TEM study on the surface white layer in two turned hardened steels

https://doi.org/10.1016/S0921-5093(01)01447-2Get rights and content

Abstract

The structure of surface white layers was examined using transmission electron microscopy. Surface specimens were machined from a BS 817M40 steel (0.4C 1.2Cr 1.4Ni 0.2Mo) of 52 HRC and a low alloy tool steel (0.8C 1.7Cr 0.4Mo) of 58 HRC, with unworn and worn alumina/TiC composite cutting tools. Thin foil specimens were prepared such that the direction of observation was normal to the machined surface. The as-tempered microstructure of both steels was lath martensite. The structure of the machined surfaces of both steels was characterised by very fine, mis-orientated cells, less than 100 nm in size. The accompanying selected area electron diffraction patterns indicated the presence of retained austenite, the volume fraction of which increased with cutting tool wear. A refinement in the size of cementite particles was also evident. In the surface of the BS 817M40 steel machined with a worn cutting tool, there was evidence to suggest a degree of recrystallization. This may be accounted for by a transition from dynamic recovery to dynamic recrystallization during surface generation; a phenomenon which is favoured by the decrease in the work materials stacking fault energy as a result of the reverse martensite transformation.

Introduction

The generation of engineering surfaces by mechanical removal of material results in an alteration to the structure and properties of the work material at and just below the geometrical surface. In the machining of hardened steels, of 45–65 HRC, with geometrically defined cutting tools, a ‘white layer’ is frequently observed in polished and etched cross-sections of the surface when examined under reflected light. The white layer is a narrow band of non-etching material extending several microns to several tens of microns from the geometrical surface, e.g. (Fig. 1). Machined surface white layers are invariably accompanied by tensile residual stresses and are usually harder than the bulk martensitic work material [1], [2], [3].

White layers have also been observed under surfaces produced by reaming, abrasion [4], grinding [4], [5] and electrical discharge machining [6] and under impacted and worn surfaces [7], [8]. White layers are invariably found on the tool side of chips formed from hardened steels and may also be found within the chip itself due to localised catastrophic failure [9]. Identical, internal non-etching white bands are frequently observed in steels deformed at high strain rates (e.g. 103–106 s−1) and are typically referred to as ‘adiabatic shear bands’ [10], [11], [12], [13], [14].

Machined surface white layers have been the subject of a number of studies. As little information regarding the structure of the white layer can be deduced from light and scanning electron microscopy, it is necessary to employ X-ray diffraction (XRD) [2], [3] or transmission electron microscopy (TEM) [4], [15], [16] techniques. Tönshoff et al. [3], using XRD, noted an increase in the volume fraction of retained austenite from 20% in the virgin work material (AISI 5115 steel case hardened to 62 HRC), to 70% in a white layer of almost 20 μm in depth. Chou and Evans [2], also using XRD, noted 33% retained austenite in a white layer of 10 μm depth compared to 11% in the virgin work material (AISI 52100 steel of 62 HRC). The high austenite content of the surface white layer clearly attests to the occurrence of the reverse martensite transformation during machining. In an extensive TEM study on the structure of hand-reamed surface white layers, Turley [4] noted a loss of temper carbides, though found no evidence to support the occurrence of the reverse martensite transformation—the electron diffraction patterns were devoid of face centred cubic (FCC) reflections. It was suggested that the carbides were ‘annihilated’ with the dispersion of carbon atoms to dislocations, vacancies and sub-grain boundaries. In transmission, the white layer was found to be composed of extremely fine cells, between 30 and 100 nm in size. A fine-grained structure was also reported by Matsumoto [15] who examined machined surfaces of AISI 4340 steel of 52 HRC using reflection electron diffraction. In contrast to the findings of Turley, FCC reflections were noted in a number of the diffraction patterns, however, there was no indication as to which machining variables influenced the reverse martensite transformation. The factors favouring the retention of austenite, upon cessation of deformation and subsequent cooling to room temperature, were identified as a fine grain size and a lowering of the martensite start (Ms) temperature due to repetition of the transformation.

Recently, Liermann [16] undertook a TEM study of the white layers produced in the surface of a ball bearing steel of 62 HRC. In agreement with previous studies [4], [15], the layer was noted to be composed of very fine cells, less than 100 nm in size. From the electron diffraction patterns, it was estimated that the layer was composed of austenite and martensite in approximately equal parts. An amorphous oxide film, of several nm thickness, was also noted on the surface and was assumed to have formed during machining. In contrast, Matsumoto [15] suggested that an oxide layer, evident as ‘dim reflections’ in the diffraction patterns, was most probably formed during specimen handling and preparation.

The current TEM study on machined surface white layers was undertaken in order to compare the structure of surface white layers in steels of different compositions, to examine the structure of surfaces machined with unworn cutting tools, in which white layers are generally not observed and to verify or otherwise the findings of previous studies. In addition, the mechanisms of white layer generation and their similarity to internal adiabatic shear bands are discussed.

Section snippets

Experimental

Two steels were employed in the study, the compositions of which are given in Table 1. The BS 817M40 steel is equivalent to the AISI 4340 grade. Both steels were austenitised at 850 °C, oil quenched and tempered for 2 h at 250 °C. The resulting hardness of the BS 817M40 steel was 52 HRC; the low alloy tool steel was 58 HRC.

Surface specimens were machined with unworn and worn alumina/28%TiC/1%MgO cutting tools of ISO designation (1832:1991): TNGA 160412 T02520. The nose radius of the unworn

Results

Table 2 shows the average cutting tool life and the corresponding flank wear levels and tool force components in machining the BS 817M40 steel and low alloy tool steel of 52 and 58 HRC, respectively. The average surface Rz values for the BS 817M40 steel and the low alloy tool steel machined with unworn cutting tools were 1.90 and 1.78 μm, respectively; the Rz values for surfaces machined with worn cutting tools were 3.6 μm (BS 817M40 steel) and 2.24 μm (tool steel).

Discussion

The structure of the surfaces of the low alloy tool steel, machined with unworn and worn cutting tools and the BS 817M40 steel machined with an unworn cutting tool is similar to the structure observed by Liermann [16] in a ‘hard turned’ ball bearing steel. Also, the diffraction patterns obtained by Liermann, in addition to those obtained by Matsumoto [15], contained continuous rings, as do many of the patterns obtained in this study. The micrographs shown above are also similar to those

Conclusions

The white layers produced in the machined surfaces of hardened steels are composed of nanocrystalline partially transformed material which bears no similarity to conventional martensites. The volume fraction of retained austenite in the white layer increases with increased flank wear, due, it is thought, to a greater degree of completion of the reverse martensite transformation.

The similarity in the structure and crystallography of machined surface white layers and adiabatic shear bands

Acknowledgements

This work was funded by Enterprise Ireland and was partly undertaken in the Electron Microscopy Laboratory in UCD. The assistance of Dr D.C. Cottell of the Electron Microscopy Laboratory, UCD and D.N. Collins of the Department of Mechanical Engineering, UCD, is greatly appreciated.

References (27)

  • Y.K. Chou et al.

    Int. J. Mach. Tools Manufac.

    (1999)
  • D.M. Turley

    Mater. Sci. Eng.

    (1975)
  • B. Zhang et al.

    Wear

    (1997)
  • C.O. Mgbokwere et al.

    Mech. Mater.

    (1994)
  • J.L. Derep

    Acta Metall.

    (1987)
  • K.-M. Cho et al.

    Acta Metall. Mater.

    (1993)
  • J.A. Hines et al.

    Acta Mater.

    (1997)
  • M.A. Elbestawi et al.

    Ann. CIRP

    (1996)
  • M.A. Davies et al.

    Ann. CIRP

    (1997)
  • H.K. Tönshoff, C.R. Arendt, R. Ben Amor, CIRP 49 (2)...
  • H.K. Tönshoff, D. Brandt, H.-G. Wobker, SME Technical Paper MR95-215,...
  • W.J. Tomlinson et al.

    Surf. Eng.

    (1989)
  • P. Kruth et al.

    Ann. CIRP

    (1995)
  • Cited by (0)

    View full text