- Email: [email protected]
The influence of cutting parameters on residual stresses and surface topography during hard turning of 18MnCr5 case carburised steel
Journal of Materials Processing Technology 174 (2006) 82–90
The influence of cutting parameters on residual stresses and surface topography during hard turning of 18MnCr5 case carburised steel Fredrik Gunnberg ∗ , Marcel Escursell, Michael Jacobson Product and Production Development, Chalmers University of Technology, Gothenburg, Sweden Received 19 March 2004; received in revised form 19 March 2004; accepted 14 February 2005
Abstract The purpose of this paper was to obtain a comprehensive understanding of the relation between cutting data and surface integrity, in terms of residual stresses and surface roughness, during hard turning with PCBN inserts in 18MnCr5 case carburised steel. In the study, fractionally reduced CCF test plan was used. It was found that the effect of cutting parameters on residual stress could only be predicted at 0–50 m below the surface. The cutting geometry and data can be changed to alter the residual stress in area referred to above. The surface roughness values were mainly influenced by the feed rate and nose radius. © 2006 Published by Elsevier B.V. Keywords: Hard turning; Residual stresses; Surface roughness
1. Introduction The possibility of machining hardened steel components by turning has generated interest within the industry. The process of hard turning has been studied extensively over the last decade [1,3–11]. Steel components often have to be machined after heat treatment in order to obtain the correct shape as well as the required surface finish. Hardened steel components are high performance parts, which are often loaded near to their physical limits. Therefore it is vital to understand how the finishing process affects the functional behaviour of the machined parts. The finishing processes commonly used today are grinding and honing, but several benefits have been reported by substituting hard turning for grinding, e.g. higher flexibility and shortened lead-times. T¨onshoff et al. [1] have described a process evaluation between hard turning and grinding; economic aspects, flexibility, ecological and quality criteria were specified and compared (Fig. 1). Obviously, hard turning is an interesting alternative to grinding operations under specific circumstances. The quality criteria consist of workpiece quality, process reliability and surface integrity. All three ∗
Corresponding author. E-mail address: [email protected] (F. Gunnberg).
0924-0136/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.jmatprotec.2005.02.262
aspects of quality have to be considered in order to successfully implement hard turning. T¨onshoff et al. [1] and several other authors [3–11] have reported how hard turning influences the surface integrity of the machined part. Field and Kahles [2] describe surface integrity as the relationship between surface geometric values and the physical properties such as residual stress, hardness and structure of the surface layers. Surface integrity has a large impact on the performance of the component. Liu and Mittal [3] report experimental evidence that the fatigue life of the hard turned surface is superior to that of a ground surface because of the more appropriate compressive residual stress profile. Liu and Mittal [3] also state that the residual stress profile can be controlled during turning, which would enable production of compressive residual stressed parts. This would enable the manufacturing of customised components e.g. bearings, with a prolonged fatigue life. The amount of residual stress beneficial for improving fatigue life differs with each application. Matsumoto et al. [4] reported that the hard turned product has as good a fatigue life as the ground product. Furthermore, it has been determined that the tool edge is the most important factor influencing the residual stress profile. Jacobson [5] investigated the effects of tool parameters and depth of cut on the residual stress in M50 steels. It was found that effective rake angle and tool
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
83
Fig. 1. Comparison of hard cutting and grinding (source: Cutting of hardened steel, CIRP 2000 [1]).
radius both affected the amount of residual stress generated. Jacobson et al. [6] also found a cutting speed that generated maximum compressive stress in bainite steel. The Thiele and Melkote [7] study showed that tool edge geometry is highly influential for surface residual stresses. In general, large edge hone tools generated compressive surface residual stresses in the axial and circumferential directions in longitudinal turning operations. Brinksmeier et al. [8] reported that residual stresses act in a component independently of external force or moment. The internal forces form a system of equilibrium. If some part of the component is removed, the state of equilibrium is disturbed. During hard machining, the amount of material removed is minimal and the residual stress only has a limited depth of penetration of some hundredths of millimetres. The hard machining process generates residual stresses in the workpiece by plastic deformation or metallurgical transformations [8]. Vomacka and Walburger [9] described the genesis of residual stress as normal force applied on the work piece by the machining tool, causing plastic deformation and as a consequence raising the compressive stress on the surface layer. Friction between the tool and the workpiece results in heat development leading to residual tensile stress on the surface layer. K¨onig et al. [10] provided a similar explanation, in that the mechanical stress induced by the cutting tool causes a transformation of the residual austenite and strain hardening in the surface layer of the workpiece, which induces compressive stress. Thermal stress, i.e. residual stress, results from the temperature due to friction on the flank face. How turning influences the generation of the residual stress level can be explained by considering the state of stress produced when the cutting tool slides across the workpiece (Figs. 2 and 3). The mechanism of mechanically generated residual stress during cutting (A) can be explained by a plastic deformation in the surface layer (1) and elastic deformation in the underlying surface layer (2) (Figs. 2 and 3 (left)). To achieve force equilibrium and geometric compatibility after the cutting processes, the elastic dilatation places the surface layer in residual compressive stress (B). The thermal resid-
Fig. 2. Generation of residual stress by turning.
ual stress mechanism is due to the heat of the cutting process, which expands the surface layer and produces compressive stress (A). The workpiece is then cooled (B) and contractions in the surface layer (1) produce tension residual stress, Fig. 3 (right). The thermal effect decreases further inside the workpiece, thus the main consequence of tension stress is on the surface. Consequently, the temperature of the cutting edge is very important for reducing tensile residual stress. Although the way in which hard turning influences surface integrity and especially residual stress has been investigated closely, the relationship between cutting conditions and product quality must be completely understood. Therefore, to improve the quality of the machined component, a
Fig. 3. The residual stress mechanism.
84
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
comprehensive understanding of the influence of the cutting parameter on the surface properties is crucial. Even though substituting hard turning for grinding would lower production costs, increase flexibility and be less harmful to the environment, it must be remembered that the quality of the machined components has to be better or at least the same. Hence a model of the effects of different cutting parameters on the generation of residual stress is required. The purpose of this paper is to investigate how different cutting parameters influence residual stress (level and magnitude) and surface topography when turning in 18MnCr5 case hardened steel. Fig. 5. The geometry of different types of CBN tools used in the experiment.
2. Experimental set-up Machining and measurements were performed at Chalmers University of Technology, Gothenburg, Sweden. 2.1. Test material and equipment The test material was DIN 18MnCr5 low carbon steel, the chemical composition of which is given in Table 1. The design of the specimens simulated small workpieces, which means short engagement times and hence more entry and exit for the insert (Fig. 4). The length of cut was fixed at 200 m and all tests were preformed dry. The case hardening depth, which is defined as the distance from the surface where a hardness of 550 HV exists, was 1.2–1.3 mm. In the study a solid CBN 100 insert from Seco Tools was used. CBN 100 has a low CBN content of 50% and the average grain size is 2 m. The different type of CBN insert geometries used in the tests can be seen in Fig. 5. Table 1 Chemical composition (%) DIN 18MnCr5 C
Si
Mn
P≤%
S≤%
Cr
0.17–0.22
≤0.40
1.10–1.40
0.035
0.035
1.00–1.30
Fig. 4. Setup used in the turning tests.
¨ An ordinary stable two-axes CNC lathe (TORSHALLA S250) was used for longitudinal turning of the specimens (Fig. 4). Surface roughness was measured by the Rubber–Replica method. A mixture of rubber and reactive was applied to the surface. Once the mixture polymerised the test samples were measured by WYKO© optical-reflecting equipment. Thereafter positive relief and roughness parameters were calculated using WYKO Vision, Version 2.210, a computer aided measuring system. Test design and regression analysis were computer aided with MODDE 6© software supplied by Umetrics AB. Residual stress is measured with the X-ray diffraction method. The principle of X-ray measurement is based on Bragg’s law, the sine-square-psi (sin2 ψ) method. A chromium tube calibrated on ferrite was used to measure the {2 1 1} peak of steel. The residual stresses were measured in tangential (cutting speed) and axial direction (feed rate). The collimator used was 2 mm, the calculation method was cross correlation and the exposure time was 7 s. In order to find the residual stress in the radial direction, the material was chemically etched away (10, 20, 30, 50, 70, 100 and 130 m depth), with 3 M NaCl-solution. The X-ray equipment was supplied by Stresstech, Finland. 2.2. Test design The purpose of the hard turning tests was to gain a comprehensive understanding of the influence of the cutting parameters on residual stress generation and surface topography. Design of experiments (DoE) was used in the study. The basic idea is to vary all relevant factors simultaneously over a set of planned experiments and then connect the results by means of a mathematical model. This model is then used for interpretation, predictions and optimisation. Five factors were considered in the investigation: tool rake angle (γ), tool nose radius (Re ), cutting speed (Vc ), cutting depth (ap ) and feed rate (f). The minimum and maximum levels were restricted by practical constraints and by previous experience (Table 2). Residual stress is a multilevel response i.e. it was measured at different depths below the surface and in two directions:
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90 Table 2 Factors
85
Table 4 Design matrix
Factor
Short
Units
Low
Medium
High
Cutting speed Feed rate Cutting depth Nose radius Rake angle
Vc f ap Re γ
m/min mm/rev. mm mm
110 0.05 0.05 0.8 6
170 0.10 0.10 1.6 15
230 0.15 0.15 4.5 21
◦
Table 3 Responses Response
Units
Level
Residual stress
MPa
Roughness
m
0, 10, 15, 20, 30, 50, 70 and 100 m below the surface (feed and cutting speed direction) Sa, St, Sz
Cutting speed direction (tangential) and feed direction (axial) (see Table 3 and Fig. 6). Three spatial parameters characterize the surface roughness of the turned samples: arithmetical average of the surface (Sa), height deviation between the lowest and highest points of the surface (St) and 10-point height of the surface (Sz). In order to maintain the model quality without doing a large number of tests, a fractional cubic centred on face (CCF) test design was used. CCF is especially efficient up to five factors, as in the present study, since it is possible to estimate main effects and quadratic effects as well as interactions between factors. In CCF, design experiments are located at each corner of the design cube and in the middle of each face. Furthermore, the design was reduced i.e. not all the experiments were performed, but only a fraction of them. The consequence of this is that high order interactions (three or more factors) cannot be calculated, although these are normally of no interest. The total number of runs was drastically reduced. The final test design had 29 runs including three replicas. Table 4 shows the design matrix where −1 denotes low level, 0 medium and 1 high level.
Fig. 6. The test specimens were etched and the residual stresses were measured in the Vc - and f-directions.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Vc
f
ap
Re
γ
−1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 0 0 0 0 0 0 0 0 0 0 0
−1 −1 1 1 −1 −1 1 1 −1 −1 1 1 −1 −1 1 1 0 0 −1 1 0 0 0 0 0 0 0 0 0
−1 −1 −1 −1 1 1 1 1 −1 −1 −1 −1 1 1 1 1 0 0 0 0 −1 1 0 0 0 0 0 0 0
−1 −1 −1 −1 −1 −1 −1 −1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 −1 1 0 0 0 0 0
1 −1 −1 1 −1 1 1 −1 −1 1 1 −1 1 −1 −1 1 0 0 0 0 0 0 0 0 −1 1 0 0 0
3. Results The results are presented in two separate sections. Firstly, the evaluation of the residual stress is reported (tangential and axial directions). Secondly, the surface roughness evaluation is presented. 3.1. Residual stress The highest compressive residual stresses were obtained in test 7, while the lowest compressive stresses were induced in test 14 (Figs. 7 and 8).
Fig. 7. Maximum and minimum residual stresses in feed direction (axial).
86
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 8. Maximum and minimum residual stresses in cutting speed direction (tangential). Fig. 11. Effects plotted on the surface (S0) in feed direction (axial).
Fig. 9. Summary of fit for residual stresses in feed direction (axial).
The model for residual stresses was fitted and improved i.e. redundant terms were removed. Model quality was evaluated by R2 and Q2 indicators as shown in Figs. 9 and 10. R2 is the fraction of variation of the response explained by the model. Q2 is the fraction of variation of the response that can be predicted by the model. R2 represents how well the model fits the raw data; while Q2 refers to how good the model is for prediction and optimisation. R2 is an overestimated and Q2 an underestimated measure of the fit of the model Both R2 and Q2 values are figures, usually between 0 and 1 (Q2 can be negative for very poor models). Values close to 1 for both R2 and Q2 indicate very exact model a with excellent predictive power. Lack of fit and reproducibility (replicates) were also taken into account in order to evaluate the model. A study of the model quality in Figs. 9 and 10 reveals that the highest values occur between 0 and 50 m below
Fig. 10. Summary of fit in cutting speed direction (tangential).
Fig. 12. Effects plotted on the surface (S0) in cutting speed direction (tangential).
the surface (high R2 and Q2 ). However, the model is poor at depths of over 70 m below the surface. In this region residual stresses are mainly influenced by pre-machining factors such as heat treatment, rather than by the cutting parameters. In addition, residual stresses within this region are more or less constant (Figs. 7 and 8). For these reasons, the analysis was restricted to depths of 0–50 m below the surface. Effects on residual stresses are shown in Figs. 11–14. Stress generation on the surface (Figs. 11 and 12), follows a different behaviour pattern than the stress generation below
Fig. 13. Effects plotted for 10–50 um below the surface (S10–S50) in feed direction (axial).
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 14. Effects plotted for 10–50 um below the surface (S10–S50) in cutting speed direction (tangential).
the surface (Figs. 13 and 14), i.e. the effects are different. This result emphasizes the fact that the mechanism by which residual stresses are generated is different on the surface than within the material. Therefore, residual stress generation on the surface is analysed separately (Figs. 11 and 12). The conclusion for the effects plotted on the surface in an axial direction (Fig. 11) is as follows: higher cutting speed (Vc ) and larger nose radius (Re ) produce more tensile stress. Feed rate (f) has a positive effect for low values but a negative effect for high values (f × f). Cutting depth (ap ) and rake angle (γ) are not significant i.e. their effect on residual stress is negligible. There is a 2nd order interaction between cutting speed and feed rate (Vc × f) although relatively insignificant. Nose radius (Re ) and cutting speed (Vc ) have a positive effect in a tangential direction (Fig. 12). Rake angle (γ) has in this case a negative effect. An interaction occurs between nose radius and rake angle (Re × ␥). Residual stress generation below the surface (Figs. 13 and 14) is dominated by feed (f), nose radius (Re ), rake angle (γ) and interactions between them. Nose radius (Re ) has a positive effect in both directions i.e. the higher the nose radius, the more tensile or less compressive are the residual stresses. Feed rate (f) and rake angle (γ) have negative effects even though their significance is lower than for the nose radius (Re ). Feed rate square (f × f) appears in an axial direction. Cutting speed (Vc ) and depth of cut (ap ) do not influence the residual stress generation below the surface. All the effects shown in Figs. 11–14 are scaled and centred for a better understanding of the results. The most significant interaction is that between nose radius and rake angle (Re × γ) and it therefore deserves special attention. Two factors interact when the effect of one of them depends on the level of the other. Hence in this study, the effect of nose radius depends on how large the rake angle is. For low rake angles, γ low in Fig. 15, nose radius has a positive effect on the residual stresses generated 30 m below the surface. In this case, the larger the nose radius, the higher the stresses. On the other hand, if the rake angle is high, γ high in Fig. 15, the effect of nose radius becomes negative. This shows a clear interaction between nose radius and rake angle.
87
Fig. 15. Interaction between rake angle and nose radius at 30 m below the surface in feed direction (axial).
Fig. 16. Surface roughness in test 8.
3.2. Surface roughness The roughest surface was obtained in test 8, and the smoothest in test 10 (Figs. 16 and 17). Fig. 17 illustrates that a smooth surface identical to that obtained in grinding can be achieved by hard turning if the cutting parameters are properly selected. Fig. 18 shows R2 and Q2 values for Sa, St and Sz. Not surprisingly, the highest model quality corresponds to Sa (arithmetical average of the surface), and the poorest to St (height deviation between the lowest and the highest points of the surface). This difference is due to the robustness in
Fig. 17. Surface roughness in test 10.
88
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 18. Summary of fit for surface roughness.
Fig. 21. Schematic illustration of the residual stress level after hard turning in 18MnCr5.
Fig. 19. Effects plotted for surface roughness.
ducted within a cutting data range that is normally used in standard production. During the testing the tool wear was kept below Vb 0.05 mm. With this knowledge it is possible to predict and control the residual stresses level down to a depth of approximately 50 m below the surface when hard turning case carburised steel (Fig. 21). Hard turning has a negligible effect on the residual stresses below 70 m. Below this depth, the stress level is due to heat treatment and/or other process parameters prior to cutting. The generation of residual stress is a complex interaction between thermal and mechanical factors. The two opposing phenomena interact with each other to create the final residual stress pattern in the component. Table 5 summarizes thermal and mechanical generation as well as residual stress generation when changing the cutting parameters. Table 5 should be seen as a general explanation of how residual stress is generated. In the following discussion the five cutting parameters will be described. Higher cutting speed increases the temperature (more heat is generated), which causes tensile residual stress on the surface. Since the chip removes most of the heat, high temperature does not penetrate deeply into the workpiece and therefore does not affect the residual stress generation below
Table 5 Residual stress generation summary Fig. 20. Nose radius effect on surface roughness.
Parameter S0 Thermal generation Mechanical generation RS() MODDE
the calculation of Sa compared to St. In other words; Sa is more predictable than St, the latter being more random. As expected, feed rate (f) has a negative influence while nose radius (Re ) has a positive effect on surface roughness, see Fig. 19. It can also be seen that feed and nose radius interact (f × Re ) although not very significantly. More important is the nose radius square effect (Re × Re ) shown in Fig. 20.
4. Discussion In this study it has been shown how different factors affect the residual stress level at various depths. The tests were con-
Vc f ap γ Re
↑ ↑ – ↑ ↑
↓ ↑ – ↑ ↑
↑+T ↓−C – ↓−C ↑+T
Vc f ap γ Re
S10–S50 Thermal generation – ↑ – ↑ ↑
Mechanical generation ↓ ↑ – ↑ ↑
RS() MODDE – ↓−C – ↓−C ↑+T
S70–S130 Hardening process
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
the surface. If a subsequent honing process follows hard turning, the tensile residual stresses on the surface can be removed. Secondly, the level of compressive stress below the surface increases constantly with higher feed rates. This can be a consequence of the fact that a higher feed rate generates increased cutting forces and therefore more plastic deformation. Thirdly, when changing the cutting depths, the axial force is only slightly increased. However, the tangential force corresponds directly to the cutting depth and consequently does not contribute to the subsurface deformation. Therefore, the residual stresses should not be affected, which is consistent with the present results. Jacobson [5] also investigated depth of cut, and the result was that depth of cut does not affect the amount of residual stress generated in the hard turning of M50 steel. This is especially useful for the precision manufacturing of products with an initial out of roundness, as it is possible to make several calibration cuts to improve the roundness without affecting the residual stresses. The product suffers no dimensional distortion due to unevenness. Fourthly, a more negative rake angle produces increased compressive stress both on the surface and below. In the test plan the rake angle changed from −6 to −21. With a more negative rake angle, the mechanical generation of the subsurface increase, which produces greater compressive stress is due to the increase in passive cutting force. The higher temperature engendered by a more negative rake angle does not affect the stress level. Thiele and Melkote [7] presented a similar relationship. Jacobson [5] stated that the effective rake angle influences the amount of residual stress induced in the workpiece. Finally, a larger nose radius produces less compressive stress on the surface and subsurface. This is due to the fact that the contact area is larger and thus the specific cutting force (force per unit of area) is lower. Jacobson [5] found the same correlation in M50 steel: a smaller nose radius generates more compressive stress. The results of this study may be used to understand the mechanism by which residual stresses are generated in hard turning. The study can also be used to predict the stress level for future applications. Our ambition is to design a mathematical tool to determine the correct cutting parameters in order to produce a specific residual stress profile. As an example, Fig. 22 shows the predicted residual stresses 30 m below the surface in a tangential direction as a function of nose radius and rake angle. From a study of the surface roughness values Sa, Sz and St, it can be concluded that they are at the same level as for a ground workpiece. As expected, feed rate and nose radius are among the most influential parameters. Fig. 23 shows the predicted Sa surface value as a function of nose radius and rake angle. However, the feed rate square does not appear to be significant as in the purely geometrical expression of Rt (Eq. (1)). The Rt equation obtained by regression is shown in Eq. (2). Rt =
f2 8Re
(1)
89
Fig. 22. Residual stress prediction at 30 m below the surface in cutting speed direction (tangential).
Rt = 5.71 + 31.483f + Re (−4.162 + 0.740Re − 5.125f )
(2)
It is important to note that even though hard turning can produce equally good surface values as grinding, it is difficult to compare their functionality. A hard turned and a ground surface with the same Ra-value do not have the same tribology behaviour. Moreover, structural surface change in the workpiece introduced by the machining process is an important consequence of the finishing process. This surface or subsurface modification occurs because of localized and rapid thermal mechanical work, resulting in metallurgical transformation and/or at times chemical interactions. T¨onshoff et al. [11] claimed that change of the workpiece surface could be essential for the fatigue life of components. When hard turning is used as the finishing operation, it is very important to
Fig. 23. Surface roughness, Sa prediction as function of nose radius and feed rate.
90
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
minimize the microstructure change. By using honing as a subsequent operation, the depth of cut is approximately 10 m and the heat effect layer is removed.
Acknowledgements The authors wish to thank Vinnova for their financial support and the AIS 13 project members for their contribution.
5. Conclusions Tests were performed to gain a comprehensive understanding of how tool geometry as well as cutting parameters affect the residual stresses and surface topography in 18MnCr5 steel. The following conclusions can be drawn from the results: • Three different sub-models were created: Surface (S0), (S10–S50) 10–50 m below the surface and (S70–S130) 70–130 m below the surface, due to the fact that different cutting parameters have a different effect on the level of residual stress generated on the surface and below the surface. • Cutting speed increases the tensile residual stress on the surface (S0). The heat generated from higher cutting speeds does not penetrate more deeply into the workpiece, and therefore does not affect the sub-model S10–S50. • Increased feed generates higher compressive stresses. • The cutting depth does not affect residual stresses. • A more negative rake angle produces more compressive stress in both S0 and S10–S50 models. • By controlling the cutting parameters, it is possible to generate tailor-made stresses in the product, which can prolong the service life of the machined component. • Feed and nose radius have the greatest effect on the geometric surface values. Sa values comparable to grinding can be achieved.
References [1] H.K. T¨onshoff, C. Arendt, R. Ben Amor, Cutting of hardened Steel, Ann. CIRP 49 (2) (2000) 547–566. [2] M. Field, J. Kahles, Review of surface integrity of machined components, Ann. CIRP 20 (1971) 153–163. [3] C.R. Liu, S. Mittal, Optimal pre-stressing the surface of a component by superfinish hard turning for maximum fatigue life in rolling contact, Wear 219 (1998) 128–140. [4] Y. Matsumoto, F. Hashimoto, G. Lahoti, Surface integrity generated by precision hard turning, Ann. CIRP 48 (1999) 59–62. [5] M. Jacobson, Surface integrity of hard-turned M50 steel, in: Proceedings of the Institution of Mechanical Engineers, Part B, J. Eng. Manuf. 216 (1) (2002) 47–54. [6] M. Jacobson, D. Patrik, G. Fredrik, Cutting speed influence on surface integrity of hard turned bainite steel, J. Mater. Proc. Technol. 128 (1–3, 6) (2002) 318–323. [7] J.D. Thiele, S.N. Melkote, Effect of cutting edge geometry and workpiece hardness on surface generation in the finish hard turning of AISI 52100 steel, J. Mater. Proc. Technol. 94 (1999) 216–226. [8] E. Brinksmeier, J.J. Cammett, P. Leskovar, J. Peters, H.K. Tonshoff, Residual stresses-measurements and causes in machining processes, Ann. CIRP 31 (1982) 491–510. [9] P. Vomacka, H. Walburger, Residual stresses due to hard-machiningindustrial experiences, in: Proceedings of the 5th European Conference on Residual stresses, Switzerland, 2000, pp. 592–597. [10] W. K¨onig, A. Berktold, K-F. Koch, Turning versus grinding: a comparison of surface integrity aspects and attainable accuracies, Ann. CIRP 42 (1993) 39–43. [11] H.K. Tonshoff, H.-G. Wobker, D. Brandt, Potential and limitation of hard turning, SME Tech. Papers (1995) MR95–MR215.
The influence of cutting parameters on residual stresses and surface topography during hard turning of 18MnCr5 case carburised steel Fredrik Gunnberg ∗ , Marcel Escursell, Michael Jacobson Product and Production Development, Chalmers University of Technology, Gothenburg, Sweden Received 19 March 2004; received in revised form 19 March 2004; accepted 14 February 2005
Abstract The purpose of this paper was to obtain a comprehensive understanding of the relation between cutting data and surface integrity, in terms of residual stresses and surface roughness, during hard turning with PCBN inserts in 18MnCr5 case carburised steel. In the study, fractionally reduced CCF test plan was used. It was found that the effect of cutting parameters on residual stress could only be predicted at 0–50 m below the surface. The cutting geometry and data can be changed to alter the residual stress in area referred to above. The surface roughness values were mainly influenced by the feed rate and nose radius. © 2006 Published by Elsevier B.V. Keywords: Hard turning; Residual stresses; Surface roughness
1. Introduction The possibility of machining hardened steel components by turning has generated interest within the industry. The process of hard turning has been studied extensively over the last decade [1,3–11]. Steel components often have to be machined after heat treatment in order to obtain the correct shape as well as the required surface finish. Hardened steel components are high performance parts, which are often loaded near to their physical limits. Therefore it is vital to understand how the finishing process affects the functional behaviour of the machined parts. The finishing processes commonly used today are grinding and honing, but several benefits have been reported by substituting hard turning for grinding, e.g. higher flexibility and shortened lead-times. T¨onshoff et al. [1] have described a process evaluation between hard turning and grinding; economic aspects, flexibility, ecological and quality criteria were specified and compared (Fig. 1). Obviously, hard turning is an interesting alternative to grinding operations under specific circumstances. The quality criteria consist of workpiece quality, process reliability and surface integrity. All three ∗
Corresponding author. E-mail address: [email protected] (F. Gunnberg).
0924-0136/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.jmatprotec.2005.02.262
aspects of quality have to be considered in order to successfully implement hard turning. T¨onshoff et al. [1] and several other authors [3–11] have reported how hard turning influences the surface integrity of the machined part. Field and Kahles [2] describe surface integrity as the relationship between surface geometric values and the physical properties such as residual stress, hardness and structure of the surface layers. Surface integrity has a large impact on the performance of the component. Liu and Mittal [3] report experimental evidence that the fatigue life of the hard turned surface is superior to that of a ground surface because of the more appropriate compressive residual stress profile. Liu and Mittal [3] also state that the residual stress profile can be controlled during turning, which would enable production of compressive residual stressed parts. This would enable the manufacturing of customised components e.g. bearings, with a prolonged fatigue life. The amount of residual stress beneficial for improving fatigue life differs with each application. Matsumoto et al. [4] reported that the hard turned product has as good a fatigue life as the ground product. Furthermore, it has been determined that the tool edge is the most important factor influencing the residual stress profile. Jacobson [5] investigated the effects of tool parameters and depth of cut on the residual stress in M50 steels. It was found that effective rake angle and tool
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
83
Fig. 1. Comparison of hard cutting and grinding (source: Cutting of hardened steel, CIRP 2000 [1]).
radius both affected the amount of residual stress generated. Jacobson et al. [6] also found a cutting speed that generated maximum compressive stress in bainite steel. The Thiele and Melkote [7] study showed that tool edge geometry is highly influential for surface residual stresses. In general, large edge hone tools generated compressive surface residual stresses in the axial and circumferential directions in longitudinal turning operations. Brinksmeier et al. [8] reported that residual stresses act in a component independently of external force or moment. The internal forces form a system of equilibrium. If some part of the component is removed, the state of equilibrium is disturbed. During hard machining, the amount of material removed is minimal and the residual stress only has a limited depth of penetration of some hundredths of millimetres. The hard machining process generates residual stresses in the workpiece by plastic deformation or metallurgical transformations [8]. Vomacka and Walburger [9] described the genesis of residual stress as normal force applied on the work piece by the machining tool, causing plastic deformation and as a consequence raising the compressive stress on the surface layer. Friction between the tool and the workpiece results in heat development leading to residual tensile stress on the surface layer. K¨onig et al. [10] provided a similar explanation, in that the mechanical stress induced by the cutting tool causes a transformation of the residual austenite and strain hardening in the surface layer of the workpiece, which induces compressive stress. Thermal stress, i.e. residual stress, results from the temperature due to friction on the flank face. How turning influences the generation of the residual stress level can be explained by considering the state of stress produced when the cutting tool slides across the workpiece (Figs. 2 and 3). The mechanism of mechanically generated residual stress during cutting (A) can be explained by a plastic deformation in the surface layer (1) and elastic deformation in the underlying surface layer (2) (Figs. 2 and 3 (left)). To achieve force equilibrium and geometric compatibility after the cutting processes, the elastic dilatation places the surface layer in residual compressive stress (B). The thermal resid-
Fig. 2. Generation of residual stress by turning.
ual stress mechanism is due to the heat of the cutting process, which expands the surface layer and produces compressive stress (A). The workpiece is then cooled (B) and contractions in the surface layer (1) produce tension residual stress, Fig. 3 (right). The thermal effect decreases further inside the workpiece, thus the main consequence of tension stress is on the surface. Consequently, the temperature of the cutting edge is very important for reducing tensile residual stress. Although the way in which hard turning influences surface integrity and especially residual stress has been investigated closely, the relationship between cutting conditions and product quality must be completely understood. Therefore, to improve the quality of the machined component, a
Fig. 3. The residual stress mechanism.
84
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
comprehensive understanding of the influence of the cutting parameter on the surface properties is crucial. Even though substituting hard turning for grinding would lower production costs, increase flexibility and be less harmful to the environment, it must be remembered that the quality of the machined components has to be better or at least the same. Hence a model of the effects of different cutting parameters on the generation of residual stress is required. The purpose of this paper is to investigate how different cutting parameters influence residual stress (level and magnitude) and surface topography when turning in 18MnCr5 case hardened steel. Fig. 5. The geometry of different types of CBN tools used in the experiment.
2. Experimental set-up Machining and measurements were performed at Chalmers University of Technology, Gothenburg, Sweden. 2.1. Test material and equipment The test material was DIN 18MnCr5 low carbon steel, the chemical composition of which is given in Table 1. The design of the specimens simulated small workpieces, which means short engagement times and hence more entry and exit for the insert (Fig. 4). The length of cut was fixed at 200 m and all tests were preformed dry. The case hardening depth, which is defined as the distance from the surface where a hardness of 550 HV exists, was 1.2–1.3 mm. In the study a solid CBN 100 insert from Seco Tools was used. CBN 100 has a low CBN content of 50% and the average grain size is 2 m. The different type of CBN insert geometries used in the tests can be seen in Fig. 5. Table 1 Chemical composition (%) DIN 18MnCr5 C
Si
Mn
P≤%
S≤%
Cr
0.17–0.22
≤0.40
1.10–1.40
0.035
0.035
1.00–1.30
Fig. 4. Setup used in the turning tests.
¨ An ordinary stable two-axes CNC lathe (TORSHALLA S250) was used for longitudinal turning of the specimens (Fig. 4). Surface roughness was measured by the Rubber–Replica method. A mixture of rubber and reactive was applied to the surface. Once the mixture polymerised the test samples were measured by WYKO© optical-reflecting equipment. Thereafter positive relief and roughness parameters were calculated using WYKO Vision, Version 2.210, a computer aided measuring system. Test design and regression analysis were computer aided with MODDE 6© software supplied by Umetrics AB. Residual stress is measured with the X-ray diffraction method. The principle of X-ray measurement is based on Bragg’s law, the sine-square-psi (sin2 ψ) method. A chromium tube calibrated on ferrite was used to measure the {2 1 1} peak of steel. The residual stresses were measured in tangential (cutting speed) and axial direction (feed rate). The collimator used was 2 mm, the calculation method was cross correlation and the exposure time was 7 s. In order to find the residual stress in the radial direction, the material was chemically etched away (10, 20, 30, 50, 70, 100 and 130 m depth), with 3 M NaCl-solution. The X-ray equipment was supplied by Stresstech, Finland. 2.2. Test design The purpose of the hard turning tests was to gain a comprehensive understanding of the influence of the cutting parameters on residual stress generation and surface topography. Design of experiments (DoE) was used in the study. The basic idea is to vary all relevant factors simultaneously over a set of planned experiments and then connect the results by means of a mathematical model. This model is then used for interpretation, predictions and optimisation. Five factors were considered in the investigation: tool rake angle (γ), tool nose radius (Re ), cutting speed (Vc ), cutting depth (ap ) and feed rate (f). The minimum and maximum levels were restricted by practical constraints and by previous experience (Table 2). Residual stress is a multilevel response i.e. it was measured at different depths below the surface and in two directions:
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90 Table 2 Factors
85
Table 4 Design matrix
Factor
Short
Units
Low
Medium
High
Cutting speed Feed rate Cutting depth Nose radius Rake angle
Vc f ap Re γ
m/min mm/rev. mm mm
110 0.05 0.05 0.8 6
170 0.10 0.10 1.6 15
230 0.15 0.15 4.5 21
◦
Table 3 Responses Response
Units
Level
Residual stress
MPa
Roughness
m
0, 10, 15, 20, 30, 50, 70 and 100 m below the surface (feed and cutting speed direction) Sa, St, Sz
Cutting speed direction (tangential) and feed direction (axial) (see Table 3 and Fig. 6). Three spatial parameters characterize the surface roughness of the turned samples: arithmetical average of the surface (Sa), height deviation between the lowest and highest points of the surface (St) and 10-point height of the surface (Sz). In order to maintain the model quality without doing a large number of tests, a fractional cubic centred on face (CCF) test design was used. CCF is especially efficient up to five factors, as in the present study, since it is possible to estimate main effects and quadratic effects as well as interactions between factors. In CCF, design experiments are located at each corner of the design cube and in the middle of each face. Furthermore, the design was reduced i.e. not all the experiments were performed, but only a fraction of them. The consequence of this is that high order interactions (three or more factors) cannot be calculated, although these are normally of no interest. The total number of runs was drastically reduced. The final test design had 29 runs including three replicas. Table 4 shows the design matrix where −1 denotes low level, 0 medium and 1 high level.
Fig. 6. The test specimens were etched and the residual stresses were measured in the Vc - and f-directions.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Vc
f
ap
Re
γ
−1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 −1 1 0 0 0 0 0 0 0 0 0 0 0
−1 −1 1 1 −1 −1 1 1 −1 −1 1 1 −1 −1 1 1 0 0 −1 1 0 0 0 0 0 0 0 0 0
−1 −1 −1 −1 1 1 1 1 −1 −1 −1 −1 1 1 1 1 0 0 0 0 −1 1 0 0 0 0 0 0 0
−1 −1 −1 −1 −1 −1 −1 −1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 −1 1 0 0 0 0 0
1 −1 −1 1 −1 1 1 −1 −1 1 1 −1 1 −1 −1 1 0 0 0 0 0 0 0 0 −1 1 0 0 0
3. Results The results are presented in two separate sections. Firstly, the evaluation of the residual stress is reported (tangential and axial directions). Secondly, the surface roughness evaluation is presented. 3.1. Residual stress The highest compressive residual stresses were obtained in test 7, while the lowest compressive stresses were induced in test 14 (Figs. 7 and 8).
Fig. 7. Maximum and minimum residual stresses in feed direction (axial).
86
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 8. Maximum and minimum residual stresses in cutting speed direction (tangential). Fig. 11. Effects plotted on the surface (S0) in feed direction (axial).
Fig. 9. Summary of fit for residual stresses in feed direction (axial).
The model for residual stresses was fitted and improved i.e. redundant terms were removed. Model quality was evaluated by R2 and Q2 indicators as shown in Figs. 9 and 10. R2 is the fraction of variation of the response explained by the model. Q2 is the fraction of variation of the response that can be predicted by the model. R2 represents how well the model fits the raw data; while Q2 refers to how good the model is for prediction and optimisation. R2 is an overestimated and Q2 an underestimated measure of the fit of the model Both R2 and Q2 values are figures, usually between 0 and 1 (Q2 can be negative for very poor models). Values close to 1 for both R2 and Q2 indicate very exact model a with excellent predictive power. Lack of fit and reproducibility (replicates) were also taken into account in order to evaluate the model. A study of the model quality in Figs. 9 and 10 reveals that the highest values occur between 0 and 50 m below
Fig. 10. Summary of fit in cutting speed direction (tangential).
Fig. 12. Effects plotted on the surface (S0) in cutting speed direction (tangential).
the surface (high R2 and Q2 ). However, the model is poor at depths of over 70 m below the surface. In this region residual stresses are mainly influenced by pre-machining factors such as heat treatment, rather than by the cutting parameters. In addition, residual stresses within this region are more or less constant (Figs. 7 and 8). For these reasons, the analysis was restricted to depths of 0–50 m below the surface. Effects on residual stresses are shown in Figs. 11–14. Stress generation on the surface (Figs. 11 and 12), follows a different behaviour pattern than the stress generation below
Fig. 13. Effects plotted for 10–50 um below the surface (S10–S50) in feed direction (axial).
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 14. Effects plotted for 10–50 um below the surface (S10–S50) in cutting speed direction (tangential).
the surface (Figs. 13 and 14), i.e. the effects are different. This result emphasizes the fact that the mechanism by which residual stresses are generated is different on the surface than within the material. Therefore, residual stress generation on the surface is analysed separately (Figs. 11 and 12). The conclusion for the effects plotted on the surface in an axial direction (Fig. 11) is as follows: higher cutting speed (Vc ) and larger nose radius (Re ) produce more tensile stress. Feed rate (f) has a positive effect for low values but a negative effect for high values (f × f). Cutting depth (ap ) and rake angle (γ) are not significant i.e. their effect on residual stress is negligible. There is a 2nd order interaction between cutting speed and feed rate (Vc × f) although relatively insignificant. Nose radius (Re ) and cutting speed (Vc ) have a positive effect in a tangential direction (Fig. 12). Rake angle (γ) has in this case a negative effect. An interaction occurs between nose radius and rake angle (Re × ␥). Residual stress generation below the surface (Figs. 13 and 14) is dominated by feed (f), nose radius (Re ), rake angle (γ) and interactions between them. Nose radius (Re ) has a positive effect in both directions i.e. the higher the nose radius, the more tensile or less compressive are the residual stresses. Feed rate (f) and rake angle (γ) have negative effects even though their significance is lower than for the nose radius (Re ). Feed rate square (f × f) appears in an axial direction. Cutting speed (Vc ) and depth of cut (ap ) do not influence the residual stress generation below the surface. All the effects shown in Figs. 11–14 are scaled and centred for a better understanding of the results. The most significant interaction is that between nose radius and rake angle (Re × γ) and it therefore deserves special attention. Two factors interact when the effect of one of them depends on the level of the other. Hence in this study, the effect of nose radius depends on how large the rake angle is. For low rake angles, γ low in Fig. 15, nose radius has a positive effect on the residual stresses generated 30 m below the surface. In this case, the larger the nose radius, the higher the stresses. On the other hand, if the rake angle is high, γ high in Fig. 15, the effect of nose radius becomes negative. This shows a clear interaction between nose radius and rake angle.
87
Fig. 15. Interaction between rake angle and nose radius at 30 m below the surface in feed direction (axial).
Fig. 16. Surface roughness in test 8.
3.2. Surface roughness The roughest surface was obtained in test 8, and the smoothest in test 10 (Figs. 16 and 17). Fig. 17 illustrates that a smooth surface identical to that obtained in grinding can be achieved by hard turning if the cutting parameters are properly selected. Fig. 18 shows R2 and Q2 values for Sa, St and Sz. Not surprisingly, the highest model quality corresponds to Sa (arithmetical average of the surface), and the poorest to St (height deviation between the lowest and the highest points of the surface). This difference is due to the robustness in
Fig. 17. Surface roughness in test 10.
88
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
Fig. 18. Summary of fit for surface roughness.
Fig. 21. Schematic illustration of the residual stress level after hard turning in 18MnCr5.
Fig. 19. Effects plotted for surface roughness.
ducted within a cutting data range that is normally used in standard production. During the testing the tool wear was kept below Vb 0.05 mm. With this knowledge it is possible to predict and control the residual stresses level down to a depth of approximately 50 m below the surface when hard turning case carburised steel (Fig. 21). Hard turning has a negligible effect on the residual stresses below 70 m. Below this depth, the stress level is due to heat treatment and/or other process parameters prior to cutting. The generation of residual stress is a complex interaction between thermal and mechanical factors. The two opposing phenomena interact with each other to create the final residual stress pattern in the component. Table 5 summarizes thermal and mechanical generation as well as residual stress generation when changing the cutting parameters. Table 5 should be seen as a general explanation of how residual stress is generated. In the following discussion the five cutting parameters will be described. Higher cutting speed increases the temperature (more heat is generated), which causes tensile residual stress on the surface. Since the chip removes most of the heat, high temperature does not penetrate deeply into the workpiece and therefore does not affect the residual stress generation below
Table 5 Residual stress generation summary Fig. 20. Nose radius effect on surface roughness.
Parameter S0 Thermal generation Mechanical generation RS() MODDE
the calculation of Sa compared to St. In other words; Sa is more predictable than St, the latter being more random. As expected, feed rate (f) has a negative influence while nose radius (Re ) has a positive effect on surface roughness, see Fig. 19. It can also be seen that feed and nose radius interact (f × Re ) although not very significantly. More important is the nose radius square effect (Re × Re ) shown in Fig. 20.
4. Discussion In this study it has been shown how different factors affect the residual stress level at various depths. The tests were con-
Vc f ap γ Re
↑ ↑ – ↑ ↑
↓ ↑ – ↑ ↑
↑+T ↓−C – ↓−C ↑+T
Vc f ap γ Re
S10–S50 Thermal generation – ↑ – ↑ ↑
Mechanical generation ↓ ↑ – ↑ ↑
RS() MODDE – ↓−C – ↓−C ↑+T
S70–S130 Hardening process
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
the surface. If a subsequent honing process follows hard turning, the tensile residual stresses on the surface can be removed. Secondly, the level of compressive stress below the surface increases constantly with higher feed rates. This can be a consequence of the fact that a higher feed rate generates increased cutting forces and therefore more plastic deformation. Thirdly, when changing the cutting depths, the axial force is only slightly increased. However, the tangential force corresponds directly to the cutting depth and consequently does not contribute to the subsurface deformation. Therefore, the residual stresses should not be affected, which is consistent with the present results. Jacobson [5] also investigated depth of cut, and the result was that depth of cut does not affect the amount of residual stress generated in the hard turning of M50 steel. This is especially useful for the precision manufacturing of products with an initial out of roundness, as it is possible to make several calibration cuts to improve the roundness without affecting the residual stresses. The product suffers no dimensional distortion due to unevenness. Fourthly, a more negative rake angle produces increased compressive stress both on the surface and below. In the test plan the rake angle changed from −6 to −21. With a more negative rake angle, the mechanical generation of the subsurface increase, which produces greater compressive stress is due to the increase in passive cutting force. The higher temperature engendered by a more negative rake angle does not affect the stress level. Thiele and Melkote [7] presented a similar relationship. Jacobson [5] stated that the effective rake angle influences the amount of residual stress induced in the workpiece. Finally, a larger nose radius produces less compressive stress on the surface and subsurface. This is due to the fact that the contact area is larger and thus the specific cutting force (force per unit of area) is lower. Jacobson [5] found the same correlation in M50 steel: a smaller nose radius generates more compressive stress. The results of this study may be used to understand the mechanism by which residual stresses are generated in hard turning. The study can also be used to predict the stress level for future applications. Our ambition is to design a mathematical tool to determine the correct cutting parameters in order to produce a specific residual stress profile. As an example, Fig. 22 shows the predicted residual stresses 30 m below the surface in a tangential direction as a function of nose radius and rake angle. From a study of the surface roughness values Sa, Sz and St, it can be concluded that they are at the same level as for a ground workpiece. As expected, feed rate and nose radius are among the most influential parameters. Fig. 23 shows the predicted Sa surface value as a function of nose radius and rake angle. However, the feed rate square does not appear to be significant as in the purely geometrical expression of Rt (Eq. (1)). The Rt equation obtained by regression is shown in Eq. (2). Rt =
f2 8Re
(1)
89
Fig. 22. Residual stress prediction at 30 m below the surface in cutting speed direction (tangential).
Rt = 5.71 + 31.483f + Re (−4.162 + 0.740Re − 5.125f )
(2)
It is important to note that even though hard turning can produce equally good surface values as grinding, it is difficult to compare their functionality. A hard turned and a ground surface with the same Ra-value do not have the same tribology behaviour. Moreover, structural surface change in the workpiece introduced by the machining process is an important consequence of the finishing process. This surface or subsurface modification occurs because of localized and rapid thermal mechanical work, resulting in metallurgical transformation and/or at times chemical interactions. T¨onshoff et al. [11] claimed that change of the workpiece surface could be essential for the fatigue life of components. When hard turning is used as the finishing operation, it is very important to
Fig. 23. Surface roughness, Sa prediction as function of nose radius and feed rate.
90
F. Gunnberg et al. / Journal of Materials Processing Technology 174 (2006) 82–90
minimize the microstructure change. By using honing as a subsequent operation, the depth of cut is approximately 10 m and the heat effect layer is removed.
Acknowledgements The authors wish to thank Vinnova for their financial support and the AIS 13 project members for their contribution.
5. Conclusions Tests were performed to gain a comprehensive understanding of how tool geometry as well as cutting parameters affect the residual stresses and surface topography in 18MnCr5 steel. The following conclusions can be drawn from the results: • Three different sub-models were created: Surface (S0), (S10–S50) 10–50 m below the surface and (S70–S130) 70–130 m below the surface, due to the fact that different cutting parameters have a different effect on the level of residual stress generated on the surface and below the surface. • Cutting speed increases the tensile residual stress on the surface (S0). The heat generated from higher cutting speeds does not penetrate more deeply into the workpiece, and therefore does not affect the sub-model S10–S50. • Increased feed generates higher compressive stresses. • The cutting depth does not affect residual stresses. • A more negative rake angle produces more compressive stress in both S0 and S10–S50 models. • By controlling the cutting parameters, it is possible to generate tailor-made stresses in the product, which can prolong the service life of the machined component. • Feed and nose radius have the greatest effect on the geometric surface values. Sa values comparable to grinding can be achieved.
References [1] H.K. T¨onshoff, C. Arendt, R. Ben Amor, Cutting of hardened Steel, Ann. CIRP 49 (2) (2000) 547–566. [2] M. Field, J. Kahles, Review of surface integrity of machined components, Ann. CIRP 20 (1971) 153–163. [3] C.R. Liu, S. Mittal, Optimal pre-stressing the surface of a component by superfinish hard turning for maximum fatigue life in rolling contact, Wear 219 (1998) 128–140. [4] Y. Matsumoto, F. Hashimoto, G. Lahoti, Surface integrity generated by precision hard turning, Ann. CIRP 48 (1999) 59–62. [5] M. Jacobson, Surface integrity of hard-turned M50 steel, in: Proceedings of the Institution of Mechanical Engineers, Part B, J. Eng. Manuf. 216 (1) (2002) 47–54. [6] M. Jacobson, D. Patrik, G. Fredrik, Cutting speed influence on surface integrity of hard turned bainite steel, J. Mater. Proc. Technol. 128 (1–3, 6) (2002) 318–323. [7] J.D. Thiele, S.N. Melkote, Effect of cutting edge geometry and workpiece hardness on surface generation in the finish hard turning of AISI 52100 steel, J. Mater. Proc. Technol. 94 (1999) 216–226. [8] E. Brinksmeier, J.J. Cammett, P. Leskovar, J. Peters, H.K. Tonshoff, Residual stresses-measurements and causes in machining processes, Ann. CIRP 31 (1982) 491–510. [9] P. Vomacka, H. Walburger, Residual stresses due to hard-machiningindustrial experiences, in: Proceedings of the 5th European Conference on Residual stresses, Switzerland, 2000, pp. 592–597. [10] W. K¨onig, A. Berktold, K-F. Koch, Turning versus grinding: a comparison of surface integrity aspects and attainable accuracies, Ann. CIRP 42 (1993) 39–43. [11] H.K. Tonshoff, H.-G. Wobker, D. Brandt, Potential and limitation of hard turning, SME Tech. Papers (1995) MR95–MR215.