- Email: [email protected]
The basic relationships between residual stress, white layer, and fatigue life of hard turned and ground surfaces in rolling contact
CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
Contents lists available at ScienceDirect
CIRP Journal of Manufacturing Science and Technology journal homepage: www.elsevier.com/locate/cirpj
The basic relationships between residual stress, white layer, and fatigue life of hard turned and ground surfaces in rolling contact Y.B. Guo a,*, A.W. Warren a, F. Hashimoto b a b
Dept. of Mechanical Engineering, The University of Alabama, 7th Ave., Tuscaloosa, AL 35487, USA The Timken Company, Canton, OH 44706, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Available online 18 January 2010
The characteristics of residual stress (RS) profiles and their effects on rolling contact fatigue life for precision turned and ground surfaces with a white layer (WL) are very controversial. This study has shown that: (a) Hard turning with a fresh tool generates a ‘‘hook’’ shaped RS profile characterized by surface compressive RS and subsurface maximum compressive RS. While gentle grinding only generates maximum compressive RS at the surface and a shallow subsurface zone of compressive RS. The ‘‘hook’’ shaped RS profile with surface compressive RS contributes to a longer (40%) fatigue life of a machined surface. (b) A turned WL surface by a worn tool generates a high tensile stress in the area of the WL, but becomes more highly compressive in the deeper subsurface than the turned one without a WL. The high tensile RS at the surface leads to a much shorter (7.6 times) fatigue life. (c) A ground WL only shifts the RS to tension but hardly affect the basic shape of the profile for a ground fresh surface. The surface tensile RS coupled with near surface peak tensile RS produces the shortest (7.8 times) fatigue life. ß 2010 CIRP.
Keywords: Residual stress White layer Surface integrity Fatigue life Hard turning Grinding
1. Introduction Hard turning and grinding are competitive finishing processes for the production of bearings, gears, cams, etc. The selection of hard turning or grinding is case by case to fit different production scenarios. Process-induced residual stress and its effect on component performance such as fatigue life are key criteria for process selection and optimization. The most significant differences in the characteristics of residual stress profiles [1–6] by ‘‘gentle’’ hard turning and grinding are manifested in two aspects: (i) hard turning with a sharp cutting edge geometry (honed or chamfered) generates a ‘‘hook’’ shaped residual stress profile characterized by compressive residual stress at the surface and maximum compressive residual stress in the subsurface. While gentle grinding only generates maximum compressive residual stress at the surface. However, Toenshoff et al. [3] reported that surface residual stresses generated by the sharp tool can also be tensile, although the magnitude was much lower than that created with a worn tool. (ii) The depth of compressive residual stress in the subsurface by hard turning is much larger than that by grinding [7]. But the magnitude of compressive residual stress at a ground surface is usually higher than that at a turned surface. Hard turning with a worn tool caused a phase transformed ‘‘white layer’’ at the surface. The white layer generated tensile
* Corresponding author. Tel.: +1 205 348 2615; fax: +1 205 348 6419. E-mail address: [email protected] (Y.B. Guo). 1755-5817/$ – see front matter ß 2010 CIRP. doi:10.1016/j.cirpj.2009.12.002
surface stress [2], but shifted to compression as the depth increased and ultimately attained larger magnitude and affected depth than the residual stress generated by a sharp tool. Similar results [3,4] about the effect of white layer on residual stress were obtained. The unique pattern of a residual stress profile has been identified as a critical factor of surface integrity for component performance. A compressive residual stress induced by hard turning and grinding was found to improve rolling contact fatigue (RCF) life [5,12,13]. Furthermore, deep compressive residual stresses in the subsurface may be more beneficial to bearing fatigue life than shallower stresses of greater magnitude. However, a white layer by hard turning may reduce RCF life as much as six times [12]. Klocke and Schlattmeier [8] reported that a slight grinding burn does not necessarily lead to an early gear fatigue. Due to the very controversial impact of white layer on fatigue life, a fundamental understanding on the nature of residual stresses associated with different surface types and their effects on fatigue life is essential. However, the literature review reveals several critical unsolved issues. First, the bulk residual stress data is limited to turned and ground surfaces free of a white layer. The nature of residual stress induced by a ground white layer is very ambiguous and lacking. Second, how residual stress resulted from heat treatment affects the final distribution of residual stress is not yet known. Third, how residual stress associated with a turned or ground white layer affects fatigue life has not been clarified yet. Therefore, a linkage between various patterns of residual stress profiles and RCF life has been poorly understood. The objective of
130
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
Fig. 1. Microstructures of the cross-sectional surfaces: (a) hard turned fresh (HTF), (b) ground fresh (GF), (c) hard turned w. white layer (HTWL), and (d) ground w. white layer (GWL).
this paper is to solve these pressing issues. This has been accomplished by determining the residual stress profiles of turned and ground surfaces with and without a white layer and their effects on fatigue life. 2. Machining tests Four types of unique surfaces were prepared: hard turned fresh (HTF) with compressive RS at the surface and maximum compressive RS in the subsurface, hard turned white layer (HTWL) with high tensile RS at the surface and maximum compressive RS in the deep subsurface, ground fresh (GF) with compressive RS at the surface, and ground white layer (GWL) with tensile RS at the surface. In addition, an as heat treated surface was prepared as a base line surface. Dry face turning AISI 52100 steel (62 HRc) discs of 76.2 mm diameter and 19.1 mm thickness were conducted on a rigid, high precision lathe with a BZN 8100 round tool (70% CBN content by volume in a TiN binder phase) with 0.15 mm/158 chamfer & 6.35 mm corner radius. The cutting velocity, feed, depth of cut are 1.78 m/s, 0.051 mm/rev, and 0.254 mm for the preparation of HTF surfaces using a sharp tool, while a worn tool (VB 0.5 mm) at speed of 2.82 m/s was used for the HTWL surfaces. High speed cutting using a worn cutting tool with large tool flank wear will produce high cutting temperatures which induce a thick phase transformed white layer with high tensile residual stress. Grinding was conducted at a surface grinder with an Al2O3 20A60H fresh grinding wheel. The GF surfaces were prepared at wheel speed 23.9 m/s, table speed 15.2 m/min, cross-feed 5.14 mm/pass, and down feed 12.7 mm/pass with water soluble coolant. The GWL surfaces were prepared at same machining condition except the down feed 26.0 mm/pass and without coolant. The measured surface roughness across feed marks of the turned and ground surfaces is Ra 0.16 mm. The 3D surface topography of the machined surfaces also has equivalent root
mean square of 0.2 mm. However, the amplitude parameter St (Rt counterpart in 3D surface topography), i.e., the sum of the maximum height of summits and the maximum depth of valleys, was 1.63 mm, 1.49 mm, 1.29 mm, and 2.11 mm, respectively for the HTF, HTWL, GF, and GWL surfaces. Microstructures of the four types of surfaces are shown in Fig. 1. It can be seen that the white layer thickness for the HTWL and GWL surfaces is about 7 mm. 3. Residual stress characterization 3.1. Residual stress measurement Residual stress profiles in Figs. 2 and 3 were measured by the Xray diffraction method coupled with layer removal by etching. Residual stress measurements were carried out at room temperature using a 4-axis Bruker-AXS X-ray machine with an area
Fig. 2. Residual stress profiles in feed direction.
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
131
Fig. 3. Residual stress profiles in cutting direction.
Fig. 5. Measurement repeatability of residual stress for HTWL surfaces.
detector. Using a cobalt source (l = 1.7889 A˚), the three-dimensional scans involved 29 scans between 908 2u 1078 for 9 tilt angles (C = 08, 108, 208, 308, and 408) at each of three rotation angles (f = 08, 458, and 908). The X-ray was generated using 35 kV and 20 mA and passed through a 0.8 mm collimator. Each scan was 15 s which was determined to be adequate for good peak intensity and definition. The gravity peak location method with 10% threshold value was used to calculate (2 1 1) peak positions using DIFFRACplus stress software. The change in lattice spacing at different tilt angles resulted in a corresponding peak shift. The residual strain was derived from the slope of both the linear and elliptical fit between the fractional change of the plane spacing (strain) and sin2 c. The Young’s modulus (E) and Poisson’s ratio (y) for AISI 52100 steel were E = 201 GPa and y = 0.27, respectively. For repeatability, measurements at three different locations at same subsurface depth for each sample were made and the results were averaged to give the residual stress profiles. The average and variance of residual stresses in Figs. 4–6 shows a good measurement repeatability (<10%) although the variation is expected due to several sources of measurement uncertainty. The error bars stand for the min/max values from the average value.
surface types, the affected depth of residual stress is rather shallow (<20 mm). In the cutting (or grinding) direction, the maximum compressive residual stress occurs at the surface for both hard turning and grinding. Considering the residual stress at the turned surface has a certain variation in Fig. 4, the maximum compressive residual stress has a great chance to be in the subsurface. It implies that a ‘‘hook’’ shaped residual stress profile may be still produced in the cutting direction. However, the magnitude of residual stress generated by hard turning is more than 3 times larger than that for grinding and the affected depth is nearly twice as deep. It should be mentioned that the magnitude of residual stress may change significantly for different machining conditions, but the patterns of residual stress profiles are very stable as far as a sharp tool or grinding wheel is used.
3.2. HTF vs. GF surfaces Several differences between gentle hard turning and grinding are observed for the residual stress profiles. In the feed direction, the maximum compressive residual stress occurs at the surface for the case of grinding while maximum compressive residual stress occurs at a shallow depth (5 mm) for hard turning. Also, the magnitude and affected depth of residual stress is greater for hard turning. Due to the gentle machining conditions used for both
Fig. 4. Measurement repeatability of residual stress for HTF and GF surfaces.
3.3. HTWL vs. GWL surfaces The residual stress profiles of the HTWL and GWL surfaces show significant differences. In the feed direction, both surfaces generated a tensile surface residual stress. Maximum tensile residual stress was located at the surface for the hard turning case, while maximum tensile residual stress occurred in the subsurface (5 mm) for grinding. While the GWL surface shows tensile stress throughout the entire measured depth (90 mm), the HTWL surface becomes compressive beyond 12 mm. At the depth of 40 mm the compressive residual stress becomes a maximum (826 MPa). The residual stress remains compressive for the HTWL surface to a depth of at least 110 mm. The affected depth of residual stress measured for HTWL surface is much greater than the GWL surface. In the cutting direction, similar trends are observed. The GWL
Fig. 6. Measurement repeatability of residual stress for GWL surfaces.
132
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
surface shows a slightly compressive (46 MPa) residual stress, while the HTWL surface has a tensile (102 MPa) residual stress. At the depth of 5 mm, both HTWL and GWL surfaces have the largest magnitude of tensile residual stress with the HTWL surface having the largest magnitude. Beyond this depth, residual stress of the GWL surface decreases slowly, becoming stable at about 20 mm. However, residual stress of the HTWL surface becomes compressive around 15 mm and attains a maximum compressive residual stress of 834 MPa at 30 mm. Beyond this depth the residual stress slowly decreases less than 200 MPa at 110 mm.
Table 1 RCF fatigue test conditions. Load
Peak Hertzian stress
Radius of contact circle
Load frequency
306 N
4629 MPa
26 mm
173 Hz
3.4. HTF vs. HTWL surfaces The HTWL surface generated a tensile surface residual stress while HTF generated a large compressive one. In addition, turning with a fresh tool imparts maximum compressive residual stresses of 638 MPa at depth of 5 mm in feed direction and 986 MPa at the surface in cutting direction but only affects residual stress in a shallow subsurface (20 mm). On the other hand, the HTWL surface showed much deeper (>100 mm) affected depth. The maximum compressive residual stress 826 MPa for the HTWL surface occurred at a depth of 40 mm in feed direction and 30 mm in cutting direction. The peak tensile residual stress at 5 mm in the subsurface for the HTWL surface is a result of the martensitic transformation which occurred in the near surface due to high cutting temperatures results from abusive turning conditions. 3.5. GF vs. GWL surfaces The difference between the residual stress profiles generated by gentle and abusive grinding is less dramatic than those for the hard turning cases. However, similar trends with a shift from surface compressive to tensile residual stress are observed. Gentle grinding resulted in a compressive surface residual stress that quickly attenuates at a depth of only 20 mm, while abusive grinding generated tensile residual stress throughout the measured depth of 90 mm. The magnitude and affected depth of the tensile residual stress is much greater for the GWL surface. 3.6. Machined vs. as heat treated surfaces By comparing the magnitudes and patterns of the residual stresses profiles from the as heat treatment surfaces and with those after machining, it is clear that machining itself is the deterministic factor for the resulting residual stress magnitudes and profiles, while the influence of heat treatment is minor. 4. Rolling contact fatigue (RCF) life tests
Fig. 7. RCF test setup.
the slave washer/spindle corresponds to half revolution of each ball in the tests, it means one revolution of spindle has 4 ball revolutions (0.5 8 balls) on the test surface. The load frequency was calculated as follows: 0:5 N n f ¼ 60 where the unit of spindle speed N is rpm, n is the number of rolling balls. The test samples are secured in a resting plate which is positioned with four rods that maintain the rigs position and rigidity. The resting plate is free to move up and down as it rests on a load cell, allowing accurate measure of the applied force. An acoustic emission sensor is mounted directly to the test sample. Prior to test startup, a thin layer of multi-purpose lithium complex grease was evenly distributed across the test surface, slave washer track, and ball bearings. Further lubrication was added at regular intervals throughout the duration of the test. Rotational speed of the spindle and washer was monitored using an optical tachometer. Parallelism and centricity between the slave washer and test surface is critical to ensure that the pressure distribution is uniform across the test surface. For this reason, a dial indicator gage was used to measure the vertical and horizontal runout of the washer and test surface. Using this method it was possible to adjust the horizontal and vertical displacements so that parallelism and centricity are 10 mm.
4.1. Fatigue test setup The experimental setup shown in Fig. 7 was used to monitor RCF tests of the ground surfaces. The test rig uses a Bridgeport milling machine to supplement as the drive motor and the shaft runs at a speed of 2600 rpm. The load is applied through the rotating shaft which also houses the slave washer. The test utilized eight chrome steel ball bearings with a diameter of 5.56 mm and a hardness of 63 HRc. A nylon bearing cage was used to hold the ball bearings and minimize noise during the tests. Slave washers of the same material as the test samples were created by turning. An 8 mm radius groove was machined on each slave washer to maintain the ball’s position during operation and provide a lower contact pressure between the ball and slave washer. The lower contact pressure in the slave washer would help ensure that fatigue failure would first occur in the test samples. The specific testing conditions are shown in Table 1. Because one revolution of
Fig. 8. Fatigue life chart for different surface types.
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
4.2. Acoustic emission (AE) monitoring of fatigue damage
6. Correlation between residual stress and fatigue life
In-situ monitoring of fatigue damage was accomplished using an AE signal acquisition and processing package AEWin [14]. An AE sensor was attached to the test sample with a holding fixture, while vacuum grease served as the coupling media between the sensor and test sample. The AE sensor has a 125 kHz resonant frequency and connects to an 18 bit PCI-2 data acquisition board that was incorporated into a PC. Before the data reached the PC it was passed through a preamplifier that was set at a 40 dB gain. A threshold of 45 dB was used. The sampled AE signal parameters in this study include: absolute energy, RMS, amplitude, counts, and average frequency. The occurrence of surface fatigue is determined by the shoot up of AE amplitude signal, which has been shown a reliable signal for monitoring surface fatigue [9]. The test was real-time monitored, in which the time, load, speed, and any observations were recorded along with application of lubricant. Also occurring at these intervals was the application of additional lithium complex grease applied manually through a spray. The monitoring continued until increased AE signals were detected due to surface pitting and spalling, at which time the test was ended and carefully disassembled to avoid any damage to the samples.
6.1. Fatigue life of HTF vs. GF surfaces
5. Fatigue life chart and damage mechanism Surface fatigue life refers to the number of stress cycles required to initiate spalling on the test surfaces. Results of the surface fatigue lives with variance are shown in Fig. 8. Three fatigue tests for each surface type were conducted for data repeatability. The HTF surfaces had an about 40% higher average fatigue life than the GF ones. The reason for the fatigue difference is due to several factors such as distinct patterns of the residual stress profiles. The surface fatigue life comparison chart for the gentle and abusive turned and ground surfaces in Fig. 8 shows that a white layer on the HTWL and GWL surfaces is very detrimental to fatigue life. In the case of hard turning, the HTF surfaces had a 760% longer average fatigue life than the HTWL surfaces. Similarly, the GF surfaces had a 780% longer fatigue life than the GWL ones. However, the HTWL surfaces had a 65% longer fatigue life than the GWL ones. Spalling is the predominant damage mechanism for the tested surfaces. A typical example of the fatigue damage is shown in Fig. 9. Since the fatigue tests were conducted at well lubricated conditions with a high-temperature multi-purpose lithium complex grease. Scuffing did not happen during the tests.
133
The HTF surfaces had an about 40% higher average fatigue life than the GF ones. The reason for the fatigue difference is due to several factors such as distinct patterns of the residual stress profiles. By comparing the residual stress profiles and fatigue life, it is clear that the life difference is concurrent with the HTF surfaces with large compressive residual stress at the surface and the maximum compressive residual stress in the subsurface. If fatigue cracks initiate from stress concentration due to surface asperities, the large surface compressive residual stress may impede crack initiation at the surface. Fatigue cracks may also be driven by the subsurface maximum shear stress experienced by the test surfaces in rolling contact. In this case, the maximum compressive residual stress in the subsurface may serve to delay the initiation of subsurface cracks. Because both surfaces have equivalent surface roughness (Ra 0.16 mm), the fatigue life is not likely to be significantly affected by surface finish. However, an important aspect of the surfaces is that the measured skewness of the GF surfaces (0.127) is less than that of the HTF ones (0.051). Negative skewness is an index to the surfaces ability to trap lubricant, as it signifies the presence of more valleys on the surface. For small or medium loads below shakedown limit, the valleys may act as reservoirs to trap lubricant and may increase fatigue life by decreasing friction between the balls and sample surface as they are in contact with one another during rolling. However, for large loads (above the shakedown limit) such as the 4.6 GPa peak Hertzian stress used in these tests, the valleys may act as notches and can promote crack growth due to the wedge effect. 6.2. Fatigue life of HTWL vs. GWL surfaces Compared with residual stresses of the HTF surfaces, HTWL surfaces have two significant changes. Surface residual stress shifts from high compression to high tension. The subsurface maximum compressive residual stress becomes more compressive and deeper in both cutting and feed directions. Obviously, the surface tensile residual stress leads to much earlier fatigue damage than the surface compressive residual stress. However, the increased subsurface maximum compressive residual stress seems not to effectively delay fatigue cracks initiation and propagation. Thus, the compound effects of deep compressive residual stresses in the subsurface and tensile residual stress at the surface may not be more beneficial to fatigue life than shallower compressive stress of
Fig. 9. Typical fatigue damage of the tested surfaces. (a) Surface spalling. (b) Subsurface fatigue cracks.
134
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
greater magnitude. It implies that fatigue cracks tend to initiate from surface rather than in the subsurface in rolling contact. It should be pointed out that, beside the compressive residual stress in the subsurface, the significant difference in hardness between the surface white layer and the beneath tempered zone causes a minor ability of the workpiece surface to compensate distortion due to external loads. This also enhances crack initiation and propagation in the subsurface as well as spalling in Fig. 9. For the GWL surfaces, the bulk residual stress is in tension and the peak tensile residual stress is in the very near surface (6 mm). Fatigue cracks induced by surface tensile residual stress are further facilitated by the peak tensile residual stress in the near subsurface. This may explains why the GWL surfaces have the shortest average fatigue life. In addition to residual stress, surface hardness of the machined surfaces in Fig. 1 could be another factor to contribute the difference in fatigue life since the GF surfaces are slightly harder than the HTF surfaces [10]. However, the effect of hardness alone cannot be separated from the compound effects of other surface integrity factors. The high hardness of white layer [3,11,12] may also contribute the shorter fatigue life of the HTWL and GWL surfaces. But the hardness effect alone is very difficult to isolate from that of residual stress. 7. Conclusions The key findings on the effects of machining-induced residual stress and white layer on rolling contact fatigue may be summarized as follow: The basic residual stress profiles by a sharp tool can be fundamentally changed by a turned white layer but not a ground one; The high tensile residual stress and hardness gradient associated with the white layer are due to high cutting temperatures induced by the large tool flank wear; The hook shaped residual stress profile of a turned surface may have about 40% more fatigue life than a ground one;
The machining-induced white layer can reduce fatigue life as much as 8 times. The turned white layer surface may have longer life than a ground white layer surface. Acknowledgment This research has been supported by the National Science Foundation under CAREER Grant No. CMMI-0447452. References [1] Matsumoto, Y., Barash, M.M., Liu, C.R., 1986, Effect of Hardness on the Surface Integrity of AISI 4340 Steel, Journal of Engineering for Industry, 180:169–175. [2] Konig, W., Klinger, M., Link, R., 1990, Machining Hard Materials With Geometrically Defined Cutting Edges-field of Applications and Limitations, Annals CIRP, 39:61–64. [3] Toenshoff, H.K., Wobker, H.G., Brandt, D., 1995, Hard Turning-influences on the Workpiece Properties, Transactions NAMRI/SME, 23:215–220. [4] Abrao, A.M., Aspinwall, D.K., 1996, The Surface Integrity of Turned and Ground Hardened Bearing Steel, Wear, 196:279–284. [5] Matsumoto, Y., Hashimoto, F., Lahoti, G., 1999, Surface Integrity Generated by Precision Hard Turning, Annals CIRP, 48/1:59–62. [6] Klocke, F., Brinksmeier, E., Weinert, K., 2005, Capability Profiles of Hard Cutting and Grinding Processes, Annals CIRP, 54/2:557–580. [7] Konig, W., Berktold, A., Koch, K.F., 1993, Turning Versus Grinding—A Comparison of Surface Integrity Aspects and Attainable Accuracies, Annals CIRP, 42:39–43. [8] Klocke, F., Schlattmeier, H., 2004, Surface Damage Caused by Gear Profile Grinding and its Effects on Flank Load Carrying Capacity, Gear Technology, 9/ 10:44–53. [9] Guo, Y.B., Schwach, D.W., 2005, An Experimental Investigation of White Layer on Rolling Contact Fatigue Using Acoustic Emission Technique, International Journal of Fatigue, 27:1051–1061. [10] Warren, A.W., Guo, Y.B., 2006, On the Clarification of Surface Hardening Mechanisms by Hard Turning and Grinding, Transactions NAMRI/SME, 34:309–316. [11] Guo, Y.B., Sahni, J., 2004, A Comparative Study of the White Layer by Hard Turning Versus Grinding, International Journal of Machine Tools and Manufacture, 44:135–145. [12] Schwach, D.W., Guo, Y.B., 2006, A Fundamental Study on the Impact of Surface Integrity by Hard Turning on Rolling Contact Fatigue, International Journal of Fatigue, 28:1838–1844. [13] Hashimoto, F., Guo, Y.B., Warren, A.W., 2006, Surface Integrity Difference Between Hard Turned and Ground Surfaces and its Impact on Fatigue Life, Annals CIRP, 55/1:81–84. [14] Physical Acoustics Corporation. 2003, PCI-2 Based ae System User’s Manual, ver. 1, .
Contents lists available at ScienceDirect
CIRP Journal of Manufacturing Science and Technology journal homepage: www.elsevier.com/locate/cirpj
The basic relationships between residual stress, white layer, and fatigue life of hard turned and ground surfaces in rolling contact Y.B. Guo a,*, A.W. Warren a, F. Hashimoto b a b
Dept. of Mechanical Engineering, The University of Alabama, 7th Ave., Tuscaloosa, AL 35487, USA The Timken Company, Canton, OH 44706, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Available online 18 January 2010
The characteristics of residual stress (RS) profiles and their effects on rolling contact fatigue life for precision turned and ground surfaces with a white layer (WL) are very controversial. This study has shown that: (a) Hard turning with a fresh tool generates a ‘‘hook’’ shaped RS profile characterized by surface compressive RS and subsurface maximum compressive RS. While gentle grinding only generates maximum compressive RS at the surface and a shallow subsurface zone of compressive RS. The ‘‘hook’’ shaped RS profile with surface compressive RS contributes to a longer (40%) fatigue life of a machined surface. (b) A turned WL surface by a worn tool generates a high tensile stress in the area of the WL, but becomes more highly compressive in the deeper subsurface than the turned one without a WL. The high tensile RS at the surface leads to a much shorter (7.6 times) fatigue life. (c) A ground WL only shifts the RS to tension but hardly affect the basic shape of the profile for a ground fresh surface. The surface tensile RS coupled with near surface peak tensile RS produces the shortest (7.8 times) fatigue life. ß 2010 CIRP.
Keywords: Residual stress White layer Surface integrity Fatigue life Hard turning Grinding
1. Introduction Hard turning and grinding are competitive finishing processes for the production of bearings, gears, cams, etc. The selection of hard turning or grinding is case by case to fit different production scenarios. Process-induced residual stress and its effect on component performance such as fatigue life are key criteria for process selection and optimization. The most significant differences in the characteristics of residual stress profiles [1–6] by ‘‘gentle’’ hard turning and grinding are manifested in two aspects: (i) hard turning with a sharp cutting edge geometry (honed or chamfered) generates a ‘‘hook’’ shaped residual stress profile characterized by compressive residual stress at the surface and maximum compressive residual stress in the subsurface. While gentle grinding only generates maximum compressive residual stress at the surface. However, Toenshoff et al. [3] reported that surface residual stresses generated by the sharp tool can also be tensile, although the magnitude was much lower than that created with a worn tool. (ii) The depth of compressive residual stress in the subsurface by hard turning is much larger than that by grinding [7]. But the magnitude of compressive residual stress at a ground surface is usually higher than that at a turned surface. Hard turning with a worn tool caused a phase transformed ‘‘white layer’’ at the surface. The white layer generated tensile
* Corresponding author. Tel.: +1 205 348 2615; fax: +1 205 348 6419. E-mail address: [email protected] (Y.B. Guo). 1755-5817/$ – see front matter ß 2010 CIRP. doi:10.1016/j.cirpj.2009.12.002
surface stress [2], but shifted to compression as the depth increased and ultimately attained larger magnitude and affected depth than the residual stress generated by a sharp tool. Similar results [3,4] about the effect of white layer on residual stress were obtained. The unique pattern of a residual stress profile has been identified as a critical factor of surface integrity for component performance. A compressive residual stress induced by hard turning and grinding was found to improve rolling contact fatigue (RCF) life [5,12,13]. Furthermore, deep compressive residual stresses in the subsurface may be more beneficial to bearing fatigue life than shallower stresses of greater magnitude. However, a white layer by hard turning may reduce RCF life as much as six times [12]. Klocke and Schlattmeier [8] reported that a slight grinding burn does not necessarily lead to an early gear fatigue. Due to the very controversial impact of white layer on fatigue life, a fundamental understanding on the nature of residual stresses associated with different surface types and their effects on fatigue life is essential. However, the literature review reveals several critical unsolved issues. First, the bulk residual stress data is limited to turned and ground surfaces free of a white layer. The nature of residual stress induced by a ground white layer is very ambiguous and lacking. Second, how residual stress resulted from heat treatment affects the final distribution of residual stress is not yet known. Third, how residual stress associated with a turned or ground white layer affects fatigue life has not been clarified yet. Therefore, a linkage between various patterns of residual stress profiles and RCF life has been poorly understood. The objective of
130
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
Fig. 1. Microstructures of the cross-sectional surfaces: (a) hard turned fresh (HTF), (b) ground fresh (GF), (c) hard turned w. white layer (HTWL), and (d) ground w. white layer (GWL).
this paper is to solve these pressing issues. This has been accomplished by determining the residual stress profiles of turned and ground surfaces with and without a white layer and their effects on fatigue life. 2. Machining tests Four types of unique surfaces were prepared: hard turned fresh (HTF) with compressive RS at the surface and maximum compressive RS in the subsurface, hard turned white layer (HTWL) with high tensile RS at the surface and maximum compressive RS in the deep subsurface, ground fresh (GF) with compressive RS at the surface, and ground white layer (GWL) with tensile RS at the surface. In addition, an as heat treated surface was prepared as a base line surface. Dry face turning AISI 52100 steel (62 HRc) discs of 76.2 mm diameter and 19.1 mm thickness were conducted on a rigid, high precision lathe with a BZN 8100 round tool (70% CBN content by volume in a TiN binder phase) with 0.15 mm/158 chamfer & 6.35 mm corner radius. The cutting velocity, feed, depth of cut are 1.78 m/s, 0.051 mm/rev, and 0.254 mm for the preparation of HTF surfaces using a sharp tool, while a worn tool (VB 0.5 mm) at speed of 2.82 m/s was used for the HTWL surfaces. High speed cutting using a worn cutting tool with large tool flank wear will produce high cutting temperatures which induce a thick phase transformed white layer with high tensile residual stress. Grinding was conducted at a surface grinder with an Al2O3 20A60H fresh grinding wheel. The GF surfaces were prepared at wheel speed 23.9 m/s, table speed 15.2 m/min, cross-feed 5.14 mm/pass, and down feed 12.7 mm/pass with water soluble coolant. The GWL surfaces were prepared at same machining condition except the down feed 26.0 mm/pass and without coolant. The measured surface roughness across feed marks of the turned and ground surfaces is Ra 0.16 mm. The 3D surface topography of the machined surfaces also has equivalent root
mean square of 0.2 mm. However, the amplitude parameter St (Rt counterpart in 3D surface topography), i.e., the sum of the maximum height of summits and the maximum depth of valleys, was 1.63 mm, 1.49 mm, 1.29 mm, and 2.11 mm, respectively for the HTF, HTWL, GF, and GWL surfaces. Microstructures of the four types of surfaces are shown in Fig. 1. It can be seen that the white layer thickness for the HTWL and GWL surfaces is about 7 mm. 3. Residual stress characterization 3.1. Residual stress measurement Residual stress profiles in Figs. 2 and 3 were measured by the Xray diffraction method coupled with layer removal by etching. Residual stress measurements were carried out at room temperature using a 4-axis Bruker-AXS X-ray machine with an area
Fig. 2. Residual stress profiles in feed direction.
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
131
Fig. 3. Residual stress profiles in cutting direction.
Fig. 5. Measurement repeatability of residual stress for HTWL surfaces.
detector. Using a cobalt source (l = 1.7889 A˚), the three-dimensional scans involved 29 scans between 908 2u 1078 for 9 tilt angles (C = 08, 108, 208, 308, and 408) at each of three rotation angles (f = 08, 458, and 908). The X-ray was generated using 35 kV and 20 mA and passed through a 0.8 mm collimator. Each scan was 15 s which was determined to be adequate for good peak intensity and definition. The gravity peak location method with 10% threshold value was used to calculate (2 1 1) peak positions using DIFFRACplus stress software. The change in lattice spacing at different tilt angles resulted in a corresponding peak shift. The residual strain was derived from the slope of both the linear and elliptical fit between the fractional change of the plane spacing (strain) and sin2 c. The Young’s modulus (E) and Poisson’s ratio (y) for AISI 52100 steel were E = 201 GPa and y = 0.27, respectively. For repeatability, measurements at three different locations at same subsurface depth for each sample were made and the results were averaged to give the residual stress profiles. The average and variance of residual stresses in Figs. 4–6 shows a good measurement repeatability (<10%) although the variation is expected due to several sources of measurement uncertainty. The error bars stand for the min/max values from the average value.
surface types, the affected depth of residual stress is rather shallow (<20 mm). In the cutting (or grinding) direction, the maximum compressive residual stress occurs at the surface for both hard turning and grinding. Considering the residual stress at the turned surface has a certain variation in Fig. 4, the maximum compressive residual stress has a great chance to be in the subsurface. It implies that a ‘‘hook’’ shaped residual stress profile may be still produced in the cutting direction. However, the magnitude of residual stress generated by hard turning is more than 3 times larger than that for grinding and the affected depth is nearly twice as deep. It should be mentioned that the magnitude of residual stress may change significantly for different machining conditions, but the patterns of residual stress profiles are very stable as far as a sharp tool or grinding wheel is used.
3.2. HTF vs. GF surfaces Several differences between gentle hard turning and grinding are observed for the residual stress profiles. In the feed direction, the maximum compressive residual stress occurs at the surface for the case of grinding while maximum compressive residual stress occurs at a shallow depth (5 mm) for hard turning. Also, the magnitude and affected depth of residual stress is greater for hard turning. Due to the gentle machining conditions used for both
Fig. 4. Measurement repeatability of residual stress for HTF and GF surfaces.
3.3. HTWL vs. GWL surfaces The residual stress profiles of the HTWL and GWL surfaces show significant differences. In the feed direction, both surfaces generated a tensile surface residual stress. Maximum tensile residual stress was located at the surface for the hard turning case, while maximum tensile residual stress occurred in the subsurface (5 mm) for grinding. While the GWL surface shows tensile stress throughout the entire measured depth (90 mm), the HTWL surface becomes compressive beyond 12 mm. At the depth of 40 mm the compressive residual stress becomes a maximum (826 MPa). The residual stress remains compressive for the HTWL surface to a depth of at least 110 mm. The affected depth of residual stress measured for HTWL surface is much greater than the GWL surface. In the cutting direction, similar trends are observed. The GWL
Fig. 6. Measurement repeatability of residual stress for GWL surfaces.
132
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
surface shows a slightly compressive (46 MPa) residual stress, while the HTWL surface has a tensile (102 MPa) residual stress. At the depth of 5 mm, both HTWL and GWL surfaces have the largest magnitude of tensile residual stress with the HTWL surface having the largest magnitude. Beyond this depth, residual stress of the GWL surface decreases slowly, becoming stable at about 20 mm. However, residual stress of the HTWL surface becomes compressive around 15 mm and attains a maximum compressive residual stress of 834 MPa at 30 mm. Beyond this depth the residual stress slowly decreases less than 200 MPa at 110 mm.
Table 1 RCF fatigue test conditions. Load
Peak Hertzian stress
Radius of contact circle
Load frequency
306 N
4629 MPa
26 mm
173 Hz
3.4. HTF vs. HTWL surfaces The HTWL surface generated a tensile surface residual stress while HTF generated a large compressive one. In addition, turning with a fresh tool imparts maximum compressive residual stresses of 638 MPa at depth of 5 mm in feed direction and 986 MPa at the surface in cutting direction but only affects residual stress in a shallow subsurface (20 mm). On the other hand, the HTWL surface showed much deeper (>100 mm) affected depth. The maximum compressive residual stress 826 MPa for the HTWL surface occurred at a depth of 40 mm in feed direction and 30 mm in cutting direction. The peak tensile residual stress at 5 mm in the subsurface for the HTWL surface is a result of the martensitic transformation which occurred in the near surface due to high cutting temperatures results from abusive turning conditions. 3.5. GF vs. GWL surfaces The difference between the residual stress profiles generated by gentle and abusive grinding is less dramatic than those for the hard turning cases. However, similar trends with a shift from surface compressive to tensile residual stress are observed. Gentle grinding resulted in a compressive surface residual stress that quickly attenuates at a depth of only 20 mm, while abusive grinding generated tensile residual stress throughout the measured depth of 90 mm. The magnitude and affected depth of the tensile residual stress is much greater for the GWL surface. 3.6. Machined vs. as heat treated surfaces By comparing the magnitudes and patterns of the residual stresses profiles from the as heat treatment surfaces and with those after machining, it is clear that machining itself is the deterministic factor for the resulting residual stress magnitudes and profiles, while the influence of heat treatment is minor. 4. Rolling contact fatigue (RCF) life tests
Fig. 7. RCF test setup.
the slave washer/spindle corresponds to half revolution of each ball in the tests, it means one revolution of spindle has 4 ball revolutions (0.5 8 balls) on the test surface. The load frequency was calculated as follows: 0:5 N n f ¼ 60 where the unit of spindle speed N is rpm, n is the number of rolling balls. The test samples are secured in a resting plate which is positioned with four rods that maintain the rigs position and rigidity. The resting plate is free to move up and down as it rests on a load cell, allowing accurate measure of the applied force. An acoustic emission sensor is mounted directly to the test sample. Prior to test startup, a thin layer of multi-purpose lithium complex grease was evenly distributed across the test surface, slave washer track, and ball bearings. Further lubrication was added at regular intervals throughout the duration of the test. Rotational speed of the spindle and washer was monitored using an optical tachometer. Parallelism and centricity between the slave washer and test surface is critical to ensure that the pressure distribution is uniform across the test surface. For this reason, a dial indicator gage was used to measure the vertical and horizontal runout of the washer and test surface. Using this method it was possible to adjust the horizontal and vertical displacements so that parallelism and centricity are 10 mm.
4.1. Fatigue test setup The experimental setup shown in Fig. 7 was used to monitor RCF tests of the ground surfaces. The test rig uses a Bridgeport milling machine to supplement as the drive motor and the shaft runs at a speed of 2600 rpm. The load is applied through the rotating shaft which also houses the slave washer. The test utilized eight chrome steel ball bearings with a diameter of 5.56 mm and a hardness of 63 HRc. A nylon bearing cage was used to hold the ball bearings and minimize noise during the tests. Slave washers of the same material as the test samples were created by turning. An 8 mm radius groove was machined on each slave washer to maintain the ball’s position during operation and provide a lower contact pressure between the ball and slave washer. The lower contact pressure in the slave washer would help ensure that fatigue failure would first occur in the test samples. The specific testing conditions are shown in Table 1. Because one revolution of
Fig. 8. Fatigue life chart for different surface types.
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
4.2. Acoustic emission (AE) monitoring of fatigue damage
6. Correlation between residual stress and fatigue life
In-situ monitoring of fatigue damage was accomplished using an AE signal acquisition and processing package AEWin [14]. An AE sensor was attached to the test sample with a holding fixture, while vacuum grease served as the coupling media between the sensor and test sample. The AE sensor has a 125 kHz resonant frequency and connects to an 18 bit PCI-2 data acquisition board that was incorporated into a PC. Before the data reached the PC it was passed through a preamplifier that was set at a 40 dB gain. A threshold of 45 dB was used. The sampled AE signal parameters in this study include: absolute energy, RMS, amplitude, counts, and average frequency. The occurrence of surface fatigue is determined by the shoot up of AE amplitude signal, which has been shown a reliable signal for monitoring surface fatigue [9]. The test was real-time monitored, in which the time, load, speed, and any observations were recorded along with application of lubricant. Also occurring at these intervals was the application of additional lithium complex grease applied manually through a spray. The monitoring continued until increased AE signals were detected due to surface pitting and spalling, at which time the test was ended and carefully disassembled to avoid any damage to the samples.
6.1. Fatigue life of HTF vs. GF surfaces
5. Fatigue life chart and damage mechanism Surface fatigue life refers to the number of stress cycles required to initiate spalling on the test surfaces. Results of the surface fatigue lives with variance are shown in Fig. 8. Three fatigue tests for each surface type were conducted for data repeatability. The HTF surfaces had an about 40% higher average fatigue life than the GF ones. The reason for the fatigue difference is due to several factors such as distinct patterns of the residual stress profiles. The surface fatigue life comparison chart for the gentle and abusive turned and ground surfaces in Fig. 8 shows that a white layer on the HTWL and GWL surfaces is very detrimental to fatigue life. In the case of hard turning, the HTF surfaces had a 760% longer average fatigue life than the HTWL surfaces. Similarly, the GF surfaces had a 780% longer fatigue life than the GWL ones. However, the HTWL surfaces had a 65% longer fatigue life than the GWL ones. Spalling is the predominant damage mechanism for the tested surfaces. A typical example of the fatigue damage is shown in Fig. 9. Since the fatigue tests were conducted at well lubricated conditions with a high-temperature multi-purpose lithium complex grease. Scuffing did not happen during the tests.
133
The HTF surfaces had an about 40% higher average fatigue life than the GF ones. The reason for the fatigue difference is due to several factors such as distinct patterns of the residual stress profiles. By comparing the residual stress profiles and fatigue life, it is clear that the life difference is concurrent with the HTF surfaces with large compressive residual stress at the surface and the maximum compressive residual stress in the subsurface. If fatigue cracks initiate from stress concentration due to surface asperities, the large surface compressive residual stress may impede crack initiation at the surface. Fatigue cracks may also be driven by the subsurface maximum shear stress experienced by the test surfaces in rolling contact. In this case, the maximum compressive residual stress in the subsurface may serve to delay the initiation of subsurface cracks. Because both surfaces have equivalent surface roughness (Ra 0.16 mm), the fatigue life is not likely to be significantly affected by surface finish. However, an important aspect of the surfaces is that the measured skewness of the GF surfaces (0.127) is less than that of the HTF ones (0.051). Negative skewness is an index to the surfaces ability to trap lubricant, as it signifies the presence of more valleys on the surface. For small or medium loads below shakedown limit, the valleys may act as reservoirs to trap lubricant and may increase fatigue life by decreasing friction between the balls and sample surface as they are in contact with one another during rolling. However, for large loads (above the shakedown limit) such as the 4.6 GPa peak Hertzian stress used in these tests, the valleys may act as notches and can promote crack growth due to the wedge effect. 6.2. Fatigue life of HTWL vs. GWL surfaces Compared with residual stresses of the HTF surfaces, HTWL surfaces have two significant changes. Surface residual stress shifts from high compression to high tension. The subsurface maximum compressive residual stress becomes more compressive and deeper in both cutting and feed directions. Obviously, the surface tensile residual stress leads to much earlier fatigue damage than the surface compressive residual stress. However, the increased subsurface maximum compressive residual stress seems not to effectively delay fatigue cracks initiation and propagation. Thus, the compound effects of deep compressive residual stresses in the subsurface and tensile residual stress at the surface may not be more beneficial to fatigue life than shallower compressive stress of
Fig. 9. Typical fatigue damage of the tested surfaces. (a) Surface spalling. (b) Subsurface fatigue cracks.
134
Y.B. Guo et al. / CIRP Journal of Manufacturing Science and Technology 2 (2010) 129–134
greater magnitude. It implies that fatigue cracks tend to initiate from surface rather than in the subsurface in rolling contact. It should be pointed out that, beside the compressive residual stress in the subsurface, the significant difference in hardness between the surface white layer and the beneath tempered zone causes a minor ability of the workpiece surface to compensate distortion due to external loads. This also enhances crack initiation and propagation in the subsurface as well as spalling in Fig. 9. For the GWL surfaces, the bulk residual stress is in tension and the peak tensile residual stress is in the very near surface (6 mm). Fatigue cracks induced by surface tensile residual stress are further facilitated by the peak tensile residual stress in the near subsurface. This may explains why the GWL surfaces have the shortest average fatigue life. In addition to residual stress, surface hardness of the machined surfaces in Fig. 1 could be another factor to contribute the difference in fatigue life since the GF surfaces are slightly harder than the HTF surfaces [10]. However, the effect of hardness alone cannot be separated from the compound effects of other surface integrity factors. The high hardness of white layer [3,11,12] may also contribute the shorter fatigue life of the HTWL and GWL surfaces. But the hardness effect alone is very difficult to isolate from that of residual stress. 7. Conclusions The key findings on the effects of machining-induced residual stress and white layer on rolling contact fatigue may be summarized as follow: The basic residual stress profiles by a sharp tool can be fundamentally changed by a turned white layer but not a ground one; The high tensile residual stress and hardness gradient associated with the white layer are due to high cutting temperatures induced by the large tool flank wear; The hook shaped residual stress profile of a turned surface may have about 40% more fatigue life than a ground one;
The machining-induced white layer can reduce fatigue life as much as 8 times. The turned white layer surface may have longer life than a ground white layer surface. Acknowledgment This research has been supported by the National Science Foundation under CAREER Grant No. CMMI-0447452. References [1] Matsumoto, Y., Barash, M.M., Liu, C.R., 1986, Effect of Hardness on the Surface Integrity of AISI 4340 Steel, Journal of Engineering for Industry, 180:169–175. [2] Konig, W., Klinger, M., Link, R., 1990, Machining Hard Materials With Geometrically Defined Cutting Edges-field of Applications and Limitations, Annals CIRP, 39:61–64. [3] Toenshoff, H.K., Wobker, H.G., Brandt, D., 1995, Hard Turning-influences on the Workpiece Properties, Transactions NAMRI/SME, 23:215–220. [4] Abrao, A.M., Aspinwall, D.K., 1996, The Surface Integrity of Turned and Ground Hardened Bearing Steel, Wear, 196:279–284. [5] Matsumoto, Y., Hashimoto, F., Lahoti, G., 1999, Surface Integrity Generated by Precision Hard Turning, Annals CIRP, 48/1:59–62. [6] Klocke, F., Brinksmeier, E., Weinert, K., 2005, Capability Profiles of Hard Cutting and Grinding Processes, Annals CIRP, 54/2:557–580. [7] Konig, W., Berktold, A., Koch, K.F., 1993, Turning Versus Grinding—A Comparison of Surface Integrity Aspects and Attainable Accuracies, Annals CIRP, 42:39–43. [8] Klocke, F., Schlattmeier, H., 2004, Surface Damage Caused by Gear Profile Grinding and its Effects on Flank Load Carrying Capacity, Gear Technology, 9/ 10:44–53. [9] Guo, Y.B., Schwach, D.W., 2005, An Experimental Investigation of White Layer on Rolling Contact Fatigue Using Acoustic Emission Technique, International Journal of Fatigue, 27:1051–1061. [10] Warren, A.W., Guo, Y.B., 2006, On the Clarification of Surface Hardening Mechanisms by Hard Turning and Grinding, Transactions NAMRI/SME, 34:309–316. [11] Guo, Y.B., Sahni, J., 2004, A Comparative Study of the White Layer by Hard Turning Versus Grinding, International Journal of Machine Tools and Manufacture, 44:135–145. [12] Schwach, D.W., Guo, Y.B., 2006, A Fundamental Study on the Impact of Surface Integrity by Hard Turning on Rolling Contact Fatigue, International Journal of Fatigue, 28:1838–1844. [13] Hashimoto, F., Guo, Y.B., Warren, A.W., 2006, Surface Integrity Difference Between Hard Turned and Ground Surfaces and its Impact on Fatigue Life, Annals CIRP, 55/1:81–84. [14] Physical Acoustics Corporation. 2003, PCI-2 Based ae System User’s Manual, ver. 1, .