Tool wear effects on white and dark layer formation in hard turning of AISI 52100 steel

Tool wear effects on white and dark layer formation in hard turning of AISI 52100 steel

Wear 286–287 (2012) 98–107 Contents lists available at ScienceDirect Wear journal homepage: www.elsevier.com/locate/wear Tool wear effects on white...

2MB Sizes 1 Downloads 20 Views

Wear 286–287 (2012) 98–107

Contents lists available at ScienceDirect

Wear journal homepage: www.elsevier.com/locate/wear

Tool wear effects on white and dark layer formation in hard turning of AISI 52100 steel A. Attanasio a,∗ , D. Umbrello b , C. Cappellini a , G. Rotella c , R. M’Saoubi d a

University of Brescia, Via Branze 38, 25123 Brescia, Italy University of Calabria, Ponte Pietro Bucci 44C, 87036 Rende (CS), Italy Politecnico di Torino, Corso Duca degli Abruzzi, 24, 10129 Torino, Italy d Seco Tools AB, R&D Materials and Processes, Fagersta, SE-73782, Sweden b c

a r t i c l e

i n f o

Article history: Received 13 November 2010 Received in revised form 1 July 2011 Accepted 1 July 2011 Available online 7 July 2011 Keywords: Tool wear White and dark layers Hard turning AISI 52100

a b s t r a c t In the present investigation a series of orthogonal hard turning tests were conducted to study the effects of tool wear and cutting parameters (cutting speed and feed rate), on white and dark layer formation in hardened AISI 52100 bearing steel, using PCBN inserts. Experimental results were presented including quantification of tool wear and microstructure analysis of the machined surfaces. The experimental results were compared with a newly developed finite elements (FE) model that enables to capture the effect of cutting conditions and tool wear on the microstructural changes occurring at the machined surface. The results showed that cutting regime parameters and, especially, tool wear affect noticeably white and dark layers formation. © 2011 Elsevier B.V. All rights reserved.

1. Introduction The meaningful developments of advanced superabrasive tool materials, have considerably increased the interest in cutting hardened steels. The comparison between hard turning and grinding operations has underlined the improvement of hard turning in terms of flexibility, ability to achieve higher metal removal rates (MRR), possibility to operate without the use of coolants and capability to manufacture complex workpiece geometry in one setup reaching comparable workpiece quality [1–3]. For the above mentioned reasons, hard turning is replacing grinding in many industrial applications. However, as reported in [4] and [5] both grinding and hard turning processes can produce material structural changes in workpiece surfaces. This metallurgical transformation is correlated to the severe plastic deformation and/or to the intense, localized, rapid thermal mechanical loading at which the workpiece material is subjected during the cutting process [6–8]. Therefore, the metallurgical structure of the machined surface and subsurface can differ

∗ Corresponding author. Tel.: +39 030 3715584; fax: +39 030 3702448. E-mail addresses: [email protected] (A. Attanasio), [email protected] (D. Umbrello), [email protected] (C. Cappellini), [email protected] (G. Rotella), [email protected] (R. M’Saoubi). 0043-1648/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.wear.2011.07.001

significantly from the bulk material structure due to the induced formation of white and dark layers. In the latter these may have negative influence on the component life [2,9,10]. White layer is the result of a microstructural modification of the martensite structure, that appears white after etching, when observed under a light optical microscope. It has been shown that white layer contains an untempered martensitic structure, with a hardness higher than the dark layer and bulk material [11]. The dark layer is a dark region under the white layer; and it consists of overtempered martensite (OTM). Such layer was generally found to be softer than the bulk material [8]. In the recent years, several experiments were performed to investigate the white and dark layers formation [12–16]. By studying several cutting conditions, it has been often shown that the higher the cutting speed the higher the white layer thickness [17]. It has been also noted that the depth of cut does not affect the white layer depth while there is a slight increase with increasing uncut chip thickness [18]. In addition, considering the 3D machining conditions, it has been also shown that the tool nose radius affects the white layer formation, in particular its thickness decreases with increasing the radius [19]. Most of the above mentioned studies have taken into account one or more of the factors affecting white and dark layer formation such as the cutting speed, tool edge condition, feed or uncut chip thickness, tool nose radius and initial workpiece hardness. It has been also reported that white and dark layers occur when tool wear reaches a critical level [18–20], in particular, both of them increase

A. Attanasio et al. / Wear 286–287 (2012) 98–107

99

Fig. 3. Experimental cutting data.

Fig. 1. Experimental test set-up.

in depth by increasing tool wear confirming the previous remarks in both orthogonal turning and three-dimensional machining cases [17–21]. In this paper, the effects of tool wear and cutting parameters on the white and dark layers formation are further examined using both experimental and numerical approaches. The experimental part consists of orthogonal turning tests on disks made of AISI 52100 (100Cr6) hardened steel using PCBN inserts where the surface and subsurface of the machined workpiece are analysed. The experimental results are compared with those obtained from newly developed 2D FE simulation model for white and dark layer prediction. Additionally, a 3D FE analysis was subsequently introduced in order to investigate the effect of process parameters on white layer formation and the numerical results were compared with experimental data available in literature. The results demonstrate the possibility of forecast the material microstructural change as a function of cutting parameters, tool geometry and wear.

2. Experimental tests The machining tests consisted of orthogonal turning of disks made of AISI 52100 hardened steel with a thickness of 1.2 mm and an initial diameter of 150 mm. The experimental test set up is shown in Fig. 1. AISI 52100 hardened steel was selected because of its industrial relevance, its hardenability and because of its attitude to generate subsurface microstructure alterations (also called

white and dark layers) after machining under aggressive cutting conditions and presence of excessive tool wear. In order to set the workpiece hardness to a defined value, a heat treatment of quenching and tempering was designed and applied. As reported in Fig. 2, a pre-heating at 550 ◦ C was followed by a cooling at 175 ◦ C. After that, a quenching treatment at 860 ◦ C was performed. Both quenching and tempering heat treatments were achieved in salt furnace. The application of this technique was necessary in order to limit the deformation of the specimens, because the low thickness of the disks increases the distortion effects due to the thermal cycle of the heat treatment. After both quenching and tempering treatments the specimens were left to cool in air. At the end of the cooling stage after the quenching treatment the hardness of the specimens was measured obtaining values ranging from 63 HRC up to 65 HRC. In order to reduce the material hardness to 57 HRC a subsequent tempering treatment in a furnace at 390 ◦ C was carried out. Finally, the disks were gently ground to obtain flat surfaces. The machining tests were performed on a computer numerical control (CNC) lathe equipped with an induction servo motor with a nominal power of 7 kW. The disks were machined using PCBN inserts (ISO TNGN 110308S-01020 CBN100). For ensuring a negative rake angle () of −8◦ , an inclination angle () of 0◦ and an entering angle () of 90◦ a toolholder ISO CTFNR3225P11-PL was employed. During the experimental campaign the feed rate and cutting speed were varied as reported in Fig. 3. All the other cutting parameters were kept constant: the tests were performed in dry condition and the depth of cut was set equal to the disks thickness (1.2 mm). It should be pointed out that the process parameters employed were chosen in order to promote rapid tool wear and facilitate the formation of white and dark layers in the machined surface. Current industrial practise requires more gentle cutting conditions so that microstructure alteration can be minimised. Metallographic inspection of polished and etched machined specimens was carried out using a scanning electron microscope

Pre-heating in furnace at 550±5°C

Quenching in salt furnace at 860±5°C

Cooling in salt furnace at 175±5°C

Air co oling

Tempering in furnace at 390±5°C

Air cooling

Measured hardness 57±1HRC

Measured hardness 64±1HRC Fig. 2. Heat treatment cycle.

100

A. Attanasio et al. / Wear 286–287 (2012) 98–107

(SEM) in order to investigate the influence of the cutting parameters (vc and f) and tool wear on the microstructural changes. 2.1. Tool wear analysis During each test the tool wear measurement in terms of both flank (VB) and crater wear (KT) was performed at regular time intervals. The catastrophic tool failure criterion was selected to identify the end of every test. In order to define time intervals for each test, the final tool life time needed to be estimated. Poulachon et al. [20] developed an analytical model that enables to estimate the tool life time. This model, that correlates the process parameters with the tool life time, is reported in Eq. (1): V · T G f E · dcF ·

 H D H0

=C

(1)

where H0 is equal to 60 HRC, H is the workpiece hardness and it is equal to 57 HRC, T is the tool-life expressed in minutes, V is the cutting speed expressed in m/min, f is the radial feed rate (orthogonal cutting conditions) in mm/rev, and dc is the depth of cut and it is equal to 1.2 mm. Since the workpiece and tool materials are similar as those investigated in [20], the exponents of Eq. (1) were selected according to the literature. In particular, G was set equal to 0.285, E equal to 0.335, F equal to 0.112, D equal to 1.07, and C equal to 172. The analytical results of Eq. (1) were found in good agreement with the experimental ones, thus, it was used to estimate the final tool life time and subsequently for defining the cutting time intervals. The estimated final tool life time was then divided in ∼10 substeps except for the case with low cutting speed where more than 20 sub-steps were employed due to longer tool life time. Flank wear was acquired by means of an optical coordinate measuring machine CMM (Mitutoyo QS200) allowing a measuring resolution of 0.5 ␮m and an accuracy lower than 2.5 ␮m. Crater wear depth was measured by means of a profilometer. Fig. 4 shows the flank (VB) and the crater (KT) wear evolution as a function of the experimental duration of each test. From these graphs it is evident that the insert breakage occurs when the crater depth is high, while the flank wear reaches low values. The severe crater wear on the PCBN inserts leading to premature tool failure

is mainly due to the severe cutting parameters used in this study that generate high load and temperatures on tool. Moreover, it can be observed that the flank wear rate is strongly influenced by the cutting speed: the higher the cutting speed, the higher the flank wear rate. While the feed rate does not show a significant influence on flank wear rate (Fig. 4a). As reported in Fig. 4b, the cutting speed shows the same influence on the crater wear rate. When considering the feed rate influence on crater wear rate, it is possible to highlight that the tests with cutting speed equal to 250 m/min and feed rate equal to 0.075 mm/rev or 0.1 mm/rev display a similar crater wear rate. On the other hand, when the feed rate is set equal to 0.125 mm/rev maintaining the same cutting speed, a higher crater wear rate value can be observed. This behaviour could be explained considering the tool geometry since the inserts utilized during the experimental tests are characterized by a chamfer equal to 0.1 mm. Hence, when the feed rate values are higher than the chamfer width, an increase of the tool crater wear rate is generally expected. Fig. 5 shows the experimental tool life (T) and the removed volume (Vol) for every cutting condition. The removed volume is calculated according to Eq. (2): Vol = f · dc · vc · T

(2) [mm3 ],

where Vol, removed volume f, feed rate [mm/rev], dc is the depth of cut, vc , cutting speed [mm/min] and T, tool life time [min]. As expected tool life and removed volume are strongly influenced by the cutting speed: the higher the cutting speed, the lower the tool life and the removed volume. It is also possible to state that with increasing of the feed rate the tool life slightly decreases. In contrast, the influence of feed rate on removed volume needs to be deeply studied with further experimental test. 2.2. White and dark layers analysis After machining, samples were sectioned from each disk specimen for microstructure analysis. The samples were polished and etched for about 5 s using 5% Nital solution to observe white and dark layers using a light optical microscope (500×) and a scanning electron microscope (SEM). The white layer thickness

Fig. 4. Tool wear evolution: (a) flank wear, VB [␮m]; (b) depth of crater, KT [␮m].

A. Attanasio et al. / Wear 286–287 (2012) 98–107

101

Fig. 5. Tool life and removed volume.

values measured using optical microscope were found to be consistent with those measured by SEM. The micrographs and SEM images for the tests carried out at 250 and 350 m/min, with feed rate equal to 0.1 mm/rev and when the critical flank tool wear (VB*) was reached, are reported in Fig. 6. From Fig. 6 it is also possible to observe that typical SEM images show a fea-

tureless structure in the white layer region, while the optical micrograph images reveal the extent of both white and dark layers. Fig. 7 shows the variation of white and dark layer thicknesses as a function of cutting parameters and tool flank wear. It can be generally observed that an increase in tool wear and cutting speed

Fig. 6. Optical observation (a) SEM image (b) for test at 250 m/min, 0.1 mm/rev when critical flank wear (VB*) was relieved. Optical (c) SEM (d) images for test at 350 m/min, 0.1 mm/rev and critical flank wear (VB*).

102

A. Attanasio et al. / Wear 286–287 (2012) 98–107

resulted in a systematic increase of white layer thickness. In contrast, white layer thickness is marginally influenced by feed rate for values lower than 0.1 mm/rev, while it rises when feed rate becomes higher than 0.1 mm/rev (i.e., the chamfer with). This is because high cutting speed, high feed rate and tool wear generate higher temperature into the workpiece leading to higher white layer thickness. As shown in Fig. 7b, the dark layer thickness increases with the tool wear as well. When the cutting speed and the feed rate are considered, their influence is not clear and further analysis based on statistical approach (ANOVA analysis) would be needed to clarify the interactions between cutting parameters, tool wear and dark layers. 3. 2D Fe analysis 3.1. 2D Fe modeling The numerical model proposed by Umbrello et al. [11,21,22] was implemented to simulate the tool wear influence on white and dark layers formation: 2D plane strain simulations were carried out using SFTC Deform 2D® V10.1. High mesh density was defined on the workpiece: in particular, the elements located around the cutting edge and along the machined surface were fifty times as dense as the other ones (average element edge length ≈ 5–6 ␮m). A hardness-based flow stress model [23] was implemented in the FE code as material constitutive law for describing the behaviour of AISI 52100 bearing steel. As far as the white and dark layer formation is concerned, two simple empirical equations based on both the thermal effects and the hardness modification (HRC) were considered. In particular, the latter variable was associated to white and dark layers formation since the material behaviour in the FE model was described by a hardness-based flow stress. Therefore, for quenching process (related to white layer formation), the empirical model implemented in the FE code by a subroutine was: HRCquenching = J

 67 − HRC

initial

1030 − TAUS



(T − TAUS )

(3)

where HRCquenching is the hardness modification due to quenching heat treatment, HRCinitial is the initial material hardness (57 HRC in such case), TAUS is the austenite-start temperature and T the current temperature in the element, 67 represents the highest

metallurgical state (in HRC hardness) that can be reached by AISI 52100 when it is quenched in oil starting from 1030 ◦ C [24]. In contrast, for tempering process (associated with the dark layer formation), the empirical model was: HRCtempering = K

 HRC



− HRCFT (TDLSTART − T ) TAUS − TDLSTART initial

(4)

where HRCtempering is the hardness modification due to tempering heat treatment, HRCFT is the fully tempered material hardness when temperature corresponds to TAUS , TDLSTART is the temperingstart temperature and T is the current temperature in the element. At the beginning the austenite-start temperature (TAUS ) was ranging between 550 ◦ C and 650 ◦ C (depending of the initial material hardness) according to the experimental study conducted by Ramesh [25], while TDLSTART was set as illustrated in ASM Handbook [24]. Then, both TAUS and TDLSTART together with material dependent constants the J and K (which depend on both the investigated material, the initial hardness, process parameters and, consequently, temperature) were empirically determined during FE model calibration [21,26] and their correlation with the material hardness and process parameters was found by a statistical approach [22]. Finally, parameter HRCFT was again derived from [24] in which the variation of hardness with tempering temperature in AISI 52100 is reported. The proposed updating strategy utilized in the FE model in order to simulate the white and dark layer formation is discussed here: each element undergoes quenching or tempering as a function of the temperature calculated during FE simulation. If temperature is higher than TAUS , Eq. (3) is invoked. On the other hand, if temperature ranges between TAUS and TDLSTART , Eq. (4) is invoked. Since quenching phenomenon is physically non-diffusive, thus it is reasonable to apply Eq. (3), if the hypothesis of fast cooling is consistent. In contrast, tempering phenomenon is here empirically modelled in order to obtain a consistent relationship based on the current temperature. It is important to underline that the temperature is checked at each time step in the simulation and for each element of the workpiece in order to update the current element hardness. The latter is stored before the next simulation step. Furthermore, it is also important to highlight that if the hardness variation occurs, the

Fig. 7. White (WL) (a) and dark (DL) (b) layer thickness vs. flank toot wear at varying of cutting speed and feed rate.

A. Attanasio et al. / Wear 286–287 (2012) 98–107

103

Fig. 8. Simulative 2D microstructural changes: (a) 250 m/min, 0.1 mm/rev, VB = 0.024 mm; (b) 350 m/min, 0.1 mm/rev, VB = 0.029 mm; (c) 350 m/min, 0.1 mm/rev, VB = VB*.

104

A. Attanasio et al. / Wear 286–287 (2012) 98–107

Table 1 Experimental conditions simulated by FE 2D analysis. Vc [m/min]

f [mm/rev]

VB [mm]

150 250 250 250 350

0.100 0.075 0.100 0.125 0.100

0.025 0.025 0.024 0.029 0.029

0.050 0.050 0.050 0.046 0.052

0.075 0.075 0.075 0.075 VB*

0.100 0.100 0.100 VB* –

VB* VB* VB* – –

Fig. 8 reports the predicted white and dark layers thicknesses, for three investigated tests when different cutting parameters and flank wear were considered. In particular, for each case reported in Fig. 8 it can be observed that the updated hardness values during the simulation which, as reported above, are related to white layer (updated hardness higher than the initial hardness) and dark layer (updated hardness lower than the initial one).

3.2. 2D FE validation and results discussion material strength is locally different (harder or softer), reflecting the reality of machining process. A series of 25 different cases (Table 1) were simulated in order to validate the capability of the proposed FE model for predicting the effect of cutting conditions and tool wear on the microstructural changes occurring at the machined surface. Simulation cutting time was set long enough in order to reach and, consequently, to maintain the mechanical and thermal steady-state conditions. In contrast, for each test the measured tool flank wear was modelled at the begin of numerical simulation and then it was kept constant, neglecting wear rates due to mechanical abrasion and thermal diffusion since the aim was to investigate the effect of known flank wear on microstructural alterations. Furthermore, the effect of crater wear on the microstructural changes in machined surface was also neglected.

The effectiveness of the proposed FE model is demonstrated by comparing numerical results from FE simulations with those experimentally found, based on microstructural changes observed on the machined surface and subsurface under different conditions employed in terms of both cutting parameters and tool flank wear. As shown in Fig. 9 the predicted results are in good agreement with those obtained experimentally. In particular, it is possible to observe that the white layer thickness increases with the increasing of the flank wear while it decreases when the cutting speed increases. Finally, the dark layer seems to be marginally influenced by the feed rate when small flank wear values were detected, while it decreases with the increasing of feed rate when higher flank wear values were measured. The reasons of these trends are related to both the heat-affected zone (HAZ) and the relative maximum temperature [22]. It is

Fig. 9. Simulated and experimental microstructural changes at varying cutting conditions and tool wear levels during orthogonal machining.

A. Attanasio et al. / Wear 286–287 (2012) 98–107

105

Fig. 10. Simulative 3D microstructural changes at Vc = 180 m/min, f = 0.05 mm/rev; d.o.c. = 2 mm, tool nose radius, rε = 1.6 mm, initial workpiece hardness = 61HRC: (a) VB = 0.1 mm; (b) VB = 0.2 mm.

well known that temperature rises with the cutting speed while HAZ decreases. Therefore, when severe cutting speeds are utilized, deeper zones in which temperature is higher than TAUS can be observed; thus larger white layer thickness is recognized. In contrast, due to smaller HAZ thickness, the region where temperature ranges within TDLSTART and TAUS is reduced, consequently a reduced amount of overtempered martensite (dark layer) can be observed. Moreover, as previously mentioned, the white layer increases with increasing of the feed rate; in contrast dark layer thickness remains almost similar for small flank wear while it decreases with increasing of the feed rate when higher flank wear values are observed. The reason is once again related to the heat-affected zone (HAZ) and the maximum reached temperature [11]. If fact, higher feed rates generate higher temperatures on the machined surface while slight rising of the HAZ are detected. Therefore, when high feed rates are utilized, deeper zones on which the temperature is higher than TAUS can be observed. In contrast, since the HAZ thickness remains almost constant or slightly increases, the dark layer is reduced. Finally, both white and dark layers increase with increasing of flank wear, because worn tools generates higher temperatures on the machined surface and deeper HAZ due to the higher frictional contact effect.

[19]. The solver employed in the simulations is the Incremental Lagrangian one and the simulation is run till mechanical and thermal steady states are reached. Concerning the mesh density, the elements located around the cutting edge and along the machined surface were fifty times as dense as the other ones (average element edge length ≈ 10 ␮m). According to the experimental data obtained by Chou and Song [19], a series of 9 different cases (Table 2) were simulated in order to validate the capability of the proposed FE model for predicting the effect of tool nose radius, rε , and tool wear on the white layer formation occurring at the machined surface. Tool material properties and tool geometry, were defined as those reported in [19], while workpiece initial hardness was equal to 61 HRC. It is important to underline that, although 3D FE model permits to predict both white and dark layers, only white layer depth was reported since no dark layer thickness data was reported in [19]. Furthermore, in 3D FE modeling (as for the 2D FE model), only measured flank tool wear was modelled and the possible effects of crater wear on the white layer in machined surface was neglected. Fig. 10 reports the predicted microstructural changes, for two of the investigated tests when different flank wear levels were considered. In particular, both predicted white layer (hardness higher than bulk material hardness) and dark layer (hardness lower than bulk material hardness) can be observed. However, as previously cited, only white layer formation was considered for results discussion and model validation.

4. 3D FE analysis 4.1. 3D FE modeling The 3D model was implemented using the same simulation parameters and models set in the 2D one. The SFTC FE software Deform 3D® V6.1 was utilized to simulate the tool wear influence on white layer formation during hard machining of cylindrical bars

Table 2 Experimental conditions simulated by FE 3D analysis. Vc [m/min]

f [mm/rev]

d.o.c. [mm]

rε [mm]

VB [mm]

180 180 180

0.050 0.050 0.050

2.000 2.000 2.000

0.800 1.600 2.400

0.100 0.100 0.100

0.200 0.200 0.200

106

A. Attanasio et al. / Wear 286–287 (2012) 98–107

higher cutting speed generates thicker white layers and thinner dark layers. In addition, smaller feed rates moderately influence the white layers thickness, while the latter rises with higher feed rate. In contrast, the dark layer thickness decreases with the increasing of the feed rate, especially when flank wear values higher than 0.075 mm were observed. Finally, a FE model, able to take into account the microstructural changes on the machined surface under different cutting conditions, was proposed. The reasonable agreement obtained between the experimental and numerical results indicates that the proposed FE model is suitable for studying the effect of cutting conditions and tool wear on white and dark layers formation during hard machining of AISI 52100 steel. Acknowledgement Fig. 11. Simulated and experimental [18] white layer depths at varying of tool nose radius and flank tool wear during three-dimensional machining (initial workpiece hardness = 61 HRC, cutting speed = 180 m/min; feed rate = 0.05 mm/rev; d.o.c. = 2 mm).

4.2. 3D FE validation and results discussion Effectiveness of the proposed 3D FE model was also demonstrated by comparing numerical results obtained from FE simulations with those experimentally found by Chou and Song [19]. White layer formation on the machined surface was investigated under different conditions in terms of both tool nose radius and tool flank wear. As illustrated in Fig. 11 the predicted results are in satisfactory agreement with those obtained experimentally. In particular, it is possible to observe that the white layer thickness increases with the increasing of flank wear, although the predicted values are always higher than those experimentally observed [19]. The reason for this discrepancy may be related to the limits in defining the 3D mesh element dimension even when maximum allowable element number was set. In other words, since the element dimension cannot be easily optimized as in FE 2D modeling, the prediction of the white layer thickness is numerically over estimated. Fig. 11 also shows white layer depth (average) versus tool nose radius when hard turning with worn tools is performed; it highlights that white layer depth tends to increase with nose radius at large wear land. This trend is also confirmed by FE 3D numerical results. White layer depths seem to be minimum for the medium tool nose radius (rε = 1.6 mm). Such circumstance is not very clear and need to be deeper studied as well as further cutting conditions and initial material hardness should be investigated in the future in order to better understand if FE model developed in 2D FE environment can utilized as it is or additional three-dimensional calibration is still required. 5. Conclusions As expected, experimental observations suggest that crater wear rate is influenced by both cutting speed and feed rate, while flank wear rate seemed to be mainly effected by cutting speed. This can be related to the wear mechanisms. When the crater wear is present, the wear mechanisms are the abrasion, deeply affected by cutting speed, and the diffusion, heavily influenced by cutting temperature. On the other hand, the flank wear mechanism is mainly due to abrasive phenomena which is strongly affected by cutting speed. The tool life and the removed material volume are mainly influenced by the cutting speed, while the influence of feed rate on material removed volume has to be further investigated with additional experimental tests. Furthermore, it was found that the thickness of white and dark layers increases with increasing of tool flank wear. Moreover,

The authors gratefully acknowledge Serena Di Renzo and Serafino Caruso for their contribution in developing this research. References [1] F. Hashimoto, S.N. Melkote, R. Singh, R. Kalil, Effect of finishing methods on surface characteristics and performance of precision components in rolling/sliding contact, International Journal Machining and Machinability of Materials 6 (1–2) (2009) 3–15. [2] F. Klocke, E. Brinksmeier, K. Weinert, Capability profile of hard cutting and grinding processes, Annals of CIRP 54 (2) (2005) 22–45. [3] G. Poulachon, B.P. Bandyopadhyay, I.S. Jawahir, S. Pheulpin, E. Seguin, Wear behavior of CBN tools while turning various hardened steels, Wear 256 (2004) 302–310. [4] F. Hashimoto, Y.B. Guo, A.W. Warren, Surface integrity difference between hard turned and ground surfaces and its impact on fatigue life, CIRP Annals 55 (1) (2006) 81–84. [5] T.I. El Wardany, H.A. Kishawy, M.A. Elbestawi, Surface integrity of die material in high speed hard machining, Part 1: Micrographical Analysis, Journal of Manufacturing Science and Engineering 122 (2000) 620–631. [6] A. Ramesh, S.N. Melkote, L.F. Allard, L. Riester, T.R. Watkins, Analysis of white layers formed in hard turning of AISI 52100 steel, Materials Science & Engineering A 390 (2005) 88–97. [7] W.J. Tomlinson, L.A. Blunt, S. Spraggett, White layers on surface of ground EN24 steel. 1. Microstructure, composition, internal stress, and corrosion properties, Surface Engineering 5 (3) (1989) 229–233. [8] M.C. Shaw, A. Vyas, Heat-affected zones in grinding steel, Annals of CIRP 43 (1) (1994) 279–282. [9] S. Smith, S.N. Melkote, L. Curzio, E.T. Watkins, R. Allard, L. Riester, Effect of surface integrity of hard turned AISI 52100 steel on fatigue performance, Materials Science & Engineering A 459 (2007) 337–346. [10] Y.B. Guo, G.M. Janowski, Microstructural characterization of white layers by hard turning and grinding, Transactions of NAMRI/SME 32 (2004) 367–374. [11] D. Umbrello, L. Filice, Improving surface integrity in orthogonal machining of hardened AISI 52100 steel by modeling white and dark layers formation, Annals of CIRP 58 (1) (2009) 73–76. [12] H.K. Tönshoff, C. Arendt, R. Ben Amor, Cutting of hardened steel, Annals of CIRP 49 (2) (2000) 547–566. [13] J. Barry, G. Byrne, TEM study on the surface white layer in two turned hardened steels, Materials Science and Engineering A 325 (2002) 356–364. [14] Y.B. Guo, J. Sahni, A comparative study of hard turned and cylindrically ground white layers, International Journal of Machine Tools & Manufacture 44 (2004) 135–145. [15] A. Ramesh, S.N. Melkote, Modeling of white layer formation under thermally dominant conditions in orthogonal machining of hardened AISI 52100 steel, International Journal of Machine Tools & Manufacture 48 (2008) 402– 414. [16] S. Akcan, W.I.S. Shah, S.P. Moylan, S. Chandrasekar, P.N. Chabra, H.T.Y. Yang, Formation of white layers in steels by machining and their characteristics, Metallurgical and Materials Transactions A 33 (4) (2002) 1245–1254. [17] B. Zhang, Y.L. Shen, X. Tang, Y. Wang, Microstructures of surface white layer and internal white adiabatic W. shear band, Wear 211 (1997) 164–168. [18] Y.K. Chou, C.J. Evans, White layers and thermal modeling of hard turned surfaces, International Journal of Machine Tools & Manufacture 39 (1999) 1863–1881. [19] Y.K. Chou, H. Song, Tool nose radius effects on finish hard turning, Journal of Materials Processing Technology 148 (2004) 259–268. [20] G. Poulachon, A. Moisan, I.S. Jawahir, Tool-wear mechanisms in hard turning with polycrystalline cubic boron nitride tools, Wear 250 (2001) 576–586. [21] D. Umbrello, A.D. Jayal, S. Caruso, O.W. Dillon, I.S. Jawahir, Modeling of white and dark layer formation in hard machining of AISI 52100 bearing steel, Machining Science & Technology 14 (1) (2010) 128–147.

A. Attanasio et al. / Wear 286–287 (2012) 98–107 [22] D. Umbrello, Influence of material microstructure changes on surface integrity in hard machining of AISI 52100 steel, International Journal of Advanced Manufacturing Technology 54 (9) (2011) 887–898. [23] D. Umbrello, J. Hua, R. Shivpuri, Hardness based flow stress for numerical modeling of hard machining AISI 52100 bearing steel, Materials Science and Engineering A 347 (2004) 90–100. [24] ASM International, Heat treating of steels and surface hardening of steel ASM Handbook, Heat Treating, vol. 4, 1991.

107

[25] A. Ramesh, Prediction of process-induced microstructural changes in orthogonal hard machining, Ph.D. Thesis, Mechanical Eng., Georgia Inst. of Technology, 2002. [26] G.L. Manco, S. Caruso, G. Rotella, FE modeling of microstructural changes in hard turning of AISI 52100 steel, International Journal of Materials Forming 3 (Suppl. 1) (2010) 447–450.