Various forms of discrete models, such as lattice or particle models, are increasingly used to model multi-contact problems at the engineering scale. Generally, the discrete model approach uses bonds between particles, which are described by constitutive equations to reproduce accurately material elasticity (through homogenization techniques) and contact (e.g., Hertz contact) behavior. Furthermore, discrete models are suitable for studies where a large fracture network might develop. Such a case is found in tribology, in which damage and third bodies appear at the contact interface during frictional sliding.
We perform discrete element simulations to conduct sliding simulations of rough bodies with and without third body generation through particle wear (see figure). As a first step, we have validated the approach by comparing our results with analytical models for normal contact (e.g. Greenwood and Williamson in 1966). Our model, which considers nonadhesive, frictionless contact between rough surfaces represented by identical spheres, is able to reproduce accurately the evolution of the real contact area as function of the applied load. Current work involves parallel discrete simulations of frictional sliding.
Figure: Study of normal elastic contact between rough surfaces is modeled by a rough rigid surface squeezes on an elastic body. The elastic body can be seen as a tensegrity structure, because the spheres initially in contact are permanently bonded.
