Most commonly used wear testing methods which use commercial abrasives allow only one sample to be tested at a time, meaning that generation of statistically valid data is very laborious. In addition, the commercial abrasive papers used do not allow true high stress abrasion to take place, as the abrasive particles do not fracture. This thesis investigates a new high stress abrasion test that uses a simple laboratory ball mill. The test involves adding samples of the materials to be tested to the ball mill, along with a ball charge and abrasive (such as sand or other mineral), and running the mill in the normal way. The wear rate of the samples can then be calculated from the weight losses. The advantages of this method are the ability to test many specimens at once, the ability to use almost any abrasive and the generation of true high stress abrasion conditions.
The validity and reproducibility of this testing method were investigated by looking at the effects of sample size, shape and edge radius on the measured wear rates of three materials. The effect of abrasive type on the performance rankings of a suite of different materials was also examined.
It was found that, in softer materials, the wear rate per unit surface area increases with increasing surface area of the sample. The wear rate, however, is independent of the aspect ratio. It was also found that samples with sharp edges initially wear more quickly than those with moderate radii, but this edge effect decreases with run time, and is usually no longer apparent after the first few minutes of testing.
The performance rankings of different materials is dependent on the abrasive used. Generally, wear rate decreases with increasing hardness, but the dependency changes with abrasive. In some abrasives, two-phase materials performed significantly better than single phase materials of similar hardnesses.
The results achieved using this testing method are promising in the correlation to data from field trials.