (2): 165-171. 34. Martin RB, Gibson VA, Stover SM, Gibeling JC, Griffin LV. Residual strength of equine bone is not reduced by intense fatigue loading: implications for stress facture. J of Biomechanics; 1997; 30:109-114. 35. Martin RB, Stover SM, Gibson VA, Gibeling JC, Griffin LV. In vitro fatigue behavior of the equine third metacarpus: remodeling and microcrack damage analysis. Journal of Orthopaedic Research; 1996; 14: 798-801. 36. McDonald DE, Palmer RH, Hulse DA, Neigut JS, Hyman WA, Slater MR. Holding power of threaded external skeletal fixation pins in the near and far cortices of cadaveric canine tibiae. Veterinary Surgery; 1994; 23:448-493. 37. Norwood GL. The bucked-shin complex in thoroughbreds. Proceedings of the American Association of Equine Practitioners; 1978; 24:319-36. 38. Nunamaker DM, Butterweck DM, Provost MT. Fatigue fractures in thoroughbred racehorse: relationships with age, peak bone strain and training. Journal of Orthopaedic Research; 1989; 8: 604-611.

PAGE 79

68 39. Pasquini Chris, Reddy V. Krishna, Ratzlaff Marc H. Altas of Equine Anatomy. Eureka, California, Sudz, 1983. 40. Rapoff AJ, Gotszen N, Cross AR, Ifju PG. Understanding stress concentration around a nutrient foramen. BED-Vol. 50, 2001 Bioengineering Conference, ASME 2001. 41. Reilly CG Observation of microdamage around osteocyte lacunae in bone. J of Biomechanics; 2000; 33: 1131-1134. 42. Rho JY, Zioupos P, Cirrey JD, Pharr GM. Microstructural elasticity and regional heterogeneity in human femoral bone of various ages examined by nano-indentation. J of Biomechanics; 2002;35:189-198. 43. Schaffler MB, Radin EL, Burr DB. Mechanical and morphological effects of strain rate on fatigue of Compact Bone. Bone; 1989; 10: 207-214. 44. Toews AR, Bailey JV, Townsend HGG, Barber SM. Effect of feed rate and drill speed on temperature in equine cortical bone. AJVR; 1999; 60(8): 942-944. 45. Vashishth D, Behiri JC, Bonfield W. Crack growth resistance in cortical bone: concept of microcrack toughening. J of Biomechanics; 1997; 30: 763-769. 46. Venkataraman S, Hafta RT, Rapoff AJ. Structural optimization using biological variables to understand how bones design holes. Structural and Multidisciplinary Optimization 2002; accepted for publication. 47. Yeni YN, Norman TL. Calculation of porosity and osteonal cement line effects on the effective fracture toughness of cortical bone in longitudinal crack growth. J Biomed Mater Res; 2000; 51: 504-509. 48. Zioupos P. On microcracks, microcracking, in-vivo, in-vitro, in-situ and other issues. J of Biomechanics; 1999; 32:209-211.

PAGE 80

BIOGRAPHICAL SKETCH Barbara Garita Canet was born in 1977. She is the second child of Donato Garita and Noemy Canet who had six children. She grew up in San Jose, Costa Rica, and attended Lincoln School from 1983 to 1995. After her graduation form Lincoln School, she started her career of mechanical engineering at the Universidad de Costa Rica; however she only studied there for one semester. She was offered a partial scholarship for playing varsity volleyball for Florida Institute of Technology (FIT.). From FIT she graduated with a bachelors degree in mechanical engineering in May 2000. In August 2000 she started a masters in biomedical engineering at the University of Florida and graduated in December 2002. She will pursue a doctoral degree in biomedical engineering. 69


Permanent Link: http://ufdc.ufl.edu/UFE0000531/00001

Material Information

Title: Foramen is Tougher than drilled hole in equine third metacarpus
Physical Description: Mixed Material
Creator: Garita, Barbara.
Publication Date: 2002
Copyright Date: 2002

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000531:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000531/00001

Material Information

Title: Foramen is Tougher than drilled hole in equine third metacarpus
Physical Description: Mixed Material
Creator: Garita, Barbara.
Publication Date: 2002
Copyright Date: 2002

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000531:00001


This item has the following downloads:


Full Text












FORAMEN IS TOUGHER THAN DRILLED HOLE IN EQUINE THIRD
METACARPUS















By

BARBARA GARITA


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2002




























Copyright 2002

by

Barbara Garita





















I dedicate this work to my family, Donato, Noemi, Esteban, Pablo, Ifigenia, Ignacio,
Natalia and my grandmother Margarita.















ACKNOWLEDGMENTS

I would like to acknowledge that this work would have been much more difficult

without the advice, company, commitment, and friendliness of the people that coexist

with me everyday. All of them in one way or the other helped this work enormously,

probably much more than they know or I can express.

I would particularly like to let Dr. Rapoff know that his counseling, patience and

support were crucial, obviously in the completion of my thesis, but also in my out of

school education. Similarly, I would like to thanks Wes Johnson for sharing so many

ideas, for coaching me through graduate school, and simply for caring for me. I feel that

Dr. Satchi Venkatarama got me closer to the reality of graduate school, and offered

immense strengthening and understanding, and Stephanie Buskirk offered valuable help

and company. Also, I would like to thank to Dr. Hafka and Dr. Jing Huang, Olivier

Fontanel, Renaud Rinaldi, Ron Brown, Matt Gabriel and Nicole Gasparina. Outside

school many friends significantly participated in my research by showing interest in my

work like Natalia P. and Ryanne C. and my family in Costa Rica.

Finally I must acknowledge the energy, tools and means behind my research: the

charm of nature and the methods of science, the available technology and the monetary

sources made available by the Biomedical Engineering Department, NASA and

Aerochem.
















TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ......... ... ............. .. ...... ... .............. ............ .. vii

LIST OF FIGURES ..................... .......... .................................. viii

A B ST R A C T ................. .......................................................................................... x

CHAPTER

1 INTRODUCTION .............. .....................................................1..

2 M ETH OD S AN D M A TERIALS............................................................................... 2

A natom y ............................................................................ .................... 2
Loading Environm ent and Fracture Incidence .............................................................3
B one Procurem ent and Identification ........................................ ............................... 5
Beam Dim tensions ...................................... ................................ ......... 5
M achining of Beam s ............... ................. ............................ ..... ..7..
D killing H oles ........................................................ ................ .. 8
M mechanical Testing Setup .................................................... ........ ............... .11
M mechanical Testing Plan .................. ...................................... .. ........ .... 13
M onotonic Tests.................... .................................17
F fatigue T ests ............................................................... ..... ............ 17
Staining ...............................................................................................18
Sectioning of Beam s for M icroscopy ........................................ ........................ 18
M icroscopy Section A nalyses Setup.................................... ........................ .. ......... 20
M orphom etric P aram eters...................................................................... ..................2 1
D am age P aram eters ......... ........... ........................................ .......... .. ...... .... 23
Specific Examples of Damage Types and Quantification ...........................................24
Statistical Analyses .................... ......................... ........... 32

3 R E SU L T S ............................................................................. 3 5

M orphom etric Param eters............................................ .................. ............... 38
Dam age Param eters .................................... .. .... ..... .. ............39

4 D IS C U S SIO N ...................................................................5 1



v









5 OSTEON TRAJECTORIES NEAR THE EQUINE METACARPUS NUTRIENT
F O R A M E N .......................................................... ................ 5 6

In tro du ctio n ...................................... .....................................................5 6
M materials an d M eth od s........................................................................... ................... 56
R e su lts ................... ......................................................................... 5 7
D iscu ssio n ...................................... ......................................................5 8

6 C O N C L U S IO N ..................................................................................... ................ .. 6 1

APPENDIX: QUANTIFICATION PROTOCOL..................................... ............... 62

L IST O F R E F E R E N C E S .......................................................................... ....................65

B IO G R A PH IC A L SK E TCH ...................................................................... ..................69
















LIST OF TABLES

Table p

2-1: T testing plan .............................................................................. 17

2-2: Staining procedure .................. ................................ ...... .. .. ............ 19

2-3: Crack types and descriptions. ............................................. ............................. 25

3-1: Results for beam specimen tested monotonically to failure .....................................36

3-2: Results for the remainder of the beams. .................. ..... ... ................. 37

3-3: Results for porosity and spatial locations ....................................... ............... 40

3-4: Results for osteon density and spatial locations. .................. ................40

3-5: Results for average osteon diameter and spatial locations................. ............41

3-6: Results for density of BM and spatial location. ................................. ............... 43

3-7: Results on crack length per area for black wispy mesocracks (BWM)...................46

3-8: Result on crack length per area for diffused stained mesocracks (DSM). ................47

3-9: Results for intra-osteonal microcracks (IOM)............................ ..... ...............48

3-10: Descriptive statistics for cement line cracks. ................................... ..................... 49

3-11: A ngles sum m ary .......................... ......... .. .. ........ .. ............50
















LIST OF FIGURES


Figure p

2-1: Skeleton of a horse. ................................ ......... ....................4

2-2: The arrow points to the location of the nutrient foramen (right)............... ...............4

2-3: Pictures identifies between right and left cannon bones. ..........................................6

2-4: Finite element model of the beam with the holes. ......................................................7

2-5: An equine MC3 (upper left), dashed lines indicate proximal and distal cutting
locations. Remnant proximal and distal cuts (top right). .................. ...............9

2-6: Mediolateral cuts being made with a diamond blade band saw (left)........................10

2-7: Drilling of the holes with nitrogen powered drill. .................................................10

2-8: The toothpick with 320 Silicon Carbide powder adhesively bonded. ....................11

2-9: Oblique view of free body diagram for a beam in four point bending (A). ..............14

2-10: Actual tested beam. Locations where the Teflon tape has worn out down by the
friction of the loading noses during cyclic testing (A). .....................................15

2-11: Testing setup (top)..................................... ............................... .... 16

2-12: Stained bone beam ............................................... .. .......... ........ .... 19

2-13: Plan view of bone beam, shown with holes near each end as they exist in the actual
b eam ............................................................................. 19

2-14: Three sections quantified for compositional and microcrack parameters ..............23

2-15: The three figures show the M -image. ........................................ ............... 26

2-16: Example of the damage type: bundle of microcracks (BM)................................27

2-17: Example of black wispy mesocracks (BWM), denoted by the arrows. .................28

2-18: Example of diffused stained mesocracks (DSM), indicated by the arrows. ...........29









2-19: Example of an intra-osteonal crack (IOM), specifically a case of a stained wedge. 30

2-20: Example of an intra-osteonal crack (IOM), specifically a case of lamellar
debonding ........................................................................ 3 1

2-21: Example of a cement line crack (CLC), it is the black wispy line surrounding the
osteon ............................................................................32

3-1: Peak strain level versus number of cycles to failure ("S-N curve")...........................36

3-2: A groove present on an 8,000t specimen (left). A permanently deformed beam
(right). .............................................................................37

3-3: Graph for porosity and spatial location ................................................................40

3-4: Graph for osteon density and spatial location. ................................ .................41

3-6: Graph of density of BM and hole type ..................................................................43

3-7: Graph of density of BM and proximity to the hole. ............... .............. 44

3-8: Number of BM split by proximity to the hole and hole type. ................................44

3-9: Graph for crack length for black wispy mesocracks (BWM) and hole type .............45

3-10: Graph for diffused stained mesocracks (DSM), hole type and proximity of
th e h o le ............. ......... .. .. ......... .. .. .......... ...................................... 4 6

3-11: Interaction plot for intra-osteonal microcracks and the effects of proximity to the
h o le ............................................................................ 4 8

3-12: Graph for intra-osteonal microcracks (IOM) and the effects of hole type ..............48

5-1: Palmar view of the equine metacarpus (left)............ ............... ...............59

5-2: Lateral view of the equine m etacarpus (left).................................. ..................... 60















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

FORAMEN IS TOUGHER THAN DRILLED HOLE IN EQUINE THIRD
METACARPUS

By

Barbara Garita

December 2002
Chair: Dr. Andrew Rapoff
Department: Biomedical Engineering

The purpose of this work was to compare the damage resistance and tolerance of

bone in the vicinity of a natural hole, the nutrient foramen and a drilled hole in the equine

third metacarpus. The nutrient foramen hole starts to form in early fetal stages, and the

microstructure around it has been adapting to its presence since then. The indefinite

adaptability is a reason why related studies have investigated the foramen and the

microstructure around it. These studies have revealed that the microstructure reduces the

local stress concentrations near the foramen, and that it increases structural strength.

Also, elastic and strength manifestations of this microstructure were mimicked in the

design and fabrication of a plate with a central hole. This plate demonstrated superior

performance over a uniform plate with a hole. The present research deepens even more

into the microstructure around the foramen and verifies that it also increases damage

resistance and tolerance ("toughness") in response to cyclic loading. Also, osteon

trajectories near the foramen were uncovered; they exist in regions of transverse tension










stress in response to global loading. Their arrangement locally increases toughness, as


well.














CHAPTER 1
INTRODUCTION

It is of interest to quantify and analyze the occurrence of physical damage in the

form of cracks in the vicinity of a natural hole, in the equine third metacarpus (MC3).

The equine MC3 is the major load bearing bone in the forelimb of the horse, and nutrient

foramen is a hole that exists in its back or palmar aspect. Most holes in structures present

themselves as stress concentrations, but the foramen is a region of stress reduction [22].

The microstructural variations near the foramen are responsible for making this hole a

site of stress reduction [22,40]. The purpose of this work is to continue investigating the

microstructure near the foramen; specifically, to evaluate its damage resistance and

tolerance to cyclic loading.

The following chapters will describe how the damage resistance and toughness of

the microstructure in the vicinity of the foramen was compared to that of a drilled hole.

Beams were procured from the palmar aspect of the equine MC3, all containing the

foramen. A drilled hole was placed a distance away from the foramen. The beams were

cyclically loaded at high strain levels effecting the same loading both holes. The cyclic

loading induced damage around the holes, and later that damage was quantify and

analyzed.

Chapter 5 deals with the latest findings on osteon trajectories near the foramen.

The location and arrangement of these osteonal trajectories reveal further localized

adaptation and damage tolerance around the foramen.














CHAPTER 2
METHODS AND MATERIALS

The equine third metacarpus (MC3) is a captivating bone. At first glance, it is

distinctly recognized because the remnants of two vestigial bones are attached to its sides.

Its shape reflects evolutionary adaptation that allows for efficient and high speed

ambulation [31]. Still, it has been repeatedly studied because of its incidence as a stress

fracture site [6,18,35,33,34,37,38]. Like many other long bones, it has a nutrient foramen

located on its palmar aspect, around a third of its length away from proximal metaphysis

(Figure 2-2).

The following pages intend to give details on specimen preparation: harvesting

bone beams, drilling holes, staining with basic fuchsin; mechanical testing, quantifying

compositional and damage parameters, and performing statistical analyses.

Anatomy

The third metacarpus bone (MC3) stretches from the elbow to the fetlock joint

(Figure 2-1). The second and fourth metacarpals are two vestigial metacarpals attached

through syndesmoses to the medial and lateral aspects of the third metacarpus; together

they are known as the cannon bone. There is no trace of the first and fifth metacarpals;

consequently, the horse walks only on the middle digit.

The nutrient foramen is a natural hole, an intrusion through the bone's cortex, on

the palmar side of the third metacarpus, located approximately 60 mm away from the

proximal end, and usually centered, equidistant from the lateral and medial sides of the

bone. Typically it has an elliptical shape and its mayor axes measure about 3 mm and 2









mm. A nutrient vessel crosses the cortex, through the foramen hole, to supply the bone

with necessary nutrients.

At a macroscopic level, cortical or cancellous bone appears solid. Cancellous

bone is highly porous and appears at the ends of long bones. Cortical or osteonal bone, is

found in the diaphyses (middle region) of long bones (like the MC3), and it is of concern

in this study. The porosity of cortical bone is usually between 5% and 10%; its pores

consist of canals (Haversian and Volkmann's canals) or temporary cavities resorptionn

spaces). Cortical bone is composed of lamellar layers, either in a parallel order

(plexiform bone) or in a circumferential arrangement (osteonal bone). Osteonal bone is

composed of osteons, which are concentric layers of lamellar wrap around a central canal

(Haversian canal). In addition, osteons are cylindrical in shape, about 200 .im in

diameter and usually oriented parallel to the longitudinal axis of long bones. Osteons are

surrounded by a weak interfacial layer, called the cement line, which arrests and deflects

cracks (physical form of damage). Cortical bone undergoes remodeling (removal and

replacement of bone), in part, to repair cracks.

Loading Environment and Fracture Incidence

The MC3 is one of the most common sites for stress fractures in racing horses,

particularly, in young thoroughbreds [38,34,35]. Peak periosteal strains, in other animal

species can reach peaks strains of 2,000 ip to 3,000 ip [34]; however, peak compressive

strains of nearly 5,000 [i have been measured on dorsal surfaces of the MC3 during

galloping [38]. Shear strains approximately measuring 44% less in magnitude were

measured during trotting, indicating a torsional loading component. Still, the equine third

metacarpus is primarily axially loaded in compression due to gravity and in vivo tests

using strain gages suggest dorsopalmar bending [34]
















Equine third
metacarpus


Proximal
direction



,oursal -* Palmar
I \. direction direction


>int )_ Distal
9 / direction
Skeleton of a horse. The ellipse surrounds the equine third metacarpus. An
expanded view elaborates on the terms describing a two dimensional
orientation of the bone. In the case of a three dimensional orientation, the
lateral direction will be out of the page, and the medial direction will be into
the page.


Proximal
direction


[it'


Lateral
direction


Medial
direction


.ck)


Distal
direction


Palmar View Image from [16]
Figure 2-2: The arrow points to the location of the nutrient foramen (right). Location of
the principal nutrient vessels, as it crosses the palmar cortex (left).


Figure 2-1:









Bone Procurement and Identification

Six pairs of metacarpi (12 bones total) were obtained from a commercial

processor (Animal Technologies, Tyler, Texas). The animals were between three and

five years of age who died from reasons unrelated to this work. The bones were stored at

-260C. Every bone was handled following the guidelines of our Institutional Animal

Care and Use Committee. After thawing, the limbs were cleared of the remaining soft

tissue and identified as right and left cannon bones.

The identification of the right and left metacarpal bones was complicated, since

the providers of the limbs included the carpal bones with the third metacarpals. In order

to anatomically differentiate between right and left metacarpal, the carpal joint (the joint

between the carpal bones and the MC3) had to be separated. A depression on the top

view of the metaphysis identified the lateral aspect [39] (Figure 2-3).

Beam Dimensions

When preparing the beam specimens, the ASTM Standards D 790M-82 were

followed as strictly as possible. However, due to the anatomical difficulties on the

palmar aspect, a support span to depth ratio of 40 to 1 was used instead of a 16 to 1 ratio,

as has been used by others [35]. The ASTM Standards recommends 25 mm width beams

for a depth of 2 mm [4]; however, the palmar aspect has a limited width (due to the

presence of the second and third vestigial metacarpals); and only 15 mm width was

possible. Considering the overhanging length, outside the supports, the beams for

experimentation measured 2 x 15 x 110 mm.




































Figure 2-3: Pictures identifies between right and left cannon bones. The black arrows
point the carpal bones, and the white arrows point on a depression on the
metaphysis of the third metacarpal bones that identify the lateral aspects.

The length of 110 mm in the bone beams was required, because of the necessity

of positioning the two holes (the foramen and a drilled hole) between the loading noses.

A finite element model of a beam with mechanical properties representative of the MC3

was constructed by a colleague (Dr. Jing Huang, Figure 2-4). This model was used to

confirm that stress fields of the drilled hole and the foramen would not interact. The

model was subjected to loading and boundary conditions representing those of the

experiments. The results demonstrated that a distance of 12 mm between the hole centers

was sufficient for the stress fields not to interact.






























Figure 2-4: Finite element model of the beam with the holes. The model confirms the
distance between the holes is enough for the stress fields from the holes to act
independently. A and B show the intricate detail of the mesh, with more
nodes in the proximity of the holes. C and D show the stress fields next to
the holes. In figure D the arrows point to a close up of the stress fields;
evidently, there is no interaction between them (they do not touch each
other).

Machining of Beams

Beams were machined from the palmar aspect using a combination of rough and

precision guided cuts. After proper anatomical identification, a normal band saw was

used to make the initial rough cuts on the cannon bones; they were cut 60 mm proximal

from the foramen and 80 mm distal from the foramen. The remaining bulks were cut

longitudinally, separating the palmar from the dorsal aspects of the bone (Figure 2-5).

The finer longitudinal cuts (slabs) were made using a precision diamond blade band saw.

The specimens were grinded and polished to the final thickness of approximately 2 mm


C



izzal I









to eliminate all the possible surface scratches introduced by the cutting blades. The

beams were kept hydrated throughout these and all other preparations.

Mediolateral cuts were made with the aid of a plexiglas fixture (Figure 2-6). The

width of the beams were cut with a diamond band saw (Gryphon C-40, Sylmar,

California). The same plexiglas assured the same width for all specimens (15 mm)

(Figure 2-6, right).

Drilling Holes

The drilled hole and the control hole were machined with a nitrogen powered drill

at a speed of 75 RPM. A powered drill was used to reduce wobble [36]. Slow drilling

will significantly decrease the maximal bone temperature on the contact surfaces [43].

Application of a moderate axial force and drilling in as short of time as possible through

the bone will reduce the damaging effects of the drilling process [3]. The plexiglas fixture

also served to create a pilot hole when using the drill and guaranteed the same location

for the drill holes in each specimen (Figure 2-7).

The drilled holes were made as similar as possible to the foramen. After drilling

the holes, the foramen was carefully measured and compared in dimensions to the drilled

holes. If any major differences were present, a milling machine was used to modify the

drilled holes. Finally, the drilled holes were meticulously ground and polished to

corresponding elliptical measures, using a toothpick to which Silicon Carbide 320 grit

powder (Buehler; Lake Buff, IL) was glued.

Before any testing, the beam specimens were polished again with 600 sand papers

and polishing cloths (Buehler Texmet Cloth; Lake Buff, IL) and water [20]. All beams

were marked with a chamfer on the lateral and distal corner to discern the correct

anatomic orientations throughout the experiments.




















































Figure 2-5: An equine MC3 (upper left), dashed lines indicate proximal and distal cutting
locations. Remnant proximal and distal cuts (top right). A longitudinal cut
being made in the Exkact saw (Exkact; Germany) (center), notice the foramen
is visible. Resulting machined sections from the six equine third metarpi
(bottom left) and a close up view of one pair (bottom right).


t -C`























Figure 2-6: Mediolateral cuts being made with a diamond blade band saw (left). Close up
fixture used to guide blade (right).

















Figure 2-7: Drilling of the holes with nitrogen powered drill. The plexiglas fixture is
used to drill the pilot hole, to diminishes wobble, and locate repeatedly the site
of the drilled holes.
































Figure 2-8: The toothpick with 320 Silicon Carbide powder adhesively bonded. Used for
meticulously sculpting the drilled hole like the foramen (top left). Plexiglas
fixture used to hold the beam and even forces while grinding and polishing
beams manually (top right). Example of a beam as it looked before testing
(bottom).


Mechanical Testing Setup

Two strain levels where chosen: a supraphysiological strain of 8,000 ipe, very

high physiological strain of 5,000 ip. Others [35] have tested in three point bending

lateral, dorsal, and medial bone beam from the MC3 for 10,000 cycles at 5,000 pi and

found no reduction in elastic modulus or yield strength. Others [21] have fatigue tested

the same aspects of the MC3, in four point bending at 10,000 pg until failure. These

beams endured between 300 and 21,000 cycles. Neither group tested the palmar aspect.

Considering that the beams in this experiment have two holes in the loaded area, the

strain levels chosen seem reasonable. Bending was accomplished by attaching a custom

fixture to a servohydraulic testing machine (MTS 858 MiniBionix with a 407 controller)









and a load cell (Entran, NJ) (Figure 2-11). Each beam was placed in the fixture so that

the periosteal surface experienced compression.

Various fixture for holding the beams where investigated. Initially the beams

moved off the supports. Possibly the anisotropy of the bone created a nonlinear response

to the stresses induced while fatigue testing. The problem was solved by restraining the

beams to the support rollers using rubber bands (Figure2-10).

A stainless steel four point bending fixture was designed and constructed. The

fixture has a pivot axis to assure that the support noses and the loading noses remain

parallel to each other when loading a specimen (Figure 2-11). Rollers where held by

rubber bands, to assured movement at the contact points, and to reduce the friction forces

against the beams surfaces, while cyclic tests took place (Figure 2-10).

The support span was 80mm and the loading span is 40mm, as recommended by

the by the ASTM 790-82 standard [4] (Figure 2-9).

Studies on cyclic testing on bone specimen using four and three point bending

fixtures [23] have noted that stainless steel fixed rollers can create grooves on the bone

beams[24]. These grooves become the primary cause of an alteration in the stiffness of

the beam. Guided by finite element results [23] the smallest diameter rollers suggested in

the ASTM standards are recommended for use (3 mm diameter rollers). Considering their

recommendation, 3.2 mm rollers were used in this experiment. Additionally, Teflon tape

was used around the rollers as well as on the bone to diminish the wear [24]. In other

cases, especially when testing bone beams at 8,000 i the Teflon tape was not enough to

eliminate the formation of grooves. For the higher strain levels, a sacrificial piece of

equine bone from the dorsal aspect, having the same thickness and width as the bone









beam but measuring only 10 mm in length, was glued with cyanoacrylate to the surface

of the bone. The bone sacrificial piece and the beam were then covered with Teflon tape

(Figure 2-9).

All tests took place in a temperature controlled aqueous solution, in a cylindrical

plexiglas container. Temperature plays an important role in fatigue testing of bone [7].

Fatigue life has been found to depend significantly upon temperature of bone [15]. These

data suggest that a linear relationship exists between the number of cycles to failure and

range of temperature. For example, the fatigue life at room temperature (220 C) is one

half that at physiologic (370 C) for bone.

Additionally, it seems reasonable that if any advantageous structural adaptation

exists near the foramen regarding toughness, then such adaptation may give the best

results at physiological temperature. All beams were tested at a physiological

temperature of 370C in a calcium buffer solution to prevent leakage of calcium ions [26];

otherwise, leakage of ions can cause a reduction in modulus of nearly 2.5% [26]. A

heater (Fisher Scientific Automerse Immersion Heater Model 199) was used to control

the solution temperature at 370C. The testing solution consisted of 0.9% saline solution

with a concentration of 57.5 mg/L of CaC12 [26]. The specimens were allow to thaw for

2 hours and acclimate at least 45 minutes in the bath prior to testing.

Mechanical Testing Plan

From the five pairs of metacarpi (12 bones) procured, a total of fourteen beams

were mechanically tested. Four were dorsal specimens used for protocol development

only. The remainders were palmar beams.
















Foramen Hole


Support Span (L)

P
Loading
I Span
a -4


1/2 (L a)


P

-F


S
1/2 (L + a)


Oblique view of free body diagram for a beam in four point bending (A).
The loading span is 80mm and the support span is 40mm (B). A transverse
shear diagram shows a region of zero shear between the loading forces and a
bending moment diagram shows a region of constant moment (D).


C p


Transverse
Shear


D Pa t


Figure 2-9:


Bending
Moment


////// //////


Control Hole


Drilled Hole


I


t


P






































Sacrificial
bone pieces



C



----- ---- I e


Figure 2-10: Actual tested beam. Locations where the Teflon tape has worn out down by
the friction of the loading noses during cyclic testing (A). Cartoon
demonstrating the manner in which the beams were held by the rubber
bands to the support noses (B), and the location of the sacrificial bone pieces
to prevent grooves from forming. Figure B shows a top view, and figure C
shows






















































Figure 2-11: Testing setup (top). Custom made four point bending fixture, note the pivot
point, and the way in which the rubbers bands hold the loading noses to allow
for rotation (bottom).












Table 2-1: Testing plan. Repeating letters under the horse column signify the bones
come from the same horse. Four beams were procured from the palmar aspect
of horse S (two from the right leg and two from the left leg.) Of importance is
the specimen ID, for it will be mentioned throughout this study.

Number Horse Side Specimen ID Test Cyclic Strain Level
1 S Right SRA Fatigue 5000
2 S Left SLA Fatigue 8000
3 S Right SRB Monotonic
4 S Left SLB none control
5 A Right AR Fatigue 8000
6 A Left AL Fatigue 5000
7 G Right GR Fatigue 5000
8 G Left GL Fatigue 8000
9 L Left LL Monotonic
10 L Right LR Fatigue 5000
11 I Left ILB Fatigue 8000
12 I Left ILA Lost ---

Monotonic Tests

The monotonic tests were done to failure, in displacement control at a rate of 1

mm/sec. The elastic moduli were calculated from the load-deflection data and used to

estimate the required deflections for cyclic strain levels.

Fatigue Tests

The fatigue tests were performed at a frequency of 2 Hz using sinusoidal

waveform. The nominal cyclic strain level was calculated form the beam theory and the

elastic modulus estimated from monotonic tests. Then the beams were loaded to

approximately 10 N for the purpose of holding them. The test machine controller was set

for the displacement corresponding to the nominal strain level desired. The data

acquisition rate was 40 Hz. The data were recorded following a burst acquisition method,

which saved 6 cycles every 100 cycles. Tests were terminated at 60,000 cycles hourss

and 20 minutes) unless the beam failed before that.









Staining

Basic fuchsin staining is widely used to separate induced from artefactual

microcracks in bone. An established protocol [9] served as a guide to stain the loaded

beams. Immediately after the specimens were tested, they were placed in 1% basic

fuchsin (J.T. Baker, Basic Fuchsin) and 70% ethanol. The ethanol concentrations were

increased in each successive step (Table 2-2). Each step lasted 24 hours, and at least the

first 2 hours were spent under a 20 mm Hg vacuum to enhance penetration of the stain

(Figure 2-12). The complete process takes 11 days. Also, it is important to note that the

solution can not be reused; since the bones become dehydrated by releasing water and

reducing the concentration of the ethanol, this can reduce the uptake of stain.

Sectioning of Beams for Microscopy

After the bones beams were stained they were sent to a commercial processing

laboratory (Pathology Associates' Advance, North Carolina), where eight 100 um thick

sections were prepared from relevant regions. Three sections were prepared in the

vicinity of both loaded holes, and two sections near or through the control hole. In this

his way, sections 1, 4, and 7 were suitable for comparison (Figure 2-13). Sections 3, 5, 6,

and 8 did not contain a hole, but are important when analyzing the microdamage in the

proximity of these holes. All the beams were chamfered on the overhanging unloaded

lateral endosteal corner, so that the commercial processing laboratory could identify the

correct orientation of the specimens. Regardless of the precautions and explanations

given to the laboratory, they were unable to provide with certainty the orientation of the

slides. The correct orientation of the slides was determined after laborious observation of

the microstructure and by looking at the reconstructed embedded beams.









Table 2-2: Staining procedure.


Figure 2-12: Stained bone beam.


3 sections at each of these holes
r-1 rA-


876 543


H "


Schamfer
2 1


5 mm
----


125 mm


glass coverslip


section


1" x 2" glass
microscope slide


Figure 2-13: Plan view of bone beam, shown with holes near each end as they exist in the
actual beam. The transverse lines represent the location of desired sections
(top). Bone section mounted on slide.


1. 24 hours inl% basic fuchsin in 70% ethanol
2. 24 hours inl% basic fuchsin in 80% ethanol.
3. 24 hours in 1% basic fuchsin in 90% ethanol.
4. 24 hours in 1% basic fuchsin in 100% ethanol.
5. 24hours in 1% basic fuchsin in 100% ethanol.
6. Remove beams and allow to air dry for 2 days.
7. Rehydrate in deionized water for 4 days.









Microscopy Section Analyses Setup

Two of the fatigued beams were analyzed with respect to morphometric and

damage parameters describe subsequently. A total of three sections per beam were

analyzed: one through the foramen and one through the drilled hole, and one through the

control hole. The morphometric parameters included microstructural characteristics

intrinsic to the bone. The damage parameters included quantitative measures of damage.

Images of the entire bone sections were taken, and put together in a sort of

mosaic, denoted as an M-image. The M-image is a paper banner composed of three

printouts and used as a map to locate important features and for organization. The

images that comprised the M-image were taken using a brightfield microscope (Olympus

BX-60 Melville, N.Y.), a digital color camera (Cohu CCD RGB), and a frame (Integral

Technologies, Flashpointl28). The microscope was used with a 4X objective, due to the

magnification of the camera the total magnification was 142.5X. Using a square grid as a

guide, images of each bone section were taken every 1 mm. A complete M-image was

comprised of 30 images, or 2 rows of 15 images. These images were cropped to square

millimeters and put together using Matlab. The M-image was used extensively:

morphometric parameters were measured with it and it served as a fixed frame of

reference where the location of every single crack was noted.

Initially, each section was divided in the mediolateral direction into 10 regions;

The 10 regions were identified considering their proximity to the hole. (medial D, medial

C, medial B, medial A, medial hole, lateral hole, lateral A, lateral B, lateral C, and lateral

D). Each region measured 1 mm in length and the height of the beam (approximately

2mm), except for the two regions next to the holes (the near regions) which measure 0.50

mm in length (Figure 2-15). The reason why the near regions are shorter in length is









because next to the foramen a more involved microstructure expected and subsequently

noted (Figure 2-15). The 10 regions were aggregated into far (medial B, C, D and lateral

B, C, D) and near (medial hole, medial A, lateral hole, and lateral A) regions based on

preliminary statistical comparisons.

Each section was divided in the endoperiosteal direction (through the thickness)

into regions according to the mode of strain experienced by the beam; tension,

compression or central (low tension and low compression). The tensile and compressive

regions measured 0.5 mm in thickness each, and the central region measured 1 mm in

thickness.

Morphometric Parameters

Morphometric parameters are intrinsic to the bone and should reflect adaptation to

its loading environment in vivo. The morphometric parameters quantified on the M-

images were porosity and osteon density and osteon diameters Detailed explanations on

the method for acquiring the images and eventually their quantification can be found in

the Appendix; herein, a brief overview of the process is provided.

Transparencies, on which a 20 pm x 20 rm grid was printed, were taped on top of

the M-images to count and mark the parameters. The M-images were scanned for the

number of Haversian canals, the length Volkmann's canals, and the area of resorption

spaces. All these values were put into the following expression to determine porosity.

SPo y area of voids
% Porosity =
total area










# of Haversians canals x n 50um2
( 2 )
+ length of Volkmann's Canals (/um) x 50/um
+ area of resorption spaces (//m2)


total area (//m2)


It was assumed that the diameter of all Haversian canals and the Volkmann's canals

were 50 [m wide. The areas of the resorption spaces were counted, explicitly if they were

larger than 0.250 [tm2

The osteon density was calculated form the number of Haversian canals in a

region. Finally, the osteon diameters were the last of the morphological parameters

quantified. Using Matlab, scan lines were placed on the M-images in the middle of each

region, in the endoperiosteal direction (Figure 2-14). Using the brightfield microscope

and 10X objective images where successively taken, following down the scan lines.

These images where not put together or cropped, they were simply printed out and

osteons were measured on them with a ruler (following some scaling). Not all osteon

were measurable, only those that followed a specified criterion [33] described below.

The reason for this is that the cement line must be clearly visible to measure osteon

diameters.

The criteria for choosing an osteon and measuring its diameter:

1. secondary or primary osteons with presence of a clear cement line

2. circular or elliptical shape

3. refilling completed or almost completed

4. all osteons were measured through the Haversian canal in perpendicular and
transverse directions only










5. two diameters were measured if their shape was elliptical





Control Section
Perioendosteal
(compression-tension)






^ Proximodistal or (hole type)
x BDrilled section



Foramen section
Mediolateral or (proximity
to the hole)

Figure 2-14: Three sections quantified for compositional and microcrack parameters.

Damage Parameters

One group [48] defined micro cracks damage in three sizes: microcracks,

mesocracks and macrocracks. Microcracks are smaller than 50 rim, mesocracks are

between 50 im and 200 rm, and macrocracks are larger than 200 rm. The same group

noted that the magnification used to observed or quantify damage is a defining factor,

because the origin of the damage could be as small as 3 .m long cracks. Before settling

on a method for quantifying cracks, the bone sections are observed and different types of

microcracks were inspected, and a protocol was prepared that included the relevant and

repeating damage characteristics.

In the six sections observed, there were five types of repeatedly, identifiable

cracks. These were bundles of microcracks (BM), black wispy mesocracks (BWM),

diffused stained mesocracks (DSM), intra-osteonal microcracks (IOM), and cracks along









cement lines (CLC) (Table 2-3, Figures 2-16 through 2-21). The BM, BWM, and DSM

are all linear cracks, and IOM and CLC are parallel to the lamellar layer of the osteons,

so these were measure in arc lengths via the radius and angle (Figures 2-19 through 2-

20).

The black wispy cracks were seen in either two ways: as small groups of cracks

less than or equal to 20 [im in length (BM) (Figure 2-16), or as single cracks usually

longer than 20 [tm (BWM) (Figure 2-17). Diffuse stained mesocracks were thicker, and

very much stained (Figure 2-18). These last cracks were carefully quantified, paying

special attention to lamellar orientations, so not to be confused with stained Volkmann's

canals. The cement line microcracks (CLC) were either deeply stained red, or looked

scalloped and black (Figure 2-21). The intra-osteonal microcracks (IOM) composed of

stained wedges (Figure 2-19) and debonding lamellar layers (Figure 2-20) were clear.

There were areas of increased stain uptake especially observed in interstitial bone, where

the population of osteocyte lacunae is high. These were not quantified, since no linear

cracks were present. Often the BM were present in these areas, and those were quantified.

Specific Examples of Damage Types and Quantification

A transparency with scaling marked on it (Figures 2-16 to Figure 2-21), worked

as an outside eyepiece that measure length and angle of the cracks. For quantifying

microdamage three essential tools used were a microscope, the frame grabber software, a

CCD camera, two computers and a transparency placed on the monitor of one computer

that served to measure the length and the crack orientation. On one computer the live

images were displayed at a magnification of 356.25X, and at this magnification the

cracks were measured. Another computer had the M-image of the section and in this one,

the coordinate location of every crack was recorded. The data recorded was length of









each crack, the angle, the type, and the x-y coordinate location for each crack. The

cracks were quantified per region by a blind observer, and all the data was put into spread

sheets, which later differentiate endosteal, periosteal, or center located cracks. Later a

custom code was used to process the data into total crack density, length per area, and

angle.

Table 2-3: Crack types and descriptions.
Crack Type Acronym Description

Bundles of
BM Damage in interstitial bone, comprised of five, 20
microcracks [lm long, wispy black cracks (Figure 2-16).

Black wispy Individual cracks in interstitial bone, from
BM interstitial space to osteons, or osteons to osteons
BWM
mesocracks or into Volkmann's Canals; longer than 20m
wispy black cracks (Figure 2-17).

Diffused Cracks from interstitial bone to osteons, or osteons
in D to osteons, or into Volkmann's Canals; longer than
stained DSM
20km wispy black cracks; longer than 40km and
mesocracks stained red (Figure 2-18).

Intra-

osteonal IOM Stained wedges (Figure 2-19)and lamellar
debonding (Figure 2-20).
microcracks

Cement lines
CLC Loopy black or stained cement line cracks
microcracks (Figure 2-21).











Region


C .- .!I ^ --










Far Near



















Figure 2-15: The three figures show the M-image. Notice the region, and how they are identified by a letter, the scan line (orange)
when counting osteons sizes. Finally, notice the transparencies and how they were used.


























I I



'c r ',







-- --- -


I .

t 'M^L^


Figure 2-16: Example of the damage type: bundle of microcracks (BM). The five
microcracks represent one bundle. Notice they are all about the same size,
and they appear perpendicular to the side of the osteons






























m b -% k." -*-


Figure 2-17: Example of black wispy mesocracks (BWM), denoted by the arrows. The
cracks measures 60 .im and it is oriented -60 with respect to the neutral axis.
Notice that the crack expands between two osteons.






































Figure 2-18: Example of diffused stained mesocracks (DSM), indicated by the arrows. It
measures about 60 .im and is oriented -30 with respect to neutral axis.






































Figure 2-19: Example of an intra-osteonal crack (IOM), specifically a case of a stained
wedge. The wedge is the crack; it is identified for an increase in stain intake
compared to its surrounding matrix. The angle of the wedge (0) is
approximately 600 and the radius (r) of the osteon is measured to be about 80
rnm. The arc length (s) is then calculated from (s=rO).































VY


Figure 2-20: Example of an intra-osteonal crack (IOM), specifically a case of lamellar
debonding. The angle measure (0) is 3600 and the radius (r) of the debonded
lamellae is about 40 jim. The arc length (s) is then calculated, (s=rO).





































Figure 2-21: Example of a cement line crack (CLC), it is the black wispy line
surrounding the osteon. The angle measure (0) is 3600 and the radius (r) of
the osteon is about 40 im. The arc length (s) is then calculated, (s=rO).

Statistical Analyses

Due the amount of details involved in analyzing these sections, an organization

scheme was developed, in which independent and dependent variables were explicitly

identified. The independent variables are those that describe the source or origin of the

bone section. They include side (right or left metacarpus), the number of cycles endured,

the strain level (5,000[e or 8,000[e), hole type foramenn, drilled or control), proximity to

the hole (far or near), and endoperiosteal location (tension, compression or central). The

dependent variables divide into morphometric and damage parameters and describe the

microstructure of the bone and changes introduced in the testing process, respectively.









Morphometric parameters are porosity, osteon density and osteon diameters. The

damage parameters crack density (#/mm2), crack length (mm/mm2), and crack angle.

The primary goal of the statistical analyses of the damage measured per region

was to compare specific damage amongst to the three holes types foramenn, drilled hole,

or control hole). All statistical analyses were performed with commercially available

software (StatView; SAS Institute; Cary, North Carolina). The confounding effects were

proximity to each hole (near or far) and applied strain level (5,000 tp, or 8,000 tp,). The

types of physical damage, the hole type, and the two confounding effects were accounted

for by performing a full interaction analysis of variance (ANOVA). If no significance

resulted in the interaction, the strain level factor was eliminated from the model and the

ANOVA repeated with only the hole type-proximity interaction active. Consequently, if

no significance resulted, then the proximity factor was eliminated, leaving only the hole

type factor. At any stage of the analysis, if a significant ANOVA resulted, Fisher's

protected least significant difference tests were performed to determine if significant

individual comparisons existed. For all analyses, significance was assumed to exist for P

< 0.05.

To perform statistical analyses on the cracks orientation, another database was

constructed in which the experimental unit was each individual crack. Angles were

measured for black wispy mesocracks (BWM), diffused stained mesocracks (DSM), and

bundles of microcracks (BM). All angles were measured with respect to the neutral axis

(NA) in each section; four levels were chosen, 00, 300, 600 and 900. By limiting the

alternative of angles to only four possible outcomes, the response variable (angle),

becomes a nominal response variable. Statistical analyses on the orientation of the linear






34


microcrack and spatial occurrence (both nominal variables) were done using Contingency

Tables Analyses (determines the relationship between nominal variables). Chi square

tests for independence are reported, which test the likelihood that response variables are

independent of each other. A small chi square P-value (significance ifP<0.05) would

show a significant relationship between certain response variables exists.














CHAPTER 3
RESULTS

The beams refer to in the testing plan (Figure 2-1) were either monotonically or

cyclically tested. In a meaningful way they were arranged for a strain level. For

example, four beam were harvested from horse S, and on was assigned to monotonic tests

(SRB), two were assigned to undergo cyclic tests, 5000 i and 8000 ig respectively

(SRA and SLA), and finally one was left as a control specimen (SLB). Table 3-1 shows

the results of the two monotonic tests performed.

Table 3-2 shows the results of the fatigue tests performed. For two specimens

relaxation tests were attempted, and the resulting data were lost. Evidently, fatigue tests

at the strain level of 5,000 [i did not affect much the modulus of the specimen. The

specimen GL (Table 3-2 and Figure 3-1) failed unexpectedly at only 3,254 cycles through

the foramen. Specimen GR (Table 3-2 and Figure 3-1) endured almost half a million

cycles and did not fail. AR specimen (Table 3-2 and Figure 3-1) also failed due to the

contact forces between rollers and the bone beam, which created grooves. A strain-

number of cycles curve (Figure 3-1), shows the strain levels at which 10 beams were

loaded, and the number of cycles each tolerate. Two beams at 8000 [i reached failure

(AL and GR), while none of the specimens tested at 5000 [i failed.

Though Teflon tape was placed at the contact areas, grooves were created on the

bone beams. The grooves were much evident on the high strained specimens (8000[e)

(Figure 3-2), and not so dramatic on the lower strained specimens (5000[s). As












mentioned before, AR failed through the site of one of the grooves before the groove


problem was solved with the addition of the sacrificial bone piece.


0 LL LR X ILA
o SRB O SLA E AR
0 AL A GR A GL
SRA


A X->I>--

*- D-M.



10000
Number of fatigue cycles


100000000


Figure 3-1: Peak strain level versus number of cycles to failure ("S-N curve"). Arrows
pointing to right indicate test stopped prior to failure.


Table 3-1: Results for beam specimen tested monotonically to failure.


Specimen ID


LL

SRA


c yield F yield a ultimate F ultimate


88 MPa 14,000 [e

142 MPa 11,000 [e


90 MPa

147 MPa


20,000 [e

20,000 ue


Beam Apparent
Elastic
Modulus


7.4 GPa

16.6 GPa


24000




16000
g
E

-J
.E 8000





0










Table 3-2: Results for the remainder of the beams.


Strain
level

8,000 pte


Number Failure Mod
Modulus
of
Sl Reduction
Cycles
60,000 yes 4.7%


Initial
Modulus

12.7 GPa


Specimen
ID

SLA

SRB

SLB

ARt

AL

GL

GR

ILA

ILB

LR


Final
Modulus

12.1 GPa





12.6 GPa



11.5 GPa


5,000 pe 60,000 no none

-------------------Control beam-----------

8,000 pe 59,900 yes 11.8% 14.3 GPa

5,000 [e 60,000 no none

8,000 pe 3,254 yes 19.6% 14.4 GPa

5,000 pe 452,836 no none ---

8,000 pe 6,500 yes Lost data ---Relax

8,000 pe 26,000 no Lost data ---Loa

5,000 pe 21,000 no Lost data ---Relax


Figure 3-2: A groove present on an 8,000g, specimen (left). A permanently deformed
beam (right).


ation tests---

d control---

ation tests---









Morphometric Parameters

The morphometric parameters are porosity, osteonal density, and osteonal

diameters per region. There is no significant relation between right and left beams and

morphometric parameters, except near the drilled hole, which is significantly different in

porosity and osteonal diameter (P = 0.0059 and P = 0.0353, respectively). The porosity

(mean 6.1% standard deviation 0.9%) near the drilled hole in the right MC3's is lower

than in the opposing MC3 (7.3% 1.4%) (Table 3-3). Also near the drilled hole, in the

right MC3, osteon diameter (13.8%) is larger than in the left MC3 (Table 3-5).

There is a significant difference between hole types for all morphometric

parameters (porosity, osteon density, and osteon diameter) (Figures, 3-3. 3-4, 3-6).

Porosity is significantly related to the hole type (P<0.0001) and proximity to the

hole (P=0.0005). Fisher PLSD demonstrated that significance existed between foramen

and drilled (P<0.0001), and foramen and control (P<0.0001), but not between control and

drilled (P=0.1078) (Figure 3-3). The foramen has the highest porosity, while the drill and

the control holes porosities are similar. Near the foramen, the highest porosity is found

near it, while farther away the porosity is significantly lower (Table 3-3).

There exists a significant difference between osteon density, hole type (P<0.0001)

and the proximity to the hole (P<0.0163). Again, the significance relies on the structural

differences found in the vicinity of the foramen hole. Fisher PLSD demonstrated that

significance existed between foramen and drilled (P<0.0001), and foramen and control

(P<0.0001), but not between control and drilled (P=0.3578) (Figure 3-4). The osteon

density is higher in the foramen section, and highest near the foramen hole. For the









drilled hole and the control hole sections the osteon density remains similar, and there is

no marked difference between their far and near regions (Table 3-4).

A significant ANOVA resulted from the interaction between hole type (P=0.001)

and average osteon diameter, and no significant for proximity to the holes (P=0.0798)

and osteon diameter. The average osteon diameter is significantly smaller in the foramen

and drill sections, when compared to the average osteon in the control section. These

values correspond to a significant difference in average osteon diameters between the

control hole and the foramen (P = 0.008), and between the control and the drilled holes

(P = 0.0295), but not between the foramen and the drill holes (P = 0.1948) (Figure 3-5).

Investigating the near and far regions in each hole type, there is no significant ANOVA

for average osteon diameter in proximity to the hole; however, the region near the

foramen has the smallest osteon diameters (near 0.152 mm 0.010 vs. far 0.171 mm +

0.012), followed by the drill section (near 0.165 mm 0.009 vs. far 0.173 mm 0.023),

while the near and far regions in the control hole are very similar (near 0.184 mm vs.

far 0.180 mm 0.019) (Table 3-5). Notice that there could be a structural trend for small

osteons in the region close to the foramen, and it could overlap longitudinally into the

region where the drill hole was placed.

Damage Parameters

As a reminder, damage parameters in the form of five different types of cracks

were measured near the foramen, drilled holes, and control holes: 1) bundles of

microcracks (BM), 2) black wispy mesocracks (BWM), 3) diffused stained mesocracks

(DSM), 4) intra-osteonal microcracks (IO), and 5) cement line cracks (CLC). Black,

wispy mesocracks, diffused stained mesocracks and bundles of microcracks are all linear










Table 3-3: Results for porosity and spatial locations.
Porosity
Far


Mean
8.00%
6.60%
9.10%


Std. Dev
1.40%
1.10%
1.70%


Near


Mean
8.00%
6.90%
15.90%


c


abc






Drilled Foramen


Std. Dev
1.20%
1.60%
5.50%


[] far
0 near


Figure 3-3: Graph for porosity and spatial location. The repeating letters symbolize
significant P- values between spatial locations. The bars represent the
standard deviation (Std. Dev).


Table 3-4: Results for osteon density and spatial locations.
Osteon Density (#/mmA2)
Far Near


Hole
Type
Control
Drilled
Foramen


Mean
18.5
16.4
23.2


Std. Dev
4.1
1.9
4.5


Mean
18.2
18.4
29.4


Std. Dev
4.6
3.1
5.6


Hole
Type
Control
Drilled
Foramen


.23
E .2
E
q .17
E
E .15
.13
o .1

.08
.05
.03
0


Control










`35

2 30
0
025
o
20

E
S15

10

5

0


liLT


ab T
_bB


Control Drilled Foramen


Figure 3-4: Graph for osteon density and spatial location.


.23
E .2
.17
E.15
.13
:3
.1
0)
O
.08
.05
.03
0


Control Drilled Foramen


Figure 3-5: Graph for average osteon diameter and spatial location

Table 3-5: Results for average osteon diameter and spatial locations..
Osteon Diameter (mm)
Far Near


Hole
Type
Control
Drilled
Foramen


Mean
0.18
0.173
0.171


Std. Dev
0.015
0.023
0.012


Mean
0.184
0.165
0.152


Std. Dev
0.023
0.009
0.01


n far
O near


* far
E near










cracks that occur predominantly in interstitial spaces, between osteons or from

interstitial areas into osteons. Intra-osteonal microcracks are made up of diffused stained

areas inside of osteons, either lamellar debonding or stained wedges.

There were 1822 cracks accounted for in the six bone sections. A total of 884

(48.51%) were found in the 5000 ig beam, and 938 (51.48%) were found in the 8000 i

beam. 679 (37.26%) were found in the foramen sections, 654 (35.89%) in the drill

section, and 489 (26.83%) in the control section. Near the holes foramenn, drilled, and

control) 470 (25.80%) cracks were quantified, and far from the holes 1352 (74.20%) were

quantified. In the same way, 1045 (57.35%) appeared in the region defined as the center

region, 422 (23.16%) in the closer to the endosteal surface (tension region), and 355

(19.48%) closer to the periosteal surface (compression region). Of the 1822 cracks 475

(26.07%) were identified as BM, 132 (7.24%) identified as BWM, 150 (8.23%) as DSM,

65 (3.58%) as IOM, and 1000 (54.88%) as CLM.

Bundles of Microcracks

Due to the characteristics describing this type of damage; that is, groups of

approximately five 20 [m long cracks (which add to 100 pm), the analysis is done on the

number of bundles of microcracks in each region, rather than on the length of this

microcrack per region.

No significance resulted from the ANOVA between strain level and the BM

(P=0.3205), as well as for the interaction between hole type and proximity to the hole

(P=0.1160). However, there is a significant relationship between BM and hole type (P=

0.0002) (Figure 3-6), and between BM and proximity to the hole (P= 0.0002) (Figure 3-

7).









Fisher's PLSD tests demonstrated that significance difference exist between

drilled and foramen holes (P=0.0003) and between the drill and control hole (P 0.0002).

No significant differences exist between the foramen and control holes (P=0.6172)

(Figure 3-6). The drilled holes contain many more cracks of this type (273) in its vicinity

compared to the foramen (121) and the drilled hole (81) (Figure 3-8). Concerning the

proximity to the hole, there are significantly more BM far (382) than near (93) the

hole (P 0.0001) (Figure 3-8). Table 3-6 accounts for the mean and standard deviation

(Std. Dev.) of the length per area for the BM in far and near regions for each hole type.

Notice that there are over twice the amount of BM in the drilled when compared to the

foramen or the control hole.


Table 3-6: Results for density of BM and spatial location.


Hole Type
Control
Drilled
Foramen

18
16
14
c 12
S 10-
8
S 6
Ie 4
2-
0


Mean
6.8
17.6
9.2


Bundle of microcracks (#/mm2)
Far Near
Std. Dev. Mean Std. Dev.
5.3 5 2.2
5.1 7.8 2.8
6.6 2.2 1.1


T a


Control


Drilled


bi


Foramen


Figure 3-6: Graph of density of bundles of microcracks (BM) and hole type. Significance
exists between hole type and density of BM.












16

14
a
12

10

8-
a
O 6

) 4
4--

2

0
far near



Figure 3-7: Graph of density of BM and proximity to the hole. A significance ANOVA
results form BM and proximity to the hole.





225
200

o 175
0 150- 121
0
E 125- far
0 81 [] near

75 -
50

25

Control Drilled Foramen

Figure 3-8: Number of BM split by proximity to the hole and hole type.










Black Wispy Mesocracks

The full interaction ANOVA for hole type, proximity to the hole and strain level

resulted not insignificant for BWM (P= 0.6812). The same is true for the interaction

between hole type and proximity to the hole (P= 0.3898). However, a significance was

seen for the hole type along with and BWM (P= 0.0126) (Figure 3-9).

Fisher's PLSD test demonstrated a clear significance between hole types and

BWM. Significance between drill and foramen sections (P=0.0151), the drill section has

more damage than the foramen section. Also, there is a significant difference between

drill and control (P= 0.0268), and the drill section also leads in damage. The drilled hole

has over 50% more crack length per area over the foramen, while the foramen only has

33% more crack length per area when compared to the control section (Table 3-7). In

addition, there no significance between the foramen and control sections, the mean

difference between them being scarcely 0.054 mm/mm2 (Table 3-7).



.6

.5
a,b
.4
E
E


a

.1

0
Control Drilled Foramen


Figure 3-9: Graph for crack length for black wispy mesocracks (BWM) and hole type.











Table 3-7: Results on crack length per area for black wispy mesocracks (BWM).
Black wispy mesocracks
(mm/mm2)
Hole
Type Mean Std. Dev
Control 0.1 0.07
Drilled 0.34 0.32
Foramen 0.15 0.13

Diffused Stained Mesocracks

The data do not reveal any significance relating this type of damage and strain

level, proximity to the hole, or hole type. A significant ANOVA results from the

interaction between hole types and proximity to the hole (P= 0.0188). This significant

interaction rejects the null hypothesis that the effect of hole types is indifferent to

proximity of the hole concerning diffused stained mesocracks (DSM) (Figure 3-10). For

this type of damage there is a 50% increase in crack length per area in the vicinity of the

foramen when compare to the drilled hole, while the mean crack length for the rest of the

hole type remain similar (Table 3-8).


.45
.4-
.35
.3
E .25 far
1 .2 [] near
.15
.1
.05
0
Control Drilled Foramen


Figure 3-10: Graph for diffused stained mesocracks (DSM), hole type and proximity of
the hole.









Table 3-8: Result on crack length per area for diffused stained mesocracks (DSM).
Diffused stained mesocracks (mm/mm2)
Far Near
Hole
Type Mean Std. Dev Mean Std. Dev
Control 0.19 0.11 0.12 0.08
Drilled 0.11 0.07 0.1 0.06
Foramen 0.11 0.07 0.26 0.19

Intra-Osteonal microcracks

Arc length per area is the measure corresponding to this type of damage. For the

case of intra-osteonal microcrack and the interaction of hole type, proximity to the hole

and strain level, there is no significant ANOVA (P=0.4896). After the strain level factor

is eliminated from the analyses, significance results from the interaction of hole type and

proximity to the hole (P=0.0165). The single ANOVA for hole type is low but not

significant (P=0.0671), as expected, the single significance analysis for intra-osteonal

cracks and proximity to the hole is relevant (P=0.0007). As a consequence Fisher's

PLSD test for proximity to the hole uncovers a significant value between near and far

regions (P=0.0002) (Figure 3-11), where more cracks of this type are found in the near

regions, or the regions closest to the holes.

The low non-significant value in concerning the hole type, further suggests that

additional analysis is necessary. Effectively, Fisher's PLSD uncovers a significance

between foramen and control sections (P= 0.0067) and between foramen and drilled

sections (P= 0.0192) (Figure 3-12). In both cases the foramen section shows more intra-

osteonal microcracks (IOM), or arc length per area than the drilled and the control

sections. There is no significant discrepancy between drill and control sections, though

the drill section exposes more length per area of this type of damage. Far away from the










hole the arc length per area is highest in the drilled section, while near the holes, the

foramen has the highest arc length per area for this type of damage (Table 3-9).


a


near


Figure 3-11: Interaction plot for intra-osteonal microcracks and the effects of proximity
to the hole.


1 -
9-
.9
CM .8
E .

~..6
1.5
< .4
.3-
.2-
1-
.1
0-


a-


Control


Drilled


Foramen


Figure 3-12. Graph for intra-osteonal microcracks (IOM) and the effects of hole type

Table 3-9: Results for intra-osteonal microcracks (IOM).
Intra-osteonal microcracks (arc length/mm2)
Far Near


Hole
Type
Control
Drilled
Foramen


Mean
0.06
0.21
0.1


Std. Dev
0.04
0.19
0.05


Mean
0.31
0.34
0.88


Std. Dev
0.32
0.31
0.39









Cement Line Cracks

As in the case of intra-osteonal cracks, arc length per area is the measure analyzed

in cement line cracks. There is no significance in the interaction of hole type, proximity

to the hole, and strain level (P=0.733). No significance results when the strain level is

eliminated from the model (P=0.184). Significance results in relation to the hole type

(P=<0.0001). Further analysis reveals a marked difference between foramen and drilled

sections, and between foramen and control regions (for both cases P= <0.0001), where

the foramen CLC lead considerably (Figure 3-13). No significance is expressed between

the control and the drill regions; unexpectedly, the control hole shows a slightly higher

value in arc length per area than the drill hole, while the foramen has the highest amount

of cracks of this type (Table 3-10).


6

5 a,b
E

a b
-3

2



0
Control Drilled Foramen


Figure 3-13: Interactive bar plot for cement line cracks and the effect of hole type.

Table 3-10: Descriptive statistics for cement line cracks.
Cement line cracks
Std.
Hole Type Mean Dev
Control 2.54 0.82
Drilled 2.5 0.74
Foramen 4.3 1.8










Linear Cracks and Angle Frequencies

Angles were measured for linear cracks only. These include BWM, DSM, and

BM. As a reminder, all angles were measured with respect to the neutral axis (NA) in

each section; and four levels were chosen, 00, 300, 600, and 900. By limiting the

alternative of angles to only four possible outcomes, the response variable angle becomes

a nominal response variable.

Contingency table analyses for angle and type of damage reveal a significantly

low Chi-square P value (P=<0.0001). The significance predominately relies on the

bundles of microcracks damage type. Out of 475 BM quantified, 469 were oriented in

the range on 60-900 from the neutral axis. The remaining linear cracks (BWM and

DSM) are more evenly distributed, with higher amounts found 0 from the NA axis (38%

and 35%, respectively).


Table 3-11: Angles summary.
Observed Frequencies for Angle, Type of Damage
0 30 60 90 Total
Bundles of Microcracks 2 4 178 291 475
Black Wispy Mesocracks 51 17 36 28 132
Diffuse Stain Mesocracks 53 26 40 31 150














CHAPTER 4
DISCUSSION

The differences in measured morphometric parameters (porosity, osteon density,

and average osteon diameter) in the vicinity of the natural and drilled holes represent

mechanical adaptation experienced while the animal was still alive. The parameters near

the foramen were unmistakably different than near the drilled and control holes. Near the

foramen existed the highest porosity, the highest osteon density, and the smallest osteons.

Furthermore, the regions nearest the foramen were regions of yet the highest porosity, the

highest osteon density, and the smallest osteons. It was expected that such morphometric

differences would exist and was further expected that they would play the role that they

did in terms of evolution of damage in each region.

The porosities reported in the current work in the vicinity of the foramen, the

drilled hole, and the control hole (approximately 12%, 7%, and 8%, respectively) are

higher than those reported by others [35] in specimens from the dorsal, medial, and

lateral aspects of the MC3 (approximately 4% to 5%). The highest porosity near the

foramen is primarily due to resorption spaces, and indication of current remodeling for a

mature animal not yet in senescence [16]. Similarly local regions of relatively high

porosity were found in earlier work on the foramen [22], and played a major role in

"routing" the higher stresses away from the foramen into regions of lower porosity and

higher strength.

The highest osteon density and the smallest osteons were found nearest the

foramen, and these parameters go hand in hand. Higher osteon density and smaller









osteons directly imply more cement lines per region of bone. A significant literature

exists that promotes cement lines as feature that arrest and deflect microcracks in

osteonal bone [2,13,15,23,28,45,47].

Mechanical testing was performed to generate damage so that correlations could

be drawn between it and the morphometric parameters. To that end, methods of test

preparation were developed that exposed the foramen and a drilled hole to the same

mechanical duress, that minimized damage due to processing, and provided a control for

existing damage. Specifically, the foramen and the drilled hole resided in the constant

moment region of a four point bending beam. The drilled holes (as well as the control

hole) were meticulously sculpted to provide the same topology as the foramen and to

remove surface damage due to their drilling. Finally, the control holes were located in an

overhanging unloaded region of each beam. In this way, the damage states near the

foramen and the drilled hole could be "normalized" with respect to the control hole.

Mechanical testing was performed on more beams than were analyzed

morphometrically and for damage. The reason for this was due to the laborious and time

consuming nature of the analyses. Nevertheless, interesting conclusions can be drawn

from the mechanical test results themselves. The apparent initial elastic modulus of the

beams containing the foramen in the current work was 14.2 + 1.4 GPa (mean standard

deviation). The apparent initial modulus of four point beams from the palmar aspect of

the equine MC3 (but not containing the foramen) has previously been reported to be 14 +

2 GPa [6]. The similarity between the current and previous results is encouraging and

reflects the local nature of the special microstructure near the foramen.









Another result from the mechanical tests is that of modulus reduction after cyclic

loading. Modulus reduction has been shown to be a measure of the mechanical

manifestation of damage in bone. The beams tested at the 5,000 i strain level exhibited

limited reduction in modulus. The beams tested at 8,000 gi exhibited a slight but clear

reduction in modulus. Others have observed similar results [1]. Three point bending

beams from the dorsal, lateral, and medial aspects of the MC3 strained at 5,000 gi were

reported to exhibit no reduction in modulus [34]. Conversely, four point bending beams

from the dorsal, lateral, and medial aspects of the MC3 strained at 10,000 i, exhibited

dramatic reduction in modulus, with failure imminent after just 10,000 cycles of loading

[21].

Another important aspect of this work was the implementation and improvement

upon existing protocols for staining and quantifying microcrack damage in bone. The

observations damage of this work agree with those of other researchers interested in

fatigue damage in equine bone [35], fractures in human metatarsals [19], and the effect of

damage on the mechanical properties in bone [8,12,43]. These other researchers

presented images and comments describing the appearance of damage that were of great

assistance in the current work [19, 30, 41].

Two types of damage bundles of microcracks and black wispy mesocracks -

clearly relate to the special microstructure near the foramen. Others [35] have reported a

type of damage very similar in character to what was defined as bundles of microcracks

in the current work. In the previous work, they were called "unstained cracks" and were

described as small cracks on the order of 25 im. These unstained cracks were

perpendicular to the periosteal surface (900 from the neutral axis of bending in parlance









of the current work), and near the periosteal facing sides of osteons. They were also

reported to be longer (by over 17%) in mechanically tested specimens compared to

untested controls. Bundles of microcrack in the current work appeared black. Nearly all

were in proximity to osteons and directed 900 from the neutral axis. Significantly more

cracks of this type were found near the drilled hole compared to the foramen.

Individually, the foramen and the drilled hole each had a decreased number of microcrack

bundles further away in regions of decreased osteon density.

It was expected that the black wispy mesocrack type of damage would reflect the

local microstructure near each of the hole types. Given that the region nearest the

foramen exhibited the highest porosity, it would be expected that more of these cracks

would be found here. However, the opposite was found: it seems as if the high porosity

is mitigated by the combination of higher osteon density and smaller osteons. The drilled

hole does not posses the specialized osteon arrangement; near it, the crack length per area

was over 2 times greater than that near the foramen. In contrast, the crack length per area

near the foramen was only approximately 1.5 times greater than near the control hole.

The other damage types (diffuse stained microcracks, intra osteonal microcracks,

and cement line cracks) did not correlate with the expectations of decreased damage

nearest the foramen. These damage types are characterized by their ability to uptake

stain in higher proportions. It is also very possible that these damage types were in

actuality stained Volkmann's canals. Other researchers have noted this and have thus

given little credence as to these type of damage as reliable indicators of the true damage

state. Finally, regions of increased remodeling (and decreased mineralization)






55


necessarily uptake more stain [42], thereby increasing the perceived level of these

damage types in regions nearer the foramen.














CHAPTER 5
OSTEON TRAJECTORIES NEAR THE EQUINE METACARPUS NUTRIENT
FORAMEN

Introduction

The purpose of this study is to describe osteon trajectories near a natural hole the

primary nutrient foramen in the equine third metacarpus (MC3).

Materials and Methods

Four equine third metacarpi were obtained from an animal tissue service for this

and a related study. The bones came from skeletally mature animals less than five years

of age and without skeletal abnormalities. All procedures involving animal tissue use

was conducted under the approval and auspices of our Institutional Animal Care and Use

Committee. The bones were cleaned of soft tissue and stored at -260C until further

preparation.

Rough cuts were made first with a diamond blade band saw (Gryphon) to separate

the palmar from the dorsal aspects. Three longitudinal sections, 2 mm thick in the

endoperiosteal direction, were cut with a precision diamond blade band saw (Exakt) from

the palmar aspect with the foramen centrally located. One parasagittal section

(approximately 5 mm thick in the mediolateral direction) was cut from the palmar aspect

with the foramen centrally located. Each section was mounted on a petrographic glass

slide, ground (Buehler Minimet) with 600 grit silicon carbide waterproof sand paper in

combination with a 3 |tm diamond slurry (Buehler METADI Supreme) and polished

manually (Buehler TEXMET 1000) until no scratches were visible under the microscope.









Successive polishings were performed on the parasagittal section so as to observe osteon

trajectories while approaching the foramen mediolaterally. The bones and sections were

kept hydrated with purified water during all the processing and storage.

All sections were observed under a reflected light microscope (Olympus BX-60).

The most periosteal surface of each longitudinal section and the most lateral surface of

the parasagittal section was observed under 40X magnification. Digital images of

regions around each foramen were captured with a CCD camera (Cohu), installed on the

microscope and connected to a frame grabbing card (Integral Technologies

FlashPointl28). The images were cropped and merged to form mosaics of larger regions

(Figures 5-1 & 5-2, center images).

Results

Osteons with endoperiosteal trajectories were evident on the most periosteal

surface, specifically on the distal and proximal apexes of the nutrient foramen (Figure 5-

1). These osteons did not disrupt the lamellar layer lining the foramen evident in the

images and reported previously by us [22]. These osteons were bundled together within a

triangular shaped region, and were similar in size and elliptical shape. Osteons with

tangential trajectories were observed elsewhere and enclosed the foramen mediolaterally.

Observations on the parasagittal section yielded diverging osteon trajectories

(Figure 5-2). Osteons become apparent in the mediolateral direction, outlining the

foramen edges. Longitudinal osteons did not reach the edge of the hole, and some turned

to become endoperiosteal osteons observed in the longitudinal sections. Many of the

osteons were elliptical in shape.









Discussion

The gross shape and heterogeneous mechanical properties of the MC3 reflect its

function in response to evolutionary adaptation for efficient and safe high speed

ambulation [31]. Its length and large cortical cross sectional area give it an inherent

resistance to both buckling and static compression failure: safety factors (SFs) exceed 5

on buckling and approach 2 on compressive strength for horses in a trot [5,16], but only

slightly exceed 1 on fatigue strength [16]. The low fatigue SF reflects the high incidence

of stress fractures in racing horses, up to 70% in young thoroughbreds [37]. These

fractures occur on the dorsal aspect of the MC3, an area of predominant tension [37],

diametrically opposite the location of the foramen. Thus, the foramen exists in a region

of predominate longitudinal compression.

It is well known that a transverse tension stress field exists in the "up-" and

"downstream" regions near an elliptical hole in a plate subjected to far field longitudinal

compression. Such regions exist near the proximodistal apexes of the MC3 foramen. It

is in these regions that we observed endoperiosteal osteon trajectories, especially evident

nearest the periosteal surface of the MC3. The periosteal surface represents the surface

of greatest compressive bending normal stress, and, thus, the greatest transverse tension

in the foramen apexes. Osteons perpendicular to this tension present their cement lines as

possible crack arrestors and, along with diverging osteons which may deflect cracks, may

increase toughness in these regions.
































Palmar view of the equine metacarpus (left). A longitudinal section
containing the foramen (center). Magnified views of proximal (P) and distal
(D) apices (right). Osteons with endoperiosteal trajectories in the apexes are
evident by the presence of Haversian canals (HC) in these views, while
osteons tangent to the foramen edge are visible elsewhere. A lamellar layer
lines the foramen.


Figure 5-1.






























Lateral view of the equine metacarpus (left). Parasagittal section through a
portion of the foramen (center). The lightly "striped" regions are where
osteons with mediolateral (ML) trajectories and, thus, tangential to the
foramen were found. Osteons with longitudinal (L) trajectories were found
in the apexes of the foramen. Osteons with longitudinal and tangential
trajectories are found in these regions as well. Magnified views of the ML
region in the center image (top right). Mediolateral osteons are evident by
the presence of Haversian canals (HC) in this view. Osteons diverge in this
view (bottom right). Labeled are Haversian canals traversing in a general
longitudinal direction (L) and those in an endoperiosteal (EP) direction.


Figure 5-2.














CHAPTER 6
CONCLUSION

The ultimate outcome of the current work was to complete the description of the

foramen as a structurally optimized feature in bone. The previous work of colleagues

clearly demonstrated the static strength advantages of the microstructure near the

foramen [22,46]. These features were then mimicked in the design of an inhomogeneous

plate containing a hole [27]. The fabrication and mechanical testing of this plate design

with twice the strength of a homogeneous panel demonstrated the obvious advantage of

biomimickry [14]. The current work describes further the microstructure near the

foramen specifically with respect to its implication on fatigue behavior and local

toughening. The demonstration of reduced damage near the foramen compared to a

drilled hole under the same mechanical duress, then, hopefully completes the description

of the foramen as an optimized structural feature in bone.














APPENDIX
QUANTIFICATION PROTOCOL

1) Placement of the Slide
a) Place bone slide on Merz grid, separated by the thickness (1.4mm) of the double
sided sticky tape, and orient the bone section parallel to one of the Merz grid
lines.
b) Use the double-sided sticky tape to secure the bone slide in a firm position.
c) Record the orientation of the bone slide, if different from the usual; lateral to the
right, and periosteal toward the observer.

2) Images are observed on the microscope with a 4X objective. An image grabber
program (Flashpoint 3.1), and a CCD camera connected to the microscope are use to look
at the bone sections magnified 142.5X on the computer screen. Images are saved with
the frame grabber program, and with the aid of the Merz grid, each image is taken 1mm
apart.
a) Focus on the bone and take image.
b) Scan the entire length taking images, the bone section fits in two
rows of 15 images each.
c) Reaching final quadrant (a quadrant boxes in the Merz grid) take
an image then focus on grid and take an image. The image of the grid will be
useful defining the appropriate area to crop off the images when putting together.

3) Highlight placement
On printout, representing entire Mertz grid, highlight the quadrants in which the bone
was placed, and its medial and endosteal corner.

4) Producing "mother image"
a) Images entered into Matlab to be cropped
b) Run program to convert to larger image
c) Name this image the "mother image"

5) Placement of grid over mother image
a) Print out mother image, made up by three 8"xl 1" printouts. Each printout
is made up by ten images, which were cropped to a square millimeter and
put together with Matlab.
b) Align grid along both sides of the hole or center of the bone
c) Secure transparencies with tape
d) Note the orientation of the bone

6) The grid is composed of boxes measuring 400 trm2. From the center, the grid is









sectioned into two columns, on both sides of the hole, measuring 50mm each. From
there, the rest of the columns are 100mm in width covering the rest of the bone. The grid
also has three rows separating periosteal, center, and endosteal sections.

7) Area per section
a) Number of squares (measuring 10,000 [tmA2) covering bone are counted
b) This number is multiplied by 10,000 giving the area of bone for each
quadrant

8) Osteon count
a) Haversian canals are marked red with marker
b) Count number of red dots in each quadrant

9) Resorption space area
a) Resorption spaces are marked blue
b) In each quadrant, the number of small boxes (400 tm^A2) colored
blue are counted
c) The number is then multiplied by 400 giving the total resorption
space area for that quadrant

10) Volkmann's canal length
a) Volkmann's canals are marked yellow
b) Number of small boxes (400 tm^A2) colored yellow are counted
c) This number is then multiplied by 20 to give the total length of the canals for the
individual quadrant

11) The data collected for steps 8-10 is then entered into an Excel spreadsheet. These are
entered into the appropriate column and then divided by the area of bone found in that
particular quadrant on the grid.

12) Preparing mother image for osteon diameter measurements
a) Open mother image (6165x822) in Paint Shop
a. Locate where the lamellar layer ends, or the drill hole start, for the medial
and lateral sides
b) Record these numbers, in pixels, in orange cells on specific chart. The yellow
cells are the location, in pixels, of the scan lines for osteon diameter
measurements. The two cells under the yellow cells are the locations of the
boundary lines.
c) The numbers in the yellow and white cells must be entered in Matlab
d) Run the program

13) Labeling the sections
a) Open mother image, now containing scan lines and boundary lines, in
Paint Shop.
b) Label each section with a letter following the below lettering system:









Medial hole Lateral
.............. 200 100 50 50 100 200 ...............
F G H I J K MM M L LL OP QR S T

14) Print out mother image with appropriate lines and labels and images of individual
sections. Preview printouts of sections, circling possible osteons that might be measured
later.

15) Getting images of osteons
a) Looking at the slide at 10X magnification and under circular polarized light
b) Each section is scanned following scan lines as guides, for osteons present in the
boundaries.
c) Take images of osteons, scanning down each section, using printouts as
references

16) Measuring osteon diameters
a) Open images of sections in Paint Shop
b) Label secondary osteons with an "s"
c) Measure osteons according to guidelines stated below:

Rules for Measuring Osteons

Circular osteons 1 diameter (x-direction)
Elliptical osteons 2 (x,y) diameters

For Secondary Osteons
Clearly see the cement line (on x axis) or (x, y axis)

For Primary Osteons
Clearly see the cement line (if not on x and y directions), then only
on 450, 1350, 2250, 3150, 900, 1800, 2700 or 360.

All measurements must go thru the Haversian Canals

17) Print out images of the osteons with labels and measurement lines with a ruler.
a) Enter the numbers in the appropriate excel sheet.















LIST OF REFERENCES


1. Akkus O, Knott DF, Davy DT, Rimnac CM. Linear microcracks may contribute little
to degradation of Young's modulus in cortical bone tissue. Bioengineering
Conference; ASME 2001; 50: 79-80.

2. Akkus O, Rimnac CM. Corical bone tissue resists fatigue fracture by deceleration
and arrest of microcrack growth. J of Biomechanics; 2001; 34:757-764.

3. Anderson MA, Mann FA, Kinden DA, Wagner-Mann CC. Evaluation of cortical
bone damage and axial holding power of nonthreaded and enhanced threaded pins
placed with and without drilling of a pilot hole in femurs form canine cadavers.
JAVMA; 1996; 208(6): 883-887.

4. ASTM 790M-82. Flexural properties of unreinforced and reinforced plastics and
electrical insulating materials. 1982.

5. Bartel DL, Schryver HF, Lowe JE, Parker RA. Locomotion in the horse: a procedure
for computing the internal forces in the digit. American Journal of Veterinary
Research 1978:39(11):1721-7.

6. Bigot G, Bouzidi A, Rumelhart C, Martin-Roset W. Evolution during growth of the
mechanical properties of the cortical bone in equine cannon-bones. Med. Eng. Phys.;
1996; 18(1):79-87.

7. Bonfield W, Li CH. The temperature dependence of the deformation of bone. J of
Biomechanics; 1968; 1:323-329.

8. Boyce MT, Fyhrie DP, Glotkowski MC, Radin EL, Schaffler MB. Damage type and
strain mode associations in human compact bone bending fatigue. Journal of
Orthopaedic Research: 1998; 16:322-329.

9. Burr DB, Hooser M. Alteration to the en bloc basic fuchsin staining protocol for the
demonstration of microdamage produced in vivo. Bone; 1995; 17(4): 431-433.

10. Burr DB, Schaffler MB, Frederickson RG. Composition of the cement line and its
possible mechanical role as a local interface in human compact bone. J of
Biomechanics; 1998: 21(11):939-945.

11. Burr DB, Stafford T. Validity of the bulk-staining technique to separate artifactual
from in vivo bone microdamage. Clinical Orthopaedics and Related Research; 1990;
260:305-308.









12. Burr DB, Turner CH, Naick P, Forwood MR, Ambrosius W, Hasan MS, Pidaparti R.
Does microdamage accumulation affect the mechanical properties of bone? J of
Biomechanics; 1998: 31: 337-345.

13. Burr DB, Turner CH, Naick P, Forwood MR, Pidaparti R. En Bloc staining of bone
under load does not improve dye diffusion into microcracks. J of Biomechanics;
1998; 31: 285-288.

14. Buskirk SR, Venkataraman S, Ifju PG, RapoffAJ. Functionally graded biomimetic
plate with hole. Proceedings of Society for Experimental Mechanics Annual
Conference & Expositions, Milwaukee, WI, 2002.

15. Carter DR, Hayes WC. Fatigue life of compact bone -I. Effects of stress amplitude,
temperature and density. J Biomechanics: 1976; 9: 27-34.

16. Cheney JA, Liou SY, Shen CK, Wheat JD. Cannon-bone fracture in the thoroughbred
racehorse. Med Biol Eng; 1973: Sep 11(5): 613-20.

17. Cowin, Stephen C. Bone Mechanics Handbook. Boca Raton, CRC Press LLC, 2001.

18. Dallap BL, Bramlage LR, Embertoson RM. Results of screw fixation combined with
cortical drilling for treatment of dorsal cortical stress fractures of the third metacarpal
bone in 56 Thoroughbred racehorses. Equine Veterinary Journal; 1999; 31(3): 252-
257.

19. Donahue SW, Sharkey NA, Modanlou KA, Sequeira LN, Martin RB. Bone strain
and microcracks at stress fracture sites in human metatarsals. Bone; 2000; 27(6): 827-
833.

20. Frost HM. Preparation of thin undecalcified bone sections by rapid manual method.
Stain Technology; 1958; 33:273-277.

21. Gibson VA, Stover SM, Martin RB, Gibeling JC, Willits NH, Gustafson MB, Griffin
LV. Fatigue behavior of the equine third metacarpus: mechanical property analysis.
Journal of Orthopaedic Research; 1995; 13(6): 861-868.

22. Gotzen N, Cross AR, Ifgu PG, Rapoff AJ. Understanding stress concentration around
a nutrient foramen. Journal of Biomechanics; 2002; invited paper accepted for
publication.

23. Gou XE, Liang LC, Goldstein SA. Micromechanics of osteonal cortical bone
fracture. J of Biomechanical Engineering; 1998; 120: 112-117.

24. Griffin LV, Gibeling JC, Gibson VA, Martin RB, Stover SM. Artifactual
nonlinearity due to wear grooves and friction in four-point bending experiments of
cortical bone. J Biomechanics; 1997; 30: 185-188.









25. Gross TS, McLeod KJ, Rubin CT. Characterizing bone strain distributions in vivo
using three triple rosette stain gages. J of Biomechanics; 1992; 25(9): 1081-1087.

26. Gustafson MB, Martin RB, Gibson V, Storms DH, Gibeling J, Griffin L. Calcium
buffering is required to maintain bone stiffness in saline solution. J of Biomechanics;
1996; 29(9): 1191-1194.

27. Huang J, Venkataraman S, Rapoff AJ, Haftka RT. Optimization of axisymmetric
elastic modulus distributions around a hole for increase strength. Structural and
Multidisciplinary Optimization 2002; accepted for publication.

28. Jepsen KJ, Davy, DT. Comparison of damage accumulation measures in human
cortical bone. J of Biomechanics; 1997; 30: 891-899.

29. Jepsen KJ, Davy DT, Krzypow DJ. The role of the lamellar interface during torsional
yielding of human cortical bone. J of Biomechanics; 1999; 32: 303-310.

30. Lee TC, O'Brien JO, Taylor D. The nature of fatigue damage in bone. International
Journal of Fatigue; 2000; 22: 847-853.

31. Les CM, Stover SM, Keyak JH, Taylor KT, Willist NH. The distribution of material
properties in the equine third metacarpal bone serves to enhance sagittal bending. J of
Biomechanics; 1997; 30: 355-361.

32. Martin RB, Burr DB, Sharkey NA. Skeletal Tissue Mechanics. New York, Springer-
Verlag, 1998.

33. Martin RB, Gibson VA, Stover SM, Gibeling JC, Griffin LV. Osteonal structure in
the equine third metacarpus. Bone; 1996; 19 (2): 165-171.

34. Martin RB, Gibson VA, Stover SM, Gibeling JC, Griffin LV. Residual strength of
equine bone is not reduced by intense fatigue loading: implications for stress facture.
J of Biomechanics; 1997; 30:109-114.

35. Martin RB, Stover SM, Gibson VA, Gibeling JC, Griffin LV. In vitro fatigue
behavior of the equine third metacarpus: remodeling and microcrack damage
analysis. Journal of Orthopaedic Research; 1996; 14: 798-801.

36. McDonald DE, Palmer RH, Hulse DA, Neigut JS, Hyman WA, Slater MR. Holding
power of threaded external skeletal fixation pins in the near and far cortices of
cadaveric canine tibiae. Veterinary Surgery; 1994; 23:448-493.

37. Norwood GL. The bucked-shin complex in thoroughbreds. Proceedings of the
American Association of Equine Practitioners; 1978; 24:319-36.

38. Nunamaker DM, Butterweck DM, Provost MT. Fatigue fractures in thoroughbred
racehorse: relationships with age, peak bone strain and training. Journal of
Orthopaedic Research; 1989; 8: 604-611.









39. Pasquini Chris, Reddy V. Krishna, RatzlaffMarc H. Altas of Equine Anatomy.
Eureka, California, Sudz, 1983.

40. Rapoff AJ, Gotszen N, Cross AR, Ifju PG. Understanding stress concentration
around a nutrient foramen. BED-Vol. 50, 2001 Bioengineering Conference, ASME
2001.

41. Reilly CG Observation of microdamage around osteocyte lacunae in bone. J of
Biomechanics; 2000; 33: 1131-1134.

42. Rho JY, Zioupos P, Cirrey JD, Pharr GM. Microstructural elasticity and regional
heterogeneity in human femoral bone of various ages examined by nano-indentation.
J of Biomechanics; 2002;35:189-198.

43. Schaffler MB, Radin EL, Burr DB. Mechanical and morphological effects of strain
rate on fatigue of Compact Bone. Bone; 1989; 10: 207-214.

44. Toews AR, Bailey JV, Townsend HGG, Barber SM. Effect of feed rate and drill
speed on temperature in equine cortical bone. AJVR; 1999; 60(8): 942-944.

45. Vashishth D, Behiri JC, Bonfield W. Crack growth resistance in cortical bone:
concept of microcrack toughening. J of Biomechanics; 1997; 30: 763-769.

46. Venkataraman S, Hafta RT, Rapoff AJ. Structural optimization using biological
variables to understand how bones design holes. Structural and Multidisciplinary
Optimization 2002; accepted for publication.

47. Yeni YN, Norman TL. Calculation of porosity and osteonal cement line effects on
the effective fracture toughness of cortical bone in longitudinal crack growth. J
Biomed Mater Res; 2000; 51: 504-509.

48. Zioupos P. On microcracks, microcracking, in-vivo, in-vitro, in-situ and other issues.
J of Biomechanics; 1999; 32:209-211.















BIOGRAPHICAL SKETCH

Barbara Garita Canet was born in 1977. She is the second child of Donato Garita

and Noemy Canet who had six children. She grew up in San Jose, Costa Rica, and

attended Lincoln School from 1983 to 1995. After her graduation form Lincoln School,

she started her career of mechanical engineering at the Universidad de Costa Rica;

however she only studied there for one semester. She was offered a partial scholarship

for playing varsity volleyball for Florida Institute of Technology (FIT.). From FIT she

graduated with a bachelor's degree in mechanical engineering in May 2000. In August

2000 she started a master's in biomedical engineering at the University of Florida and

graduated in December 2002. She will pursue a doctoral degree in biomedical

engineering.