A descriptive phenomenological model for white layer formation in hard turning of AISI 52100 bearing steel

https://doi.org/10.1016/j.cirpj.2021.01.014Get rights and content

Abstract

This paper investigates the characteristics and the formation of white layers and dark layers induced by hard turning of through-hardened AISI 52,100 steel. The investigation showed that different types of white layers exist e.g., formed predominantly through excessive thermal or mechanical energy loading. The thermally induced white layer is formed when the cutting temperature is above the critical austenitisation temperature for the material. The nano-sized microstructure is initiated through dynamic recovery, which transitions to dynamic recrystallisation when the temperature rises above the onset temperature for dynamic recrystallisation. The corresponding white layer is characterised by a higher retained austenite content compared to the unaffected material, and the presence of a dark layer beneath the white layer. The white layer and the adjacent dark layer are found to be ∼12% harder and 14% softer, respectively, compared to the unaffected material. On the other hand, the mechanically induced white layer is formed through severe plastic deformation, where the formation is controlled by dynamic recovery and results in an elongated and broken-down substructure. Neither austenite nor an adjacent dark layer could be found for such white layers. The mechanically induced white layer is ∼26% harder than the unaffected material. For both types of white layer, (Fe, Cr)3C carbides are found in the microstructure. The investigation shows that the heating rate, cooling rate, pressure, and duration of contact between the cutting tool and workpiece surface should also be considered to understand the underlying formation mechanisms. The characteristics of the examined white layers and the cutting conditions are summarised in a descriptive phenomenological model in order to create a systematic approach for the definition of the different types of white layers.

Introduction

Hard turning has been extensively studied for almost four decades, and the process has shown promising results in manufacturing complex parts with high quality and an improved economic and environmental impact. During hard turning, the surface integrity is greatly affected by the thermo-mechanical material removal mechanisms. When studying the surface integrity in finish hard turning of case-hardened steels, Rech and Moisan [1] concluded that the feed rate (f) and the cutting speed (Vc) had a major effect on the surface roughness and the residual stress level, both being important aspects of the surface integrity. For parts that are machined in the hardened state, the surface integrity is often the main concern since it affects their fatigue life performances. When Grzesik [2] studied the surface topography after various machining operations, he reported that the surface topography has a direct impact on the functional properties such as friction, corrosion, and fatigue. As reported by Tönshoff et al. [3], the hard-turning process introduces microstructural changes in the surface layer, depending on the cutting conditions, which are metallurgical transformations and known as white layer (WL) and dark layer (DL). The WL and DL are visible under an optical microscope after a cross-section of the machined surfaces has been polished and etched. The WL has a nanocrystalline microstructure which possesses a higher hardness than the bulk material (unaffected material), whereas the DL, usually compared to an over-tempered martensitic structure, is commonly softer than the bulk material. The phenomenon of a harder surface layer (in the WL) followed, subsequently, by s softer region (in the DL) that converges to the hardness of the bulk material, has previously been reported by several researchers. For example, regarding the AISI O1 tool steel, Navas et al. [4] reported an approximately 5% higher hardness in the WL, and an approximately 10% lower hardness in the DL, compared to the unaffected material. With respect to the WL thickness and how it can be affected by the Vc and the tool flank wear (VB), Chou and Evans [5] observed a parabolic behaviour of the WL thickness when investigating the AISI 52100 bearing steel. They reported a maximum WL thickness of about 10 μm when machining at 180 m/min using a VB ∼0.3 mm. Similar observations were also reported by Tanaka et al. [6] for a low alloyed steel (SCM415).

When it comes to grain size, Bedekar [7] reported that the grain size could be between ∼10 nm and ∼150 nm, depending on the initial microstructure and cutting conditions. Even grains as small as 5 nm have been reported by Ramesh et al. [8] when studying hard turning of AISI 52100 steel at 90 m/min. To investigate whether a phase transformation has taken place during the formation of WLs, the volume fraction of the retained austenite content is usually studied. Chou and Evans [5] observed a three-fold increase of the retained austenite, i.e., from ∼10 to 30 vol. %. Despite of the high austenite content, the authors did not report on any carbide dissolution. For a similar material and process settings, Akcan et al. [9] concluded that the large strains accompanied by dynamic recrystallisation cause a destabilisation of the carbides, which eventually leads to their dissolution. The authors reported that the austenite content was less than 10 vol. %, which can be compared to the ∼30 vol. % as reported by Chou and Evans [5]. When studying hard turning of AISI 52,100 steel at cutting speeds ranging from 200 to 300 m/min, both Ueda et al. [10] and Tanaka et al. [6] reported temperatures as high as 950 °C, which could explain the high austenite content that is usually reported for the WLs that are generated at the higher cutting speeds. However, as the duration of contact in hard turning, t, is extremely short (μs), other mechanisms such as severe plastic deformation, nano-sized grains, carbon diffusion, stress and strain, heating- and cooling-rate, and pressure must be considered to explain the microstructural evolution and consequently, the changes in the carbide morphology and the retained austenite content at room temperature. The effect of the contact time between the cutting tool and the workpiece material on the white layer formation has already been reported by Aramcharoen and Mativenga [11] for a H13 tool steel. The authors studied different cutting tool inserts and cutting speeds and found that when machining at 200 m/min, a ∼2 μm thick WL can be formed. However, no WLs were formed when machining at 800 m/min. Using a single colour Impac Pyrometer, they observed an increase by a factor of 2 in the chip temperatures when machining at 200 m/min compared to 800 m/min, but only minor changes in the workpiece surface temperatures. This indicates that a greater portion of the heat is dissipated with the chips and that the shorter duration of contact suppresses the WL formation. Similar observations were also reported by Bosheh and Mativenga [12] when conducting in-depth analyses of the white layer formation in tool steel using CBN inserts at three different cutting speeds (100, 400 and 700 m/min). They reported a thinner WL when the cutting speed was increased. The authors also connected the suppression of the WL to both contact time and temperature.

When comparing the characteristics of the white and dark layers with martensite and bainite, there is a likelihood that the hard-turned surfaces containing WLs will behave differently. Whether the hard-turned surfaces have superior or inferior properties is strongly dependent on the final microstructural constituents that are present after machining. As the hard-turned parts are high-performance components and usually loaded close to their physical strength, the mechanical performance is often studied by conducting rolling contact fatigue (RCF), tensile, and wear investigations. For example, Guo et al. [13] reported that the WLs created during hard turning can reduce the fatigue life of AISI 52,100 steel by as much as 8 times compared to a surface free from WLs. The authors studied the relation between the surface residual stresses and WLs after hard turning and grinding. A significant shift from compressive residual stresses for surfaces free from WLs, to tensile residual stresses for surfaces with WLs, was reported. In a series of investigations, Jouini et al. [14], [15] revealed that there is a clear connection between the surface topography and the fatigue life when studying RCF performance of precision hard turned surfaces, where the fatigue life was improved as the surface topography Ra decreased. Recently, when conducting RCF tests on a twin-disc machine, Jouini et al. [16] reported that precision hard turned surfaces have a fatigue life four times higher than those generated through grinding. The results were explained through the hook-shaped residual stress profile with subsurface compressive residual stresses for the hard-turned surfaces, and the low surface roughness (Ra ∼0.1 μm). When investigating the wear properties of the hard-turned surfaces with WLs under both dry and lubricated conditions during sliding friction, Zhang et al. [17] observed improved adhesive and abrasive wear properties, respectively, for the studied surfaces. The high hardness of the WL and the nano-crystalline grain morphology was reported to increase the strength of the surface and thereby improve the wear performance of hard turned surfaces containing WLs.

Despite the extensive amount of research considering the surface integrity after hard turning, there is still a lack of knowledge regarding the phase transformations that take place. This hinders a broader industrial acceptance of hard turning as a finishing process. The aim of this work is to provide a novel and comprehensive descriptive phenomenological model explaining the nature of WLs, and to classify the different types of WLs that can potentially offer new insights regarding the WL formation in hard machining. The descriptive phenomenological model is based on our previous findings when examining the temperatures at which the reverse martensitic transformation and/or severe plastic deformation occurs when turning, and the microstructural development and the associated thermodynamics and kinetics which change the residual stress state and austenite content [18], [19], [20], [21], [22]. New findings about cutting forces and hardness depth profiles are also taken into consideration. The comprehensive data of the WL-characteristics connected to the used process settings have enabled the descriptive phenomenological model to be presented in the current work. Fig. 1 illustrates how the studied parameters are linked to each other to enable i) an in-depth understanding of the underlying mechanisms that cause the surface microstructure to alter, ii) what the microstructural constituents are of the corresponding white and dark layers, and iii) whether the white and dark layers can be avoided or generated under controlled conditions. The workflow in Fig. 1 is the basis of being able to create the descriptive phenomenological model.

Section snippets

Work material

A high carbon-chromium steel, AISI 52,100, which was supplied in a spheroidised-annealed condition, was used as the work material. The chemical composition of the work material is provided in Table 1. The raw material was austenitised followed by either i) isothermal treatment or ii) quenching and tempering (QT) to produce either bainitic or martensitic structures, both of which are studied herein. The hardness of the steels was between 59 and 62 HRC. The retained austenite content after

Results

In Table 2, some of the surface responses for the studied cutting conditions are summarised. For example, the cutting forces tend to increase as much as three times as the tool wear reaches VB ∼0.2 mm during machining at 30 and 110 m/min. This can be compared to 260 m/min, where only a two-fold increase was measured. The X-ray diffraction analyses revealed that at 30 m/min, the progress of the VB generated even higher surface compressive residual stresses (in the axial direction, f), whereas for

Discussion

This investigation showed that when machining with fresh (unworn) cutting tools at cutting speeds ranging from 30 to 110 m/min, no WLs were generated, and both the surface and subsurface stresses were compressive. Also, only minor changes in the austenite content were recorded. On the other hand, as the Vc was increased to 260 m/min, discontinuous WLs covering parts of the feed grooves were detected on the machined surfaces. The observation of the discontinuous WLs is in line with previous

Conclusions

In this paper, we investigated the characteristics of white and dark layers (WL and DL) that are usually found at the machined surfaces after hard turning. The two main driving forces, i.e. thermal and mechanical load, for promoting the formation of the corresponding layers are strongly linked with the variables that were studied herein, namely cutting speed (Vc) and tool flank wear (VB). With help of extensive analyses of the machined surfaces using i) thermal analysis, ii) microstructure

Conflict of interest

None.

Acknowledgements

The research was carried out under the scope of the Centre for Metal Cutting Research (MCR) at Chalmers University of Technology. The authors would like to thank the ÅForsk Foundation for funding (Research grant No. 19-487). The Area of Advanced Production at Chalmers University of Technology is also acknowledged for financial support.

References (39)

Cited by (0)

View full text