Local Plastic Indentation Resistance of Retained Austenite in Bearing Steel

Abstract

This paper concerns the function of retained austenite to resist local plastic indentation deformation in bearing steel. Three austenitizing temperatures were used to adjust the morphology, volume fraction and carbon content of retained austenite in samples. The volume fraction of retained austenite was measured by magnetic method. The higher austenitizing temperature result in higher fraction of retained austenite and higher carbon content in it, and changes the morphology of retained austenite from film-like to blocky. Resistance to plastic indentation of retained austenite was measured by nanoindentation on samples with different heat treatments. Strain induced martensite transformation was observed by pop-in phenomena on indentation curves. Results showed that indentation resistance depends mostly on the local stability of retained austenite rather than the fraction and morphology. Furthermore, the local stability of retained austenite under indentation deformation is mainly attributed to the carbon content.

1. Introduction

GCr15 bearing steel, the most widely exploited material for ball and roller bearing applications, is routinely treated by quenched and tempered (QT) heat treatment with microstructures of tempered martensite, retained austenite (RA) and spherical carbides.¹⁾ The quantity and stability of RA have a crucial role in determining the fatigue performance and dimensional stability in bearing steel.^2,3) There are suggestions based on modeling that any amount of RA transformation leads to a deterioration in fatigue performance because the volume expansion accompanying its transformation introduces localized stresses.⁴⁾ Contradictory studies believed that martensite transformed from RA during fatigue should act to relieve local stresses through variant selection.⁵⁾ Meanwhile, it has been demonstrated that RA leads to cyclic hardening through transformation, and enhances the development of a mean compressive stress.⁶⁾ Volume expansion due to the transformation of RA is an important issue in respect of dimension precision of components in aerospace industry and machining tools.⁷⁾

When bearing under severe conditions, such as contaminated lubrication, the particles introduced by air or debris creates the indentation that exacerbate fatigue damage. Carlson et al.⁸⁾ reported that the smaller amount of RA results in sharper and somewhat larger upheavals surrounding the indent. This may cause the local penetration of the lubricant film and enhanced contact with the rolling elements, whose passage would lead to large stress pulses below the upheavals and hence enhanced spalling. Oppositely, the upheavals profile is smoother when the steel has a larger quantity of RA. However, Martin⁹⁾ believed that RA content is not the only factor to determining the indentation resistance.

Hardness is generally applied as an evaluate parameter of material’s ability against indentation deformation.¹⁰⁾ Plastic deformation of steel surfaces under indenting loads, such as ball to race contact in rolling bearings, depends on properties of the material which cannot be adequately expressed by hardness testing techniques, such as Rockwell or Vickers hardness. For the multiphase steel, a Rockwell or Vickers hardness value was the cumulative resistances to indentation offered by the martensite and austenite phases. The extremely high strains generated by the diamond indent causes austenite, the soft constituent, to transform to martensite which is much harder and contribution to indentation resistance from austenite will be abnormally high. Indentation testing with low contact strains, therefore, tends to reflect the presence of retained austenite more than Rockwell or Vickers hardness testing.^9,11) Moreover, for the size of RA at micro-scale, it is difficult to be measured with the traditional optical microscopy because of a too low resolution. Thus, a sensitive nanoindentation test is useful for investigating the initial deformation behavior and enables the load-displacement behavior of the entire process to be observed. The detailed introduction of nanoindentation technology can be seen elsewhere.

The objective of the present study is to explore micro-deformation behavior and plastic indentation resistance of RA in bearing steel with the aim that it can elucidate the relationship between RA and plastic indentation resistance.

2. Materials and Methods

The material for the present study is GCr15 bearing steel with the chemical composition of 1.05C–0.27Si–1.46Cr–0.33Mn–0.002P (wt %), which was supplied as spheroidized annealing condition. The samples quenched into oil (50°C) after austenitizing 10 minutes at 860°C, 920°C and 1050°C, then tempered at 150°C for 2 h, designated as 860QT, 920QT and 1050QT.

Microstructures were examined by field-emission-gun-assisted scanning electron microscope (SEM, JEOL, JSM-7600F). Samples for SEM observation were prepared in the normal procedure and etched with 3% nital for 5 s. Transmission electron microscopy (TEM) experiments were carried out on a JOEL 2100 microscope operated at an accelerating voltage of 200 kV. TEM samples were prepared using a twin-jet technique in a 10% perchloric acid+90% acetic acid solution at room temperature.

The volume fraction of RA was calculated by the method proposed by Zhao et al.¹²⁾ Samples were cut by wire-electrode with a size of 2×2×1.4 mm and measured in a Quantum Design Physical Property Measurement System (PPMS-9T (EC-II)). The applied magnetic field was changed from 5 T to 0 T in steps of 0.25 T at room temperature. The standard sample was annealed at 1000°C for 0.5 h and furnace cooling to room temperature for obtained the austenite-free sample as reference. Thermo-Calc software is used to calculate the equilibrium carbon content in RA using TCFE 6.

Microindentation experiments were carried out on a Zwick Nanoindentation equipments, following previous studies,^13,14,15) with load-controlled mode at a loading rate of 0.1 mN/s and maximum load set to 20 mN. The nanoindentation system (MTS XP) consisted of a Berkovich three-sided pyramidal diamond indenter with a nominal angle of 65.3° and indenter tip radius of 200 nm. The samples were prepared by mechanical polishing down to 0.5 μm and electropolishing with a solution of 5% perchloric acid and 95% ethanol at room temperature.¹⁶⁾ The microhardness was measured from Vickers indents obtained by a microhardness tester machine (Zwick/Rockwell). For electron backscattering diffraction (EBSD) observation after indentation tests, samples were additionally polished with 0.05 μm colloidal silica suspension.

3. Results and Disscussion

3.1. Microstructure and RA Content

Figure 1(a) shows the typical microstructures after traditional QT heat treatment, the microstructures consist of tempered martensite (dark, plate structure), RA (light constituent) and spherical carbides. These spherical carbides were determined as (FeCrMn)-C type mixed carbides partially dissolved or retained during austenitizing.¹⁷⁾ The retention of such primary spherical carbides is beneficial to resist abrasive wear in bearing applications and the amount of retained austenite is not likely to be high in the final microstructure after quenched. Figure 1(b) shows the similar microstructures as Fig. 1(a) at 920°C austenitizing. The difference between the 860QT and 920QT sample is that spherical carbides decreased and fractions of RA increased. Figure 1(c) shows the microstructures of 1050QT sample. The microstructures consist of tempered martensite and a large amount of RA and spherical carbides disappeared. The morphology of RA was confirmed by TEM in the Fig. 2. It is clearly shown that the RA has two types of morphologies: film-like and blocky austenite. For clarification purpose, all the TEM micrographs revealing austenite are shown in dark-field mode. By comparing Figs. 2(a) and 2(b), with austenitizing temperatures increasing from 860°C to 1050°C, the morphology of RA changes from film-like to blocky. This is agreed with results reported by Luo et al.¹⁸⁾

Fig. 1.

The microstructures of SEM micrograph: (a) 860QT, (b) 920QT, (c) 1050QT. It is consist of martensite (M), RA, and Spherical carbides (SC).

Fig. 2.

TEM investigates the morphology of RA: (a) 860QT, (b) 1050QT.

Figure 3 is the magnetic curves for the different samples. The standard sample contains no austenite, which is used as the reference. The volume fraction of the RA calculated from the saturation magnetization is 13.2%, 21.6% and 23.2% at 860°C, 920°C and 1050°C austenitizing temperature, respectively. The volume fraction of RA increased with austenitizing temperature increasing, and reaches a maximum value. This is due to the increased dissolution of spherical carbides result in more carbon enriched in austenite and martensite starting temperature (Ms) decreased. This tendency is in accordance with results reported by Stickels.¹⁹⁾ The carbon content in RA calculated based on thermodynamic equilibrium conditions determined as 0.7 wt%, 0.83 wt%, 1.05 wt% in 860QT, 920QT, 1050QT samples, respectively.

Fig. 3.

The magnetic saturation curves as a function of the applied magnetic field for all samples.

3.2. Vickers Hardness Curve and Nanoindentation Test

The typical load-displacement (P–h) curves of Vickers hardness and nanoindentation for the 1050QT sample are presented in Fig. 4. The large variations of the residual indentation depths show the result of different load. Clearly, there is no abrupt change in indentation curves under the load of Vickers (2 N) and nanoindentation (2 mN). Further evidence of the indentation resistance of RA in 1050QT sample is shown in Fig. 5. When a 2 mN load was applied on the area of RA (Fig. 5(b)), no obviously change of load curves was observed. Meanwhile, RA adjacent to the indent was stable and underwent no phase transformation according to the observation of Euler map (Fig. 5(c)) and phase map (Fig. 5(d)). However, a sudden displacement burst, called a pop-in, was observed on the load curve of 20 mN. It is indicate that judiciously choose load could be used to analyze the deformation behavior of RA during indentation.

Fig. 4.

(a) The vickers hardness curve of 1050QT sample at load of 2N. (b) and (c) Nanoindentation P–h curves of 1050QT samples with load of 20 mN, 2 mN, respectively.

Fig. 5.

Nanoindentation results of 1050QT sample. (a) P–h curve; (b) Contrast map of severalindents; (c) Euler map of the indent; (d) Phase map (austenite is shown in red color and white color refers to martensite).

The P–h curves under 20 mN load for the different samples are presented in Fig. 6(a). The P–h curves of samples were chosen from numerous nanoindentation tests. The detailed analysis of the pop-ins marked by arrows in the three P–h curves can give an indication of the mechanical behavior of RA during indentation test.

Fig. 6.

(a) P–h curves of different samples. Arrows indicate the position of the pop-ins. (b) The (P/h)–h curves of different samples. Solid black lines indicate the slope of the curve.

The elastic deformation region was identified by fitting the Hertz contact theory to the load–displacement (P–h) curves obtained by nanoindentation. The relation between the load, P, and penetration depth, h, during pure elastic deformation is expressed as:   

$$\text{P = }\frac{4}{3}{\text{E}}^{*}{R}_i^{\frac{1}{2}}{h}^{\frac{3}{2}}$$(1)

  

$$\frac{1}{{\text{E}}^{*}}=\frac{(1-{v_s}^2)}{E_s}+\frac{(1-{v_i}^2)}{E_i}$$(2)

where R_i is the curvature of the indenter tip, E* is the reduced modulus given in Eq. (2), ν_s and ν_i are the Poisson’s ratios of the sample and the indenter, and E_s and E_i are the Young’s moduli of the sample and the indenter, respectively. The Hertzian solution line in Fig. 6(a) is calculated from Eq. (1). Given a tip radius of 200 nm, the Young’s modulus and Poisson’s ratio for the indenter and the samples are 1140 GPa, 0.07²⁰⁾ and 210 GPa, 0.3,²¹⁾ respectively. Three P–h curves deviate from the Hertzian solution at the early stage of plastic deformation.

A change in the plastic deformation mode was analyzed from the (P/h)–h plots, as shown in Fig. 6(b). The slope of (P/h)–h curve of 860QT sample increase from 0.432 to 0.649 μN nm⁻² after the first pop-in. This increase in slope indicates the change of predominant deformation mode from dislocation slip to martensitic transformation, as suggested in the literatures.²¹⁾ When the martensitic transformation is finished, the slope changes back to a similar value of 0.441 μN nm⁻². In other words, pop-ins accompanied with an increase in the slope of the P/h–h curve corresponding to the martensitic transformation. A large increase in the slope after the pop-ins suggests that martensitic transformation results in an increase in hardness. The similar results appeared in 920QT and 1050QT sample.

A large pop-in plateau (about 100 nm) and lower load (about 2.5 mN) is observed at the load curve of 860QT sample. It is worth noting that with austenitizing temperature increasing, the indentation resistance of RA remarkably improved by not only increasing the load required for the occurrence of pop-in and but also decreasing the pop-in plateau. In this study, higher austenitizing temperatures result in dissolution of carbides, lead to higher carbon contents in RA and morphology of RA changed from film-like to blocky. The higher carbon content in RA will result in higher yield strength, thus improve the ability of indentation resistance. Moreover, the level of compressive stress is increasing with austenitizing temperatures increased from −60 MPa at 860QT to −235 MPa at 1050QT.^22,23) It can be explained that RA is strengthen by the surrounding martensitic plates which is related to the residual compressive stresses. With more enriched carbon, the RA stability increases and the critical value of plastic strain required to trigger martensitic transformation also increases. This result demonstrated that the ability of RA resist indentation is mainly attributed to carbon content other than phase fractions or morphology. The indentation resistance of RA is related to the mechanical stability of itself. Although the film-like of RA is more stable than that of blocky during uniform tensile deformation.^23,24,25) The inverse mechanical stability of RA during nanoindentation compared to uniform tensile test may imply different mechanisms, which needs to be further studied.

4. Conclusions

The local plastic indentation resistance of retained austenite is studied in high carbon GCr15 bearing steel subjected to different austenitizing temperature. The higher austenitizing temperature result in higher retained austenite fractions and changes the morphology of retained austenite from film-like to blocky. The experiment result shown that the fraction and morphology of retained austenite are not the main factor for indentation resistance. The improved plastic indentation resistance of retained austenite is mainly attributed to the higher carbon content.

Acknowledgements

This research was supported by the Ministry of Industry and Information Technology of China under the project of LNG shipbuilding and National Natural Science Foundation of China No. 51571141 and No. 51201105.

We also would like to thank Special fund for the development of talents in Minhang District, Shanghai.

References

• 1)   H.  Burrier: ASM Handbook Properties and Selection of Iron Steels and High Performance Alloys, vol. 1, ASM, Materials Park, OH, (1987), 380.
• 2)   H. K. D. H.  Bhadeshia: Prog. Mater. Sci., 57 (2012), 268.
• 3)   K. I.  Sugimoto,  N.  Usui,  M.  Kobayashi and  S. I.  Hashimoto: ISIJ Int., 32 (1992), 1311.
• 4)   E. S.  Alley and  R. W.  Neu: Int. J. Fatigue, 32 (2010), 841.
• 5)   H.  Bhadeshia,  H.  Abreu and  S.  Kundu: Int. J. Mater. Res., 99 (2008), 342.
• 6)   R.  Richman and  R.  Landgraf: Metall. Trans. A, 6 (1975), 955.
• 7)   C. H.  Surberg,  P.  Stratton and  K.  Lingenhöle: Cryogenics, 48 (2008), 42.
• 8)   D.  Carlson,  R.  Pitsko,  A.  Chidester and  J.  Imundo: ASTM Spec.Tech. Publ., 1419 (2002), 330.
• 9)   J.  Martin: ASLE Trans., 5 (1962), 341.
• 10)   D.  Tabor: The Hardness of Metals, Oxford University Press, London, (1951).
• 11)   T.  Tallian,  O. G.  Gustafsson and  H. O.  Walp: ASLE Trans., 3 (1960), 244.
• 12)   L.  Zhao,  N.  Van Dijk,  E.  Brück,  J.  Sietsma and  S.  Van der Zwaag: Mater. Sci. Eng. A, 313 (2001), 145.
• 13)   C.  Begau,  A.  Hartmaier,  E. P.  George and  G. M.  Pharr: Acta Mater., 59 (2011), 934.
• 14)   T. H.  Ahn,  C. S.  Oh,  D.  Kim,  K.  Oh,  H.  Bei,  E. P.  George and  H.  Han: Scr. Mater., 63 (2010), 540.
• 15)   B.  He,  M.  Huang,  Z.  Liang,  A.  Ngan,  H.  Luo,  J.  Shi,  W.  Cao and  H.  Dong: Scr. Mater., 69 (2013), 215.
• 16)   Q.  Furnémont,  M.  Kempf,  P.  Jacques,  M.  Göken and  F.  Delannay: Mater. Sci. Eng. A, 328 (2002), 26.
• 17)   J.  Chakraborty,  D.  Bhattacharjee and  I.  Manna: Scr. Mater., 61 (2009), 604.
• 18)   H.  Luo,  J. J.  Liu and  B. L.  Zhu: Wear, 174 (1994), 57.
• 19)   C. A.  Stickels: Metall. Trans., 4 (1974), 865.
• 20)   S.  Shim,  H.  Bei,  E. P.  George and  G. M.  Pharr: Scr. Mater., 59 (2008), 1095.
• 21)   K.  Sekido,  T.  Ohmura,  T.  Sawaguchi,  M.  Koyama,  H.  Park and  K.  Tsuzaki: Scr. Mater., 65 (2011), 942.
• 22)   M.  Villa,  K.  Pantleon and  M. A.  Somers: Acta Mater., 65 (2014), 383.
• 23)   X. H.  Lu,  W.  Li,  C. L.  Wang,  H. S.  Zhao and  X. J.  Jin: Acta Metall. Sin., 28 (2015), 787.
• 24)   X.  Xiong,  B.  Chen,  M.  Huang,  J.  Wang and  L.  Wang: Scr. Mater., 68 (2013), 321.
• 25)   M.  Mukherjee,  O. N.  Mohanty,  S. I.  Hashimoto,  T.  Hojo and  K. I.  Sugimoto: ISIJ Int., 46 (2006), 316.