The Pennsylvania State University. The Graduate School. College of Engineering

Transcription

1 The Pennsylvania State University The Graduate School College of Engineering A STUDY OF MACHINABILITY IN MILLING OF AUSTEMPERED DUCTILE IRON (ADI) ADI 900, ADI 1050, AND ADI 1200 WITH CARBIDE TOOLS A Thesis in Industrial Engineering by Abibu Sinlah Jr Abibu Sinlah Jr. Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science May 2014

2 The thesis of Abibu Sinlah Jr. was reviewed and approved* by the following: Christopher Saldana Harold and Inge Marcus Assistant Professor Thesis Adviser Robert Voigt Professor of Industrial Engineering Thesis Reviewer Paul Griffin Professor of Industrial Engineering Head of the Department of Industrial Engineering *Signatures are on file in Graduate School ii

3 Abstract Austempered Ductile Iron (ADI) is a relatively new and high strength material produced from Ductile Iron (DI) by means of a unique heat treatment, austempering. The heat treatment produces a uniform microstructure of acicular ferrite and carbon-stabilized austenite, known as Ausferrite, with graphite nodules. ADI exhibits grade material properties including a high strength-weight ratio, good ductility and toughness, high fatigue strength, and wear resistance. It is also approximately 10% less dense than steel with similar strength levels and three times stronger than aluminum at only 2.5 times the mass. These material properties make ADI an attractive alternative for applications that require light weight flexible designs, where density matter and the maintaining of high strength and toughness are important. However, the difficulties experienced when machining ADI remain one of the major restricting factors in the growth of the market for ADI. In many cases, because of the high strength and hardness values, manufacturers have deemed ADI un-machinable. The main objective of this study was to assess the machinability of ADI grades 900, 1050, and 1200 (GR900, GR1050, GR1200) in order to establish cost-effective processing conditions for the machining of ADI. This was accomplished by conducting face and end milling studies for a range of machining conditions. More specifically, the effects of cutting speed on tool life and surface roughness in face milling were investigated as well as the effects of chip load on cutting force and cutting stiffness during end milling. The face milling operations were carried out on 25 mm thick plates using the following cutting conditions with a three tooth, 80 mm face milling tool: constant chip load of 0.08 mm/tooth and depth of cut of 1 mm, with varying cutting speed of m/min, m/min, and m/min for GR1200, GR1050, and GR900, respectively. A Taylor tool life model was also developed, using the five cutting speeds, for the GR900, GR1050, and GR1200. In addition, end milling operations were carried out on 25 mm thick plates using the following cutting conditions with a single tooth 19 mm end milling tool: constant cutting speed of 15 m/min and depth of cut of 1 mm, with varying chip load of mm/tooth. In addition to the analysis of machinability of the grades of ADI, the machinability was compared to that of AISI/SAE 4340 steel (S-4340), under similar cutting conditions. The cutting stiffness of the grades of ADI were found to be generally in the expected cutting stiffness range for hardened steels, such as S-4340, during conventional milling. Empirical testing also showed that high cutting speeds accelerate the rate of tool wear, decreasing tool life and increasing surface roughness when machining ADI. In general, GR900 exhibited the best tool life, surface finish, and iii

4 cutting stiffness with higher machinability than the higher strength grades of ADI. Furthermore, Taylor tool life models were developed that represented 95%, for GR1200 and GR1050, and 85%, for GR900, of the tool life data collected. iv

5 CONTENTS LIST OF FIGURES... VII LIST OF TABLES... XII ACKNOWLEDGMENTS... XVI 1 INTRODUCTION PROBLEM STATEMENT ORGANIZATION OF THE THESIS MACHINING MACHINING MECHANICS CHIP FORMATION IN MACHINING MILLING Mechanics in milling MACHINABILITY Chip formation Forces Tool life Surface finish AUSTEMPERED DUCTILE IRON (ADI) DEVELOPMENT OF ADI PRODUCTION OF ADI MACHINING OF ADI SURROGATE MATERIALS EXPERIMENTAL WORK MATERIAL CHARACTERIZATION EXPERIMENTAL PLATFORM DYNAMOMETER CALIBRATION MACHINABILITY METRICS RESULTS CHIP FORMATION CUTTING FORCES TOOL LIFE SURFACE ROUGHNESS DISCUSSION v

6 6.1 MACHINABILITY EVALUATION SURROGATE MATERIAL COMPARISON CONCLUSION AND FUTURE WORK REFERENCES APPENDIX A. ADI CUTTING FORCE DATA SUMMERY FIGURES APPENDIX B. TOOL WEAR MEASUREMENTS APPENDIX C. TOOL LIFE MEASUREMENTS APPENDIX D. SURFACE ROUGHNESS MEASUREMENTS APPENDIX E. INSERT IDENTIFICATION CHARTS vi

7 List of Figures Figure 2.1: Orthogonal and oblique cutting, (a) orthogonal cutting; (b) oblique cutting [1] Figure 2.2: Terms used in metal cutting [1] Figure 2.3: Cutting force diagram - Merchant circle [1] Figure 2.4: Discontinuous chip formation [1] Figure 2.5: Continuous chip formation [1] Figure 2.6: Continuous chip with BUE formation [1] Figure 2.7: Types of milling operations [6] Figure 2.8: Milling configurations, (a) up/conventional and (b) down/climb milling, adapted from [6]. 22 Figure 2.9: Chip thickness variation for up and down milling, where a = r [8] Figure 2.10: Exit and start angle geometry for up and down milling [8] Figure 2.11: Cutting force geometry for tangential and normal force [8] Figure 2.12: Regions of tool wear in metal cutting [1] Figure 2.13: Features of single-point-tool wear in turning operations. (after International Standards Organisation, ISO 3635, ISO Geneva, 1993) [1] Figure 2.14: Definition of parameters used to compute the average roughness Ra [6] Figure 2.15: The shape, waviness, and roughness of a surface (after Whitstone) [6] Figure 3.1: ADI market distribution - North America 2008 [48] Figure 3.2: ADI heat treatment process [4] Figure 4.1: Results for XRD analysis at various grades of ADI, (a) GR900, (b) GR1050, and (c) GR1200 [49] Figure 4.2: Tool-holder used in face milling experiments [50] Figure 4.3: Post preparation process workpiece for face milling test Figure 4.4: Dynamometer calibration setup Figure 4.5: Results for dynamometer calibration Figure 4.6: Experimental flow diagram Figure 4.7: Face milling experimental setup Figure 4.8: Wear measurement setup with stereoscope and insert fixture Figure 4.9: End milling insert geometry [51] Figure 4.10: End milling experimental setup with dynamometer and data acquisition system Figure 4.11: Machine center and dynamometer alignment diagram vii

8 Figure 5.1: Effect of cutting speed on chip form for ADI grades and S-4340 (a = 1 mm, f = 0.08 mm/tooth) Figure 5.2: Results for collected cutting forces, Fx and Fy, and calculated resultant cutting force F for the grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, (d) S-4340 (a = 1 mm, f = 0.15 mm/tooth, SS = rpm, and 20% radial emersion) Figure 5.3: Results for cutting force calculations for GR1200 cut at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure 5.4: Close-up of results for cutting force calculations for GR1200 cut at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure 5.5: Effect of chip load on resultant cutting forces for grades of ADI and S Figure 5.6: Effect of chip load on cutting stiffness for all grades of ADI and S Figure 5.7: Approximately constant portion of effect of chip load on the cutting stiffness at 1 mm Axial Depth of Cut 20% Radial Cut for grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, and (d) S Figure 5.8: Approximately constant portion of the effect of chip load on the beta angle at 1 mm Axial Depth of Cut 20% Radial Cut for grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, and (d) S Figure 5.9: Example of tool wear development for ADI GR1050 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth), at cutting lengths (a) 0 mm, (b) 1829 mm, (c) 2743 mm, and (d) 3658 mm, 63x magnification Figure 5.10: Example of tool wear development for ADI GR1050 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth), at cutting lengths (a) 0 mm, (b) 610 mm, (c) 914 mm, and (d) 1219 mm, 63x magnification Figure 5.11: Tool wear progressions for GR1200 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.12: Tool wear progressions for GR1050 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.13: Tool wear progressions for GR900 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.14: Tool wear progressions for S-4340 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) viii

9 Figure 5.15: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 240 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.16: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 360 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.17: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 360 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.18: Effect of cutting speed on tool life of for grades of ADI (GR1200, GR1050, GR900), a = 1 mm, f = 0.08 mm/tooth, (a) tool life in terms of cutting time, (b) tool life in terms of cutting length Figure 5.19: Log-log tool life plot used to develop the Taylor tool life equation Figure 5.20: Effect of cutting length on surface roughness (Ra) for ADI GR900 at various cutting speeds, (a) 240 m/min, (b) 300 m/min, (c) 360 m/min, (d) 420 m/min, and (e) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Figure 5.21: Effect of cutting length on surface roughness (Ra) for ADI GR1050 at various cutting speeds, (a) 240 m/min, (b) 300 m/min, (c) 360 m/min, (d) 420 m/min, and (e) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Figure 5.22: Effect of cutting length on surface roughness (Ra) for ADI GR1200 at various cutting speeds, (a) 120 m/min, (b) 180 m/min, (c) 240 m/min, (d) 300 m/min, and (e) 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Figure 5.23: Effect of cutting length on surface roughness (Ra) for ADI GR1050 at various cutting speeds, (a) 240 m/min, (b) 360 m/min, and (c) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Figure 5.24: Effect of cutting speed [m/min] on surface roughness (Ra) for all grades of ADI (GR1200, GR1050, GR900) and S-4340, a = 1 mm, f = 0.08 mm/tooth Figure 6.1: Acceptable range of cutting stiffness (specific cutting force) per material category [52] Figure A.1: Results for cutting force calculations for GR900 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.2: Results for cutting force calculations for GR900 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time ix

10 Figure A.3: Results for cutting force calculations for GR900 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.4: Results for cutting force calculations for GR900 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.5: Results for cutting force calculations for GR900 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.6: Results for cutting force calculations for GR1050 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.7: Results for cutting force calculations for GR1050 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.8: Results for cutting force calculations for GR1050 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.9: Results for cutting force calculations for GR1050 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.10: Results for cutting force calculations for GR1050 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.11: Results for cutting force calculations for GR1200 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time x

11 Figure A.12: Results for cutting force calculations for GR1200 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.13: Results for cutting force calculations for GR1200 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.14: Results for cutting force calculations for GR1200 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.15: Results for cutting force calculations for S-4340 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.16: Results for cutting force calculations for S-4340 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.17: Results for cutting force calculations for S-4340 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.18: Results for cutting force calculations for S-4340 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time Figure A.19: Results for cutting force calculations for S-4340 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time xi

12 List of Tables Table 2.1: Advantages and disadvantages of conventional and climb milling [7] Table 3.1: Comparison of ISO, ASTM, SAE, and China standard grades for ADI [29] Table 3.2: Comparison of the Brinell (HBW) hardness ranges for the various ADI grades [18] Table 4.1: Chemical composition of pre-heat treat Ductile Iron Table 4.2: Ferrite and austenite volume fractions Table 4.3: Austenite lattice parameter results Table 4.4: Results for bulk hardness of grades of ADI, Brinell hardness Table 4.5: Dynamometer calibration results Table 4.6: Face milling study machining conditions for ADI grades Table 4.7: Face milling study machining conditions for S Table 4.8: End milling study machining conditions for ADI grades and S Table 5.1: Cutting force measurements in the X direction of dynamometer for grades of ADI (GR1200, GR1050, and GR900) Table 5.2: Cutting force measurements in the Y direction of dynamometer for grades of ADI (GR1200, GR1050, and GR900) Table 5.3: Cutting force measurements in the X and Y direction of dynamometer for S Table 5.4: Effect of chip load on resultant cutting force for grades of ADI (a = 1 mm, V = 15 m/min). 78 Table 5.5: Effect of chip load on cutting stiffness for grades of ADI (a = 1 mm, V = 15 m/min) Table 5.6: Effect of chip load on resultant cutting force and cutting stiffness for S Table 5.7: Tool wear progression cubic polynomial fit equations for work materials Table 5.8: Effect of cutting speed on tool life for ADI grades in terms of cutting time and cutting length (a = 1 mm, f = 0.08 mm/tooth) Table 5.9: Effect of cutting speed on tool life in terms of cutting time for S-4340 (a = 1 mm, f = 0.08 mm/tooth) Table 5.10: Average surface roughness (Ra) for grades of ADI (GR1200, GR1050, GR900), a = 1 mm, f = 0.08 mm/tooth Table 5.11: Average surface roughness (Ra) for S-4340 (a = 1 mm, f = 0.08 mm/tooth) Table 6.1: Estimated cutting speeds for desired tool life of 10, 30, and 60 minutes for present study (a = 1 mm, f = 0.08 mm/tooth, 63% immersion) xii

13 Table 6.2: Estimated cutting speeds for desired tool life of 10, 30, and 60 minutes for modified cutting conditions (e.g. depth of cut, immersion and chip load) with constant material removal rate Table B.1: Wear measurements for ADI GR900 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth). 113 Table B.2: Wear measurements for ADI GR900 at 420 m/min (a = 1 mm, f = 0.08 mm/tooth). 114 Table B.3: Wear measurements for ADI GR900 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth). 115 Table B.4: Wear measurements for ADI GR900 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth). 116 Table B.5: Wear measurements for ADI GR900 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth). 117 Table B.6: Wear measurements for ADI GR1050 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth) 117 Table B.7: Wear measurements for ADI GR1050 at 420 m/min (a = 1 mm, f = 0.08 mm/tooth) 118 Table B.8: Wear measurements for ADI GR1050 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) 119 Table B.9: Wear measurements for ADI GR1050 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth) 120 Table B.10: Wear measurements for ADI GR1050 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.11: Wear measurements for ADI GR1200 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.12: Wear measurements for ADI GR1200 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.13: Wear measurements for ADI GR1200 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.14: Wear measurements for ADI GR1200 at 180 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.15: Wear measurements for ADI GR1200 at 120 m/min (a = 1 mm, f = 0.08 mm/tooth) Table B.16: Wear measurements for S-4340 (a = 1 mm, f = 0.08 mm/tooth) Table C.1: Effect of cutting speed on tool life for ADI grades in terms of cutting time for all trials (a = 1 mm, f = 0.08 mm/tooth) Table D.1: Average surface roughness [µm] for ADI GR900 (a = 1 mm, f = 0.08 mm/tooth) Table D.2: Average surface roughness [µm] for ADI GR1050 (a = 1 mm, f = 0.08 mm/tooth) Table D.3: Average surface roughness [µm] for ADI GR1200 (a = 1 mm, f = 0.08 mm/tooth) Table D.4: Surface roughness measurements [µm] for ADI GR900 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.5: Surface roughness measurements [µm] for ADI GR900 at 420 m/min (a = 1 mm, f = 0.08 mm/tooth) xiii

14 Table D.6: Surface roughness measurements [µm] for ADI GR900 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.7: Surface roughness measurements [µm] for ADI GR900 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.8: Surface roughness measurements [µm] for ADI GR900 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.9: Surface roughness measurements [µm] for ADI GR1050 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.10: Surface roughness measurements [µm] for ADI GR1050 at 420 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.11: Surface roughness measurements [µm] for ADI GR1050 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.12: Surface roughness measurements [µm] for ADI GR1050 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.13: Surface roughness measurements [µm] for ADI GR1050 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.14: Surface roughness measurements [µm] for ADI GR1200 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.15: Surface roughness measurements [µm] for ADI GR1200 at 300 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.16: Surface roughness measurements [µm] for ADI GR1200 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.17: Surface roughness measurements [µm] for ADI GR1200 at 180 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.18: Surface roughness measurements [µm] for ADI GR1200 at 120 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.19: Surface roughness measurements [µm] for S-4340 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.20: Surface roughness measurements [µm] for S-4340 at 360 m/min (a = 1 mm, f = 0.08 mm/tooth) Table D.21: Surface roughness measurements [µm] for S-4340 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth) Table E.1: Identification chart for inserts used to machine GR Table E.2: Identification chart for inserts used to machine GR Table E.3: Identification chart for inserts used to machine GR xiv

15 Table E.4: Identification chart for inserts used to machine S xv

16 Acknowledgments I would like to first acknowledge my utmost appreciation for my adviser, Dr. Saldana, whose patience, wisdom, and guidance over the past year have been pivotal to the execution of this project. I would also like to thank and recognize the contributions of labmate Cesar Moreno, and support specialists Dan Supko and Travis Richner to the design and fabrication of various components, as well as calculations, used in this study. My gratitude also goes out to Dr. Voigt for his extremely valuable input resulting from the reviewing process. This work was supported by Applied Process, Quaker Chemical Corp., and SECO Tools. A special thanks to Kathy Hayryen from Applied Process for her expertise and guidance. I also want to thank my parents, Abibu Sr. and Ayeshata Sinlah, and the rest of my family, who have supported all my endeavors. Finally, I would like to extend my deepest thanks to the love of my life, Janee Minor, for her unwavering faith and support over the last several years. xvi

17 1 Introduction Machining continues to play a pivotal role in the production of performance components in many industries. As developmental material systems evolve to meet more stringent performance standards, the balance between higher processing costs and systems performance gains becomes harder to maintain. This gives rise to need for research on methods to effectively and efficiently process materials. The term machinability refers to the relative ease or difficulty a given material can be processed by subtractive mechanical methods. Machinability is an aggregate measure used to evaluate a candidate material system based on effects that the material has on machining time, tool cost, power consumption, and surface quality. The adoption of developmental materials in system designs is highly dependent on material machining cost and the availability of machinability information. This is a particular challenge for Austempered Ductile Iron (ADI), which is a high specific strength iron material produced from a family of ductile irons through an innovative heat treatment process. ADI microstructure consists of an ausferrite microstructure that includes ferrite needles within an austenite background with spheroidal graphite nodules. The unique microstructure promotes tensile strength in the range of MPa at a density of approximately 7.1 g/cm 3. Due to the high strength to weight ratio, ADI is a promising replacement for steel and aluminum in many applications where its use in lieu of conventional cast irons results in substantial weight savings. In automotive applications, these weight savings can be transferred to consumers in the form of lighter vehicles and better fuel efficiency, which reduces CO2 emissions. ADI was first developed in the 1960 s and commercialized in the 1970 s. A number of investigations since that time have explored the machinability of its various grades for a limited set of machining conditions. Generally, the machining of ADI is attributed with rapid tool wear in comparison to conventional ductile irons primarily due to its high hardness. Further, machinability also is accompanied by strain-induced phase transformations that occur during machining, wherein the retained austenite in the ADI microstructure transforms to hard and difficult-to-machine martensite. Even though numerous studies have evaluated ADI material properties and address the machinability of various ADI grades in turning, little work can be found on the machinability of ADI during milling operations. Hence, manufacturers are apprehensive when it comes to using ADI in systems applications, especially when milling would be required. The purpose of the present study is to develop effective milling guidelines for ADI grades compared to steels at similar hardness level. 1

18 1.1 Problem statement The machinability of ADI in milling configurations is not well understood and has limited the use of ADI in new system designs. The present study was undertaken to investigate machinability of several ADI grades (GR900, GR1050, GR1200) with respect to various performance metrics in a milling configuration. Several methods were used to measure the effects of machining parameters on tooling performance in milling through a series of experiments measuring cutting forces, tool life, and surface roughness during processing. More specifically, the effects of cutting speed on tool life and surface roughness in face milling were investigated, as well as the effects of chip load on cutting force and cutting stiffness during end milling. 1.2 Organization of the thesis The report details the investigation carried out on the study of the machinability of ADI. The report is divided into seven chapters. Chapter 1 includes introductory remarks on the current study and chapter 2 describes relevant machining concepts and characteristics of the material systems used in the research. These machining concepts include orthogonal and milling machining mechanics as well as discussion of various machinability metrics. The evolution, production, properties and machining characteristic of ADI, from literature, are discussed in chapter 3. Chapter 4 explains the experimental design and chapter 5 presents the results and the analyzed data. Chapter 6 includes the discussion of the results. Chapter 7 is comprised of the conclusions drawn from the research, a summarization of the whole research, and recommendations for future work. 2

19 2 Machining Machining is a material removal process used for rough and final shaping of raw metallic forms into functional components. In machining, a wedge-shaped tool is moved relative to a workpiece such that a layer of material is removed in the form of a chip at its cutting edge, shown in Figure 2.1 [1]. The cutting edge is formed by the intersection of two surfaces, the tool face (or rake face) and the tool flank (or flank face). The rake face is the surface along which the chip flows, and the flank face is the surface ground at an angle to clear the new or machined workpiece surface [1]. In orthogonal cutting, Figure 2.1a, the cutting edge is orthogonal to the cutting direction, idealizing the problem with deformation occurring in two dimensions. This simplifies modeling treatment of the machining problem and a number of analytical solutions of the deformation is widely discussed in the literature. Oblique cutting is the most general case of the machining configuration, Figure 2.1b, the deformation that occurs in this case varies in three-dimensions. 2.1 Machining mechanics In cutting, chip formation occurs by the physical interaction of a tool and a workpiece, shown in Figure 2.2. The tool is engaged in the workpiece at an undeformed chip thickness of b and the tool is moved relative to the work at a velocity Vc. The tool is inclined with respect to the vertical at a rake angle γ ne and is relieved from the free surface of the workpiece at a relief angle α ne. The chip forms by a process of concentrated shear occurring in a narrow zone, noted as the shear band. The shear band is inclined with respect to the cutting direction at a shear angle φ. The deformation that occurs during the chip formation process causes the chip to thicken to a deformed chip thickness h. The strain imposed in this process can be estimated using Equation 2.1 [2] γ = λ tan(γ cos(γ n e ) λ cos(γ n e ) n e ) [Eq. 2.1] where λ, the chip thickness ratio (h/b), is typically greater than 1. The strain rate can be estimated using Equation 2.2 [3] γ = v s (y s ) max [Eq. 2.2] 3

20 where (y s ) max is the estimate of the maximum thickness of the shear zone (typically μm) and v s is the shear velocity vector. Note that the shear velocity vector is the vector sum of the cutting velocity and the chip velocity. The temperature rise during machining, assuming that all of the plastic work at the shear plane is entirely converted into heat (adiabatic temperature), can be estimated using Equation 2.3 [3] T ad = U ρc [Eq. 2.3] where U is the specific cutting energy, ρ is the density of the workpiece, and c is the specific heat of the workpiece. Note that significant temperature rise occur at large plastic strains and very high strain rates, even though the cutting is normally carried out at ambient temperature. Also, at high strain rates almost all the plastic work is converted into heat due to the lack of time needed for the heat to dissipate. The resulting high temperatures creates a secondary deformation zone. During machining, cutting forces initiate the process of chip formation and control the flow of the chip and thermal gradient involved in machining through plastic deformation and friction [4]. In orthogonal cutting, two resultant forces, the force between the tool face and the chip and the force between the workpiece and the chip, are in equilibrium [5]. Figure 2.3 depicts theoretical forces derived from the first complete analysis for predicting the shear angle presented by Ernst and Merchant [1]. Their analysis showed that the resultant cutting force acting at the tool cutting edge could be resolved into two components, known as the instantaneous feed component Fx and the instantaneous normal component Fy, which can be measured by experimental means. The instantaneous feed component is characterized as the force in the direction of cutting and the instantaneous normal component is normal to the direction of the cutting. These orthogonal cutting forces are especially useful for researchers because they can be found experimentally and used to calculate the resultant cutting force by Equation 2.4 F = (F x ) 2 + (F y ) 2 [Eq. 2.4] The resultant cutting force can also be resolved into components Fn and Ff, which occur normal to and along the tool face, respectively, and into components Fns and Fs, which occur normal to and along the shear plane, respectively. The trigonometric relations between the measureable components (Fx and Fy) and the other components of the resultant force during orthogonal machining provide an empirical 4

21 means of computing the major forces in the primary and secondary deformation zones. These forces include the force required to shear the workpiece and form the chip Fs in the primary deformations zone and the force between the tool and the chip that resist the flow of the chip along the rake face of the tool in the secondary deformation zone Ff. The trigonometric relationships are given by Equation 2.5 and 2.6. F s = (F x cos ϕ) (F y sin ϕ) [Eq. 2.5] F f = (F x sin γ ne ) + (F y cos γ ne ) [Eq. 2.6] where ϕ is the shear angle and γ ne is the rake angle. 2.2 Chip formation in machining The type of chip formed by the orthogonal cutting of metal is dependent on the workpiece material and the cutting conditions used. In practical machining cases, the three basic types of chip formation, defined by Ernst and Merchant [1], are continuous chips, continuous chips with built-up edges (BUE), and discontinuous chips. Discontinuous chips are characterized by the occurrence of facture in the primary deformation zone, which is defined as the zone spanning from the tool cutting edge to the junction between the surface of the chip and workpiece when the chip is only partly formed. The severe straining of the material during machining causes the fracture to propagate from the cutting edge to the free surface in a cyclic manner, forming small segments of chips, as shown in Figure 2.4. The formation of discontinuous chips typically occurs when machining brittle materials. Machining of materials such as cast iron or cast brass always produce discontinuous chips but may also be produced when machining ductile materials at very low speeds and high feeds [1]. The formations of discontinuous chips do not necessarily degrade machining performance, but can characterize the processing of brittle materials. On the other hand, continuous chips, without BUE, are formed by the shearing of the work material to form a chip, which slides along the cutting tool remaining in contact with the tool face for a short distance (the contact length) before curling away, as shown in Figure 2.5 [1]. The formation of the chip also occurs in the primary deformation zone. However, in order to deform the material in this manner the forces are applied to secondary deformation (the interface between the chip and the tool), which sufficiently deforms the lower layers of the chip as it slides along the tool face [1]. Continuous chips can be the most desirable chips because in ductile materials they reduce cutting forces and result in generally 5

22 good surface finish. Furthermore, continuous chips are most common when machining ductile materials, such as mild steel, copper, and aluminum, at relatively high cutting speeds in the essentially steady-state process. Similarly, continuous chips with BUE are observed in the machining of soft, ductile metals. However, BUEs typically occur at relatively low cutting speeds. BUE tend to form at speeds where the temperature at the chip-tool interface is relatively low. Fracture may occur within the chip along a plane approximately at right angles to the shear plane, leaving behind a portion of the chip attached to the tool face, which then acts as a cutting edge, shown in Figure 2.6 [5]. Although the formation of BUE occurs at relatively low cutting speeds, the cutting speed must be high enough to cause the chip face to behave in a ductile manner. The resultant plastic flow creates a strong weld, which forms between the chip and tool known as a BUE [5]. As the cutting speed increases the size of the BUE tends to decrease. BUE results in a blunt cutting edge, which increases the cutting forces relative to continuous chip formation without BUE. A BUE tends to grow to a critical size, where it breaks off and is carried away by the underside of the chip or the workpiece surface, resulting in large continuous variation in size of BUE during machining. The main influences of BUE are an unusually high friction stress (fracture vs. sliding stress) at the tool point and increasing the effective rake angle [5]. The outward and downward nature of the growth of BUE also causes variation in the depth of cut, which greatly degrades the surface roughness of the new workpiece surface when the BUE is very large. However, as the rough particles of hard highly worked BUE passing off with the chip, which tend to increase abrasive wear, the BUE actually protects the cutting edge from wear, thus decreasing tool wear [5]. As machining speeds have been increased and more material systems have been studied, researchers have observed a fourth type of chip in special cases, macroscopically continuous chips, which consist of narrow bands of heavily deformed material alternating with large regions of relatively undeformed material [6]. The type and size of chip formation that occurs when machining a material, such as ADI, can vary based on the machining operation used (i.e. turning, milling, drilling, or boring) and cutting conditions. 2.3 Milling Milling is a material removal process that consists of a coordinated linear or multiple-axis feeding motion of a multi-edged cutter as it rotates across and into the workpiece [6, 7]. Unlike turning, which is 6

23 basically a continuous machining operation, a milling operation is an intermittent cutting action with discrete cutting events occurring as each cutting edge continuously enters and exits the workpiece. Milling can be broken down into three basic configurations: peripheral or plain milling, end milling and face milling, shown in Figure 2.7. Peripheral or plain milling (Figure 2.7c) generates a surface parallel or inclined to the axis of rotation, which produces a variable undeformed chip thickness [5]. Furthermore, the undeformed chip width, in peripheral milling operations, is measured parallel to the cutting edge. Face milling (Figure 2.7b) produces a surface normal to the axis of rotation, which generates a constant undeformed chip thickness throughout the cut [5]. In the case of face milling the undeformed chip thickness corresponds to the chip load (feed/tooth) and the undeformed chip width corresponds to the depth of cut. End milling (Figure 2.7a), which resembles face milling with a much smaller cutter, is a type of peripheral milling used for profiling and slotting operations. Much like peripheral milling, in end milling the undeformed chip thickness is variable throughout the cutting operation, but the undeformed chip width corresponds to the depth of cut [5]. End milling, as well as face milling, consists of a major cutting edge that is parallel to the cutting axis and a secondary edge that is perpendicular to the cutting axis connected by a nose radius. All three milling configurations can be divided into two milling orientations, up (or conventional) milling and down (or climb) milling, shown in Figure 2.8. The milling operation being used is determined by either fixing the feed direction of the workpiece or the spindle rotating direction (clockwise or anti-clockwise). If the rotation direction of the spindle is fixed, shown in Figure 2.8, the work feed direction will determine which cutting orientation is being used. In climb milling (Figure 2.8b) the feed direction of the workpiece is the same direction as the cutting rotation but in conventional milling (Figure 2.8a) the feed direction is opposite of the rotation. However, when the axis of the cutter intersects the workpiece, both conventional and climb milling occur at different stages of the rotation [6]. The milling configuration preferred for a process greatly depends on the operation requirements. Conventional milling is usually desirable, over climb milling, when the spindle and feed drive exhibit backlash and when the part has large variations in height or hardened outer layer due to sand casting or flame cutting [6]. Therefore, conventional milling is preferred when milling castings and forgings with very rough surfaces, while climb milling is preferred when milling heat treated alloys and stainless steel because of the accompanying reduction in work hardening of the workpiece material [7]. However, in climb milling chips have the tendency to become wedged between the insert and the cutter, causing tool breakage [6]. Table 2.1 shows some additional facts and advantages and disadvantages for the two milling orientations. 7

24 All milling operations are also configured using a combination of tool geometry and cutting conditions, which in some cases are functions of tool geometry or other cutting conditions. The cutting speed Vc (m/min) for the diameter of the tool at the cutting depth d (mm) and spindle speed SS (rpm) is given by Equation 2.7. V c = π d SS [Eq. 2.7] The linear speed or table feed rate Vf (mm/min) is the linear feed of the table, which holds the workpiece, with relation to the cutting tool. For a given feed per tooth ft (mm/tooth) and number of effective teeth (N), Vf is given by Equation 2.8. V f = f t SS N [Eq. 2.8] The feed per revolution, which corresponds to the movement of the tool in a single revolution, is given by Equation 2.9. f n = V f SS [Eq. 2.9] The material removal rate Q (mm 3 /min), which characterizes the volume of the workpiece that is removed per unit time is given by Equation 2.10 Q = b a e V f [Eq. 2.10] where ae is the work engagement or radial depth of cut. These cutting conditions play a major role in the ability to machine a material, known as machinability Mechanics in milling Unlike in turning where the chip thickness and chip width are constant, in milling chip thickness varies with time and the location of the cutting edge [8]. The time dependent chip thickness h is given by Equation 2.11 h = f t sin(ϕ) [Eq. 2.11] where ϕ is the tool s rotational angle and f t is the chip load (mm/tooth). The tool s rotational angle ϕ is zero when it is equal to 0 and 180 and is greatest when it equals 90, as shown in Figure 2.9. As the tool 8

25 rotates, the chip thickness increases during up milling and decreases in down milling, but in both cases it is zero for 180 < ϕ < 360 because no cutting occurs between these angles [8]. In up milling the start, or entry, angle (θ s ) is zero, while the exit angle (θ e ) depends on the radial depth of cut and the tool radius. On the other hand, in down milling the exit angle is 180, while the entry angle depends on the radial depth of cut and the tool radius. Figure 2.10 depict the exit and start angle geometry for up and down milling. The exit angle for up milling can be found using Equation 2.12, and the start angle for down milling can be found using Equation Both are dependent on the radial depth of cut, a, and tool radius, r. The entry and exit angles identify the occurrence of each cutting event for each tooth of the cutter. θ e = cos 1 ( (r a) ) [Eq. 2.12] r θ s = 180 cos 1 ( (r a) ) [Eq. 2.13] r The changing angle between the start and exit of the tooth characterizes the cutting event and the time dependent area of the chip. The time dependent tool angle ϕ(t) can be computed using Equation 2.14 ϕ(t) = θ e + 6ω(t t e ) [Eq. 2.14] where ω is the spindle frequency, t is the dynamic time, and te is the start time of the cutting operation. The specific force, which is analogous to the cutting stiffness, is the force observed per material removed and is a function of the cutting force and the chip area, shown in Equation 2.15 K s = F bh = F bf t sin(ϕ) [Eq. 2.15] where b is the axial depth of cut, in end milling. The specific cutting energy (SPCE), which is also analogous to the cutting stiffness, can be defined as the total energy used in removing a unit volume of work material [9]. Therefore, the cutting stiffness includes the energy expended in the primary and secondary deformations zones, interfacial friction between the tool and workpiece, and the energy needed to generate the new surface [9]. The energy is transferred from the cutting tool s rake and flank surfaces to the chip, workpiece, and heat, during the cutting operation. Thus cutting stiffness is known to 9

26 vary considerably for a given material system and is affected by changes in cutting conditions and geometries such as cutting speed, feed, tool rake, etc [1]. Researchers have also found that SPEC, or cutting stiffness, decreases as the chip thickness increases and that cutting speed has minimal effect on the energy at higher cutting speeds, which can be attributed to the increased ratio of the plowing force or size effect as the chip thickness decreases [9]. The plowing force is a function of the nose radius of the tool and the deformations of the tool in high stress conditions. Therefore, since the chip thickness is a function of chip load, the chip load can be assumed to have a similar effect on the cutting stiffness as the overall chip thickness. Thus, cutting stiffness is expected to decrease as the chip load is increased, which is also attributed to the size effect. In this case, the size effect or ratio of the plowing force refers to the ratio of the nose radius to the chip load. Furthermore, research has found that for a given tool rake at high cutting speeds and chip loads, the cutting stiffness tends to be constant [1]. When this mostly constant cutting stiffness is used as a machinability index, much like cutting forces, higher energy or cutting stiffness indicates lower machinability and vice versa [9]. Favorable cutting conditions will improve cutting energy and machinability. Therefore, the constant cutting stiffness of given material systems under certain conditions can be used to compare the machinability of the material systems. Additionally, the normal force (Fn) and tangential force (Ft) are the force components in the coordinate frame that rotates with the tool. Equations 2.16 and 2.17 show the normal and tangential forces as a function of the resultant cutting force and cutting stiffness. F t = sin(β) F = sin(β)k s bh [Eq. 2.16] F n = cos(β) F = cos(β) K s bh [Eq. 2.17] where β is the angle between the resultant force and the normal force, known as the beta angle. These forces are a function of the beta angle, which relates the position of the tool relative to the workpiece, shown in Figure In order to determine the beta angel, Equation 2.19, the angle between Fy and F, e.g., the alpha angle, must be determined, using Equation (t) = tan 1 ( F x (t) ) [Eq. 2.18] F y (t) 10

27 β(t) = 180 (ϕ(t)+ (t)) [Eq. 2.19] The beta angle is considered a property of the cutting tool geometry and workpiece, therefore, it should remain mostly constant as the cutting conditions change. 2.4 Machinability Machinability is defined as the relative ease or difficulty a given material can be processed by subtractive mechanical methods to a satisfactory surface finish and dimensional accuracy. Materials with good machinability exhibit low cutting power consumption and minimum tool wear. Many of the material characteristics of the work material (such as microstructure, hardness, tensile strength, etc.) that can be modified to improve and define the performance of a material can also influence machinability. Thus, engineers are often balancing the opposing economical drives of cheaper machining, or improved machinability, with maintaining performance. Consequently, machinability is commonly viewed as a material property that is dependent upon the chemical composition, mechanical and thermophysical properties of the work material. However, machinability is also influenced by the rigidity of the machine tool, part, and fixture, the tool material, and cutting speed and feed, which makes machinability more of a property of a machining system operating under a given set of conditions [6]. In order to quantify, determine, and compare the machinability of materials, engineers use metrics such as tool life, chip formation, cutting forces and power requirements, and surface roughness Chip formation The form and size of the chip produced can indicate the machinability of the material being tested under a set of machining conditions. Typically, materials that produce short, easy to manage, chips are considered to have better machinability than those that produce long, unbroken, chips or small, powder like, chips. Chip morphology is especially important in operations, such as drilling, where chip breaking and disposal concerns can be factors limiting production rate. Materials that form long, unbroken chips, such as ductile materials, also have greater tendency to form burrs, especially as the tool wears [6]. Therefore, the machinability of soft, ductile alloys, such as aluminum alloys, can be assessed based on the chip morphology and burr behavior, if tool wear is not an issue (such as when diamond tools are used). The forces that generate the formation of the chip can also be used to evaluate machinability. 11

28 2.4.2 Forces The measurement of cutting forces is a method of evaluating machinability that can provide insight into the behavior of material systems during machining. In general, chip load and width of cut have the greatest effect on cutting forces, while cutting speed has a lesser effect on cutting forces for most materials and cutting conditions [6]. The force per unit length, derived by dividing the cutting forces by the width of the cut of the effective edge length, is important to determine whether the tool will chip or break and is primarily dependent on chip load. Furthermore, increasing the rake angle tends to reduce cutting forces. Forces are usually lower for a tool with chip break geometry than for flat-face tools, since the effective rake angle is larger [6]. Additionally, flank wear tends to significantly increase cutting forces, thus machinability increases as cutting force and power consumption decreases for a given set of cutting conditions [6]. Consequently, lower cutting forces generally result in an increase in tool life, due to reduced loads on bearings and ways, lower tool wear rates, and improved dimensional accuracy, due to decreased deflections [6]. The two measureable cutting force components, measured using a dynamometer, can be used to calculate the resultant cutting force by Equation 2.4, where Fx is the measured force in the instantaneous feed direction and Fy is the measured force in the instantaneous normal direction (perpendicular to Fx). When collecting cutting force data, the component forces are measured by the specially designed dynamometers, which utilized technologies such as hydraulic, pneumatic, strain gage instruments, and most recently piezoelectric. Most dynamometers determine the tool force by measuring the deflections or strain in the elements supporting the cutting tool, while minimizing tendency for chatter, or vibrations, to occur during cutting [1]. Dynamometers are designed with high rigidity and natural frequencies to insure dimensional accuracy of the cutting operation is maintained and so that accurate, sufficiently large, strain or displacement measurements can be obtained. An advantage to collecting cutting forces is real-time data acquisition during machining. Commonly used today, piezoelectric dynamometers utilize quartz load measuring elements, which have high stiffness and broad frequency response, are very thermally stable and exhibit little static crosstalk between measurements in different directions [6]. However, since piezoelectric crystals produce a charge with leakage effects, piezoelectric dynamometers are best suited for dynamic measurements, rather than static measurements. 12

29 2.4.3 Tool life The study of tool life as a measure of machinability holds immense value in manufacturing, because tool life has direct implications in production cost and time, affecting both the quality and cost of machining. Tool life is the measure of time or cutting length that correspond to the growth of a specified wear mechanism reaching a critical tool wear value, which signifies tool failure. Although tool life studies can be very time and material consuming, it is one of the most direct methods to evaluate the machinability of a material. Conversely, machinability will increase as tool life increases for a set of cutting conditions. The machinability based on tool wear is restricted to the range of cutting conditions used because when conditions are changed (such as a drastic increase or decrease in cutting speed), the dominant tool wear mechanism and tool wear rate may change. In metal cutting, the three main tool wear mechanisms that occur are adhesion, abrasion, and diffusion wear [1]. Adhesion wear is caused by the formation of a strong bond, due to friction between the chip and tool material, which results in particles being transferred from one surface to another [5]. This particle transfer occurs when the bonds generated are stronger than the material at the cutting interface, resulting in a loss of fragments from the tool surface in the form of wear particles. The lost wear particles are typically particles from the tool material, creating wear growth on the tool. Abrasion wear also results from hard particles on the underside of the chip, that are highly strain-hardened fragments from an unstable build-up edge [1]. As a result, fragments of hard tool material removed by adhesion wear, or hard constituents in the work material such as oxide scale, pass over the tool face and remove tool material by mechanical action [1]. Solid state diffusion is defined as the movement of atoms in a metallic crystal lattice from a region of high atomic concentration to one of low concentration [1]. Solid state diffusion results in wear at high surface temperature and low surface velocity. This is due to the rate of diffusion increasing exponentially as temperature increases and the extended exposure of the tool material and workpiece to the temperature, at low velocities. In this regard, diffusion wear is the movement of atoms from the tool material to the work material due to the high temperature that exist at the intimate contact between the tool and work materials [1]. The form of wear observed on the tool depends on the cutting conditions, primarily cutting speed and undeformed chip thickness (depth of cut), as well as the tool material and work material. Although there are numerous types of tool wear, such as notch wear, nose radius wear, thermal and mechanical cracking, edge buildup, plastic deformation, edge chipping or frittering, chip hammering and 13

30 tool fracture or breakage, the two most prevalent types of wear are flank and rake wear. Rake wear occurs on the tool face resulting from the chip flow along the face of the tool forming a crater [1]. Flank wear occurs on the relief, or flank, of the tool creating a wear land caused by the friction between the newly generated workpiece surface and the contact area on the tool flank [1]. Figure 2.12 is a depiction of these main two mechanisms of wear for the two dimensional orthogonal machining model. Crater wear is normally a result of a combination of abrasion and diffusion wear mechanisms [10]. It can be evaluated by measuring the maximum depth of the crater formed on the tool face. The greatest depth of crater wear is usually near the midpoint of the contact length because this is where the tool face temperature is typically at its maximum, shown in Figure 2.13a [5]. However, since diffusion wear is the dominant wear mechanism in the development of crater wear at high cutting speeds, the temperature increases quickly and greatly increases the rate of crater wear. Crater wear tool life criterion can be defined as when the crater approaches the cutting edge, just short of total destruction, or when it reaches a critical value of KT, as seen in Figure 2.13a [5]. At very high cutting speeds (high temperatures) crater wear often becomes the factor that determines the life of the cutting tool because cratering becomes so severe that the tool edge is weakened and eventually fractures [1]. However, this only occurs at extreme conditions. When economic functional cutting conditions are used, the flank wear is typically the controlling tool wear factor. Flank wear, or wear-land wear, usually results from abrasive wear mechanisms and results in a loss of relief angle on the clearance face of the tool giving rise to increased friction resistance [5, 7]. The width of the wear land is normally used as a measure of the amount of wear and can be easily determined using a toolmaker s microscope [1]. Figure 2.13b depicts the major flank wear seen on a single point tool, where VC is the flank wear at the tool corner (zone C) and VN, notch wear, is at the opposite end of cutting activity (zone N) and VB, uniform flank wear, is the central portion (zone B) of cutting activity. In zone N a groove or wear notch often forms due to the tendency of work materials to work harden in this region [1]. Zone B is the location of a fairly uniform wear land, where the average wearland width is known as VB and the maximum wear land width is VBmax. The flank wear criterion specified for the width of the flank wear, which typically corresponds to VBmax, identifies the failure of a tool. In milling, a commonly used flank wear criterion value is 0.3 mm for uniform wear (average over all teeth) and 0.5 mm for localized wear (on any individual tooth) [11]. As machinability degrades, or the depth of cut decreases, the tool material used is more refractory and hence brittle so it is generally advantageous to select a smaller limiting flank wear value [5]. Flank wear, measured by wear land width, plotted versus time or cutting distance for a given cutting speed, is the typical tool life 14

31 progression curve reported, which consist of three stages. First, the quick break down of the sharp cutting edge and the establishment of a finite wear land, then a uniform rate of wear progression, and finally a sharply increasing rate of wear to failure. Since cutting speed typically has the largest effect on tool life, a fundamental relationship can be formulated between cutting speed and tool life that can be used as a guide for machining a given tool-material combination. The development of qualitative methods for predicting tool life, using the relationship between cutting speed and tool life can be used to optimize the economic impact of tool life on production operations. Tool life is largely dependent on tool material and cutting conditions, which makes it difficult to develop an empirical method of predicting tool life [6]. However, through a series of many experiments, F. W. Taylor (1907) [12] developed one of the most used tool life equations, which he first used to characterize machining with high-speed steel (HSS) cutting tools. Taylor, using the empirical relationship he developed, was able to increase the production of the Bethehem Steel Company s machine shop by 500% [1]. The equation developed, Equation 2.20, relates the tool life T in minutes to the cutting speed V through empirical tool life constants, C and n [6]. V T n = C [Eq. 2.20] The constants C and n correspond to a unique work and tool material combination, which is also influenced by the machining conditions other than cutting speed (feed, depth of cut, fluid, tool geometry, etc.). The n value depends primarily on the tool material, indicating the sensitivity of tool life to cutting speed (the higher the n value, the less sensitive) [13]. The C value, in Equation 2.20, is the cutting speed corresponding to one minute of tool life [13]. This C value varies widely with tool material, work material, and tool geometry and is typically in the range of 100 m/min for rough machining low-carbon steels [6]. The constant n is based on the tool material, typically in the range of 0.1 to 0.7 for HSS tools, 0.2 to 0.25 for uncoated tungsten carbide (WC) tools, 0.3 for TiC- or TiN-coated WC tools, 0.4 for Al2O3-coated WC tools, and 0.4 to 0.6 for solid ceramic tools [6]. The n value is typically small (normally less than one), thus a large value (near or above one) suggest that tool life is only linearly proportional to cutting speed and one might operate at higher speeds primarily to get the machining done in less time [13]. When the tool life (T) based on either total destruction or a limiting wear land is plotted against cutting speed using log-log coordinates the log equivalent equation (Equation 2.21) can be readily determined, where the constant n is the slope of the generally linear curve [5, 6]. Equation 2.21 can be solved to predict tool life as a function of cutting speed, Equation

32 ln T = ln C 1 ln V [Eq. 2.21] n T = C V 1 n [Eq. 2.22] However, the constant C in Equation 2.22 no longer corresponds to the cutting speed at which one minute of tool life is observed, it is simply an empirical constant relating the cutting speed and tool life for a particular work and tool material combination. In order to determine the constants and predict tool life due to a change in the cutting speed, machining conditions such as chip load and depth of cut must be held constant. In practice the tool life shows a significant variation for repeated tests with the same set of conditions, therefore requiring a considerable number of tests to establish the constants for a particular tool-work material combination with an appropriate degree of confidence [1]. Once appropriately developed, the Taylor tool life equations can be used to predict tool life due to cutting speed and guide the machining of a particular tool/work combination. In addition to tool life considerations when machining a particular tool/work combination, the surface generated is often of interest Surface finish A major factor of surface integrity, surface finish requirements plays a pivotal role in machining productivity and tooling cost, since the machined surface finish and geometric consistency are concerns as the tool wears. The effect of tool wear and the surface finish gives rise to stringent finish requirements that may limit tool life. In order to quantify the surface finish for comparison analysis, surface finish is commonly expressed as a single factor or index referred to as surface roughness. In engineering applications, surface roughness is usually characterized through the use of statistical parameters, which include the arithmetical mean value, Ra, maximum valley depth Rv, maximum peak height Rp, and peak to valley height Rt [6]. The arithmetical mean value Ra, the most commonly used roughness index, is defined as the average absolute deviation of the workpiece from the centerline, given in Equation 2.23 and shown in Figure 2.14 R a = 1 L y dx L o [Eq. 2.23] where y is the profile deviation from the centerline and L is the length of the measurement trace. 16

33 Surface roughness can be idealized for a given tool shape and feed rate, with the assumption that there is no BUE, chatter, or inaccuracies in machine-tool movement [1]. The ideal finish can be simply calculated from the feed per tooth, the tool nose radius, and the tool lead angle [6]. Equations 2.24 and 2.25 are examples of a method of calculating roughness in the ideal finish case, known as ideal roughness, R a = f 4(cot κ r e cot κ re ) [Eq. 2.24] R a = f2 r n [Eq. 2.25] where f is the chip load, κ re and κ r e are the working major and minor cutting-edge angles, respectively, and r n is the nose radius. Deviations form model prediction of surface finish are expected in practice due to effects related to tool wear, vibration, machining motion error, and work material effects such as inhomogeneity, BUE formation, and rupture at low cutting speeds [6]. In practice, surface roughness is a measure of surface texture that is quantified by the vertical deviations of a real surface from its ideal form, as shown in Figure These measurements are most commonly collected with a stylus-type profile meter known as a profilometer, which amplifies the vertical motion of the stylus as it is drawn across the surface [6]. Profilometers provide a two-dimensional profile of the traced surface segment, which are amplified in the normal and surface directions to accentuate surface contours and irregularities. In order to measure surface roughness, the effect of other types of deviations from a perfectly smooth surface, such as surface flaws and waviness must be minimized. However, surface flaws, which are widely separated irregularities that occur random over the surface such as scratches, cracks, or similar flaws, are difficult to eliminate but can be minimized by large surface profiles. Waviness refers to a form of regular deviation where the wavelength is greater than a specified magnitude (usually about 1 mm), which results from workpiece or tool deflections, vibrations, or run out [1]. Modern profilometers calculate the R a value continuously over a selected cutoff distance for a set wavelength (typically 0.75 mm), which is selected to eliminate waviness greater than the wavelength from the measurement. The arithmetical-mean surface roughness Ra, which is well suited for monitoring the consistency of a machining process, tends to increase with tool wear and be sensitive to changes in conditions of the 17

34 coolant or work material [6]. Typically, for a given set of cutting conditions, machinability increases as the achievable surface roughness improves. In milling, additional factors of the surface finish from that of turning or boring mainly result from differences in tooling and process kinematics. Face milling, using multitooth cutters, introduces setup and grinding errors that cause the cutting edges to all cut at slightly different chip load and depths of cut, especially with indexable cutters [6]. The effective chip load varies as a function of the vibrations, which are produced by vibrations caused by the interruptive nature of the process and the changes in cutting position caused by spindle and cutter run out, and the changing angle of the cutting edge from the feed direction [6]. These additional variations cause the surface roughness in milling to be less uniform than in turning or boring, where a single point cutting edge yields comparatively constant chip load and depth of cut. Therefore, additional roughness measurements are needed to develop the surface profile due to the increased variability for milled surfaces. Figure 2.1: Orthogonal and oblique cutting, (a) orthogonal cutting; (b) oblique cutting [1] 18

35 Figure 2.2: Terms used in metal cutting [1] Figure 2.3: Cutting force diagram - Merchant circle [1] 19

36 Figure 2.4: Discontinuous chip formation [1] Figure 2.5: Continuous chip formation [1] 20

37 Figure 2.6: Continuous chip with BUE formation [1] Figure 2.7: Types of milling operations [6] 21

38 Figure 2.8: Milling configurations, (a) up/conventional and (b) down/climb milling, adapted from [6] Figure 2.9: Chip thickness variation for up and down milling, where a = r [8] 22

39 Figure 2.10: Exit and start angle geometry for up and down milling [8] Figure 2.11: Cutting force geometry for tangential and normal force [8] 23

40 Figure 2.12: Regions of tool wear in metal cutting [1] Figure 2.13: Features of single-point-tool wear in turning operations. (after International Standards Organisation, ISO 3635, ISO Geneva, 1993) [1] 24

41 Figure 2.14: Definition of parameters used to compute the average roughness Ra [6] Figure 2.15: The shape, waviness, and roughness of a surface (after Whitstone) [6] 25

42 Table 2.1: Advantages and disadvantages of conventional and climb milling [7] CONVENTIONAL MILLING up MILLING (Feed movement opposite to tool rotation) Width of chip starts from zero and increases. Tooth meets the workpiece at the bottom of cut. Upward force tends to lift up workpiece. More power required - rubbing provoked by chip beginning at minimum width. Surface finish marred (spoiled) due to the chips being carried upward by tooth. Chips fall in front of cutter - chip disposal difficult. Faster wear on tool than climb milling. Climb milling down milling (Feed movement and tool rotation same direction.) Width of chip starts at maximum and decreases. Tooth meets workpiece at top of cut. Easier chip disposal - chips removed behind cutter. Less wear - increases tool life up to 50%. Improved surface finish - chips less like to be carried by the tooth. Less power required - cutter with high rake angle can be used. Climb milling exerts a downward force on workpiece - fixtures simple and less costly. 26

43 3 Austempered Ductile Iron (ADI) Austempered Ductile Iron (ADI) is a relatively new and unique material developed from Ductile Iron (DI) that is given a unique heat treatment, austempering. The heat treatment produces a uniform matrix microstructure of acicular ferrite and carbon-stabilized austenite, known as Ausferrite, with graphite nodules [14]. Due to advantageous mechanical properties, the uses of ADI has expanded to a large range of applications replacing metals such as steel, aluminum, and other light weight alloys. A recent survey, conducted in 2008, on the North American market distribution of ADI applications determined the distribution of ADI in various fields of engineering, shown in Figure 3.1. The agricultural and heavy vehicle industries account for the largest portion of the distribution with 27 and 28%, respectively. Other industries ADI is currently in are general manufacturing, railroad, light vehicles, and construction/milling industries accounting for 10, 11, 10, and 7%, respectively, of the ADI market. The increased market for ADI and the research into the nature of the material has developed in tandem with each other. Needless to say, researchers have conducted experiments, published technical papers and journals in order to explore and discuss the nature of the material and its mechanical properties. 3.1 Development of ADI Prior to the development of ADI, the austempering process was developed in the early 1930 s using steel. The equipment and process knowledge for the implementation of the process on DI was not developed until the 1960 s. In the 1960 s, companies such as International Harvester and General Motors began to develop ADI. In 1972, when the first commercial application of ADI for crankshafts was implemented as hermetically sealed Type AE compressors by Tecumseh Products [15]. In 1977, General Motors replaced its carburized and hardened steel with ADI hypoid ring and pinion gear in Pontiac rear-drive automobiles [15, 16]. Cummins Engine Co. developed ADI timing gears in its B and C series diesel engines, producing its ADI very similar to GM, in 1984 [16]. By the 1900 s, automotive companies such as Navistar, Freightliner, Kenworth, GM, Iveco, Volvo, and other heavy truck manufacturers, as well as many other industries, have adopted ADI in various components [16]. In 2003, the annual worldwide production of ADI was estimated as 125,000 tons and is expected to exceed 300,000 tons by 2020 [15]. Today ADI has expanded to applications in agricultural equipment, construction equipment, gears or powertrain, heavy truck or trailers, light vehicles and buses, mining or forestry equipment, railway 27

44 equipment, farm and oilfield machinery, conveyor and tooling equipment, defense, energy generation, and even sporting goods [17]. Due to the growth of the market, a standard of ADI was issued in 2002 (revised in 2006) by ASTM (A897/A879M). By 2003, two more standards for ADI were approved and published by SAE (J2477) and ISO (17804). In 2009, China also issued a standard for ADI. A comparison of all four standards (ISO, SAE, ASTM, and China) for ADI is shown in Table 3.1. The standards identify the following properties in the following format: tensile strength (MPa) yield strength (MPa) elongation (%) [18]. However, the ISO standard is only designated for sections less than 30 mm. The standards also compared the Brinell hardness numbers for the various grades of ADI, which is shown in Table 3.2. The ISO, SAE, ASTM, and China standards developed for ADI are similar but not identical. ISO and China developed standards for grades 800, 900, 1050, 1200, and 1400, while SAE and ASTM developed standards for grades 750, 900, 1050, 1200, 1400, and Because the hardness values for the various grades overlap, a given ADI material may satisfy two different grade requirements. In many cases, the same DI chemical composition can be heat treated to the various grade requirements simply by adjusting the austempering temperature and time. 3.2 Production of ADI The production of ADI is a two-phase heat treatment process, which involve austenization and isothermal austempering. The austempering reaction is characterized by the decomposition of austenite to ferrite and enriched austenite in stage one followed by further decomposition of enriched austenite to ferrite and carbide in the stage two as austempering proceeds [19]. The heat treatment procedure is shown in Figure 3.2. In the austenization phase (A-B and B-C in Figure 3.2) the casting is first heat treated to a temperature range of F ( C) and then sustained at that temperature for one to three hours. This step in the process forces the DI matrix to become fully austenitic and saturates the austenite matrix with carbon. The casting is then elevated temperature quenched (C-D in Figure 3.2) in a quenching medium at a temperature range of F ( C), which corresponds to the austempering temperature. A high quenching rate is required to avoid the formation of pearlite and to reach the targeted austempering temperature rapidly [20]. The isothermal austempering phase (D-E in Figure 3.2) consist of holding the casting at the austempering temperature for a period of time. The period of time the casting should be held at the austempering temperature is restricted to a window of time known as the heat treatment processing window. This is due to the fact that an insufficient holding time does not allow for sufficient carbon 28

45 stabilization of the austenite, which results in the undesirable transformation to martensite upon cooling to room temperature. On the other hand, excessive holding time at the austempering temperature results in the undesirable decomposition of the stabilized austenite to form ferrite and carbide, also known as bainite. The size of the heat treatment processing window (in terms of an acceptable austempering time) is dependent on several factors such as alloying composition, alloying-element segregation patterns and austempering temperature [21, 22]. The final step of the process is the cooling of the casting to room temperature (E-F in Figure 3.2), resulting in a matrix known as ausferrite consisting of ferrite needles in retained (carbon stabilized) austenite [4]. The main factors that affect the austempering success are austempering time and temperature, austenitizing time and temperature, chemical composition, alloy segregation, DI quality, and section size of the casting. The heat treatment process parameters influence the microstructure and properties of the ADI. The austenitizing temperature and time is chosen to ensure a fully austenitic matrix structure with uniform carbon content. The austenitizing time must also be sufficiently long enough to completely modify the initial microstructure to an austenite matrix saturated with carbon [23, 24]. The thermal equilibrium carbon content of the austenite is a function of the austenitizing temperature. The austenitizing temperature and time are typically chosen as 900 C for 1 hour. However, the upper critical austenitizing temperature (UCT) is strongly dependent on the chemical composition, especially silicon and manganese of the pre heat treated DI [23, 24]. Silicon increases the UCT, while manganese causes a decrease in the UCT. The three main factors drive the selection of the austenitizing time are the section thickness, nodule count and the initial microstructure. The section thickness is proportional to the austenitizing time, while nodule count is inversely proportional to the austenitizing time. Increasing the austenitizing temperture causes an increase in dissolved carbon leading to higher hardness and strength. At higher austenitizing temperature the ADI structure tends to be more like upper bainitic with an increase in elongation and impact strength. In the second phase, austempering temperature and time are the processing parameters with the greatest effect on the final microstructure. Decreasing the austempering temperature delays the austempering transformation, which increases the austempering time needed to obtain the maximum impact energy [21]. At lower austempering temperature the structure may not completely stabilize, leading to a martensite content up to 12% upon cooling, which results in an increase in hardness [14]. Lower austempering temperature also creates a finer ausferrite that contains low austenite content with very fine ferrite needles. At higher austempering temperature a coarser ausferrite microstructure is produced, which consist of coarse and distinct bainitic ferrite 29

46 needles. This results in more ductility but smaller increase in strength and hardness than transformed at lower bainite temperature range [22]. The dissolved carbon, austenite grain size and reduction in the rate of austempering have a proportional relationship with the austenitizing temperature. Austempering temperature and time, which have the largest effect on the final microstructure, are chosen based on the desired mechanical properties. At lower austempering temperatures, the diffusion rate of carbon is decreased and the driving force for the formation of ferrite is increased [23]. At higher austempering temperatures, the increased rate of diffusion of carbon creates more stabilized austenite. Therefore, higher austempering temperatures (350 to 450 C) lead to lower strength and hardness but higher elongation, toughness, and in many cases better fatigue characteristics. While, lower austempering temperature (250 to 350 C) lead to higher strength, hardness and abrasion resistance but lower elongation and toughness [24]. The austempering time is critical because it determines where in the processing window the casting is being produced. Shorter austempering times drive the operations toward a martensitic reaction, when cooled to room temperature, thus resulting in an increase in strength and hardness but a significant decrease in ductility. Very short austempering time can create a hard but brittle casting with poor mechanical properties. Longer austempering times drive the operation towards an upper bainitic reaction, which results in an increase in ductility and a decrease in strength. Austempering time is also extremely critical because it determines which phase reaction occurs. The first reaction (stage 1 reaction) corresponds to the phase transformation seen in the heat treat processing window, while the second reaction (stage 2 reaction) represents the reaction passed the upper bound of the processing window. At longer austempering times, due to the carbide formation, the second reaction (stage 2 reaction) leads to an undesirable reduction in ductility and toughness. On the other hand, shorter austempering times lead to poor mechanical properties due to the insufficient transformation of the first reaction. Stage 1 Reaction: γ (Austenite) α (Ferrite) + γ hc (enriched Austenite) Stage 2 Reaction: γ hc (enriched Austenite) α (Ferrite) + ε (Carbide) The size of the heat treatment processing window is one of the many factors of ADI that are effected by carbon and different alloying elements as well as the segregation of those elements. Silicon (Si), manganese (Mn), sulfur (S), copper (Cu), nickel (Ni), molybdenum (Mo), magnesium (Mg), phosphorus 30

47 (P), chromium (Cr), vanadium (V), and titanium (Ti) are examples of elements that can have a significant effect on ADI. The main effects alloying elements in ADI have on the microstructure and materials properties are changes in carbide formation and segregation of elements, hardenability, and the processing window. Si levels above 2% delay the formations of carbides, allowing a high toughness ferrite and stabilized austenite microstructure to form and effectively increasing the processing window size up to 3.2% Si. Mn acts as a carbide stabilizer and hardenability promoter by delaying the end of the martensitic reaction but this can effectively shrink the processing window due to the severe segregation of Mn to the intercellular regions of the DI microstructure. The segregation of alloying elements to intercellular regions of the DI microstructure leads to poor mechanical properties, such as low fracture toughness. Cu increases hardenability and acts as a barrier to carbon diffusion, at the austenite-graphite interface, to stabilize the austenite, thus increases both the transformation rate and the carbon content in the matrix during austenitizing and restricting carbide formation during austempering [25]. Ni and Mo are both hardenability promoters but Ni widens the processing window, while Mo slightly shrinks the processing window due to strong carbide formation from segregation. Mn and Mo are typically limited to % and %, respectively. Cr, V, Ti, Cerium and Antimony have tendency to form carbides at cell boundaries but have less effect on mechanical properties and are rarely intentionally added to ductile irons [4]. Casting quality factors, such as poor nodularity and low nodule count, deteriorate ADI properties. However, undissolved carbides or porosity have more effect on the properties. Nodularity is defined as: Nodularity (N) = 100 D d where D is the diameter of the major axis of the nodule and d is the diameter of the minor axis of the nodule. The nodule count is defined as the number of graphite nodules per unit area. The importance of nodular count increases with the addition of alloying elements because many alloying elements segregate to the intercellular regions between the graphite nodules. Low nodule counts lead to larger spacing between the graphite nodules and larger regions of segregation, which promote the formation of low carbon austenite or even martensite after austempering [26]. On the other hand, higher nodule counts minimize micro segregation, micro shrinkage and cell boundary carbides. Properties such as toughness, strength and machinability improve with high nodule counts. The minimum quality characteristics for good ADI quality are consistent chemical analysis, <100 nodules/mm 2 nodule count>, 85% nodularity, with a maximum of 0.5% carbides and inclusions, a maximum of 1% micro shrinkage 31

48 and consistent austenite and ferrite ratios [26]. Fluctuations in DI composition can affect the heat treatment times, temperature and hardenability of the iron. A consistent austenite/ferrite ratio is important to the minimizing of the variations in mean volumetric expansion during austempering. The addition of copper (0.6%) increases the strength [27]. Strength, elongation and impact energy strongly depend on amounts of bainitic ferrite and retained austenite [25]. An austempering temperature range of C produces high ductility, yield strength of approximately 500 MPa, good fatigue strength and impact strength. Lower austempering temperature produces high hardness, excellent wear resistance and contact fatigue strength [28]. In order to facilitate the formation of a stable microstructure the proper austempering time must be chosen to optimize the parameters. Insufficiently long austempering time results in a lack of carbon stabilized austenite and martensite in the microstructure, which causes high hardness and low ductility [28]. In monotonic testing, unlike steel and aluminum, ADI does not exhibit necking (extreme plastic deformation local to the fracture site) when pulled to failure in a test bar, instead the cross section is reduced along the entire gage length of the bar [29]. Consequently, ADI tends to exhibit equivalent and/or superior wear performance to competitive materials in applications where this strain transformation to martensite can occur [29]. The high strength-to-weight ratio, good stiffness, 100% recyclability and low embodied energy of ADI allow it to rank well amongst sustainable engineering materials [29]. ADI has found application in most durable goods industries and into products ranging from bulldozers, to automobile camshafts, to power tools, to engine timing gears, to dry cleaning conveyors, etc [30]. Specifically, ADI is currently used as ground engagement components-digger bucket teeth and dozer blades, truck suspension brackets, differential spiders, tow hooks, cylinder liners, engine con rods and crank shafts [31]. The ability of ADI to be easily deformed into various shapes makes it applicable not only for thin products which cannot be fabricated from conventional casting materials, but also for products which are used under severe conditions [32]. However, the specific application ADI is applicable to typically depends on the grade of ADI. The structural grades, GR900 and GR1050, are often used in suspension components and many other dynamic applications; while the wear grades, GR1200 and GR1400, are used in applications where wear resistance is the primary criteria for material selection because of their high hardness [14]. The lower grades of ADI compete favorably with forged steels as a cost and weight saver, while the higher grades compete with carburized and hardened steels for wear resistance in addition to better noise damping capabilities and a generally lower manufacturing cost [16]. In general, 32

49 the energy required to produce an ADI component is typically reduced by 50% compared to steel casting and forgings, which results in a reduction in manufacturing cost and often substantial savings [33]. However, the substantial increase in strength and wear resistance from the austempering process, which creates ADI, logically presents some machining problems [28]. For example, gears hubbing of ADI gears is very difficult because of the occurrence of tool failure in a short time [32]. The difficulties in the machining of ADI have limited the mass adoption of the material in the current industries it is being used as well as new industries. The myth that ADI is un-machinable, due to its high hardness, is still held by many industries today [14]. 3.3 Machining of ADI Regardless of the advantages of a material, the ability, as well as the cost, of processing the material will always be a factor in the adoption of the material. Over the last thirty years, various studies on the machinability of ADI have been performed, but the reasons for the difficulties of machining ADI remain a mystery and one of the major restricting factors in the growth of the market for ADI. Three different strategies for the cost effective machining of ADI components have been proposed. The first option is to machine prior to heat-treatment, which circumvents the challenges of machining ADI. This option can be very effective depending on the design of the component if the dimensional growth during the austempering process can be predicted. The second option is to perform a rough machining operation prior to heat treatment and a finishing operation after heat treatment. This process is applicable when tight tolerances and surface finish are required that cannot be held when heat treating after machining [14]. However, the second option can result in the logistical challenge of removing components from the machining flow, for heat treatment, then returning them after heat treatment. The final option is to completely machine the components after heat treatment. This allows for the maintaining of tight tolerances and surface finish without the logistical considerations of the second option. When machining ADI, in any capacity, the hardness of the material and strain-induced surface transformation must be taken into account. Early machining guidelines for ADI have recommended in order to get acceptable tool life, to set machining parameters 50% less speed and 50% deeper compared to materials with similar hardness values well as use slower feed rate [14]. In addition to this recommendation, overall machinability has been investigated by many researchers utilizing some combination of machinability characteristics, which include tool life, cutting energy, surface roughness, and chip formation. Although, the machinability of ADI, in the post heat treatment state, has been investigated for many years, 33

50 unanswered questions remain. There has been far more study of the machinability of the lower, softer grades of ADI than the higher, harder grades. One of the most researched phenomenons, with respect to the machinability of ADI, is the effect of the various machining parameters on the cutting forces. The cutting forces can be used to determine the cause of deflections of the part, tool, or machine structure, and supply energy to the machining system which could result in excessive cutting temperature or unstable vibrations [6]. Additionally, cutting force can be used to derive machining information such as machine power requirements and bearing loads. Cutting force measurements are also a means of comparing the machinability of materials, especially in cases where time and material are too limited to conduct tool life tests. In general, chip load or depth of cut has been found to have the largest effect on cutting forces. In the case of ADI, researchers have reported that the chip load is proportional to cutting forces [35, 36]. Therefore, theoretically, if the chip load or depth of cut is increased the machinability, with respect to the cutting forces, will decrease. However, in the case of ADI, decreasing the depth of cut (reduced chip load) does not always improve machinability [34]. Avishan, Yazdani, and Vahid [34], in turning studies, found that ADI machining at deeper depths of cut, such as 0.5 and 1 mm, exhibits better machinability than lower depths, such as 0.1 mm, compared to that of most alloys. Other machining parameters, such as cutting speed, are reported to have little to no influence on the cutting forces. However, cutting speed does demonstrate a measureable effect on the cutting force, in some investigations. Typically, in turning, as cutting speed is decreased, especially for cutting speeds less than 200 m/min, the ADI cutting force increased [36, 38]. Conversely, in turning, at extremely high cutting speeds a drastic increase in the cutting forces can be observed. In several investigations researchers found that the machinability of ADI, in terms of tool life, degrades as the cutting speed increases [35, 36, 38, 39, 40]. Increasing the cutting speed leads to decreasing tool life by rapid wear development [35]. The rapid wear development, in many cases, can be attributed to the effect that cutting speed has on the temperature at the cutting interface. Large tool wear rates, thus short tool life and high cutting forces, indicate high temperatures at the cutting interface. Therefore, increase of cutting speed, which also increases tool wear rate and shortens tool life, increases the temperature at the cutting interface as well. In metal cutting, the most common type of tool wear measured is flank wear due to its effect on the dimensional accuracy and surface quality. The shape of the tool life curve, with respect to cutting length (or time), tends to have nonlinear behavior as it approaches the wear limit for flank wear. In one case of ADI turning, reported by Akdemir, Yazman, 34

51 Saglam, and Uyaner [36], the increase in the flank wear of the cutting tool with respect to cutting speed was approximately uniform up to a machining length of 200 mm and beyond a machining length of 300 mm the maximum wear limit for flank wear was exceeded due to the temperature generated in the cutting area [36]. Even though the behavior of the tool life remained similar, the transition points between the phases of wear depended on the cutting tool and grade of ADI. In a turning study by Katuku, Koursaris, and Sigalas [37] the machinability of ADI, ASTM grade 2 (ADI 900), with polycrystalline cubic boron nitride (PcBN) cutting tools under cutting conditions of constant depth of cut 0.2 mm, feed rate of 0.05 mm/rev, and cutting speeds ranging from 50 to 800 m/min was investigated as a finishing operation. This investigation found that cutting speeds between 150 m/min and 500 m/min were optimum for the production of the workpiece with acceptable cutting tool life, flank wear rate and low dynamic cutting forces [37]. Another turning study by Kacal and Gulesin [38] utilized the Taguchi Method (TM) to determine the optimal finishing cutting conditions by evaluating various cutting speeds (600, 700, and 800 m/min), chip loads (0.05, 0.08, and 0.12 mm/rev), cutting tools (ceramic and CBN) and workpiece materials (DI, ADI at 380 C, and ADI at 290 C). This investigation found that the best cutting conditions with respect to surface roughness (ADI austempered at 290 C) were ceramic cutting tool, 800 m/min cutting speed, and a feed rate of 0.05 mm/rev [38]. Additionally, the effect of cutting speed and depth of cut was investigated by Akdemir, Yazman, Saglam, and Uyaner [36], in a turning operation, by varying the cutting speed, depth of cut, and machining length in the ranges of m/min, 1-3 mm, and mm, respectively, at a constant chip load of 0.12 mm/rev. With use of the response surface methodology techniques, the optimum cutting parameters were identified to be a depth of cut of 1 mm and cutting speed of m/min for surface roughness [36]. The optimum cutting parameters for the tangential force and flank wear were found to be 1 mm, m/min and 1 mm, m/min, respectively [36]. The disparity of the optimum conditions found by these studies can be attributed to the different cutting conditions and cutting material used in the studies. Understanding the variations in the machinability of different grades of ADI compare to each other and other materials is pivotal to improving the machinability of ADI. In a turning study by Aslantas and Ucun [39] two cutting tool materials, ceramic (Al2O3 based) and cermet (TiCN+TiN coated), were compared for various cutting speeds (100, 200, 300, and 500 m/min), constant depth of cut of 1 mm and chip load of 0.1 mm/rev in the machining of ADI-250 and ADI-375. In terms of cutting speed and surface roughness, the ceramic tools were found to be unsuitable for machining at low cutting speeds (V 35

52 < 300 m/min) and the cermet tools were not appropriate for machining at higher cutting speeds (V>240 m/min) with respect to tool life. However, both cutting materials are suitable for machining at high cutting speeds (V> 400 m/min) with respect to surface roughness [39]. Often, research suggests that high wear resistance cutting tools for the machining of ADI such as K-grade carbide tools with the use of cutting fluid and P-grade for dry cutting as well as Al2O3 ceramics for continuous cutting processes but Si3N4 ceramics and PcBN were not recommended for machining of ADI [14]. Carbide cutting tools, of various coatings, was found to demonstrate favorable machinability characteristics on various grades of ADI for a wide range of cutting conditions [36, 44, 45]. Reported by De Carvalho, Montenegro, and Gomes [40], the chip formation and surface roughness tend to be functions of the cutting tool and its geometry, as well as the cutting conditions. In turning, at higher feed rates, chips exhibit a tendency to increase in thickness and at greater depths of cut the chip tends to be flat with a round end [40]. Although surface roughness typically improves as the cutting speed decreases, in the machining of ADI researchers found better surface finish at higher cutting speeds, in some cases [36, 43]. The shape of the roughness curve, with respect to cutting speed and feed rate, is expected to be a parabola but in some cases can be linear, increasing with machining roughness as chip load increases. This suggest that other aspects such as good rigidity in the tool holder and turret disk and tribology between insert and workpiece could contribute to better surface quality [40]. Even though the surface roughness improved when greater nose radius were used, as expected, little to no improvement was seen between nose radius 0.8 and 1.2mm. Coupled with the aspects stated above, this suggest that kinematic roughness is not the only main contributor to the roughness value [40]. However, when the cutting speed and depth of cut were varied from m/min and 1-3 mm, respectively, Akdemir, Yazman, Saglam, and Uyaner [36] observed the expected parabola shaped surface roughness curve. The surface roughness value decreased with increasing cutting speed until a certain cutting speed and then begun to increase as the average surface roughness increased with increasing depth of cut [36]. Furthermore, surface roughness values for the lower grades, ADI GR900 and GR1050, tend to be high compared to higher grades, ADI grades 1200 and 1400 [41]. The grade of ADI also plays a role in machinability. Due to the different hardness and structural composition, the assessment of the machinability of the difference of ADI can vary tremendously. Several researchers found that when machining the softer grades of ADI, GR900 and below, that the machinability is equal to or superior to that of steel with equivalent strengths [14, 19]. When the higher grades are studied, such as GR1200 and above, researchers found that as hardness increases so does tool 36

53 wear and cutting force [28]. Along with the hardness, the machinability of ADI, especially of the softer grades, is degraded by plastically induced phase transformation in the microstructure. It has been widely concluded that a strain-induced phase transformation occurs, in the machining of ADI when a sufficient normal force is applied transforming the austenite to martensite [14, 38, 46]. The strain induced transformation (SIT) that occurs during machining takes place when the surface of the workpiece is subjected to a high rate of plastic deformation. Since residual stresses or the internal stresses remaining after the normal force is removed are associated with plastic deformation logically it would also play a role in martensite formation. High thermal energy, the temperature in the cutting zone, can relieve the residual stress and speed up the SIT reaction by transforming the unstable retained austenite to a stable martensite phase [4]. Research has shown that the susceptibility of ADI to this phase transformation is a function of the content and morphology of retained austenite and carbon, which also plays a role in the mechanical properties of ADI [47, 48]. For example, at higher austempering temperatures, the addition of nickel and copper increases the level of austenite remaining in the matrix and imparts excellent toughness and fatigue strength in the ADIs [42]. At high austempering temperatures the percentage of retained austenite increases, increasing the tendency of strain-induced martensite to form [43]. Garin and Mannhein [44] found that with 20% cold work, the content of retain austenite will gradually decrease as the formation of martensite increases as a result of further cold work. Since the susceptibility of ADI is dependent on the amount of retained austenite and carbon, the softer standard grades of ADI would be more susceptible to the transformation to martensite. Although the SIT from austenite to martensite degrades the machinability of ADI, due to the increased hardness, the high hardness grades of ADI, which are less susceptible to SIT, also exhibit poor machinability. This is because as the hardness of the workpiece increases, tool wear increases, which corresponds to poor machinability [28]. 3.4 Surrogate materials In order to further evaluate the machinability of ADI, as an alternative material, a widely used material with similar properties and applications was also evaluated. AISI/SAE 4340 steel (S-4340) is considered the standard by which other ultrahigh-strength steels are compared, combining deep hardenability with high ductility, toughness, and strength [45]. S-4340 is considered a general-purpose steel with a wide range of applications in automotive and allied industries by virtue of its good hardenability enabling it to be used in fairly large sections [46]. It is also used in aircraft components where strength and toughness 37

54 are fundamental design requirements [47]. More specifically, S-4340 has been implemented in applications which include bolts, screws, and other fasteners; gears, pinions, shafts, and other machinery components; crankshafts and piston rods for engines; and landing gear and other critical structural members of aircraft [45]. Many of the industries S-4340 is widely utilized, ADI could be a viable alternative that could exhibit equal or improved performance with substantial weight saving. The ductility, toughness, good fatigue strength, and wear resistance or hardenability properties of ADI would likely provide, at least, equal performance to S-4340 in applications where wear resistance is required. On the other hand, the high strength-to-weight ratio, among other properties, could allow ADI to outperform S-4340 in many applications. Similar to ADI, S-4340 exhibits relatively poor machinability with machinability ratings of 55% for cold drawn material and 45% for annealed material, based off 100% machinability of cold-rolled B1112 [45]. The combinations of the material properties and similar machinability characteristics, along with the common applications, makes S-4340 a great material to investigate the machinability of in tandem with ADI. Investigating the machinability of S-4340 and ADI in tandem provided a means of evaluating the validity of ADI as an alternative material to S Figure 3.1: ADI market distribution - North America 2008 [48] 38

55 Figure 3.2: ADI heat treatment process [4] Table 3.1: Comparison of ISO, ASTM, SAE, and China standard grades for ADI [29] ISO Issued 2005 SAE J2477 Issued 2003 Revised 2004 ASTM A897/A879M Issued 2002 Revised 2006 China Std. GB/T24733 Issued Table 3.2: Comparison of the Brinell (HBW) hardness ranges for the various ADI grades [18] Grades (TS MPa) ISO SAE J2477 ASTM A897/A897M

56 4 Experimental The main objective of the experiments was to assess and compare the machinability of ADI GR900, GR1050, and GR1200 to S-4340, which was accomplished by conducting face and end milling studies across a range of machining parameters. This chapter describes the experimental methodology used in the study. The description includes machining methods and configurations used and the machinability metrics that were used (tool life, surface roughness, and cutting forces/stiffness). This chapter also describes the machining center operation data, which includes machine preparation, the cutting tools used, machining parameters (speed, chip load and depth of cut), use of coolant and its preparation, and the overall aim of the experiments. In addition, the chapter describes the instruments used to collect and analyze data obtained for hardness, tool life, surface roughness, and cutting force. The experimental design for the research work is divided into two sections: the characterization of the work materials and the machinability methods. 4.1 Work material characterization The DI casting used to create the ADI work materials was poured in two separate commercial heats, resulting in slightly different chemical compositions. Table 4.1 displays the chemical composition of the DI. All the DI were commercial heat treated to the following grades of ADI GR900, GR1050, and GR1200. The same austenization conditions were used for all grades, austenizing at 1625 F (885ºC) for 128 minutes. The second phase of the heat treatment process, austempering, varied between grades. GR900 was austempered at 720 F (382ºC) for 134 minutes, GR1050 was austempered at 670 F (354ºC) for 171 minutes, and GR1200 was austempered at 600 F (316ºC) for 225 minutes. The retained austenite content in the post heat treatment microstructure of the ADI grades were characterized using X-ray diffraction (XRD) analysis. Invented by Bragg, XRD analysis is governed by Bragg s Law, shown in Equation 4.1 λ = 2 d sin θ [Eq. 4.1] where θ is the angle of reflection, d is the lattice spacing between the reflecting planes and λ is the wavelength of the reflecting X-ray. In this analysis, quantitative diffraction peak area was determined by profile filling with a Pearson VII function for each peak identified in the respective patterns. The 40

57 diffraction patterns for each sample were then background corrected using a cubic-spline function. The volume fraction calculations were determined using values for each of the ferrite and austenite diffraction peaks [49]. In order to correct the positional changes in the austenite diffraction peaks, prior to the volume fraction analysis, the lattice parameter was determined based on carbon content. The lattice parameter determination of the austenite phase used a least square analysis for the three diffraction peaks collected from each sample. The XRD data was collected using a Scintag XDS-2000 θ/θ diffractometer equipped with a copper target x-ray tube (Superior XRD Consulting) from 70 to 105 2θ, with a step size of 0.05, and 30 second count time [49]. The x-ray optics used consisted of 1 and 2 mm beam slits, 0.5 and 0.3 mm receiving slits, and a graphite crystal monochromator [49]. Solid samples were polished and etched with Nital three times, then mounted to sample holder. Figure 4.1 displays the X-ray diffraction pattern for the grades of ADI (GR900, GR1050, and GR1200) along with the superimposed JCPDS-ICDD powder diffraction data, showing the phases ferrite in red and austenite in black. The large deviation in angular position between the austenite diffraction peaks and the powder diffraction data, shown in black, can be attributed to large amounts of carbon present in the austenite for the collected data [49]. Additionally, several very low intensity diffraction peaks, at angles such as 77.5, 84, and 102 2θ, can be attributed to the presence of iron carbides. However, further data collection and evaluation would be needed to positively identify those diffraction peaks. The volume fraction calculations (% of matrix) were determined using the direct comparison method, along with a temperature correction and area values for each of the ferrite and austenite diffraction peaks [49]. The results of the XRD analysis showed that the grades of ADI consisted of the following ferrite and austenite volume fractions, GR900 (40% austenite and 60% ferrite), GR1050 (37% austenite and 63% ferrite), and GR1200 (30% austenite and 70% ferrite), as shown in Table 4.2 and austenite lattice parameters of for GR900, for GR1050, and for GR1200, as shown in Table 4.3. In order to characterize the hardness of the grades of ADI and S-4340, Brinell hardness test were conducted on 101x76x25 mm 3 workpieces. The Brinell hardness tests were conducted on the 3000 kg scale with a 10 mm steel ball indenter. The results from the bulk hardness tests for the grades of ADI (GR900, GR1050, and GR1200) and S-4340 are given in Table 4.4. GR1200 has the highest bulk hardness, 429 HBW, GR900 has the lowest value, 320 HBW, and GR1050 fell in the middle, 357 HBW. The bulk hardness data was used to validate the hardness measurements collected with respect to ASTM 41

58 standard data, which classifies the grades of ADI. The bulk hardness values obtained for all the ADI grades fell within the ASTM standard hardness ranges. As expected, the hardness observed exhibited an increasing trend from GR900 to GR1200. Additionally, the hardness of S-4340 was found to be HBW, which falls in between GR900 and GR1050. The hardness values were used to determine test conditions from the Machining handbook [45] for the initial cutting force experiments. All experiments were conducted on post heat treated materials with the cast and heat treated outer layer removed. The workpieces were cut and/or milled into plates of dimensions 305x151x25 mm 3, for face milling, and 101x76x25 mm 3, for end milling. Tool life and surface roughness data were collected from the face milling operations and cutting force data was collected from the end milling operations. The ADI grades were characterized in terms of chemical composition, heat treatment parameters, microstructure, and hardness. Since S-4340 is a well-known material, it was only characterized in terms of hardness, which was used to compare it with the ADI grades. 4.2 Experimental platform All experiments were carried out on a HAAS VF-2 vertical CNC Mill (with a maximum spindle speed of 7500 rpm) or a HAAS VF-2SS (with a maximum spindle speed of 12,000 rpm), under various setups and configurations. The face milling operations were conducted using a large modular vise on the HAAS VF-2, while the end milling operations were conducted on a dynamometer mounted to the machining center table of the HAAS VF-2SS. A SECO Double Octomill S R SA (diameter-80mm) was used, in 3 and 6 insert configurations, for the face milling operations along with a CAT40 arbor 2.5" GL - E EDP# The tool holder has axial and radial rake angles of -8, a cutting rake angle of -11 and a tool cutting edge angle of 40, shown in Figure 4.2. Additionally, due to the cutting insert geometry, an effective relief angle of 15 was used. The cutting tool was a SECO carbide coated face mill insert (ONMU090520ANTN-M14 MK2050 PVD) with PVD deposited TiSiN-TiAlN Nanolaminate coating. A new cutting edge was used prior to each test. Prior to conducting any experiments and analyses, ADI specimens were machined to remove the cast and heat treated outer layer, shown in Figure 4.3. Two 305 mm long parallels were used to add support and reduce chatter at the ends of the specimens. The cutting conditions used for workpiece preparation were 500 surface feet per minute (sfm), feed rate of 10.8 inches per minute, roughing depth of cut of 0.03 inches and a finishing depth of cut of 0.02 inches. 42

59 The coolant used in all tool life experiments was Quakercool 7020-CG (Quaker Chemical Company). The main function of coolant is to reduce the tool-workpiece interfacial temperature. The coolant concentration was between 7% and 8% and two measurement methods were used to verify these concentration ranges. The first method used was a Westover Portable Refactometer, model RHB-32, which measures ratio of water and non-water of a solution present on a slide. This method provided an approximation of the concentration of the coolant solution of 8.5%. Titration was also used, which can estimate the concentration of a coolant solution by determining the required amount of acid (0.5 normal Hydrochloric acid) to reach a critical ph value (e.g., 4.4 ph). Once the amount of acid to reach the critical value is obtained, Equation 4.2 is used to calculate the concentration. Concentration = ml of HCL [%] [Eq. 4.2] Five samples were taken and the average concentration was used to determine the per volume concentration of 7.58%. The concentration was within the desired range for the coolant, 7 to 8%. Prior to using the coolant, the tank and coolant lines of the machine were cleaned to avoid cross contamination. The machine was populated with the desired coolant and periodically checked with regard to level and concentration. 4.3 Dynamometer calibration The dynamometer used was a Kistler type 9273 and the charge amplifier was a Kistler type To calibrate the dynamometer short static measures can be used, in some capacity, to verify operation and accuracy. Three weights of known mass (0.5 kg, 1 kg, and 2 kg) were used in a fixture designed to statically isolate two of the desired channels, x-component (Fx) and y-component (Fy) (Figure 4.4). The dynamometer was orientated in the direction of each channel such that a load-bearing wire was directly in-line with the isolated channel. The results from the static calibration of the dynamometer are given in Figure 4.5. The measured values did not reach the theoretical force, mass times the force of gravity (9.8 m/s 2 ), which can likely be attributed to the static nature of the measurement. Piezoelectric dynamometers are designed for measuring dynamic events, static loading is not ideal and will be subject to charge dissipation. Table 4.5 shows the dynamometer calibration results. When measuring the force generated by a 0.5 kg (4.9 N) weight, the measured force was N. When measuring the force generated by a 1 kg (9.8 N) weight, the measured force was N. When measuring the force generated by a 2 kg (19.6 N) weight, the measured force was N. Due to the fairly small deviations observed, the dynamometer can be assumed to be functioning properly. 43

60 4.4 Machinability metrics Machinability of GR900, GR1050, GR1200, and S-4340 was evaluated by measurement of tool life and surface roughness in a face milling configuration and cutting force in an end milling configuration (Figure 4.6). The setup for the face milling operations is shown in Figure 4.7. The workpiece was mounted in the vise with the coolant nozzle in the direction of the tool feed, to facilitate coolant delivery during machining. The tool life test was conducted according to ISO In this regard, 5 cutting speeds were investigated per work material with 3 replications per condition. Useful tool life was defined as maximum flank wear penetration (VBmax) measured from the uniform wear of 0.3 mm (average wear across all teeth) or localized wear of 0.5 mm (on any individual tooth) [11]. Wear land measurements were taken over intervals corresponding to constant volume of material removed (e.g., after each tool pass). Measurements were taken by removing each insert from the tool and placing them in the measurement fixture, shown in Figure 4.8, then returning them to the same location in the holder for the next pass. Measurements were taken at the maximum magnification of the stereoscope (Nikon SMZ800) of 63x. The magnification was verified using a standard glass calibration scale. A 3-insert configuration (equal angular spacing) was used in climb milling with a radial immersion of 63%. Preliminary experiments indicated that the different grades required different cutting speeds due to the wide range of hardness across the set of work materials. The cutting conditions used in face milling for the ADI grades and S-4340 are provided in Table 4.6 and Table 4.7, respectively. The ADI GR900, GR1050, and GR1200 were tested at speeds of 240 to 480 m/min, 240 to 480 m/min, and 120 to 360 m/min, respectively. The S-4340 tests were limited to cutting speeds of 240, 360, and 480 m/min. Chip load and depth of cut were held constant at 0.08 mm/tooth and 1 mm, respectively. For each cutting speed, a spindle speed and table feed rate was calculated using Equations 2.8 and 2.9. Surface roughness data was also obtained for each pass. A Mahr portable profilometer (Pocketsurf III) was used to collect the surface roughness data. The profilometer was set to a travel length of 5.0 mm and a cutoff wavelength of 0.8 mm. Five readings were taken for each surface measured. Cutting force experiments were conducted to determine machinability metrics related to machining energy/power and cutting stiffness. The cutting conditions used for the cutting force measurements are provided in Table 4.8. In these cutting tests, the chip load was varied between 0.05 and 0.15 mm/tooth to determine any dependence of cutting stiffness on chip load. The cutting speed (m/min) and depth of cut were held constant at 15 m/min and 1 mm, respectively, and 20% radial immersion was used. The cutting tool used was a single insert in an indexable holder. The holder was an indexable square- 44

61 shoulder end mill tool holder (19 mm diameter, Kennametal ) and the insert was a TiAlN-PVD coated carbide insert (Kennametal SDEB KC725M, 0 rake, 15 relief, nose radius of 0.1 mm), shown in Figure 4.9. Figure 4.10, the setup for the end milling operations, shows the force measurement system, which includes the dynamometer and charge amplifiers. Figure 4.11 shows the alignment of the machine center axes (X1, Y1, and Z1) with the axes of the dynamometer (Fx, Fy, and Fz). The alignment of the axes consisted of making X and Fx parallel and making Y and Fy parallel, which effectively aligned Fx and Fy with the machine center feed and normal directions, respectively. To minimize dynamometer charge drift, the charge amplifiers were in the reset position until moments before collecting data and a long time constant was used. Data was collected using a data acquisition system (NI cdaq-9178) through a LabVIEW user interface. A Matlab subroutine was coded to compute instantaneous values of chip thickness, cutting stiffness, and beta angle. 45

62 Figure 4.1: Results for XRD analysis at various grades of ADI, (a) GR900, (b) GR1050, and (c) GR1200 [49] 46

63 Figure 4.2: Tool-holder used in face milling experiments [50] Figure 4.3: Post preparation process workpiece for face milling test Figure 4.4: Dynamometer calibration setup 47

64 Figure 4.5: Results for dynamometer calibration 48

65 Figure 4.6: Experimental flow diagram Figure 4.7: Face milling experimental setup 49

66 Figure 4.8: Wear measurement setup with stereoscope and insert fixture Figure 4.9: End milling insert geometry [51] 50

67 Figure 4.10: End milling experimental setup with dynamometer and data acquisition system Figure 4.11: Machine center and dynamometer alignment diagram 51

68 Table 4.1: Chemical composition of pre-heat treat Ductile Iron Element Heat 1 Heat 2 [%] [%] Method C L S L P ISO1 Si ISO1 Mn ISO1 Cr ISO1 Ni ISO1 Mo <0.01 <0.01 ISO1 V < ISO1 Al ISO1 Cu ISO1 Mg ISO1 Ti ISO1 Sb <0.005 <0.005 ISO1 Ce ISO1 Sn <0.005 <0.005 ISO1 Table 4.2: Ferrite and austenite volume fractions Sample Ferrite Vol. Fraction 95% Confidence Austenite Vol. Fraction 95% Confidence GR GR GR Table 4.3: Austenite lattice parameter results Sample Lattice Parameter (A ) GR GR GR Table 4.4: Results for bulk hardness of grades of ADI, Brinell hardness Grade 900 HBW Grade 1050 HBW Grade 1200 HBW Steel 4340 HBW Average Standard Dev ASTM-A Range in HBW

69 Table 4.5: Dynamometer calibration results Mass [kg] Theoretical force [N] Measured force [N] Standard deviation [N] Table 4.6: Face milling study machining conditions for ADI grades GR1200 GR1050 GR900 # of Teeth Chip load [mm/tooth] Cutting Feed [mm/rev] Meters per Minute [m/min] Spindle Speed [Rpm] Feed Rate [mm/min] Depth of Cut Material Removal rate [cm 3 /min] # of Teeth Chip load [mm/tooth] Table 4.7: Face milling study machining conditions for S-4340 Cutting Feed [mm/rev] Meters per Minute [m/min] Spindle Speed [Rpm] Feed Rate [mm/min] Depth of Cut Material Removal rate [cm 3 /min]

70 # of Teeth Table 4.8: End milling study machining conditions for ADI grades and S-4340 Chip load [mm/tooth] (in/tooth) Meters per Minute [m/min]/ (in/min) Spindle Speed [RPM] Feed Rate [mm/min]/ (in/min) Depth of Cut / (in) Material Removal rate [cm 3 /min]/ (in 3 /min) (0.002) 15 (49.213) (0.4934) 1 (0.0394) (0.0029) (0.003) 15 (49.213) (0.7895) 1 (0.0394) (0.0047) (0.004) 15 (49.213) (0.9868) 1 (0.0394) (0.0058) (0.005) 15 (49.213) (1.2829) 1 (0.0394) (0.0076) (0.006) 15 (49.213) (1.4803) 1 (0.0394) (0.0087) 54

71 5 Results This chapter presents the results from the experiments conducted as part of the machinability investigations of ADI and S The experimental results are divided into four parts. The first part contains the analysis of chip formation as a result of the face milling experiments. The second part consists of the results from the end milling experiments, which were used to measure the cutting force and stiffness. The third, and fourth parts consists of results from the face milling experiments, which were used to measure the tool life and surface roughness, respectively. 5.1 Chip formation Figure 5.1 displays the effect of cutting speed on the chip form for the ADI grades and S-4340 after machining at cutting speeds ranging from 240 to 480 m/min. The chips formed after machining the ADI grades were mostly discolored and varied in length as the cutting speed changed. The length of the chips produced after machining GR1200 and GR1050 increased as the cutting speed increased. The length of the chips produced after machining GR900 decreased as the cutting speed increased. In general, the chips produced from machining the ADI grades were flat and varied between short continuous chips and discrete chips. However, the machining of S-4340 produced much longer continuous chips, slightly increasing in length as the cutting speed increased. The chips formed after machining the S-4340 only displayed partial discoloration in small regions of the chip. 5.2 Cutting forces Figure 5.2 shows a comparison of force traces for a single cutting event (Fx, Fy, F) for the grades of ADI (GR900, GR1050, GR1200) and S-4340 at a chip load of 0.15 mm/tooth. The shapes of the curves are typical for a single tool conventional end milling operation. Cutting forces increase steadily as the tool rotates until the tool exits the workpiece and the forces fall rapidly to zero. Oscillation in the force trace at exit is due to system compliance. Figure 5.3 and Figure 5.4 depict the chip width, tooth angle, cutting stiffness and beta angle in addition to the cutting force. It should be noted that since conventional milling was used, it can be assumed that the maximum cutting forces are experienced at the exit angle of the cutter. Therefore, the maximum cutting force in the Y direction, corresponding to Y2 of the dynamometer axes shown in Figure 4.11, occurs at θ e. This was used to detect the end of the first cutting event in the force trace and used as the starting point of the data set to calculate dynamic chip width h(t), beta angle β(t), and cutting stiffness Ks, relative to the tool angle as a function of time. The exit angle θe 55

72 was found to be 53.13, using Equation 2.13, due to a tool radius of mm and a radial immersion of 3.81 mm, this is consistent with the observations in Figure 5.4b. The dynamic tooth angle ϕ(t) was computed using Equation The dynamic tooth angle was used to compute the dynamic chip width, shown in Figure 5.3c, using Equation 2.11, where ϕ = ϕ(t). In order to determine the cutting stiffness, shown in Figure 5.3d, the dynamic chip width h(t) was multiplied by the axial depth of cut (b), to find dynamic chip area, which was used in Equation The beta angle, the angle between resultant force and the normal force, was determined by first finding the angle between Fy and F, e.g., the alpha angle, using Equation 2.18, and plugging it into Equation The beta angle value exhibited approximately constant peak behavior, shown in Figure 5.3e. A single cutting event is isolated in Figure 5.4. The dynamic chip width, shown in Figure 5.4c, very closely mimics the shape of the cutting force, shown in Figure 5.4a, which indicate that the cutting forces are proportional to the chip width. The measured forces, Fx and Fy for the ADI grades and S-4340 are shown in Table 5.4 and Table 5.5 show the peak average values for the resultant cutting force and cutting stiffness, respectively, for the grades of ADI. The peak average values for the resultant cutting force and cutting stiffness for S-4340 are shown in Table 5.6. A general increase in the values of resultant cutting force, shown in Figure 5.5, can be seen as the chip load increases for all grades of ADI and S In general, GR1050 exhibited the largest resultant cutting forces of the ADI grades, ranging from 188 to 435 N. GR900 had the smallest forces, ranging from 169 to 399 N, across all chip loads. The resultant cutting force of GR1200 exceeded that of GR1050 only for chip load of 0.13 mm/tooth. S exhibited a slightly higher resultant cutting force than the ADI grades at lower chip loads up to 0.08 mm/tooth, but was similar to that of GR900 at the higher chip loads. In contrast to cutting force, cutting stiffness exhibited a decreasing trend as chip load increased, shown in Figure 5.6. The cutting stiffness values for GR1200, GR1050, and GR900 ranged from 4060 to 6094 N/mm 2, 3677 to 6344 N/mm 2, and 3829 to 6061 N/mm 2, respectively, as chip load decreased. Cutting stiffness for S-4340 was similar, ranging from 3888 to 6551 N/mm 2 as chip load decreased. To determine an average cutting stiffness, measurements in the range of 0.1 to 0.15 mm/tooth were used, as the behavior was steady in this range. Figure 5.7 shows the cutting stiffness values over this range of chip loads for which this average was determined. The average cutting stiffness for the work materials are as follows: GR900 ( N/mm 2 ), GR1050 ( N/mm 2 ), GR1200 ( N/mm 2 ) and S-4340 ( N/mm 2 ). The average beta angles for each material system were determined over the same 56

73 selected range, as in Figure 5.8. The average beta angles for the work materials are: GR900 (66.62 ), GR1050 (68.61 ), GR1200 (67.32 ) and S-4340 (69.7 ). 5.3 Tool life Tool life is widely considered the most important machinability metric. Useful tool life was defined as maximum flank wear penetration (VBmax) by uniform wear of 0.3 mm (average wear across all teeth) or localized wear of 0.5 mm (on any individual tooth). Average tool wear was measured in cutting length intervals of mm. In some cases, tool life was limited by the occurrence of catastrophic tool failure, rather than progressive tool wear. Catastrophic tool failure occurred several times at the lower cutting speeds when machining GR1200 and GR900. In all other test conditions, stable progressive tool wear was observed. Figure 5.9 and Figure 5.10 display examples of images taken of the flank wear of insert after machining GR1050 at 240 and 480 m/min (a= 1 mm and f = 0.08 mm/tooth), respectively, for a certain cutting length until tool failure. These wear profiles were typical tool wear profiles observed when machining other grades of ADI across all test conditions. Figure 5.11 displays VBmax as a function of effective cutting time for GR1200 over cutting speeds ranging from 120 to 360 m/min. Cutting speed expectedly had a major effect on tool wear rate, with higher cutting speeds yielding faster wear rates. A cubic polynomial fit was applied to the tool wear data to represent the tool wear progression. The fitted curves are accurate representations of the tool wear progression for cutting speeds 180, 240, 300, and 360 m/min due to the high R 2 values, as shown in Table 5.7. The cubic polynomial equations for GR1200 are also displayed in Table 5.7. For cutting speeds 180 to 360 m/min, wear rate followed mostly uniform and gradual progression of tool wear. For a cutting speed of 120 m/min, tool wear did not reach steady state prior to catastrophic tool failure. Figure 5.12 displays VBmax as a function of effective cutting time for GR1050 over cutting speeds ranging from 240 to 480 m/min. Similar to GR1200, cutting speed once again has a major effect on tool wear progression in the same manner. The R 2 values, shown in Table 5.7, for cutting speeds 300, 360, 420, and 480 m/min indicate an accurate representation of the tool wear progression by the fitted model. The cubic polynomial equations for GR1050 are also displayed in Table 5.7. The tool wear curve for a cutting speed of 300 m/min followed the expected behavior for most cost-effective machining operations in that the profile consists of a rapid increase in the wear rate at the beginning of cutting that is followed by a linear wear rate until rapid tool wear beyond a threshold cutting time. Figure 5.13 displays VBmax, as a function of effective cutting time for GR900 over cutting speeds ranging from

74 to 480 m/min. Unlike the other grades of ADI, GR900 exhibited significant variability in tool wear progression. The cubic polynomial equations and R 2 values for GR900 are displayed in Table 5.7. For S- 4340, a more limited test (240, 360, 480 m/min) was conducted. Figure 5.14 displays VBmax as a function of effective cutting time for S Once again, a cubic polynomial fit was applied to the tool wear data to model the tool wear progression, which exhibited very high R 2 values, shown in Table 5.7, for 240, 360, and 480 m/min. The test at a cutting speed of 240 m/min was terminated prior to tool failure due to the fact that it vastly outperformed the ADI grades with minimal tool wear of mm after over 30 minutes of machining. Figure 5.15, Figure 5.16, and Figure 5.17 display the comparative tool wear progressions for GR900, GR1050, and S-4340 at cutting speeds 240, 360, and 480 m/min. Tool life estimates were extracted from the tool wear progressions by computing the time or length machined when the tool reached the tool wear criterion established above. Summary tool life data is shown in Table 5.8 and displayed in Figure Figure 5.18a displays tool life in terms of cutting time and Figure 5.18b displays tool life in terms of cutting length. The effective cutting time computed is a function of the material removal rate and volume removed, thus taking into account the effect of cutting speed on the time interval between measurements. GR1200 exhibited tool life ranging from 43.8 to 2.2 minutes, with rapid decline in tool life from 120 to 180 m/min and more linear degradation with cutting speed from 180 to 360 m/min. Tool life for GR1050 and GR900 gradually decreased from 16.2 to 2.4 min and 22.7 to 3.9 min, respectively, as cutting speed increased. From these observations, it is clear that for cutting speeds common to all grades, GR900 exhibited the best tool life, followed by GR1050 and GR1200. Tool life in terms of cutting length for GR1200, GR1050, and GR900 ranged from 4957 to 760 mm, 3841 to 1070 mm, and 4726 to 1724 mm, respectively, as cutting speed increased. Tool life data for machining S-4340 is shown in Table 5.9 and also displayed in Figure The tool life for machining S-4340 at 360 and 480 m/min cutting speed was 10.9 and 3.4 minutes, respectively. The tool life of S-4340 at 360 m/min cutting speed, exceeded that of GR900 and GR1050 by 3.5 and 5.3 minutes, respectively. The differences between S-4340 and the GR900 and GR1050 were very small at 480 m/min cutting speed. The equation developed by Taylor to quantify the relationship between cutting speed and tool life (Equation 2.22) was calibrated for the ADI grades. The tool life tests above were used to determine the empirical constants, C and n, for each ADI grade. In order to derive the constants, the relationship between tool life and cutting speed were plotted on a log-log graph, shown in Figure 5.19, and Equation 2.22 was developed. From the linear fit of the data the constant n was extracted from the slope and the 58

75 constant C was extracted from the y-intercept, of the linear regression equation. The results of the fit are provided in Equation 5.4 (GR1200), Equation 5.5 (GR1050) and Equation 5.6 (GR900). From these relationships C varied between and , while n varied from to T GR1200 = V 1/0.381, R 2 = [Eq. 5.1] T GR1050 = V 1/0.376, R 2 = [Eq. 5.2] T GR900 = V 1/0.397, R 2 = [Eq. 5.3] 5.4 Surface roughness Figure 5.20, Figure 5.21, and Figure 5.22 display surface roughness for GR900, GR1050, and GR1200, respectively, as a function of cutting length and cutting speed. The data obtained for the effect of cutting speed on surface roughness for GR1200, GR1050, and GR900 are shown in Table 5.10 and displayed in Figure The average surface roughness values for S-4340, as a function of cutting speed, were obtained and are shown in Table 5.11 and also displayed in Figure The average surface roughness of GR1200 displayed an upward trend from 0.38 to 0.55 μm as the cutting speed increased. Surface roughness in machining GR1050 only exhibited a slight upward trend with cutting speed, from 0.36 to 0.49 μm. In comparison, surface roughness in machining GR900 increased from 0.36 to 0.47 μm when the cutting speed was increased from 240 to 360 m/min and decreased from 0.47 to 0.38 μm at 480 m/min. The average surface roughness observed for S-4340 at cutting speeds of 240, 360, and 480 m/min were 0.51, 0.41, and 0.5 μm, respectively. These values are not significantly different than those from the ADI grades at similar material removal conditions. 59

76 Figure 5.1: Effect of cutting speed on chip form for ADI grades and S-4340 (a = 1 mm, f = 0.08 mm/tooth) 60

77 Figure 5.2: Results for collected cutting forces, Fx and Fy, and calculated resultant cutting force F for the grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, (d) S-4340 (a = 1 mm, f = 0.15 mm/tooth, SS = rpm, and 20% radial emersion) 61

78 Figure 5.3: Results for cutting force calculations for GR1200 cut at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 62

79 Figure 5.4: Close-up of results for cutting force calculations for GR1200 cut at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 63

80 Figure 5.5: Effect of chip load on resultant cutting forces for grades of ADI and S-4340 Figure 5.6: Effect of chip load on cutting stiffness for all grades of ADI and S

81 Figure 5.7: Approximately constant portion of effect of chip load on the cutting stiffness at 1 mm Axial Depth of Cut 20% Radial Cut for grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, and (d) S

82 Figure 5.8: Approximately constant portion of the effect of chip load on the beta angle at 1 mm Axial Depth of Cut 20% Radial Cut for grades of ADI, (a) GR900, (b) GR1050, (c) GR1200, and (d) S

83 Figure 5.9: Example of tool wear development for ADI GR1050 at 240 m/min (a = 1 mm, f = 0.08 mm/tooth), at cutting lengths (a) 0 mm, (b) 1829 mm, (c) 2743 mm, and (d) 3658 mm, 63x magnification Figure 5.10: Example of tool wear development for ADI GR1050 at 480 m/min (a = 1 mm, f = 0.08 mm/tooth), at cutting lengths (a) 0 mm, (b) 610 mm, (c) 914 mm, and (d) 1219 mm, 63x magnification 67

84 Figure 5.11: Tool wear progressions for GR1200 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.12: Tool wear progressions for GR1050 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) 68

85 Figure 5.13: Tool wear progressions for GR900 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.14: Tool wear progressions for S-4340 at various cutting speeds, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) 69

86 Figure 5.15: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 240 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.16: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 360 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) 70

87 Figure 5.17: Tool wear progressions for GR900, GR1050 and S-4340 at Vc = 360 m/min, maximum wear values (VBmax) as a function of effective cutting time (a = 1 mm, f = 0.08 mm/tooth) Figure 5.18: Effect of cutting speed on tool life of for grades of ADI (GR1200, GR1050, GR900), a = 1 mm, f = 0.08 mm/tooth, (a) tool life in terms of cutting time, (b) tool life in terms of cutting length 71

88 Figure 5.19: Log-log tool life plot used to develop the Taylor tool life equation 72

89 Figure 5.20: Effect of cutting length on surface roughness (Ra) for ADI GR900 at various cutting speeds, (a) 240 m/min, (b) 300 m/min, (c) 360 m/min, (d) 420 m/min, and (e) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) 73

90 Figure 5.21: Effect of cutting length on surface roughness (Ra) for ADI GR1050 at various cutting speeds, (a) 240 m/min, (b) 300 m/min, (c) 360 m/min, (d) 420 m/min, and (e) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) 74

91 Figure 5.22: Effect of cutting length on surface roughness (Ra) for ADI GR1200 at various cutting speeds, (a) 120 m/min, (b) 180 m/min, (c) 240 m/min, (d) 300 m/min, and (e) 360 m/min (a = 1 mm, f = 0.08 mm/tooth) 75

92 Figure 5.23: Effect of cutting length on surface roughness (Ra) for ADI GR1050 at various cutting speeds, (a) 240 m/min, (b) 360 m/min, and (c) 480 m/min (a = 1 mm, f = 0.08 mm/tooth) 76

93 Figure 5.24: Effect of cutting speed [m/min] on surface roughness (Ra) for all grades of ADI (GR1200, GR1050, GR900) and S-4340, a = 1 mm, f = 0.08 mm/tooth Table 5.1: Cutting force measurements in the X direction of dynamometer for grades of ADI (GR1200, GR1050, and GR900) GR1200 GR1050 GR900 Chip load [mm/tooth] (in/tooth) Average cutting force, Fx [N] Standard Deviation [N] Average cutting force, Fx [N] Standard Deviation [N] Average cutting force, Fx [N] Standard Deviation [N] 0.05 (0.002) (0.003) (0.004) (0.005) (0.006)

94 Table 5.2: Cutting force measurements in the Y direction of dynamometer for grades of ADI (GR1200, GR1050, and GR900) GR1200 GR1050 GR900 Chip load [mm/tooth] (in/tooth) Average cutting force, Fy [N] Standard Deviation [N] Average cutting force, Fy [N] Standard Deviation [N] Average cutting force, Fy [N] Standard Deviation [N] 0.05 (0.002) (0.003) (0.004) (0.005) (0.006) Table 5.3: Cutting force measurements in the X and Y direction of dynamometer for S-4340 Chip load [mm/tooth] (in/tooth) Average cutting force, Fx [N] Standard Deviation [N] Average cutting force, Fy [N] Standard Deviation [N] 0.05 (0.002) (0.003) (0.004) (0.005) (0.006) Table 5.4: Effect of chip load on resultant cutting force for grades of ADI (a = 1 mm, V = 15 m/min) GR1200 GR1050 GR900 Chip load [mm/tooth] (in/tooth) Average resultant cutting force [N] Standard Deviation [N] Average resultant cutting force [N] Standard Deviation [N] Average resultant cutting force [N] Standard Deviation [N] 0.05 (0.002) (0.003) (0.004) (0.005) (0.006)

95 Table 5.5: Effect of chip load on cutting stiffness for grades of ADI (a = 1 mm, V = 15 m/min) Chip load [mm/tooth] (in/tooth) Average cutting stiffness [N/mm 2 ] GR1200 GR1050 GR900 Standard Deviation Average cutting stiffness [N/mm 2 ] Standard Deviation Average cutting stiffness [N/mm 2 ] Standard Deviation 0.05 (0.002) (0.003) (0.004) (0.005) (0.006) Table 5.6: Effect of chip load on resultant cutting force and cutting stiffness for S-4340 Chip load [mm/tooth] (in/tooth) Average resultant cutting force [N] Standard Deviation [N] Average cutting stiffness [N/mm 2 ] Standard Deviation [N/mm 2 ] 0.05 (0.002) (0.003) (0.004) (0.005) (0.006) Table 5.7: Tool wear progression cubic polynomial fit equations for work materials Work Material Cutting speed [m/min] Cubic polynomial fit equations R 2 GR x x x GR x x x GR x x x GR x x x GR x x x GR E 05x x x GR x x x GR x x x GR x x x GR x x x GR E 07x 3 2E 05x x GR x x x GR x x x GR x x x GR x x x S E 06x x x S x x x S x x x

96 Table 5.8: Effect of cutting speed on tool life for ADI grades in terms of cutting time and cutting length (a = 1 mm, f = 0.08 mm/tooth) Grade Cutting speed [m/min] Average Tool life, cutting time [min] Standard Deviation [min] Average Tool life, cutting length Standard Deviation Table 5.9: Effect of cutting speed on tool life in terms of cutting time for S-4340 (a = 1 mm, f = 0.08 mm/tooth) Cutting speed [m/min] Tool life, cutting time [min] Tool life, cutting length 240 > 35 >

97 Table 5.10: Average surface roughness (Ra) for grades of ADI (GR1200, GR1050, GR900), a = 1 mm, f = 0.08 mm/tooth Grade Cutting Speed Average Surface Standard [m/min] Roughness [µm] Deviation [µm] Table 5.11: Average surface roughness (Ra) for S-4340 (a = 1 mm, f = 0.08 mm/tooth) Cutting Speed [m/min] Average Surface Roughness [µm] Standard Deviation [µm] 81

98 6 Discussion The present study provided a detailed evaluation of ADI machinability characteristics during machining and a baseline comparison with respect to a conventional surrogate material, AISI/SAE 4340 steel. This chapter discusses implications of the results for the machining of ADI in production environments. 6.1 Machinability evaluation Conventional milling experiments at different chip loads were used to characterize the relationship between chip load, cutting forces, cutting stiffness, and beta angle. From these experiments, a primary observation of the present work is that cutting stiffness of all materials decreased as the chip load was increased. This relationship can be attributed to the size effect or ratio of the plowing force. Past research has found that for a given tool rake at high cutting speeds and chip loads, the cutting stiffness tends to be constant [1]. Researchers have also found that SPEC, thus cutting stiffness, decreases as the chip thickness increases which can be attributed to the increased ratio of the plowing force or size effect as the chip thickness decreases [1, 6, 10]. The plowing force refers to the friction force acting in the toolflank region that occurs when the tool material is deformed due to high stressed when sharp cutting tools are used. This deformation causes contact between the tool and workpiece surface over the area of the tool flank, which generates force acting on the tool edge that does not contribute to the removal of the chip. If the force generating the chip is held constant, at relatively small chip thickness the plowing force becomes a larger portion of the forces acting on the cutting edge. This results in a rapid increase in the cutting stiffness value as the chip thickness is decreased. Therefore, since the chip thickness is a function of chip load, the chip load can be assumed to have the same effect on the cutting stiffness as the overall chip thickness. Thus, cutting stiffness is also expected to decrease as the chip load is increased. Representative cutting stiffness values for GR900, GR1050, and GR1200 were found to be 4028, 4258, and 4398 N/mm 2, respectively. These cutting stiffness values are significantly outside the cast iron range (1000 to 1500 N/mm 2 ) provided by Sandvik [52], denoted with a K (Figure 6.1). The cutting stiffness fell within the hardened steel range (2500 to 5000 N/mm 2 ), denoted with an H. The cutting stiffness observed was also significantly higher than the cutting stiffness reported by De Carvalho, Montenegro, and Gomes [40], for GR900 ( N/mm 2 ) and GR1050 ( N/mm 2 ) resulting from a turning study investigating the effect of chip load ( mm/tooth) on cutting stiffness at different depths of cut (2-5 mm) and cutting speed of 80 m/min with no coolant. Note that Kienzle s method 82

99 (1952) for modeling the cutting force with respect to cutting stiffness, chip thickness, and chip width was used to calculate the cutting stiffness from the measured cutting forces. Additionally, a wider range of chip loads were studied during turning operations, resulting in very different force behavior. Therefore, the disparity between the cutting stiffness values can be attributed to the machining configuration and the chip loads investigated. Tool wear measurements provide insight into the machinability of ADI in face milling configurations. Figure 5.9 and Figure 5.10 display examples of images taken of the flank wear of insert after machining GR1050 at 240 and 480 m/min (a= 1 mm and f = 0.08 mm/tooth), respectively, for a certain cutting length until tool failure. These wear patterns were typical wear patterns observed when machining all the grades of ADI across all test conditions. Figure 5.9 depicts progressive wear typically observed at lower cutting speeds [1, 6]. When the cutting speed is fairly low, the cutting edge maintains its proper function at a broader range of cutting time and the abrasive wear progresses slowly, shown in Figure 5.9a and b [53]. This continues until a sufficient number of mechanical impact loading cycles occur, resulting in the tool coating being worn away. Abrasive wear accelerates causing tool failure, as seen in Figure 5.9c and d. It has been reported that the main wear mechanism in dry milling ADI 900 is a combination of abrasive wear and partial destruction of the cutting edge [35]. Similarly in this study with cutting fluid, primary failure modes observed were abrasive wear followed by breakage of the flank. The Taylor tool life relationship [12] was used to understand the parameters that describe the dependent relationship between cutting speed and useful tool life. The n values of the three grades were found to be fairly close, indicating mostly constant sensitivity of the tool material to changes in cutting speed. The tool material was most sensitive to cutting speed when machining GR900, n = 0.397, followed by GR1200, n = 0.381, and GR1050, n = The tool material (SECO TiSiN-TiAlN coated carbide tool) exhibited n values larger than TiC- or TiN-coated uncoated tungsten carbide (WC) tools, 0.3, and smaller than AL2O3-coated WC tools, 0.4, [6]. Therefore, in general, it can be assumed that the SECO TiSiN-TiAlN Nanolaminate coated carbide tool (when cutting ADI) is less sensitive to cutting speed than TiC- or TiN-coated WC tools, 0.3, and more sensitive than Al2O3-coated WC tools, 0.4. Katuku, Koursaris, and Sigalas [37] reported an n value of 0.51 and a C value of for the tool-workpiece combinations of PcBN and GR900 at 0.2 mm depth of cut and 0.05 mm/tooth chip load, with cutting speeds ranging from 50 to 800 m/min. Note that the purpose of the study was to investigate the toolworkpiece combination under finishing conditions in dry turning. The n value is mostly based on the tool material, PcBN, indicating lesser sensitivity of tool life to cutting speed in the cutting operation 83

100 studied. The C values for tool-workpiece combination of the SECO TiSiN-TiAlN Nanolaminate coated carbide tool and GR900 ( ), developed in this study, was 26% smaller than the tool-workpiece combinations of PcBN and GR900 developed by Katuku, Koursaris, and Sigalas [37]. Since similar material removal rates were studied, the disparity between C values can be attributed to the tool material and the machining operation studied. The Taylor tool life equations developed for the ADI grades were used to estimate cutting conditions that would provide desirable tool life values. Table 6.1 show the estimated cutting speeds that should yield tool life of 10, 30, and 60 minute, under similar cutting conditions (i.e. similar depth of cut, immersion, and chip load). When machining GR900, under similar cutting conditions, a cutting speed 330 m/min should provide 10 min of tool life, a cutting speed of 213 m/min should provide 30 min of tool life, and a cutting speed of 162 m/min should provide 60 min of tool life. When machining GR1050, under similar cutting conditions, a cutting speed 289 m/min should provide 10 min of tool life, a cutting speed of 191 m/min should provide 30 min of tool life, and a cutting speed of 147 m/min should provide 60 min of tool life. When machining GR1200, under similar cutting conditions, a cutting speed 207 m/min should provide 10 min of tool life, a cutting speed of 136 m/min should provide 30 min of tool life, and a cutting speed of 105 m/min should provide 60 min of tool life. If similar material removal rates are retained as the cutting conditions (depth of cut, immersion, and chip load) are manipulated, the various recommended cutting speeds should continue to provide the desired tool life. For example, as shown in Table 6.2, if the depth of cut is doubled, the cutting speed should be divided by approximately four in order to maintain constant material removal rate and should provide similar tool life. However, the effectiveness of constant material removal rate is limited to a small range of cutting condition because the influence of depth of cut, immersion, and chip load on tool life tend to be disproportionate. Therefore, further investigation would be needed to determine the specific influence of depth of cut, immersion, and chip load on tool life. 6.2 Surrogate material comparison S-4340 is a common automotive material system and its bulk hardness is between that of GR900 and GR1050, making it a useful surrogate material to baseline machinability of ADI material systems. S displayed similar behavior to the ADI grades with respect to both cutting force and cutting stiffness. As expected, S-4340, a hardened steel, exhibited a cutting stiffness within the hardened steel range [52], displayed in Figure 6.1 denoted with a circle, of approximately 2500 to 5000 N/mm 2 and 84

101 denoted with a H. The similar cutting stiffness between the three materials (S-4340, GR900, and GR1050) can be attributed to hardness, since the three materials all have similar hardness and cutting stiffness is proportional to hardness. When machining S-4340, similar to ADI, tool failure occurred at cutting speeds of 360 and 480 m/min in less than 10 minutes. S-4340 outperformed GR1050 and GR900 at the lower cutting speed, 360 m/min, but performed similarly to GR1050 and GR900 at the high cutting speed, 480 m/min. The main difference between the ADI grades and S-4340, prior to machining, was microstructure. After machining, the retained austenite in ADI can transform into much harder martensite under the thermomechanical loading in machining, which increases tool wear rate and reduces tool life. Since GR900 and GR1050 are susceptible to this transformation, due to their retained austenite content, it is possible the transformation may have occurred and degraded the tool life [48, 50]. However, further investigation would be needed to determine how much transformation occurred. In regards to S-4340, researchers have also found that a phase transformation, from pearlite or tempered martensite to untempered martensite, can occur during machining [61, 62, 63]. This transformation tends to occur throughout the machined surface, leaving little to no retained austenite, resulting in a hard and brittle martensite microstructure. Once again, further investigation would be needed to determine how much transformation occurred. Nevertheless, the similar tool life behavior between the three materials can be attributed to the hardness and phase transformation, since the three materials tend to undergo similar transformations that degrade tool life. 85

102 Figure 6.1: Acceptable range of cutting stiffness (specific cutting force) per material category [52] Table 6.1: Estimated cutting speeds for desired tool life of 10, 30, and 60 minutes for present study (a = 1 mm, f = 0.08 mm/tooth, 63% immersion) Grade Desired tool life [min] Estimated Cutting speed [m/min] Material Removal rate (cm 3 /min)

103 Table 6.2: Estimated cutting speeds for desired tool life of 10, 30, and 60 minutes for modified cutting conditions (e.g. depth of cut, immersion and chip load) with constant material removal rate Grade Depth of cut Radial Immersion Chip load [mm/tooth] Desired tool life [min] Material Removal rate [cm 3 /min] Estimated Cutting speed [m/min]

104 7 Conclusion and Future Work A lack of understanding regarding machinability of ADI in milling configurations has limited use of ADI in new system designs. The present study was undertaken to investigate machinability of several ADI grades (GR900, GR1050, GR1200) with respect to performance metrics in a milling configuration. Performance in the ADI system was compared to S Several methods were used to understand the effects of machining parameters on tooling performance in milling through a series of experiments measuring cutting forces, tool life, and surface roughness during processing. More specifically, the effects of cutting speed on tool life and surface roughness during face milling was investigated, as well as the effects of chip load on cutting force and cutting stiffness during end milling. All the material systems investigated demonstrated the characteristic rapid decrease of cutting stiffness as the chip load and chip thickness increased. It was found that S-4340 exhibited the largest cutting stiffness, followed by GR1050, GR1200, GR900, respectively. All the representative cutting stiffness values fell within the expected range of cutting stiffness for hardened steel. The uncharacteristically high cutting stiffness values, for the ADI cast irons, can be attributed to the hardness increase with heat treatment and possible strain-induced phase transformation from retained austenite to martensite. The tool wear analysis showed the characteristic relationship between cutting speed and tool life, where higher cutting speed accelerates tool wear. In the machining of ADI, the main flank wear mechanisms were shown to be abrasive wear and partial destruction of the cutting edge during aggressive machining. Chipping and cracking were shown to occur at all cutting speeds, indicating that the cutting configuration (i.e. climb milling) had a larger influence on wear mode. Further, discoloration and premature tool breakage indicate high temperatures at the tool-workpiece interface when aggressively machining ADI. In general, GR900 was shown to have the greatest tool life across common cutting speeds of GR1200 and GR1050. Similarly, S-4340 displayed slightly better tool life than GR900. In general, GR900 also exhibited the best surface finish of the test materials. Overall, the machinability of the ADI grades, in terms of cutting stiffness, tool life, and surface roughness, is very similar to that of S Additionally, Taylor tool life coefficients were derived for the grades of ADI investigated. The constant C was shown to be dependent on the grade of ADI, while the constant n was fairly similar for all grades investigated. Future extensions of this work should include the investigation of the alternative milling orientation, up (conventional) milling for face milling operations and down (climb) milling for end milling operations; 88

105 exploration of the effect of other cutting conditions, such as depth of cut and radial immersion, on the machinability of ADI; and extensive characterization of the surface integrity and martensite transformation experienced during the machining of the ADI grades and S

106 References [1] G. Boothroyd and W. Knight, Fundamentals of Machining and Machine Tools. [2] C. Saldana, S. Swaminathan, T. L. Brown, W. Moscoso, J. B. Mann, W. D. Compton and S. Chandrasekar, "Unusual Applications of Machining: Controlled Nanostructuring of Materials and Surfaces," Journal of Manufacturing Science and Engineering, vol. 132, pp to 12, [3] T. Udomphol, Machining of metals, Suranaree University of Technology, [4] A. Polishetty, "Machinability and microstructural studies on phase transformations in Austempered Ductile Iron," Doctorial Thesis, Auckland University of Technology, [5] M. C. Shaw, Metal Cutting Principles, New York: Oxford University Press, [6] D. A. Stephenson and J. S. Agapiou, Metal Cutting Theory and Practice, Boca Raton, FL: Taylor & Franics Group, LLC, [7] "Technical Support Page: Conventonal vs. climb milling," Innovative Tool Sales, Anaheim, CA, [8] T. L. Schmitz and K. S. Smith, Machining Dynamics: Frequency Responce to Improved Productivity, New York : Springer, [9] A. E. Bayoumi, G. Yucesan and D. V. Hutton, "On the Closed Form Mechanistic Modeling of Milling: Specific Cutting Energy, Torque, and Power," Journal of Materials Engineering and Performance, vol. 3, no. 1, pp , [10] G. Smith, Cutting Tool Technology, London: Springer-Verlag, [11] ISO 8688, Interational Standard, [12] F. W. Taylor, On the are of cutting metals, New York: The American society of mechanical engineers, [13] P. Hu, C. Wilson and Z. J. Pei, "An experimental investigation into machinability of A refractory composite material," in IIE Annual Conference.Proceedings, [14] K. Brandenberg, "Machining austempered ductile iron.," Society of Manufacturing Engineering., vol. 128, no. 5, pp , [15] K. L. Hayrynen and J. R. Keough, "Austempered Ductile Iron-The State of the Industry in 2003," in Keith D. Millis Symposium, Louisville, [16] J. R. Keough and K. L. Hayrynen, "Automotive Applications of Austempered Ductile Iron (ADI): A Critical Review," Society of Automotive Engineers, [17] "ADI Applications," Austempered Ductile Iron Treatments, [Online]. Available: [Accessed 26 April 2013]. [18] T. Dorn, J. R. Keough, T. Schroeder and T. Thoma, "The Current State of Worldwide Standards for Ductile Iron," in Keith Millis Symposium on Ductile Cast Iron,

107 [19] A. Polishetty, G. Littlefair and S. Singamneni, "A comparative assessment of austempered ductile iron as a substitute in weight reduction applications.," in Proceedings of the ASME International Manufacturing Science and Engineering Conference, [20] B. V. S. Kovacs, "Austempered Ductile Iron: Fact and Fiction," Modern Casting, pp , [21] R. C. Voigt, "Austempered Ductile Iron - Processing and Properties," Cast Metals, vol. 2, no. 2, pp , [22] K. Hanzlikova, S. Vichet and J. Kohout, "Influence of microstructure composition on mechanical properties of austempered ductile iron," Metallic Materials, vol. 46, no. 2, pp , [23] R. Harding, "The production, properties and automotive applications of austempered ductile iron.," Metallic Materials, vol. 45, no. 1, pp. 1-19, [24] K. L. Hayrynen, "The Production of Austempered Ductile Iron (ADI)," in World Conference on ADI, [25] A. Abedi, S. Marashi, K. Sohrabi, M. Marvastian and S. Mirbagheri, "The effect of heat treatment parameters on microstructure and toughness of austempered ductile iron (ADI).," Advanced Materials Research, Vols , pp , [26] Y. Kim, H. Shin, H. Park and J. Lim, "Investigation into mechanical properties of austempered ductile cast iron (ADI) in accordance with austempering temperature.," Materials Letters, no. 62, pp , [27] A. Vasko, "Microstucture and mechanical properties of Austempered Ductile Iron," Annals of the Faculty of Engineering Hunedoara, vol. 10, no. 53, [28] O. Eric, M. Jovanovic, L. Sidanin, D. Rajnovic and S. Zec, "The austempering study of alloyed ductile iron," Materials & Design, no. 27, pp , [29] U. Seker and H. Hasirci, "Evaluation of machinability of austempered ductile irons in terms of cutting forces and surface quality," Journal of Material Processing Technology, vol. 173, pp , [30] M. C. Cakir, A. Bayram, Y. Isik and B. Salar, "The effects of austempering temperature and time onto the machinability of austempered ductile iron," Materials Science & Engineering A, vol. 407, no. 1-2, pp , [31] J. R. Keough, K. L. Hayrynen and V. M. Popovski, "Continuing Development of the Science and Application of Austempered Ductile Iron (ADI)," World Foundry Congress, [32] J. R. Keough and K. L. Hayrynen, "Wear Properties of Austempered Ductile Irons," SAE International, [33] K. L. Hayrynen, K. R. Brandenberg and J. R. Keough, "Applications of Austempered Cast Irons," American Foundry Society (AFS) Transactions, vol. 02, no. 084, pp. 1-10, [34] H. Matsuoka, Y. Tsuda and H. Ono, "Fundamental Research on Hobbing of Austempered Ductile Iron Gear (Influence of Graphite Particle o Machinability)," JSME Internations Journal, vol. 46, no. 3, pp ,

108 [35] M. Goldberg, J. Berry, T. Graham and G. Littlefair, "Machinability assessment and surface integrity characteristics of austempered ductile iron (ADI) using ultra-hard cutting tools," Society of Manufacturing Engineers, pp. 1-20, [36] A. Akdemir, S. Yazman, H. Saglam and M. Uyaner, "The effects of cutting speed and depth of cut on machinability characteristics of austempered ductile iron," Journal of Manufacturing Science and Engineering-Transactions of the Asme, vol. 134, [37] B. Avishan, S. Yazdani and D. Jalali Vahid, "The influence of depth of cut on the machinability of an alloyed austempered ductile iron.," Materials Science & Engineering, no. 523, pp , [38] K. Aslantas and I. Ucun, "Evaluation of the performance of CBN tools when turning austempered ductile iron material," Journal of Manufacturing Science and Engineering, vol. 130, no. 5, pp , [39] F. Klocke, M. Arft and D. Lung, "Material-related aspects of the machinability of Austempered Ductile Iron," Production Engineering, vol. 4, pp , [40] K. Katuku, A. Koursaris and I. Sigalas, "Wear, cutting forces and chip characteristics when dry turning ASTM grade 2 austempered ductile iron with PcBN cutting tools under finishing conditions," Journal of Materials Processing Tech, vol. 209, no. 5, pp , [41] I. Sadik, K. Hamberg and H. Borgstrom, "The Performance of PVD Coated Grade in Milling of ADI 900," in CIRP International Conference on High Performance Cutting, [42] A. Kacal and M. Gulesin, "Determination of optimal cutting conditions in finish turning of austempered ductile iron using Taguchi design method," Journal of Scientific and Industrial Research, vol. 70, no. 4, pp , [43] K. Aslantas and I. Ucun, "The performance of ceramic and cermet cutting tools for the machining of austempered ductile iron," The International Journal of Advanced Manufacturing Technology, vol. 41, no. 7, pp , [44] A. Meena and M. El Mansori, "Study of dry and minimum quantity lubrication drilling of novel austempered ductile iron (ADI) for automotive applications," Waer, vol. 271, no. 9-10, pp , [45] A. Polishetty and G. Littlefair, "Assessment of machining characteristics of austempered ductile iron," Advanced Materials Research,, Vols , pp , [46] M. V. De Carvalho, D. M. Montenegro and J. D. Gomes, "An analysis of the machinability of ASTM grades 2 and 3 austempered ductile iron," Journal of Materials Processing Technology, vol. 213, no. 560, [47] C. Z. Wu, Y. J. Chen and T. S. Shih, "Phase transformation in austempered ductile iron by microjet impact," Materials Characterization, vol. 48, no. 1,, pp , [48] S. Daber and P. Prasad Rao, "Formation of strain-induced martensite in austempered ductile iron.," Journal of Materials Science, vol. 43, no. 1, pp , [49] C. Z. Wu and T. S. Shih, "Phase transformation and fatigue properties of austempered ductile irons," International Journal of Cast Metals Research, vol. 15, no. 2, pp ,

109 [50] S. Daber, K. Ravishankar and P. Prasad Rao, "Influence of austenitising temperature on the formation of strain induced martensite in austempered ductile iron.," Journal of Materials Science, vol. 43, no. 14, pp , [51] J. Garin and R. Mannheim, "Strain-induced martensite in ADI alloys.," Journal of Materials Processing Technology, Vols , pp , [52] A. H. Program, ASM Handbook, Volume 1: Properties and Selection: Irons, Steels, and High- Performance Alloys, Materials Park: ASM International, [53] P. S. Paul and A. S. Varadarajan, "Performance evaluation of hard turning of AISI 4340 steel with minimal fluid application in the presence of semi-solid lubricants," Journal of Engineering Tribologhy, vol. 227, no. 7, pp , [54] M. P. Nascimento, R. C. Souza, W. L. Pigatin and H. J. Voorwald, "Effects of surface treatments of fatigue strength of AISI 4340 aeronautical steel," International Journal of Fatigue, vol. 23, pp , [55] K. L. Hayrynen, "Appliedprocess.com," Applied Process, [Online]. Available: [Accessed 12 November 2013]. [56] E. Laitila, "Applied Process: Ferrite - Austenite Volume Fractions," Superior XRD Consulting, Houghton, [57] 2012 Milling catalog, SECO Tools, [58] "Kennametal," Kennametal, [Online]. Available: [Accessed 3 April 2014]. [59] "The specific cutting force," Sandvik Coromant, [Online]. Available: [Accessed 35 March 2014]. [60] A. Li, J. Zhao, H. Luo, Z. Pei and W. Zeming, "Progressive tool failure in high-speed dry milling of Ti-6Al-4V alloy with coated carbide tools," The International Journal of Advanced Manufacturing Technology, vol. 58, pp , [61] S. Akcan, S. Shah, S. P. Moylan, P. N. Chhabra, S. Chandrasekar and T. N. Farris, "Characteristic of White Layers Formed in Steels by Machining," in Proceedings of ASME, IMECE, Nashvile, TN. [62] Y. Matsumoto, M. M. Barash and C. R. Liu, "Cutting Mechanism During Machining of Hardened Steel," Material Science Technology, vol. 3, pp , [63] W. Kim and P. Kwon, "Phase transformation and its effect on flank wear in machining steels," Journal of Manufacturing Science and Engineering, vol. 124, pp ,

110 Appendix A. ADI cutting force data summery figures Summary: The following figures display the results from the cutting force data collection and associated calculations for each grade of ADI and S-5450 at each chip load investigated. Each figure includes the collected cutting forces, cutting stiffness, tangential cutting force, dynamic chip thickness, tooth angle during cutting, and beta angle during cutting, all as a function of time. Figure A.1: Results for cutting force calculations for GR900 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 94

111 Figure A.2: Results for cutting force calculations for GR900 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 95

112 Figure A.3: Results for cutting force calculations for GR900 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 96

113 Figure A.4: Results for cutting force calculations for GR900 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 97

114 Figure A.5: Results for cutting force calculations for GR900 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 98

115 Figure A.6: Results for cutting force calculations for GR1050 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 99

116 Figure A.7: Results for cutting force calculations for GR1050 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 100

117 Figure A.8: Results for cutting force calculations for GR1050 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 101

118 Figure A.9: Results for cutting force calculations for GR1050 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 102

119 Figure A.10: Results for cutting force calculations for GR1050 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 103

120 Figure A.11: Results for cutting force calculations for GR1200 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 104

121 Figure A.12: Results for cutting force calculations for GR1200 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 105

122 Figure A.13: Results for cutting force calculations for GR1200 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 106

123 Figure A.14: Results for cutting force calculations for GR1200 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 107

124 Figure A.1: Results for cutting force calculations for S-4340 at 0.05 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 108

125 Figure A.2: Results for cutting force calculations for S-4340 at 0.08 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 109

126 Figure A.3: Results for cutting force calculations for S-4340 at 0.10 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 110

127 Figure A.4: Results for cutting force calculations for S-4340 at 0.13 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 111

128 Figure A.5: Results for cutting force calculations for S-4340 at 0.15 mm/tooth chip load, 1 mm depth of cut, 20% radial cut, V = 15 m/min, (a) collected cutting forces, (b) tooth angle during cutting, (c) dynamic chip thickness, (d) cutting stiffness, and (e) beta angle during cutting, all as a function of time 112