Material properties affecting cutting forces Master’s thesis in Materials Engineering Alexander T. Bengtsson Daniel Johansson Department of Industrial and Materials Science CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2019 Master’s thesis 2019 Material properties affecting cutting forces ALEXANDER T. BENGTSSON DANIEL JOHANSSON Department of Industrial and Materials Science Chalmers University of Technology Gothenburg, Sweden 2019 Material properties affecting cutting forces ALEXANDER T. BENGTSSON, DANIEL JOHANSSON © ALEXANDER T. BENGTSSON, DANIEL JOHANSSON, 2019. Supervisor: Amir Malakizadi, Chalmers University of Technology Supervisor: Sören Hägglund, Seco Tools AB Examiner: Peter Krajnik, Chalmers University of Technology Master’s Thesis 2019 Department of Industrial and Materials Science Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Typeset in LATEX Printed by Chalmers Reproservice Gothenburg, Sweden 2019 iv Material properties affecting cutting forces ALEXANDER T. BENGTSSON, DANIEL JOHANSSON Department of Industrial and Materials Science Chalmers University of Technology Abstract The aim of this master thesis is to evaluate the feasibility of using readily available material properties to estimate the constants in the proposed models that describe cutting resistance and therefore cutting force. The study is carried out for two types of workpiece materials, each from a different ISO-group. The investigated materials are 316L, an austenitic stainless steel, and 100Cr6, a high carbon through harden- ing steel. 316L is delivered by two different suppliers while 100Cr6 is delivered in three different hardening conditions, where the latter significantly alters the mate- rial characteristics. The study includes characterization of the workpiece materials with activities in- cluding grain size estimation, inclusion analysis, tensile testing and hardness test- ing. Machining experiments are performed using a CNC-lathe and the cutting re- sistance is calculated based on the measured force response for a certain theoretical chip thickness. The data is generated by using the stepwise increased feed-rate test method. The relation between properties such as hardness and tensile strength with the cutting resistance is presented for the 100Cr6 material. Since there is a con- nection between the cutting resistance and the cutting force, it is thus feasible to calculate the cutting forces under arbitrary cutting conditions and for different tool geometries. It is also observed that, while the hardening condition of 100Cr6 has a significant effect on its cutting resistance, only a slight difference exists between 316L produced by different suppliers. Keywords: Cutting forces, cutting resistance, machining, cutting resistance model- ing, main cutting force modeling, 316L, 100Cr6, material characterization. v Acknowledgements First of all we want to show appreciation to our supervisors Amir Malakizadi and Sören Hägglund, without whose support this project would not have been possible. We would also like to thank Peter Krajnik for being our examiner during this project. We would also like to express our most sincere gratitude towards Seco Tools AB and the Department of Industrial and Materials Science at Chalmers University of Technology for trusting us with this project. The study included many activities that provided valuable hands-on experience for us that will be useful in our future careers. Furthermore we want to thank Seco Tools with everyone involved for supplying all the necessary tooling and workpiece materials as well providing general support. We also want to thank the Department of Industrial and Materials Science for pro- viding all the necessary lab equipment, including the CNC-machine itself. Special thanks to Håkan Millqvist from the department for helping us extracting samples for the material characterization. Alexander T. Bengtsson, Gothenburg, June 2019 Daniel Johansson, Gothenburg, June 2019 vii Contents List of Figures xiii List of Tables xvii List of symbols and abbreviations xxi 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 Specification of issue under investigation . . . . . . . . . . . . . . . . 2 2 Theory 3 2.1 Forces in metal cutting . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1.1 Analytical approach to forces . . . . . . . . . . . . . . . . . . 5 2.1.2 Empirical approach to forces . . . . . . . . . . . . . . . . . . . 5 2.1.2.1 Empirical models describing cutting resistance . . . . 7 2.2 Process recommendations . . . . . . . . . . . . . . . . . . . . . . . . 8 2.3 Cutting tool geometry . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3.1 Orthogonal cutting . . . . . . . . . . . . . . . . . . . . . . . . 11 2.3.2 Chip geometry . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3.2.1 Equivalent chip thickness . . . . . . . . . . . . . . . 12 2.4 Machinability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.5 Material properties related to metal cutting . . . . . . . . . . . . . . 14 2.6 Material structure and composition related to metal cutting . . . . . 15 2.7 Workpiece materials . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.7.1 Selected workpiece materials . . . . . . . . . . . . . . . . . . . 17 2.8 Previous studies on links between cutting force and materials . . . . . 18 3 Methods 21 3.1 Workpiece materials . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 3.1.1 316L workpieces . . . . . . . . . . . . . . . . . . . . . . . . . . 21 3.1.2 100Cr6 workpieces . . . . . . . . . . . . . . . . . . . . . . . . 22 3.2 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.2.1 316L sample preparation . . . . . . . . . . . . . . . . . . . . . 23 3.2.2 100Cr6 sample preparation . . . . . . . . . . . . . . . . . . . . 24 3.3 Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 ix Contents 3.3.1 Optical microscopy . . . . . . . . . . . . . . . . . . . . . . . . 25 3.3.1.1 Grain size estimation . . . . . . . . . . . . . . . . . . 26 3.3.2 Inclusion analysis . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.3.2.1 Identifying inclusions . . . . . . . . . . . . . . . . . . 27 3.3.2.2 Quantifying inclusions . . . . . . . . . . . . . . . . . 27 3.4 Monotonic testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.5 Hardness testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.6 Machining experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.6.1 Face turning 316L workpieces . . . . . . . . . . . . . . . . . . 33 3.6.1.1 Face turning 316L workpieces with recommended feeds 33 3.6.1.2 Face turning 316L workpieces without nose engage- ment . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.6.1.3 Face turning 316L workpieces with special inserts . . 34 3.6.2 Face turning 100Cr6 workpieces . . . . . . . . . . . . . . . . . 35 3.7 Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.7.1 Obtaining the average forces . . . . . . . . . . . . . . . . . . . 36 3.7.2 Force normalization . . . . . . . . . . . . . . . . . . . . . . . . 36 3.7.3 Cutting resistance as a function of equivalent chip thickness . 38 3.7.4 Investigating connection between material properties and cut- ting resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.8 Verifying result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.8.1 Machining experiments at Seco Tools . . . . . . . . . . . . . . 40 3.8.2 Approximating cutting resistance . . . . . . . . . . . . . . . . 40 3.8.2.1 Approximating based on a previous model . . . . . . 40 3.8.2.2 Approximating cutting resistance of 316L with em- pirical relation determined for 100Cr6 . . . . . . . . 41 4 Results 43 4.1 Material characterization . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.1.1 316L characterization . . . . . . . . . . . . . . . . . . . . . . . 43 4.1.1.1 Inclusions 316L . . . . . . . . . . . . . . . . . . . . . 43 4.1.1.2 Grain size estimation 316L . . . . . . . . . . . . . . . 44 4.1.1.3 Hardness testing 316L . . . . . . . . . . . . . . . . . 45 4.1.2 100Cr6 characterization . . . . . . . . . . . . . . . . . . . . . 46 4.1.2.1 Inclusions 100Cr6 . . . . . . . . . . . . . . . . . . . . 46 4.1.2.2 Microstructure 100Cr6 . . . . . . . . . . . . . . . . . 47 4.1.2.3 Hardness testing 100Cr6 . . . . . . . . . . . . . . . . 47 4.1.2.4 Monotonic testing 100Cr6 . . . . . . . . . . . . . . . 48 4.2 Machining result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.2.1 Face turning 316L result . . . . . . . . . . . . . . . . . . . . . 50 4.2.2 Face turning 100Cr6 result . . . . . . . . . . . . . . . . . . . . 53 4.3 Verifying cutting resistance . . . . . . . . . . . . . . . . . . . . . . . . 55 4.4 Empirical relations between material properties and cutting resistance 57 4.5 Approximating cutting resistance . . . . . . . . . . . . . . . . . . . . 59 4.5.1 Approximation of Cr1 for 100Cr6 based on a previous model . 59 x Contents 4.5.2 Approximation of Cr1 for 316L with constants based on the relation with hardness obtained from 100Cr6 . . . . . . . . . . 61 5 Discussion 63 5.1 Quality of the data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5.1.1 Force measurements . . . . . . . . . . . . . . . . . . . . . . . 64 5.1.2 Force normalization . . . . . . . . . . . . . . . . . . . . . . . . 64 5.2 Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.3 Modeling cutting resistance . . . . . . . . . . . . . . . . . . . . . . . 66 5.3.1 Approximation of the main cutting force . . . . . . . . . . . . 67 5.3.1.1 Approximation based on existing model . . . . . . . 67 5.3.1.2 Approximation based on only the Vickers Hardness . 67 6 Conclusion 69 6.1 Conclusions for the initial research questions . . . . . . . . . . . . . . 70 6.2 Emerging questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 6.3 Further studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Bibliography 73 A Microstructure 316L I B Microstructure 100Cr6 V C Inclusion analysis IX D Tensile testing 100Cr6 XI E Cutting forces 316L XV F Cutting forces 100Cr6 XVII G Average normalized forces 316L XIX H Average normalized forces 100Cr6 XXI I Scatter in force between repetitions XXIII J Cutting resistance plots 316L XXVII K Cutting resistance plots 100Cr6 XXXIII xi Contents xii List of Figures 2.1 Forces in turning. a) [2, pp. 105], Figure 3.4, b) [2, p. 56], Figure 3.2. 4 2.2 Example of a process window. This particular window generated by Seco’s web page for a specific insert geometry combined with 316L as material choice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3 Model of equivalent chip thickness as proposed by Woxén. [4, pp. 134], Figure 5-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4 Aspects that affects machinability. . . . . . . . . . . . . . . . . . . . . 14 3.1 The figure illustrates the radial cross section and how the 316L sam- ples are cut from the workpiece. . . . . . . . . . . . . . . . . . . . . . 24 3.2 The figure illustrates the radial cross section and how the 100Cr6 samples are cut from the workpiece. . . . . . . . . . . . . . . . . . . . 25 3.3 The figure illustrates how the samples for ISO-M are divided into areas at different radial depths for the grain size characterization. . . 26 3.4 The figure illustrates the area investigated for inclusions on the 316L samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.5 The figure illustrates the areas investigated for inclusions on the 100Cr6 sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.6 All feeds and cutting speeds used for the respective machining exper- iment. (316L special inserts refers to the machining of 316L with the special made ISO-P inserts). . . . . . . . . . . . . . . . . . . . . . . . 31 3.7 Schematical representation of face turning operation. [4, pp. 124] . . 32 3.8 The figure illustrates the experimental setup for the face turning of the 316L workpieces without the nose engaged. Note that the insert is censored. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.9 The figure illustrates the experimental setup for the face turning of the 100Cr6 workpieces as well as the 316L. Note that the insert is censored. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.10 Model of the plane definitions in turning. [2, p. 41], Figure 2.7. . . . 38 3.11 Flow chart schematically representing how the data for the cutting forces are treated towards calculating the cutting resistance. Note that Insert 3 is not included due to it not being considered. . . . . . 39 4.1 Graph that shows how the grain size varies with radial depth for the 316L workpieces from the two different suppliers. The average standard deviation over all data points is 0, 70µm for Supplier A and 0, 69µm for Supplier B, respectively. . . . . . . . . . . . . . . . . . . . 45 xiii List of Figures 4.2 The hardness variation with radial depth for the 316L workpieces. The average standard deviation over all data points is 4, 0HV for Supplier A and 3, 0HV for Supplier B, respectively. . . . . . . . . . . 46 4.3 Comparison of the microstructure of the 100Cr6 samples in three conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.4 Engineering stress-strain curves for the 100Cr6 workpieces. . . . . . . 49 4.5 The cutting resistance for the 316L workpieces, according to the Woxén-Johansson model. Plotted for all machining scenarios against the equivalent chip thickness, in Log-Log. . . . . . . . . . . . . . . . . 51 4.6 The main cutting force, according to the Woxén-Johansson model, for the 316L workpieces. Plotted for all machining scenarios against the equivalent chip thickness, in Log-Log (derived from the cutting resistance). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.7 The cutting resistance for the 100Cr6 workpieces plotted against equivalent chip thickness, in Log-Log. . . . . . . . . . . . . . . . . . . 54 4.8 The main cutting force for the 100Cr6 workpieces plotted against equivalent chip thickness, in Log-Log (derived from the cutting resis- tance). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.9 The cutting resistance of the 100Cr6 workpieces obtained at Seco’s facility plotted against equivalent chip thickness, in Log-Log. . . . . . 56 4.10 The main cutting force plotted against equivalent chip thickness, in Log-Log. Based on data from measured at Seco’s facility. . . . . . . . 57 4.11 The empirical relations between selected mechanical properties and Cr1 for 100Cr6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 4.12 The empirical relations between selected mechanical properties and Cr2 for 100Cr6, no curve fitting is attempted. . . . . . . . . . . . . . 60 A.1 Microscopy image showing microstructure of a 316L sample from Sup- plier A, captured on Sample 1 from a random spot within the radial depth of 0, 0 − 3, 5mm. . . . . . . . . . . . . . . . . . . . . . . . . . . I A.2 Microscopy image showing microstructure of a 316L sample from Sup- plier A, captured on Sample 2 from a random spot within the radial depth of 16, 0 − 19, 5mm. . . . . . . . . . . . . . . . . . . . . . . . . . II A.3 Microscopy image showing microstructure of a 316L sample from Sup- plier B, captured on Sample 1 from a random spot within the radial depth of 0, 0 − 3, 5mm. . . . . . . . . . . . . . . . . . . . . . . . . . . III A.4 Microscopy image showing microstructure of a 316L sample from Sup- plier B, captured on Sample 2 from a random spot within the radial depth of 16, 0 − 19, 5mm. . . . . . . . . . . . . . . . . . . . . . . . . . IV B.1 Microstructure of 100Cr6 in Annealed condition. . . . . . . . . . . . . V B.2 Microstructure of 100Cr6 in Hardening 1 condition. . . . . . . . . . . VI B.3 Microstructure of 100Cr6 in Hardening 2 condition. . . . . . . . . . . VII C.1 Example image showing inclusions on a 316L sample from Supplier B. Captured on Sample 2, closer to the center of the workpiece. . . . IX xiv List of Figures C.2 Example image of the sample from the 100Cr6 sample in annealed condition, captured on Area 2 close to the center of the tube wall. . . X D.1 Stress strain curve for 100Cr6 in Annealed condition. . . . . . . . . . XI D.2 Stress strain curve for 100Cr6 in Hardening 1 condition. . . . . . . . XII D.3 Stress strain curve for 100Cr6 in Hardening 2 condition. . . . . . . . XIII J.1 Comparison between the models when machining the 316L workpieces from Supplier A within the recommended feed window, plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXVII J.2 Comparison between the models when machining 316L workpieces from Supplier B within the recommended feed window, plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXVIII J.3 Comparison between the models when machining 316L workpieces from Supplier A without nose contact within recommended feeds, plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXIX J.4 Comparison between the models when machining 316L workpieces from Supplier B without nose contact within recommended feeds, plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXX J.5 Comparison between the models when machining 316L workpieces from Supplier A based on two different types inserts (the same used for 100Cr6), plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . XXXI J.6 Comparison between the models when machining 316L workpieces from Supplier A based on two different types inserts (the same used for 100Cr6), plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . XXXII K.1 Comparison between the models when machining 100Cr6 Annealed workpieces, plotted in Log-Log. . . . . . . . . . . . . . . . . . . . . . . XXXIII K.2 Comparison between the models when machining 100Cr6 Hardening 1 workpieces, plotted in Log-Log. The W/J extended model is not included here. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXXIV K.3 Comparison between the models when machining 100Cr6 Hardening 2 workpieces, plotted in Log-Log. The W/J extended model is not included here. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXXV xv List of Figures xvi List of Tables 3.1 Chemical composition in wt% from Supplier A of the 316L workpieces. 21 3.2 Chemical composition in wt% from Supplier B of the 316L workpieces. 22 3.3 Chemical composition in wt% for the 100Cr6 workpieces. . . . . . . . 22 3.4 Table showing the selected feeds for the feed steps for both machining operations for the 316L workpieces using Insert 4, in the unit [mm/rev]. 33 3.5 Table showing the selected feeds for the feed steps for both machining operations for the 316L workpieces with Insert 1, Insert 2 and Insert 3, in the unit [mm/rev]. . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.6 Table showing the selected feeds for the feed steps used for the 100Cr6 workpieces, in the unit [mm/rev]. . . . . . . . . . . . . . . . . . . . . 36 3.7 Table containing the constants for the approximation. As presented in [11]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 4.1 Table showing the average area fraction of inclusions on the images within the respective area at different depths comparing the two sup- pliers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.2 Table showing the average area fraction of inclusions on the images within each area at different depths on the sample of the annealed 100Cr6 workpieces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.3 Table containing the average hardness for each material condition followed by the respective standard deviation. . . . . . . . . . . . . . 48 4.4 Table over the different material parameters obtained through the tensile test for the 100Cr6 workpieces. . . . . . . . . . . . . . . . . . 49 4.5 Table containing the model parameters, according to the Woxén- Johansson model, for the cutting resistance of the 316L workpieces for all different experiments. . . . . . . . . . . . . . . . . . . . . . . . 52 4.6 Table containing the model parameters, according to the Woxén- Johansson model, for the cutting resistance of the 100Cr6 workpieces. 54 4.7 Table containing the model parameters for the cutting resistance of the 100Cr6 workpieces obtained at Seco’s facility. . . . . . . . . . . . 56 4.8 Table containing a comparison for the model parameters between those obtained at Chalmers and at Seco. . . . . . . . . . . . . . . . . 56 4.9 Table containing the constants for the empirical relations shown for 100Cr6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.10 Table containing the model parameters of 100Cr6 for the approxima- tion of the cutting resistance. . . . . . . . . . . . . . . . . . . . . . . 60 xvii List of Tables 4.11 Table containing the result of the approximation. The error is the comparison with the result obtained at Chalmers. . . . . . . . . . . . 61 4.12 Table containing the result of the approximation based on experi- ments at Seco’s facility. . . . . . . . . . . . . . . . . . . . . . . . . . . 61 E.1 The measured main cutting forces for the 316L workpieces when machining with recommended feeds and without nose contact re- spectively, using Insert 4, where each repetition is presented. Vc = 230m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . . . . XV E.2 The measured main cutting forces for the 316L when machining with special inserts where each repetition is presented. Vc = 230m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XVI F.1 The measured main cutting forces for the 100Cr6 Annealed workpieces where each repetition is presented. Vc = 150m/min, ap = 2mm. . . . XVII F.2 The measured main cutting forces for the 100Cr6 Hardening 1 work- pieces where each repetition is presented. Vc = 90m/min, ap = 2mm. XVIII F.3 The measured main cutting forces for the 100Cr6 Hardening 2 work- pieces where each repetition is presented. Vc = 60m/min, ap = 2mm. XVIII G.1 The normalized average cutting force for each feed level for 316L when machining with recommended feeds and without nose contact respectively, using Insert 4. Vc = 230m/min, ap = 2mm. . . . . . . . XIX G.2 The normalized average cutting force for each feed level for 316L when machining with special inserts. Vc = 230m/min, ap = 2mm. . . . . . . XIX H.1 The normalized average main cutting forces for each feed level for the 100Cr6 Annealed workpieces. Vc = 150m/min, ap = 2mm. . . . . . . XXI H.2 The normalized average main cutting forces for each feed level for the 100Cr6 Hardening 1 workpieces. Vc = 90m/min, ap = 2mm. . . . . . XXI H.3 The normalized average main cutting forces for each feed level for the 100Cr6 Hardening 2 workpieces. Vc = 60m/min, ap = 2mm. . . . . . XXII I.1 The variation between repetitions divided by the average value pre- sented in percent when machining the 316L workpieces with recom- mended feeds and without nose contact respectively, using Insert 4. Vc = 230m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . XXIII I.2 The variation between repetitions divided by the average value pre- sented in percent for machining the 316L workpieces with special inserts. Vc = 230m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . XXIV I.3 The variation between repetitions divided by the average value pre- sented in percent for the 100Cr6 Annealed workpieces. Vc = 150m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXIV I.4 The variation between repetitions divided by the average value pre- sented in percent for the 100Cr6 Hardening 1 workpieces. Vc = 90m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . . . . . XXIV xviii List of Tables I.5 The variation between repetitions divided by the average value pre- sented in percent for the 100Cr6 Hardening 2 workpieces. Vc = 60m/min, ap = 2mm. . . . . . . . . . . . . . . . . . . . . . . . . . . . XXV xix List of Tables xx List of symbols and abbreviations ap [mm] Depth of cut A [mm2] Chip area b1 [mm] Theoretical chip width be [mm] Equivalent chip width bmax [mm] Maximum allowed chip width C1 − Cutting force constant in Fc = C2 + C1 · h1 C2 − Cutting force constant in Fc = C2 + C1 · h1 Cp [J/K] Specific heat capacity Cr [N/mm2] Cutting resistance Cr1 − Cutting resistance constant in Cr = Cr1 + Cr2 he and Cr = Cr1 + Cr2 · he Cr3 Cr2 − Cutting resistance constant in Cr = Cr1 + Cr2 he and Cr = Cr1 + Cr2 · he Cr3 Cr3 − Cutting resistance constant in Cr = Cr1 + Cr2 · he Cr3 E [GPa] Young’s modulus ef [%] Elongation at rupture f [mm/rev] Cutting feed Fc [N ] Main cutting force Ff [N ] Feed force Fp [N ] Passive force h1 [mm] Theoretical chip thickness h2 [mm] Mean chip thickness he [mm] Equivalent chip thickness heW [mm] Equivalent chip thickness proposed by Woxén h0min [mm] Minimum orthogonal distance hemax [mm] Maximum allowed equivalent chip thickness HV [HV ] Vickers hardness kc [N/mm2] Specific cutting force kc1.1 - Constant in kc = kc1.1 · hmc 1 k [W/mK] Thermal conductivity l [mm] Contact length mc - Constant in kc = kc1.1 · hmc 1 rn [mm] Nose radius re [mm] Edge radius Rp [MPa] Yield strength Rp0.2 [MPa] Yield strength, at 0,2% plastic deformation Rm [MPa] Tensile strength Vc [m/min] Cutting speed xxi List of symbols and abbreviations α [◦] Clearance angle α′ - Constant in Equation 2.10 β′ - Constant in Equation 2.10 δ′ - Constant in Equation 2.10 εb [%] Elongation at rupture, in Equation 2.10 η′ - Constant in Equation 2.10 γ [◦] Rake angle γ′ - Constant in Equation 2.10 κ [◦] Major cutting edge angle κb [◦] Minor cutting edge angle λ [◦] Inclination angle ν ′ - Constant in Equation 2.10 ω′ - Constant in Equation 2.10 φ [◦] Shear plane angle ξ′ - Constant in Equation 2.10 xxii 1 Introduction This research project is intended to improve the models used when calculating rec- ommendations for cutting data applied for different types of workpiece materials. The work include analyses of material properties/characteristics, force measure- ments during machining experiments, data analysis and modeling. 1.1 Background This master’s thesis is carried out on the behalf of Seco Tools AB, a world leading provider of cutting tools. Seco is part of Sandvik Machining Solutions and they have their main facility located in Fagersta, Sweden. They offer metal cutting solutions for milling, stationary tools, holemaking and tooling systems. In addition to this Seco also offer software applications that based on cutting data guides the user to the best possible results. If the available cutting data is more precise and adapted depending on the choice of material, the cutting process will be more efficient. A more efficient cutting process leads to longer tool life and better surface quality of the machined material. 1.2 Aim The purpose of the master’s thesis is to evaluate the feasibility of using readily available material properties to estimate the constants in order to fine tune a model that describes the cutting resistance. The aim is to be able to predict the cutting forces during arbitrary cutting conditions. With good feasibility it means that this approach should be valid for at least similar materials to those included in the study. The choice of materials are from two ISO groups, one material from ISO-P and the other from ISO-M. The intention for ISO-P is to isolate the effect of mechanical properties by performing experiments on the same material subjected to different heat treatments. For ISO-M the focus is to vary the material supplier in order to study the effects of batch-to-batch variations of the same material on the cutting forces. The study is done by analyzing how specific material properties are connected to the cutting resistance during turning. The main intention for the work is to investigate if readily available material properties such as tensile strength, hardness, chemistry 1 1. Introduction and/or the physical microstructure can be used to calculate the constants in a pro- posed model describing the cutting resistance. The analysis includes measurement of static forces during primarily face turning operation. The forces are measured while varying the feed rates and maintaining all other parameters constant. By this method the energy per cut volume of ma- terial can be plotted, i.e. the cutting resistance. The aim is to predict the cutting forces by analyzing the connection between material properties and the constants that describe the cutting resistance. 1.3 Limitations The purpose of this master’s thesis is limited to validate the feasibility of using readily available material properties to estimate the cutting resistance. The study is limited to only two different materials, each from a separate material group. From the ISO-P group the material is limited to a type of bearing steel, 100Cr6, which is a high carbon through hardening steel. Meanwhile from the ISO-M group the material is limited to an austenitic stainless steel, 316L. Material characterization is limited to Environmental scanning electron microscopy, ESEM, Light optical microscopy, LOM, hardness testing and tensile testing. These are applicable methods to attain results regarding the material parameters that are believed to possibly affect the cutting resistance. Only static forces are measured during machining since only they are relevant to the models intended to be used in this study. This means that the dynamic forces that occurs during machining are not measured or accounted for in any way. Forces during the cutting tool’s entry and exit in the material are also not analyzed. The forces are measured using the unworn tools, thus the effect of tool wear on the forces are not investigated. 1.4 Specification of issue under investigation In the list below the general research questions are formulated. • How is the cutting resistance and consequently the main cutting force con- nected to the selected material properties? • Can the cutting resistance and thereby the main cutting force be approximated based on material properties? • Which material properties have sufficient impact on cutting resistance to be considered in the approximation? • Which established model should be used as the base for approximation? 2 2 Theory In the following sections, theories of metal cutting and turning in particular are presented as well as theories regarding material properties/characteristics related to metal cutting and information about the materials included in this study. 2.1 Forces in metal cutting Knowledge about the forces in metal cutting is relevant for several reasons. From a practical point of view it is for example used to estimate the power needed for the process. Cutting forces can also be used to estimate the machinability of a material and is the easiest and most well-established response to measure. The methods for measuring the forces have been developed and improved during the last century. All methods used for measurement are based on the deflection of the tool during machin- ing with a dynamometer. In turning there is three force components in orthogonal directions. The direction of the force components in the case of longitudinal turning is illustrated in Figure 2.1a. The directions may change when changing the mode of turning, although they still remain orthogonal with each other. Furthermore in Figure 2.1b the force components in relation to the the cutting edge of the tool is presented. The components are the main cutting force (Fc), feed force, (Ff ) and the passive force (Fp). The main cutting force is in most cases the largest component and is acting on the rake face of the tool, perpendicular to the cutting edge. The feed force is acting parallel to the direction of the cutting feed and is thus normal to the main cutting force. The third component, the passive force is the smallest and often ignored component. [1, pp. 58] Cutting forces can have both a dynamic and a static character. Static forces, which is of focus in this study, can be treated as the average cutting forces over time. Mo- mentarily the forces vary as a result of local variations in cutting resistance which is caused by variations in the workpiece material. [2, pp. 103] The force required to form a chip is dependent on two major factors. The shear strength of the material under cutting conditions and the area of the shear plane. Provided that the latter remains constant, the force is then increased by any heat treatment or alloying that increases the yield and shear strength, more on this in 3 2. Theory (a) Force components in general turning. (b) Force components in relation to the chip. Figure 2.1: Forces in turning. a) [2, pp. 105], Figure 3.4, b) [2, p. 56], Figure 3.2. Chapter 2.5. In practice the area of the shear plane varies and is explained to be the most significant influence on the main cutting force often outweighing the influence of shear strength. During orthogonal cutting this area is geometrically related to the undeformed chip thickness, the width of cut and the shear plane angle, φ. The forces are directly affected by an increase in these parameters since they all affect the area. The shear plane angle can however not be controlled by the operator and is found to vary greatly under different conditions. When using a rake angle of zero degrees, it has been found that the shear plane angle may vary from a maximum of approximately 45 degrees to a minimum of five degrees. Subsequently, when the angle is very small the shearing force becomes large, and as shown in the literature, up to five times larger than the minimum which occurs at 45 degrees. Thus, much of the research has been focused on predicting this parameter. [1, pp. 58-60] The other region where forces arise is in the flow zone which is on the rake face of the tool. In a case where the rake angle is zero degrees the feed force is simply the drag force which the chip exerts on tool as it flows away. The contribution to the feed force caused by friction in the non seized areas is believed to be relatively small during most cutting conditions. Thus, the feed force can be considered dependent on the product of the workpiece material shear strength and the area of seized contact on the rake face. The latter parameter is explained to be very difficult to determine though, mainly since it is impossible to observe this area during engagement. [1, pp. 60-62] As explained [1, pp. 40], it is important to understand that during seizure in metal cutting, relative movement continues. This is said to be possible because the area of seizure is small in combination with a sufficient force to shear the work material near the seized interface. The relative movement does not take place at the interface between the rake face and the workpiece material. This is because the force required 4 2. Theory to overcome seizure is typically greater than the force needed to shear the material adjacent to it. It is debated whether or not the shear strength of the material at the rake face is of much significance due to several reasons. One of which being inaccuracies in the measurement of contact length in combination with the extreme conditions present in this region. Thus, the values of shear strength at the rake face are unlikely to be the same as those present in the shear plane. 2.1.1 Analytical approach to forces To further understand the influence of the shear plane angle it is relevant to study the theories that attempt to describe this. These models are noted as Analytical, although Semi-Analytical would be more a accurate term since they still rely on experimental data for certain parameters. The first model which is considered one of the milestones in metal cutting theory is Merchant’s force circle. However, this model to predict forces is considered inadequately accurate, mainly due to the inap- propriate use of friction relationships which are only relevant to sliding conditions. Another model, later developed by Oxley and colleagues, accounts for parameters such as work hardening and treats the frictional problem as shear within a layer of the chip adjacent to the rake face of the tool. In this layer, near seizure conditions are said to exist with the velocity of the workpiece material approaching zero at the interface. [1, pp. 62-65] However, two major challenges or limitations are presented for Oxley’s model. Mainly, there is no good data available for the stress-strain relations near the interface be- tween the chip and the tool. In particular not for the amounts of strain, extreme strain rates, time and temperatures at which material is deformed in the flow zone at the interface. Another issue is, while the importance of the contact area is recog- nized, the estimations of this area rely on calculations of a mean contact length and the basis for this calculation are deemed to be inadequate. No alternative to exper- imental measurement of this area has been found and thus it present difficulties for the model. [1, pp. 67] 2.1.2 Empirical approach to forces As is evident, there are several difficulties attempting to determine the cutting forces with the semi-analytical approach. An alternative to that, employed by the indus- try, is predicting the forces with an empirical approach. This method is based on the use of constants and curve fitting. All the unknowns are covered by constants and coefficients that relies on experimental data. The static forces, as explained in Chapter 2.1 above, are mainly a function of chip area, the depth of cut and properties of the material being cut. Less significant, 5 2. Theory but not negligible, are geometrical aspects such as rake angle but also cutting speed and the temperature of the process. As mentioned, these complex relationships can be treated as constants in a function depending on the theoretical chip thickness, h1, see Equation 2.1. In the simplest of cases this can be modelled as a linear rela- tionship. This linear behavior is valid especially for applications with constant rake angle. [2, pp. 105-106] Fc = C2 + C1 · h1 (2.1) In many machining scenarios it is stated that the coefficient C1 can be expressed as constant. However, several factors contribute to an disproportional behavior. The coefficient may either be increasing or decreasing as a function of h1 which for example can be a result of deformation hardening of the material or built up edge. Thus, it is more accurate with this approach to express the main cutting force as above but including C1 as a function of h1, Equation 2.2. [2, pp. 112-113] Fc = C2 + C1(h1) · h1 (2.2) The main cutting force can be treated as a function of the cutting resistance. It is defined in [2, pp. 142] as the resistance per total area of chip that the workpiece material shows in a given application involving a particular cutting tool and cutting data. This means that cutting resistance is dependent on the cutting process and should thereby not be viewed as a simple material constant. It can be described as the energy consumption per cut volume of workpiece material, i.e. the specific cutting energy. This parameter is directly tied to the force acting on the tool. It is computed by dividing the main cutting force with the chip area, Equation 2.3. [2, pp. 142] Cr = Fc h1 · b1 (2.3) This simple equation can be extended by expressing Equation 2.2 introduced earlier in this section. This extended model is presented below in Equation 2.4. It consists of two parameters, Cr1 and Cr2, where Cr1 describes load and energy consumption on the rake face. The parameter Cr1 is thus linked to the material dependency. Whereas Cr2 describes load and energy consumption on the clearance face and is dependent on process and the micro geometry of the cutting tool. This means that Cr2 is dependent on the contact between workpiece material and clearance face. Thus, increased flank wear results in an increase in the Cr2 parameter. [2, pp. 142- 144] Cr = Cr1 + Cr2 h1 (2.4) 6 2. Theory As explained in [2, p. 144], the cutting resistance is preferably determined by use of the method involving incremental feed in a stepwise manner, generally known as feed step. Important remark is that the geometry of the tool remains constant during the process. [2, pp. 142-144] A term that is fundamentally the same as cutting resistance is the specific cut- ting force, kc which is introduced by Kienzle. The specific cutting force is often described by the exponential curve fitting of experimentally obtained data. As with cutting resistance, it is based on describing the dependence of theoretical chip thick- ness h1 by the use of two constants kc1.1 and mc, see Equation 2.5. It is pointed out that the specific cutting force cannot be assigned a physical attribute since it is a combination of effects from both the rake face and clearance face of the tool. [2, pp. 145] kc = kc1.1 · hmc 1 (2.5) The way of determining the cutting resistance or specific cutting force is by experi- mentally measuring the main cutting force. Approximate result can also be obtained by measuring the power consumption of the motor in the process. [2, pp. 146] It is pointed out that from a research perspective the empirical models are flawed in the sense that they all rely on original testing data. This is problematic since that data would be obtained under a certain condition which may not translate well to slightly different conditions. Even though the empirical model, as all other models, can be criticized, they can still be useful in the industry. Keeping in mind that they may not be perfectly accurate, they can be utilized as “starting points” for operation. [1, pp. 374] 2.1.2.1 Empirical models describing cutting resistance This study considers three models for describing the specific cutting force or cut- ting resistance that are commonly used in the industry. The two first models are presented earlier in Chapter 2.1.2 although here an extended version is also intro- duced. This chapter provides some comparison between these models based on the literature study. [3] The specific cutting force, or the pressure, in the cutting zone can be expressed as either the specific cutting force kc or the cutting resistance Cr. Both are defined as the force required per unit area of the chip. Kienzle, which is traditionally the most common model uses the specific cutting force kc. The models introduced more recently are the Woxén-Johansson model, Equation 2.6, and the extension of this model proposed by Hägglund, Equation 2.7, referred to as the Hägglund model or simply Woxén-Johansson extended model. 7 2. Theory Cr = Fc A = Cr1 + Cr2 he (2.6) Cre = Fc A = Cr1 + Cr2 · he Cr3 (2.7) In order to obtain the constants needed for Equation 2.6 and 2.7, a feed step proce- dure can be used. This procedure is termed SIFT (stepwise increased feed-rate test) and is explained in Chapter 3.6. By using this procedure it is possible to determine the constants Cr1, Cr2, Cr3, kc1.1 and mc. There are some differences between the three models presented here. One differ- ence that can be noted is mainly what occurs when the equivalent chip thickness he is equal to 1.0. What happens is that neither the mc constant in the Kienzle model, nor the Cr3 constant in the Hägglund model, have any effect. The specific cutting force or cutting resistance becomes equal to the value kc1.1 in the Kienzle model and Cr1 + Cr2 in the Woxén-Johansson and Hägglund model. A way to further analyze the differences between the models is to multiply each model with the chip area. The result from this multiplication can be expected to be the main cutting force Fc. Performing this multiplication while the equivalent chip thickness approaches 0 shows some distinct differences between the models. The Kienzle model would show that no energy is consumed when he = 0, which can not be true in either fact or theory, since the deformation of the surface consumes energy. This can be considered the main limitation of the Kienzle model. [3] According to [3], the error rates between the different models relatively complex. Kienzle model is shown to provide good curve fitting for some ISO-P materials that are brittle and have a strong tendency to strain hardening. The study also showed favourable curve fitting for some ISO-M, ISO-N materials, and for the difficult ma- terials Alloy 718 and Ti6Al4V. However, for the majority materials analyzed in the study, the Hägglund model was shown to be the most favourable for curve fitting. 2.2 Process recommendations In order to support the customer, tool suppliers typically offer a so called working range model. This is a process window for suitable ranges of depths of cut and feeds for a specific tool insert. Such model considers several parameters with one being the fact that different materials generates different mechanical loads. Generally, as is explained in [4, pp. 129], such model is difficult to develop. Therefore the aim of such model is to provide, perhaps not the best, but recommendations that are "good enough". The working range model is intended to provide a window of operation where the process works relatively good, but outside the boundaries the result may 8 2. Theory be sub-optimal and in some cases even catastrophic. An example of such window is presented below in Figure 2.2. Figure 2.2: Example of a process window. This particular window generated by Seco’s web page for a specific insert geometry combined with 316L as material choice. The boundaries in the working range have different explanations. As presented in [4, pp. 129], the left most boundary is defined by the minimum orthogonal distance, h0min, from the cutting edge which defines the minimum allowable feed with chang- ing the depth of cut. The right boundary, declining with feed, is defined by the maximum allowed equivalent chip thickness hemax while changing feed and depth of cut. The upper boundary is defined by bmax as a function of the insert geometry maximizing the depth of cut ap. The bottom limit is then instead the minimum depth of cut considering the nose radius. The right boundary is the maximum feed as function of the nose radius which then abruptly limits the right sloped boundary. Presented in [4, pp. 48], the main logical way of optimizing cutting data is con- sidered in three steps whilst considering physical constraints: • Maximizing the depth of cut. • Maximizing the feed. • Optimizing the cutting speed. An example of constraints that must be accounted for is force restrictions. Account- ing for such constraint may involve balancing the depth of cut with feed. There 9 2. Theory are however some constraints as to how much the feed can be increased. Provided that the absolute feed limit is not exceeded and the surface finish constraint is met, the main concern is the models based on chip thickness he. One of which being the model for main cutting force which is directly affected by the value of he. The maximum cutting force that corresponds to a maximum equivalent chip thick- ness can be related to the term breaking feed. As explained in [4, pp. 132] the maximum value of the chip thickness needs to be within the breaking feed since if exceeded, catastrophic failure will occur. 2.3 Cutting tool geometry Cutting tool geometry refers to the form and dimensions that characterize a tool. There are essentially two aspects for the term, namely macro and micro geometry. The latter refers primarily to the form of the cutting edge. It is essential for the process to have a favourable cutting geometry since it directly affects the results. [2, pp. 44] It should be mentioned that various sets of standards are used in the field of metal cutting. However, for this work the terminology presented in [2, pp. 46] is mainly used. The rake angle and the clearance angle, γ and α, which are commonly mentioned, depend on the cutting tool’s position in relation the workpiece. The rake angle affects chip radius so that by decreasing the rake angle the radius of the chip is also decreased. Regarding how the rake angle affects the forces, it affect the size and direction of the main cutting force. For example, increasing the rake angle will reduce the force on the insert but in the process also weakens the cutting edge. This parameter has a very direct effect on the dynamics of the process. The other angle, the clearance angle, is what provides the tool access to the surface of the workpiece and as a result gives it freedom to move to the new surface being generated in the process. This angle is affected by whether or not the engagement is internal or external. It is also affected by the feed since, if increased, the effective clearance angle decreases. It should be noted that in order to provide clearance, the feed angle needs to be smaller than the clearance angle. [2, pp. 46] The major and minor cutting edge angles, κ and κb, are also dependent on the positioning of the tool. As explained in [2, pp. 46] the theoretical chip thickness, h1, together with the theoretical chip width, b1, both are depending on this. In the case of orthogonal engagement κ is set to 90◦, more on this in Chapter 2.3.1. The inclination angle, λ, also has an effect on the main cutting force as well as the chip removal process. As explained in [2, pp. 47], the most beneficial loading scenario is obtained if this angle is negative. However, it should be noted that this also leads to chips being directed onto the workpiece. 10 2. Theory The edge radius re is commonly described as an ideal radius. However, in reality it frequently deviates from the true radius along the edge line as presented in [2, pp. 45]. The exact form of the edge line is what the term micro geometry often refers to. The other radius, introduced in [2, pp. 47], is the nose radius, rn which is the rounded point of the insert. In essence, this feature together with the feed, determines the theoretical surface roughness as well as the strength of the nose area. 2.3.1 Orthogonal cutting Even though Orthogonal cutting is not used in this study it is still relevant to under- stand the fundamental requirements and benefits of this. This is intended to provide insight into how that can alter the results and why it is generally the method of choice when setting up experiments. Orthogonal cutting is the basic process variant that is commonly used for research in the field of metal cutting. The conditions that are required for orthogonal cutting can be expressed in the following way according to [5, pp. 46]: • The cutting edge angle κ is 90◦. • The inclination angle λ is 0◦. • Only the major cutting edge is being engaged, i.e. the nose and the minor cutting edge is not in contact. Orthogonal cutting is also described in [1, pp. 24]. According to [1, pp. 24] or- thogonal cutting entails simplified conditions that is beneficial for laboratory in- vestigations. To achieve orthogonal cutting, the cutting edge must be straight, i.e normal to both the cutting and feed direction. If the workpiece is in the form of a tube and the wall thickness is the depth of cut, only the major cutting edge will be in contact with the workpiece. In this scenario the cutting speed varies along the edge of the cutting tool. Although, if the diameter of the tube is large enough the variation in speed is considered negligible. Orthogonal cutting can also be achieved on a planing or shaping machine where the material is in the form of a plate. The edge of the plate is machined in an orthogonal manner. However, a shaper or planer has limited cutting speed and time of continuous machining. Due to this, the lathe based method is more convenient. Three methods for orthogonal cutting are mentioned in [6, pp. 24] as the following. • Orthogonal plate machining (OPM), i.e. machining a plate in a milling ma- chine. • Orthogonal tube turning (OTT), i.e. end-cutting a tube wall in a turning setup. • Orthogonal disc machining (ODM), i.e. end-cutting a plate with a tool feeding in facing direction. 11 2. Theory 2.3.2 Chip geometry The most important aspect of the chip is the mean chip thickness, h2, which can also be described as t2. This parameter can for example be determined, as explained in [1, pp. 26], by measuring the length and weight of a chip. The other factors such a width are assumed equal to the depth of cut, ap. The density of the workpiece material is also assumed to be unchanged during chip formation. Inserting these values into the equation provides an estimation of the mean chip thickness. As is further pointed out, the chip thickness can never be smaller than the theoretical chip thickness h1. This can also be described as t1 and is in the case of orthogonal cutting equal to the feed. The ratio between h1 and h2 is called the chip thickness ratio, which is commonly occurring in the literature. As is presented in [1, pp. 27], a low value relates to a low shear plane angle, something that is briefly explained in Chapter 2.1 above. 2.3.2.1 Equivalent chip thickness The equivalent chip thickness is a parameter used to approximate the chip area. This parameter is of greater significance when the depth of cut, ap, is in the same order or smaller than the nose radius of the insert. It can, as presented in [2, pp. 69], to a lesser degree of accuracy be described by the following basic relation, where h1 is the theoretical chip thickness. A ≈ ap · f ≈ b1 · h1 (2.8) However, as further discussed in [2, pp. 69] this relationship often yields too large errors. In order to obtain a more accurate representation of the chip area Woxén introduced the equivalent chip thickness. This parameter is aimed to describe a theoretical average chip thickness for the length of the active part of the edge line. Since the active part of the edge line is never straight, unless orthogonal conditions apply, this length is curved in one end due to the nose radius. Fundamentally, what is done by this model is straightening out the nose radius which is allowing the chip area to be treated as a rectangle, see Figure 2.3. The model, as proposed by Woxén, is presented in Equation 2.9. It is noted as heW to distinguish it from another way of representing this parameter which is referred to as the true equivalent chip thickness. heW = AW lW = ap · f ap−r(1−cosκ) sinκ +κ · rǫ + f 2 (2.9) 12 2. Theory (a) Approximated area, Ae com- pared to the true cutting area, AD. (b) Length of the edge in cut di- vided in sections. Figure 2.3: Model of equivalent chip thickness as proposed by Woxén. [4, pp. 134], Figure 5-10. As is explained in [2, pp. 82] the true equivalent chip thickness accounts for a loss in accuracy that is inherent to Woxén’s approximation. The inaccuracy is amplified for large nose radii or small feeds. However, for the purpose of this study Woxén’s model is considered sufficient. 2.4 Machinability Machinability is a term frequently mentioned in subsequent chapters. Machinability is an ambiguous term that according to [2, pp. 391] can be described as: The workpiece materials behaviour in the cutting process and its influence on the machining result. The following table lists some process behaviours that are related to the machin- ability term. The table is derived from [2, pp. 393]. 13 2. Theory Power and cutting forces Chip formation Surface quality Environmental factors Tool deterioration Energy consumption Workpiece deformation Equipment deformation Clamping robustness Entry and exit damages Process stability Chip type and dynamics Chip form Hardness Micro geometry properties Topography Residual stresses Structure Chemical composition Dust generation Allergic reactions Process additives Sound level Uniform flank wear Localized wear Crater wear Chipping and flaking Crack formation Plastic deformation Diffusion and chemical reactions Figure 2.4: Aspects that affects machinability. Machinability refers to all the above mentioned aspects of the process. However, those that are most relevant for this study are the ones related to the power and cutting forces. Especially the energy consumption which is directly related the load and thereby the forces the cutting tool is subjected to. [2, p. 142] 2.5 Material properties related to metal cutting As presented in [6, p. 59], the term work hardening is described as the phenomenon where a materials strength increases during plastic deformation. This effect is a result of dislocation movement and when the amount of dislocations which in turn increases the encounters and interactions between them. These encounters and in- teractions impedes movement of the dislocations and thereby increase the resistance to plastic deformation. As is further explained in [2, p. 396], the work hardening of the material affects the properties of the chip and the surface of the material which leads to an increased load on the edge of the cutting tool. The adhesion that takes place between the workpiece and the cutting tool is one aspect that affects machinability and adds complexity to the cutting process. Em- pirically there is a connection between adhesion and ductility, where higher ductility entails greater adhesion. This can cause the removed material fusing to the cutting tool and in doing so, creating a built up edge. This phenomenon can be both bene- ficial and detrimental. The built up edge can provide protection to the cutting edge 14 2. Theory which increases the tool life. However, if the built up edge is removed at a high frequency the tool can suffer increased wear. The built up edge can also result in an increased rake angle, γ, which can lead to an favourable chip formation although it also worsens the surface quality. [2, p. 397] The thermal conductivity, k, of the workpiece material is important to the tem- perature of the process. The Specific heat capacity, Cp, of the workpiece material is the particular factor that has the largest relative effect on the process temperature. [2, p. 397] The Hardness of the material is typically directly connected to the deformation re- sistance. Higher deformation resistance increases the main cutting force and thereby the cutting resistance. Thus, increased hardness generally also results in increased main cutting force and cutting resistance. [2, p. 497] Multiphase materials can have an effect on the machinability due to the fact that different phases in a material can have largely varying properties. One phase can be very brittle while the other is more ductile. Achieving exceptional machinability is also greatly dependent on the structure distribution of different phases and particles in the material. The distribution of abrasive particles is for example something that can have a significant effect on the tool wear. Also, in the structure distribution of the material porosity can contribute to low machinability due to increased material flow in the periphery of the tool. Material porosity can also cause the tool to have varying contact with the material during cutting. [2, pp. 397-398] Chemical reactions between the workpiece material and the cutting tool is also something that can offer limitations to the machinability. Diffusion is an example of something that can be considered a chemical reaction. [2, pp. 397-398] 2.6 Material structure and composition related to metal cutting The previously mentioned material properties are governed by the material struc- ture. The previous occurrence of a specific material structure and associated proper- ties is something that is heavily controlled by the constituents of the material. The different alloying elements have a large influence on the process from a machinability point of view. In the subsequent part some alloying elements effects on steel alloys will be described. [2, pp. 412-420] A low alloyed steel with a low carbon content is characterized by large ductility and thereby adhesion to to the cutting tool. Additions of nickel (Ni) and cobalt (Co) increases the risk of material build up on the cutting tool. Additions of ele- ments that are carbide or oxide formers e.g. chromium (Cr), vanadium (V), tungsten (W) and aluminum (Al) can increase the wear on the cutting tool. There are also 15 2. Theory alloying elements that facilitates machinability by providing a lubricating effect or by favouring chip breaking. Lead (Pb) and sulfur (S) have a lubricating effect on the cutting tool while phosphorus (P) and sulfur (S) can improve chip breaking. The addition of manganese (Mn) can also improve machinability due to the formation of manganese-sulfide (MnS) if sulfur (S) is present, which also have a lubricating effect. Small additions of calcium (Ca) can form soft oxides that offer even better lubrication. [2, pp. 412-420] Furthermore, as presented in [7, pp. 766], inclusions can play an important role in the process of chip formation as they can serve several functions. Not only can they serve as lubrication for the process but they can as well become diffusion bar- riers which can isolate the tool from chemical wear. They can also act as localized stress raisers in shear plane and thus can cause a crack formation which will be beneficial for breaking the chips. The inclusions are also playing an active role in the flow zone where they contribute to shearing the material. In the case of non-metallic inclusions it is further discussed in [7, pp. 756] that the following aspects are deemed relevant for steel materials. The composition, number, size and morphology of the inclusions. Furthermore as presented in [7, pp. 760], the properties as hardness, deformability and distribution are also deemed important. Thermal conductivity k is of great importance to machinability. In a material with high k heat is transported away from the cutting zone into the rest of the workpiece and most importantly into the chips. In a case where a material has lower k, heat is instead transported into the insert. The balance of the thermal conductivity be- tween the workpiece material and the insert is thereby important to the function of the cutting process. The thermal conductivity of a material is largely a result of the material constituents. Adding alloying elements with low k like titanium (Ti) will decrease the overall thermal conductivity, while adding copper (Cu) which has high k will increase the overall thermal conductivity. [2, pp. 412-420] Inclusions or impurities in the material can be both added by choice or added un- intentionally. Inclusions added by choice are often there to improve machinability and steels with these type of inclusions are commonly called free machining steel. These types of steels have higher content of sulphur (S) and lead (Pb). Calcium can also be added to change oxides and sulfides into aluminates which are encapsulated in calcium sulfide. The machinability of stainless steels can also be improved with the addition of sulphur (S), lead (Pb) and calcium (Ca). [2, pp. 412-420] The cutting process produces lasting deformation and increase of hardness in the surface of the workpiece material due to work hardening. Austenitic steels, duplex stainless steels and steels with a high content of manganese are especially susceptible to work hardening. [2, pp. 412-420] 16 2. Theory Hardness is not an indication of a material machinability except within a narrow range of a material group with similar properties and composition. The strength of the material has a more direct effect on its machinability. Generally it can be said that the higher the strength of the material the more energy is required for processing. However, this statement is a basic simplification of a much more com- plex connection. There is no simple formula to calculate how the machinability is dependent on the strength of the material. The energy that is needed to separate the material is mostly transformed to heat and transported away from the cutting process area by the insert, workpiece material and chips. This means that the re- lation between the materials strength and thermal conductivity is of importance. Ductility is the property that causes the material to flow while its influenced by shear stress. The materials ductility can be indicated by its elongation at fracture, which can be measured. High ductility in materials can cause problems since the chips might not break in a adequate manner. If the material is also of high strength further difficulties can be experienced. [2, pp. 412-420] 2.7 Workpiece materials Workpiece materials are divided in to six major material groups to support the choice of cutting tool geometry, grade and cutting data. The material groups are in accordance with the ISO standard and all have unique properties in regards to machinability. The two material groups mentioned in the following parts are those that are included in this study. ISO-P is the largest material group and consist of steel alloys of slightly varying types. The types range from low alloy to high alloy together with steel castings, as well as ferritic and martensitic stainless steels. The machinability of ISO-P materi- als are generally good but may vary depending on material properties. [8] ISO-M consist of stainless steel alloys which have a chromium (Cr) content above 12%. These alloys can also include alloying elements such as nickel (Ni) and molyb- denum (Mo) and can exist in several different conditions e.g. ferritic, martensitic, austenitic and duplex phase. Common characteristics between all ISO-M materials are that there is significant heat generation, notch wear and that they are prone to built-up edge. [8] 2.7.1 Selected workpiece materials The selected ISO-P material, 100Cr6, is a through hardening steel which is mostly used for bearings and similar applications with rolling contact and high fatigue. The steel in its hardened condition has high hardness, strength and cleanliness which helps the material to withstand high cycle and high stress fatigue. It can also be used for other machine components requiring high strength and hardness. [9] 17 2. Theory The ISO-M material of choice, 316L, is an austenitic stainless steel. It is less suscep- tible to corrosion and pitting than more traditional nickel chromium stainless steels. 316L is characterized by high creep resistance, excellent formability as well as cor- rosion and pitting resistance. It also maintains high rupture and tensile strength at elevated temperatures. Some common applications for 316L varies from structural building components to industrial equipment and even cookware and cutlery. Gener- ally, any component or equipment that require corrosion resistance. 316L has lower carbon content compared to the regular 316 stainless steel. This slightly lowers the strength but makes the material resistant to sensitization during heat treatments and significantly easier to weld. [10] 2.8 Previous studies on links between cutting force and materials In a paper from Lund University, [11], the influence of the workpiece material prop- erties on the cutting forces are investigated. It focuses on modelling the cutting resistance as a function of the properties of the workpiece material. The aim of that study, as well as for this, is to be able to predict the forces without relying on experimental cutting. The model proposed by [11] suggests that four different material properties should be used in order to estimate the cutting resistance. These are, hardness, yield strength, elongation at rupture, and thermal conductivity. Cr1 = α′ · HV δ′ + β′ · Rν′ p + γ′ · εη′ b + ξ′ · kω′ (2.10) Where HV is the Vickers hardness [Kp/mm2], Rp is yield strength [MPa], εb is the elongation at rupture [%], and finally k as the thermal conductivity [W/mK]. The following, α′, β′, δ′, γ′, η′, ν ′, ξ′, and ω′ are constants. These constants are de- termined by experimentally obtained values for Cr1 together with knowledge about the material properties. When the experimental value for Cr1 is compared to the value generated by the model the constants are determined by minimizing the dif- ference between these values. The presented variation for the results by the model is roughly 13% for all 98 entities used in the study. The coefficient of variation used to evaluate the result in the study, [11], was determined by the following equation: V = 1 n n ∑ i=1 ∣ ∣ ∣ ∣ ∣ Cr1,input − Cr1,model Cr1,input ∣ ∣ ∣ ∣ ∣ (2.11) The author of the study presented in [11], discuss the possibility of obtaining a better model by modelling Cr1 for each material ISO-group individually. It is motivated by the fact that all the materials within a group exhibit similar properties in terms of 18 2. Theory machinability, even though their mechanical or thermal properties may vary. This was tried and the conclusion from their study is that it seems advantageous with this approach for some ISO-groups, mainly ISO P where the coefficient of variation was roughly 5%. The model was on the contrary worse for determining Cr1 in ISO-M and ISO-S. However, they further conclude that the general model for all workpiece materials appears to be comparatively good and thus it could be speculated that it might be appropriate for all scenarios. It should be pointed out that some of the error obtained is thought to be due to tribological characteristics on the rake face in combination with potential inconsistency of material properties of the workpiece samples. The final conclusion for this study is that the model are relatively accurate for estimating the cutting resistance. Even though the error is relatively small it is not negligible and the result should thus be viewed as only an estimation. 19 2. Theory 20 3 Methods Presented in this section are the methods used to acquire the results. 3.1 Workpiece materials Described in the following sections is the workpiece materials used for the study. They are as mentioned from two different ISO groups and thus vastly different in properties and morphology. The general description of the materials are presented in Chapter 2.7.1. 3.1.1 316L workpieces The workpiece material selected from the ISO-M group is 316L from two different suppliers, Supplier A and Supplier B. Even though the workpieces are of the same material they vary slightly in chemical composition and manufacturing process re- sulting in somewhat different material properties. Both 316L workpiece materials are, according to their material certificates, hot rolled and annealed. The bars are then peeled for both suppliers, meaning that they are machined to remove surface cracks, cooled layers of "skin", and oxides. From the certificate it also reads that for Supplier B, an additional polishing is added after peeling. Since both of them are hot rolled, the process parameters of this method have a direct effect on the grain size. The following annealing step also plays a role, mainly since the time and temperature of this impacts the resulting microstructure. The differences in alloying content, or chemical composition, also play an impor- tant role on the properties. Below in Table 3.1 the chemical composition as stated in the material certificate of Supplier A can be viewed. It can be compared to that of Supplier B which is presented in the following table, Table 3.2. C Si Mn Cr Mo Cu Ni P S N 0,015 0,58 1,79 16,69 2,05 0,53 10,14 0,030 0,027 0,072 Table 3.1: Chemical composition in wt% from Supplier A of the 316L workpieces. 21 3. Methods C Si Mn P S Cr Ni Mo N 0,012 0,29 1,71 0,033 0,025 16,82 10,10 2,04 0,047 Table 3.2: Chemical composition in wt% from Supplier B of the 316L workpieces. The alloying content can cause some tendencies for the process behavior as explained in Chapter 2.6. In addition to the low amount carbon in this material type, the rela- tively high amounts of nickel is expected to make the material prone to form a built up edge. Hence a higher cutting speed is preferred for these workpiece materials in order to avoid this. 3.1.2 100Cr6 workpieces The selected workpiece material from the ISO-P group is 100Cr6 which is a high car- bon through hardening steel commonly used in bearing applications. The material supplier states the chemical composition as presented in Table 3.3. This material is due to the relative high amount of carbon, based on Chapter 2.6, expected to have less ductility and adhesion. This causes the process being less prone to suffer from issues related to built up edge. C Si Mn P S Cr Ni Mo Cu Al Ti O 0,98 0,28 0,33 0,014 0,007 1,43 0,14 0,04 0,178 0,022 7ppm 3,6ppm Table 3.3: Chemical composition in wt% for the 100Cr6 workpieces. In the case of ISO-P, instead of varying the supplier of the material, its condition is varied by different heat treatments. For this study the material are supplied in three different conditions resulting in differences in microstructure and consequently also for properties such as the hardness. The three conditions are designated Annealed, Hardening 1, and Hardening 2. Where Hardening 2 is the most hardened condition. 3.2 Sample preparation To analyze the workpiece materials, accurate preparation of the samples is necessary. Specifically the sample preparation is performed in order to study the microstruc- ture of the material with microscopy. Different methods of sectioning and organizing the samples are used between the two ISO groups since the workpieces are of different geometries. The following chapters, Chapter 3.2.1 and 3.2.2, describes how the samples are extracted from the different workpieces. 22 3. Methods The cut samples are mounted in thermosetting bakelite hot mounting resin with carbon filler. The samples are mounted in such way that the exposed faces are from the circular cross sections of the workpieces. After mounting the samples the surface is ground and polished, creating an even surface suitable for microscopy. The samples are etched in order to reveal microstructural details that would oth- erwise not be visible. Features such as porosity, cracks, and inclusions are however visible by only polishing. A properly etched sample reveals properties such as grain size, segregation, as well as the shape, size, and distribution of the phases and inclu- sions. Other aspects such as mechanical deformation and thermal treatments may also be observed. The 316L samples are difficult to etch. Due to this the etching is performed elec- trochemically with Oxalic acid 10% at 10V . Since the 100Cr6 are on the contrary relatively easy to etch, they are simply exposed to Nital 3% by submerging the sample for a short period of time. The time is selected arbitrarily until the surface appears non-reflective, paying close attention that they do not become over etched. 3.2.1 316L sample preparation The ISO-M materials are supplied in cylindrical bars. Since the total depth of ma- chining for this material group is decided to be a maximum of 30mm, only samples until this radial depths are deemed necessary to analyze. Figure 3.1, illustrates how the samples are extracted from the 316L workpieces. 23 3. Methods Figure 3.1: The figure illustrates the radial cross section and how the 316L samples are cut from the workpiece. 3.2.2 100Cr6 sample preparation The ISO-P materials are supplied in a slightly different geometry, hollow cylindrical bars, with the reason being that they are hardened. Since the wall thickness is only 20mm the full radial depth can be covered in one sample. In Figure 3.2, it is illustrated how the samples are cut for all three workpieces of 100Cr6. 24 3. Methods Figure 3.2: The figure illustrates the radial cross section and how the 100Cr6 samples are cut from the workpiece. 3.3 Microscopy Described in these sections are the methods used for studying the microstructure of the radial cross sections. Different methods are employed depending on the specific features being studied. 3.3.1 Optical microscopy The Light Optical Microscope used is a LEITZ DMRX with a AxioCAM MRc 5 for image acquisition. This equipment is used to capture the microstructure, more specifically the grains. This is described in detail in the following chapter, Chapter 3.3.1.1. This is also used as part of the inclusion analysis in order to capture im- ages to complement and verify the result from ESEM, described below in Chapter 3.3.2.2. Note that the LOM-images of the inclusions are captured prior to etching. 25 3. Methods 3.3.1.1 Grain size estimation The grain size estimations are carried out with a method described in ASTM stan- dard E112 - 12. The method in this standard is the intercept method which involves placing a line over a plane of the microstructure and determining the number of intersections made between the line and the grain boundaries. Two diagonal lines are placed on each respective image from the LOM where the lines intersect at least 50 grain boundaries. The intersections are manually counted for each line and the average number of intersects for the two lines are determined. The number of intersections made with each line are divided by the length of the line which provides a value for the average grain size. The areas analyzed are presented in Figure 3.3. Each sample is divided into four radial depths, meaning that the increment is 3, 5mm for each area. One image is captured in a random manner within each grid square and two grid squares are an- alyzed at each increment in radial depth. This means that a total of two images are captured for each step from the surface in order to obtain an average for each level. The average grain sizes are then plotted for the respective radial depth providing an approximation of the radial variation in grain size. Figure 3.3: The figure illustrates how the samples for ISO-M are divided into areas at different radial depths for the grain size characterization. 26 3. Methods 3.3.2 Inclusion analysis The inclusion analysis is performed with an Environmental scanning electron mi- croscope, ESEM. This is carried out before etching since the etchant can alter the result. The equipment used is a PHILIPS XL 30 ESEM. 3.3.2.1 Identifying inclusions The Energy dispersive spectroscopy, EDS, is done with an INCA x-sight from Oxford Instruments. For the analysis, three random areas containing inclusions are stud- ied. The inclusions within each area are identified and the result is a graph with characteristic peaks where the the presence of certain peaks represents the presence of certain elements. At least three inclusions for a specific morphology are analysed in order to reach some degree of certainty that there is no variation between them. Even though a rel- atively small amount och inclusions are analyzed with EDS, a large area is visually inspected. Based on the morphology of those inclusions analyzed, the conclusion may be drawn that it is unlikely that there are any unidentified types of inclusions that constitutes a large volume fraction of the material. 3.3.2.2 Quantifying inclusions The relative amount of inclusions are captured in a systematic manner roughly in- spired by the ASTM E45-18a standard. This is done for both materials with the aim to observe how the amount and size of the inclusions vary with the radial depth of the workpieces. The inclusions are quantified by image capturing with back scat- tered electrons and the inclusions can be distinguished by their shape and contrast. A total of 25 images per sample area are captured at a magnification of 500x. Each sample area is a predetermined 5x5mm square positioned in the center of each sample for the 316L workpieces. The images are evenly spaced with five images cap- tured per row in an alternating right to left manner. The total captured area per square is 3, 6mm2 although it is sampled from a total area 25mm2. The placement of the analyzed areas can be viewed in Figure 3.4. 27 3. Methods Figure 3.4: The figure illustrates the area investigated for inclusions on the 316L samples. Since the geometry of the 100Cr6 samples are different and because the whole radial cross section can be covered in one sample, two areas are analyzed at different radial depth. The method is the same as for the 316L samples but the placements of the squares are different. This is illustrated in Figure 3.5. 28 3. Methods Figure 3.5: The figure illustrates the areas investigated for inclusions on the 100Cr6 sample. The images are analyzed with the open source software Image-J. The area covered by inclusions is selected by contrast, utilizing a threshold tool. The software then quantifies the area covered and the size distribution of the particles. Since this is done on both sample areas for the respective material it is thus possible to see if this changes with radial depth. 3.4 Monotonic testing The monotonic testing, i.e. the tensile testing, is carried out at Seco’s facility in Fagersta for the 100Cr6 workpieces. It is conducted in accordance with ISO 6892- 1:2016 with the exception that the specimens are 60mm in total length together with a gauge length of 22mm. The nominal diameter of the tests are 6mm. Two monotonic are performed for each hardening condition. 3.5 Hardness testing The hardness evaluation is performed with a Struers DuraScan 70 G5. The method used is the Vickers hardness test which involves creating an indent with a diamond pyramid-shape indenter. The diagonals of the indent is measured optically and based on the lengths of these a value for the hardness is obtained. The indents are made on the polished (and etched) surface of the samples. The specific test method HV10 is employed which refers to the applied load of 10kgf . 29 3. Methods For all samples a total of ten indentations on each are done. In order to detect a variation of hardness with the radial depth of the workpieces, see chapter 3.2, the total length of each sample is divided into five depth. Two lines of five equally spaced indentations are made on each sample in order to obtain two values for each depth. By using two indentations for each step it provides an average hardness at each level of radial depth. The lines are placed manually on the sample using the software meaning that they vary slightly in length. When the lines are drawn, the software distributes the marks for the indentations with equal spacing. Since the length may vary slightly between lines a variation is introduced in spacing as well. In addition to this, the position of the starting point is also manually placed potentially causing a slight offset between the rows of indents. Due to this, the spacing of indents as well the starting points, that together provides the radial depth, is approximated based on averages between lines. The radial starting depth is approximated to 1, 7mm and the average step length is calculated to 2, 8mm. The average hardness, from two data points, is plotted against the respective ra- dial depth in order to obtain a graph that approximates the hardness against radial depth. 3.6 Machining experiments The method used for all experiments to gather data about the cutting resistance is the feed step method. This method involves a stepwise increment of feed whilst maintaining constant cutting velocity, Vc, and depth of cut, ap. The maximum feed allowed for each material is for this study varied depending on the hardness. In addition to this the cutting speed is also varied between material, with a higher cutting speed for the more ductile workpieces. The cutting data is explained more in detail in the respective section however in Figure 3.6 the plan over feeds and cutting speeds is presented. 30 3. Methods Figure 3.6: All feeds and cutting speeds used for the respective machining exper- iment. (316L special inserts refers to the machining of 316L with the special made ISO-P inserts). All experiments that are carried out to generate the data for the study are done in face turning operation. This mode of turning involves feeding the tool in the radial direction, as is schematically shown in Figure 3.7. The machine used for all experiments is an EMCO Turn 365 which is a conventional three axis CNC-lathe. A Kistler force gauge module is used to measure all three force components in their respective direction. 31 3. Methods Figure 3.7: Schematical representation of face turning operation. [4, pp. 124] To assure good quality of the data, all measurements are performed with pristine cutting edges which eliminates variations induced by preceeding tool wear. All tests are carried out with coolant, emulsion with 6wt% mineral oil (Castrol CareCut S 600), measured with a refractometer. All experiments are repeated at least once in order to show repeatability and thus assure that no significant variation occurs. If the second repetition is within a rea- sonable tolerance of the previous it is considered good enough. The values used as data for analysis are the average values of forces obtained at each feed level of the machining experiments. The average value is calculated from the first five seconds of the experiment where the forces stabilize. The reason behind having a fixed time is mainly because the time of engagement is varying between all tests and that this should further reduce influence of tool wear. However, it should be noted that with increased feed a greater distance in radial direction is covered in the time used for the average forces. This means that if microstructural differences depending on radial depth are present in the workpiece, the effect of this can intro- duce a greater variation with the higher feeds. The same tool holder is used for all experiments meaning that the geometry of the tool holder is constant. The inserts are of four different variants with varying rake angles and nose radius as well as chip breakers. There are three special made ISO-P inserts which are called inserts 1, inserts 2 and inserts 3. There is one com- mercial ISO-M insert called inserts 4. Since the inserts themselves have no clearance angle, all angles except the rake angle is constant and thus carried over from the tool holder. The tool holder has a slight clearance angle together with a negative inclination angle. The cutting edge angle is however 90◦ meaning that the cutting edge is parallel to the axis of rotation in the case of face turning. 32 3. Methods The main cause of the inclination angle for the general process is that orthogo- nal cutting, as described in Chapter 2.3.1, will not take place. Thus the mode of cutting is for all tests is oblique cutting. 3.6.1 Face turning 316L workpieces 3.6.1.1 Face turning 316L workpieces with recommended feeds The first experiments in the study are carried out with commercial inserts used within their recommended application and process window. More information about this be found in Chapter 2.2 where this particular process window is presented in Figure 2.2. The feed steps are selected, with some margin, within the recommended minimum and maximum feed for a certain depth of cut while taking the cutting speed into account. The selected feed steps are presented below in Table 3.4. The cutting speed is set to 230m/min which is about the average recommended cutting speed for all feeds selected. The setup is identical to that of turning ISO-P and Figure 3.9 in Chapter 3.6.2 is thus representative. Feed 1 Feed 2 Feed 3 Feed 4 Feed 5 Feed 6 Feed 7 0,06 0,07 0,08 0,09 0,10 0,11 0,12 Table 3.4: Table showing the selected feeds for the feed steps for both machining operations for the 316L workpieces using Insert 4, in the unit [mm/rev]. Three repetitions are performed at each level of feed in order to obtain average values for the force components. For practical reasons this approach is only applied for the first experiments on the 316L workpieces using the commercial ISO-M tool, Insert 4. 3.6.1.2 Face turning 316L workpieces without nose engagement In order to avoid influences from having the nose in engagement, a setup involving equally spaced flanges are used where the width of the flange is the depth of cut. To achieve this setup, 20mm deep grooves are cut into the workpieces, spaced evenly to accommodate sufficient space for the nose in front of the flange being cut. This experiment is considered complementary for the conventional face turning experi- ments in order to study the effect on the result by not engaging the nose. It can also provide an indication of the degree of inaccuracy possibly introduced by using Woxén’s model for equivalent chip thickness. 33 3. Methods The same cutting conditions as presented in the previous chapter, Chapter 3.6.1.1 are used. This is done in order to have the ability to directly compare the results. The setup for this experiment is pictured in Figure 3.8. Figure 3.8: The figure illustrates the experimental setup for the face turning of the 316L workpieces without the nose engaged. Note that the insert is censored. 3.6.1.3 Face turning 316L workpieces with special inserts Tests are performed on the 316L workpieces with the same special made inserts and feeds as for the 100Cr6, described in Chapter 3.6.2. The same levels of feeds are selected for machining the 316L workpieces as for the annealed 100Cr6. The cutting speed is set to the higher value of 230m/min which is used for all experiments on this material. The feeds for this experiment are presented in Table 3.5 below. Feed 1 Feed 2 Feed 3 Feed 4 Feed 5 0,05 0,10 0,15 0,20 0,25 Table 3.5: Table showing the selected feeds for the feed steps for both machining operations for the 316L workpieces with Insert 1, Insert 2 and Insert 3, in the unit [mm/rev]. 34 3. Methods 3.6.2 Face turning 100Cr6 workpieces Three special made inserts are used for the machining experiments of mainly the 100Cr6 workpieces from ISO-P. However, they are also used for ISO-M as explained in the precious chapter, Chapter 3.6.1.3. The special made inserts are called In- sert 1, Insert 2 and Insert 3. The reason why these are special made is to have the same grade (coating and substrate) for all inserts. These inserts are based on commercial geometries as well as grade, although they do not exist commercially in these combinations. This is considered important for assuring good quality of the data since tribological effects from the grade would be constant between tests. The intention behind varying the geometry and chip breakers is to observe the effect of this and to calculate an average cutting resistance between the normalized cutting forces. This would further reduce the impact by variations that the normalization does not account for, more in this in Chapter 3.7.2. It should be noted that these inserts have cutting edge reinforcements where the length of which depends on the variant. Since the feeds are always less or equal to the length of these edge reinforcements, limited contact with the insert is assumed where the angle of these effectively acts as the rake angle. For one geometry however the last feed exceeds the length of the edge reinforcement. In Figure 3.9, the setup is shown. The mounted workpiece material is 100Cr6 from ISO-P, however the image is representative for the 316L workpieces from ISO-M as well. Figure 3.9: The figure illustrates the experimental setup for the face turning of the 100Cr6 workpieces as well as the 316L. Note that the insert is censored. 35 3. Methods The workpieces are machined in the order of hardening degree and with a different cutting speed and maximum feed depending on the level of hardness. The selected window of feed for each condition is divided into five levels. The reason for subse- quently lowering the maximum feed is to avoid problems with vibrations and plastic deformation and excessive wear on the cutting tools since much greater forces are expected for the hardened workpieces. The selected levels for the feeds are presented in Table 3.6 below. In addition to lowering the maximum level of feed, the cutting speed is also lowered for increasing hardness. The cutting speed is lowered from 150m/min for 100Cr6 Annealed, down to 90m/min for Hardening 1 and the lowest value of 60m/min for the hardest condition, Hardening 2. Condition Feed 1 Feed 2 Feed 3 Feed 4 Feed 5 Annealed: 0,05 0,10 0,15 0,20 0,25 Hardening 1: 0,05 0,09 0,13 0,16 0,20 Hardening 2: 0,05 0,08 0,10 0,13 0,15 Table 3.6: Table showing the selected feeds for the feed steps used for the 100Cr6 workpieces, in the unit [mm/rev]. 3.7 Data analysis 3.7.1 Obtaining the average forces The average forces for each combination of workpiece material, insert geometry, and feed level is determined. This is necessary in order to be able to study the static forces generated by said combinations. The raw data for the force measurements together with time is imported and refined in MATLAB. A script is written that saves the average forces based on a manually selected point in the force-time graph where the forces have stabilized. From the selected point, and five seconds ahead, the average force is calculated and saved. Since two repetitions are carried out, the average between two tests are also calcu- lated. This is saved in a vector and is plotted against the respective feed. The result of this is then normalized to eliminate the effect of insert geometry as explained in the following chapter. 3.7.2 Force normalization To further refine the data and to obtain a result independent of tool geometry, or rather a normalized result that can be transformed for other tool geometries than for those used in the study, a transformation matrix is employed. This method is described in detail in [2, p. 117-127]. By determining the load function from the 36 3. Methods measured forces for each cutting tool combination and by knowledge of the geome- try of the tools, the forces can be transformed to a value which can be considered a baseline. From this baseline forces can be approximated for another geometry if desired. This method is applied for all cutting force data in the study. The baseline geometry where all data is normalized to is presented in the following list. • Rake angle, γ = 0◦ • Clearance angle, α = 0◦ • Inclination angle, λ = 0◦ • Major cutting edge angle, κ = 90◦ Load functions are determined from experimental data where the force is plotted against the equivalent (theoretical) chip thickness, heW . Ideally the data points fall on a straight line and the forces as a function of theoretical chip thickness is ex- pressed as a linear equation. This is carried out for all three force components, Fc, Ff and Fp. Based on the load functions, the force components can be computed in their re- spective direction for any tool geometry that is defined in terms of clearance angle, rake angle, major cutting edge angle, and the inclination angle. For more details see [2, p. 123]. The illustration presented in Figure 3.10 indicates the normal planes that the force components are transferred to. This is done by multiplying the force components with transformation matrices that are calculated for each tool geometry. In the end this provides a cutting force component that is in the same plane for all different tools. Some assumptions are being made for the modeling. For example when per- forming the load transformations, the clearance angle is assumed zero degrees for reasons explained in [2, p. 123]. 37 3. Methods Figure 3.10: Model of the plane definitions in turning. [2, p. 41], Figure 2.7. 3.7.3 Cutting resistance as a function of equivalent chip thickness For each insert and material combination the average force response, now indepen- dent of insert geometry, can for a certain feed be plotted against the theoretical chip thickness using Equation 2.1. Since the nose of the insert is in engagement the equivalent chip thickness, heW , is used for a more accurate representation of this pa- rameter. Note that this equation does not account for build up edge or deformation hardening. See Chapter 2.1.2 for a more detailed description. The result presented in 4.2 is the Cutting resistance as a function of equivalent chip thickness. The cutting resistance is directly dependant on the cutting force since it is defined as the static main cutting force divided by the chip area. The result of this can, as explained in the previous chapter, then be transformed for a specific insert geometry. Since the aim of this study is to analyze how the cut- ting resistance, and consequently the main cutting force, depends on the material properties, only the average cutting resistance based on the normalized forces as a function of equivalent chip thickness is further used. How the data is treated in this study is schematically shown in Figure Flowchart special inserts. The figure illustrates the data for the non commercial inserts used for all workpiece materials. The same approach is applied for the standard inserts used for the 316L workpieces as well, following only one branch. 38 3. Methods Figure 3.11: Flow chart schematically representing how the data for the cutting forces are treated towards calculating the cutting resistance. Note that Insert 3 is not included due to it not being considered. 3.7.4 Investigating connection between material properties and cutting resistance Since the refined data from the previous steps, normalized for a baseline tool geom- etry, the cutting resistance is compared between materials. Comparing the cutting resistance with material is however not trivial, mainly since the term material can be divided into several parameters. Thus it is relevant to analyze which parameters have the strongest correlation with the cutting resistance and by that the main cut- ting force. 39 3. Methods 3.8 Verifying result 3.8.1 Machining experiments at Seco Tools In order to verify the result, a series of experiments are performed at Seco’s facility in Fagersta. In this case the operation is changed to longitudinal turning and no cutting fluid is used. Furthermore a different tool holder and different type of inserts are used. In addition to this, a different machine is obviously used, which in this case is more stable. The experiments are carried out for the 100Cr6 workpieces in all three conditions. The feed step method is carried out in a different manner. A more rapid method is used, instead of making a new cut for each feed level with a pristine edge, the machine is programmed to increase the feed after a set distance meaning that all feeds are covered in one consecutive cut. This has the major benefit of being a faster way of generating data, with the drawback that the tools can wear out during the experiment from one feed step to another.. Relative difference = Cr1,Seco − Cr1,Chalmers Cr1,Chalm