Wear of ceramic interfaces: Multiscale statistical simulation

When developing new tribological materials, such as high performance ceramics, the accurate estimation of their expected wear rate is a challenging goal due to the complexity of wear mechanisms and surface morphology and ultimately the multiscale nature of wear. The present work proposes a combination of modelling techniques, addressing these challenges. A dual scale model is formulated, with single-asperity interactions in the microscale and the interaction between stochastic surface morphologies in the macroscale. The microscale model is based on numerical simulations or alternatively to analytical models. The macroscale model, presented in this work, transforms surface topography into a multivariate distribution of asperity parameters and maps the microscale model's response: then it simulates the wear process via a Monte Carlo simulation, by querying the map and integrating over the interface. Challenges such as variable selection, mapping and profile evolution are addressed. The dry sliding interfaces of silicon nitride, tungsten carbide and zirconia, exhibiting abrasive wear mechanisms, are simulated and compared to a macrogeometric wear model based on Archard's law and experimental results. (C) 2012 Elsevier B.V. All rights reserved.