Development and application of smoothed particle hydrodynamics for studying surface erosion due to solid particle impacts

Surface erosion due to repetitive impacts of solid particles was numerically investigated by using smooth particle hydrodynamics (SPH). On the one hand, erosion is desirable and, hence, optimized in polishing processes such as abrasive flow machining. On the other hand, it is avoided due to its adverse effects in, for example, pneumatic conveyors. When dealing with repetitive high velocity impacts, it is of paramount importance to effectively prevent numerical instabilities that might otherwise propagate throughout the whole numerical domain of a target material. Therefore, a stabilization technique, namely hourglass control, was successfully applied for the first time to Eulerian SPH. This numerical stabilization forms the basis for the following applications within this thesis. A unified numerical method was employed to model material removal during an abrasive flow machining process where a slurry of low viscosity polishes the surface of a micro-channel as a result of random high velocity impacts of abrasive particles. Additionally, the effect of external magnetic fields on the efficiency of the polishing process was investigated by tracking the change of surface roughness at the presence of magnetic fields of different strengths which interact with magnetizable abrasive particles. The results showed that an external magnetic field gradient with a carefully adjusted magnitude can enhance the performance by increasing the material removal rate. This is achieved by forcing abrasive particles to move within regions close to the surface while maintaining enough kinetic energy to abrade the surface. The resistance of a surface against solid particle erosion is typically studied by measuring the ratio of total mass loss to total mass of solid particles for a given particle type and a given nominal impact velocity at different impact angles. In this thesis, next, the surface erosion of a stainless steel was investigated. To that end, a numerical methodology was proposed for reliable predictions of surface erosion. The methodology was then employed to systematically study the effect of particle shape. Based on the simulation results it was possible to extend a widely used erosion model that originally lacks an explicit description of the effect of particle shape.