Numerical Modelling of Fretting Wear and Debris Particles

Abstract

Fretting wear has been observed for decades and can be found in structures or machines with contacting surfaces which experience oscillatory motion. Debris particles generated during fretting wear was recently known to be one critically important factor affecting the progress of fretting wear. Correctly predicting fretting wear inside structures can be very helpful to avoid structural failure and reduce maintenance cost. However, the current method to predict fretting wear relies on empirically measured coefficients and does not consider the debris particles. This PhD thesis investigates the effect of the debris particles via a hybrid Finite-Discrete Element Modelling (FDEM) method developed in this work. The hybrid method takes the advantages of Finite Element (FE) method and Discrete Element (DE) method by simulating large solid bodies/continuum material in FE while simulating discrete bodies/granular media in DE. Thus, the debris particles can be included in the simulation. Via the FDEM model, the effect of debris particles is studied via comparing the numerical results obtained from the models with and without debris particles. The introduction of debris particles is found to affect the contact between surfaces via changing the local contact force especially reducing the maximum force value. The physical mechanism of fretting wear is also explored in this PhD thesis by implementing seven different material evolution criteria including the most popular one: the Dissipated Energy method by Fouvry. It was found that the Dissipated Energy method is likely to agree with the experimental data at the cycle where the empirical coefficient is calculated but disagree before and after this point. The shear stress evolution criterion uses a damage indicator to reflect applied shear stress as a fraction of ultimate shear stress, and as such is a threshold method. Compared to the other criteria, it predicts fretting wear well vs experimental results.