Prediction of abrasive weight wear rate using machine learning methods

The paper applies machine learning methods to the prediction of the abrasive weight wear rates for a set of wear resistance experiments with model samples made of electrical copper and pure aluminum. To assess the effect of rotation frequency and abrasive grit sizes on the results of the evaluation of wear resistance and the determining quantities, the samples are tested with various disc rotation frequencies and various sandpaper grit sizes. It is noteworthy that, during experiments with materials comparable to copper in hardness and with harder ones, the path specified by the test design is implemented completely. However, this test design is not always suitable for the analysis of the mechanical characteristics of aluminum since, in some cases, the sample becomes fully worn before the end of the test. To overcome this problem, machine learning (ML) methods are proposed.