Multigrid Methods for Anisotropic Heterogeneous Material Simulation in Dry and Elasto-Hydrodynamically Lubricated Contacts

The objective of this thesis is to develop an efficient multigrid based algorithm, which can allow a dense grid to consider the effect of complex heterogeneous anisotropic microstructure for stress field calculation and fatigue life.
Firstly, a multigrid algorithm has been developed for a 2D/line dry contact problem. The effects of coating and material inclusions on the stress field and contact pressure distribution are studied. Then, the developed 2D multigrid algorithm is extended to solve a 3D/point contact problem where the efficiency of the multigrid method is highly needed. It is proven that the developed multigrid algorithm has the required accuracy and high efficiency for polycrystalline anisotropic material and can be applied for relevant industrial applications on small scale computers.
In order to advance the understanding of the mutual influences between the EHL pressure and the subsurface stress field and their combined effects on the RCF, a multigrid algorithm is developed for simultaneous solution of a 3D elastic problem with a 2D lubrication problem as the boundary condition. The time cost for solving the 3D stress field and the EHL pressure further proves the high efficiency of the developed multigrid algorithm. Finally, the effect of material anisotropy on the predicted rolling contact fatigue life is investigated based on the Ioannides-Harris fatigue life model. The effect of contact pressure, shear stress, grain size, and rotation angles on stress fields and fatigue life is analysed.
This thesis develops efficient multigrid algorithms and studies the effects of heterogeneous anisotropic material variation on the stress field and the fatigue life prediction of bearing material. In order to extend the fatigue life, rolling bearing material should be manufactured with fine grain size and optimized distribution of rotation angles. The developed multigrid algorithms can be applied for material optimization and computational diagnostics using actual data from electron back-scatter diffraction analysis of bearing material in future research.