Academia.eduAcademia.edu

Figure from:Experimental and finite element analysis of machinability of AL-6XN super austenitic stainless steel

See full PDFdownloadDownload figure
Fig. 6 Applying boundary conditions on the model parts boundaries using section assignment tool as shown in Fig. 5. The benefit of using a meshed microstructure during the sim- ulation is to identify the real deformation (which is reflected in the stress, strain, and temperature values and their distribution) in the workpiece microstructure ahead the cutting tool tip.

Figure 6 Applying boundary conditions on the model parts boundaries using section assignment tool as shown in Fig. 5. The benefit of using a meshed microstructure during the sim- ulation is to identify the real deformation (which is reflected in the stress, strain, and temperature values and their distribution) in the workpiece microstructure ahead the cutting tool tip.

Related Figures (23)

Table 1 Chemical composition of the AL-6XN alloy Johnson-Cook constitutive equation was used in both simula- tions to represent the material plastic behaviour. The stress, strain and temperature values as well as chip formation obser- vation were obtained through the macro-simulation. The micro-simulation was used to reveal the influence of loads on the morphology and orientation of the voids created due to macro-simulation. Klocke et al. evaluated the stress in regard to strain and temperature effect during the simulation process [31]. Ben Moussa, Sidhom and Braham used the material prop- erties of AISI 316L ASS alloy in a calibrated Johnson-Cook model to calculate stresses according to the applied cutting parameters [8]. The FEA results (cutting forces, cutting tem- perature and chip morphology) were compared to experimental results to assess the validation of the calibrated model. Ozel and Zeren conducted research on modelling the machinability test of the AISI 1045 steel [32]. A Johnson-Cook model was utilised to calculate the elastic-plastic deformations in the chip shear zones. Stress and temperature distribution around the rounded tool tip were the main results gained in their work. Umbrello, M’Saoubi and Outeiro applied a FEA study on the machinability of AISI 316L stainless steel alloy [33]. They explored the outcomes when five various groups of material constants were used as material inputs to estimate cutting forces, temperature and chip formation during the simulation process. A correlation between the numerical and experimental outcomes was carried out based on the material constants used in the FEA study. Other recent research studies on FEA in metal cutting of other alloys and materials have also been reviewed and conducted by researchers [34-39].
Table 1 Chemical composition of the AL-6XN alloy Johnson-Cook constitutive equation was used in both simula- tions to represent the material plastic behaviour. The stress, strain and temperature values as well as chip formation obser- vation were obtained through the macro-simulation. The micro-simulation was used to reveal the influence of loads on the morphology and orientation of the voids created due to macro-simulation. Klocke et al. evaluated the stress in regard to strain and temperature effect during the simulation process [31]. Ben Moussa, Sidhom and Braham used the material prop- erties of AISI 316L ASS alloy in a calibrated Johnson-Cook model to calculate stresses according to the applied cutting parameters [8]. The FEA results (cutting forces, cutting tem- perature and chip morphology) were compared to experimental results to assess the validation of the calibrated model. Ozel and Zeren conducted research on modelling the machinability test of the AISI 1045 steel [32]. A Johnson-Cook model was utilised to calculate the elastic-plastic deformations in the chip shear zones. Stress and temperature distribution around the rounded tool tip were the main results gained in their work. Umbrello, M’Saoubi and Outeiro applied a FEA study on the machinability of AISI 316L stainless steel alloy [33]. They explored the outcomes when five various groups of material constants were used as material inputs to estimate cutting forces, temperature and chip formation during the simulation process. A correlation between the numerical and experimental outcomes was carried out based on the material constants used in the FEA study. Other recent research studies on FEA in metal cutting of other alloys and materials have also been reviewed and conducted by researchers [34-39].
The quick-stop device used to generate the frozen chip roots samples has a geometry consisted of three major parts (shooter
The quick-stop device used to generate the frozen chip roots samples has a geometry consisted of three major parts (shooter
2.2 Temperature measurements Temperature measurements during the cutting process were evaluated using an infrared thermal camera. Thermal imagers are devices that calculate the emitted infrared energy from heated objects and convert this into a thermal picture. Micro Epsilon’s Thermo-IMAGER TIM 160 was used for tempera- ture analysis during the cutting process, as seen in Fig. 4. The thermal camera has the ability to measure temperature within the range —20 to 900 °C. The infrared camera was controlled using the thermal imager TIM connect software (Release. 2.9.2147.0). The camera was attached to the lathe facing the cutting tool and the workpiece contact zone. When the cutting process was started, the cutting temperature was recorded and measured for 20 s during the cutting process. These measure- ments were repeated five times per cutting trial to ensure pre- cision, and the average temperature values were used during the analysis stage.
2.2 Temperature measurements Temperature measurements during the cutting process were evaluated using an infrared thermal camera. Thermal imagers are devices that calculate the emitted infrared energy from heated objects and convert this into a thermal picture. Micro Epsilon’s Thermo-IMAGER TIM 160 was used for tempera- ture analysis during the cutting process, as seen in Fig. 4. The thermal camera has the ability to measure temperature within the range —20 to 900 °C. The infrared camera was controlled using the thermal imager TIM connect software (Release. 2.9.2147.0). The camera was attached to the lathe facing the cutting tool and the workpiece contact zone. When the cutting process was started, the cutting temperature was recorded and measured for 20 s during the cutting process. These measure- ments were repeated five times per cutting trial to ensure pre- cision, and the average temperature values were used during the analysis stage.
The finite element (FE) analysis was performed using ABAQUS 6.14-1 software to measure stresses, plastic strain, the temperature of the cutting process and the residual stresses beneath the cutting tool. In this modelling, simulation of the real microstructure was considered. The alloy microstructure obtained using the EBSD analysis was exported to OOF2 software (v. 2.1.12) in order to create the mesh. OOF2
The finite element (FE) analysis was performed using ABAQUS 6.14-1 software to measure stresses, plastic strain, the temperature of the cutting process and the residual stresses beneath the cutting tool. In this modelling, simulation of the real microstructure was considered. The alloy microstructure obtained using the EBSD analysis was exported to OOF2 software (v. 2.1.12) in order to create the mesh. OOF2
Fig. 4 Setup of the infrared thermal camera soltware 1s tree software provided by the National Institute of Standards and Technology. The software is used to read and simulate the real image of a material microstructure [40]. Recently, the software was used in the machining field to simulate a two-phase metal cutting process [41]. OOF2 helps to create a real meshed microstructure instead of building a traditional meshed model using ABAQUS software drawing tools. The EBSD microstructure image of the workpiece were coloured in two colours: white (grains) and black (grains boundaries). The reason behind colouring the image is that the OOF2 software has the ability to define and recognise the alloy phases, grains and grains boundaries from their col- ours. When the selection process of the grains and the grain boundaries were finished, the elements within each part were accounted and assigned later to the meshed parts separately. The meshing process was started and repeated as iterations until the adequate size of the elements is reached. Uniform mesh size was created first across all the microstructure image regardless of the size of the grains and the grain boundaries, then the mesh size starts to change after few iterations (more information on the mesh size and meshing iteration procedure are found in [42]). Therefore, the mesh size and type (triangle or rectangle) were automatically generated according to the meshed part size. In other words, rectangle-shaped meshed elements were created inside the grains whereas fine triangle-shaped meshed elements were found inside the grain boundaries due to their small size compared to the grains’ size. After completing the meshed parts, these meshes were exported to be used in ABAQUS software where the material properties were assigned to the grains and the grains’
Fig. 4 Setup of the infrared thermal camera soltware 1s tree software provided by the National Institute of Standards and Technology. The software is used to read and simulate the real image of a material microstructure [40]. Recently, the software was used in the machining field to simulate a two-phase metal cutting process [41]. OOF2 helps to create a real meshed microstructure instead of building a traditional meshed model using ABAQUS software drawing tools. The EBSD microstructure image of the workpiece were coloured in two colours: white (grains) and black (grains boundaries). The reason behind colouring the image is that the OOF2 software has the ability to define and recognise the alloy phases, grains and grains boundaries from their col- ours. When the selection process of the grains and the grain boundaries were finished, the elements within each part were accounted and assigned later to the meshed parts separately. The meshing process was started and repeated as iterations until the adequate size of the elements is reached. Uniform mesh size was created first across all the microstructure image regardless of the size of the grains and the grain boundaries, then the mesh size starts to change after few iterations (more information on the mesh size and meshing iteration procedure are found in [42]). Therefore, the mesh size and type (triangle or rectangle) were automatically generated according to the meshed part size. In other words, rectangle-shaped meshed elements were created inside the grains whereas fine triangle-shaped meshed elements were found inside the grain boundaries due to their small size compared to the grains’ size. After completing the meshed parts, these meshes were exported to be used in ABAQUS software where the material properties were assigned to the grains and the grains’
The resulting meshed model obtained from OOF2 was exported as an orphan mesh to ABAQUS in order to complete the model setup and apply the boundary conditions as well as the cutting parameters. The total number of the elements cre- ated in OOF2 was 73,965. These elements consisted of 34,401 CPE4RT four-node plane strain thermally coupled linear quadrilateral, bilinear displacement and temperature, reduced integration and hourglass control inside grains elements. The remaining 39,564 elements were three-node plane strain ther- mally coupled triangle, linear displacement and temperature and located within and around the grain boundaries. It should be mentioned that the behaviour of the grain boundaries could be similar to the behaviour of amorphous materials, and co- hesive element-type COH2D4 might be more suitable during the simulation and would enhance the expected results. However, as mentioned above, the imported meshed micro- structure is an orphan mesh. The element types within this orphan mesh were limited to the elements used in the current work. In addition, the cohesive elements do not consider the temperature measurements and this is also proved and men- tioned in [43].
The resulting meshed model obtained from OOF2 was exported as an orphan mesh to ABAQUS in order to complete the model setup and apply the boundary conditions as well as the cutting parameters. The total number of the elements cre- ated in OOF2 was 73,965. These elements consisted of 34,401 CPE4RT four-node plane strain thermally coupled linear quadrilateral, bilinear displacement and temperature, reduced integration and hourglass control inside grains elements. The remaining 39,564 elements were three-node plane strain ther- mally coupled triangle, linear displacement and temperature and located within and around the grain boundaries. It should be mentioned that the behaviour of the grain boundaries could be similar to the behaviour of amorphous materials, and co- hesive element-type COH2D4 might be more suitable during the simulation and would enhance the expected results. However, as mentioned above, the imported meshed micro- structure is an orphan mesh. The element types within this orphan mesh were limited to the elements used in the current work. In addition, the cohesive elements do not consider the temperature measurements and this is also proved and men- tioned in [43].
Table 3 Tensile test parameters When the original JC coefficients of the AL-6XN al- loy (Table 2) were directly applied in the simulation pro- cess of this work, the stress values obtained were unre- alistic. The reason behind the unrealistic stresses could
Table 3 Tensile test parameters When the original JC coefficients of the AL-6XN al- loy (Table 2) were directly applied in the simulation pro- cess of this work, the stress values obtained were unre- alistic. The reason behind the unrealistic stresses could
De ALUIOUULEG LO UIE PLOUEss POLLO alld Usea Wl Geliv~ ing the JC coefficients, as these constants were previous- y used to simulate other operations than machining to evaluate the equivalent plastic strain only. Therefore, a MATLAB Curve fitting tool was applied to calibrate the JC equation depending on the strain-stress curve of an executed tensile test. The engineering stress and strain parameters obtained by an executed tensile test listed in Table 3 were used for the curve-fitting process to obtain the JC coefficients (A and B) at a strain rate of 0.00027 s'. Table 2 lists the modified JC coefficients (A and B). Figure 7 shows the original and modified JC curves as well as the experimental curve. Tables 4 and 5 list the main properties of the workpiece and the cutting tool (as supplied by the manufacturer), respec- tively. Stress-strain values of the modified JC curve were submitted to ABAQUS and the material anisotropy was considered in the material model. The elastic constants of the AL-6XN are Cll, C12 and C44 and the values of these constants were imported from [50] and listed in Table 4. oe ee Vin Ee OSG 5; A eS SES Cot, PMR ee cr ES SE. (Age: Smee Sy fae
De ALUIOUULEG LO UIE PLOUEss POLLO alld Usea Wl Geliv~ ing the JC coefficients, as these constants were previous- y used to simulate other operations than machining to evaluate the equivalent plastic strain only. Therefore, a MATLAB Curve fitting tool was applied to calibrate the JC equation depending on the strain-stress curve of an executed tensile test. The engineering stress and strain parameters obtained by an executed tensile test listed in Table 3 were used for the curve-fitting process to obtain the JC coefficients (A and B) at a strain rate of 0.00027 s'. Table 2 lists the modified JC coefficients (A and B). Figure 7 shows the original and modified JC curves as well as the experimental curve. Tables 4 and 5 list the main properties of the workpiece and the cutting tool (as supplied by the manufacturer), respec- tively. Stress-strain values of the modified JC curve were submitted to ABAQUS and the material anisotropy was considered in the material model. The elastic constants of the AL-6XN are Cll, C12 and C44 and the values of these constants were imported from [50] and listed in Table 4. oe ee Vin Ee OSG 5; A eS SES Cot, PMR ee cr ES SE. (Age: Smee Sy fae
Fig.7 Stress-strain curves for the experimental, original and modified JC parameters As the AL-6XN is a ductile material, a damage for ductile material (ductile damage) criterion and chip separation was used to represent the model failure and chip formation. This ductile damage criterion requires the user to specify values for equivalent fracture strain and the displacement at the fracture. In the current model, for a scalar damage variable D = 0, the equivalent fracture strain is equal to 0.41 which was selected based on the executed tensile test. For a D = | at which the elements fail, the value of the displacement at failure (based on repeated observations for the model outcomes) was equal to 1.7x 10° m.
Fig.7 Stress-strain curves for the experimental, original and modified JC parameters As the AL-6XN is a ductile material, a damage for ductile material (ductile damage) criterion and chip separation was used to represent the model failure and chip formation. This ductile damage criterion requires the user to specify values for equivalent fracture strain and the displacement at the fracture. In the current model, for a scalar damage variable D = 0, the equivalent fracture strain is equal to 0.41 which was selected based on the executed tensile test. For a D = | at which the elements fail, the value of the displacement at failure (based on repeated observations for the model outcomes) was equal to 1.7x 10° m.
Table 5 Cutting tool physical properties as supplied by the manufacturer
Table 5 Cutting tool physical properties as supplied by the manufacturer
3.3.2 Stress and strain analysis The JC model was used to evaluate the stress and the strain during the cutting process. In Fig. 13, all the feature lines were hidden in order to reveal stress and strain distribution along the workpiece. Due to size limitation, only two images each of stress and strain distribution are cited in this paper. Figure 13 shows the first contact point between the tool tip and the workpiece. Maximum stresses were accumulated around the cutting tool tip and then penetrated through the grains and grains’ boundaries, which weaken the non-machined layers (Fig. 13a). After the tool was moved to cut two chips, the stresses were concentrated in the subsurface beneath the cut- ting tool flank face (Fig. 13b). The concentrations of the stresses were varied regarding the depth of cut. It can be seen that the maximum stresses were concentrated in the hardened layer, whereas the grain boundaries of the undeformed grains were affected by stress propagation. Shear localization during the cutting process was found while calculating the strain as shown in Fig. 13c, d. It can be seen that the maximum strain localised around the cutting tool tip and propagated through the primary shear zone (Fig. 13a) whereas after cutting two chips, the shear localization was concentrated at the chips edges (along the primary shear plane of each chip) as viewed in Fig. 13d. Also, Fig. 12c shows the localised shear strain accumulated along the formed chip shear band.
3.3.2 Stress and strain analysis The JC model was used to evaluate the stress and the strain during the cutting process. In Fig. 13, all the feature lines were hidden in order to reveal stress and strain distribution along the workpiece. Due to size limitation, only two images each of stress and strain distribution are cited in this paper. Figure 13 shows the first contact point between the tool tip and the workpiece. Maximum stresses were accumulated around the cutting tool tip and then penetrated through the grains and grains’ boundaries, which weaken the non-machined layers (Fig. 13a). After the tool was moved to cut two chips, the stresses were concentrated in the subsurface beneath the cut- ting tool flank face (Fig. 13b). The concentrations of the stresses were varied regarding the depth of cut. It can be seen that the maximum stresses were concentrated in the hardened layer, whereas the grain boundaries of the undeformed grains were affected by stress propagation. Shear localization during the cutting process was found while calculating the strain as shown in Fig. 13c, d. It can be seen that the maximum strain localised around the cutting tool tip and propagated through the primary shear zone (Fig. 13a) whereas after cutting two chips, the shear localization was concentrated at the chips edges (along the primary shear plane of each chip) as viewed in Fig. 13d. Also, Fig. 12c shows the localised shear strain accumulated along the formed chip shear band.
The stress and strain values at the beginning of the cutting process were extracted and are graphically represented in Fig. 14. It can be seen that for cutting speeds of 94 and 65 m/min, the recorded stress values within distances between
The stress and strain values at the beginning of the cutting process were extracted and are graphically represented in Fig. 14. It can be seen that for cutting speeds of 94 and 65 m/min, the recorded stress values within distances between
Fig. 12 a Microstructure deformation and Von Mises stress (Pa) distribution during the cutting process. b Crack initiation and propagation along the primary shear zone. ¢ The formation of the first chip shows the shear band and primary shear zone
Fig. 12 a Microstructure deformation and Von Mises stress (Pa) distribution during the cutting process. b Crack initiation and propagation along the primary shear zone. ¢ The formation of the first chip shows the shear band and primary shear zone
Fig. 13. Observation of the formed chip during the cutting process at 65-m/min. a, b Von Mises stresses (Pa). ¢, d Equivalent plastic strain (mm/mm
Fig. 13. Observation of the formed chip during the cutting process at 65-m/min. a, b Von Mises stresses (Pa). ¢, d Equivalent plastic strain (mm/mm
strain at high and low cutting speeds was measured and is presented in Fig. 15a. Strain values collected from the exper- imental and FEA study are also presented in Fig. 15b. 3.3.3 Residual stresses obtained by FEA study The relationship between the residual stresses due to the cut- ting process and the depth beneath the cutting tool was eval- uated through the FEA study. A path was drawn below the flank face of the cutting tool down to the bottom of the work- piece to extract the values of the residual stresses at the two cutting speeds. The path location was selected at 0.2 mm in he X-direction as shown in Fig. 16c. This location was pre- ferred to estimate the stresses as the tool was cutting the last chip and the stresses were residing in the machined work- piece. Figure 16a, b shows the residual stresses in term of axial stresses (in the cutting feed direction) S,; for both 94- and 65- m/min cutting speed models. Maximum tensile and
strain at high and low cutting speeds was measured and is presented in Fig. 15a. Strain values collected from the exper- imental and FEA study are also presented in Fig. 15b. 3.3.3 Residual stresses obtained by FEA study The relationship between the residual stresses due to the cut- ting process and the depth beneath the cutting tool was eval- uated through the FEA study. A path was drawn below the flank face of the cutting tool down to the bottom of the work- piece to extract the values of the residual stresses at the two cutting speeds. The path location was selected at 0.2 mm in he X-direction as shown in Fig. 16c. This location was pre- ferred to estimate the stresses as the tool was cutting the last chip and the stresses were residing in the machined work- piece. Figure 16a, b shows the residual stresses in term of axial stresses (in the cutting feed direction) S,; for both 94- and 65- m/min cutting speed models. Maximum tensile and
[58]. In addition, the compressive residual stresses are directly proportional to the PEEQ strain values. For a cutting speed of 94 m/min, the PEEQ strain value equal to 6 and a —120MPa was recorded as a compressive stress while at 65 m/min, the stress value was equal to —90 MPa for a PEEQ strain value of 5.4. 94 m/min, the PEEQ strain value equal to 6 and a —-120MPa was recorded as a compressive stress while at 65 m/min, the
[58]. In addition, the compressive residual stresses are directly proportional to the PEEQ strain values. For a cutting speed of 94 m/min, the PEEQ strain value equal to 6 and a —120MPa was recorded as a compressive stress while at 65 m/min, the stress value was equal to —90 MPa for a PEEQ strain value of 5.4. 94 m/min, the PEEQ strain value equal to 6 and a —-120MPa was recorded as a compressive stress while at 65 m/min, the
Fig. 18 Temperature measurements (°C) during simulation of the cutting process at the 65-m/min cutting speed. a At the beginning of the cutting process. b After forming the chips The validity of the FEA model was assessed according to the strain, temperature measurements and shear plane angle (chip morphology). The percentage difference (error) between the results obtained from the experiments and the FEA model were estimated and listed in Fig. 20. The percentage difference The FEA study in this work included the estimation of the temperature during the cutting process. Results obtained from the FE analysis were compared with the experimental results carried out by the thermal camera. The model shows the tem- perature distribution around the cutting tool tip and in the primary and secondary shear zones as shown in Fig. 18a, b. As the AL-6XN alloy is considered a low-thermal conductiv- ity material, a large amount of the heat at the beginning of the cutting process was accumulated around the tool tip. However, the heat was propagated into the primary and sec- ondary shear zones when the first chip was created because of the friction between the chip and the tool rake face. When the second chip was formed, the temperature was transformed to the chip segments and to the cutting insert. As the temperature of the cutting edge increased, the probability of the BUE 3.5 Model validation
Fig. 18 Temperature measurements (°C) during simulation of the cutting process at the 65-m/min cutting speed. a At the beginning of the cutting process. b After forming the chips The validity of the FEA model was assessed according to the strain, temperature measurements and shear plane angle (chip morphology). The percentage difference (error) between the results obtained from the experiments and the FEA model were estimated and listed in Fig. 20. The percentage difference The FEA study in this work included the estimation of the temperature during the cutting process. Results obtained from the FE analysis were compared with the experimental results carried out by the thermal camera. The model shows the tem- perature distribution around the cutting tool tip and in the primary and secondary shear zones as shown in Fig. 18a, b. As the AL-6XN alloy is considered a low-thermal conductiv- ity material, a large amount of the heat at the beginning of the cutting process was accumulated around the tool tip. However, the heat was propagated into the primary and sec- ondary shear zones when the first chip was created because of the friction between the chip and the tool rake face. When the second chip was formed, the temperature was transformed to the chip segments and to the cutting insert. As the temperature of the cutting edge increased, the probability of the BUE 3.5 Model validation
workpiece is characterised by low thermal conductivity and heat diffusivity, a smaller amount of heat value will be dissi- pated through the workpiece and the tool whereas a large amount of the heat was propagated to the formed chips [59-61]. formation increased. A heated cutting edge assists the forma- tion of thermal softening at the shear zone, which causes wear to be created and spread along the cutting edges. Also, a heat- ed cutting edge sticks and welds the segments of the formed chips to form BUE elements as shown in the SEM images of the shear zone in the “Microstructure analysis” section.
workpiece is characterised by low thermal conductivity and heat diffusivity, a smaller amount of heat value will be dissi- pated through the workpiece and the tool whereas a large amount of the heat was propagated to the formed chips [59-61]. formation increased. A heated cutting edge assists the forma- tion of thermal softening at the shear zone, which causes wear to be created and spread along the cutting edges. Also, a heat- ed cutting edge sticks and welds the segments of the formed chips to form BUE elements as shown in the SEM images of the shear zone in the “Microstructure analysis” section.
Fig. 20 Obtained errors during validating the FE model in the strain measurements between the experiment and the FEA results were equal to 5.25 and 7.5% at the 65- and 94-m/ min cutting speeds, respectively. The reason for the high per- centage difference of 7.5% could be related to the indexing factor during the EBSD scan in the experimental work. In some areas, where high dislocations were accumulated, the boundaries of the deformed grain may not be visible during the indexing process. Therefore, in the postrocessing stage, the HKL Channel5 software considered the deformed grain as a set of small deformed grains. As a result, when the true strain is calculated, the reference area will be larger than the deformed areas and causes the strain values to be increased. The percentage difference in the measured temperature at 65m/min between the experimental and FEA studies was equal to 10% (at the beginning of the cutting process) and 2.5% (after 20 s of the cutting process). On the other hand, at 94 m/min, the percentage difference was 6.7% at the begin- ning of the cutting speed and 3% after 20 s of the cutting process. The differences between the experimental and FEA results could be related to several factors. For example, the setup of the infrared thermal camera on the lathe may affect the readings. Vibration produced by the lathe, as well as the distance between the workpiece-cutting tool contact area and the lens of the camera, can increase noise and reduce the efficiency of the thermal camera. Also, the accuracy of the obtained results from the FEA model was limited to some factors such as the friction coefficient, element size and type. The shear plane angle of the frozen chip samples were also used to validate the FEA models. For a 94-m/min cutting speed, the measured shear plane angles from the experiment and the model were equal to 43° and 41°, respectively, and the
Fig. 20 Obtained errors during validating the FE model in the strain measurements between the experiment and the FEA results were equal to 5.25 and 7.5% at the 65- and 94-m/ min cutting speeds, respectively. The reason for the high per- centage difference of 7.5% could be related to the indexing factor during the EBSD scan in the experimental work. In some areas, where high dislocations were accumulated, the boundaries of the deformed grain may not be visible during the indexing process. Therefore, in the postrocessing stage, the HKL Channel5 software considered the deformed grain as a set of small deformed grains. As a result, when the true strain is calculated, the reference area will be larger than the deformed areas and causes the strain values to be increased. The percentage difference in the measured temperature at 65m/min between the experimental and FEA studies was equal to 10% (at the beginning of the cutting process) and 2.5% (after 20 s of the cutting process). On the other hand, at 94 m/min, the percentage difference was 6.7% at the begin- ning of the cutting speed and 3% after 20 s of the cutting process. The differences between the experimental and FEA results could be related to several factors. For example, the setup of the infrared thermal camera on the lathe may affect the readings. Vibration produced by the lathe, as well as the distance between the workpiece-cutting tool contact area and the lens of the camera, can increase noise and reduce the efficiency of the thermal camera. Also, the accuracy of the obtained results from the FEA model was limited to some factors such as the friction coefficient, element size and type. The shear plane angle of the frozen chip samples were also used to validate the FEA models. For a 94-m/min cutting speed, the measured shear plane angles from the experiment and the model were equal to 43° and 41°, respectively, and the
percentage difference between the two angles was 4.7%. Similarly, for a 65-m/min cutting speed, the angles were equal to 37.5° and 35°, respectively, and their percentage difference was 6.8%.
percentage difference between the two angles was 4.7%. Similarly, for a 65-m/min cutting speed, the angles were equal to 37.5° and 35°, respectively, and their percentage difference was 6.8%.