The fretting of mild steel in air

Abstract

A series of fretting experiments has been undertaken on mild steel in air, at room temperature using flat contact areas 15 mm square with direct measurement of specimen relative movement and a high speed hydraulic drive system. Loads of 1000 and 200 N were employed with peak-to-peak amplitudes between 5 and 115 μm at frequencies generally around 190 Hz. At 1000 N there are two main types of fretting behaviour. Above 20 μm, the wear behaviour appears to be unrelated to the amplitude over a wide range of fretting distances. This type of wear produces copious quantities of loose oxide debris, pitting and peak formation. Although the relationship between specimen weight loss and fretting distance may be conveniently represented by two straight lines, a fast early rate and a slower rate following on after about 4 × 10 6 cycles, there are good reasons to believe that this presentation is misleading. It suggests a transition in behaviour at around a large number of cycles. However, there is no external evidence to support the idea of a change in mechanism here. The concept that, in fretting, wear is being thwarted by the products of its own making provides a better fit to the data and is more consistent with other observations. At amplitudes below 10 μm there is far less wear and only minor pitting. There is little evidence for a fast initial wear rate and the oxide formed is more coherent. Even in long-term experiments, the oxide only covers a small part of the surfaces and the wear rate is much lower; it may even approach zero. Between 10 and 20 μm, the behaviour is transitional with some evidence of surface roughening and rapid initial material loss. Fairly coherent oxide forms in the longer term, when wear rates may be similar to those found below 10 μm. However, in this region there is far more variability. The results, and the appearance of fretted specimens, suggest that the amount of wear will be a function of specimen geometry and surface texture. It is concluded that the function of the debris is so central as to undermine the concept of a specific wear rate which may be defined in advance of testing under representative conditions. Thus, to a large extent, the variability in results between different workers represents an underlying truth of the fretting process.