The study of friction and wear is a crucial element in the effort to reduce carbon footprint in technology. It is evident that friction and wear are responsible for a significant amount of global energy losses, emphasizing the need for research on the topic. However, due to the complexities associated with multi-physics phenomena and surface roughness at the micro-scale, it can become challenging to understand the tribological processes involved. Apart from friction and wear, this interaction also gives rise to phenomena like frictional heating and the generation of third-body wear particles due to both adhesion and abrasion. These phenomena can lead to reductions in performance, efficiency and durability in mechanical systems. 

The aim of present work is the development of advanced numerical tools with the purpose of studying friction and wear processes in detail. Wear, friction and the associated heating can be found in nearly all types of sliding mechanical systems. Typical examples include, but are not limited to, bearings, gears, shafts and cams. The numerical methods which exist currently are usually simplified, using idealized assumptions and non-realistic boundary conditions. For this reason, many of the models are not able to account for the various mechanisms involved in multi-asperity contacts. 

This research presents a multi-scale and multi-asperity thermo-mechanical model to study the temperatures at the interface due to surface roughness, while accounting for wear with Archard’s wear law. Realizing that the thermo-mechanical behavior is influenced by the post-necking behaviour of stresses and their respective states, the subsequent work focuses on incorporating these in single asperity wear simulations. The research has provided valuable insights into the wear mechanisms, revealing various issues within classical models, such as Archard’s wear law and resulting in the development of more advanced tools. Specifically, in the context of asperity-to-asperity interaction, where non-linear effects are more prominent, a linear relation between the wear volume and load may no longer hold. To address this, the research introduces thermo-mechanical models that combine the Boundary Element Method with non-linear Finite Element methods to study temperatures, deformations and wear in asperity-to-asperity contacts.

Key findings suggests that the average interface temperature is independent of roughness, unlike the maximum temperature which increases with increasing high frequency cut-off values and decreasing Hurst exponent values. Recognizing the significant influence of strain-softening and stress-states on the thermo-mechanical behavior, subsequent studies have been directed towards addressing this aspect, while focusing on single asperity collisions. The work presents an advanced three-dimensional Finite Element and a meshfree particle method to simulate large deformations and fracture in colliding asperities, accounting for stress triaxiality and lode parameters. It is shown that the maximum temperature rise and total wear volume are both affected by the triaxiality values and strain-softening. Simulations conducted with a model based on the meshfree particle method reveals a critical parameter that signals the transition from mild to severe wear, leading to the creation of a wear particle at the interface. More importantly, the findings have reveal the limitations in Archard’s wear law, serving as motivation for improvement and resulting in the final paper. In the final study, an improved wear coefficient is presented, resulting in more accurate wear predictions than the traditional Archard’s wear law. The improved wear coefficient is deduced from the contact area and the accumulation of crack energy along the direction of frictional force, resulting in a spatially varying and non-linear relation between wear volume and load. This model is coupled with the Boundary Element Method, which assumes that the surfaces are flat and semi-infinite and that the interacting surfaces are perfectly-plastic. This advancement eliminates the necessity of resorting to large, complex, and often time-consuming finite element based methods. The work also highlights deficiencies in the classical Archard’s wear model in correctly predicting the wear particle formation.