Degree Name

Doctor of Philosophy


Department of Materials Engineering


The investigation reported in this thesis was carried out to study the abrasive wear behaviour of three groups of ferrous alloys, comprising ten carbon steels, three tool steels and four high-strength low-alloy steels, with the particular emphasis on the effect of microstructure. Wear tests were conducted using an unlubricated pin-on-drum machine, with silicon carbide, alumina and garnet abrasive papers, and a standard test condition of 20N applied load, 50mm/s sliding speed and 6m wear path.

The inter-relationships between wear resistance, microstructure, hardness and carbon content were studied using the carbon steels containing 0.10%C to 1.4%C. For constant hardness and carbon content less than 1.0%, the results showed that bainite had the highest wear resistance, followed by tempered martensite and the annealed structures. For a steel containing carbon > l.0%C, the annealed structure had wear resistance superior to the quenched and tempered structure or the spheroidized structure. Additionally, the relationship between wear resistance and hardness was linear for annealed steels, but the slope for hypoeutectoid steels was lower than for h~pereutectoid steels. Further, normalizing of these steels increased both hardness and wear resistance for the eutectoid steel, but hardness only for hypoeutectoid steel.

For quenched steels, the relationship between wear resistance and hardness was rather complicated, while for tempered martensite at the same hardness level, the higher the carbon content, the higher was the wear resistance. For a particular steel, a non-linear relationship between wear resistance and hardness of tempered martensite was confirmed. This wear behaviour indicates that abrasive wear resistance is not simply related to the hardness of materials, but is determined also by the microstructural properties.

The effect of heat treatment on abrasive wear behaviour was examined using a Ni-Cr-Mo- C tool steel, with specimens being either single-quenched or double-quenched from 900°C followed by tempering at l00°C, 200°C, 300°C and 400°C. The results showed that wear resistance was optimized by double-quenching followed by 200°C tempering, for which the microstructure comprised highly dispersed fine carbides in a matrix of tempered martensite. The relationship between wear resistance and hardness was non-linear for both single- and double-quenched specimens; however, for singlequenched specimens, wear resistance increased with hardness to Hv610, then decreased with further increase of hardness. For double-quenched specimens, wear resistance increased non-linearly with hardness to Hv660, which was the highest value available in the study.

The effect of massed carbides in the microstructure on abrasive wear behaviour was measured as a function of applied load using two high-carbon high-chromium tool steels designated Chrome and XW-5. The result that Chrome was more wear resistant than XW-5 under high applied load can be attributed to the,difference in the carbides in the two steels. In XW-5, only massed M7C3 type carbides were identified by energy dispersive spectroscopy, while in Chrome, mixed massed M7C3 and MC carbides were identified. Optical microscopy indicated that the carbide sizes and volume fractions were similar for both steels. It is concluded that wear resistance should be optimized for microstructures comprising mixed massed M7C3 and MC carbides randomly distributed in tempered martensite.

The effects of the concentrations of carbon and molybdenum on wear behaviour were studied using four high-strength low-alloy steels designated Bisalloys, which were available in the quenched condition as with martensite or bainite (high molybdenum alloy), or the quenched and tempered condition. The relationship between wear resistance and hardness after tempering a particular steel was non-linear, due to the effect of carbides precipitated during the tempering process. Also, for the specimens at the same hardness level, bainite had higher wear resistance than tempered martensite, and for tempered martensite, the higher the carbon content, the higher was the wear resistance, consistent with results for the plain carbon steels. However, after high temperature (>300°C) tempering, tempered bainite had lower wear resistance than tempered martensite due to the large inter-carbide spacing in the bainitic structure.

A thin white surface layer was generated under the standard abrasive wear test condition on specimens with prior microstructures of bainite or low temperature tempered martensite. For plain carbon steels, the thickness of the white layer increased non-linearly with carbon content. The generation of the layer can be attributed as the prior microstructures which comprised tempered martensite or bainite containing retained austenite and fine carbides, and the severe plastic deformation which occurred during abrasion. The structure of the white layer was possibly severely deformed martensite or bainite, containing extremely fine carbides with ultra-fine-grained structures.

Microscopical studies of the worn surfaces and of wear debris indicated that microcutting was the dominant mechanism of metal removal and that rnicroploughing, which formed grooves with prows and bulges, was a necessary precursor to microcutting, fracture and side-cut chip formation. Steels with high carbon content (annealed 1.2%C) or brittle rnicrostructure (quenched martensite) have a great tendency for microcracking. Microploughing was significant for low carbon steels. Additionally, some secondary wear mechanisms such as adhesion and delamination, were involved in the formation of small wear debris.

It is clear that the inter-relationships between wear resistance, hardness, composition and microstructure are complex. However, it is also clear that microstructure is as important as hardness or as composition in determining wear resistance. For a particular steel, wear resistance can be optimized by applying an appropriate heat treatment to generate the most suitable microstructure in relation to the tribological requirements of the application.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.