Evaluation of a model for polymer composite wear behavior, including preferential load support and surface accumulation of filler particles

Polymer composites are attractive tribomaterials as they have relatively low steady state rate of wear and low coefficients of friction in the absence of fluid lubricant, and also have mechanical properties compatible with metals. With many of these composites, steady state rates of wear have been so effectively reduced that initial time-dependent run-in wear effects are now more important.; Many researchers have studied the tribological behavior and wear reduction of polymeric composite materials by experimental methods. However, few of them have been concentrated on theoretical modeling, especially time dependent wear. Numerical modeling has not been explored yet to predict wear behavior based on matrix and filler properties and volume fraction.; In this research, a proposed steady state and time dependent model for wear behavior was evaluated experimentally. The model described the effect of volume fraction and wear resistance of the filler and matrix material on composite wear behavior by using the rules-of-mixture theory and the conventional Archard's wear model. This model accounts for both preferential load support by the filler, and surface accumulation of filler. In order to evaluate the model's predictions of wear, pin-on-disk sliding tests were performed over a range of volume fraction for various polymeric composite systems prepared in the laboratory. To evaluate the accumulation of filler at sliding surface predicted by the model, Back Scattered Electron compositional contrast imaging was used. Image processing techniques were developed to quantify the volume fraction of filler and matrix at the sliding surface.; The steady state model was in accordance with experimental results. However, the transient model was not suitable to explain wear behavior during the initial transient period. In order to overcome this drawback, a simulation algorithm for time-dependent wear behavior was designed and developed. Wear-estimating equations were obtained by applying the regression technique to the computational results. The wear estimating equations can be used for prediction of wear volume change, surface composition change and composition profile change as a function of sliding distance.