Wear and Friction Modelling of Carbon Fiber Reinforced Epoxy Composites

The wear resistance of unidirectional carbon fiber reinforced epoxy under severe abrasive sliding conditions was studied. It was found that unidirectional laminates tested with the fibers parallel to the sliding direction (UDp) were more wear resistant under the same set of test conditions when compared to the laminates tested with fibers transverse to the sliding direction (UDap). A novel energy-based model was developed to explain the difference in the wear rate of unidirectional laminates. The model predicts that wear volume per sliding distance is proportional to the total frictional energy, the average volume to surface area ratio of the debris, the energy required to generate new surfaces, and a novel k factor that represents the fraction of the total friction energy used for creating wear particles. It was found that while the wear mechanisms of UDp specimens enabled the generation of debris particles with higher volume to surface area ratio, a lower fraction of the frictional energy was used for wear particle generation (a much lower k), which ultimately lead to a lower wear volume per sliding distance compared to UDap specimens. Heterogeneous surface wear models were proposed to predict the specific wear rates of woven carbon fabric reinforced epoxy laminates. Derivation of the models assumed that each fiber orientation was either worn at the same rate due to normal load redistribution (EW model) or worn at different rates but under equal pressure (EP model). Under mild abrasive sliding, the EW model was able to predict the specific wear rates of both laminates well. However, the models grossly under-estimated the specific wear rates of the woven laminates due to wear mechanisms that are specific to woven fabric laminates that involved the out-of-plane fibers at the cross-over points in the carbon weave. Finally, a model was proposed for predicting the coefficient of friction of the woven fabric laminates. This model was derived using the same set of assumptions as the EW model for specific wear rate. The model results matched the experimental data very well in all tested conditions.Ph.D