Journal of Materials Processing Technology 211 (2011) 773–783 Contents lists available at ScienceDirect Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec Surface integrity in cryogenic machining of nickel based alloy—Inconel 718 F. Pusavec a,∗ , H. Hamdi b , J. Kopac a , I.S. Jawahir c a University of Ljubljana, Faculty of Mechanical Engineering, Askerceva 6, SI 1000, Ljubljana, Slovenia LTDS/ENISE, 58 rue Jean Parot, Saint Etienne Cedex 02 42023, France c University of Kentucky, Institute for Sustainable Manufacturing (ISM), 414C UK, CRMS Building, Lexington, KY 40506, USA b a r t i c l e i n f o Article history: Received 21 July 2010 Received in revised form 15 December 2010 Accepted 16 December 2010 Available online 24 December 2010 Keywords: Cryogenic machining Inconel 718 Surface integrity X-ray diffraction Sustainability a b s t r a c t In machining processes, a major quality related output is integrity of the machined part surface. In machining of difficult-to-cut materials, a drastic decrease in tool-life makes the machining process even more difficult. By considering the broader perspective of the machining system tailored towards sustainable operations, in this work an alternative—cryogenic machining is evaluated for machining performance. The surface integrity characteristics of machined surface as a function of depth have been analyzed for different combinations of cooling/lubrication machining conditions. The residual stresses on the machined surface and sub-surface, surface hardness, and surface roughness are among the significant characteristics studied in this work. The results show that cryogenic machining processes can be implemented to improve all major surface integrity characteristics, thus improving the final product quality level. © 2010 Elsevier B.V. All rights reserved. 1. Introduction Recent developments in aerospace, automotive, and power industry have seen increased application of novel hightemperature alloys. Nickel alloys are one such material that meets the need of resisting mechanical loads under high operating temperatures. These materials contain larger volume fractions of strengthening ′ precipitates and refractory elements than conventional superalloys, thereby offering higher strength at high temperatures. However, these superior mechanical and thermal properties, in addition to poor thermal conductivity, are making these material difficult to machine. The heat generated in the cutting zone stays there, and builds up to the extreme values, thus making the softening of the cutting tool material. This usually leads to rapid tool-wear, and consequently shorter cutting tool-life, and at the end to the deteriorated machined surface integrity. Surface roughness, hardness, residual stresses, etc., characteristics are considered as surface integrity factors, determining functionality and fatigue life of the final product. As the surface integrity is one of the most relevant performance characteristic of the final product machined surface quality, particularly when critical structure components are machined, numerous studies have been conducted to determine the correlation between the machining conditions and residual stresses. The earliest attempt was made ∗ Corresponding author. Tel.: +386 1 4771 769; fax: +386 1 4771 768. E-mail address: franci.pusavec@fs.uni-lj.si (F. Pusavec). 0924-0136/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2010.12.013 in the 1950s, examining the effect of varying rake angles on residual stresses, and concluding that residual stresses decreased with increasing rake angle (Henriksen, 1951). In the 1970s there was another notable study on the mechanical state of the machined layer affected by tool-wear, which describes the deformation in the machined layer by considering stress–strain curves and the associated impact of thermal effects due to tool-wear (Liu and Barash, 1976). Later on, there have been quite a few researches reporting the residual stresses induced by machining. Authors in Outeiro et al. (2004) reported that both thermal and mechanical loads affect residual stresses. Thermal loads for austenitic structures promote tensile stress, while the mechanical loads suppress them. The same observations were reported in Wardany et al. (2000), for high-speed machining. This is even more dominant in machining of high-temperature alloys, such as Inconel, where high temperatures are present and are causing undesired tensile residual stresses (Outeiro et al., 2008). Considering the rake angle, it was reported (Arunachalam et al., 2004a,b) that negative rake angle tends to generate compressive residual stress, whereas positive rake angle tends to induce tensile residual stresses. Further, the work of M’Saoubi et al. (1999) proves that creating a chamfer on the cutting edge has a similar effect as creating a negative rake angle, by which the maximum compressive residual stress in the sub-surface can be increased. Similar observations were reported from FEM studies in Liang and Su (2007) for orthogonal cutting and in Ozel and Zeren (2006) for oblique machining process case. It was concluded that a large cutting edge radius is associated with a deep temperature and stress fields compared to small cutting edge 774 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 radius. This is supported with the observations in work of Sasahar et al. (2004) where concluded that geometrically increasing cutting edge radius to a point that the edge roundness is larger than the uncut chip thickness, actually decreases the effective rake angle towards the negative value. Hence, a large cutting edge radius has a similar effect as a negative rake angle. Related to residual stresses, it is shown in Hua et al. (2005) on bearing steel and in Jang et al. (1996) on stainless steel that increased cutting edge radius also positively influences the compressive residual stresses. As increasing flankwear alters the cutting edge geometry, in Ordas et al. (2003) for hard turning and Sharman et al. (2006) for Inconel turning, authors proved that progressive tool-wear increases also the tensile residual stress at the surface as well as the compressive residual stress in the sub-surface. Additionally, excessive tool-wear, or the tool chippage, drastically affects the sub-surface metallurgy/white etching layer as reported in Axinte et al. (2006) and Li et al. (2009) for sintered and casted nickel alloys, respectively. With the tool-wear also the temperature in the cutting zone is increased. While it is known that temperature affects the surface integrity, authors in Ordas et al. (2003) proved that the significance of microstructure change in the machined layer due to the high temperature (mostly related to worn tools), can be identified by the diffraction peak full width at half maximum (FWHM). And as reported in newer work (Li et al., 2006), the distribution of plastic work using calibrated diffraction peak width (FWHM) can be determined. Nevertheless, the residual stresses in machined surface are affected also by machining parameters, cutting tools and cooling/lubrication fluid (CLF). Authors in both work (M’Saoubi et al., 1999; Outeiro et al., 2002) transparently reported that machining parameters highly influence residual stresses induced on the machined surface and the sub-surface (AISI 316L). General findings are pointing out that the sub-surface stress values also tended to be less compressive with increase in the cutting speed. In finding the influence of feed rate on residual stresses, the minimal stress values were more compressive and were located deeper in the workpiece as the feed rate increases, although the surface residual stresses barely varied with feed rate. In newer work on machining of high-temperature nickel alloys (Outeiro et al., 2008), authors did a step further and reported that the use of coated or uncoated WC tools also has an influence on residual stresses. They proved that coated WC tools result in slightly improved residual stress field in machined surface, on account of reduction in cutting temperatures, even in the case of difficult-to-cut materials. Additionally, in Arunachalam et al. (2004a,b) CBN cutting tools were analyzed in Inconel 718 under different CLFs. Results show that ceramic cutting tools induce tensile residual stresses with a much higher magnitude than CBN cutting tools. The use of coolant either results in compressive residual stresses or lowers the magnitude of the tensile residual stresses, whereas dry cutting always resulted in tensile residual stresses. Although numerous studies have analyzed the effects of machining parameters, tool geometry, etc., on machining-induced surface integrity characteristics, there is a gap in understanding the effects of cooling/lubrication on the surface integrity (Pusavec et al., 2010). There are some work on grinding process proving that cryogenic cooling can reduce the magnitude of the tensile residual stresses, in machining of different tool steels (Paul and Chattopadhyay, 1995) and in machining of stainless steels (Fredj et al., 2006), due to substantially reduced temperature in grinding zone. However, all this work is related to the nondefined tool geometry machining processes and is not related to high-temperature alloys, such as Inconel. As seen in overall review of published work, little work has been reported on residual stresses in machining of Inconel 718 and influence of CLFs. Additionally, there has been no published work related to residual stresses in cryogenic machining of this alloys. Therefore, this paper presents a study of the influence of cryogenic machining process on surface integrity measures, Fig. 1. Kinematics of cutting tool—work material engagement and the CLF delivery methods. aimed to evaluate the effect of varying cooling/lubrication conditions on the residual stress distribution in the machined layer on Inconel 718. Additionally, investigated measures are also hardness and surface roughness, all for the case of finish machining operations, as the final product surface and sub-surface quality characteristics (fatigue life, tool-life, etc.) are influenced by all three measures. 2. Experimental work 2.1. Experimental setup Machining experiments were performed on Inconel 718 alloy used for the jet engines in the aerospace industry. Experiments consisted of four conditions, each corresponding to different cooling/lubrication fluid usage: dry machining (dry), minimum quantity lubrication machining (MQL – 120 ml/h Coolube 2210EP), cryogenic machining (cryo – under 1.5 MPa pressure and a flow rate of ∼0.6 kg/min per nozzle) and the combination cryo + MQL machining. The delivery for each of the CLF is shown in Fig. 1 – dry machining and near-dry machining (MQL) – where oil mist is applied with one nozzle on the tool rake face in the direction of cutting zone shown by arrow A, Cryogenic machining—where liquid nitrogen (LN) is applied with two nozzles: one nozzle delivers LN onto the rake face in the direction of cutting zone (Arrow A), while the second nozzle delivers LN onto the workpiece to cool it just before the actual cut (Arrow C), and cryogenic–lubrication combination where liquid nitrogen is applied on the flank face of the cutting tool (Arrow B) and oil mist on the rake face (Arrow A). The turning experiments were conducted on the centerlessground Inconel 718 round bars with a diameter of 32 mm and length of 150 mm in a CNC lathe under constant cutting parameters: vc = 60 m/min, f = 0.05 mm/rev, and ap = 0.63 mm. The parameters employed were found as optimal (Pusavec et al., 2008) for minimizing tool-wear, minimizing cutting forces and maximizing material removal rate, and were therefore kept constant throughout the whole investigation. The major outputs investigated were residual stresses with the corresponding depth profiles in two different directions: the hoop direction corresponding to cutting velocity,   , and the axial direction corresponding to feed motion,  ⊥ . Additional outputs studied are: hardness profiles (HV) and machined surface roughness (Ra ). For adequate comparison, each experiment employed a new tool edge (carbide insert – CNMG120404, 890 grade with a MF1 chip breaking geometry offering the edge radius rn = 10 ␮m) carrying out L = 20 mm long cut. On account of using new cutting edge for each experiment and a short length of the cut, there was practically no tool-wear on cutting edges. Therefore, the tool-wear effect on the residual stresses is negligible. The bars were cleaned prior to the experiments by removing approximately 775 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 Table 1 Chemical composition of Inconel 718 AMS 5663M [wt.%]. C 0.030 Al 0.48 Mn 0.07 B 0.005 Si 0.09 Ta 0.005 P 0.007 Cu 0.04 S 0.001 Fe 17.65 Cr 18.45 Ca <0.010 Ni 53.53 Mg 0.0018 Co 0.25 Pb <0.0001 Mo 3.02 Bi 0.00001 Nb + Ta 5.32 Se <0.0001 Ti 0.95 Nb 5.31 Table 2 Mechanical and physical properties of Inconel 718 AMS 5663M. Temp. [◦ C] Ultimate tensile strength [MPa] Young’s modulus (0.2%) [MPa] Density [kg/m3 ] Melting point [◦ C] Thermal conductivity [W/mK] Room temp. 650 1594 1366 1926 1596 8190 1300 11.2 0.5 mm thickness of the top surface of each bar, in order to eliminate any surface defects and wobbling that can adversely affect the machining results. Kinematics of turning and machining directions (axial, hoop, and radial) are shown in Fig. 1. For residual stresses measurements, an X-ray diffraction technique was used, due to its accuracy and the ability to concurrently acquire the information about the changes in plasticity. In order to evaluate the residual stress distribution with the depth, the X-ray diffraction stress measurements were carried out in the sub-surface, as well as on the surface of samples. For removing small amounts of material, to measure in depth characteristics, electro-polishing material removal technique was used. For the other two measures, commercially available testing equipment was employed: a micro-indentation hardness testing system and a portable surface roughness measurement. 2.2. Workpiece material Inconel 718 is a nickel-based high-temperature alloy. The alloy used in this work is designated as AMS 5663M, which is hot rolled, solution-treated and aged. The workpiece material properties are given in Tables 1 and 2. The rough material used for this study was solution treated at 968 ◦ C for 1 h and 63 min, removed from the furnace and water quenched for 20 min. The material was then aged at sequence of 718 ◦ C for 8 h 16 min, then furnace-cooled at the rate of −38◦ /h to 621 ◦ C and held at 621 ◦ C for 8 h 13 min, then air cooled. Total aging time therefore was 16 h 17 min. Fig. 2. X-ray measuring head with the measurement setup and the corresponding angles of oscillations during the measurement procedure. ing successive layers of material up to the depth of 150 ␮m by electro-polishing, thus avoiding the reintroduction of additional residual stresses. The polishing solution was a mixture of 15% ether, 55% methanol and 30% perchloric acid. Considering the size of the workpiece in relation to the stressed layer (<150 ␮m), the difference between the redistributed stress due to layer removal and the true stress state (relaxation effect) is negligible. Therefore, no corrections were made for layer removal. 2.3. X-ray diffraction 2.4. Surface roughness and hardness measurements The methods of X-ray diffraction residual stress measurement were based on principles described in Cullity and Stock (2001) and Noyan and Cohen (1987). In this study residual stresses were measured on a portable ProtoXRD unit using manganese K␣ radiation tube ( = 2.1031 Å) at 18 kV, 4 mA to acquire the (h k l = 3 1 1) diffraction peak at a Bragg’s angle 2 = 151.88◦ , using a collimator that produces the 1 mm diameter spot size. sin2 method measurements were performed at total seven angles = ± 30◦ (ˇ angles with ProtoXRD) using a 6◦ oscillations at each ˇ angle. Additionally, for enhanced measurement accuracy, simultaneously axial measurement oscillations were performed,  = ± 10◦ (Fig. 2). Young’s modulus and Poisson’s ratio used for the (3 1 1) reflection were 196 MPa and 0.31, respectively. Stresses in the hoop () and axial (⊥) direction were evaluated using the usual assumption that because of the penetration depth being limited, the residual stress in the direction normal to the free-machined surface face is zero. Considering the accumulation of heat, which might affect residual stresses during machining as a function of axial positions, all stress measurements were carried out at a fixed machining length, Lmeasure = 3/4 Lmachining (Fig. 3). The residual stresses vs. depth profiles were measured by remov- The surface roughness measurements were performed with a non-contact, three-dimensional, interferometry profiler ZYGO 3D—New View 5000. On each of machined surface, 5 measuring points were taken three times across the surface and averaged for the sake of statistical significance of the measurements. In case of hardness measurements, hardness profiles were obtained by Vickers hardness test method, consisting of indenting the test material with a diamond indenter, in the form of a End point Start point Workpiece Cutting length, Lmachining f RS measuring point (¾ Lmachining) Fig. 3. Residual stress measuring point on the machined surface. 776 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 right pyramid with a square base and an angle of 136◦ between opposite faces subjected to a load of m = 50 g. The full load has been applied for 10 s. Hardness measurement location is in accordance with residual stresses measurement position. 3. Measurement results 3.1. Surface roughness The machined surface roughness is measured for all four CLF delivery cases. The results are shown in Fig. 4. From the results, it can be seen that CLF significantly influences the machined surface finish, when machining with the same parameters. While using MQL, better surface roughness is obtained in comparison with dry machining, due to improved lubrication effect. A decrease in surface roughness is achieved by using cryogenic machining as against conventional dry and MQL machining processes. At this point, it has to be emphasized that a large difference was noted between last two cases (cryo vs. cryo-MQL). For these two cases, the cutting forces were also followed and are analyzed in our previous work (Pusavec et al., 2008). It has been observed that the highest force Fig. 4. Machined surface roughness for different cooling/lubrication conditions (carbide cutting tool, vc = 60 m/min, f = 0.05 mm/rev, and ap = 0.63 mm). level appears in cryo CLF case. The reason for this is that in this CLF condition, the workpiece was frozen before the actual cut. This increased the hardness of the workpiece surface, thus adversely affecting cutting forces, and consequently causing higher surface roughness in magnitude as well as in variations. For improved machining performance, the LN delivery to the workpiece should be removed, using just the cutting zone deliveries. However, the Fig. 5. Residual stresses measurements on and beneath the surface (hoop and axial directions), machined under different CLF conditions (carbide cutting tool, vc = 60 m/min, f = 0.05 mm/rev, and ap = 0.63 mm). F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 777 Fig. 6. Phase space of measured residual stresses on the surface and beneath the surface of machined sample (hoop and axial), machined under different cooling lubrication conditions. results prove that cryo-lubrication CLF is providing some lubrication into the process, besides cooling mechanisms and improves the machining performance. 3.2. Residual stresses The residual stress depth profiles (  and  ⊥ ), measured on the workpiece surface, are shown in Fig. 5. It is evident that the subsurface residual stresses differ dramatically between hoop and axial directions. On the surface, the hoop direction stresses are much more tensile in comparison with axial direction (∼+500 MPa), while at the deepest point  ⊥ are about 200 MPa more compressive than in   direction. The hoop stresses are predominantly tensile on the surface, decreasing monotonically for all four CLF delivery trials with the depth, while magnitude dependent on the CLF condition. However, the amplitudes shift from tensile to compressive approximately 10 ␮m beneath the surface, going to the extreme values before stabilizing at the level corresponding to that found in the workpiece material before machining (around 0 MPa). Observing the direction of primary motion—hoop, for all the range of machining conditions investigated, the hoop residual stresses on the surface were comparable reaching around 700 MPa. In observing the compressive part beneath the surface when MQL is used instead of dry machining, slightly higher compressive values (from   = −280 MPa to   = −300 MPa) are reached. When further switching to one of the cryogenic machining cases, the profile shows a much larger compressive residual stresses (up to   = 450 MPa). The other important measure in surface integrity refers to the affected zone depth (da ) that presents the material zone that is affected by mechanical and thermal loads of machining, and has therefore changed hardness, residual stress, microstructure, etc. Same trend is observed in the stress magnitudes, in comparing different CLF conditions, for the affected depth. In case of dry machining, the affecting zone reaches da = 40 ␮m, while for other conditions, it goes deeper beneath the surface, reaching maximum for cryogenic machining at da = 70 ␮m. This presents a thicker compressive zone beneath the machined surface. The axial stresses ( ⊥ ) show significantly lower tensile stresses near the surface and larger compressive stresses beneath the surface in comparison with the residual stresses in the hoop direction. However, the correlation trend between different CLF conditions stays unchanged. The results in Fig. 5 show that machining under cryogenic machining conditions results in higher compressive stresses beneath the machined surface, as well as thicker compressive zone of material beneath the surface. These two entities in general improve the machining quality, and present an upgraded characteristic to the one mentioned at the beginning—better surface roughness of machined surface, when using cryogenic CLF. 4. Analyses and discussion 4.1. Correlation between residual stresses and plastic work While residual stresses in two directions are measured (  and  ⊥ ), the relation between these two interrelated components can be analyzed. For this a phase space representation methodology is established, show in Fig. 6. The plot specifies points along the measured depth beneath the surface, which describes the state of the 778 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 Fig. 7. Diffraction peak intensity of measured residual stresses on and beneath the machined surface for different cooling lubrication conditions. stresses. The phase states for all four CLF conditions are shown in Fig. 6. The x-axis corresponds to the hoop direction, while y-axis to the feed direction. In parallel to the axis, corresponding residual stress depth profiles are shown respectively. From these relations, it can be seen that residual stresses in the hoop direction are dispersed over a wider space, covering compressive to tensile values. In the case of  ⊥ , the stress space is more compressive than tensile. This means that for obtaining compressive values, stresses in hoop direction are more critical. By correlating both directions, it is possible to conclude that compressive values in the cutting direction (  ), yield also compressive values in the feed direction ( ⊥ ). The converse is not true. Comparing CLF conditions, it is obvious that CLF affects both stress directions simultaneously. Considering that thin layers in the radial direction of workpiece are analyzed, one dimension (thickness of those cylinders) is very small compared to the other two. In this case, the stress in the radial direction is negligible compared to the in-plane stress. Therefore, no load is applied to the element face, and the structural elements can be analyzed as two-dimensional (thin walled structures). The stress state can then be approximated by:      11 12 13   11 12 0   =   =  21 22 23   21 22 0   0 0     0 31 32 (1) 33 and hydrostatic stress: m = I1 = 1 1 1 tr() = (11 + 22 + 0) = ( + ⊥ ) 3 3 3 (2) where I1 is the first stress invariant as a trace of the stress tensor ( ij ). As it is shown in Fig. 6(a), the negative values for mainly any depth are observed. Therefore, the mean stress is compressive and the stress state can be considered compressive. Based on this, it is possible to claim that both cryogenic CLF conditions lead to higher compressive mean stresses, and therefore higher compressive stress state. This is evident from the phase space quadrants, where the IVth quadrant is the most desired one, and cryogenic machining cases are the ones closest to this. As generally known, the thermal load yields tensile stresses whereas the mechanical load suppresses it (Liu and Barash, 1976; M’Saoubi et al., 1999). The use of cryogenic cooling conditions decreases the thermal effect, and consequently reduces the tendency for tensile stressing. From the X-ray diffraction measurement procedure, beside primary measurement of residual stresses, it is also possible to analyze diffraction peak signals (Outeiro et al., 2008). Using calibrated diffraction peak width, it is possible to determine the distribution of plastic work induced by machining. The sample progression of X-ray diffraction peak width measurements, developed from the surface to the region beneath the workpiece surface that is not affected by machining, is shown in Fig. 7. A comparison is made for the extreme two CLF cases, dry and cryo-machining. On the x and y axes there are intensity and the angle theta values, while z axis corresponds to the depth beneath the surface. For the ease of comparison, the 2D projections of the peak width are shown on the lower plane. The projections are normalized peak heights that can be directly compared. In observing the signal width, right on the surface, the blunt peaks, that quickly narrow when penetrating into the material, can be seen. And, when comparing dry and cryogenic machining, the peak widths beneath the surface do not differ significantly, while on the surface they are wider for the case of cryogenic machining. This is related to higher hardness of machined surface in cryogenic machining (will be analyzed in the next section). The anisotropy observed in the hoop and axial direction highlights the dominant role that the mechanical load plays. Under normal machining conditions, the plastic deformation takes place primarily in three regions: the plastic shear between the cut and uncut chip, the compressive deformation ahead of the cutting edge, and tensile deformation behind the passing edge. This results in higher hardness of the machined surface. To prove this, the FWHM (full width at half maximum) for all the cases are measured and shown in Fig. 8. Considering FWHM, in both directions the values are in the range between 2.5◦ and 5◦ . The difference between CLF conditions is not as significant as it is present with the depth of cut. The FWHM F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 779 Fig. 8. Diffraction peak full width at half maximum (FWHM) for different cooling lubrication conditions. values rapidly increase at the surface, indicating the plastification effect caused by rubbing of the cutting tool over the machined surface. The levels of plastic deformation were estimated in Li et al. (2006) and Prevey (1987, 2000), and are used here in this analysis (Fig. 8). Peak broadening can be affected by a large number of factors such as low angle grain boundaries, very small particle or grain size, plastic deformation, compositional inhomogeneity, as well as variations in macro and micro-lattice strain. In this case, the grain size and other broadening factors besides plastic work are taken to be insignificant across the sample. Therefore, FWHM is considered to be a good indicator of plastic work (Prevey, 1987). By comparing the measured diffraction peak widths against those for uniaxially tensile deformed calibration samples (Fig. 9), it can be expected that high surface FWHM level corresponds to dominant plastic work. For quantitative evaluation, the explicit case study calibration between diffraction peak and plastic work would be needed. However, going beneath the surface, the falling value of FWHM, to that representative of the base material at the depth of 40 ␮m, reflects also the reduction in plastic work. Nevertheless, the FWHM is comparable for all the CLF conditions and so the plastic work is expected to be comparable too. The machining process affected material zone depth is same for all the CLF cases, while it differs from residual stress profiles (Fig. 5). The reason for this may be that the plastic work is more related to the mechanical loadings, while residual stresses, in addition to this are also a significant function of both thermal and mechanical effects. This shows that with extreme cooling in cryogenic machining (liquid nitrogen evaporation point −196 ◦ C), even if the plastic work is comparable between different CLF conditions, it is possible to get more compressive residual stresses. In this way, the results prove the conclusion that cryogenic machining is more effective in improving machined surface integrity. 4.2. Correlation between residual stresses and hardness Hardness and compressive residual stresses are two important criteria in material applications where a high demand is placed on wear and fatigue performance in the finished product. High surface and sub-surface hardness reduces product wear, while larger compressive residual stresses improve resistance to fatigue failure, and these two improved properties extend the life of finished products. Conventionally, pre-machining and post-machining techniques (shot peening, laser peening, roller burnishing, etc.) are generally used to improve both hardness and compressive residual stresses. However, this requires additional processes to be performed. To comprehensively cover the surface integrity characteristics, also the hardness was measured and correlated with the microstructure of the material beneath the surface. Fig. 10 shows conventionally estimated Vickers hardness profiles obtained for Inconel 718 samples, machined under different CLF conditions. The plot is made in such a way that hardness profile (left side) can be directly correlated with corresponding microstructure (right side). This is the case also for the scale (scale of microstructure figures is y-axis of hardness plot). Observing the hardness profiles, they are in direct relation to FWHM. The hardness is reaching extreme values on the surface. More specifically, the first measured point is 10 ␮m beneath the surface, due to the limitations in measurements. The hardness value reaches up to 800 HV, and decreases to the hardness values of the base material of approximately 500 HV. The decreasing trend in the sub-surface is rapid, 780 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 Fig. 9. Diffraction peak width broadening with plastic work (Li et al., 2006; Prevey, 1987, 2000): (a) broadening of (3 1 1) peak width with increasing plastic work in RR1000; (b) percent of plastic work in correlation to the FWHM for different nickel alloys. reaching the lowest value 40 ␮m beneath the machined surface. If CLF conditions are compared, again it can be seen that the highest machined surface hardness is achieved in the cryogenic machining processes. This fact can contribute to extended final product fatigue life, due to higher wear resistant surface. Furthermore, the decreasing trends are more dominant in cryogenic than in dry or MQL machining. Compared with the micro hardness, no drastic change in microstructure is evident. By comparing two cryogenic machining cases (cryo and cryoMQL), it can be seen that cryo-machining induces a higher hardness on the machined surface. From this observation, it would be preferable to move the tool holder into position for the machining pass and begin jetting the cryogenic fluid onto the workpiece before the actual cut. This “pre-cooling” step reduces the temperature of the workpiece (as well as the cutting tool), which results in increased hardness and increased compressive residual stresses in the finished product. However, this increases also the cutting forces and a compromise is required between cutting forces and the hardness of machined surface. 4.3. Correlation between residual stresses and metallographic parameters For understanding of the correlation between mechanical hardness and thermal flux properties, the microstructure modifications due to the hardening process were studied using the scanning electron microscopy (SEM), under higher magnifications (Fig. 11). The figure shows the microstructure under different CLF conditions for the bulk material far beneath the surface and on the machined surface. In all cases, it is possible to see grains with carbides at the grain boundaries. Surprisingly, the cut in all the cases is relatively clean. In machining process, there are severe deformations of the material just beneath the surface, in which case, this zone is hard to be recognized from the SEM pictures. There is a slight orientation of the grains, and borders are observed in the direction of cutting speed. However, this can be observed just up to 5–10 ␮m beneath the surface, while for machining cutting tools with really sharp cutting tool edge (rn = 10 ␮m) were used. Another effect that is present just under machined surface is the presence of smaller carbide parti- Fig. 10. Hardness measurements beneath the machined surface (profile), machined under different cooling lubrication conditions and their correlation with the metallographic structure of the material beneath the surface. F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 781 Fig. 11. Metallographic structure (SEM) of the material on the machined surface and far beneath the surface (not affected by machining). cles. Most probably, the reason for this is in its fractures under high deformation work. In their nature, these carbide flakes are thin and long. Therefore, they hardly resist undergoing deformation work. These size reductions of carbides are observable at about 30 ␮m depths in the sub-surface. Beyond this depth, no evident difference can be observed. Comparing different machining CLF conditions, it is hard to recognize any difference between microstructure of the machined sub-surface in this study. Based on that, it is possible to conclude that if using one of cryogenic machining alternatives instead of conventional machining, process can improve the surface integrity of machined surface through lower surface roughness, more compressive residual stresses just beneath the surface (down to 75 ␮m from the surface), and with higher hardness of the surface. It has to be noted, that even though the hardness has been measured from 10 to 120 mm beneath the surface, based on the trend, the hardness on the surface can be predicted. However, even when better integrity characteristics are achieved, it is not on the account of the microstructure of the machined part beneath the surface. Due to the wider results of previous work that showed that the cooling liq- 782 F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 Fig. 12. SEM microstructure, with corresponding chemical composition and EDS spectra. Fig. 13. SEM micrographs of specimens and X-ray mapping of carbide particle in metal matrix. uids may influence the microstructure, a more detailed study would be necessary to see the actual effect of material transformation resulting from the CLF conditions. The role of carbides in superalloys is often questioned. However, in many cases it has been noted that carbides in microstructure of nickel-based alloys have to be tolerated due to their influence on the stability of the matrix and ductility of a material (Cesnik et al., 2008). Therefore, the carbides were analyzed through SEM/EDS in an effort to find the composition of those particles (Fig. 12). Carbides, such as geometrically closely packed phases, are dispersed in austenitic matrix. Checking the atomic and weight percentages and the spectrum, it can be seen that Nb (21 wt.%) and Mo (2 wt.%) carbides are recognized in Ni matrix. The microstructures shown in Fig. 11 were chosen for the precipitate identification, because it contains typical particles observed from SEM. Beside SEM, corresponding EDS spectrum of the particles is shown, from which it can be preliminarily concluded that the precipitate contains Nb and Mo, which are carbide formers, but contained also appreciable amounts of Cr, Fe and Ni in addition to Ti. The presence of Cr, Fe and Ni in the EDS spectrum may be associated with matrix elements, or could be attributed to the nature of the carbides. Further, the confirmation of the carbides was needed, and Xray mapping was carried out for carbide particle in metal matrix, to determine the chemical composition. This mapping is shown in Fig. 13. In the analysis of the microstructure of Inconel 718, beside niobium and molybdenum carbides (Nb, Mo, and C), some elements of titanium carbides (TiC) are also present, while the rest of the elements are confirmed to be a part of the matrix. However most dominant Nb-rich carbides were observed, and are of irregular shape. Going back on the influence of the machining process on the microstructure, in this study, it was hard to distinguish between different CLF conditions. In all cases, the Nb carbides are shorter just beneath the surface in comparison to the microstructure far beneath the surface. This proves that plastic work and high deformations of the material, due to the nature of machining process, is present and is breaking these carbides down. Comparing CLF conditions, the only evident difference recognized in the material deformation zone is the direction of cut-hoop (cutting speed direction). In case of cryo-machining, the deformations are hardly observable, while reaching the depth of 1–2 ␮m beneath the surface. In contrast, in dry and MQL machining, these deformations are evident down to 5–10 ␮m beneath the surface, through reorientated carbides. 5. Conclusions The effects of CLF conditions, when using cryogenic machining, on surface integrity characteristics of the turned Inconel 718 nickel-based alloy have been investigated. Surface roughness, hardness depth profile and residual stress depth profile were obtained. For the measurement of residual stresses, X-ray diffraction method is used. For deeper understanding, the diffraction peak width and SEM microstructure were also investigated with the aim of evalu- F. Pusavec et al. / Journal of Materials Processing Technology 211 (2011) 773–783 ating plastic deformation and the work hardening of the machined surface. The following conclusions can be drawn: – Comparing CLF conditions, cryogenic machining presents sustainable machining alternative, and provides lower surface roughness in machining under the same machining parameters (f, ap , and Vc ). – The compressive zone beneath the surface is thicker for the case of cryogenic CLF conditions, extending the compressive zone from da = 40 ␮m to da = 70 ␮m (for 185%). – The FWHM has been analyzed and correlated with the measured hardness depth profile. Good correlations are seen with the increase in hardness on the surface from 500 HV to 800 HV. – The work hardened sub-surface layer was found to have a depth of about 40 ␮m, while a decreasing trend for this value is observed in the case of cryogenic machining. This means that in case of cryogenic machining a higher hardness on the surface is achieved, and this hardened layer is thinner than in dry or MQL machining. – Cryogenic machining slightly influences the final product microstructure (smaller grain size), and induces less plastic deformation on the machined surface. Plastic deformations on the machined surface, in the direction of cutting speed, can be recognized just up to 1–2 ␮m beneath the surface in cryogenic machining, while this plastic deformation zone is thicker and reaches up to 5–10 ␮m in dry and MQL machining. The results from this work show that cryogenic machining increases hardness and compressive residual stress, and reduces surface roughness in the workpiece/part. Hence, it offers more than just a clean process. It offers a process that improves the final part performances through enhanced machined surface integrity, emphasizing the benefits and the need for implementing cryogenic machining in industrial applications. References Arunachalam, R.M., Mannan, M.A., Spowage, A.C., 2004a. Residual stress and surface roughness when facing age hardened Inconel 718 with CBN and ceramic cutting tools. International Journal of Machine Tools & Manufacture 44 (9), 879–887. Arunachalam, R.M., Mannan, M.A., Spowage, A.C., 2004b. Surface integrity when machining age hardened Inconel 718 with coated carbide cutting tools. International Journal of Machine Tools & Manufacture 44, 1481–1491. Axinte, D.A., Andrews, P., Li, W., Gindz, N., Withers, P.J., 2006. Turning of advanced Ni based alloys obtained via powder metallurgy route. CIRP Annals - Manufacturing Technology 55 (1), 117–120. Cesnik, D., Bratus, V., Kosec, B., Rozman, J., Bizjak, M., 2008. Heat treatment and fineblanking Inconel 718. RMZ – Materials and Geoenvironment 55 (2), 163–172. Cullity, B.D., Stock, S.R., 2001. Elements oy X-ray Diffraction. Prentice-Hall, Inc., New York. Fredj, N.B., Sidhom, H., Braham, C., 2006. Ground surface improvement of the austenitic stainless steel AISI 304 using cryogenic cooling. Surface and Coatings Technology 200 (16–17), 4846–4860. 783 Henriksen, E.K., 1951. Residual stresses in machining surfaces. Journal of Applied Mechanics. 73, 69–76. Hua, J., Shivpuri, R., Cheng, X., Bedekar, V., Matsumoto, Y., Hashimoto, F., Watkins, T.R., 2005. Effect of feed rate, workpiece hardness and cutting edge on subsurface residual stress in the hard turning of bearing steel using chamfer + hone cutting edge geometry. Materials Science & Engineering 394, 238–248. Jang, D.Y., Watkins, T.R., Kozaczek, K.J., Hubbard, C.R., 1996. Surface residual stress in machined austenitic stainless steel. Wear 194, 168–173. Li, W., Preuss, M., Withers, P.J., Axinte, D., Andrews, P., 2006. Characterization of residual stresses in machined surface of a high strength nickel-based superalloy. Material Science Forum 524–525, 587–592. Li, W., Withers, P.J., Axinte, D., Preuss, M., Andrews, P., 2009. Residual stresses in face finish turning of high strength nickel-based superalloy. Journal of Materials Processing Technology 209, 4896–4902. Liang, S.Y., Su, J.C., 2007. Residual stress modeling in orthogonal machining. CIRP Annals – Manufacturing Technology 56 (1), 65–68. Liu, C.R., Barash, M.M., 1976. The mechanical state of the sublayer of a surface generated by chip removal process. Part 1. Cutting with a Sharp Tool. Transaction of the ASME, 1192–1201. M’Saoubi, R., Outeiro, J.C., Changeux, B., Lebrun, J.L., Dias, A.M., 1999. Residual stress analysis in orthogonal machining of standard and resulfurized AISI 316L Steel. Journal of Materials Processing Technology 96 (1–3), 225–233. Noyan, I.C., Cohen, J.B., 1987. Residual Stress—Measurement by Diffraction and Interpretation. Springer-Verlag, New York. Ordas, N., Penalva, M.L., Fernandez, J., Garcia-Rosales, C., 2003. Residual stresses in tool steel due to hard turning. Journal of Applied Crystallography 36, 1135–1143. Outeiro, J.C., Dias, A.M., Lebrun, J.L., 2004. Experimental assessment of temperature distribution in three dimensional cutting process. Machining Science and Technology 8 (3), 357–376. Outeiro, J.C., Dias, A.M., Lebrun, J.L., Astakhov, V.P., 2002. Machining residual stresses in AISI 316L steel and their correlation with the cutting parameters. Machining Science and Technology 6 (2), 251–270. Outeiro, J.C., Pina, J.C., M’Saoubi, R., Pusavec, F., Jawahir, I.S., 2008. Analysis of residual stresses induced by dry turning of difficult-to-machine materials. CIRP Annals – Manufacturing Technology 57 (1), 77–80. Ozel, T., Zeren, E., 2006. Finite element modeling the influence of edge roundness on the stress and temperature fields induced by high-speed machining. International Journal of Advanced Manufacturing Technology 35 (3–4), 255–267. Paul, S., Chattopadhyay, A.B., 1995. A study of effects of cryo-cooling in grinding. International Journal of Machine Tools & Manufacture 35 (1), 109–117. Prevey, P.S., 1987. The measurment of subsurface residual stress and cold work distributions in nickel base alloys. In: Proceedings of ASM’s Conference on Residual Stress in Design, Process and Materials Selection , OH, US, 11–19. Prevey, P.S., 2000. The effect of cold work in the thermal stability of residual compression in surface enhanced IN718. In: Proceedings of the 2nd ASM Materials Solutions Conference & Exposition , MO, USA. Pusavec, F., Deshpande, A., M’Saoubi, R., Kopac, J., Dillon Jr., O.W., Jawahir, I.S., 2008. The modeling and optimization of machining of high temperature nickel alloy for improved machining performance and enhanced sustainability. In: Proceedings of the 11th CIRP conference on Modeling of Machining Operations , Gaitherburg, USA, 21–28. Pusavec, F., Krajnik, P., Kopac, J., 2010. Transition to sustainable production. Part I. Application on machining technologies. Journal of Cleaner Production 18 (2), 174–184. Sasahar, H., Obikawa, T., Shirakashi, T., 2004. Prediction model of surface residual stress within a machined surface by combining two orthogonal plane models. International Journal of Machine Tools & Manufacture 44, 815–822. Sharman, A.R.C., Hughes, J.I., Ridgeway, K., 2006. An analysis of the residual stresses generated in Inconel 718 when turning. Journal of Materials Processing Technology 173, 359–367. Wardany, T.I.E.l., Kishawy, H.A., Elbestawi, M.A., 2000. Surface integrity of die material in high speed machining. Part 2. Microhardness variations and residual stress. Journal of Manufacturing Science and Engineering – ASME 122, 632–641.