Mechanical Properties of cBN-Al Composite Materials Dependence on Grain Size of cBN and Binder Content Amanda Lynne McKie A dissertation submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science in Engineering. October 2009 Declaration - ii - DECLARATION I declare that this dissertation is my own unaided work. It is being submitted to the degree of Master of Science in Engineering to the University of the Witwatersrand, Johannesburg. It has not been submitted before for any degree or examination to any other University. ------------------------ Amanda Lynne McKie --------------- day of ------------------------ 2009 Abstract - iii - ABSTRACT PcBN materials have been widely used as cutting tools for ferrous materials for which PcD materials have limitations. Like diamond, cBN has excellent properties, a hardness in excess of 40 GPa, good thermal and chemical stability and good thermal conductivity. Several research studies have gone into the development of PcBN materials, but very little has been done on determining the mechanical properties affecting them. It is important to know the mechanical properties of polycrystalline cBN materials such as hardness, fracture toughness and strength (?f) in order fully to understand the behaviour of these materials in application. There has not yet been an extensive mechanical testing out on cBN-Al composites. The aim of this project is to investigate the relationship between the microstructure and mechanical properties for a wide range of composite materials based on polycrystalline cubic boron nitride and aluminium as a binder phase (PcBN-Al). The PcBN-Al composites were made using high-pressure high-temperature (HPHT) sintering methods, yielding materials with grain sizes of cBN of between 2 ? 20 ?m and an amount of Al binder of between 15 ? 25 vol.%. Mechanical properties tested were hardness, fracture toughness, KIC, R-curve behaviour and transverse rupture strength (TRS). Hardness ranged between 15 ? 40 Abstract - iv - GPa, while fracture toughness and strength were between 6.4 ? 8.0 MPa.m1/2 and 355 ? 454 MPa, respectively. Fractography was employed to rationalize the scatter in fracture strengths, to understand the nature of the flaws and correlate fracture strength with fracture toughness through the size of the fracture origins. The main findings concerning structure properties relationships of the PcBN-Al composite materials are: There is a strong dependence of hardness on grain size and binder phase, but a weaker dependence exists for fracture toughness and strength. The results confirm that the hardness increases with increasing cBN grain size and decreasing binder content. Fracture toughness generally increases with increasing cBN grain size for low binder contents and decreases with increasing binder content. R-curve behaviour was also found. It is suggested that the toughening mechanisms involved in the cBN-Al composites are due to crack and grain bridging. Strength decreases with increasing cBN grain size and there is no relationship with a change in binder content. Large strength limiting flaws were found to exist in the materials, these flaws ranged between 100 ? 500 ?m. The strength limiting flaws are caused mainly by large binder pools. Dedication - v - DEDICATION I dedicate this work to the most important people in my life; my mom, dad and brother. Thank you for all your help and support. Acknowledgements - vi - ACKNOWLEDGEMENTS I would like to acknowledge the following people for their valuable contributions which facilitated the completion of this research work. a) Prof. I. Sigalas for suggestions and supervising this work. b) Dr. M. Herrmann for his advice, useful comments, time and constructive criticism. c) Prof. J. R?del for the use of the NAW facilities and his advice. d) Element Six (Pty) Ltd ? Diamond Research Laboratory for their financial support. e) NRF Centre of Excellence in Strong Materials for financial assistance. f) The staff at Element Six (Pty) Ltd ? Diamond Research Laboratory, especially Josias, Rod, Festus, Themba, Nedret and Hester for their help and advice. g) Jami, Ludwig and the rest of the staff of NAW, Technische Universitate of Darmstadt, for their help and advice. h) K. Katuku for his time and assistance in achieving good SEM images and useful advice. i) My family, especially my mom, dad and brother, for their support and sacrifices throughout this period. Table of Contents - vii - TABLE OF CONTENTS Declaration .................................................................................................................... ii Abstract ........................................................................................................................ iii Dedication ..................................................................................................................... v Acknowledgements ...................................................................................................... vi Table of contents ......................................................................................................... vii List of figures ................................................................................................................ x List of tables ............................................................................................................... xiv List of symbols: ........................................................................................................... xv Chapter 1: Introduction and Motivation ........................................................................ 1 Chapter 2: Overview of Boron Nitride Materials ......................................................... 4 2.1. History of Boron Nitride (BN) Materials ........................................................... 4 2.2. Synthesis of Cubic Boron Nitride (cBN) ........................................................... 5 2.3. Production of Polycrystalline Cubic Boron Nitride (PcBN) Materials .............. 7 2.4. Synthesis of cBN-Al Composite Materials ...................................................... 10 Chapter 3: Strengthening & Toughening of cBN Materials ....................................... 17 3.1. Fracture Toughness .......................................................................................... 18 3.2. R-Curve Behaviour .......................................................................................... 19 3.3. Toughening Mechanisms ................................................................................. 22 3.3.1. Crack Deflection ....................................................................................... 23 3.3.2. Zone Shielding .......................................................................................... 24 3.3.3. Contact Shielding ...................................................................................... 25 3.4. Relationship of the Structure to Properties of the Polycrystalline Cubic Boron Nitride Composites ........................................................................... 27 3.4.1. Hardness .................................................................................................... 28 3.4.2. Fracture Toughness ................................................................................... 31 3.4.3. Transverse Rupture Strength ..................................................................... 34 Chapter 4: Experimental Procedure ............................................................................ 40 4.1. Starting Materials ............................................................................................. 40 4.2. Experimental Equipment for Powder Processing ............................................ 43 4.2.1. Wet Sieving of Aluminium Powder .......................................................... 43 4.2.2. Turbula Mixer ........................................................................................... 45 4.2.3. Uniaxial Pressing ...................................................................................... 46 4.2.4. Vacuum Degassing.................................................................................... 46 4.2.5. High Temperature High Pressure Sintering .............................................. 46 4.2.6. Lapping, Surface and OD Grinding Machines .......................................... 48 4.2.7. Laser Cutting ............................................................................................. 48 4.2.8. Polishing Equipment ................................................................................. 50 Table of Contents - viii - 4.3. Powder Processing Procedure .......................................................................... 51 4.3.1. Preparation of cBN Powders ..................................................................... 51 4.3.2. Preparation of the cBN-Al Powders .......................................................... 54 4.4. Characterisation and Analysis .......................................................................... 58 4.4.1. Malvern Particle Size Analysis ................................................................. 58 4.4.2. XRD Phase Analysis ................................................................................. 59 4.4.3. Microstructural SEM and EDS Analysis .................................................. 60 4.4.4. Density ...................................................................................................... 60 4.4.5. Grain Size and Phase Composition ........................................................... 63 4.5. Mechanical Properties and Testing Techniques ............................................... 64 4.5.1. Hardness .................................................................................................... 64 4.5.2. Fracture Toughness and R-curve Behaviour ............................................. 69 4.5.3. Transverse Rupture Strength (TRS) .......................................................... 77 Chapter 5: Results of High Pressure High Temperature Sintering ............................. 83 5.1. G 2 cBN ? Al Materials ................................................................................... 83 5.1.1. Microstructure ........................................................................................... 83 5.2 G 6 cBN ? Al Materials .................................................................................... 88 5.2.1. Microstructure ........................................................................................... 88 5.3. G 10 cBN ? Al Materials ................................................................................. 93 5.3.1. Microstructure ........................................................................................... 93 5.4.Grade 20 cBN ? Al Materials ........................................................................... 94 5.4.1.Microstructure ............................................................................................ 94 5.5. Grain Size and Composition ............................................................................ 99 5.6. Density Measurements of PcBN-Al Composite Materials ............................ 100 Chapter 6: Results of Mechanical Properties of cBN-Al Composites ...................... 102 6.1. Hardness ??????????????????????????.102 6.2. Fracture Toughness ........................................................................................ 103 6.3. R-Curve Behaviour (Compact Tension Measurements) ................................ 104 6.4. Transverse Rupture Strength (TRS) Measurements ...................................... 107 6.4.1. Strength Results ...................................................................................... 107 6.4.2 Weibull Analysis ...................................................................................... 108 Chapter 7: Discussion of Results .............................................................................. 110 Hardness ...... ??????????????????????????.110 Fracture Toughness ............................................................................................... 115 R-curve Behaviour ................................................................................................ 124 Transverse Rupture Strength ................................................................................. 129 Fractography 139 Chapter 8: Conclusion ............................................................................................... 149 References: ................................................................................................................ 152 Table of Contents - ix - Appendices: ??????????????????????????.161 Appendix A: Rietveld Analysis ............................................................................ 161 Appendix B: Determination of the Grain Size and Binder Phase Composition using Image Analysis Tools ............................................................ 163 Appendix C: Hardness Estimation ....................................................................... 171 Appendix D: Results of the Hardness Measurements .......................................... 172 Appendix E: Results of the SEVNB Fracture Toughness ..................................... 173 Appendix F: Results of the Compact Tension Measurements for Determining R- Curve Behaviour. ............................................................................ 176 Appendix G: Results of the Flexural Strength Tests ............................................. 181 Appendix H: Results of the Weibull Statistics .................................................. 187 Appendix I: Flaw Size ....................................................................................... 190 Appendix J: Statistical Analysis of Properties .................................................. 201 List of Figures - x - LIST OF FIGURES Figure Page Figure 1. 1: Flowchart describing the sample preparation and testing for the research project. ........................................................................................................ 3 Figure 2. 1: Crystal structures of Boron Nitride allotropes ....... ................................... 5 Figure 2. 2: SEM image of the microstructure of surface polished Amborite (courtesy of Element Six Ltd). ................................................................................. 11 Figure 2. 3: Comparison of the Knoop hardness of the cBN-TiN-Al system and the cBN-Al system. ........................................................................................ 13 Figure 2. 4: XRD graph showing the phases present in the cBN-Al materials at temperature of 1300?C, 1400?C and 1500?C . .......................................... 16 Figure 3. 1: Fracture modes of crack deformation . .................................................... 19 Figure 3. 2: A graphical representation of the ideal dependence of fracture toughness with flaw size of a ceramic material. The top curve shows the rising R- curve behaviour and the bottom curve a flat R-curve behaviour . ........... 21 Figure 3. 3: Toughening mechanisms . ....................................................................... 23 Figure 3. 4: Bridging grains: Ligament formation allowed by residual stress . .......... 26 Figure 3. 5: Toughening of material by bridging of crack . ........................................ 27 Figure 3. 6: Fracture toughness as a function of the average grain size of diamond for PCD materials with Co-binder . ............................................................... 32 Figure 3. 7: Transverse rupture strength as a function of the nominal grain size of diamond for PCD materials . .................................................................... 36 Figure 3. 8: Strength as a function of the grain size relationship for PCD materials. . 37 Figure 4. 1: SEM micrographs of the cBN and Aluminium powders used in the PcBN-Al composite; a) grade 2 cBN powder; b) grade M2-4 cBN powder; c) grade 4 cBN powder; d) grade 9 cBN powder; e) grade M20- 40 cBN powder and f) Aluminium powder. ............................................. 42 Figure 4. 2: Particle size distribution of the cBN and Aluminium original powders used in the PcBN-Al Composites. ............................................................ 43 Figure 4. 3: Wet sieving apparatus .............................................................................. 44 Figure 4. 4: Image of a rotary evaporator. .................................................................. 45 Figure 4. 5: Load and power profile used to sinter the PcBN-Al composites. ........... 47 Figure 4. 6: Laser cutting of bend bars from PcBN sintered disk. .............................. 49 Figure 4. 7: Laser cutting of compact tension (CT) samples from PcBN sintered disk. .................................................................................................................. 50 Figure 4. 8: SEM images of the mixed cBN powders; a) G6 cBN powder; b) G10 cBN powder and c) G20 cBN powder. .................................................... 53 Figure 4. 9: Particle size distribution of the mixed cBN powders. ............................. 54 List of Figures - xi - Figure 4. 10: SEM image of the grade G2 cBN-Al mixed powders; a) G2 cBN + 15 vol. % Al powder; b) G2 cBN + 20 vol. % Al powder; c) G2 cBN + 25 vol. % Al powder; d) G6 cBN + 15 vol. % Al powder; e) G6 cBN + 20 vol. % Al powder and f) G6 cBN + 25 vol. % Al powder. .................... 57 Figure 4. 11: SEM image of the G10 and G20 cBN-Al mixed powders; a) G10 + 15 vol. % Al powder; b) G20 cBN + 15 vol. % Al powder; c) G20 cBN + 20 vol. % Al powder and d) G20 cBN + 25 vol. % Al powder. ............ 58 Figure 4. 12: The configuration of the Vickers diamond indenter. ............................. 66 Figure 4. 13: The configuration of the Knoop Hardness indenter. ............................. 68 Figure 4. 14: The SEVNB arrangement in four-point bend configuration. ................ 71 Figure 4. 15 a and b: a) The razor blade cutting machine to cut V-notches into the sample and b) a V-notch cut into the surface of the PcBN-Al sample. . 72 Figure 4. 16: The compact tension (CT) sample dimensions. .................................... 74 Figure 4. 17: Cross section of the CT sample ............................................................. 75 Figure 4. 18: Schematic of the testing device used for R-curve testing . .................... 76 Figure 4. 19: The 4-point bend configuration for the Transverse Rupture Strength (TRS) tests. ............................................................................................... 79 Figure 5. 1: Microstructure of the G2 cBN-Al sintered materials. a and b) G2 cBN+15vol.% Al; c and d) G2 cBN+20vol.% Al; and e and f) G2 cBN+25vol.% Al. ..................................................................................... 84 Figure 5. 2: XRD scan of the PcBN G 2 + 15 vol. % Al sintered material. ............... 85 Figure 5. 3: XRD scan of the PcBN G 2+ 20 vol. % Al sintered material. ................ 85 Figure 5. 4: XRD scan of the PcBN G 2 + 25 vol. % Al sintered material. ............... 86 Figure 5. 5: SEM micrograph of the G 2 cBN + 25 vol. % Al sintered material turbula for a) 6 hours, b) 10 hours and c) 1 hour. ................................................. 88 Figure 5. 6: SEM images of the microstructure of the PcBN G 6 ?Al composite materials: a) and b) G6 cBN+15% Al; c ) and d) G6 cBN+20% Al; ....... 90 Figure 5. 7: SEM images of the microstructure of the PcBN G 6 +25vol.%Al composite materials: a) and b) micrographs c) EDS of a cBN grain and d) EDS image showing unreacted Aluminium. ............................................ 91 Figure 5. 8: XRD scan of the G 6 cBN + 15 vol. % Al sintered material. .................. 91 Figure 5. 9: XRD scan of the G 6 cBN + 20 vol. % Al sintered material. .................. 92 Figure 5. 10: XRD scan of the G 6 cBN + 25 vol. % Al sintered material. ................ 92 Figure 5. 11: SEM micrograph of the G 10 cBN + 15 vol. % Al sintered material. ... 93 Figure 5. 12: XRD scan of the G 10 cBN + 15 vol. % Al sintered material. .............. 94 Figure 5. 13: SEM images of the microstructure of the PcBN G 20 ?Al composite materials: a and b) G20 cBN+15vol.% Al; c and d) G20 cBN+20vol.% Al. ............................................................................................................. 96 Figure 5. 14: SEM images of the microstructure of the PcBN G 20 + 25vol.% Al composite materials: a and b) microstructure, c) EDS of unreacted Aluminium and d) EDS of AlB2. ............................................................. 97 Figure 5. 15: XRD scan of the G 20 cBN + 15 vol. % Al sintered material. .............. 97 Figure 5. 16: XRD scan of the G 20 cBN + 20 vol. % Al sintered material. .............. 98 Figure 5. 17: XRD scan of the G 20 cBN + 25 vol. % Al sintered material. .............. 98 List of Figures - xii - Figure 6. 1: R-Curve behaviour with crack length for G2 PcBN-Al composite materials. ................................................................................................ 106 Figure 6. 2: R-Curve behaviour with crack length for G6 and 10 PcBN-Al composite materials. ................................................................................................ 106 Figure 6. 3: R-Curve behaviour with crack length for G20 PcBN-Al composite materials. ................................................................................................ 107 Figure 7. 1: Hardness as a function of the grain size of the cBN particles. .............. 111 Figure 7. 2: Hardness as a function of the inverse square root of the cBN grain size. ................................................................................................................ 111 Figure 7. 3: Hardness as a function of the binder content. ........................................ 112 Figure 7. 4: Hardness as a function of the cBN grain size and the binder content. .. 114 Figure 7. 5: Main effect plot of the Hardness with cBN grain Size and Al content. 115 Figure 7. 6: Fracture toughness of the PcBN-Al composite materials as a function of grain size of cBN, using SEVNB method. ............................................. 116 Figure 7. 7: Fracture toughness of PcBN-Al composite materials as a function of the inverse square root of the cBN grain size. ............................................. 116 Figure 7. 8: Fracture toughness of the PcBN-Al composite materials as a function of binder %, using SEVNB method............................................................ 117 Figure 7. 9: Crack propagation in fracture toughness tests showing the fracture path on polished G20 cBN + 25 vol.% Al sample exhibiting slightly dominant crack propagation in the binder phase and transgranular fracture through the cBN phase......................................................................................... 121 Figure 7. 10: Crack propagation in fracture toughness tests. Fracture surface of a) G2 cBN + 15 vol.% Al and b) G20 cBN + 25 vol.% Al, exhibiting mostly transgranular crack propagation in the cBN and binder phase. .............. 121 Figure 7. 11: Fracture toughness as a function of the cBN grain size and the binder content. ................................................................................................... 123 Figure 7. 12: Main effects plot of the fracture toughness with cBN grain Size and Al content. ................................................................................................... 123 Figure 7. 13: R-curves of the G2 and G6 PcBN-Al composite materials. ................ 125 Figure 7. 14: R-curves of the G10 and G20 PcBN-Al composite materials. ............ 126 Figure 7. 15: Weibull distributions of bending strength data (a) showing more or less signs of bimodal distributions, batches G2 cBN+15vol.%Al, G6cBN+15vol.%Al (b) showing monomodal distributions of the batches G6cBN+20vol.%Al and G20cBN+15vol.%Al with Weibull moduli of 15 and 25. .................................................................................................... 131 Figure 7. 16: Weibull plots of the 15 vol.% binder PcBN-Al composite materials. . 132 Figure 7. 17: Weibull plots of the 20 vol.% binder PcBN-Al composite materials. . 133 Figure 7. 18: Weibull plots of the 25vol.% binder PcBN-Al composite materials. .. 133 Figure 7. 19: Characteristic strength plot of the PcBN-Al composite materials with cBN grain size. Error bars show the 80% confidence intervals. ............ 134 Figure 7. 20: Characteristic strength plot of the PcBN-Al composite materials with binder %. Error bars show the 80% confidence intervals. ................... 135 List of Figures - xiii - Figure 7. 21: Flexural Strength as a function of the cBN grain size and the binder content. ................................................................................................. 136 Figure 7. 22: Main effects plot of the flexural strength with cBN grain Size and Al content. ................................................................................................. 136 Figure 7. 23: Comparisons between the fracture toughness and hardness, and their dependence on grain size and binder content. ..................................... 137 Figure 7. 24: Comparisons between strength and hardness, depending on grain size and binder content. ............................................................................... 138 Figure 7. 25: Comparisons between strength and fracture toughness, depending on grain size and binder content. .............................................................. 139 Figure 7. 26: Typical flaws (a) with features of ductile fracture, sample from batch G2cBN+15vol.%Al, (b) with very large grains of binder phase, sample from batch G6cBN+20vol.%Al. .......................................................... 142 Figure 7. 27: Model crack size calculated from fracture toughness and strength limiting flaws; (a) G6cBN+15vol.%Al, and (b) G10cBN+15vol.%Al. ............................................................................................................. 143 Figure 7. 28: Flaws in G2 cBN +15vol.%Al composite: a) large surface flaw resulting in low strength and b) small binder flaw resulting in high strength. ... 144 Figure 7. 29: Flaws from materials with low strength, a) G20 cBN+25vol.%Al showing large surface defect with large cBN grain and b) G20+20vol.%Al showing binder surface defect. ................................. 144 Figure 7. 30: Flaws from materials with high strength, a) G2 cBN+25vol.%Al showing a smaller binder pool and b) G6cBN+25vol.%Al showing a small pore. ............................................................................................ 145 Figure 7. 31: SEM images of internal cracking in the PcBN-Al composite materials; a) G2 cBN + 20vol.% Al and b) G2 cBN + 15vol.% Al. .................... 145 Figure 7. 32: SEM images of various types and sizes of flaws in the PcBN-Al composite materials; a) large grain; b) elongation of binder with pore; c) larger binder pool with oxidized brittle layer and d) EDS showing O and Al. ........................................................................................................ 147 Figure 7. 33: SEM images of various types and sizes of flaws in the PcBN-Al composite materials; a and b) binder pool with pore; c) EDS showing ductile Al; d and e) larger brittle binder pool; and f) EDS showing binder is AlN. ....................................................................................... 148 List of Tables - xiv - LIST OF TABLES Table Page Table 2. 1: Properties of cBN, PCD and diamond . ...................................................... 7 Table 2. 2: Properties of PcBN and other cutting tool materials ............................... 11 Table 2. 3: Properties of the main phases found in PcBN-Al products after sintering .... .............................................................................................................. 14 Table 3. 1: Changes in Hardness with grain size of ceramics . ................................... 30 Table 4. 1: Powders used for processing. .................................................................... 40 Table 4. 2: Composition of the mixed cBN powders. ................................................. 51 Table 4. 3: Sample materials. ...................................................................................... 55 Table 4. 4: Composition of the mixed cBN-Al composite powders ........................... 56 Table 4. 5: The theoretical densities of the cBN-Al powder mixtures........................ 62 Table 4. 6: The theoretical densities of the cBN-Al sintered material. ....................... 62 Table 5. 1: Predicted volume % of the binder and cBN phases for the PcBN-Al composite materials. ................................................................................. 99 Table 5. 2: Composition and grain size of PcBN-Al composite materials. .............. 100 Table 5. 3: The density of the PCBN-Al composite materials. ................................. 101 Table 6. 1: The Vickers hardness measurements results for PcBN-Al composite materials. ................................................................................................ 103 Table 6. 2: SEVNB fracture toughness measurements results for PcBN-Al composite materials. ................................................................................................ 104 Table 6. 3: The KR results for the R-curve behaviour. .............................................. 105 Table 6. 4: Average flexural strength results for the PcBN-Al composite materials. ................................................................................................................ 108 Table 6. 5: Characteristic strength ?o, Weibull modulus m, 80% confidence intervals (C.I) and average flaw size of the PcBN-Al composite materials. ........ 109 Table 7. 1: Comparison of KIC and KIR values for the PcBN-Al composite materials. ................................................................................................................ 127 Table 7. 2: R-curve results of some common ceramics materials ............................ 128 Table 7. 3: Transverse rupture strength for commercially available cutting tool materials . ............................................................................................... 130 Table 7. 4: The estimated average flaw size, characteristic strength and SEVNB fracture toughness for the PcBN-Al composite materials. ..................... 141 List of Symbols - xv - LIST OF SYMBOLS: A ? Area of plastic contact (m2) a ? notch length (?m) a ? length of the moment arm (mm) aC ? critical flaw size (?m) aO ? length of pre-crack (?m) ? ? thermal expansion coefficient (10-6K-1) B, b ? specimen thickness (mm) c ? Crack length (?m) d ? Grain size (?m) d ? Average length of diagonals (mm) E ? Young?s modulus (GPa) f ? Volume fraction F ? Fracture load (MN) G ? Grain size (?m) H ? Hardness (GPa) K1 & 2 ? Strength coefficient KIC ? Fracture toughness (MPa.m 1/2) KIF ? Crack resistance (MPa.m 1/2) KSE ? Steady state crack resistance (MPa.m 1/2) l ? Length of crack tip (?m) m ? Weibull modulus mi ? mass % of component i n ? Notch length (?m) ? ? Density (g/cm3) P ? Applied load (N) Pf ? Probability of failure P(V) ? Probability of occurrence of a critical defect. List of Symbols - xvi - S1 & 2 ? Outer and Inner support spans ? ? Applied stress (MN) ?f ? Fracture strength (MPa) ?O ? material constant for starting strength (MPa) ?? ? Characteristic strength (MPa) T ? Temperature (K) ? ? Poisson?s ratio V ? Volume (cm3) VO ? Mean Volume (cm 3) W ? Specimen width (m) Wdry ? Dry Weight (g) Wsat ? Saturation Weight (g) WSusp ? Suspended Weight (g) Y ? Stress intensity factor Chapter 1: Introduction & Motivation - 1 - Chapter 1: Introduction and Motivation Sintered superhard materials such as diamond and cubic boron nitride are used extensively in machining due to their excellent properties such as high hardness (cBN has a hardness of 45 GPa, half that of diamond), good thermal conductivity and chemical and thermal stability. Diamond compacts are widely used as cutting tools for materials such as glass, stone, concrete, ceramics and other hard abrasive materials. Diamond has its limitations; it cannot be used for the machining of ferrous materials, as it reacts with iron at temperatures above 700?C (1). However, cBN compacts are relatively chemically inert to iron and thus thermally stable at temperatures as high as 1000?C. This makes cBN a potential material for the machining of hardened ferrous materials. Intensive research has been carried out to develop cBN materials, but there are only a few papers (2,3,4,5,6,7,8) dealing with the mechanical properties of cBN material such as hardness, fracture toughness and strength and most of them focused on hardness. These properties are especially necessary for materials that are used in applications where these properties are crucial such as in cutting tools. In cutting applications, a material is needed which is very hard, with excellent strength and which has a high wear resistance and good thermal stability to withstand high operating temperatures without decomposing or significantly weakening. Good Chapter 1: Introduction & Motivation - 2 - fracture toughness and strength are also required, as they will determine the extent to which a material will fail through flaws and crack propagation. The aim of this project is to determine the mechanical properties of PcBN-Al composite materials based on variations in the microstructure. The systematic change of microstructure will be achieved by varying the cBN grain size and the volume content of the binding aluminium metal. The effect of these changes to the mechanical properties will be analysed. A detailed flowchart describing the project process is shown in Figure 1.1 A detailed overview of the cBN-Al materials is presented in Chapter 2. Chapter 3 contains a detailed description of the mechanical properties of cBN materials, such as the toughening mechanisms and effects of grain size and binder content. Chapter 4 contains a detailed description of all the powder processing and experimental procedures, equipment and testing techniques used in the project. Chapter 5 summarizes the results of the sintered composites analysis. Chapter 6 summarizes the results of the mechanical properties. The effect of microstructure on the mechanical properties is discussed at the end of the chapter. The conclusions and recommendations for future work are discussed in Chapter 7. Chapter 1: Introduction & Motivation - 3 - Figure 1. 1: Flowchart describing the sample preparation and testing for the research project. Powder Characterisation Mixing Sintering Preparation High Pressure High Temperature Sintering Material Processing Chain & Process Optimisation Surface Preparation: Laser cutting, grinding lapping & polishing Material Characterisation: XRD, SEM, Phase composition & Grain size analysis Mechanical Properties Measurements [HV, KIc (SEVNB), KR & TRS] Integration of properties with Microstructure Material Preparation & Characterisation Properties Testing & Analysis Chapter 2: Overview of Boron Nitride Materials - 4 - Chapter 2: Overview of Boron Nitride Materials 2.1. History of Boron Nitride (BN) Materials Boron nitride (BN) is a commercially synthesized material discovered in the early 19th century, but not used commercially until the 20th century. It consists of equal amounts of boron and nitrogen atoms and has a structure similar to carbon. Thus like carbon, BN has different allotropes; hexagonal BN (hBN), cubic BN (cBN) and wurzite BN (wBN), although hBN and cBN are the most common forms (9). Hexagonal boron nitride (hBN) is similar to graphite, consisting of layers of nitrogen and boron atoms combined in a hexagonal network. Between the layers only a weak bonding exists, therefore it has a soft material structure (10). hBN is produced by nitridation or the ammonolysis of boron trioxide (B2O3) at elevated temperatures. Due to its graphite-like structure hBN is used as a lubricant in various applications where the chemical reactivity and the electrical conductivity of graphite would cause problems if used (10,11). The second most common BN form is cubic boron nitride (cBN); cBN has a zinc blende structure which resembles that of diamond. It consists of boron and nitrogen atoms in a three-dimensional network where B and C have 4 equivalent strong covalent bonds. It is extremely hard, although less than diamond. cBN is used as an Chapter 2: Overview of Boron Nitride Materials - 5 - abrasive and as a cutting tool. cBN is formed by the conversion of hBN under high pressures (4 ? 6 GPa) and temperatures (1200 - 1700?C) similar to the synthesis of diamond from graphite (11,12). Figure 2.1 shows the crystal structure of BN allotropes. Figure 2. 1: Crystal structures of Boron Nitride allotropes (13). 2.2. Synthesis of Cubic Boron Nitride (cBN) Cubic boron nitride (cBN) was first synthesized in 1957 by Wentorf (14).It was found that high temperature and high pressure conditions similar to those needed for the synthesis of diamond could be used to convert hexagonal boron nitride to the cubic form. Wentorf (14) showed that cBN can be produced by the direct conversion of hBN to cBN by high-pressure high temperature treatment. The direct conversion of hBN to cBN can be done with pressures of up to 18 GPa and temperatures between 1750 and 3023?C. Additions of small amounts of boron oxide can lower the pressure to 4 ? 7 Chapter 2: Overview of Boron Nitride Materials - 6 - GPa and temperature to 1500?C. Commercially cBN is synthesized by a mixture of hBN and various catalysts/solvents, which are subjected to pressures in the range of 4.5 ? 5.5 GPa and temperatures between about 1200 - 1700?C (11). The various catalysts/solvents that can be used are lithium nitride, calcium nitride and magnesium nitride. Diamond and cBN have excellent properties such as high hardness, thermal conductivity and chemical stability. They are therefore used widely in industry applications, such as in cutting tools. Diamond is widely used as a cutting tool for hard materials such as concrete, stones, ceramics, non ferrous metals and other hard materials that have a low chemical reactivity with carbon. Due to the tendency of iron and its alloying elements such as nickel to react with diamond at high temperatures when machined and cause deterioration of the cutting tool (15), cBN was introduced as the alternative. cBN is an extremely hard material, second only to diamond and it is chemically inert to iron and is thermally stable to temperatures as high as 1000?C, better than that of diamond. It can also form passive oxide layers at high temperatures when in contact with oxygen. It is thus suitable for the machining of hard ferrous alloys. cBN particles also bind well to metals, forming interlayer?s of metal borides or nitrides. The mechanical properties of single crystal cBN and natural diamond are given in Table 2.1. Chapter 2: Overview of Boron Nitride Materials - 7 - Low-pressure methods can also be used to manufacture thin films of cBN on a substrate; such as CVD, PVD and ion beam deposition (16). These methods result in materials with good properties and are also more cost effective when used on coatings. Table 2. 1: Properties of cBN, PCD and diamond (11). Property Diamond (natural) cBN Density (g/cm3) 3.5 3.48 Melting point (?C) - 2700 Fracture Toughness (MPam0.5) 3.4 5 Knoop Hardness (GPa) 57-104 43-47 Young?s Modulus (GPa) 1141 600-800 Thermal expansion (10-6K-1) 1.5-4.8 4.9 Thermal conductivity 20oC(Wm-1K-1) 500-2000 150-700 2.3. Production of Polycrystalline Cubic Boron Nitride (PcBN) Materials Although single crystal cBN is useful, it is also limited in its use as only small particles of single crystal cBN can be produced. It is necessary for application purposes to produce materials containing polycrystal crystals of cBN joined together with a binder. Higher fracture toughness is introduced into polycrystalline cubic boron nitride (PcBN) materials with the addition of different binder phases. However, fully densified PcBN can only be manufactured through high pressure (~ 7 GPa) and high temperature conditions with the aid of binders. This is due to cBN?s strong Chapter 2: Overview of Boron Nitride Materials - 8 - covalently bonded structure, which is very difficult to compress and sinter at low pressures (11). Still, research has been done on the sintering of PcBN materials at low pressures (17,18,19). Wentorf and Rocco (20,21) sintered the first PcBN materials. The sintered PcBN material contained alloys of nickel, iron, cobalt and aluminium as binders. In the study of cBN to cBN bonding, it was believed to have been achieved by a liquid phase sintering process, which involved the binder alloys. Hibbs and Wentorf (22) produced high cBN-content cutting tools by infiltrating molten Al-Co from WC-Co substrate into cBN layers during high-pressure and high temperature sintering. In 1980 a patent was filed by Sumitomo Electric Industries Ltd (23) that covered the manufacturing of any sintered compact for a machining tool with 10-80% cBN and various binders comprising carbides, nitrides, borides or silicates of many of the metals. Other composites that have been formed recently include cBN-TiN (24,25), cBN-Al (1,3,8,26,27), cBN-WC-Co (28), cBN-Ti (29) and cBN-TiN-Al (6). The composites of the cBN-Al system will be described in more detail in section 2.4. High purity polycrystalline cBN single phase (binderless) materials have also been produced; these materials have the advantage of excellent mechanical and thermal properties free from the effects of the binder phase. Wakatsuki et al,(30) reported on the direct conversion of hBN to PCBN at low pressures of 55 kbar and temperatures of 1100-1400?C with no catalyst. Chapter 2: Overview of Boron Nitride Materials - 9 - The commercial method of PcBN production is done by sintering together cBN with sintering additives or binder materials at pressures of 4-6 GPa and temperatures of between 1200 and 1500?C (11). PcBN is sintered by cold and hot compaction stages and liquid phase sintering mechanisms. Reaction sintering plays a major role in the driving force of the densification process. A reaction between the BN and the metal binder occurs to form borides and nitrides. These act as the binder phase. The metal binders added influence the final properties of the material. Composite metals of the elements in groups IV-VI of the Periodic table are often used (31). Other metals can also be added as a binder such as aluminium (group III), cobalt and nickel (3,17) as well as binders that can increase the thermal stability or wear resistance of the material such as titanium nitride (TiN) and titanium carbide (TiC). Aluminium has the lowest melting point at ambient pressures and can react with cBN under a mild temperature (8). Studies by Hara and Yazu (25), Morgiel et al (29) and Rong et al (6) on the sintering of cBN-Ti based composites have shown that Ti can be used as an important binder to produce a material with thermal stability and wear resistance by forming TiN or TiC. The properties of the PcBN compact are determined by the final microstructure and the phases formed during sintering. From chemical equilibrium calculations in the BN-Al system it is known that Al reacts with BN in a wide temperature and pressure range. The type and number of new phases formed through sintering depends on the temperature and pressure. The reaction mechanism is still unclear. Chapter 2: Overview of Boron Nitride Materials - 10 - 2.4. Synthesis of cBN-Al Composite Materials The synthesis of cBN with Aluminium was first reported in 1987 (1) using high temperatures and high pressures. In this study, solid cBN reacted with molten Al to form residual binder phases of Aluminium nitride (AlN) and Aluminium diboride (AlB2). The AlN and AlB2 phases protected the internal surfaces from chemical attack and gave the material good thermal stability. cBN-Al type sintered materials have been commercially marketed and are known as Amborite. Figure 2.2 shows the microstructure of Amborite (De Beers/ Element Six Ltd). Some property comparisons of Amborite and other cutting tool materials are listed in Table 2.2 below. When compared to the PCD material Syndite 010 or Tungsten Carbide (WC), as can be seen in the table above, Amborite has good properties. The strong covalent bonds in Amborite give it its good properties such as high hardness 31.5 GPa and good fracture toughness 6.36 MPa.m1/2. The original cBN particles provide the hardness and wear resistance of the material, thus giving it use in cutting tool applications. Chapter 2: Overview of Boron Nitride Materials - 11 - Figure 2. 2: SEM image of the microstructure of surface polished Amborite (courtesy of Element Six Ltd). Table 2. 2: Properties of PcBN and other cutting tool materials (11). Amborite SYNDITE 010 WC/Co Density [g cm-3] 3.41 4.127 14.7 Compressive strength [GPa] 2.73 4.74 4.5 Fracture toughness [MPa.m0.5] 6.36 8.39 10.8 Knoop Hardness [GPa] 31.5 50 13.0 Young modulus [GPa] 653 993 620 Modulus of rigidity [GPa] 228 453 258 Bulk modulus [GPa] 297 412 375 Poisson?s ratio 0.13 0.102 0.22 Thermal expansion coefficient [10-6K-1] 4.9 3.8 5 Thermal conductivity [W.m-1K-1] 100 120 100 In a study conducted by Rong and Fukunaga (24) on the sintering behaviour of the cBN-Al system they described the reactions occurring during high-temperature and high-pressure conditions in which cBN powder (1-3 ?m) was mixed with Aluminium Chapter 2: Overview of Boron Nitride Materials - 12 - powder (1-4 ?m) in various compositions ranging between 50 and 90 mol% cBN. The powders were sintered at temperatures of between 1200 and 1400?C and a pressure of 5.75 GPa. The residual binder phases formed were Aluminium nitride (AlN), Aluminium diboride (AlB2) and ?-AlB12. The reaction of the cBN-Al system can be observed in the following reactions: )2(121312 )1(232 12 2 AlNAlBAlBN AlNAlBAlBN ???? ??? ? Rapid grain growth of AlN and AlB2 was observed during sintering, consequently lowering the mechanical strength. TEM analysis of the cBN-Al composites showed that the AlN phase existed around the cBN grains, while the AlB2 and AlB12 phases existed between the AlN-AlN phase. The diffusion of B ion was faster than the N ion within the aluminium liquid. Knoop Hardness measurements were done on the sintered cBN-Al compacts (26) sintered at 1200?C and 1400?C. Later Rong and Fukunaga (6) added TiN as a secondary phase to the cBN-Al system to prevent the rapid growth of the AlN and AlB2 grains and to improve its properties. TiN has an intermediate thermal expansion coefficient between that of cBN and Al; this prevents cracking in the microstructure. The sintering of cBN-TiN-Al resulted in the cBN grains reacting with the TiN to Chapter 2: Overview of Boron Nitride Materials - 13 - form TiB2 and AlN. The TiB2 and AlN formed around the cBN and TiN grains and reinforced the strength of the matrix. Figure 2. 3: Comparison of the Knoop hardness of the cBN-TiN-Al system and the cBN-Al system (6). Figure 2.3 also shows the Knoop hardness values for the cBN-TiN-Al samples sintered at 1200 - 1400?C. The samples sintered at 1200?C have a higher hardness than that of the 1400?C samples across the composition range. This resulted in the TiN reducing the sintering temperature of the system. Compared with the hardness values of the cBN-Al system (Figure 2.3), the hardness of the cBN-TiN-Al system was always higher than the cBN-Al system at 1200?C and 1400?C. The addition of the relatively hard TiN enhanced the hardness of the binder phase especially in the region of the lower cBN compositions. While in the cBN-Al Chapter 2: Overview of Boron Nitride Materials - 14 - system at the lower cBN compositions significant amounts of AlB2 were formed which was softer and thus lowered the hardness of the system. Thus by adding TiN to the composite the hardness was increased at a lower sintering temperature. Table 2.3 shows the main phases found in PcBN-Al composites after sintering and the properties associated with each phase. Table 2. 3: Properties of the main phases found in PcBN-Al products after sintering (11). Properties cBN Diamond AlN AlB2 AlB12 Crystal structure Cubic Cubic Hexagonal Hexagonal Tetragonal Density [g/cm3] 3.48 3.51 3.26 3.17 2.57 Melting Point [?C] 2700 - 2300 975 2070 Knoop Hardness [GPa] 47 57-104 12 24 Young Modulus [GPa] 700-800 1141 318 4.3 Thermal expansion coefficient [10-6K-1] 3.2 1.5-4.8 3.9 Thermal Conductivity [W.m-1K-1] 150-700 500-2000 200 Research on the phases present in cBN-Al materials sintered under HP-HT conditions was looked at by Zhao and Wang (27) in which cBN powder (3-5 ?m) was mixed with Al powder (38 ?m) in a volume ratio of cBN:Al = 7:3. Sintering was done at about 5.5 GPa at temperatures ranging from 1300?C - 1500?C for 10 min. Figure 2.4 shows the XRD results which show that at 1300?C no AlN or AlB2 was observed therefore no reaction had taken place between the cBN and Al. At 1400?C, Chapter 2: Overview of Boron Nitride Materials - 15 - new phases (AlN and AlB2) were formed while retaining some Al. At 1500?C all Al disappeared and only AlN and AlB2 remained with the cBN. The XRD and TEM results showed that the Al reacted with cBN at 1400?C to produce AlN and AlB2. Al atoms diffused into the hBN layer of cBN particles and reacted with hBN to form AlN, while B atoms diffused into the Al zone and reacted with Al to form AlB2. Recently Li et al (8) investigated the high pressure sintering behaviour of cBN-Al composites on WC/Co substrates at various temperatures and cutting performance. Hardness results showed a decrease with increasing binder content. Benko et al (3) also did research on cBN-Al composite materials under hot pressing conditions. In this research composites with a BN:Al ratio of 9:1 was sintered under a pressure of 10 MPa and at a temperature of 1750?C. The samples were then thermally treated in a vacuum of 3 x 10-3 Pa for 1 hour. Knoop hardness of the sintered compacts was measured before and after annealing. The study showed that the hardness increased from 10 to 20 GPa after annealing, which showed that in thermal annealing the compacts could increase the hardness of the material. Sithebe (18) investigated the preparation of cBN-Al composites at low pressures by hot pressing 50vol.%Al with 50 vol.% cBN at different temperatures. Reaction between the cBN and Al was observed at temperatures between 1000?C and 1400?C under N2 and Ar. AlB2 started forming at 1100?C, but disappeared above 1300?C where AlB12 started to form. In this study Sithebe et al (18,19) also looked at pressure infiltration of Chapter 2: Overview of Boron Nitride Materials - 16 - molten Al in a porous hBN or cBN matrix at 800?C and 12 MPa. Fully infiltrated cBN performs were made with 12 ?m cBN, while a limited extent of infiltration was achieved for 3 ?m cBN grains. Results showed that the hardness was higher for the infiltrated cBN than the hBN. Figure 2. 4: XRD graph showing the phases present in the cBN-Al materials at temperature of 1300?C, 1400?C and 1500?C (27). Chapter 3: Strengthening & Toughening of cBN Materials - 17 - Chapter 3: Strengthening & Toughening of cBN Materials Ceramic materials and their composites are mostly brittle materials. Ceramics and ceramic composites are used in a wide variety of advanced engineering applications. Knowledge of how these materials behave is important in understanding their behaviour during application. It is essential to understand the structures to properties relationships for these materials. Hardness, strength and fracture toughness are crucial in understanding when a material will fail in application. Failure of these materials during application can have consequences on productivity and operations. Failure of ceramic materials is caused by the extension of flaws, which are usually introduced into the material during processing or surface treatment such as grinding and polishing. Flaws in the material can be inclusions, pores and cracks. The strength of a material is related to the maximum flaw size ac, and the critical fracture toughness K1C, of a material. So in order to improve the strength (?f,) of a material a reduction in the maximum flaw size ac, and/or an increase in the critical fracture toughness K1C, are required. Chapter 3: Strengthening & Toughening of cBN Materials - 18 - The Griffith equation shows the dependence of the strength (?f) of a material on the critical flaw size ac and the fracture toughness K1C (32). C C f aY K1?? (3.1) Where Y depends on the position and shape of the crack, and can usually be taken to be Y ? 1 for most ceramics, but can also be calculated using Y = ??/2 for penny- shaped cracks (32). Since the strength of a material is both related to the flaw size and its fracture toughness, it is important to have a material with high fracture toughness. Fracture toughness is mostly influenced by microstructure. Improving fracture toughness therefore needs a tailoring of the microstructure. 3.1. Fracture Toughness The fracture toughness of a material is a very important parameter, it is the material?s resistance to crack propagation, but ultimately it is characterized by the critical stress intensity factor K1C as the limiting value of stress intensity which can be applied to a material above which crack propagation will occur. K1C corresponds to the critical stress intensity factor of the mode I fracture mode. There are 3 basic modes of crack tip deformation: the opening (mode I), the in-plane shear (mode II) and the out-of- Chapter 3: Strengthening & Toughening of cBN Materials - 19 - plane shear (mode III). Figure 3.1 shows the various fracture modes. Mode I is used for crack opening and tension applications. Mode I: Tension, Opening; Mode II: In-plane shear, sliding and Mode III: Out-of-plane shear, tearing. Figure 3. 1: Fracture modes of crack deformation (33). 3.2. R-Curve Behaviour Since ceramics such as cBN-based materials are brittle they generally fail by brittle fracture. This puts some major limitations on applications. This has motivated researchers to develop ways of increasing the toughness of ceramics. The R-curve behaviour has thus been a characteristic feature of toughened ceramics. The R-curve or Resistance curve behaviour concept of crack propagation arose from the studies of fracture in metals and alloys described by Broek (1986) (34), but was Chapter 3: Strengthening & Toughening of cBN Materials - 20 - later applied to ceramics (34,35,36,37). The R-curve behaviour of a material is the enhanced resistance to crack growth during crack extension. R-curve behaviour is typically indicated by a rising fracture resistance curve. The crack tip is screened or prevented from propagating, causing an increase in the toughness of the material. The mechanisms involved in toughening materials will be described in Section 3.3. The R-curve behaviour of a material can be seen in Figure 3.2, which shows the resistance of a crack to propagation with increasing crack length of a V-notched sample under 4-point bend loading. n symbolizes the length of the V-notch, which is constant for all specimens, and l and a0 denote the length of the crack tip and length of the pre-crack respectively. With increasing crack length the crack resistance increases due to shielding mechanisms in the precracked zone. After a certain distance the shielding and the distortion due to large crack openings are equal and a steady state of resistance is reached (KSE). Crack propagation occurs from the crack tip until steady-state toughness (KSS) is reached. This remains so until fracture occurs. Ceramic materials exhibiting R-curve behaviour are more flaw-tolerant than ceramics that do not (38). In Figure 3.2, the dotted line shows a flat curve which means that no R-curve behaviour occurs. Chapter 3: Strengthening & Toughening of cBN Materials - 21 - Figure 3. 2: A graphical representation of the ideal dependence of fracture toughness with flaw size of a ceramic material. The top curve shows the rising R-curve behaviour and the bottom curve a flat R-curve behaviour (38). R-curve behaviour results in (32): 1) Materials with a rising R-curve show an increase in their strength compared with a material with the same crack tip toughness and a flat R-curve. Materials with coarse grains and two phases with an elongated or needle-like phase lead to a rising R-curve and thus higher strength values. 2) A rising R-curve decreases the scatter in the measured strength from sample to sample compared to a material with the same crack tip toughness. 3) R-curve behaviour can be determined by macrocracks. The various types of toughening mechanisms involved in the toughening of materials are given in Section 3.3. In a study on R-curve behaviour by various researchers and summarised by Munz (32), non-phase transforming ceramics with rising R-curves have Chapter 3: Strengthening & Toughening of cBN Materials - 22 - crack face interactions which are responsible for the rising crack growth resistance observed. These crack face interactions are usually crack deflecting or bridging in nature. 3.3. Toughening Mechanisms One of the objectives of developing advanced ceramic composite materials is a high reliability. To achieve this, the flaw size distribution and toughening mechanisms acting in the material need to be controlled. The flaw size control approach takes into account the brittleness of the material and attempts to control the processing flaw, while the toughening control approach attempts to create microstructures that impart sufficient fracture resistance so that the strength becomes insensitive to the flaw size. The resistance of brittle materials to crack propagation can be strongly influenced by the microstructure and by adding various reinforcements such as secondary phase particles, whiskers and fibres. It is possible that a material can have more than one toughening mechanism; this thus generates a net toughness larger than that resulting from the separate mechanisms. In multiple mechanism systems the dominant mechanism usually changes as the microstructure varies. Toughening mechanisms that result in the enhancement of the fracture toughness of ceramic materials can be divided into three basic types shown in Figure 3.3. Each toughening mechanism is described below (39,40,41). Chapter 3: Strengthening & Toughening of cBN Materials - 23 - Figure 3. 3: Toughening mechanisms (41). 3.3.1. Crack Deflection This mechanism of toughening occurs by impeding the growth of the crack by placing obstacles in the crack path thus increasing the toughness of the material. The necessary conditions of crack deflection are that the particles have weak interfaces and high residual stresses between the different phases, so that crack growth occurs along the grain boundaries. These obstacles can include secondary phase particles, whiskers, fibres or platelets. The crack takes the path of least resistance, the obstacles pin the crack thus causing the crack to either bow around the obstacle or deflect out Chapter 3: Strengthening & Toughening of cBN Materials - 24 - of the crack plane. The extent of crack deflection does not in general increase with crack length so that long R-curve behaviour and crack tolerance cannot be expected. PcBN materials are brittle materials and thus fail in a brittle manner with rapid crack propagation throughout a stressed material. The material contains very little plastic deformation before failure and the cracks run perpendicularly to the applied stress. Brittle fracture can be classified into two main groups: transgranular and intergranular fracture. Transgranular fracture is the fracture through the grain of a material. The fracture path changes direction from grain to grain and is sometimes dependent on the orientation of the grain. The crack chooses the path of least resistance. Intergranular fracture occurs when a crack travels along the grain boundaries of the material and usually occurs when the phase between the grains is weak. 3.3.2. Zone Shielding This toughening mechanism results from shielding the crack tip from applied stress as the crack grows. There are two types of zone shielding toughening effects: transformation toughening and microcrack toughening. 3.3.2.1. Transformation Toughening Transformation toughening is seen mostly in Zirconia-containing ceramic materials, resulting from a volume increase caused by the phase transformation of zirconia from tetragonal into its monoclinic allotrope. This transformation induces high Chapter 3: Strengthening & Toughening of cBN Materials - 25 - compressive stresses in the vicinity of the crack, which prevent the cracks from propagating. PcBN materials do not go through any phase change during cracking, thus they do not experience any transformational toughening. 3.3.2.2. Microcack Toughening Microcrack formation occurs at the grain boundaries of a material as a result of high residual stresses generated by thermal expansion anisotropy in single phase ceramics or thermal expansion mismatch in multiphase ceramics. Microcracking as a toughening mechanism has not been seen in PcBN materials but microcrack development has been linked to one of the major causes of replacements of cutting tools (42). 3.3.3. Contact Shielding Contact shielding is the shielding of the crack tip by forming a crack closure behind the crack, either by the friction of serrated grains or by the bridging of the crack by ligaments. 3.3.3.1. Crack Bridging Toughness can be increased by the presence of reinforcing elements that bridge the crack surfaces. Bridging can be both ductile and brittle (39). In ductile bridging as it pertains to metal-toughened ceramics, the ductility and high toughness of the metal ligaments provide the toughening through plastic dissipation. Large local residual Chapter 3: Strengthening & Toughening of cBN Materials - 26 - stresses caused by thermal expansion mismatch can suppress local crack propagation, which allows the intact ligament to exist behind the crack front as can be seen in Figure 3.4. Once the ligaments have failed the energy is dissipated as acoustic waves. Figure 3.5 shows a schematic of effects the bridging has on the crack. Figure 3. 4: Bridging grains: Ligament formation allowed by residual stress (39). For brittle materials the toughening is more subtle and is similar to non-bridging materials. The bridging requires microstructural residual stress or weak interfaces. The low energy interfaces and/or grain boundaries can cause the crack to deflect along those interfaces, also permitting intact ligaments. As the crack extends, further debonding can occur and eventually the bridges fail either by debonding around the end or fracture. After failure, frictional sliding may occur along the debonded surface. Chapter 3: Strengthening & Toughening of cBN Materials - 27 - Figure 3. 5: Toughening of material by bridging of crack (39). 3.4. Relationship of the Structure to Properties of the Polycrystalline Cubic Boron Nitride Composites The relationship of the microstructure of the PCBN composites on mechanical properties is an important aspect in understanding the materials. The effect of the grain size of the cBN particles and the amount of binder phase have on the hardness, fracture toughness and strength of the material will be discussed in this chapter. The binder chemically reacts with the cBN grains under sintering conditions to form bonds between the cBN grains. The residual binder phase has properties that will affect the properties of the whole composite material. In the PcBN-Al composite system the Aluminium added reacts with the cBN to produce a residual binder phase made up of AlN and AlB2 as shown in Chapter 2. These residual binder phases are ceramic and thus quite brittle. Chapter 3: Strengthening & Toughening of cBN Materials - 28 - In explaining the effects of the microstructure on the mechanical properties of the PcBN materials, there is insufficient information on fracture toughness and strength (2,11). There is a close relationship between the properties of PcBN and polycrystalline diamond (PCD). The most significant difference between PcBN and PCD materials is the binder phase; in PcBN materials the binder phase is normally ceramic and brittle as in the case of PcBN-Al composites, while the PCD materials contain Co, which is more ductile. As there are much more data on the mechanical properties of PCD materials (43,44,45), the effect of the grain size and binder content on the properties of PCBN will be explained using the corresponding information available for PCD materials. Extrapolation from this body of information into PcBN will be made, as part of an effort to understand the mechanical properties of PcBN materials and to interpret the results obtained in this work. 3.4.1. Hardness In a study on the hardness of PcBN materials by Hooper and Brookes (1986) (46) they suggested that the highly deformed structure formed through the synthesis of high content cBN materials results in a high level of hardness. Hooper (47) indicated that the hardness of the PcBN materials could be accurately modelled using the rule of mixtures relationship between hardness and cBN content as shown in equation 3.2. bindercBN fHHfH ??? )1( (3.2) Where f is the volume fraction of the binder phase and H is the hardness. Chapter 3: Strengthening & Toughening of cBN Materials - 29 - Since the ultra hard cBN particles are the main contributors to the hardness of the PCBN composite, the size of these particles will affect the hardness of the material. Similarly to other brittle materials, the relationship between the hardness and the grain size of the cBN grains can be approximated by the Hall-Petch equation shown below. This relationship shows that the hardness is inversely proportional to the square root of the grain size (d) (48). d H 1 ? (3.3) Thus the hardness increases with the decreasing particle size of the cBN grains. This occurs because the smaller the grain size, the higher the applied stress needed to cause dislocation movement through the material. With a coarse grained material a lower stress is needed to move dislocations, therefore reducing the hardness of the material as more plastic flow can occur. Rice (1994) (48) reported conflicting trends in literature where for some materials of large grain sizes the hardness decreases with decreasing grain size. In a study on the hardness-grain size relationships in ceramics Rice (48) tested the two trends, concluding that the confusion was due to lack of sufficient data to properly cover the two trends. Of the two conflicting trends, the first was the normal inverse trend of hardness with grain size (H?G-1/2) applied to finer grain sizes (1 ? 50 ?m). The second trend was for a minimum hardness value occurring at intermediate grain sizes (20 ? 50 ?m) (where cracking or spalling occur along grain boundaries) and also for Chapter 3: Strengthening & Toughening of cBN Materials - 30 - hardness increases towards single crystal values for large grain sizes (>50 ?m). Table 3.1 shows variations in hardness with grain size for some ceramic materials. The Hall-Petch relationship can only be used as a guideline in determining the grain size effect on the hardness of cBN materials as it was developed for single phase materials. Since PcBN materials consist of two phases the effect of the binder phase can change the relationship between the hardness and the grain size. Thus the effect of the binder phase needs to be looked at. Table 3. 1: Changes in Hardness with grain size of ceramics (48). Hardness (GPa) (100g load) % Change 2 ?m 50 ?m 100 ?m (2 -50 ?m) (50-100 ?m) Al2O3* 24.7 23.5 23 4.8% 2.1% B4C** - 37 36.5 - 1.4% SiC* - 34.5 34 - 1.5% Si3N4* 17.5 21 - - 20% - *HV (100g load) **HK (100g load) Both the composition of binder material and the amount of binder will affect the hardness of the composite material. The binder phase is usually a softer material than the ultra-hard cBN phase and thus will decrease the hardness of the composite material. Low binder content leads to more contact between the cBN grains, generating high stresses in the material structure due to the ultra high pressures used for sintering, thus increasing the hardness of the material. Increasing the binder content causes a decrease in the hardness due to the decrease in the harder grain Chapter 3: Strengthening & Toughening of cBN Materials - 31 - abating against each other and a decrease in the flow stress (see Figure 2.3, Chapter 2). 3.4.2. Fracture Toughness The resistance to crack propagation plays an important role in understanding a material?s behaviour and enhancing application limits. The fracture toughness and toughening mechanisms are important in understanding the material. As stated previously there is very little data on the mechanical properties of PcBN materials (2,11). Thus some of the trends for fracture toughness will be discussed using PCD adapted to PcBN materials. Studies by Lammer (1988) (43) and later Meiss (1996) (44) investigated the changes of fracture toughness of PCD materials with diamond grain size. The technique used to determine the fracture toughness was the diametral compression test or Brazilian disk test. In this study, the fracture toughness increased with increasing grain size to a maximum value and then remained constant (2 ? 6 MPa.m0.5). Figure 3.6 shows the fracture toughness as a function of the average grain size of the diamond grains for Miess? (44) work. He explained this trend on the basis of the work of Rice (48,49) on the changes of fracture energy with differences in thermal expansion coefficients between the primary and secondary phases, which cause microcracking to occur and propagate at the interface. Microcracking has not however been seen in PcBN materials and thus is unlikely to cause toughening. Chapter 3: Strengthening & Toughening of cBN Materials - 32 - Huang (1997) (45) also did a study on the fracture toughness of PCD materials; in this study the fracture toughness was measured using a modified single-edge-precracked beam method test. The investigation also showed that the fracture toughness increased with the increasing grain size of the diamond particles. Figure 3. 6: Fracture toughness as a function of the average grain size of diamond for PCD materials with Co-binder (44). Rice et al (1979) (49) reported that the fracture toughness of polycrystalline ceramic depended on the grain size, i.e. fracture toughness increasing with the particle size of the hard phase. It was suggested that the internal stresses that existed between the components due to different thermal expansions resulted in local tensile and compressive stresses. The stresses near flaws or along the applied stress resulted in failure by adding to the applied stress. Rice et al (49) suggested that the contribution of Chapter 3: Strengthening & Toughening of cBN Materials - 33 - the internal stresses to failure increased with decreasing grain size. Other factors such as impurities in the matrix material, the size and amount of the secondary phase and deformation of the grains during synthesis could also have affected the fracture toughness of the material. As described in section 3.3 the most likely toughening mechanisms to occur in PcBN materials are crack deflection, grain bridging and to lesser extent crack bridging due to residual Al metal left after sintering. As PcBN materials are phases made up from man with hard cBN grains surrounded by one or more ceramic binder phases, these cBN grains will act as obstacles to crack propagation and thus cause a certain amount of toughening in the material. Crack propagation is generally affected by the residual stresses in the material. In a multi-phase material such as PcBN the differences in the thermal expansion cause residual stresses between the different phases. This will likely affect crack propagation. The magnitude of residual stresses and their local distribution is determined by the volume fraction of the ceramic phases. Estimation of the internal stresses formed in the material by thermal expansion mismatch of the various phases is given in equation 3.4 (50). The internal stresses are modelled for the ceramic binder phase. ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ? ?? ?? ? ? ? ? ? ?? ? ? ? ? ? ?? p p m m pm tr EE T where r R r R ?? ??????? 21 2 12 ; 33 (3.4) Chapter 3: Strengthening & Toughening of cBN Materials - 34 - Where ?r and ?t are the radial and tangential stresses, the tangential stress is half the radial stress. Em, ?m and Ep, ?p are the Young?s modulus and Poisson?s ratio for the matrix (cBN) and secondary phase (AlN). R and r are the secondary phase particle size and the distance from the centre of the particle, at the interface of the phases R = r. Poisson?s ratio for cBN and AlN is 0.2 (51) and 0.22 (52) respectively. The result from equation 3.4 gives the radial stress as 442 MPa and the tangential stress as ? 221 MPa. A residual radial matrix compressive stress is generated around the AlN phase, as the thermal expansion of the AlN is higher than the matrix phase. The compressive stresses can prevent or reduce intergranular cracking and crack propagation. The binder phase also affects the fracture toughness and acts typically as a crack blunting. A more ductile binder phase can increase the blunting effect on the crack. The higher the content of the binder phases the greater the effect on fracture toughness, which should increase with increasing binder content. 3.4.3. Transverse Rupture Strength The strength of a material is important as it indicates the material?s ability to resist a force and thus the strength at which a material will fail. As stated previously there is limited data on the mechanical properties of PcBN materials (2,11). Thus some of the trends for the strength will be discussed using PCD and adapted to PcBN materials. Chapter 3: Strengthening & Toughening of cBN Materials - 35 - A study by Lammer (43) on the mechanical properties of PCD materials showed that strength decreased with increasing grain size for PCD materials as can be seen in Figure 3.7. He explained this trend according to two models, the first using the classical Griffith-Orowan strength relationship shown in equation 3.5, the second using the Hall-Petch strength relationship shown in equation 3.6. The Griffith- Orowan relationship shows that the largest grain size had the most severe flaw, which leads to failure, with a failure mechanism caused by the propagation of cracks originating in these existing flaws. Although this was applied to an ideal material, in reality there were pores in the material, which were larger than the grain size and could have resulted in the most severe flaws. In the second model the Hall-Petch strength relationship was used for smaller grain size materials and the mechanism was based on the microplastic flow or dislocation movement which leads to crack initiation, propagation and then failure. Microplastic flow did not occur in the cBN material since the binder was brittle, so a different mechanism needed to be proposed for PcBN materials. Since the flaw size was equivalent to the particle size due to flaws introduced during the machining process, Figure 3.7 should be redrawn with respect to 1/?d, as shown in Figure 3.8, where region 1 and 2 are represented by equations 3.5 and 3.6 respectively. Griffith-Orowan Strength relationship (Region 1): dK f / 1 ?? (3.5) Where ?f is the fracture strength, K1 a constant and d the average grain size. Chapter 3: Strengthening & Toughening of cBN Materials - 36 - Hall-Petch Strength relationship (Region 2): dK Of / 2 ???? (3.6) Where ?f is the fracture strength; ?O the material constant for the starting strength due to dislocation movement, K2 the constant (strength coefficient) and d the mean grain size. Figure 3. 7: Transverse rupture strength as a function of the nominal grain size of diamond for PCD materials (43). Chapter 3: Strengthening & Toughening of cBN Materials - 37 - Figure 3. 8: Strength as a function of the grain size relationship for PCD materials (43). Rice (48) also suggested that the bimodal behaviour of the curve of TRS versus grain size was more likely to be due to the grain size d versus crack size c relationship. He showed that in ceramic materials the inherent flaw introduced by machining and surface finishing procedures were independent of grain size. The surface flaw resulted in a sharp strength reduction in the brittle region (45). Rice suggested that the fracture strength ?f of fine grain size materials was controlled by the polycrystalline surface energy, while for coarse-grained materials the fracture strength was controlled by the single crystal crack surface energy or the size of the defects since the flaw is contained within a single grain. Therefore, when the critical stress intensity factor corresponding to the single crystal was reached, the crack propagated and would lead to failure. According to this theory, the transition grain size occurred at c ? d. From experimental measurements Lammer (43) calculated that for PCD materials the critical ratio of c/d = 1 should be reached between 12 and 30 ?m. This can be seen in Figure Chapter 3: Strengthening & Toughening of cBN Materials - 38 - 3.8, the two lines intersect where the flaw size is equivalent to the particle size, approximately 30 ?m. Lammer (43) also showed that the fracture mode of the coarse grained areas (?30 ?m) was transgranular while the fine grained areas (? 30 ?m) was microplastic flow and microcracking where dislocations lead to crack nucleation. The strength of cBN materials appears to increase with the decreasing grain size of the cBN particles. However, strength is also affected by the largest flaw in the material. Large flaws can form due to agglomerated particles, either as cBN particles sticking to each other or the binder materials can act as defects in the microstructure. These defects are fracture initiation sites, which can lead to the lowering of the strength of the material. It is thus important to have good mixing techniques to produce uniform distributions of all the phases throughout the composite material (53, 54). The strength of the material is also affected by the binder phase; higher binder phase content will increase the chances for flaws to be present in the material and thus increase the risk of premature failure. Lammer (43) suggested that the strength of PCD materials depended also on the amount of secondary phase, cobalt. The strength increases with increasing secondary phase content. Consequently there is a weigh-up with the mechanical properties dependence on grain size and binder content. Large grain sizes are favoured when fracture toughness is the Chapter 3: Strengthening & Toughening of cBN Materials - 39 - critical property, while materials with smaller grain sizes are used where hardness and strength are the most important properties. Low binder content is favoured for materials that require high hardness and medium strength. Thus a balance must be reached depending on application. The dominating effect controlling the mechanical properties needs to be determined as well as the influence of binder content. All these aspects will be analyzed in subsequent chapters. Chapter 4: Experimental Procedure - 40 - Chapter 4: Experimental Procedure This chapter describes in detail the experimental procedures used in the making, characterisation and mechanical testing for PCBN-Al composite materials. A detailed description of all the equipment used for making, characterisation and mechanical testing is also provided in this chapter. The starting powders, mixing procedures and powder characterisation are also described in this chapter. 4.1. Starting Materials All the powders used in this study are given in Table 4.1. The average particle size and manufacturing company are also provided. Table 4. 1: Powders used for processing. Material: Particle Size, d50 (?m) Company Grade M2-4 cBN powder 2.45 Element Six (Pty) Ltd Grade 2 cBN powder 1 Element Six (Pty) Ltd Grade 4 cBN powder 6.37 Element Six (Pty) Ltd Grade 9 cBN powder 14.5 Element Six (Pty) Ltd Grade M20-40 cBN powder 14.75 Element Six (Pty) Ltd Aluminium powder 4.2 Saarchem The starting materials used for this work were crushed cBN powders grades 2, M2-4, 4, 9 and M20-40 (average particle size ranges from: 1 ? 40 ?m) in various grade size Chapter 4: Experimental Procedure - 41 - combinations and Aluminium powder (average particle size: 1 ? 4 ?m). The Aluminium powder was wet sieved before use to remove all large particles (section 4.2.1). The powders were analysed using a scanning electron microscope (SEM) to determine the morphology of the particles. The SEM images of the various cBN grades and the Aluminium powder are shown in Figure 4.1. The images show that the cBN grains are facetted and the Aluminium powder consists of spherical particles. A Malvern Mastersizer 2000 (Section 4.3) was used to measure the particle size and particle size distribution of the powders. Figure 4.2 shows the particle size distribution of the cBN powders. d (10), d(50) and d(90) means that 10, 50 and 90% of the particles are less than the particle size indicated. Chapter 4: Experimental Procedure - 42 - Figure 4. 1: SEM micrographs of the cBN and Aluminium powders used in the PcBN-Al composite; a) grade 2 cBN powder; b) grade M2-4 cBN powder; c) grade 4 cBN powder; d) grade 9 cBN powder; e) grade M20-40 cBN powder and f) Aluminium powder. (a) (b) (c) (d) (e) (f) (e) Chapter 4: Experimental Procedure - 43 - Particle Size Distribution 0 2 4 6 8 10 12 14 16 0.01 0.1 1 10 100 1000 Particle Size (um) Vo lu m e % Al powder G M2-4 G 4 G 9 GM 20 -40 Figure 4. 2: Particle size distribution of the cBN and Aluminium original powders used in the PcBN-Al composites. 4.2. Experimental Equipment for Powder Processing 4.2.1. Wet Sieving of Aluminium Powder The Aluminium powder was wet sieved before use, to remove any particles that were larger than 4?m. Wet sieving was done by mixing small amounts of the Aluminium powder (20 g) with hexane (C6H14) and placing the mixture in a container with a filter cloth. The filter cloth had an aperture of 4?m to allow all the fine particles to pass through and retain the larger particles. The mixture was then mixed using a rotating Aluminium: d(10) = 2.9 ?m d(50) = 4.2 ?m d(90)= 10.5 ?m G M2-4: d(10) =1.12 ?m d(50) = 2.45 ?m d(90)= 4.65 ?m G 4: : d(10) = 1.28 ?m d(50) = 6.37 ?m d(90)= 11.7 ?m G 9: : d(10) = 9.9 ?m d(50) = 14.5 ?m d(90)= 21.25 ?m G M20-40: d(10) = 4.07 ?m d(50) = 14.75 ?m d(90)= 29.0 ?m Chapter 4: Experimental Procedure - 44 - stirrer to assist the movement of particles through the filter cloth. After 1 hour of stirring, the mixture was drained into a container; extra amounts of hexane were added to the container to wash the fine particles through the aperture. The fine particle slurry was then dried in a rotary evaporator. The filter cake on the filter cloth was dried and placed into a container. Figure 4.3 illustrates the apparatus used for wet sieving the Al powder. A rotary evaporator was used to evaporate the solvent (hexane). The hexane gas was re-condensed and the liquid hexane captured in a flask. The temperature and pressure of evaporation were 40?C and 335 mbars respectively. Figure 4.4 shows an image of a rotary evaporator. Figure 4. 3: Wet sieving apparatus Al powder & Hexane Filter cloth Stirrer Filtered slurry Drained filtered slurry for rotary evaporator Chapter 4: Experimental Procedure - 45 - Figure 4. 4: Image of a rotary evaporator. 4.2.2. Turbula Mixer A turbula mixer was used to assist in the mixing of the dry powders. This was done to prepare the powders for sintering. The powders were weighed into their correct proportions and placed into a plastic container. 10 mm stainless steel balls were used as milling media to aid in the mixing and breaking up of agglomerates in the powders. The number of balls depended on the amount of powder that was being made (8 steel balls were used for 30 g of powder). The mixing time was 1 hour. After mixing, the steel balls were removed and cleaned and the mixed powder was decanted into a labelled storage container and placed in a vacuum chamber set at 20?C to prevent oxidation and moisture from affecting the powders. The powders were kept in the vacuum chamber until use. Chapter 4: Experimental Procedure - 46 - 4.2.3. Uniaxial Pressing After the powders had been mixed and placed in to the vacuum oven overnight, the powders were then used for sintering. The powders were placed into the required 50 mm diameter metal containers used in the sintering process. The powders were then uniaxially pressed using a 50 MPa pressure into smooth surfaced compacted powder disks. The containers were then closed with a metal lid and sealed with a metal wire. 4.2.4. Vacuum Degassing The sealed compact containers were then placed in a Torvac vacuum furnace for outgassing and subsequent sealing. The vacuum heat treatment is used to remove all absorbed gases and surface impurities from the powders so as to prevent oxidisation during sintering. After the outgassing stage the containers are heated to a higher temperature to melt the sealing metal wire to seal them from the outside environment. The process of vacuum degassing and sealing takes 10 hours. 4.2.5. High Temperature High Pressure Sintering After the compacts had been outgassed and sealed, the compacts were then prepared for sintering. The compacts were placed into the required containers used for sintering (the exact procedure and type of sintering containers used in the sintering process are classified by Element Six (Pty) Ltd and thus cannot be divulged). Chapter 4: Experimental Procedure - 47 - Sintering was done at high pressure high temperature (HPHT) to facilitate full densification. HPHT belt apparatus was used. This operation was carried out at the pressing facilities at Element Six (Pty) Ltd in Springs, South Africa. The sintering process is information proprietary to Element Six (Pty) Ltd and thus most aspects of this process cannot be divulged. The sintering profile consisted of first by applying pressure and then heating the material to sinter and form fully densified compacts. The approximate temperature of sintering was 1400?C. Figure 4.5 shows the sintering powder and load profile. After sintering the sintered disks were removed and taken for surface processing. The metal container surrounding the sintered PcBN disk was removed by surface and OD grinding (outer diameter) of the PcBN disk. Figure 4. 5: Load and power profile used to sinter the PcBN-Al composites. 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 500 1000 1500 2000 2500 3000 3500 4000 Time (seconds) L o ad (MN ) 0 2 4 6 8 10 12 Load (Ton) Power (kW) Power (kW) Chapter 4: Experimental Procedure - 48 - 4.2.6. Lapping, Surface and OD Grinding Machines Lapping and surface grinding of the sintered compacts were done to remove the outside metal container and to machine the components to the correct dimensions and surface quality for testing. The sample disks were first taken to the surface grinding machine to remove the metal container on the flat surfaces of the disk. Then the disk was taken to the OD (outer diameter) grinding machine to remove the metal container on the circumference of the disk. After the metal container had been removed from the disks, the PcBN disks were lapped in two stages; the first stage was aimed at grinding to the correct disk thickness and the second to prepare the correct surface quality. The first stage of lapping was done using a coarse lapping grit of 120-grit size; this was done for about 1 ? 2 hours to reduce the thickness of the disks to approximately 3.2 mm. The samples were then fine lapped using a fine lapping grit of 300-grit size to further reduce the thickness to 3 mm and to make sure that the surface contained no visible scratches and had a smooth surface quality. 4.2.7. Laser Cutting Test samples were cut to the required dimensions for each of the mechanical tests, using laser cutting. The sample dimensions were 25 x 4 x 3 mm for fracture toughness and strength bend bars. Figure 4.6 shows the arrangement of the bend bars Chapter 4: Experimental Procedure - 49 - laser cut-out of the PcBN disks, with 8 samples being cut out per disk. 4 disks per material were sintered for the bend bar samples, and a total of 28 samples were cut out; 20 for TRS and 5 for SEVNB with the rest used for testing. The compact tension samples measured 36 x 33.5 x 3 mm and hardness samples 10 x 10 x 3 mm. Figure 4.7 shows the arrangement of the CT samples laser cut-out of each PCBN disk, 1 sample being cut out for each disk. 2 PcBN disks per material were sintered for the CT samples. The laser cut samples had a heat affected zone due to the high temperatures of the laser, thus the samples needed to undergo lapping in order to remove the heat affected zone. Figure 4. 6: Laser cutting of bend bars from PcBN sintered disk. 25 mm 4 mm 50 mm 3 mm 1 mm Chapter 4: Experimental Procedure - 50 - Figure 4. 7: Laser cutting of compact tension (CT) samples from PcBN sintered disk. After laser cutting of the samples the bend bars were fine lapped again to remove any residual heat affected zone left from the laser cutting and to straighten out the edges of the bars. The lapping was done using a 120 mesh size. 4.2.8. Polishing Equipment The samples prepared for SEM analysis and hardness testing needed to be polished. The polished PcBN discs were prepared by lapping a cut surface flat and then presenting the flat surface to a diamond scaife wheel. Speeds, loading conditions and details regarding the operation of the scaife wheels for the polishing process are information proprietary to Element Six (Pty) Ltd. 50 mm 33.6 mm 35 mm 3 mm Chapter 4: Experimental Procedure - 51 - 4.3. Powder Processing Procedure 4.3.1. Preparation of cBN Powders The various cBN powders used in this work were mixed together to form multimodal powders. This was done in order to facilitate particle packing and to prevent the larger grains from breaking during sintering on application of pressure. The process of mixing is described in section 4.2.2. Table 4.2 shows the composition of the mixed cBN powders, where G2, G6, G10 and G20 are the names of the mixed multimodal cBN powders used for the sample materials. Table 4. 2: Composition of the mixed cBN powders. cBN powders: Particle Size (?m) Composition: Average Grain Size (?m) G 2 Bi-modal ~ 2 ?m 100 % grade M2-4 2.45 G 6 Multi-modal with 10 ?m as the largest size 80 % grade 4 20 % grade 2 5.96 G 10 Multi-modal with 20 ?m as largest particle size 80 % grade 9 20 % grade 2 11.8 G 20 Multi-modal with < 20 ?m as largest particle size 75 % grade M20/40 10 % grade 4 15 % grade 2 16.19 Aluminium powder, 99.9% pure 4.2 ?m 4.2 Chapter 4: Experimental Procedure - 52 - The powders were mixed in a turbula mixer for 30 minutes; Figure 4.8 shows the mixed cBN powders. The particle size and particle size distribution of these powders were measured using the Malvern Mastersizer 2000 particle size analyser. Figure 4.9 shows the particle size distribution, d10, d50 and d90 of the mixed powders. The multimodality of the G6, G10 and G20 cBN mixed powders is evident in the double peaks. The average particle size of the G6 and G10 mixed powders decreased from the original cBN grade powder due to the high amount (20%) of fine grade 2 cBN powder added. The G20 mixed cBN powder showed an increase in the average particle size due to the addition of the grade 4 cBN powder. This could also be attributed to breakage of the larger particles during mixing. Chapter 4: Experimental Procedure - 53 - Figure 4. 8: SEM images of the mixed cBN powders; a) G6 cBN powder; b) G10 cBN powder and c) G20 cBN powder. (a) (b) (c) G10 mix Chapter 4: Experimental Procedure - 54 - Particle Size Distribution 0.01 0.1 1 10 100 1000 3000 Particle Size (?m) 0 2 4 6 8 10 12 Vo lum e ( %) Gr 9 cBN mix - Average, Monday, May 07, 2007 11:39:49 AM Gr M20/40 cBN mix - Average, Saturday, June 02, 2007 11:28:58 AM Gr 4 cBN mix - Average, Saturday, June 02, 2007 12:00:17 PM Figure 4. 9: Particle size distribution of the mixed cBN powders. 4.3.2. Preparation of the cBN-Al Powders The 10 materials prepared for this study are given in Table 4.3. The samples were prepared in various cBN grade sizes and volume percentages of the Al binder phase. The names of the produced samples consisted in the name of the multimodal powders used, shown in Table 4.2, followed by the volume percentage of Al added. The various cBN powders were mixed with the correct amount of Al to give a 15, 20 and 25 vol. % of Al. Table 4.4 shows the composition of the mixed cBN-Al powders used for the cBN-Al composites. Gr 10 cBN mix ? Average Gr 20 cBN mix ? Average Gr 6 cBN mix ? Average G6 cBN: d (10) = 1.16 ?m d (50) = 5.96 ?m d (90) = 11.18 ?m G10 cBN: d (10) = 1.41 ?m d (50) = 11.8 ?m d (90) = 21.7 ?m G20 cBN: d (10) = 2.27 ?m d (50) = 16.19 ?m d (90) = 31.58 ?m Chapter 4: Experimental Procedure - 55 - Table 4. 3: Sample materials. Sample G2 G6 G10 G20 Average grain size (?m) of cBN 2.45 5.96 11.8 16.2 Vol. % of Al binder phase 15 X X X X 20 X X X 25 X X X Figure 4.10 ? 4.11 shows the SEM images of the mixed cBN-Al powders of the different cBN grades. The white Aluminium particles were homogeneously distributed throughout the cBN powder. Some of the Aluminium powder particles were not spherical but slightly elongated. Although the Aluminium powder was wet sieved it can be seen from the figures that the process was not completely efficient as there were still some Al particles larger than 4 ?m. Chapter 4: Experimental Procedure - 56 - Table 4. 4: Composition of the mixed cBN-Al composite powders Material Super Hard Phase cBN Binder Phase Al Grade Mass % Volume % Mass % Volume % G2 + 15% Al M2-4 87.87 85 12.13 15 G2 + 20% Al M2-4 83.64 80 16.36 20 G2 + 15% Al M2-4 79.31 75 20.69 25 G6 + 15% Al G2 17.57 17 12.13 15 G4 70.29 68 G6 + 20% Al G2 16.73 16 16.36 20 G4 66.91 64 G6 + 25% Al G2 15.86 15 20.69 25 G4 63.45 60 G10 + 15%Al G2 17.57 17 12.13 15 G9 70.29 68 G 20 + 15% Al G2 13.18 12.75 12.13 15 G4 8.79 8.5 M20-40 65.9 63.75 G 20 + 20% Al G2 12.55 12 16.36 20 G4 8.36 8 M20-40 62.73 60 G 20 + 25% Al G2 11.9 11.25 20.69 25 G4 7.93 7.5 M20-40 59.48 56.25 Chapter 4: Experimental Procedure - 57 - Figure 4. 10: SEM image of the grade G2 cBN-Al mixed powders; a) G2 cBN + 15 vol. % Al powder; b) G2 cBN + 20 vol. % Al powder; c) G2 cBN + 25 vol. % Al powder; d) G6 cBN + 15 vol. % Al powder; e) G6 cBN + 20 vol. % Al powder and f) G6 cBN + 25 vol. % Al powder. (a) (d) (c) (b) (f) (e) Chapter 4: Experimental Procedure - 58 - Figure 4. 11: SEM image of the G10 and G20 cBN-Al mixed powders; a) G10 + 15 vol. % Al powder; b) G20 cBN + 15 vol. % Al powder; c) G20 cBN + 20 vol. % Al powder and d) G20 cBN + 25 vol. % Al powder. 4.4. Characterisation and Analysis 4.4.1. Malvern Particle Size Analysis The Malvern Mastersizer 2000 Particle Size Analyser (PSA) was used to determine the particle size distribution of all powders, raw and mixed, before use. The powders were suspended in a mixture of water and a dispersant (Sodium hexametaphosphate, (a) (b) (c) (d) Chapter 4: Experimental Procedure - 59 - (NaPO3)6). A small amount of the suspended solution was poured into the PSA analysis tank where an ultrasonic probe was used to disperse the particles into the water. The PSA measures the light intensity as a function of the scattering angle of the suspended solution and determined the particle size and particle size distribution of the suspended powder particles. The particle size analysis of each powder was performed at least 3 times to obtain an average of the results. 4.4.2. XRD Phase Analysis X-Ray diffraction (XRD) was used to qualitatively determine the phases present in the composite material. The samples were examined using a PW 1710 Philips powder diffractometer, using monochromatic Cu (K?) radiation produced at 40 KV and 20 mA. Diffractograms were collected over a range of 2? between 20? and 80o, with a step size of 0.02o 2?, together with a scan step time of 4.0 s. Sintered samples found to contain peaks of unreacted-Al were subjected to Rietveld refinement analysis to determine the amount of elemental Al left in the samples. Rietveld refinement was done using Topas A version 4.1, and structural models used were taken from the ICSD - version 1.4.4 Powder X-ray diffraction data were collected using a Bruker AXS D8 equipped with a primary beam G?bel mirror, a radial Soller slit, a V?ntec-1 detector and using Cu- K? radiation (40kV, 40mA). Data were collected in the 2? range of 5? to 100? in Chapter 4: Experimental Procedure - 60 - 0.021? steps, using a scan speed resulting in an equivalent counting time of 73.2 s per step. 4.4.3. Microstructural SEM and EDS Analysis A Scanning Electron Microscope (SEM) (LEO1525FE) was used to study the microstructure of the composite materials. An EDX (energy dispersing X-ray) spectrometer was used to identify the elemental composition of the various phases in the microstructure. This was used to support the evidence of the XRD analysis and to find evidence of free aluminium that could not be found in the XRD analysis. The samples were non-conductive; an acceleration voltage of 2 ? 20 keV was used. A low voltage was used for the fracture surface analysis while high voltages were used for the study of the microstructure. Double-sided carbon tape was used to mount the samples onto the aluminium sample holders and also to provide the electrical conductivity. The samples were not coated. An assortment of magnifications of micrographs were taken ranging between 100 ? 10 000X. SEM was also used in the fractographic analysis of the fracture surfaces. 4.4.4. Density The density of the samples was measured using the Archimedes principle. The sample?s dry weight, Wdry, was measured. The samples were immersed in water, and Chapter 4: Experimental Procedure - 61 - boiled for 3 h in order to displace air from the pores in the samples. After boiling, the samples in the water were allowed to cool to room temperature. The samples were then removed from the water and lightly dried using some cloth or tissue paper to remove excess water on the surface of the samples. The weight of the samples was measured and the saturated weight Wisat was recorded. The samples were then suspended in water and the weight of the saturated samples Wsusp were measured. The temperature of the water was also measured to determine its density. These measured quantities were then used to calculate the densities of the samples. The sample density is determined through equation 4.1. isussat OofHTempOHdrydry WW W V W )( 2@2 ? ? ?? ?? (4.1) 3 2 .99756.0)23@( ?? cmgCOH ?? The density was determined at least twice per sample. Between each measurement the samples were dried in an oven. The theoretical density of a composite material depends on the amount of the phases present in it. In this work, the amount of aluminium added to the boron nitride ranged between 15 and 25 vol. %. The density of the cBN used for the calculation of the theoretical density of the composite was taken from the data in Table 2.1 in Section 2.2. The densities of Al (2.701 g/cm3) and cBN (3.48 g/cm3) were used for the calculation of the theoretical density. Chapter 4: Experimental Procedure - 62 - Using the following equation: i im ? ? ? ? 100 (4.2) Where ?, ?i and mi is the theoretical density of the composite, the density and mass % of each component i respectively. This formula is based on the rule of mixtures. The masses of each starting component, as well as the calculated theoretical density are given in Table 4.5 and 4.6 for the powder mixture and sintered materials respectively. Table 4. 5: The theoretical densities of the cBN-Al powder mixtures. Vol. % of Al mass % of Al mass % of cBN Density ? (g/cm3) 15 12.13 87.87 3.36 20 16.36 83.64 3.32 25 20.69 79.31 3.28 Table 4. 6: The theoretical densities of the cBN-Al sintered material. Vol. % cBN Vol.% AlN Vol.% AlB2 Density ? (g/cm 3) 85 9.37 5.63 3.41 80 12.50 7.50 3.40 75 15.62 9.38 3.38 Chapter 4: Experimental Procedure - 63 - 4.4.5. Grain Size and Phase Composition Image analysis techniques were been used to quantify and characterise the microstructure of the sintered PCBN-Al composite materials. The microstructures of the sintered materials were looked at in a Phillips-ESEM XL30 and 10 SEM micrographs of the microstructure were taken. The magnification of the micrographs depended on the grain size of the cBN materials; a high magnification was used for G2 and G6 materials while a low magnification was used for G10 and G20 materials. The micrographs included in this work represented the average microstructure for each material. Using the image analysis software program (AnalySIS ?Pro?) the grain size and the percentage of the binder and cBN phases were determined. In the micrographs analysed, areas with different light intensity were ascribed to different phases, thus allowing for a numerical determination of different phase constants. The cBN phase shows a lower light intensity than the binder phase. The grain size and grain size distribution were determined by measuring the grain size for each grain and deriving a distribution. The data from the AnalySIS program was then processed automatically by using the Mathcad and Mathconnex programs. The program module was designed to calculate the grain size distribution and mean free path of both the cBN and binder phases for the PcBN materials. A detailed Chapter 4: Experimental Procedure - 64 - analysis of the Image analysis techniques for the grain size and phase composition determination is given in Appendix A. 4.5. Mechanical Properties and Testing Techniques 4.5.1. Hardness Hardness, H, is generally defined as a material?s resistance to plastic deformation usually due to a static penetration by a harder material (11). Hardness is an important parameter to describe the mechanical properties of a material, and can be determined by measuring the depth of penetration of a harder indenter, usually diamond. It influences a number of performance aspects of ceramics including compressive strength, wear, as well as machining and ballistics performance (48). The indenter is loaded perpendicularly to the planar surface of the material, the area of indentation is measured and the hardness is thus related to the load divided by the area of plastic contact. Thus hardness can be defined as the ratio of the applied load P to the area of plastic contact A. The relationship defining hardness is given below (55): 2d P A P H ??? (4.3) Where H is the hardness P ? Applied load Chapter 4: Experimental Procedure - 65 - A ? Area of plastic contact d ? Characteristic size of the indentation impression ? ? Constant dependent on the indenter geometry The hardness testing of a material is an easy and simple way to determine the bulk mechanical properties of a material. The use of rigid indenters causes the material to have a highly strained crack system, which develops beneath the indentation. This could cause errors in the results. There are various test techniques to determine the hardness of a material, namely the Vickers, Knoop, Brinell and Rockwell tests. They are all based on the penetration of indenter into a softer material, but differ in the indenter geometry. Hardness of a material depends on the test method and indenter geometry (i.e. Brinell, Vickers, Knoop), and varies with applied load, indenter shape and dimensions, as well as the microstructure of the material. There are two major indenter geometries used widely for brittle materials like ceramics, namely Vickers and Knoop. The Vickers testing method is the standard method used to determine the hardness of various materials, especially those that have hard surfaces, as it is a reliable method of measuring the hardness and a type of indenter that can be used for various types of materials. Knoop testing can also be used for materials that are very brittle ceramics; it is used to define the relative microhardness of a material. Chapter 4: Experimental Procedure - 66 - Knoop hardness measurements generally provide a lower value than the Vickers hardness. Vickers hardness HV, is measured on the contact area of the indentation impression, while Knoop hardness HK, is calculated on the projected area of the indentation impression. Equation 3.1 can be used to calculate the hardness for both methods with ? corresponding to 1.8544 for Vickers and 14.229 for Knoop. The characteristic size of the resultant indentation d changes for the two methods, for Vickers d is defined as the average of the two diagonals of the resultant squared shaped impression, while for Knoop it is the length of the diagonal of the resultant rhombic impression. 4.5.1.1. Vickers Hardness The Vickers method to determine the hardness of ceramic materials was used in this study. The Vickers indentation method uses a square based, pyramid shaped diamond indenter with 136? angle between the two opposite faces, as shown in Figure 4.12. Figure 4. 12: The configuration of the Vickers diamond indenter. Chapter 4: Experimental Procedure - 67 - A load is applied to the indenter which then causes an indentation in the sample material. After the load is removed, the two diagonals of the indentation left in the material are measured using a microscope and the averages are taken. The Vickers hardness, HV, of the material can then be calculated by the following equation (56): 2 854.1 d L HV ?? (4.4) Where L is the normal load (kgf) and d is the average length of the two diagonals (mm) of the Vickers indentation (56). The Vickers hardness results are given in kgmm- 2, but are more commonly expressed in SI units, GPa. 4.5.1.2. Knoop Hardness The Knoop testing method uses an elongated four-sided pyramid, with the angles between the two of the opposite faces being approximately 170? and the angle between the other two sides being 130?. The configuration of the Knoop diamond indenter is shown in Figure 4.13. Similar to the Vickers method a load is applied to the indenter which causes an indentation in the sample material. After the load is removed, the longest of the two diagonals of the indentation left in the material is measured using a microscope. Knoop hardness, HK, of a material can be calculated from the following equation (57): 2 229.14 d L H K ?? (4.5) Chapter 4: Experimental Procedure - 68 - Where d is the length of the longest diagonal of the rhombic impression, which is approximately 7 times the width of the indentation, w. Figure 4. 13: The configuration of the Knoop Hardness indenter. For this project the Vickers hardness technique was used to determine the hardness of the PcBN-Al composite materials. Samples need to be correctly polished so that the microstructure can be seen before hardness measurements can be done. The procedure for this is given in section 4.2.8. At least 10 hardness measurements were done per sample to get an average hardness value and a standard deviation. The loading conditions for the hardness measurements were a 5 kg load and a dwell time of 10 seconds. The average hardness and standard deviation for the hardness measurements were determined. Chapter 4: Experimental Procedure - 69 - 4.5.2. Fracture Toughness and R-curve Behaviour There are a variety of different methods worldwide that can be used for measuring the fracture toughness of brittle materials and thus there is variability between the results of the various testing methods. The methods used for measuring the fracture toughness of brittle material range from indentation techniques to those involving crack propagation and the fracture of larger samples such as the double cantilever, double torsion, single edge notched beam, chevron notched beam or compression disk methods. The standard testing methods for fracture toughness of advanced ceramics is described in ASTM C1421-09 (58). The choice of an appropriate technique for PCBN materials is limited to the size of the sintered disk produced, which tends to be small. Preference of the fracture toughness test for this study was the single edge v-notch beam method (SEVNB) as it is a simple method based on the notch root radius, which gives relatively accurate results for all brittle materials, and samples can be made to the correct sample dimensions. The SEVNB technique has also been described in various standards and been subjected to extensive round robin practices (59). 4.5.2.1.Single Edge V-Notch Beam (SEVNB) The single edge V-notch beam method was used to determine the mode I fracture toughness, KIc, of the brittle materials made in this work. The test involved loading a Chapter 4: Experimental Procedure - 70 - rectangular bar specimen which had a sharp V-shaped notch in one edge of the sample to failure. The test jig used for the SEVNB fracture toughness testing was the four-point flexure test jig. Figure 4.14 shows the four-point bend test jig and mechanical testing machine used for flexural testing. The samples? dimensions were 25 x 4 x 3 mm3, thus the load and support span needed to be 10 and 20 mm apart respectively. The SEVNB method used is based on the fracture resistance method by K?bler (1999) (59, 60). The SEVNB samples were arranged so that the notched edge was in the tensile position (i.e. the notched surface opposite to the loading span) as seen in Figure 4.14. The samples were placed into the 4-point bend test jig and loaded to fracture in the universal servohydraulic testing machine MTS 810.22, 100 kN load frame, a 5 kN load cell being used to measure the load of fracture. The loading rate was 5 ?m/s, equilvalent to ~ 100 N/s. TestStar was the control program used to measure the loading force; the maximum load to fracture was recorded. For fracture toughness 4 ? 6 specimens were tested per materials composition. Chapter 4: Experimental Procedure - 71 - Figure 4. 14: The SEVNB arrangement in four-point bend configuration. 4.5.2.1.1. Machining of the V-Notch Samples for the SEVNB fracture toughness testing required a 1 mm sharp V-shaped notch cut into the tensile surface. This was done in two stages; in the first stage 4 samples were aligned and glued onto a metal sample holder next to each other with the 3 mm tensile polished edge facing upwards. A 0.3 mm thick diamond saw was used to cut a 800 ?m deep starter notch into the sample surface. The second stage involved cutting a sharp V-notch into the starter notch using a razor blade cutting machine. The razor blade edge in the razor blade holder was aligned with the middle of the starter notch and adjusted until it was straight and secured tightly in the holder. The notch was then filled with a fine diamond paste between 1 and 6 ?m to facilitate cutting. The razor blade was put back into the starter notch and the blade was moved 10mm 20 mm 3 mm 4 mm Applied Force F ? 5 mm 25 mm Chapter 4: Experimental Procedure - 72 - backwards and forwards by a motor. The razor blade was changed every 30 minutes and extra diamond paste was used every 15 minutes. The larger size diamond paste was used first to facilitate the cutting process, followed by 3 and 1 ?m paste to develop a sharp well-defined V-notch. The sample was removed every 1 hour to measure the depth and sharpness of the V- notch by optical microscopic examination. The length of the notch was measured at three evenly spaced intervals across the sample width and then averaged. The relative deviations between the three measurements were not to exceed 0.1mm, to prevent a skew notch depth so as not to obtain inaccurate results. The notch tip radius needed to be between 10 ? 30 ?m and the total length of the notch 1 mm. Figure 4.15 shows the razor cutting machine and a V-notch cut into the PcBN-Al sample. Figure 4. 15 a and b: a) The razor blade cutting machine to cut V-notches into the sample and b) a V-notch cut into the surface of the PcBN-Al sample. (a) (b) Chapter 4: Experimental Procedure - 73 - The mode I fracture toughness, K1c (MPa. m ) can be calculated for a 4-point flexure by the following equations (55, 59): ? ? Y W SS WB F YaK Ic 5.1 21 12 3 .. ? ?? ? ? ?? (4.6) With: ? ? ? ? ? ?2 2 1 135.168.049.3 326.19887.1 ? ????? ? ??? ???Y (4.7) W a ?? (4.8) Where ? is the relative notch depth, a is the average notch length (m), W is the specimen width (m), B is the specimen thickness (m), F is the fracture load (MN), Y stress intensity shape factor and S1 & S2 are the support spans (S1>S2). The relative notch depth (?) is defined as the ratio of the notch length ( a ) to the width of the sample (W) and should be between 0.2 < ? < 0.3. 4.5.2.2. R-curve Behaviour Compact Tension (CT) Binders and other materials are used to improve the fracture toughness of brittle materials. These additives can provide toughening mechanisms, which will blunt the crack tip from the applied stress. Fracture toughness resistance is determined by the toughening mechanism of the material. Open displacement of the crack front can be Chapter 4: Experimental Procedure - 74 - used to determine the toughening mechanisms of a composite material by looking at its R-curve behaviour. The Compact tension (CT) method was used to determine the crack opening displacement for R-curve, v-K curve and fracture toughness, KIC. The sample dimensions for the compact tension were 35 x 33.6 x 3 mm. A notch and 2 holes were laser cut into the surface of the sample. Figure 4.16 shows the sample dimensions of the CT sample. A wire saw cut a diagonal cut into the end of the notch, such that the angle between the cut and the surface was approximately 30?. Figure 4.17 shows the cross section of the CT sample. A Knoop hardness indentation was positioned at the edge of the notch. Figure 4. 16: The compact tension (CT) sample dimensions. 33.5 mm 35 mm 3 mm 16 mm ? 10 mm Chapter 4: Experimental Procedure - 75 - Figure 4. 17: Cross section of the CT sample The Compact tension samples were carefully placed into the Compact Tension testing device as illustrated in Figure 4.18. A 1 kN load cell was used to measure the loading force. A preload of 20 N was applied to the sample to cause a crack to form from the end of the indentation; the crack was then propagated to a length of 1- 2 mm. The sample was removed from the CT device and a new cut was made horizontally such that the diagonal cut was removed, but not the tip of the crack. This would ensure the stable, symmetrical loading of the crack. Drilled hole 30? angle Position of the Knoop indentation End of the notch Diagonal cut Chapter 4: Experimental Procedure - 76 - Figure 4. 18: Schematic of the testing device used for R-curve testing (61). The sample was placed back into the CT device and a preload of 20 N was applied to the sample. The length of the crack was measured from the notch end to the crack tip. The sample was then loaded slowly until the crack started to grow at a steady rate and the crack length and load were also measured for every 0.2 mm. The load was applied until the sample broke. The equation to calculate the fracture toughness resistance for the Compact Tension samples is given by (61): ? ? ? ? ? ? ? W c f BW F K IR 2 1 (4.9) 2 3 432 1 6.572.1432.1364.4886.02 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ? ?? ? ? ? ? ? W c W c W c W c W c W c W c f (4.10) Chapter 4: Experimental Procedure - 77 - Where: W c is the ratio of the crack length to the width of the sample and is valid for: 0.2 < W c < 1 4.5.3. Transverse Rupture Strength (TRS) Transverse Rupture Strength (TRS) or Flexural strength is the property that determines the strength of a material. It is the maximum flow stress that occurs during the deformation, cracking and crack propagation process before the material fractures and is determined by measuring the flexural strength (the measure of the ultimate strength) of a bar of a material in bending. The flexural strength of a material is dependent on both its inherent resistance to fracture and the presence of defects. The size of the flaws in the material can greatly decrease its strength. Failure in ceramic materials is caused by the extension of flaws, which are usually introduced in it during processing or surface treatment such as grinding and polishing. Flaws in the material can be inclusions, pores or cracks. There are various internationally recognized standards that exist for the determination of flexural strength. The standard test method to determine the flexural strength of advanced ceramic materials at ambient temperatures is given in ASTM C 1161 (62) and ISO 14704:2000 (63). The common specified dimensions for the 3PB and 4PB tests are 20, 40, 80 mm for the support span and 10, 20, 40 mm for the loading span Chapter 4: Experimental Procedure - 78 - respectively. The dimensions used for this study were 25 x 4 x 3 mm with loading and support spans of 10 and 20 mm respectively. The TRS of a material can be measured by flexural bend tests. These tests are easy to perform and they reveal selectively the worst flaws present in the stressed volume. Ultimately TRS is the stress required to make the longest crack in the sample propagate unstably and relate this to the statistical distribution of the stresses and flaws. TRS is the stress calculated from the flexural formula required to break a specimen in the shape of a beam. The specimen is supported near its ends and a force is applied midway between the supports (in a 3 or 4-point bending test configuration). For this study a 4-point bend configuration was used and is illustrated in Figure 4.19. The sample dimensions were 25 x 4 x 3 mm and a 0.1 mm chamfer was machined into the edges of the sample to ensure that there were no machining defects at the edge of the sample to cause premature failure. The transverse rupture strength was determined using the standard methods for flexural testing as given in ASTM C 1161 (62) and ISO 14704:2000 (63). The bending tests were carried out on a universal servohydraulic testing machine MTS 810.22, 100 kN load frame and a load cell of 5 kN. The crosshead loading speed of 5 ?m/s, equivalent to ~ 100 N/s was used. For the strength testing 20 Chapter 4: Experimental Procedure - 79 - samples per material composition were tested. The control program used to measure the loading force was TestStar; the maximum load to fracture was recorded. Figure 4. 19: The 4-point bend configuration for the Transverse Rupture Strength (TRS) tests. TRS is determined by measuring the flexural strength of a material in a 4-point bending flexure, as shown in Figure 4.19. The flexural strength calculation can be determined from the ISO 14704:2000 (63) standards. Flexural strength is the measure of the ultimate strength of a specific beam in bending. The standard formula for the strength of a beam in a 4-point flexure is as follows (63): 2 3 bd Fa f ?? (4.11) Where 2 12 SSa ? ? Where: 25 mm 10 mm 20 mm 4 mm 3 mm ? 5 mm Applied Force F Chapter 4: Experimental Procedure - 80 - F = load at failure (MPa), a = length of the moment arm (mm) S1 and S2 = outer and inner support span (mm), b = specimen width (mm), and d = specimen thickness (mm) Since standard flexural strength data do not follow normal distribution, a statistical analysis of the strength data needs to be done to determine the statistical distribution. For the statistical analysis of the strength data the Weibull distribution approach was utilised. Section 4.5.3.2 details the use of the Weibull distribution. 4.5.3.1. Weibull Distribution Brittle materials exhibit scatter of failure strengths. Thus with brittle materials, the strength property needs to be analysed using probabilistic approaches. The probability of existence of a critical crack size is higher in a larger volume of test specimen. Therefore brittle materials tend to exhibit a volume dependent strength behaviour; this means that the higher the loaded volume of test specimens the lower the strength. Weibull (1939) (64) developed a relationship that would relate the probability of failure to the applied stress. In the Weibull analysis it is assumed that the material is Chapter 4: Experimental Procedure - 81 - isotropic and statistically homogeneous and that the most critical flaw leads to failure. This implies that the probability of finding a flaw in the loaded volume of the material is equal in every direction. The second assumption is based on the concept of the ?weakest-link? theory; the largest flaw will cause failure. Weibull proposed a two- parameter distribution function, which would characterise the strength of brittle materials and relate it to the probability of failure. The probability of an occurrence of a critical defect in a volume V is given by equation 4.12 (54). ? ? ? ? ? ? ? ? ??? OV V VP exp1)( (4.12) Where VO is the mean volume occupied by a defect and thus 1/VO is the concentration of the critical defects. The concentration of critical defects increases with increasing applied stress according to a powder law. Weibull developed an expression to relate the probability of occurrence of a defect with applied stress. m OOV ? ? ? ? ? ? ? ? ? ? ?1 (4.13) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??? m O VVP ? ? exp1)( (4.14) Chapter 4: Experimental Procedure - 82 - The probability of failure of a sample with a constant volume V is given by the following equation (65): ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??? m O fP ? ? exp1 ? > 0 (4.15) Where ?O is the Weibull characteristic strength (failure stress at 63.2% failure probability), ? is the applied uniaxial tensile stress and m is the Weibull modulus. The Weibull modulus m indicates the nature, severity and dispersion of the flaws; a low m value indicates a broad distribution of defects and therefore strength. A high m value indicates a narrow distribution of flaws with a narrower strength distribution (66). Typically for ceramic materials m varies between 3 and 25 (53, 66), depending on the processing conditions. The minimum amount of valid tests that need to be performed to meet the statistic requirements is 20, although more should be used to get better statistics. Weibull analysis of the strength data yields characteristic strength, ?O, and Weibull modulus, m. These parameters were evaluated by the Maximum Likelihood method according to the European Standard EN 843-5 (67). The 80% confidence intervals of ?O and m were determined according to a study on the ?Interferences on the parameters of the Weibull distribution? by Thoman (1969) (68). Chapter 5: Results of HPHT Sintering - 83 - Chapter 5: Results of High Pressure High Temperature Sintering This chapter describes the results of the characterisation of the sintered PcBN-Al composite materials. The characterisation of the PcBN-Al composites includes SEM micrographs, XRD scans and density measurements. The grain size distribution and phase composition of these composites is also described in this chapter. 5.1. G 2 cBN ? Al Materials 5.1.1. Microstructure The cBN-Al powders were sintered using the HTHP method as explained in chapter 4.2.5. The sintered materials were then cut and polished according to the procedures described in Chapter 4. The microstructure of the polished sintered cBN-Al materials were examined in the SEM at various magnifications. Figure 5.1 shows the SEM images of the microstructure of the sintered G2 cBN-Al composites. The SEM images of show that these materials were nearly fully densified. The microstructures were homogeneous although larger binder pools can be seen in both the 20% and 25%Al materials. These binder pools were in the region of 5-10 ?m and 10-20 ?m respectively. Chapter 5: Results of HPHT Sintering - 84 - Figure 5. 1: Microstructure of the G2 cBN-Al sintered materials. a and b) G2 cBN+15vol.% Al; c and d) G2 cBN+20vol.% Al; and e and f) G2 cBN+25vol.% Al. The microstructure of the G2 cBN-Al composite materials consisted of cBN grains (dark phase) surrounded by the residual binder phase. The binder phase consisted of aluminium nitride (AlN) and aluminium boride (AlB2). The AlN phase is the lighter phase while the AlB2 is the darker grey phase. These phases are confirmed by XRD cBN AlN AlB2 AlN cBN AlB2 cBN AlN (a) (b) (f) (e) (c) (d) Chapter 5: Results of HPHT Sintering - 85 - scans shown in Figures 5.2-5.4. These phases are consistent with Walmsley (1987) (1) and Zhao (2007) (27). 20 30 40 50 60 0 500 1000 1500 2000 2500 o? ? ooo o C ou nt s Position [ O 2 Theta] ? BN o AlN Figure 5. 2: XRD scan of the PcBN G 2 + 15 vol. % Al sintered material. 0 10 20 30 40 50 60 70 80 0 500 1000 1 2000 * * * ** * * oooo oo o ? ? ? C ou nt s Position [ o 2 Theta] ? BN o AlN * AlB 2 Figure 5. 3: XRD scan of the PcBN G 2+ 20 vol. % Al sintered material. Chapter 5: Results of HPHT Sintering - 86 - 20 30 40 50 60 70 80 0 500 1000 1500 ** * * ? ? ? ? BN o AlN * AlB 2 OO O O OO O Co u n ts Position [ o 2 Theta] Figure 5. 4: XRD scan of the PcBN G 2 + 25 vol. % Al sintered material. Tests were done to try to reduce the size of the binder pools for the G 2 + 25%Al; this was done by increasing the energy of mixing. Turbula mixing was carried out for longer time periods ranging from 1 ? 10 hours and also increasing the number (10 balls) and/or size of the steel balls (20 mm). The increase in the energy intensity of mixing was used to deform the aluminium particles and to increase the contact area between the cBN grains and the aluminium particles. The intention was to decrease the size of the pools through this process. After turbula mixing of the cBN and Al powders for 2, 4, 6, 8, and 10 hours, SEM analysis was done on the mixed powders. The results show very little difference in the cBN-Al contact. Two materials were then sintered with the 6-hour and 10-hour mixed powders to determine whether there was any change of obtaining an improved microstructure with increased mixing time. Both sintered materials showed only a Chapter 5: Results of HPHT Sintering - 87 - small reduction in the binder pool size and population as can be seen in Figure 5.5 a and b, compared with the 1 hour mixed sample in Figure 5.5 c. Increasing the number and size of the steel balls also had very little effect on the reduction of the binder pools. The reduction of the binder pools could be achieved by changing the mixing process to a wet mixing process like attrition milling, but wet milling contains its own problems like contamination of the powders from the milling media and the oxidation of the Aluminium powder. It was thus decided to keep the original 1-hour mixing time and maintain the large binder pools. When amount of binder phase was too large, there was insufficient space between the cBN grains for the binder to infiltrate into the green compact when melting, thus causing large build-ups of the binder phase. All materials made with a binder content of 25 vol. % Al would suffer from these large binder pools. Chapter 5: Results of HPHT Sintering - 88 - Figure 5. 5: SEM micrograph of the G 2 cBN + 25 vol. % Al sintered material turbula for a) 6 hours, b) 10 hours and c) 1 hour. 5.2 G 6 cBN ? Al Materials 5.2.1. Microstructure The microstructure of the polished sintered G6 cBN-Al materials were examined in the SEM at various magnifications. Figures 5.6 ? 5.7 show the SEM images of the microstructure of the sintered G6 cBN-Al composites. The SEM images of the (a) (b) Chapter 5: Results of HPHT Sintering - 89 - microstructure of the G6 cBN-Al materials show that they were almost fully densified. The microstructures were homogeneous although an increase in the size and amount of the larger binder pools can be seen in both the 20 and 25%Al materials. The binder pools for the G6 cBN + 15%, 20% and 25% Al materials were in the region of 3-5 ?m, 5-15 ?m and 10-20 ?m respectively. The microstructure of the G6 cBN-Al composite materials consisted of cBN grains (dark phase) surrounded by the residual binder phase. The binder phase consists of aluminium nitride (AlN) and aluminium boride (AlB2). The AlN phase is the lighter phase while the AlB2 is the darker grey phase. These phases are confirmed by XRD traces shown in Figures 5.8-5.10. The free Al was confirmed in the XRD scan although the quantity was very small. These phases were consistent with the findings of Walmsley (1987) (1) and Zhao (2007) (27). A small amount of free (unreacted) aluminium was found in the G6 cBN + 25vol.% sample with EDS, Figure 5.7 d. The amount of unreacted Aluminium was determined to be approximately 1.3 wt.%, determined by Rietveld refinement analysis outlined in chapter 4.4.2. The result of the Rietveld analysis is shown in Appendix B. Chapter 5: Results of HPHT Sintering - 90 - Figure 5. 6: SEM images of the microstructure of the PcBN G 6 ?Al composite materials: a) and b) G6 cBN+15% Al; c ) and d) G6 cBN+20% Al; cBN AlN AlB2 AlN cBN AlB2 (a) (b) (c) (d) Chapter 5: Results of HPHT Sintering - 91 - Figure 5. 7: SEM images of the microstructure of the PcBN G 6 +25vol.%Al composite materials: a) and b) micrographs c) EDS of a cBN grain and d) EDS image showing unreacted Aluminium. 20 30 40 50 60 0 500 1000 1500 2000 2500 o ? ? * * * * o o o o C o u n ts Position [ o 2 Theta] ? BN o AlN * AlB 2 Figure 5. 8: XRD scan of the G 6 cBN + 15 vol. % Al sintered material. cBN AlN Free Al AlB2 (a) (b) (c) (d) Chapter 5: Results of HPHT Sintering - 92 - 10 20 30 40 50 60 70 80 0 1000 2000 3000 4000 5000 6000 7000 ? ? ? *** * * ** oo o o o o o C ou nt Position [ o 2 Theta] ? BN o AlN * AlB 2 Figure 5. 9: XRD scan of the G 6 cBN + 20 vol. % Al sintered material. 10 20 30 40 50 60 70 80 0 1000 2000 300 4000 5000 6000 * o x ? ? ? * x**** * oo o o oo o C o u n t Position [ o 2 Theta] ? BN o AlN * AlB 2 x Al Figure 5. 10: XRD scan of the G 6 cBN + 25 vol. % Al sintered material. Chapter 5: Results of HPHT Sintering - 93 - 5.3. G 10 cBN ? Al Materials 5.3.1. Microstructure The microstructure of the sintered G 10 cBN + 15 vol. % Al material can be seen in Figure 5.11. This shows that the microstructure was almost fully densified and had a homogeneous structure. The average binder pool was about 3-8 ?m. The microstructure consisted of cBN grains (dark phase) surrounded by aluminium nitride (AlN) (lighter phase) and aluminium boride (AlB2) (grey phase). This was confirmed by XRD phase analysis as seen in Figure 5.12. These phases were consistent with Walmsley(1987) (1) and Zhao (2007) (27). No traces of unreacted Aluminium were found in XRD scans suggesting that all the Aluminium had reacted. Figure 5. 11: SEM micrograph of the G 10 cBN + 15 vol. % Al sintered material. AlN AlB2 cBN (a) (b) Chapter 5: Results of HPHT Sintering - 94 - 10 20 30 40 50 60 70 80 0 500 1000 1500 2000 2500 3000 3500 * ? ? ? * * ** ooo o oo o C oun ts Position [ o 2 Theta] ? BN o AlN * AlB 2 Figure 5. 12: XRD scan of the G 10 cBN + 15 vol. % Al sintered material. 5.4. Grade 20 cBN ? Al Materials 5.4.1. Microstructure The microstructure of the polished sintered G20 cBN-Al materials was examined in the SEM at various magnifications. Figures 5.13-5.14 show the SEM images of the microstructure of the sintered G20 cBN-Al composites. The SEM images of these materials show that they had been sintered to full density. The microstructures were homogeneous although an increase in the size and amount of the larger binder pools can be seen in both the 20 and 25%Al materials. The binder pools for the G20 cBN + 15%, 20% and 25% Al materials were in the region of 5-10 ?m, 8-13 ?m and 12-20 ?m respectively. Chapter 5: Results of HPHT Sintering - 95 - The microstructure of the G20 cBN-Al composite materials consisted of cBN grains (dark phase) surrounded by the residual binder phase. The binder phase consisted of aluminium nitride (AlN) and aluminium boride (AlB2). The AlN phase is the lighter phase while the AlB2 is the darker grey phase. These phases were confirmed by XRD scans shown in Figures 5.15-5.17. These phases were consistent with Walmsley (1987) (1) and Zhao (2007) (27). A small amount of free Al was found by SEM in combination with EDS and would be seen in the G20 cBN + 20vol.% and 25vol.%Al composite materials. Figure 5.14 c shows the unreacted Al in the G20 cBN + 25vol.% Al sample. The amount of unreacted Aluminium was determined using Rietveld refinement analysis (Chapter 4.4.2) and was approximately 1.1wt.% and 2.7 wt.% for the 20vol.% and 25vol.% respectively. The result of the Rietveld analysis is shown in Appendix A. Chapter 5: Results of HPHT Sintering - 96 - Figure 5. 13: SEM images of the microstructure of the PcBN G 20 ?Al composite materials: a and b) G20 cBN+15vol.% Al; c and d) G20 cBN+20vol.% Al. cBN AlN AlB2 AlN AlB2 cBN (d) (c) (a) (b) Chapter 5: Results of HPHT Sintering - 97 - Figure 5. 14: SEM images of the microstructure of the PcBN G 20 + 25vol.% Al composite materials: a and b) microstructure, c) EDS of unreacted Aluminium and d) EDS of AlB2. 0 20 40 60 80 0 1000 2000 3000 4000 5000 6000 ? ? ? *** * * o oo oo o o o ? BN o AlN * AlB 2 C ou nt s Position [ o 2 Theta] Figure 5. 15: XRD scan of the G 20 cBN + 15 vol. % Al sintered material. cBN Free Al AlN AlB2 (a) (b) (c) (c) (d Chapter 5: Results of HPHT Sintering - 98 - 10 20 30 40 50 60 70 80 0 2000 4000 6000 8000 10000 *x ? ? ? * **** o* oooo oo o C ou nt Position [ o 2 Theta] ? BN o AlN * AlB 2 x Al Figure 5. 16: XRD scan of the G 20 cBN + 20 vol. % Al sintered material. 10 20 30 40 50 60 70 80 0 2000 4000 6000 8000 10000 x x ? ? ? * ** * * ooo o o oo o C ou nt Position [ o 2 Theta] ? BN o AlN * AlB 2 x Al Figure 5. 17: XRD scan of the G 20 cBN + 25 vol. % Al sintered material. Chapter 5: Results of HPHT Sintering - 99 - 5.5. Grain Size and Composition The grain size distribution, volume fraction and mean free path of the cBN and binder phases, as well as the size of the Al pools were determined using image analysis as described in Appendix B. The maximum grains sizes as well as the d(10), d(50) and d(90) of the phases present in these composites were measured. Table 5.1 shows the predicted volume % of the binder and cBN phases for the PcBN-Al composite materials based on the mass balance calculation using cBN-Al reaction 1 (Section 2.4). Table 5.2 shows the composition and grain size of the PcBN-Al composites. The phase composition was very close to the predicted values. A detailed table of the cBN grain size, binder size and mean free paths (MFP) of both the cBN and binder phases is shown in Appendix B. Table 5. 1: Predicted volume % of the binder and cBN phases for the PcBN-Al composite materials. Volume % Of Al Added Predicted Volume % of the binder phase Predicted Volume % of the cBN hard phase 15 20.86 79.14 20 27.98 72.02 25 35.18 64.82 Chapter 5: Results of HPHT Sintering - 100 - Table 5. 2: Composition and grain size of PcBN-Al composite materials. Sample Volume % cBN phase Volume % Binder phase cBN grain Size (Mean d50) (?m) G2 cBN+15%Al 82.1 ? 1.5 17.9 ? 1.5 1.5 ? 0.3 G2 cBN+20%Al 70.2 ? 1.8 29.8 ? 1.8 1.7 ? 0.3 G2 cBN+25%Al 60.0 ? 1.7 40.0 ? 1.7 1.8 ? 0.3 G6 cBN+15%Al 74.3 ? 1.9 25.70 ? 1.9 4.7 ? 1.2 G6 cBN+20%Al 67.9 ? 3.0 32.0 ? 3.0 4.7 ? 1.1 G6 cBN+25%Al 59.3 ? 4.1 40.7 ? 4.1 4.7 ? 0.9 G10 cBN+15%Al 80.7 ? 1.4 19.3 ? 1.4 10.8 ? 3.0 G20 cBN+15%Al 75.1 ? 2.3 24.9 ? 2.3 12.5 ? 3.2 G20 cBN+20%Al 70.5 ? 2.3 29.5 ? 2.3 13.2 ? 2.4 G20 cBN+25%Al 60.2 ? 2.7 39.8 ? 2.7 12.8 ? 2.8 5.6. Density Measurements of PcBN-Al Composite Materials Density measurements were done on the sintered PcBN-Al composites. The Archimedes principle was used as described in Chapter 4.4.4. Table 5.3 shows the results of the density measurements. It can be seen that all the sintered materials were fully densified, with a density of 98.7-99.6% of the theoretical. Chapter 5: Results of HPHT Sintering - 101 - Table 5. 3: The density of the PcBN-Al composite materials. Sample Grain size of cBN Density (?m) (g/cm3) (%) Gr 2 cBN + 15 % Al 1.5 3.360 99.2 Gr 2 cBN + 20 % Al 1.7 3.350 99.6 Gr 2 cBN + 25 % Al 1.8 3.345 99.6 Gr 6 cBN + 15 % Al 4.7 3.392 99.6 Gr 6 cBN + 20 % Al 4.7 3.356 99.5 Gr 6 cBN + 25 % Al 4.7 3.347 99.7 Gr 10 cBN + 15 % Al 10.8 3.388 99.5 Gr 20 cBN + 15 % Al 12.5 3.362 98.7 Gr 20 cBN + 20 % Al 13.2 3.358 99.5 Gr 20 cBN + 25 % Al 12.8 3.308 99.1 Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 102 - Chapter 6: Results of Mechanical Properties of cBN-Al Composites This chapter describes the results of the measurement of mechanical properties of the sintered PcBN-Al composite materials. The mechanical properties results described are hardness, fracture toughness, R-curve behaviour, transverse rupture strength and strength distribution. A discussion of the results will be given in Chapter 7. 6.1. Hardness Details of the sample preparation for the hardness testing are described in Chapter 4.5.1. The hardness measurements were carried out using a Vicker?s hardness indenter with a load of 5 kgf (49 N) and a dwell time of 10 s. The results of the Vicker?s hardness measurements are given in Table 6.1, all data from the hardness measurements in Appendix D. The results from Table 6.1 show that the hardness ranges between 15.6 ? 0.38 ? 40.7 ? 1.2 GPa. G2 cBN + 15%Al had the highest hardness of 40 ? 1.2 GPa, while G20 cBN + 25 % Al had the lowest hardness of 15.6 ? 0.38 GPa. Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 103 - Table 6. 1: The Vickers Hardness measurements results for PcBN-Al composite materials. Sample Grain Size (?m) Hardness HV5 (GPa) Standard deviation G2cBN + 15%Al 1.5 40.7 1.20 G2cBN + 20%Al 1.7 33.5 1.15 G2cBN + 25%Al 1.8 32.0 1.65 G6cBN + 15%Al 4.7 36.0 1.39 G6cBN + 20%Al 4.7 27.2 0.38 G6cBN + 25%Al 4.7 22.0 1.47 G10cBN + 15%Al 10.8 35.8 2.51 G20cBN + 15%Al 12.5 23.2 1.74 G20cBN + 20%Al 13.2 19.8 1.50 G20cBN + 25%Al 12.8 15.6 0.38 6.2. Fracture Toughness The details of the sample preparation and test procedure for the SEVNB fracture toughness measurements are described in Chapter 4.5.2. The results of the SEVNB average fracture toughness measurements for the cBN-Al composite materials and the grain size of the cBN particles and binder content are shown in Table 6.2; Appendix E contains all the SEVNB fracture toughness results. The fracture toughness for the cBN-Al composite materials was in the range of between 6.4 ? 8.0 MPa.m1/2. G10 cBN + 15vol.% Al had the highest fracture toughness of 8.0 ? 0.05 MPa.m1/2, while G2 cBN + 15vol.% Al had the lowest fracture toughness of 6.4 ? 0.4 MPa.m1/2. Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 104 - Table 6. 2: SEVNB fracture toughness measurements results for PcBN-Al composite materials. Sample Grain Size (?m) Fracture Toughness KIC [MPa.m 1/2] Standard deviation G2cBN + 15%Al 1.5 6.4 0.4 G2cBN + 20%Al 1.7 7.0 0.5 G2cBN + 25%Al 1.8 7.1 0.2 G6cBN + 15%Al 4.7 7.6 0.7 G6cBN + 20%Al 4.7 7.6 0.3 G6cBN + 25%Al 4.7 7.0 0.3 G10cBN + 15%Al 10.8 8.0 0.1 G20cBN + 15%Al 12.5 7.9 0.2 G20cBN + 20%Al 13.2 7.8 0.3 G20cBN + 25%Al 12.8 6.9 0.1 6.3. R-Curve Behaviour (Compact Tension Measurements) The details of the sample preparation and testing procedure for the R-curve behaviour measurements are given in Chapter 4.5.2.2. Table 6.3 shows the KIR results of the CT measurements for R-curve behaviour. The results show that KIR ranges between 3.8 ? 9.1 MPa.m1/2. The detailed results of the CT measurements are given in Appendix F. There could be inaccuracies with the resultant data from the CT measurements, due to large pre-cracks in the material, which were initiated in the materials during the laser machining of the notch and holes into the materials. These pre-cracks caused difficulty in initiating the crack used for the R-curve behaviour as described in Chapter 4.5.2.2. The pre-cracks were situated along the notch tip thus the pre-cracks Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 105 - were used as the R-curve crack. The material was also brittle and once force was exerted to lengthen the crack, the crack propagated extremely quickly resulting in only a few data points per experiment. For some of the sample materials no data points could be extracted as the crack propagated too quickly through the material resulting in premature failure. Table 6. 3: The KR results for the R-curve behaviour. Sample KIR (MPa.m 1/2) G2cBN15Al 5.1 ? 8.7 G2cBN20Al 6.5 ? 7.6 G2cBN25Al 5.5 ? 6.5 G6cBN15Al 3.8 ? 7.6 G6cBN20Al 5.8 ? 8.3 G6cBN25Al 5.0 ? 9.1 G10cBN15Al 4.6 ? 5.1 G20cBN15Al 5.8 ? 9.5 G20cBN20Al 4.8 ? 9.0 G20cBN25Al 5.3 ? 8.6 Figures 6.1 ? 6.3 show the R-curve behaviour with increasing crack lengths for the G2, G6 and G10 and the G20 PcBN-Al composite materials respectively. The cracks propagated stably for approximately 2-5 mm before unstable propagation occurred and the samples fractured. Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 106 - 9 10 11 12 13 14 15 16 0 2 4 6 8 10 K IR , M P a. m 1/ 2 Crack length, mm G2+15% Al(1) G2+15% Al(2) G2+20% Al(1) G2+20% Al(2) G2+25% Al(1) G2+25% Al(2) Figure 6. 1: R-Curve behaviour with crack length for G2 PcBN-Al composite materials. 9 10 11 12 13 14 15 16 0 2 4 6 8 10 G6+15% Al G6+20% Al(1) G6+20% Al(2) G6+25% Al G10+15% Al K IR , M P a. m 1/ 2 Crack length, mm Figure 6. 2: R-Curve behaviour with crack length for G6 and 10 PcBN-Al composite materials. Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 107 - 9 10 11 12 13 14 15 16 0 2 4 6 8 10 12 K IR , M P a. m 1/ 2 Crack length, mm G20+15% Al(1) G20+15% Al(2) G20+20% Al(1) G20+20% Al(2) G20+25% Al Figure 6. 3: R-Curve behaviour with crack length for G20 PcBN-Al composite materials. 6.4. Transverse Rupture Strength (TRS) Measurements 6.4.1. Strength Results The details of the sample preparation and testing procedure for the transverse rupture strength measurements are given in Chapter 4.5.3. The samples were loaded to failure at a rate of 0.05 mm/min under a 4-point bend loading. The average flexural strength and standard deviation results for the PcBN-Al composite materials are shown in Table 6.4. These values were in the region of between 346.7 ? 19.7 ? 443.9 ? 22.7 MPa. All the results of the flexural strength tests are given in Appendix G. Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 108 - Table 6. 4: Average flexural strength results for the PcBN-Al composite materials. Sample: Grain Size (?m) Mean Flexural Strength (MPa) Standard Deviation G 2 cBN + 15 % Al 1.5 372 41 G 2 cBN + 20 % Al 1.7 406 61 G 2 cBN + 25 % Al 1.8 426 42 G 6 cBN + 15 % Al 4.7 380 72 G 6 cBN + 20 % Al 4.7 382 30 G 6 cBN + 25 % Al 4.7 418 23 G 10 cBN + 15 % Al 10.8 444 23 G 20 cBN + 15 % Al 12.5 384 25 G 20 cBN + 20 % Al 13.2 347 20 G 20 cBN + 25 % Al 12.8 380 23 6.4.2 Weibull Analysis The results of the Weibull analysis on the bending strength measurements are shown in Table 6.5, which shows the characteristic strength values (?o), the Weibull modulus (m), the 80% confidence intervals and the average flaw size of the composite materials. All the Weibull statistical data results per material can be seen in Appendix H. The average characteristic strength values for the cBN composite materials were in a narrow range between 354 ? 454 GPa. The G10 + 15 vol.% Al material had the highest average characteristic strength value of 454 GPa, while the G20 + 20 vol.% Chapter 6: Results of Mechanical Properties of cBN-Al Composites - 109 - Al material had the lowest average characteristic strength of 355 GPa. The strength values for the cBN-Al composite materials were low compared to other commercially available PcBN materials. The range of the Weibull moduli between 6.1 - 24.9 was quite big. Table 6. 5: Characteristic strength ?o, Weibull modulus m, 80% confidence intervals (C.I) and average flaw size of the PcBN-Al composite materials. Sample Characteristic Strength ?O (MPa) 80% C.I M 80% C.I G2cBN+15%Al 390 377 ? 405 9.8 7.2 ? 11.7 G2cBN+20%Al 433 413 ? 454 7.0 5.3 ? 8.3 G2cBN+25%Al 445 431 ? 459 10.7 8.1 ? 12.7 G6cBN+15%Al 409 389 ? 432 6.1 4.7 ? 7.3 G6cBN+20%Al 395 386 ? 403 15.7 11.8 ? 18.7 G6cBN+25%Al 429 420 ? 438 16.2 12.3 ? 19.3 G10cBN+15%Al 454 448 ? 460 24.0 18.2 ? 28.5 G20cBN+15%Al 396 388 ? 404 16.6 12.6 ? 19.8 G20cBN+20%Al 355 250 ? 340 24.9 18.9 ? 29.6 G20cBN+25%Al 390 383 ? 396 19.9 14.9 ? 23.7 Chapter 7: Discussion of Results - 110 - Chapter 7: Discussion of Results Hardness The results show high hardness values in the expected range with a standard deviation of less than 10%. Expected hardness values can be calculated by using the assumption of the rule of volume mixtures using equation 3.2, section 3.4.1. The results can be viewed in Appendix C. Figure 7.1, 7.2 and 7.3 shows the results of the Vickers hardness measurements of the sintered polycrystalline cBN-Al composites at room temperature plotted against the grain size of the cBN hard phase, the inverse of the square root of cBN grain size and the binder content respectively. This shows the effect the grain size and binder content had on the hardness values. The hardness of the PCBN-Al composite materials produced from Table 6.1 ranges between 15.6 ? 0.38 ? 40.7 ? 1.2 GPa. Compared to other PcBN materials these values are quite reasonable. The hardness of single crystal cBN grains is 43 GPa (7) and Amborite (11) (Element Six (Pty) Ltd commercially produced material which contains 90vol.% of cBN with aluminium nitride and aluminium borides as the binder phases) has a Knoop hardness of 31.5 GPa. Rong and Funkunaga (2002, 1994) (6, 26) measured the Knoop hardness of their cBN-Al composite to be between 15 ? 35 GPa. Comparisons with results obtained by Rong and Funkunaga (6, 26) are limited, as different hardness measurement tests have been done (i.e. Knoop and Vickers). A conversion between the Knoop and Vickers methods can be done, but there will still Chapter 7: Discussion of Results - 111 - be an error associated with it. Hardness results of cBN-Al composites obtained by Li (8) showed 20 - 32.7 GPa for 70 - 95 wt.% cBN. 0 5 10 15 10 20 30 40 50 H ar dn es s, H V (G P a) Grain size, G (?m) 15% Binder 20% Binder 25% Binder Figure 7. 1: Hardness as a function of the grain size of the cBN particles. 0.2 0.4 0.6 0.8 1.0 10 20 30 40 50 H ar d n es s, H (G P a) Inverse square root grain size, G -1/2 (?m -1/2 ) 15 vol.% Binder 20 vol.% Binder 25 vol.% Binder Figure 7. 2: Hardness as a function of the inverse square root of the cBN grain size. Chapter 7: Discussion of Results - 112 - 15 20 25 30 35 40 45 10 20 30 40 50 H ar dn es s, H V (G P a) Binder Content, vol.% G2 cBN G6 cBN G10 cBN G20 cBN Figure 7. 3: Hardness as a function of the binder content. From Figures 7.1 ? 7.3 it can be seen that hardness decreases with increasing cBN grain size and increasing binder content. This is expected since the hardness of the material is dependent mainly on the hardness of the ultra-hard phase. Since the binder, consisting of AlN and AlB2, is softer than cBN there will be weaker phases bridging the cBN grains, thus reducing the hardness of the material. At lower binder content there is more cBN-cBN contact, giving rise to a rigid mechanically locked skeleton which leads to higher hardness values. The hardness also decreases with the increasing cBN grain size as predicted by the Hall-Petch relationship (48) (i.e. dH /1? given in equation 3.3 in Chapter 3.4.1). Except for the lowest binder content where it deviates for the largest grain size, as Chapter 7: Discussion of Results - 113 - seen in Figure 7.2 (shown by circle), the reason for this behaviour is probably that the materials consisting of larger grains have poorer sinterability and some residual porosity. The slope of the lines dH /1? reduces with reducing binder content. The decrease in hardness with increasing grain size suggests that the hardness is dependent on the grain size of the cBN particles. The smaller the cBN grains the larger the grain boundary area, therefore there is less slip between the grains. The decrease in hardness with increasing binder phase per material grade is not quite linear, also the hardness doesn?t decrease linearly with increasing grain size. A statistical study into the effect of both the grain size and binder content has been carried out. Minitab, a statistical software program, was used to determine the correlation between each of the properties and the affected parameters. Appendix J shows the results of the General linear model and the Two-Way ANOVA tests. The results of the Hardness measurements as a function of both the cBN grain size and the binder content are shown in Figure 7.4, while in Figure 7.5, the main effects of the hardness measurements are shown. It is clear that Hardness decreases with increasing binder content and cBN grain size, although there is a maximum observed at G10. It is unclear whether the maximum at G10 is real or is a result of error in the statistics. Chapter 7: Discussion of Results - 114 - The contributions of each effect on the hardness were determined using a general linear model and the Two-way ANOVA model, shown in Appendix J. The results from the Two way ANOVA model shows the grain size had a 67% effect on the hardness and the binder content had a 28%, which shows that the grain size had a larger effect on the hardness than the binder content in the investigated concentration range of the binders. Grain size Al-content % G20G10G06G02 252015252015252015252015 45 40 35 30 25 20 15 Ha rd ne ss (G Pa ) 95% CI for the Mean Hardness Measurement Results for PcBN-Al Composites Figure 7. 4: Hardness as a function of the cBN grain size and the binder content. Chapter 7: Discussion of Results - 115 - G20G10G06G02 36 34 32 30 28 26 24 22 20 252015 Grain size M ea n Ha rd ne ss (G Pa ) Al-content Main Effects Plot for Hardness of PcBN-Al Composites Fitted Means Figure 7. 5: Main effect plot of the Hardness with cBN grain size and Al content. Fracture Toughness The fracture toughness for the cBN-Al composite materials is in the range between 6.4 ? 8.0 MPa.m1/2 (Table 6.2). The measured fracture toughness values are above those reported for commercially available PcBN materials of 3.5 ? 6.6 MPa.m1/2 (11), and are also mostly higher than the values reported for Amborite, with a fracture toughness of 6.4 MPa.m1/2 (11). It is difficult to directly compare the data of the PcBN-Al composite and the data from literature as they were determined by different testing methods (i.e. SEVNB and indentation methods). The results obtained for the PcBN-Al composite materials exhibit a low standard deviation, less than 10 %. Figure 7.6, 7.7 and 7.8 show the plots of the SEVNB fracture toughness results as a Chapter 7: Discussion of Results - 116 - function of grain size of cBN particles, the inverse of the square root of cBN grain size and binder content respectively. 0 5 10 15 4 6 8 10 Fra cture T ou ghn es s, K IC (M P a. m 0. 5 ) Grain size, G (?m) 15% Binder 20% Binder 25% Binder Figure 7. 6: Fracture toughness of the PcBN-Al composite materials as a function of grain size of cBN, using SEVNB method. 0.2 0.4 0.6 0.8 1.0 4 6 8 10 Fra cture T ou ghn es s, K IC (M P a. m 0. 5 ) Inverse square root grain size, G -1/2 (?m -1/2 ) 15 vol.% Binder 20 vol.% Binder 25 vol.% Binder Figure 7. 7: Fracture toughness of PcBN-Al composite materials as a function of the inverse square root of the cBN grain size. Chapter 7: Discussion of Results - 117 - 15 20 25 30 35 40 45 4 6 8 10 Fr ac tu re T ou gh ne ss , K IC (M P a. m 0. 5 ) Binder Content, vol.% G2 cBN G6 cBN G10 cBN G20 cBN Figure 7. 8: Fracture toughness of the PcBN-Al composite materials as a function of binder %, using SEVNB method. The fracture toughness for all composite materials studied in this work generally increases with increasing cBN grain size, except for the composites with 25vol.% binder where the KIC is constant within the errors of the determinations (Figure 7.6 and 7.8). The increase in fracture toughness with increasing grain size is consistent with work done on Al2O3 materials (70, 71, 72). There is no significant grain size dependence on the fracture toughness, although larger grains act as crack deflection sites resulting in a higher fracture toughness. Considering the binder content, the fracture toughness of the batches with 6 and 20 ?m cBN grain size does not show a significant dependence on binder content at 15 ? 20vol.%, but decreases at 25vol.%. However, at low binder level, the batches with 2 Chapter 7: Discussion of Results - 118 - ?m grain size show a slight increase of fracture toughness with increasing binder content. The toughness generally decreases with increasing binder content, except for the G2 cBN composites where the opposite trend can be seen, although these values are within the experimental error, suggesting constant values or a slight increase. This suggests that at fine grain size the material is very brittle due to a higher cBN- cBN contact. Increasing the binder content provides the material with more weaker phases is more prone to plastic deformation. The distance between the particles is influenced by the amount of binder content; for smaller particles the distance between the grains is smaller than for larger particles. If the distance between the particles is too high (i.e. high binder content) a negative effect results due to the weakness of the binder phases, as seen from the decrease in fracture toughness for the higher binder contents. At larger grain sizes there is a high amount of crack deflection, by increasing the binder content to 25vol.%, a decrease in the fracture toughness occurs. As the distance between two grains increases the less stress is required for crack propagation therefore cracking is easier long the binder phase with fewer grains providing crack deflection. Analysis of the crack path and fracture surfaces of the PCBN-Al composites was done. Figure 7.9 and 7.10 show crack propagation paths through a G20 cBN + 25vol.% Al composite material and fracture surfaces of fracture toughness samples of G2cBN + 15vol.% Al and of G20 cBN + 25vol.% Al respectively. The fracture mode Chapter 7: Discussion of Results - 119 - is transgranular, both in the cBN phase and in the binder. From the crack path in Figure 7.9, a tendency of predominant crack propagation in the binder phase might be deduced; this is valid at least for that material composition and microstructure. The crack propagates mostly through the binder, and along the grain boundaries between the different binder phases. The high fracture toughness values obtained was substantiated by crack deflection/bridging observed during crack propagation, Figure 7.9. Crack deflection/bridging is generally associated with R-curve behaviour, i.e. increase of fracture resistance with increasing crack length. The fracture toughness values determined in this study were calculated from maximum stress and notch depth. Crack deflection can lead to a small increase in the crack tip toughness, such as was determined in the studies of whisker-reinforced alumina by R?del (1991) (73), but mainly aids in setting up efficient crack bridges leading to R-curve behaviour (74). Unreacted aluminium may lead to further ductile bridging, but this effect will be minimal due to the low amount of residual Al (up to 2.7 wt.%). Samples with residual Al (G 6 cBN + 25vol.% Al, G20 cBN + 20 vol.%Al and G20 cBN + 25vol.% Al) show a lower toughness because the amount of binder content between the grains for these materials has exceeded the limit (maximum 20vol.% shown in Figure 7.12) which leads to easier crack propagation between grains. G20 cBN + 25vol.%Al has the lowest fracture toughness and the highest amount of residual Al, showing that the effect of aluminium on the fracture toughness is not a major one. Chapter 7: Discussion of Results - 120 - At low binder content the grain size has a larger effect on the toughness, but at higher binder content it appears that there is no sensitivity to grain size and the binder starts controlling the toughness. Crack propagation is generally affected by the residual stress state in the material. In multi-phase materials with phases of different thermal expansion residual stresses always occur (49). Internal stresses exist in all polycrystalline materials that include a secondary phase of different thermal expansion coefficient. The height of residual stresses and their local distribution is determined by the volume ratio of the two phases and their thermal and elastic constant. At cBN grain sizes of 6 ? 20 ?m and a binder content of 15 ? 20vol.% the residual stresses are obviously in an optimum range with regard to fracture toughness. At low binder content there is more cBN-cBN contact and the material is more brittle. Increasing the amount of binder will provide more opportunities for crack deflection and prevent crack propagation, thus increasing the fracture toughness. Further consideration on fracture toughness, based on real and estimated flaw sizes will be made after discussions of strength and strength limiting flaws. Chapter 7: Discussion of Results - 121 - Figure 7. 9: Crack propagation in fracture toughness tests showing the fracture path on polished G20 cBN + 25 vol.% Al sample exhibiting slightly dominant crack propagation in the binder phase and transgranular fracture through the cBN phase. Figure 7. 10: Crack propagation in fracture toughness tests. Fracture surface of a) G2 cBN + 15 vol.% Al and b) G20 cBN + 25 vol.% Al, exhibiting mostly transgranular crack propagation in the cBN and binder phase. The results of the fracture toughness measurements as a function of both the cBN grain size and the binder content are shown in Figure 7.11, while in Figure 7.12, the main effects of the hardness measurements are shown. It is clear that fracture toughness increases with increasing cBN grain size until 10 ?m and then decreases, which suggests that 10 ?m is the optimum grain size for toughening. The fracture toughness also increases with increasing binder content up until 20vol.% then 50 ?m 2 ?m a) 10 ?m b) Chapter 7: Discussion of Results - 122 - decrease below the 15vol.% level for the 25vol.% material, which suggests that 20vol.% binder is the maximum binder content; anything higher will result in a decrease in the fracture toughness. The contributions of each effect on the fracture toughness were determined using a general linear model and the two-way ANOVA model, shown in Appendix J. The results from the two-way ANOVA model show that the grain size had a 27% effect on the fracture toughness and the binder content had 13%, which means that there was no clear dependence on the fracture toughness with both cBN grain size and binder content, although it does show that the grain size had a larger effect than the binder content. From the statistical results it can be deduced that the optimal composition for the fracture toughness would be G10 cBN and 20vol.% Al, although this theory could not be proved due to experimental time constraints. Chapter 7: Discussion of Results - 123 - Grain size Al-content % G20G10G06G02 252015252015252015252015 9 8 7 6 5 Fr ac tu re T ou gh ne ss , K 1c , ( M Pa .m ^ 0. 5) 95% CI for the Mean Fracture Toughness, K1c Results for PcBN-Al Composites Figure 7. 11: Fracture toughness as a function of the cBN grain size and the binder content. G20G10G06G02 8.2 8.0 7.8 7.6 7.4 7.2 7.0 252015 Grain size M ea n Fr ac tu re to ug hn es s, K Ic (M Pa .m ^ 0. 5) Al-content Main Effects Plot for Fracture toughness, K1c of PcBN-Al Composites Fitted Means Figure 7. 12: Main effects plot of the fracture toughness with cBN grain size and Al content. Chapter 7: Discussion of Results - 124 - R-curve Behaviour The results of the R-curve behaviour tests done using the Compact tension (CT) method described in Chapter 4.5.2.2 are shown in Chapter 6.3. Figure 7.13 and 7.14 show the R-curves of the PcBN-Al composite materials. The R-curve starts approximately at KIO, which is the crack tip toughness (i.e. the onset of crack growth) (69) and extends with increasing crack length. Due to measurement errors it was difficult to determine the correct value of KIO for each material. The KIO was taken as the point in the data where constant crack growth resulted. Sample data for the R-curve behaviour had a very low reproducibility; variability in the data could have been due to measurement errors as explained in Chapter 6.3. Figure 7.13 shows the R-curves for the G2 and G6 PcBN-Al materials where the KIR rises to between 3.8 and 9.1 MPa.m1/2, these values being higher than the KIc for these materials. There was a large difference in the KIO for the different materials of the G2 family of materials, the difference being attributed to errors in measurement. The KIR for most of the samples was not the same for each material. The KIR for the G2+15vol.%Al was between 5.1 ? 8.7 MPa.m1/2, the KIR for the G2+20vol.%Al between 6.5 ? 7.6 MPa.m1/2 and the KIR for the G2+25vol.%Al is between 5.5 ? 6.5 MPa.m1/2, shows that the material had no toughening before failure as it was below the KIC values. The KIR for the G6+15vol.%Al was between 3.8 - 7.6 MPa.m 1/2. It can also be seen that the second sample showed no rise in toughness. The KIR for the Chapter 7: Discussion of Results - 125 - G6+20vol.%Al was between 5.8 ? 8.3 MPa.m1/2 and the KIR for the G6+25vol.%Al 5.0 ? 9.1 MPa.m1/2. 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 G2+15vol.% Al(1) G2+15vol.% Al(2) K IR , M P a .m 1 /2 ?a, cm G2+20vol.% Al (1) G2+20vol.% Al (2) K IR , M P a .m 1 /2 ?a, cm G2+25vol.% Al(1) G2+25vol.% Al(2) K IR , M P a .m 1 /2 ?a, cm G6+15vol.% Al(1) G6+15vol.% Al(2) K IR , M P a .m 1 /2 ?a, cm G6+20vol.% Al (1) G6+20vol.% Al (2) K IR , M P a .m 1 /2 ?a, cm K IR , M P a .m 1 /2 ?a, cm G6+25vol.% Al (1) Figure 7. 13: R-curves of the G2 and G6 PcBN-Al composite materials. Chapter 7: Discussion of Results - 126 - Figure 7.14 shows the R-curves for the G10 and G20 PcBN-Al materials, the KIR rises to between 5.3 and 9.5 MPa.m1/2, these values are higher than the KIc for these materials. The KIR for G10+15vol.%Al was 4.6 ? 5.1 MPa.m 1/2, which shows that there was no toughening. While the KIR for the G20+15vol.%Al was between 5.8 - 9.5 MPa.m1/2, the KIR for the G20+20vol.%Al between 4.8 - 9.0 MPa.m 1/2 and the KIR for the G20+25vol.%Al 5.3 ? 8.6 MPa.m1/2. 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 K IR , M P a .m 1 /2 ?a, cm G10+15vol.% Al (1) G20+15vol.% Al (1) G20+15vol.% Al (2) K IR , M P a .m 1 /2 ?a, cm G20+20vol.% Al (1) G20+20vol.% Al (2) K IR , M P a .m 1 /2 ?a, cm K IR , M P a .m 1 /2 ?a, cm G20+25vol.% Al (1) Figure 7. 14: R-curves of the G10 and G20 PcBN-Al composite materials. Chapter 7: Discussion of Results - 127 - Table 7.1 shows the comparisons between the KIC and the KIR values. It can be seen that the compared results show similar toughness values. There are a few exceptions such as G10 cBN + 15vol. % Al which has a lower KIR value than the KIC value or G6 cBN + 25vol.% Al has a lower KIC value than KIR. These discrepancies would have been a result of the errors explained in the beginning for the CT measurements. Although it does show that samples with a small amount of unreacted Al left in the material (e.g. G6 cBN + 25 vol.% Al, G20 cBN + 20 vol.% Al and G20 cBN + 25 vol.% Al) had a higher K IR than KIC values, this could suggest that unreacted Al will increase the R-curve behaviour of the material, but only for long crack lengths. Table 7. 1: Comparison of KIC and KIR values for the PcBN-Al composite materials. Sample KIC (MPa.m 1/2) KIR (MPa.m 1/2) G2cBN15Al 6.4 ? 0.4 5.1 ? 8.7 G2cBN20Al 7.1 ? 0.5 6.5 ? 7.6 G2cBN25Al 7.1 ? 0.2 5.5 ? 6.5 G6cBN15Al 7.6 ? 0.7 3.8 ? 7.6 G6cBN20Al 7.6 ? 0.3 5.8 ? 8.3 G6cBN25Al 7.0 ? 0.3 5.0 ? 9.1 G10cBN15Al 8.0 ? 0.1 4.6 ? 5.1 G20cBN15Al 7.9 ? 0.2 5.8 ? 9.5 G20cBN20Al 7.8 ? 0.3 4.8 ? 9.0 G20cBN25Al 6.9 ? 0.1 5.3 ? 8.6 The R-curve behaviour of the PCBN-Al composites was compared with some common ceramic materials (Al2O3, Si3N4, SiC and ZrO2). Table 7.2 shows the R- Chapter 7: Discussion of Results - 128 - curve results of these ceramic materials. Al2O3, Si3N4 and SiC have R ? curves with KIR between 2 ? 9 MPa.m 1/2, while ZrO2 doped materials have a KIR between 5 ? 16 MPa.m1/2. The increasing fracture toughness of Al2O3 was caused by crack tip shielding by microcracking and crack bridging and grain bridges (70), while Si3N4 exhibited crack bridging effects. The increase in fracture toughness of ZrO2 ceramics was due to transformation toughening brought about by stresses around the crack tip due to the transformation from the tetragonal to the monoclinic form. Table 7. 2: R-curve results of some common ceramics materials (41, 70). Sample: KIR (MPa.m 1/2) KIC (MPa.m 1/2) Al2O3 (depends on grain size) 2.3 ? 7 3 ? 4 Si3N4 (elongated grains) 5 ? 11 4 ? 7 SiC (elongated grains) 2 ? 8 2 ? 3 ZrO2 (Mg, Ce doped) 5 ? 16 7 ? 10 There was a small amount of toughening over a small section of the measured samples, 0.2 ? 0.8 cm extension, as can be seen in the R-curves. The slight rise in R- curve behaviour could be attributed to crack deflection, observed during crack propagation in the material. Crack deflection can also cause partial bridging by grains and secondary phase and is a prerequisite in the setting up of crack bridging effects (68). Crack bridging and frictional grain bridging toughening mechanisms can also play a role in increasing the toughening mechanisms. Crack deflection/bridging are Chapter 7: Discussion of Results - 129 - generally associated with R-curve behaviour, i.e. increase of fracture toughness with increasing crack length. The fracture toughness from the SEVNB and R-curve behaviour tests both show a slight toughening caused by crack deflection which as stated in the SEVNB fracture toughness section can lead to a small amount of crack tip toughness and thus assist in setting up crack bridging leading to R-curve behaviour. Unreacted aluminium can further lead to ductile bridging, thus increasing toughness. In the samples which contained the unreacted Al, the amount of binder was too high causing a deterioration of the properties. Frictional grain bridges also result in the toughening of the materials. Transverse Rupture Strength Chapter 6.4 shows the results of the transverse rupture strength tests. Table 6.5 gives the characteristic strength, Weibull modulus and 80% confidence intervals for the PcBN-Al composite materials. The strength ranges from 355 ? 454 MPa. Strength values for commercially available PcBN cutting tools are in the region of about 600 - 800 MPa (11). The strength values for the PcBN-Al composites are lower than those found in literature and for similar tool materials as shown in Table 7.3. The range of the Weibull moduli between 6.1 - 24.9 is quite large. However, considering the strength distributions of the batches with low Weibull moduli more Chapter 7: Discussion of Results - 130 - closely, the strength data of these batches more or less clearly show signs of bimodal distributions (two different m slopes), whereas the strength data of the batches with high m values belong to monomodal distributions, see Figure 7.15. Figure 7.15a shows the two different m slopes for the G2 cBN +15vol.%Al and G6 cBN+15vol.%Al bimodal distribution. G6 cBN + 15vol.%Al shows a more pronounced bimodal distribution, exhibiting different fracture origins for the various m slopes. The differences in the fracture origins can be seen in the Fractography section: Figure 7.28 shows the fracture origins for the low and high strength distributions for the G2 cBN + 15vol.%Al samples; Figure 7.27a shows the fracture origin of a G6 cBN + 15vol.%Al sample with a high strength (binder phase) and Figure 7.29a and 7.30a shows the fracture origins of samples with a low strength (large pores) for the G6 cBN +15 vol.%Al samples. Therefore the statistical parameters ?O and m of the batches with low m values need to be considered with caution. Table 7. 3: Transverse rupture strength for commercially available cutting tool materials (76, 77). Material: Transverse Rupture Strength (MPa) (3 point bend test) AMBORITE (AMB90) 664 AMBORITE (DBC50) 570 Polycrystalline diamond tools (PCD) 750 ? 1000 WC-Co tools 1000 ? 2500 Figure 7.16 ? 7.18 shows the Weibull plots for the 15, 20 and 25 vol.% binder materials respectively. These results show the probability for each material to fail. Chapter 7: Discussion of Results - 131 - Refer to Appendix H for the full range of characteristic strength and Weibull moduli for various confidence intervals. Figure 7.16 shows that the G 10 cBN + 15vol.% Al material had the highest strength values and the least amount of scatter, while G 6 cBN + 15vol.% Al had the highest amount of scatter which could have been due to its bimodal distribution. This suggests that the strength results were low and the material could have contained a high amount of flaws in this material. Both G 10 and G 20 cBN + 15vol.% Al materials had very little scatter and corresponding graphs were relatively straight. 200 300 400 500 600700 -4 -3 -2 -1 0 1 2 200 300 400 500 600700 -4 -3 -2 -1 0 1 2 G2cBN15 G6cBN15 ln ln (1/ (1- P) ) Strength (MPa) G6cBN20 G20cBN20 ln ln (1/ (1- P) ) Strength (MPa) Figure 7. 15: Weibull distributions of bending strength data (a) showing more or less signs of bimodal distributions, batches G2 cBN+15vol.%Al, G6cBN+15vol.%Al (b) showing monomodal distributions of the batches G6cBN+20vol.%Al and G20cBN+15vol.%Al with Weibull moduli of 15 and 25. (b) (a) Low strength region ? large fracture origin High strength region ? small fracture origin Chapter 7: Discussion of Results - 132 - 100 1000 -5 -4 -3 -2 -1 0 1 2 3 ln ln (( 1- P )-1 ) Strength (MPa) G2cBN+15% binder G6cBN+15% binder G10cBN+15% binder G20cBN+15% binder Figure 7. 16: Weibull plots of the 15 vol.% binder PcBN-Al composite materials. Figure 7.17 shows that G 20 cBN + 20vol.% Al had the lowest strength values and lowest amount of scatter. G 2 cBN + 20vol.% Al materials had a high amount of scatter and the highest strength values also showing a bimodal distribution, while G 6 cBN + 20vol.% Al had a moderate amount of scatter. Figure 7.18 shows that there was a very low amount of scatter for the 3 materials made with 25% Al. G 20 cBN has the lowest strength values, G 2 cBN + 25vol.% Al had the highest strength values and G 6 cBN + 25vol.% Al had the least amount of scatter. Chapter 7: Discussion of Results - 133 - 100 1000 -5 -4 -3 -2 -1 0 1 2 3 ln ln (( 1- P )-1 ) Strength (MPa) G2cBN+20% binder G6cBN+20% binder G20cBN+20% binder Figure 7. 17: Weibull plots of the 20 vol.% binder PcBN-Al composite materials. 100 1000 -5 -4 -3 -2 -1 0 1 2 3 ln ln (( 1- P )-1 ) Strength (MPa) G2cBN+25% binder G6cBN+25% binder G20cBN+25% binder Figure 7. 18: Weibull plots of the 25vol.% binder PcBN-Al composite materials. Chapter 7: Discussion of Results - 134 - Figure 7.19 and 7.20 show the variation of the characteristic strength values of the PcBN composite materials as a function of the grain size of the cBN phase and binder content with 80% confidence intervals. The strength generally decreased with increasing grain size. It was observed that both the 20 and 25vol.% Al composites had characteristic strength values which decreased with increasing grain size. This decrease with grain size is a common feature for most ceramics (74) and is related to the size of the initial flaw which is related to the grain size (75). With the 15vol.% Al there was no clear trend. Also there were no clear trends of strength dependence on binder content. The increase in binder content increased the strength of the G2 cBN, while for G6 and G20 cBN the strength decreased to a minimum at 20 vol.% and then increased. 0 5 10 15 200 300 400 500 600 S tr en g th , ? O (M P a) Grain size, G (?m) 15% binder 20% binder 25% binder Figure 7. 19: Characteristic strength plot of the PcBN-Al composite materials with cBN grain size. Error bars show the 80% confidence intervals. Chapter 7: Discussion of Results - 135 - 15 20 25 30 35 40 45 200 300 400 500 600 S tr en g th , ? O (M P a) Binder Content, vol.% G2 cBN G6 cBN G10 cBN G20 cBN Figure 7. 20: Characteristic strength plot of the PcBN-Al composite materials with binder %. Error bars show the 80% confidence intervals. The results of the strength measurements as a function of both the cBN grain size and the binder content are shown in Figure 7.21, while in Figure 7.22, the main effects of the hardness measurements are shown. Figures 7.21 and 7.22 reveal the strength increased with increasing binder content, except for the G20, while there is no clear trend with grain size. The contributions of each effect on the strength were determined using a general linear model and the Two-way ANOVA model, shown in Appendix J. The results of the two-way ANOVA model showed that the grain size had a 9% effect on the strength and the binder content had a 10%, which means that there was only a weak dependence of the strength on either cBN grain size or binder content. Chapter 7: Discussion of Results - 136 - Grain Size Al-content % G20G10G06G02 252015252015252015252015 450 425 400 375 350 Fl ex ur al S tr en gt h (M Pa ) 95% CI for the Mean Flexural Strength Results of PcBN-Al Composites Figure 7. 21: Flexural Strength as a function of the cBN grain size and the binder content. G20G10G06G02 460 450 440 430 420 410 400 390 380 370 252015 Grain Size_1 M ea n St re ng th (M Pa ) Al-content Main Effects Plot for Flexural Strength of PcBN-Al Composites Fitted Means Figure 7. 22: Main effects plot of the flexural strength with cBN grain Size and Al content. Chapter 7: Discussion of Results - 137 - Comparisons of the hardness, fracture toughness and strength results with each other were also done. This was done in order to determine whether there were any correlations between each of the properties. Figure 7.23 ? 7.25 show the correlations between the tested mechanical properties with grain size and binder content. Figure 7.23 shows the correlation between fracture toughness and hardness. For the G20 samples the fracture toughness increased with increasing hardness, while for the G2 samples the fracture toughness decreased with increasing hardness. There was no real correlation between fracture toughness and hardness dependence on binder content, although there was a slight increase in the fracture toughness with increasing hardness for 25vol.%, while there was a decrease in fracture toughness with increasing hardness for 20vol.% binder. This suggests that the materials with the highest hardness have the lowest KIC. 403530252015 8.25 8.00 7.75 7.50 7.25 7.00 6.75 6.50 Hardness (GPa) F ra ct u re T o u g h n e ss , K 1 c, ( M P a .m ^ 0 .5 ) 2 6 12 17 Size Grain cBN Fracture Toughness vs Hardness of PcBN-Al Composites 403530252015 8.25 8.00 7.75 7.50 7.25 7.00 6.75 6.50 Hardness (GPa) F ra ct u re T o u g h n e ss , K 1 c, ( M P a .m ^ 0 .5 ) 15 20 25 Al Content Fracture Toughness vs Hardness f r PcBN-Al Composites Figure 7. 23: Comparisons between the fracture toughness and hardness, and their dependence on grain size and binder content. Chapter 7: Discussion of Results - 138 - Figure 7.24 shows that there was no correlation between strength and hardness dependence on grain size, except at G2, where the strength decreased with increasing hardness. Strength increased with increasing hardness for both the 20 and 25vol.% binder content. Figure 7.25 shows that there was no trend in the strength and fracture toughness comparisons with both dependence on grain size and binder content, although for G2 strength increased with increasing fracture toughness. The strength for 20 vol.% binder decreased with increasing fracture toughness, while for 25vol.% binder the strength increased with increasing fracture toughness. 403530252015 450 425 400 375 350 Hardness (GPa) S tr e n g th ( M P a ) 2 6 12 17 Size Grain cBN Flexural Strength vs Hardness of PcBN-Al Composites 403530252015 450 425 400 375 350 Hardness (GPa) S tr e n g th ( M P a ) 15 20 25 Al Content Flexural Strength vs Hardness f r PcBN-Al Composites Figure 7. 24: Comparisons between strength and hardness, depending on grain size and binder content. Chapter 7: Discussion of Results - 139 - 8.258.007.757.507.257.006.756.50 450 425 400 375 350 Fracture Toughness, K1c, (MPa.m^0.5) S tr e n g th ( M P a ) 2 6 12 17 Size Grain cBN Strength vs Fracture Toughness for PcBN-Al Composites 8.258.007.757.507.257.006.756.50 450 425 400 375 350 Fracture Toughness, K1c, (MPa.m^0.5) S tr e n g th ( M P a ) 15 20 25 Al Content Strength vs Fracture Toughness f r PcBN-Al Composites Figure 7. 25: Comparisons between strength and fracture toughness, depending on grain size and binder content. Fractography It has been well established that strength depends on the size, shape and surface finish of the specimens i.e. the flaw distribution in the material. It is thus important to determine the size, shape and type of the flaws present in the material. This will help to determine the failure behaviour of the material and to determine strategies for its improvement. The flaw size of the various materials can be estimated using the Griffiths equation (equation 3.1 in Chapter 3) relating the strength and fracture toughness of the material to the flaw size of failure. Based on the fracture toughness, KIc, the size of a crack initiating flaw can be calculated for a definite strength value f? according to (32) Chapter 7: Discussion of Results - 140 - 2 ? ? ? ? ? ? ? ? ? ? Y K a f Ic ? , 2/??Y , a = radius of flaw This can be used to determine whether the results obtained from the strength and fracture toughness measurements are in the correct range by comparing them with the results of the actual flaw size measured by microscopy. The size of the flaws in each material will be determined using optical microscopes or scanning electron microscopes to determine the accuracy of the estimated results. Using equation 3.1 and the strength and fracture toughness data, an estimation of the flaw size can be made (refer to Appendix I). The estimated average flaw size for each material calculated from both the average strength values and the measured flaw size is given in Table 7.4. The results show the flaw sizes for the PcBN-Al composite materials in the range between 300 - 650 ?m. Crack initiating flaws were representative of 10 - 15 % of the tested specimens of each batch, from specimens of low, medium or high strengths. The reason for this was that only 10-15% of the samples were fractographically examined. In most cases the crack initiating flaws were clearly identified. As far as they were identified the flaws were quite big, of 100 ? 500 ?m maximum diameter (Figure 7.26). These large flaws were responsible for the low strength measured. Based on the Griffith?s equation, 3.1, the comparison between the size of the calculated cracks and crack Chapter 7: Discussion of Results - 141 - initiating flaws shows a close agreement, Figure 7.27. This clearly affirms the fracture toughness values tested by the SEVNB tests. The initiating flaws were elongated in an aspect ratio from 1.5 to 5, as shown in the examples in Figure 7.26, the largest diameter always oriented perpendicularly to the specimen length. They either exposed features of ductile rupture or contained very large grains of the binder phase (either AlN or AlB2), sometimes showing features of powder granules not fully sintered together. These characteristics were observed irrespective of the location of the flaws, in the bulk or at the tensile surface of the specimens. They clearly pointed to deficiencies in the powder processing. The flaws exhibiting ductile fracture obviously originated from big particles or agglomerates of unreacted Al binder as proven by EDX analysis. Table 7. 4: The estimated average flaw size, characteristic strength and SEVNB fracture toughness for the PcBN-Al composite materials. Sample Estimate Flaw Size (?m) Measured Flaw Size (?m) G2 cBN + 15% Al 382 100 ? 400 G2 cBN + 20% Al 386 100 ? 420 G2 cBN + 25% Al 356 85 ? 280 G6 cBN + 15% Al 505 200 ? 490 G6 cBN + 20% Al 506 200 ? 500 G6 cBN + 25% Al 361 140 ? 400 G10 cBN + 15% Al 417 200 ? 420 G20 cBN + 15% Al 543 150 ? 300 G20 cBN + 20% Al 646 200 ? 370 G20 cBN + 25% Al 427 100 ? 250 Chapter 7: Discussion of Results - 142 - Figure 7. 26: Typical flaws (a) with features of ductile fracture, sample from batch G2cBN+15vol.%Al, (b) with very large grains of binder phase, sample from batch G6cBN+20vol.%Al. The microstructure and size of the flaws indicate poor deagglomeration and homogenization of cBN and the binder powder during powder processing. The high aspect ratio and parallel orientated of the flaws point out that they originate before the hot pressing step and underscored the attribution to powder processing. The poor correlation of strength and cBN grain size and binder content was mainly due to the very broad scatter of flaw sizes and poor correlation to the processing of the individual material batches. A more complete characterisation of the size and nature of the crack initiating flaws might elucidate the flaw size distribution of the individual materials and thus allow a better correlation of strength and material microstructure. However, the effort of such investigations is probably not appropriate at the current state of processing. 50 ?m 100 ?m (a) (b) (b) 100 ?m Surface defects Chapter 7: Discussion of Results - 143 - Figure 7. 27: Model crack size calculated from fracture toughness and strength limiting flaws; (a) G6cBN+15vol.%Al, and (b) G10cBN+15vol.%Al. As mentioned in the discussion G2 cBN +15vol.%Al and G6 cBN +15 vol.%Al had bimodal strength distribution (see Figure 7.15). Figure 7.28 shows the flaw origins of the low and high strength distribution regions for the G2 cBN +15vol.%Al sample. Figure 7.27a, 7.30b and 7.31a show G6 cBN + 15vol.%Al samples with high strength. The results show flaw size and type differences between the samples in the low and high strength distribution regions within the samples materials showing bimodality. At low strength regions there were very large flaws mainly characterized by large surface defects (pores or large binder pools). While in the high strength distribution regions the flaws are smaller and were characterized by defects near the surface (binder pools or pores). The bimodality of these materials indicated that better processing technology was needed. Figure 7.26a and b, Figure 7.28a and Figure 7.29 show flaws from materials with low strength, flaws that are mainly large surface defects containing large cBN grains or 100 ?m 100 ?m (a) (b) Chapter 7: Discussion of Results - 144 - binder pools, while Figure 7.27b and 7.30 show flaws from materials with high strength. The flaws consist mainly of smaller binder pools and pores. Figure 7. 28: Flaws in G2 cBN +15vol.%Al composite: a) large surface flaw resulting in low strength and b) small binder flaw resulting in high strength. Figure 7. 29: Flaws from materials with low strength, a) G20 cBN+25vol.%Al showing large surface defect with large cBN grain and b) G20+20vol.%Al showing binder surface defect. Figure 7.31 shows SEM images of internal cracking in the PcBN-Al composite materials indicating high internal stresses in the material. Figure 7.26 and 7.33 show Surface defect (a) (b) (a) (b) Chapter 7: Discussion of Results - 145 - images of the failure originating flaws of some of the PCBN-Al composites. The flaw sizes in the materials were quite large (100-500 ?m), confirming the lower strength values. Various flaw types were evident in the composite materials; large surface flaws can be seen in Figure 7.26 and 7.28a; flaws containing large cBN grains can be seen in Figure 7.32a. Figure 7. 30: Flaws from materials with high strength, a) G2 cBN+25vol.%Al showing a smaller binder pool and b) G6cBN+25vol.%Al showing a small pore. Figure 7. 31: SEM images of internal cracking in the PcBN-Al composite materials; a) G2 cBN + 20vol.% Al and b) G2 cBN + 15vol.% Al. (a) (b) (a) (b) Chapter 7: Discussion of Results - 146 - Figures 7.32b and 7.33a and b show flaws containing large pores surrounded by binder phase. Figure 7.32c shows a large binder pool consisting of a brittle oxide. EDS shows an oxidized layer on Al in binder phase, (Figures 7.32 c and d,) which could have come from oxidation of the aluminium powder before sintering. Flaws consisting of larger binder pools are evident in Figure 7.33 d and e. These binder pools also consisted of smaller pores. The appearance of ductile dimple ruptures can be seen in Figures 7.30 a and 7.32 a, b and Figure 7.33 signifying plastic deformation around the edges due to small amounts of free Aluminium within the binder phase found through EDS analysis. Chapter 7: Discussion of Results - 147 - Figure 7. 32: SEM images of various types and sizes of flaws in the PcBN-Al composite materials; a) large grain; b) elongation of binder with pore; c) larger binder pool with oxidized brittle layer and d) EDS showing O and Al. (b) (b) (a) (d) (c) (c) Large cBN grain Pore with binder Large binder pool (a) Chapter 7: Discussion of Results - 148 - Figure 7. 33: SEM images of various types and sizes of flaws in the PcBN-Al composite materials; a and b) binder pool with pore; c) EDS showing ductile Al; d and e) larger brittle binder pool; and f) EDS showing binder is AlN. (b) (e) (d) (a) (c) (f) Pore Large binder pool Large binder pool Chapter 8: Conclusion - 149 - Chapter 8: Conclusion A study into the mechanical properties of PcBN-Al materials was conducted in this project. Materials of high content cBN with an Aluminium binder phase were sintered using High Pressure High Temperature methods. The grain size and binder content were varied to determine the effects each parameter had on the mechanical properties. These materials were fully densified and contained residual binder consisting of AlN and AlB2. Some unreacted aluminium was also found in some of the samples with large amount of additives. The PcBN-Al composite materials made and tested in this study reveal the following in respect of the structure to properties relationships: The hardness of the PcBN materials prepared with Al binder is high as expected due to the hard cBN phase. It decreases with increasing cBN grain size and increasing binder content. From the statistical analysis of the hardness data, the grains size of the cBN particles has a larger effect on the hardness than the binder phase. For an optimum material in reference to hardness, a small cBN grain size and low binder phase content should be used. Fracture toughness is high compared with other PcBN materials (2,11). Fracture toughness also depends on grain size and binder content, but to a much lesser extent Chapter 8: Conclusion - 150 - than hardness and with less clear tendencies. Contrary to the hardness, high values of the fracture toughness are associated with larger grain sizes and medium binder content (20 vol.%). The cracks propagate mostly through the binder; there are signs of crack deflection and crack bridging, as well as grain bridging. Crack deflection may be associated with residual stresses in the constituent phases. The high fracture toughness values are in good correlation with the observed sizes of fracture origins and the corresponding strength values. R-curve behaviour was found, but the data scatters strongly. Crack deflection, grain bridging and crack bridging caused toughening in the materials. Unreacted aluminium in samples with high binder content (up to 2.7 wt.%) can act as a ductile bridging phase, but despite this possible additional toughening mechanism these samples showed low fracture toughness. Strength values are low compared to literature values for commercially available PcBN cutting tools. They show no clear dependence on grain size and binder content. This is due to very large strength limiting flaws (100-500?m) in the materials. These large flaws prevent us from accurately determining the effect of the grain size and binder content on the strength. Chapter 8: Conclusion - 151 - Fractography showed that the materials consisted of various types of flaws, the flaws varying between large binder pools to small pores surrounded by large binder segregations. Some flaws exposed features of ductile rupture (presumably due to unreacted aluminium in the materials) or contained large grains some of which showed features of powder granules not fully sintered together. Surface defects were also found. The largest flaws were the large binder pools that were formed due to the agglomeration of powder particles that had not been properly distributed in the system. These flaws originated from poor powder processing, i.e. poor deagglomeration and homogenization. Better powder processing techniques need to be used to prevent agglomeration and to improve homogenization and to reduce the size of the binder pools which are the largest flaws leading to failure. The technique of attrition milling can be used to improve homogenization and deagglomeration. To achieve this however, it would be necessary to use non-water-based milling fluids with surfactants which would be impossible to burn-out prior to sintering. Comparisons between each of the mechanical properties reveal little correlation between them. References - 152 - References: 1. Walnsley, J.C. and Lang, A.R. A transmission electron microscope study of a cubic boron nitride-based compact material with AlN and AlB2 binder phases, Journal of Materials Science, 22, (1987) pp. 4093. 2. Pipkin, N.J, Roberts, D.C and Wilson, W.I, Amborite ? A remarkable new cutting material, Advances in Ultrahard Materials Applications Technology, Vol. 1, edited by Paul Daniel, DeBeers Industrial Diamond Division, Hornbeam Press limited, (1982), pp. 72 ? 80. 3. Benko, E, et al, BN Sintered with Al: Microstructure and Hardness, Ceramics International, 23, (1997), pp. 89-91 4. Benko, E, et al, cBN-metal/metal nitride composites, Ceramics International, 26, (2000) pp. 639-644 5. Sumiya, H et al, Mechanical properties of high purity polycrystalline cBN synthesized by direct conversion sintering method, Journal of Material Science, 35, (2000), pp. 1181 ? 1186. 6. Rong, X.Z. and Fukunaga, O. High-pressure sintering of cBN-TiN-Al composite for cutting tool application, Diamond and Related Materials, 11, (2002), pp. 280- 286 7. Harris, T.K,, Brookes, E.J, Taylor, C.J, The flow stress of PcBN cutting tool materials at high temperatures, International Journal of Refractory Metals and Hard Materials, 19 (2001), pp. 267 ? 273. References - 153 - 8. Li, Y et al, Study of high-pressure sintering behaviour of cBN composites starting with cBN-Al mixtures, Journal of Materials Research, 23 (9), (2008), pp. 2366 ? 2372 9. Pauleau, Y, Materials surface processing by direct energy techniques, European Materials Research Society, Elsevier Science and Technology, (2005), pp. 375. 10. Eichler, J and Lesniak, C, Boron nitride (BN) and BN composites for high- temperature applications, Journal of the European Ceramics Society, 28, (2008), pp. 1105 - 1109. 11. Riedel, R. Handbook of Ceramic Hard Materials, Wiley-VCH, 2002, Vol 2. 12. Casanova, C.A.M, et al, Experimental study of plastic deformation during sintering of cubic boron nitride compacts, Diamond and Related Materials, 8 (1999), pp. 1451 ? 1454. 13. Crystal structure of Boron Nitride allotropes, Courtesy of Michigan University, Materials Science Department, www.eecs.umich.edu/USL/MaterialScience/MotivationBN.html 14. Wentorf, R. H., Jr., Journal of Chemical Physics, 26, (1957), pp. 956 15. Wentorf, R. H., Jr., Sintered Superhard Materials, Science, 208, (1980), pp. 880-812 16. Wada, T and Yamashita, N, Formation of cBN films by ion beam assisted deposition, Journal of Vacuum Science and Technology, A 10 (3), (1992), pp. 515 ? 520. References - 154 - 17. Benko, E, et al, Microstructure and Hardness of cBN-Zr Composite, Journal of European Ceramics Society, 18, (1997), pp. 389-393.Benko, E, CBN-TiH2 composites: chemical equilibria, microstructure and mechanical testing, Diamond and Related Materials, 6, (1997), pp. 1192-1197 18. Sithebe, H. S. L, The preparation and characterisation of cBN ceramics with Al-based binder phase, Thesis, University of the Witwatersrand, Johannesburg, South Africa, (2006). 19. Sithebe, H. S. L, McLachlan, D, Sigalas, I and Herrmann, M; Pressure infiltration of boron nitride performs with molten aluminium, Ceramics international, (2007), article in press. 20. Wentorf, R. H., Jr. and Rocco W. A, Method Converting Hexagonal BN into cubic BN, Japan. Patent Applied. No. 65392 (1972) 21. Wentorf, R. H., Jr., Synthesis of Cubic Form of Boron Nitride, Journal of Chemical Physics, vol. 34 (3), (1961), pp. 809-812 22. Hibbs L E, Jr, Wentorf, R H, Jr, Borazon and diamond compact tools, High Temperature ? High Pressures, 6, (1974) pp. 409 ? 413 23. Sumitomo Electric Industries, Ltd. patented (US Patent 4 334 063, 1982) 24. Hara, A. and Yazu, S. Sintered Cutting Tool and Manufacturing Process, Japan Patent Application Number 21633 (1979). 25. Yazu, S., Kohno, Y., Satu, S. and Hara, A. New cBN-TiN composite sintered under ultra high pressure, Sumitomo Electric Industries limited. References - 155 - 26. Rong, X.Z. and Fukunaga, O. Sintering of cubic boron nitride with added aluminium at high pressure and high temperatures, Transactions: Materials Research Society of Japan, 14B, (1994), pp. 1455. 27. Zhao, Yu-Cheng and Wang, Ming-Zhi, Cubic BN with Al under High Temperature and High Pressure, Chin. Phys. Lett., Vol. 24, No. 8, (2007), pp. 2412- 2414. 28. Rong, X.Z. and Fukunaga, O. Cubic BN-WC-Co composite sintered at high pressures and high temperatures, Diamond Film Technology, 3, (1993) pp. 65. 29. Morgiel, J and Benko, E, Microstructure of boron nitride sintered with titanium, Materials Letter, 25, (1995) pp .49-52 30. Wakatsuki M et al, Synthesis of Polycrystalline Cubic BN, Mater. Res. Bull. 7 (9), (1970), pp. 999 ? 1004 31. Benko, E, et al, CBN-TiSiC2 composites, Diamond and Related Materials, 13, (2004), pp. 521-525 32. Munz, D, What can we learn from R-curve Measurements? Journal of American Ceramics Society, 90[1], (2007), pp 1-15. 33. Types of Fracture modes of crack deformation, University of Santa Cruz, earth 150 lecture notes, http://ic.ucsc.edu/~casey/eart150/Lectures/Joints/joints.htm 34. Broek, D, Elementary Fracture Mechanics, 4th edition, (1986) Nijhof, The Hague 35. H?bner and Jillek, Subcritical crack extension and crack resistance in polycrystalline alumina, Journal of Materials Science, 11 (1976), pp. References - 156 - 36. Knehans and Steinbrech, Quantitative Correlation of Acoustic Emission to the Brittle Fracture of Porous Sintered Glass, Materials Science and Engineering, 61 (1983), pp. 17 ? 22. 37. Evans, A.G and Marshall, D.B, Mechanical Behaviour of Ceramic Matrix Composites, In: Fibre Reinforce Ceramic Composites: Materials, Processing and Technology, edited by K.S. Mazdiyasni, Noyes Publications (1990), pp. 1 - 38 38. Sarkar, D, R-curve behaviour of Ti3SiC2, Ceramics International, 2006. 39. Evans, A.G, Perspective on the Development of High-Toughness Ceramics, Journal of American Ceramics Society, 73 [2] (1990) pp. 187 ? 206 40. Meguid, S.A, Mechanics and Mechanisms of Toughening of Advanced Ceramics, Journal of Materials Processing Technology, 56, (1996), pp 978 ? 989 41. Steinbrech, R.W, Toughening Mechanisms for Ceramic Materials, Journal of the European Ceramics Society, 10, (1992), pp. 131 ? 142 42. Wang, Z.Y, et al, J-Integral analysis of PcBN tool crack propagation in hybrid machining of advanced materials, Journal of Manufacturing Processes, vol. 6 [2] (2004), pp. 196 ? 200. 43. Lammer, A. Mechanical Properties of polycrystalline diamonds, Materials Science and Technology, 4, (1988), pp. 949 ? 955. 44. Meiss, D and Rai, G, Fracture toughness and thermal resistance of polycrystalline diamond compacts, Materials Science and Engineering A, 209, (1996), pp. 270 ? 276. References - 157 - 45. Huang, B. L, et al, Fracture toughness of Sintered Polycrystalline Diamond (PCD), Int. Conf. on Advanced Particulate Materials and Processes, 5th, (1997), pp. 431 ? 438. 46. Hooper R.M and Brookes C.A, Microstructure and wear of cubic boron nitride aggregate tools. In: Almond AE, Brookes CA, Warren R, editors. Science of Hard materials. Institute of Physics Conference Series No 75. Bristol: Adam Hilger; 1986, pp. 907 ? 917. 47. Hooper, RM, Guilliou, MO, Henshall, JL, Indentation studies of cBN-TiC Composites, J. Hard Mater, 2 (1991), pp. 223 ? 231. 48. R. Rice, Hardness ? Grain-Size Relationship in Ceramics, Journal of American Ceramics, Vol 77 [10], (1994), pp. 2539 ? 2553 49. Rice, R.W, Effect of Stresses from Thermal expansion Anisotropy, Phase transformation and Second Phases on the Strength of Ceramics, Journal of American Ceramics Society, 63, (1979), pp 703 ? 710. 50. Ishizaki, K; Yanai, T, Si3N4 Grain Boundary Reinforcement by SiC Nanocomposites, Silicates Industrials, 7-8 (1995), pp. 215 - 222. 51. Materials for High-temperature Semiconductor Devices, National Research Council (U.S.), Committee on Materials for High temperature semiconductor devices, Commission on Engineering and Technical Systems, National Academies Press, (1995), pp. 25. 52. Valley design Corp., Properties of Aluminium nitride substrates, http://www.edgepolishing.com/aluminum-nitride.htm References - 158 - 53. Sung-Tag Oh, et al, Effect of particle size distribution and mixing homogeneity on microstructure and strength of alumina/copper composites, NanoStructured Materials, vol. 10 [2] 1998, pp. 327 ? 332. 54. Warren, R, Ceramic-Matrix Composites, Chapman Hall, New York, 2007, pp. 4-7, 66-69. 55. Gong, J et al, A comparison between Knoop and Vickers hardness of silicon nitride ceramics, Materials Letters, 56, (2002), pp. 941-944 56. ASTM C1327 -08: Standard Test Method Vickers Indentation Hardness of Advanced Ceramic Materials. Annual Book of ASTM Standards, American Society for Testing Materials, Pennsylvania (2008). 57. ASTM C1326 -08e1: Standard Test Method Knoop Indentation Hardness of Advanced Ceramic Materials. Annual Book of ASTM Standards, American Society for Testing Materials, Pennsylvania (2008). 58. ASTM C1421-09: Standard Test Method for Fracture Toughness of Advanced Ceramics at Ambient Temperatures. Annual Book of ASTM Standards, American Society for Testing Materials, Pennsylvania (2009) 59. K?bler, J, Fracture toughness of Ceramics using the SEVNB Method; Round Robin, VAMAS Report No. 37 (1999) 60. Moon, R, K, Bowman, Trumble, K and R?del, R; A comparison of R-curves from SEVNB and SCF Fracture Toughness Tests Methods on Multi-layered Alumina-Zirconia Composites, Journal of the American Ceramics Society, 83 [2] (2000), pp. 445 ? 447 References - 159 - 61. Raddatz, O, Modeling of R-curve behaviour in Ceramic/metal Composites, Acta mater. Vol 46 [8] (1998), pp. 6381 ? 6395. 62. ASTM C1161 - 02c: Standard Test Method for Flexural Strength of Advanced Ceramics at Ambient Temperature. Annual Book of ASTM Standards, American Society for Testing Materials, Pennsylvania (2002). 63. ISO 14704:2000 (2000): Fine ceramics (advanced ceramics, technical ceramics) Test methods for flexural strength of monolithic ceramics at room temperature. 64. Weibull, W, The Phenomenon of Rupture of Solids, Ingeniors Ventenskaps Akademien, No. 153 (1939) 65. ASTM standard C 1239 ? 95: Standard Practice for reporting Uniaxial strength data and Estimating Weibull Distribution Parameters for Advanced Ceramics. Annual Book of ASTM Standards, American Society for Testing Materials, Pennsylvania (1995). 66. Basu, B et al, Is Weibull distribution the most appropriate statistical strength distribution for brittle material, Ceramics International, (2008), Article in press. 67. European Standard EN 843-5: Advanced Technical Ceramics ? Monolithic Ceramics Mechanical Properties at Room Temperature ? Part 5: Statistical evaluation of strength values, 1997. 68. Thoman, D.R, C.J. Bain, C.E. Antle, "Interferences on the parameters of the Weibull distribution," Technometrics, 11, [3], (1969), pp. 445-460 References - 160 - 69. Wachtman, J, Mechanical Properties of Ceramics, Wiley & Sons, (1996), pp.159 ? 175, 357 ?366. 70. Munz, D, Fett, T, Ceramics: Mechanical Properties, Failure Behaviour, Materials Selection, Springer, 1999, pp. 19-73 71. Thomaszewski, H, Boniecki, M and Weglarz, H, Effect of grain size on R- curve behaviour of alumina ceramics, Journal of the European Ceramics Society, 20 (2000), pp. 2569 ? 2574. 72. R?del, J, Fuller, Jr. E.R, and Lawn, B.R, "In-situ Observations of Toughening Processes in Alumina Reinforced with Silicon Carbide Whiskers", Journal of the American Ceramics Society, 74 [12] (1991) pp. 3154-3157. 73. R?del, J, "Interaction Between Crack Deflection and Crack Bridging", Journal of the European Ceramics Society, 10 [3] (1992) pp.143-150. 74. Seidel, J. Claussen, N and R?del, J, "Reliability of Alumina: Effect of Grain Size", Journal of the European Ceramics Society, 15 (1995) pp. 395-404. 75. A. Zimmermann, M. Hoffman, B. D. Flinn, R. K. Bordia, T. Chuang, E. R. Fuller Jr. and J. R?del, "Fracture of Alumina with Controlled Pores", Journal of the American Ceramics Society, 81 ?9? (1998) pp. 2449?2457. 76. Element Six (Pty) Ltd, Private Correspondence, Mr J. Reid. 77. Drozda, T, Tool and manufacturing engineers handbook, Society of Manufacturing Engineers, Vol. 1, 4th edition, (1983), chapter 3, pp. 19 Appendices - 161 - Appendices: Appendix A: Rietveld Analysis The percentages associated with each plot are for crystalline materials only. Any amorphous materials present could significantly alter the validity of these results. The results have been calculated based on a simple analytical peak shape model ? this could be improved but would most likely not be worthwhile based on the amount of effort required. Rietveld refinement was done using Topas A version 4.1 and Structural models used were taken from the ICSD - version 1.4.4. Powder X-ray diffraction data were collected using a Bruker AXS D8 equipped with a primary beam G?bel mirror, a radial Soller slit, a V?ntec-1 detector and using Cu-K? radiation (40kV, 40mA). Data were collected in the 2? range 5 to 100? in 0.021? steps, using a scan speed resulting in an equivalent counting time of 73.2 s per step. G20_cBN_25Al 00-039-1483 (*) - Aluminum Boron - AlB2 - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Hexagonal - a 3.00540 - b 3.00540 - c 3.25276 - alpha 90.000 - beta 90.000 - gamma 120.000 - Primitive - P6/mmm (191) - 1 - 25.444 00-004-0787 (*) - Aluminum, syn - Al - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Cubic - a 4.04940 - b 4.04940 - c 4.04940 - alpha 90.000 - beta 90.000 - gamma 90.000 - Face-centered - Fm-3m (225) - 4 - 66.4006 - I/Ic 00-025-1133 (*) - Aluminum Nitride - AlN - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Hexagonal - a 3.11140 - b 3.11140 - c 4.97920 - alpha 90.000 - beta 90.000 - gamma 120.000 - Primitive - P63mc (186) - 2 - 41.7447 - 00-035-1365 (*) - Boron Nitride - BN - Y: 50.00 % - d x by: 1. - WL: 1.5406 - Cubic - a 3.61580 - b 3.61580 - c 3.61580 - alpha 90.000 - beta 90.000 - gamma 90.000 - Face-centered - F-43m (216) - 4 - 47.2730 - F6= 5 Operations: Import G20_cBN_25Al - File: G20 cBN 25Al_new.raw - Type: 2Th/Th locked - Start: 5.000 ? - End: 99.997 ? - Step: 0.014 ? - Step time: 73.2 s - Temp.: 25 ?C (Room) - Time Started: 16 s - 2-Theta: 5.000 ? - Theta: 2.500 ? - C Lin ( Co unts ) 0 1000 2000 3000 4000 5000 6000 7000 8000 2-Theta - Scale 26 30 40 50 60 70 80 90 Figure A 1: XRD scan of G20 cBN +25 vol.% Al showing various phases. Rieveld refinement shown in Figure A2- A4 shows that there is approximately 1.28 wt.% Al in G6 cBN + 25 vol.% Al sample, 1.05 wt.% in G20 cBN + 20vol.%Al Appendices - 162 - and 2.73wt.% in G20 cBN + 25vol.%Al samples. Figure A 2: Rietveld analysis of G6 cBN + 25vol.% Al. Figure A 3: Rietveld analysis of G20 cBN + 20vol.% Al. Figure A 4: Rietveld analysis of G20 cBN + 25vol.% Al. Appendices - 163 - Appendix B: Determination of the Grain Size and Binder Phase Composition using Image Analysis Tools Image analysis techniques have been used to quantify and characterise the microstructure of the sintered PCBN-Al composite materials. The microstructures of the sintered materials were looked at in a Phillips-ESEM XL30 and 10 SEM micrographs of the microstructure were taken. The magnification of the micrographs depended on the size of the cBN materials; a high magnification was used for G2 and G6 materials while a low magnification was used for G10 and G20. The micrographs represented the average microstructure for each material. Using image analysis software program (AnalySIS ?Pro?) the grain size and the percentage of the binder and cBN phases were determined. The method used to determine the percentage of the phases was by measuring the area of light intensity of the different phases. The cBN phase shows a darker intensity than the binder phase. The grain size and grain size distribution was determined by measuring the grain size for each grain and calculating the distribution. The data from the AnalySIS program was then processed automatically by using Mathcad and Mathconnex programs. The program module was designed to calculate the grain size distribution and mean free path of both the cBN and binder phase for the PcBN materials. Grain Size and Composition of the Sintered Materials: The grain size and composition of the various PcBN-Al composite materials are shown in tables B 1-8. The volume percentage of the cBN and binder phase with standard deviation of the various PcBN-Al composite materials are shown in tables B 1, 3, 5 and 7. The grain size distribution and mean free path (MFP) of both the cBN grains and the binder phase for the various PcBN-Al composite materials are shown in tables B2, 4, 6 and 8. Appendices - 164 - Figures B 1-4 shows the modified SEM micrographs obtained from the AnalySIS program. These micrographs show colour differences in the cBN and binder phases and the various grain sizes. Figure B1a-f show the micrographs of the G2 cBN-Al composite materials. The grain size of the different composites can be seen in Figure B1a, c and e; the red and green phases show the different grain sizes while the black phase shows the binder phase. The phase composition is shown in Figure B1 b, d and f; the green phase is the cBN phase and the blue phase is the binder phase. A clear difference in the amount of binder phase can be seen. Similarly for Figures B2-4; although in Figure B2 b, d and f the cBN phase is shown in blue and the binder phase is green. The grain size images in Figures B3 and B4 show the larger cBN grain sizes in yellows and blues. G2 cBN Materials: Table B. 1: Phase composition of the G2 cBN-Al composite materials Volume % of Phases Material cBN Binder Standard deviation G2 cBN + 15%Al 82.12 17.88 1.45 G2 cBN + 20%Al 70.21 29.79 1.84 G2 cBN + 25%Al 60.02 39.98 1.65 Table B. 2: Grain size of the G2 cBN-Al composite materials G2-15 G2-20 G2-25 Mean cBN grain size (?m) 1.49 1.72 1.78 D10 (?m) 0.89 0.97 1.04 D50 (?m) 1.46 1.70 1.72 D90 (?m) 2.10 2.24 2.37 Maximum cBN grain size (?m) 2.79 3.30 3.21 cBN MFP (?m) 0.79 0.39 0.50 Standard deviation (?m) 0.90 0.71 0.81 Maximum cBN MFP (?m) 9.58 10.39 8.96 Binder pool size (?m) 0.78 1.29 1.90 Standard deviation (?m) 0.40 0.56 0.69 Maximum Binder pool size (?m) 2.14 3.12 4.70 Binder Mean Free Path (MFP) (?m) 0.21 0.26 0.36 Standard deviation (?m) 0.24 0.50 0.69 Appendices - 165 - Maximum binder MFP (?m) 2.5 9.73 10.64 Figure B. 1: Modified SEM micrographs for the determination of grain size and phase composition of the G2 cBN-Al composite materials. a) G2 cBN + 15vol.% Al grain size and b) G2 cBN + 15vol.% Al phase composition; c) G2 cBN + 20vol.% Al grain size and d) G2 cBN + 20vol.% Al phase composition; e) G2 cBN + 25vol.% Al grain size and f) G2 cBN + 25vol.% Al phase composition. G6 cBN Materials: Table B. 3: Phase composition of the G6 cBN-Al composite caterials: Volume % Phases cBN Binder Standard deviation G6cBN+15%Al 74.25 % 25.75 % 1.89 G6cBN+20%Al 67.97 % 32.03 % 2.95 G6cBN+25%Al 59.29 % 40.71 % 4.13 Appendices - 166 - Table B. 4: Grain size of the G6 cBN-Al composite materials: G6-15 G6-20 G6-25 Mean cBN grain size (?m) 4.68 4.73 4.72 D10 (?m) 2.62 2.57 2.9 D50 (?m) 4.73 4.9 4.93 D90 (?m) 6.44 6.41 5.98 Maximum cBN grain size (?m) 7.38 7.23 6.95 cBN MFP (?m) 1.56 1.59 1.88 Standard deviation (?m) 2.09 2.02 2.15 Maximum cBN MFP (?m) 17.58 17.84 19.18 Binder pool size (?m) 3.13 4.46 4.80 Standard deviation (?m) 1.18 1.53 1.75 Maximum Binder pool size (?m) 7.36 9.36 9.76 Binder Mean Free Path (MFP) (?m) 0.90 1.26 1.55 Standard deviation (?m) 1.2 1.77 2.11 Maximum binder MFP (?m) 13.22 18.73 19.18 Appendices - 167 - Figure B. 2: Modified SEM micrographs for the determination of grain size and phase composition of the G6 cBN-Al composite materials. a) G6 cBN + 15vol.% Al grain size and b) G6 cBN + 15vol.% Al phase composition; c) G6 cBN + 20vol.% Al grain size and d) G6 cBN + 20vol.% Al phase composition; e) G6 cBN + 25vol.% Al grain size and f) G6 cBN + 25vol.% Al phase composition. G10 cBN Materials: Table B. 5: Phase composition of the G10 cBN + 15%Al composite material: Volume % Phases cBN Binder Standard deviation G10cBN+15%Al 80.73 % 19.27 % 1.39 Appendices - 168 - Table B. 6: Grain csze of the G10 cBN + 15%Al composite material: G10-15 Mean cBN grain size (?m) 10.79 D10 (?m) 5.90 D50 (?m) 10.96 D90 (?m) 14.86 Maximum cBN grain size (?m) 17.18 cBN MFP (?m) 2.85 Standard deviation (?m) 4.50 Maximum cBN MFP (?m) 47.34 Binder pool size (?m) 4.35 Standard deviation (?m) 1.45 Maximum Binder pool size (?m) 12.15 Binder Mean Free Path (MFP) (?m) 1.15 Standard deviation (?m) 1.61 Maximum binder MFP (?m) 20.09 Figure B. 3: Modified SEM micrographs for the determination of grain size and phase composition of the G10 cBN-Al composite material. a) G10 cBN + 15vol.% Al grain size and b) G10 cBN + 15vol.% Al phase composition; G20 cBN Materials: Table B. 7: Phase omposition of the G20 cBN ?Al composite materials: Volume % Phases cBN Binder Standard deviation G20cBN+15%Al 75.08% 24.92% 2.27% G20cBN+20%Al 70.55% 29.45% 2.27% G20cBN+25%Al 60.21% 39.79% 2.73% Appendices - 169 - Table B. 8: Grain size of the G20 cBN-Al composite materials: G20-15 G20-20 G20-25 Mean cBN grain size (?m) 12.52 13.25 12.76 D10 (?m) 5.05 5.12 5.48 D50 (?m) 12.27 13.20 12.88 D90 (?m) 18.72 18.93 18.81 Maximum cBN grain size (?m) 25.39 27.75 22.64 cBN MFP (?m) 4.81 5.25 5.46 Standard deviation (?m) 5.85 5.95 5.64 Maximum cBN MFP (?m) 54.09 63.53 48.90 Binder pool size (?m) 7.70 7.70 10.99 Standard deviation (?m) 3.14 3.14 4.31 Maximum Binder pool size (?m) 17.99 17.99 23.11 Binder Mean Free Path (MFP) (?m) 2.12 2.59 4.02 Standard deviation (?m) 2.39 3.02 4.66 Maximum binder MFP (?m) 25.03 32.15 43.39 Appendices - 170 - Figure B. 4: Modified SEM micrographs for the determination of grain size and phase composition of the G20 cBN-Al composite materials. a) G20 cBN + 15vol.% Al grain size and b) G20 cBN + 15vol.% Al phase composition; c) G20 cBN + 20vol.% Al grain size and d) G20 cBN + 20vol.% Al phase composition; e) G20 cBN + 25vol.% Al grain size and f) G20 cBN + 25vol.% Al phase composition. Appendices - 171 - Appendix C: Hardness Estimation Using the assumption of the rule of volume mixtures, the hardness can be estimated for the materials also assuming that the major binder phase consists of AlN. HV = vol. % cBN x HV (cBN) + vol. % AlN x HV (AlN) 85% cBN HV = 0.85 x 45 + 0.15 x 12 = 40.05 GPa 80% cBN HV = 0.80 x 45 + 0.20 x 12 = 38.4 GPa 75% cBN HV = 0.75 x 45 + 0.25 x 12 = 36.75 GPa 40 50 60 70 80 90 20 25 30 35 40 45 Hard n ess, H v (G P a) Initial vol % cBN Figure C. 1: Hardness estimation with initial vol.% cBN. Appendices - 172 - Appendix D: Results of the Hardness Measurements Table D shows the results of the Vickers Hardness measurements. The Hardness procedure is described in section 4.5. Table D. 1: Results of the Vickers mardness measurements. Hardness, HV5 (GPa) S a m p le G 2 cBN + 1 5 %A l G 2 cBN + 2 0 %A l G 2 cBN + 2 5 %A l G 6 cBN + 1 5 %A l G 6 cBN + 2 0 %A l G 6 cBN + 2 5 %A l G 1 0 cBN + 1 5 %A l G 2 0 cBN + 1 5 %A l G 2 0 cBN + 2 0 %A l G 2 0 cBN + 2 5 %A l 1 42.07 32.38 31.19 34.29 27.51 21.78 37.68 26.03 21.66 15.54 2 41.18 34.97 32.38 34.29 27.51 20.51 31.68 24.44 21.20 16.17 3 39.48 32.38 33.64 37.12 27.51 21.00 36.38 21.86 20.26 15.34 4 42.07 33.64 29.00 37.12 26.58 24.75 37.88 22.16 18.46 15.96 5 38.67 34.29 30.62 36.38 27.04 22.21 35.67 23.25 19.08 15.65 6 40.31 34.79 33.64 37.12 27.04 21.84 32.43 21.53 18.04 15.14 7 41.08 33.81 33.00 37.42 26.82 20.50 36.53 21.59 18.40 15.28 8 41.08 32.02 33.00 34.70 27.57 23.43 38.47 25.05 21.50 16.02 9 10 Mean 40.74 33.54 32.06 36.06 27.20 22.00 35.84 23.24 19.83 15.64 std dev 1.20 1.15 1.65 1.39 0.38 1.47 2.51 1.74 1.50 0.38 Appendices - 173 - Appendix E: Results of the SEVNB Fracture Toughness The results of the Fracture toughness testing using the Single-edge V notch beam (SEVNB) method is shown the Tables E1-10. The fracture toughness procedure is described in Section 4.5. The fracture toughness K1c is calculated using the following equations: ? ? Y W SS WB F YaK Ic 5.1 21 12 3 .. ? ?? ? ? ?? ? ? ? ? ? ?2 2 1 135.168.049.3 326.19887.1 ? ????? ? ??? ???Y and W a ?? Table E. 1: G2cBN+15%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1305.39 0.00374 0.349 246.91 0.00305 1323746.4 2.674 1.687 1.099 6.565 2 1251.25 0.00372 0.336 256.22 0.00304 1381867.1 2.688 1.609 1.116 6.671 3 923.67 0.00398 0.232 324.13 0.00304 1690083.4 2.513 1.074 1.281 5.842 4 1064.30 0.00377 0.282 297.93 0.00302 1606717.5 2.653 1.311 1.195 6.674 Mean 6.438 Std dev 0.400 Table E. 2: G2cBN+20%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1180.08 0.00410 0.288 374.47 0.0029 2016607 2.439 1.339 1.186 7.811 2 1155.08 0.00406 0.285 341.37 0.00289 1853807 2.463 1.322 1.191 7.191 3 978.28 0.00412 0.237 366.54 0.00286 1996650 2.427 1.098 1.271 6.763 4 987.97 0.00407 0.243 347.23 0.00289 1883312 2.457 1.122 1.262 6.547 5 971.09 0.00410 0.237 385.50 0.00289 2083201 2.439 1.095 1.272 7.079 Mean 7.078 Std dev 0.483 Appendices - 174 - Table E. 3: G2cBN+25%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 973.20 0.00410 0.237 418.59 0.00303 2157528 2.439 1.097 1.271 7.342 2 960.70 0.00409 0.235 404.80 0.00303 2089000 2.445 1.086 1.276 7.079 3 1007.58 0.00410 0.246 380.67 0.00303 1962068 2.439 1.135 1.256 6.824 4 934.84 0.00411 0.227 420.83 0.00301 2180833 2.433 1.054 1.290 7.211 5 1027.34 0.00411 0.250 373.95 0.00287 2032396 2.433 1.155 1.249 7.129 Mean 7.117 Std dev 0.191 Table E. 4: G6cBN+15%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1115.86 0.00411 0.271 369.47 0.00278 2073045 2.433 1.257 1.212 7.685 2 1004.30 0.00414 0.243 403.60 0.00276 2272676 2.415 1.121 1.262 7.764 3 1195.47 0.00408 0.293 369.98 0.00278 2083564 2.451 1.366 1.178 8.217 4 925.86 0.00407 0.227 343.78 0.00273 1973897 2.457 1.054 1.290 6.592 Mean 7.564 Std dev 0.690 Table E. 5: G6cBN+20%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1093.97 0.00420 0.260 387.57 0.00277 2158949 2.381 1.204 1.231 7.614 2 1087.11 0.00422 0.258 396.87 0.00274 2229691 2.370 1.190 1.235 7.769 3 1077.11 0.00418 0.258 383.77 0.00275 2158508 2.392 1.191 1.235 7.595 4 1076.72 0.00422 0.255 408.59 0.00276 2278907 2.370 1.179 1.240 7.891 5 1117.11 0.00422 0.265 359.12 0.00273 2025002 2.370 1.224 1.223 7.186 Mean 7.611 Std dev 0.267 Table E. 6: G6cBN+25%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1070.00 0.00411 0.260 350.33 0.00285 1917411 2.433 1.203 1.231 6.908 2 1099.30 0.00411 0.267 340.85 0.00286 1859003 2.433 1.237 1.219 6.821 3 1035.86 0.00411 0.252 366.54 0.00287 1992112 2.433 1.164 1.245 7.025 4 1040.63 0.00413 0.252 386.01 0.00285 2107576 2.421 1.164 1.245 7.395 Mean 7.038 Std dev 0.253 Appendices - 175 - Table E. 7: G10cBN+15%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1091.95 0.00408 0.268 380.84 0.00275 2168119 2.451 1.238 1.219 8.017 2 1051.88 0.00409 0.257 401.01 0.00279 2247446 2.445 1.188 1.236 8.071 3 1101.41 0.00418 0.263 396.70 0.00274 2239363 2.392 1.218 1.225 7.997 4 950.86 0.00408 0.233 416.35 0.00276 2361677 2.451 1.078 1.279 7.984 5 1071.41 0.00416 0.258 404.80 0.00274 2290580 2.404 1.190 1.236 8.095 Mean 8.033 Std dev 0.048 Table E. 8: G20cBN+15%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1071.48 0.00415 0.258 406.70 0.00279 2262788 2.410 1.193 1.234 8.030 2 1068.36 0.00415 0.257 410.66 0.0028 2276688 2.410 1.189 1.236 8.063 3 1042.81 0.00414 0.252 389.12 0.002758 2192727 2.415 1.163 1.245 7.674 4 1059.61 0.00415 0.255 404.29 0.00278 2257453 2.410 1.179 1.239 7.952 Mean 7.930 Std dev 0.177 Table E. 9: G20cBN+20%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1091.02 0.00416 0.262 368.78 0.00283 2020369 2.404 1.212 1.228 7.227 2 1094.94 0.00410 0.267 393.94 0.00287 2143678 2.439 1.235 1.219 7.877 3 1116.02 0.00406 0.275 374.98 0.00285 2064917 2.463 1.274 1.207 7.816 4 1310.86 0.00418 0.314 369.47 0.00288 1984238 2.392 1.477 1.148 8.046 5 1292.81 0.00418 0.309 369.64 0.00284 2013122 2.392 1.453 1.154 8.075 Mean 7.808 Std dev 0.343 Table E. 10: G20cBN+25%Al S a mp le a , ? m W , m a /W L o a d ( F ), N B i J k Y K 1 C (M P a .m 0 .5 ) 1 1205.23 0.00420 0.287 334.48 0.00286 1804567 2.381 1.335 1.187 6.809 2 1200.55 0.00418 0.287 345.51 0.00287 1862031 2.392 1.336 1.187 7.064 3 1097.19 0.00417 0.263 357.92 0.00288 1924517 2.398 1.216 1.226 6.883 4 1039.22 0.00416 0.250 373.95 0.00289 2006165 2.404 1.154 1.249 6.950 5 1042.50 0.00419 0.249 381.53 0.00287 2053724 2.387 1.149 1.251 7.045 Mean 6.950 Std dev 0.108 Appendices - 176 - Appendix F: Results of the Compact Tension Measurements for Determining R-Curve Behaviour. The results of the Compact tension measurements to determine the R-curve behaviour of the PcBN-Al composite material is shown in Tables F 1-10. The R-curve behaviour procedure is described in Section 4.5. The resistance toughness KR is calculated using the equations below: ? ? ? ? ? ? ? W c f BW F K IR 2 1 2 3 432 1 6.572.1432.1364.4886.02 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ?? ? ? ? ? ? ? ?? ? ? ? ? ? W c W c W c W c W c W c W c f Table F. 1: G2cBN+15% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 39.3 0.0393 11.304 1.1304 0.3804 39.1 0.0391 0.40 0.07 7.35 0.48 11.797 1.1797 0.4297 367.1 0.3671 0.42 0.61 7.71 4.71 12.321 1.2321 0.4821 343.5 0.3435 0.44 0.57 8.11 4.64 12.783 1.2783 0.5283 329.1 0.3291 0.46 0.55 8.50 4.66 13.254 1.3254 0.5754 344.2 0.3442 0.47 0.57 8.92 5.12 2 7.5 0.75 0 116.3 0.1163 10.286 1.0286 0.2786 115.5 0.1155 0.37 0.19 6.68 1.29 10.662 1.0662 0.3162 646.5 0.6465 0.38 1.08 6.92 7.46 11.13 1.113 0.363 673 0.673 0.40 1.12 7.23 8.11 11.513 1.1513 0.4013 701 0.701 0.41 1.17 7.50 8.76 Appendices - 177 - Table F. 2: G2cBN+20% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 60.9 0.0609 10.901 1.0901 0.3401 60.4 0.0604 0.39 0.10 7.08 0.71 11.42 1.142 0.392 119.5 0.1195 0.41 0.20 7.43 1.48 11.939 1.1939 0.4439 143.6 0.1436 0.43 0.24 7.81 1.87 12.413 1.2413 0.4913 364.5 0.3645 0.44 0.61 8.19 4.97 12.878 1.2878 0.5378 372.4 0.3724 0.46 0.62 8.58 5.32 13.357 1.3357 0.5857 383.1 0.3831 0.48 0.64 9.01 5.76 13.838 1.3838 0.6338 381.5 0.3815 0.49 0.64 9.49 6.03 14.328 1.4328 0.6828 376.6 0.3766 0.51 0.63 10.02 6.29 14.807 1.4807 0.7307 372.3 0.3723 0.53 0.62 10.58 6.57 2 7.5 0.75 0 151.1 0.1511 11.512 1.1512 0.4012 150.4 0.1504 0.41 0.25 7.50 1.88 11.966 1.1966 0.4466 510.3 0.5103 0.43 0.85 7.83 6.66 12.438 1.2438 0.4938 516.2 0.5162 0.44 0.86 8.21 7.06 12.889 1.2889 0.5389 530.1 0.5301 0.46 0.88 8.59 7.59 Table F. 3: G2cBN+25% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 51.7 0.0517 11.139 1.1139 0.3639 50.6 0.0506 0.40 0.08 7.24 0.61 11.577 1.1577 0.4077 306.7 0.3067 0.41 0.51 7.54 3.86 12.096 1.2096 0.4596 430 0.43 0.43 0.72 7.93 5.69 12.563 1.2563 0.5063 429.9 0.4299 0.45 0.72 8.31 5.95 13.049 1.3049 0.5549 428.4 0.4284 0.47 0.71 8.73 6.23 13.521 1.3521 0.6021 423.8 0.4238 0.48 0.71 9.17 6.48 Table F. 4: G6cBN+15% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 53.2 0.0532 12.337 1.2337 0.4837 52.5 0.0525 0.44 0.09 8.12 0.71 13.205 1.3205 0.5705 82 0.082 0.47 0.14 8.87 1.21 13.642 1.3642 0.6142 217.9 0.2179 0.49 0.36 9.29 3.37 16.254 1.6254 0.8754 118.3 0.1183 0.58 0.20 12.67 3.50 19.159 1.9159 1.1659 108.4 0.1084 0.68 0.18 19.86 3.59 19.665 1.9665 1.2165 104.8 0.1048 0.70 0.17 21.82 3.81 20.174 2.0174 1.2674 93.4 0.0934 0.72 0.16 24.14 3.76 20.62 2.062 1.312 76.3 0.0763 0.74 0.13 26.51 3.37 21.075 2.1075 1.3575 69.2 0.0692 0.75 0.12 29.36 3.39 21.535 2.1535 1.4035 61.8 0.0618 0.77 0.10 32.77 3.38 2 7.5 0.75 0 58.1 0.0581 10.19 1.019 0.269 121.5 0.1215 0.36 0.20 6.63 1.34 Appendices - 178 - 10.656 1.0656 0.3156 320.9 0.3209 0.38 0.53 6.92 3.70 10.886 1.0886 0.3386 505 0.505 0.39 0.84 7.07 5.95 11.313 1.1313 0.3813 603.7 0.6037 0.40 1.01 7.36 7.40 11.699 1.1699 0.4199 601.3 0.6013 0.42 1.00 7.63 7.65 Table F. 5: G6cBN+20% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 73.7 0.0737 10.244 1.0244 0.2744 72.6 0.0726 0.37 0.12 6.66 0.81 10.743 1.0743 0.3243 203.9 0.2039 0.38 0.34 6.97 2.37 11.201 1.1201 0.3701 364.5 0.3645 0.40 0.61 7.28 4.42 11.085 1.1085 0.3585 462 0.462 0.40 0.77 7.20 5.54 11.563 1.1563 0.4063 543.9 0.5439 0.41 0.91 7.53 6.83 12.25 1.225 0.475 535.7 0.5357 0.44 0.89 8.05 7.19 12.713 1.2713 0.5213 539.3 0.5393 0.45 0.90 8.44 7.58 13.157 1.3157 0.5657 2.3 0.0023 0.47 0.00 8.83 0.03 2 7.5 0.75 0 97.7 0.0977 9.458 0.9458 0.1958 96.9 0.0969 0.34 0.16 6.19 1.00 10.026 1.0026 0.2526 185.4 0.1854 0.36 0.31 6.53 2.02 10.579 1.0579 0.3079 214.3 0.2143 0.38 0.36 6.87 2.45 11.067 1.1067 0.3567 333.4 0.3334 0.40 0.56 7.19 3.99 11.203 1.1203 0.3703 506.6 0.5066 0.40 0.84 7.28 6.15 11.689 1.1689 0.4189 557.7 0.5577 0.42 0.93 7.63 7.09 12.132 1.2132 0.4632 570.1 0.5701 0.43 0.95 7.96 7.56 12.585 1.2585 0.5085 566.8 0.5668 0.45 0.94 8.33 7.87 13.008 1.3008 0.5508 563.2 0.5632 0.46 0.94 8.69 8.16 13.452 1.3452 0.5952 550.5 0.5505 0.48 0.92 9.10 8.35 13.94 1.394 0.644 130.4 0.1304 0.50 0.22 9.60 2.09 Table F. 6: G6cBN+25% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 60.1 0.0601 9.431 0.9431 0.1931 59.7 0.0597 0.34 0.10 6.18 0.61 9.937 0.9937 0.2437 175.5 0.1755 0.35 0.29 6.47 1.89 10.435 1.0435 0.2935 270 0.27 0.37 0.45 6.78 3.05 10.44 1.044 0.294 232.8 0.2328 0.37 0.39 6.78 2.63 10.709 1.0709 0.3209 436.7 0.4367 0.38 0.73 6.95 5.06 11.209 1.1209 0.3709 562.7 0.5627 0.40 0.94 7.28 6.83 11.665 1.1665 0.4165 573 0.573 0.42 0.96 7.61 7.27 12.091 1.2091 0.4591 600.7 0.6007 0.43 1.00 7.93 7.94 12.529 1.2529 0.5029 595.3 0.5953 0.45 0.99 8.28 8.22 12.982 1.2982 0.5482 587.3 0.5873 0.46 0.98 8.67 8.49 13.459 1.3459 0.5959 581.6 0.5816 0.48 0.97 9.11 8.83 13.854 1.3854 0.6354 556.1 0.5561 0.49 0.93 9.51 8.81 Appendices - 179 - 14.279 1.4279 0.6779 546.7 0.5467 0.51 0.91 9.96 9.08 Table F. 7: G10cBN+15% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 39.3 0.0393 11.304 1.1304 0.3804 39.1 0.0391 0.40 0.07 7.35 0.48 11.797 1.1797 0.4297 367.1 0.3671 0.42 0.61 7.71 4.71 12.321 1.2321 0.4821 343.5 0.3435 0.44 0.57 8.11 4.64 12.783 1.2783 0.5283 329.1 0.3291 0.46 0.55 8.50 4.66 13.254 1.3254 0.5754 344.2 0.3442 0.47 0.57 8.92 5.12 Table F. 8: G20cBN+15% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 87.1 0.0871 10.397 1.0397 0.2897 83.5 0.0835 0.37 0.14 6.75 0.94 10.885 1.0885 0.3385 645.5 0.6455 0.39 1.08 7.07 7.60 11.31 1.131 0.381 681 0.681 0.40 1.14 7.35 8.35 11.838 1.1838 0.4338 670 0.67 0.42 1.12 7.74 8.64 12.3 1.23 0.48 651.5 0.6515 0.44 1.09 8.09 8.79 2 7.5 0.75 0 148.9 0.1489 10.187 1.0187 0.2687 146.7 0.1467 0.36 0.24 6.62 1.62 10.677 1.0677 0.3177 503.1 0.5031 0.38 0.84 6.93 5.81 11.122 1.1122 0.3622 709.6 0.7096 0.40 1.18 7.23 8.55 11.549 1.1549 0.4049 703.4 0.7034 0.41 1.17 7.52 8.82 11.987 1.1987 0.4487 701.2 0.7012 0.43 1.17 7.85 9.17 12.215 1.2215 0.4715 711 0.711 0.44 1.19 8.03 9.51 Table F. 9: G20cBN+20% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 56.1 0.0561 9.627 0.9627 0.2127 55.8 0.0558 0.34 0.09 6.29 0.59 10.112 1.0112 0.2612 165.8 0.1658 0.36 0.28 6.58 1.82 10.19 1.019 0.269 419.1 0.4191 0.36 0.70 6.63 4.63 10.642 1.0642 0.3142 599.6 0.5996 0.38 1.00 6.91 6.90 11.135 1.1135 0.3635 609 0.609 0.40 1.02 7.23 7.34 11.626 1.1626 0.4126 620.4 0.6204 0.42 1.03 7.58 7.84 2 7.5 0.75 0 103.3 0.1033 10.094 1.0094 0.2594 102.3 0.1023 0.36 0.17 6.57 1.12 10.567 1.0567 0.3067 423.1 0.4231 0.38 0.71 6.86 4.84 11.018 1.1018 0.3518 613.4 0.6134 0.39 1.02 7.16 7.31 11.458 1.1458 0.3958 620.8 0.6208 0.41 1.03 7.46 7.72 11.851 1.1851 0.4351 617.3 0.6173 0.42 1.03 7.75 7.97 Appendices - 180 - 12.051 1.2051 0.4551 613.9 0.6139 0.43 1.02 7.90 8.08 12.492 1.2492 0.4992 603.2 0.6032 0.45 1.01 8.25 8.29 12.914 1.2914 0.5414 602.3 0.6023 0.46 1.00 8.61 8.64 12.924 1.2924 0.5424 601.2 0.6012 0.46 1.00 8.62 8.64 13.463 1.3463 0.5963 588.9 0.5889 0.48 0.98 9.12 8.95 13.975 1.3975 0.6475 560.3 0.5603 0.50 0.93 9.63 9.00 Table F. 10: G20cBN+25% Al Sample Length [mm] Length, c [cm] ?c (cm) Load [N] Load, P (kN) (c/W) P/(B.W0.5) f (c/W) KR (MPa.m0.5) (calc) 1 7.5 0.75 0 483 0.0483 20.874 2.0874 1.3374 474 0.0474 0.75 0.08 28.04 2.22 21.345 2.1345 1.3845 930 0.093 0.76 0.16 31.29 4.85 21.808 2.1808 1.4308 1053 0.1053 0.78 0.18 35.11 6.16 22.321 2.2321 1.4821 956 0.0956 0.80 0.16 40.29 6.42 22.87 2.287 1.537 844 0.0844 0.82 0.14 47.35 6.66 23.345 2.3345 1.5845 666 0.0666 0.83 0.11 55.21 6.13 23.854 2.3854 1.6354 504 0.0504 0.85 0.08 66.23 5.56 24.442 2.4442 1.6942 451 0.0451 0.87 0.08 84.11 6.32 24.94 2.494 1.744 352 0.0352 0.89 0.06 106.30 6.24 25.399 2.5399 1.7899 247 0.0247 0.91 0.04 136.59 5.62 2 7.5 0.75 0 977 0.0977 10.136 1.0136 0.2636 970 0.097 0.36 0.16 6.59 1.07 10.418 1.0418 0.2918 3715 0.3715 0.37 0.62 6.77 4.19 10.799 1.0799 0.3299 4689 0.4689 0.39 0.78 7.01 5.48 11.295 1.1295 0.3795 5568 0.5568 0.40 0.93 7.34 6.82 11.757 1.1757 0.4257 5757 0.5757 0.42 0.96 7.68 7.37 12.332 1.2332 0.4832 5750 0.575 0.44 0.96 8.12 7.78 12.819 1.2819 0.5319 5698 0.5698 0.46 0.95 8.53 8.10 13.326 1.3326 0.5826 5539 0.5539 0.48 0.92 8.98 8.29 13.826 1.3826 0.6326 5458 0.5458 0.49 0.91 9.48 8.62 Appendices - 181 - Appendix G: Results of the Flexural Strength Tests The results of the flexural strength testing for the PcBN-Al composite materials are given in Tables G 1-10 below. The samples dimensions; length (l), width (b) and thickness (d); load to failure (F) and flexural strength (?f) are shown. The mean strength and standard deviation for each material composition are also shown. Sample dimensions were approximately 25x4x3 mm3. The flexural strength (?f) for a 4-point bend testing is calculated based on the equation: mm SS awhere bd Fa f 5 2 1020 2 3 12 2 ? ? ? ? ??? Table G. 1: G2 cBN + 15% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.27 3.92 2.86 756.1 353.7 2 25.33 3.87 2.97 714.6 313.9 3 25.48 3.74 3.08 1045.3 441.9 4 25.38 3.75 3.00 773.2 343.6 5 25.39 3.96 3.05 1080.2 439.8 6 25.28 3.75 2.99 759.5 339.8 7 25.39 3.79 3.06 828.8 350.3 8 25.25 3.92 2.86 704.6 329.6 9 25.40 3.78 3.08 953.5 398.8 10 25.33 3.99 2.86 744.9 342.3 11 25.29 3.76 2.97 766.8 346.7 12 25.47 3.74 3.12 953.1 392.6 13 25.37 3.78 2.84 754.5 371.2 14 25.32 4.00 2.84 827.1 384.5 15 25.27 4.00 2.85 700.9 323.6 16 25.42 3.93 3.07 903.8 366.0 17 24.49 3.96 2.93 972.2 428.98 18 25.24 3.98 3.03 1029.6 422.6 Mean 371.7 Std dev 40.8 Appendices - 182 - Table G. 2: G2 cBN + 20% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.31 4.06 2.95 874.3 371.1 2 25.19 3.96 2.96 872.8 377.3 3 25.37 4.04 2.89 769.5 342.0 4 25.36 4.01 2.90 875.3 389.3 5 25.47 4.01 2.92 1179.4 517.4 6 25.43 4.00 2.89 968.9 435.0 7 25.48 3.94 3.03 904.1 374.9 8 25.37 4.00 2.96 1029.6 440.6 9 25.31 4.06 2.88 760.9 338.9 10 25.23 4.00 3.03 1121.5 458.0 11 25.38 3.99 3.00 933.8 390.0 12 25.31 4.04 2.87 1095.7 493.8 13 25.59 3.99 2.88 1175.8 532.9 14 25.17 3.99 2.92 1023.8 451.3 15 25.23 4.06 2.86 703.0 317.5 16 25.20 3.98 2.94 885.9 386.2 17 25.31 4.04 2.92 940.4 409.4 18 25.28 4.03 2.88 875.9 393.0 19 25.26 4.04 2.92 874.9 380.9 20 25.26 4.06 2.97 774.4 324.3 Mean 406.2 Std dev 61.0 Table G. 3: G2 cBN + 25% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.19 4.01 2.90 994.8 442.4 2 25.41 4.07 2.99 1265.1 521.5 3 25.44 4.11 2.97 1085.5 449.1 4 25.48 4.07 2.85 912.8 414.1 5 25.36 4.10 3.07 1135.6 440.8 6 25.39 4.10 2.92 897.4 385.0 7 25.45 4.04 3.03 1048.2 423.9 8 25.34 4.08 2.92 894.7 385.7 9 25.49 4.09 2.98 1079.5 445.8 10 25.23 4.05 2.88 1068.9 477.3 11 25.36 4.09 2.90 791.6 345.2 12 25.36 3.99 2.88 1059.8 480.3 13 25.25 4.02 2.93 1032.4 448.7 14 25.29 4.10 3.00 1108.4 450.5 15 25.33 4.08 2.90 917.6 401.1 16 25.31 4.09 3.06 919.5 360.1 17 25.60 4.06 2.94 929.3 397.2 18 25.38 4.10 2.92 961.6 412.5 19 25.18 4.10 2.89 950.5 416.3 20 25.21 4.11 2.94 987.9 417.1 Mean 425.7 Std dev 41.8 Appendices - 183 - Table G. 4: G6 cBN + 15% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.46 4.11 2.97 1202.02 497.3 2 25.41 4.11 2.85 949.1 426.4 3 25.36 4.09 2.92 758.7 326.3 4 25.40 4.10 2.91 748.3 323.3 5 25.37 4.10 2.92 699.2 300.0 6 25.28 4.10 2.91 973.1 420.4 7 25.17 4.12 2.86 1047.1 466.0 8 25.13 4.11 2.84 1000.5 452.7 9 25.38 4.08 2.90 730.4 319.3 10 25.39 4.09 2.88 762.5 337.1 11 25.25 4.11 2.97 1087.2 449.8 12 25.11 4.08 2.92 706.1 304.4 13 25.34 4.14 2.90 881.2 379.6 14 25.49 4.09 2.90 670.8 292.5 15 25.20 4.10 2.82 1001.03 460.5 16 25.14 4.09 2.92 669.7 288.0 17 25.09 4.06 2.86 719.2 324.8 18 25.50 4.15 2.82 970.2 440.9 19 25.44 4.11 2.92 759.02 324.8 20 25.19 4.11 2.85 1026.7 461.3 Mean 379.8 Std dev 71.7 Table G. 5: G6 cBN + 20% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.45 4.19 2.95 925.5 380.7 2 25.42 4.22 2.93 952.2 394.2 3 25.28 4.18 2.83 836.1 374.6 4 28.17 4.22 2.90 991.2 418.9 5 25.46 4.23 2.93 879.2 363.1 6 25.43 4.22 2.94 893.8 367.5 7 25.23 4.18 2.77 876.9 410.1 8 25.39 4.22 2.87 891.9 384.8 9 25.40 4.21 3.02 945.8 369.5 10 25.40 4.21 2.84 965.3 426.4 11 25.13 4.17 2.87 822.8 359.3 12 25.17 4.17 2.76 884.1 417.5 13 25.29 4.22 2.76 632.2 294.9 14 25.43 4.22 2.96 907.9 368.3 15 25.40 4.22 2.94 923.5 379.7 16 25.17 4.23 2.93 920.01 380.0 17 25.37 4.18 2.83 819.5 367.2 18 25.59 4.18 2.78 904.3 419.9 19 25.19 4.18 2.89 873.3 375.2 Mean 381.7 Std dev. 30.1 Appendices - 184 - Table G. 6: G6 cBN + 25% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.37 4.13 2.85 931.9 416.7 2 25.45 4.14 2.90 982.1 423.0 3 25.39 4.18 2.84 915.3 407.2 4 25.45 4.14 2.90 908.3 391.3 5 25.37 4.17 2.87 1012.9 442.3 6 25.44 4.10 2.87 863.8 383.6 7 25.39 4.14 2.90 934.5 402.5 8 25.35 4.14 2.85 1080.5 481.9 9 25.40 4.12 2.88 893.6 392.2 10 25.28 4.11 2.88 979.8 431.1 11 25.39 4.12 2.89 967.9 421.9 12 25.44 4.12 2.93 982.7 416.7 13 25.40 4.11 2.93 982.8 417.7 14 25.33 4.13 2.90 979.8 423.1 15 25.38 4.14 2.87 1029.8 452.9 16 25.47 4.12 2.89 921.6 401.7 17 25.43 4.11 2.92 994.8 425.8 18 25.48 4.15 2.93 974.1 410.1 19 25.34 4.18 2.85 888.8 392.6 20 25.39 4.12 2.95 1001.03 418.7 Mean 417.7 Std dev 23.0 Table G. 7: G10 cBN + 15% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.44 4.18 2.93 1046 437.2 2 25.34 4.10 2.89 1020 446.7 3 25.44 4.18 2.96 1044.8 427.9 4 25.40 4.09 3.03 947.2 378.4 5 25.39 4.09 3.00 1154.4 470.4 6 25.52 4.10 3.02 1073.1 430.4 7 25.43 4.18 2.98 1111.2 449.0 8 25.37 4.17 2.91 1068.6 453.9 9 25.46 4.07 2.85 951.6 431.7 10 25.40 4.16 2.96 1114.5 458.6 11 25.45 4.17 2.95 1095.7 452.8 12 25.23 4.13 2.94 1059.5 445.1 13 25.35 4.08 2.86 964.8 433.6 14 25.45 4.12 2.78 915.5 431.3 15 25.38 4.17 3.02 1204.9 475.2 16 25.44 4.17 2.95 1164.4 481.3 17 25.40 4.16 2.98 1137.6 461.8 18 25.37 4.11 2.79 912.9 428.0 19 25.43 4.13 2.79 981.5 457.9 20 25.35 4.08 2.94 999.8 425.2 Mean 443.8 Std dev 22.7 Appendices - 185 - Table G. 8: G20 cBN + 15% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.35 4.11 2.98 893.1 367.0 2 25.41 4.12 2.93 916.6 388.7 3 25.51 4.19 2.91 810.05 342.4 4 25.38 4.13 3.00 951.7 384.0 5 25.40 4.13 2.90 892.1 385.2 6 25.45 4.14 2.82 945.7 430.8 7 25.40 4.14 2.86 900.02 398.6 8 25.46 4.12 3.00 1035.3 418.8 9 25.34 4.11 2.93 864 367.3 10 25.47 4.12 2.94 899.3 378.8 11 25.43 4.12 2.86 759.9 338.2 12 25.39 4.11 2.88 879.5 386.9 13 25.52 4.11 2.84 829.7 375.4 14 25.42 4.11 2.94 891.7 376.5 15 25.34 4.10 2.96 855.2 357.1 16 25.37 4.09 2.99 1022.4 419.4 17 25.42 4.11 2.98 885.9 364.0 18 25.45 4.12 2.87 950.1 419.9 19 25.41 4.12 2.98 954.1 391.1 20 25.34 4.10 2.95 921.7 387.5 Mean 383.9 Std dev 25.06 Table G. 9: G20 cBN + 20% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.44 4.13 2.97 846.9 348.7 2 25.38 4.14 2.73 753 366.0 3 25.37 4.12 3.05 902.8 353.3 4 25.46 4.19 2.98 829.2 334.2 5 25.44 4.11 2.99 869.3 354.8 6 25.34 4.12 2.98 708.7 290.5 7 25.39 4.13 2.90 803.3 346.9 8 25.36 4.18 2.90 786.9 335.7 9 25.32 4.17 2.88 728.7 316.0 10 25.34 4.10 2.93 815.6 347.5 11 25.37 4.16 2.96 801.7 329.9 12 25.34 4.19 2.91 796.8 336.8 13 25.40 4.14 3.02 900.5 357.7 14 25.40 4.19 2.87 835.9 363.3 15 25.44 4.08 2.89 800.4 352.3 16 25.36 4.18 2.97 904.9 368.1 17 25.39 4.14 2.73 690.9 335.8 18 25.43 4.13 2.91 860.7 369.1 19 25.40 4.12 2.89 846.2 368.8 20 25.41 4.15 2.72 731.6 357.4 Mean 346.6 Std dev 19.6 Appendices - 186 - Table G. 10: G20 cBN + 25% Al Sample length (l) width (b) thickness (d) F (load) (N) Flexural strength (?f) [MPa] 1 25.38 4.17 2.78 824.7 383.8 2 25.34 4.17 3.04 978.9 381.0 3 25.33 4.19 3.07 1036.02 393.5 4 25.36 4.17 3.06 1078.8 414.4 5 25.35 4.18 2.81 838.5 381.0 6 25.33 4.23 2.97 846.9 340.4 7 25.47 4.19 2.80 914.7 417.6 8 25.31 4.17 2.99 935.5 376.4 9 25.21 4.17 3.00 854 341.3 10 25.34 4.18 2.92 956.4 402.5 11 25.35 4.17 2.90 920.02 393.5 12 25.35 4.16 2.92 915.7 387.2 13 25.30 4.19 2.94 890.9 368.9 14 25.27 4.17 2.94 938.6 390.6 15 25.39 4.13 2.93 842.3 356.3 16 25.23 4.18 3.03 850.04 332.2 17 25.25 4.16 3.00 943.5 377.9 18 25.34 4.19 2.94 925.9 383.4 19 25.23 4.19 2.85 878.7 387.2 Mean 379.4 Std dev 23.2 Appendices - 187 - Appendix H: Results of the Weibull Statistics The Tables H 1-10 show the results from the Weibull-statistical analysis of the flexural strength data of the PcBN-Al composite materials, which use the ?Maximum Likelihood? method. The results show the characteristic strength values (?O) and the Weibull modulus (m) for the various confidence intervals. The average strength (?), Weibull modulus (m) and standard deviation (s) are also shown as a comparison. Weibull Statistics of the PcBN-Al Composite Materials. Table H. 1: G 2 cBN + 15 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 390.1 9.77 2 % 367.7 5.99 5 % 372.6 6.57 10 % 376.8 7.25 25 % 383.6 8.15 75 % 397.7 10.44 90 % 404.5 11.71 95 % 408.6 12.47 98 % 413.9 13.31 m average 9.08 ? 371.7 Standard deviation s 40.81 Table H. 2: G 2 cBN + 20 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 432.7 6.97 2 % 400.4 4.42 5 % 407.4 4.82 10 % 413.3 5.29 25 % 423.2 5.89 75 % 443.6 7.44 90 % 453.8 8.3 95 % 459.8 8.81 98 % 467.6 9.4 m average 6.53 ? 406.3 Standard deviation s 61.08 Table H. 3: G 2 cBN + 25 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 444.5 10.03 2 % 422.6 6.78 5 % 427.4 7.4 10 % 431.4 8.12 25 % 438.1 9.05 75 % 451.9 11.43 90 % 458.5 12.75 95 % 462.4 13.54 98 % 467.5 14.41 m average 10.03 ? 425.8 Standard deviation s 41.9 Appendices - 188 - Table H. 4: G 6 cBN + 15 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 409.4 6.14 2 % 374.8 3.89 5 % 382.3 4.25 10 % 388.6 4.66 25 % 399.2 5.19 75 % 421.3 6.56 90 % 432.1 7.31 95 % 438.5 7.76 98 % 447.0 8.27 m average 5.75 ? 379.8 Standard deviation s 71.7 Table H. 5: G 6 cBN + 20 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 394.5 15.67 2 % 380.6 9.77 5 % 383.7 10.7 10 % 386.3 11.76 25 % 390.5 13.16 75 % 399.1 16.73 90 % 403.2 18.72 95 % 405.6 19.9 98 % 408.8 21.21 m average 14.62 ? 381.7 Standard deviation s 30.1 Table H. 6: G 6 cBN + 25 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 429.1 16.2 2 % 414.9 10.26 5 % 418.0 11.2 10 % 420.7 12.29 25 % 424.9 13.7 75 % 433.7 17.3 90 % 437.9 19.3 95 % 440.4 20.5 98 % 443.6 21.8 m average 15.2 ? 417.7 Standard deviation s 23.04 Table H. 7: G 10 cBN + 15 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 453.8 23.97 2 % 443.6 15.18 5 % 445.8 16.56 10 % 447.7 18.17 25 % 450.8 20.25 75 % 457.1 25.58 90 % 460.1 28.53 95 % 461.8 30.29 98 % 464.1 32.25 m average 22.44 ? 443.9 Standard deviation s 22.7 Appendices - 189 - Table H. 8: G 20 cBN + 15 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 395.5 16.6 2 % 382.8 10.51 5 % 385.6 11.47 10 % 387.9 12.58 25 % 391.8 14.03 75 % 399.7 17.71 90 % 403.5 19.77 95 % 405.7 20.98 98 % 408.6 22.33 m average 15.54 ? 383.9 Standard deviation s 25.06 Table H. 9: G 20 cBN + 20 % Al ?u C. I. Limits ?c M Statistical Characteristic (63%) 354.8 24.88 2 % 347.2 15.76 5 % 348.9 17.2 10 % 350.3 18.87 25 % 352.6 21.03 75 % 357.3 26.56 90 % 359.5 29.62 95 % 360.9 31.45 98 % 362.6 33.48 m average 23.3 ? 346.7 Standard deviation s 19.67 Table H. 10: G 20 cBN + 25 % Al ?u C. I. Limits ?c m Statistical Characteristic (63%) 389.7 19.88 2 % 378.9 12.39 5 % 381.3 13.56 10 % 383.3 14.92 25 % 386.6 16.7 75 % 393.2 21.22 90 % 396.4 23.74 95 % 398.3 25.24 98 % 400.8 26.91 m average 18.5 ? 379.5 Standard deviation s 23.28 Appendices - 190 - Appendix I: Flaw Size The flaw size of the various materials can be estimated using the Griffiths equation (equation 3.1 in Chapter 3) relating the strength and fracture toughness of the material to the flaw size of failure. Based on the fracture toughness, KIc, the size of a crack initiating flaw can be calculated for a definite strength value f? according to 2 ? ? ? ? ? ? ? ? ? ? Y K a f Ic ? , 2/??Y , a = radius of flaw, [Munz (2007)] This can be used to determine whether the results obtained from the strength and fracture toughness measurements are in the correct range by comparing them with the results of the actual flaw size measured by microscopy. The flaw size of the flaws in each material will be measured using optical microscopes or scanning electron microscopes to determine the accuracy of the estimated results. Table I1 shows the estimated flaw size for the PcBN-Al composites, in the range between 300 -650 ?m. Table I. 1: Estimation of the flaw size of the PcBN-Al composites. Sample Strength (MPa) Fracture toughness, KIC (MPa.m0.5) Estimated Flaw size (?m) G2 cBN + 15% Al 371.7 6.44 382 G2 cBN + 20% Al 406.3 7.08 386 G2 cBN + 25% Al 425.8 7.12 356 G6 cBN + 15% Al 379.8 7.56 505 G6 cBN + 20% Al 381.7 7.61 506 G6 cBN + 25% Al 417.7 7.04 361 G10 cBN + 15% Al 443.9 8.03 417 G20 cBN + 15% Al 383.9 7.93 543 G20 cBN + 20% Al 346.7 7.81 646 G20 cBN + 25% Al 379.5 6.95 427 The type and size of each flaw from each sample material was estimated using microscopy. Figures I 1-10 show the probability graphs for each material. Table I 2- 11 show the flaw type, strength and measured flaw size. Not all the flaws could be identified and only a few of the flaws were measured, 10-15% of the tested specimens of each batch; of a high, medium and low strength. Appendices - 191 - The type of flaw is shown: B - binder SD - Surface defect B/P - binder or pore P - Pore P & B - binder surrounding pore Figure I. 1: Strength distribution of G2 cBN + 15% Al Table I. 2: Flaw type and size for G2 cBN + 15% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 3 B 442 5 Large P 440 400 17 P 429 18 P 423 9 B 399 12 Large B 393 14 Large B 385 13 ? 371 16 ?SD 366 1 Large B 354 110 7 B 350 11 ? 347 4 Large B 344 10 Large B 342 6 Large B 340 8 Large P SD 330 15 Large P SD 324 208 2 P 314 208 Appendices - 192 - Figure I. 2: Strength distribution of G2 cBN + 20% Al Table I. 3: Flaw type and size for G2 cBN + 20% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 13 B 533 5 B 517 410 12 ? 494 10 ? 458 14 B 451 8 B 441 6 Oxide inclusion 435 421 17 B 409 18 P 393 11 B 390 4 B 389 16 B 386 19 P 381 2 ED 377 7 B 375 1 P / large grain 371 200 3 P 342 9 P 339 190 20 SD 324 100 15 B/P 318 Appendices - 193 - Figure I. 3: Strength distribution of G2 cBN + 25% Al Table I. 4: Flaw type and size for G2 cBN + 25% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 2 B 522 180 12 P 480 10 B 477 14 P 451 3 B 449 13 B 449 9 B 446 1 ED P & B 442 85 5 ? 441 7 ? 424 20 SD 417 19 Large P 416 275 4 B 414 18 P 413 15 B with small P 401 150 17 B/P 397 8 SD 386 6 B 385 16 B 360 11 ?B 345 Appendices - 194 - Figure I. 4: Strength distribution of G6 cBN + 15% Al Table I. 5: Flaw type and size for G6 cBN + 15% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 8 B 482 15 B 453 5 SD B 442 263 10 ? 431 17 B 426 14 B 423 487 2 B 423 11 B 422 20 B 419 13 B 418 12 P 417 1 B 417 18 ED 410 3 B 407 7 B 403 16 B with P 402 359 19 B 393 9 SD 392 4 P 391 6 P 384 Appendices - 195 - Figure I. 5: Strength distribution of G6 cBN + 20% Al Table I. 6: Flaw type and size for G6 cBN + 20% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 11 SD Large P 426 257 19 Large B 420 4 B 419 13 SD P 418 226 7 B 410 2 B 394 8 B 385 1 P 381 17 B 380 16 B 380 20 B 375 3 P 375 500 10 B 370 15 B/P 368 6 B 368 18 B 367 5 B 363 12 B 359 14 B 295 298 Appendices - 196 - Figure I. 6: Strength distribution of G6 cBN + 25% Al Table I. 7: Flaw type and size for G6 cBN + 25% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 1 B 497 7 P 466 370 20 P 461 15 P 461 8 B 453 11 ? 450 18 B 441 2 B 426 6 P 420 148 13 B 380 10 P 337 3 P 326 19 P 325 17 B 325 4 B 323 216 9 P 319 12 SD P 304 188 5 P 300 14 P 293 162 16 P 288 Appendices - 197 - Figure I. 7: Strength distribution of G10 cBN + 15% Al Table I. 8: Flaw type and size for G10 cBN + 15% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 16 ? 481 15 ? 475 270 5 ? 470 17 ? 462 10 ? 459 19 SD 458 310 8 B 454 11 P 453 7 B 449 2 ? P 447 12 B 445 1 B 437 13 B 434 9 B 432 14 B 431 6 B 430 18 P 428 3 B 428 417 20 B 425 4 B with small P 378 237 Appendices - 198 - Figure I. 8: Strength distribution of G20 cBN + 15% Al Table I. 9: Flaw type and size for G20 cBN + 15% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 6 ? 431 18 B 420 312 16 ED P 419 8 ? 419 7 P 399 19 B 391 2 B 389 236 20 SD 388 12 P 387 5 B 385 325 4 B 384 203 10 B 379 14 P 377 13 B 375 9 B 367 1 ? 367 17 P 364 163 15 P 357 3 ? 342 11 ?B 338 265 Appendices - 199 - Figure I. 9: Strength distribution of G20 cBN + 20% Al Table I. 10: Flaw type and size for G20 cBN + 20% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 18 ED P 369 19 B 369 16 SD P 368 365 2 B 366 14 ? 363 13 ? 358 20 Large B 357 290 5 P 355 3 B 353 15 ? 352 1 B 349 10 B 348 7 B 347 12 ? 337 17 SD P 336 237 8 B 336 4 ED B with P 334 212 11 ? 330 9 B 316 6 B 291 Appendices - 200 - Figure I. 10: Strength distribution of G20 cBN + 25% Al Table I. 11: Flaw type and size for G20 cBN + 25% Al Sample No: Flaw type Strength (MPa) Flaw Size (?m) 7 B 418 4 P 414 208 10 P 403 3 B @ edge 394 246 11 B 394 14 ? P 391 19 ? P 387 12 B 387 1 B 384 18 B 383 5 P 381 2 P 381 17 P 378 8 P 376 13 ?P 369 15 B 356 9 ? 341 6 B 340 16 Large P/B in middle 332 126 Appendices - 201 - Appendix J: Statistical Analysis of Properties Hardness: General Linear Model: Hardness versus Grain size, Al-content Factor Type Levels Values Grain size fixed 4 G02, G06, G10, G20 Al-content fixed 3 15, 20, 25 Analysis of Variance for Hardness, using Adjusted SS for Tests Source DF Seq SS Adj SS Adj MS F P Grain size 3 3502.48 3075.66 1025.22 275.50 0.000 Al-content 2 1260.02 1260.02 630.01 169.30 0.000 Error 74 275.37 275.37 3.72 Total 79 5037.88 S = 1.92906 R-Sq = 94.53% R-Sq(adj) = 94.16% Test for Equal Variances: Hardness_1 versus Al-content One-way ANOVA: Hardness_1 versus Grain Size_1 Source DF SS MS F P Grain Size_1 9 4887.61 543.07 252.98 0.000 Error 70 150.27 2.15 Total 79 5037.88 S = 1.465 R-Sq = 97.02% R-Sq(adj) = 96.63% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ---------+---------+---------+---------+ G02-15 8 40.740 1.195 (*-) G02-20 8 33.535 1.151 (-*) G02-25 8 32.060 1.648 (-*) G06-15 8 36.058 1.385 (-*) G06-20 8 27.198 0.380 (-*) G06-25 8 22.003 1.473 (*-) G10-15 8 35.840 2.514 (*-) G20-15 8 23.239 1.743 (*-) G20-20 8 19.828 1.505 (*-) G20-25 8 15.638 0.378 (*-) ---------+---------+---------+---------+ 21.0 28.0 35.0 42.0 Pooled StDev = 1.465 This shows that there are significant differences between the grain size and binder content of the various materials and their effect on the hardness. The grain size has a larger effect on the hardness than the binder content. Appendices - 202 - Fracture Toughness: General Linear Model: K1c versus Grain size, Al-content Factor Type Levels Values Grain size fixed 4 G02, G06, G10, G20 Al-content fixed 3 15, 20, 25 Analysis of Variance for K1c, using Adjusted SS for Tests Source DF Seq SS Adj SS Adj MS F P Grain size 3 5.6544 5.7495 1.9165 5.96 0.002 Al-content 2 1.7688 1.7688 0.8844 2.75 0.076 Error 41 13.1863 13.1863 0.3216 Total 46 20.6095 S = 0.567112 R-Sq = 36.02% R-Sq(adj) = 28.22% Unusual Observations for K1c One-way ANOVA: K1c versus Grain size Source DF SS MS F P Grain size 3 5.654 1.885 5.42 0.003 Error 43 14.955 0.348 Total 46 20.610 S = 0.5897 R-Sq = 27.44% R-Sq(adj) = 22.37% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev --------+---------+---------+---------+- G02 14 6.9092 0.4646 (-----*------) G06 14 7.2376 0.8238 (------*-----) G10 5 8.0331 0.0480 (----------*---------) G20 14 7.5364 0.5052 (------*-----) --------+---------+---------+---------+- 7.00 7.50 8.00 8.50 Pooled StDev = 0.5897 This shows that there is only a significant difference in the fracture toughness of the G2 and G10 and G20 materials. One-way ANOVA: K1c versus Al-content Source DF SS MS F P Al-content 2 1.674 0.837 1.94 0.155 Error 44 18.936 0.430 Total 46 20.610 S = 0.6560 R-Sq = 8.12% R-Sq(adj) = 3.94% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -------+---------+---------+---------+-- 15 18 7.3754 0.9508 (----------*---------) 20 15 7.4991 0.4715 (----------*----------) Appendices - 203 - 25 14 7.0347 0.1870 (----------*-----------) -------+---------+---------+---------+-- 6.90 7.20 7.50 7.80 Pooled StDev = 0.6560 This shows that there is very little significant difference between the fracture toughness for the 15, 20 and 25% materials, they all overlap. Strength: General Linear Model: Flexural Strength versus Grain Size_1, Al-content Factor Type Levels Values Grain Size_1 fixed 4 G02, G06, G10, G20 Al-content fixed 3 15, 20, 25 Analysis of Variance for Flexural Strength, using Adjusted SS for Tests Source DF Seq SS Adj SS Adj MS F P Grain Size_1 3 88126 94502 31501 18.50 0.000 Al-content 2 33645 33645 16822 9.88 0.000 Error 191 325302 325302 1703 Total 196 447072 S = 41.2692 R-Sq = 27.24% R-Sq(adj) = 25.33% Unusual Observations for Flexural Strength One-way ANOVA: Flexural Strength versus Grain size Source DF SS MS F P Grain size 9 151284 16809 10.63 0.000 Error 187 295788 1582 Total 196 447072 S = 39.77 R-Sq = 33.84% R-Sq(adj) = 30.65% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ------+---------+---------+---------+--- G02-15 18 371.70 40.81 (----*----) G02-20 20 406.25 61.08 (----*----) G02-25 20 425.77 41.90 (----*----) G06-15 20 379.81 71.76 (----*----) G06-20 20 381.10 29.42 (----*----) G06-25 20 417.71 23.04 (----*----) G10-15 20 443.87 22.74 (----*----) G20-15 20 383.92 25.06 (----*----) G20-20 20 346.69 19.67 (----*----) G20-25 19 379.47 23.28 (----*-----) ------+---------+---------+---------+--- 350 385 420 455 Pooled StDev = 39.77 Appendices - 204 - One-way ANOVA: Flexural Strength versus Al-content Source DF SS MS F P Al-content 2 27268 13634 6.30 0.002 Error 194 419804 2164 Total 196 447072 S = 46.52 R-Sq = 6.10% R-Sq(adj) = 5.13% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ------+---------+---------+---------+--- 15 78 395.42 52.53 (------*------) 20 60 378.01 47.02 (-------*-------) 25 59 408.13 36.40 (-------*-------) ------+---------+---------+---------+--- 375 390 405 420 Pooled StDev = 46.52 This shows that there is very little significant difference between the strength for the 15, 20 and 25% materials, they all overlap. One-way ANOVA: Flexural Strength versus Grain Size_1 Source DF SS MS F P Grain Size_1 3 88126 29375 15.79 0.000 Error 193 358946 1860 Total 196 447072 S = 43.13 R-Sq = 19.71% R-Sq(adj) = 18.46% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev +---------+---------+---------+--------- G02 58 402.26 53.10 (---*---) G06 60 392.87 49.21 (---*---) G10 20 443.87 22.74 (-----*-----) G20 59 369.87 28.01 (--*---) +---------+---------+---------+--------- 360 390 420 450 Pooled StDev = 43.13 This shows that the strength of the G10 and G20 are significantly different, while very little difference between G2 and G6. Calculation of the % Contribution: Grain Size and Binder content contributions to Hardness: Appendices - 205 - %47.501.2552.69100% %01.25100 88.5037 02.1260 % %52.69100 88.5037 48.3502 _% ???? ??? ??? error Binder SizeGrain R-Sq = 94.53% This suggests that the grain size and binder content have a good fit to the general linear model for Hardness. The grain size has a larger contribution to the hardness than the binder content. Grain Size and Binder content contributions to Fracture Toughness: %98.6358.843.27100% %58.8100 6095.20 7688.1 % %43.27100 6095.20 6544.5 _% ???? ??? ??? error Binder SizeGrain R-Sq = 36.02% This suggests that the cBN grain size and binder content have a very poor fit to the general linear model of the Fracture toughness results. Grain Size and Binder content contributions to Strength: %76.7252.771.19100% %52.7100 447072 33645 % %71.19100 447072 88126 _% ???? ??? ??? error Binder SizeGrain R-Sq = 27.24% This suggests that the cBN grain size and binder content have a very poor fit to the general linear model of the Strength results.