| Crystal plasticity models have been successfully used to take into account the effect of microstructure in modeling and designing crystalline materials. Up to now, several computational methods for crystal plasticity models have been proposed. The main objectives of these computational methods have been to overcome the problem arising from the non-uniqueness of active slip systems during the plastic deformation of a crystal and to increase the efficiency and computational speed. The problem with most current models is that either they are not efficient for all strain paths or are very expensive. In this thesis, three new computational methods are developed for crystal plasticity which are believed to be the most efficient and the fastest crystal plasticity models currently available. In the first model, the current single crystal yield function is modified to decrease the degree of nonlinearity of the model and increase its computational speed. The second model, which is a temperature and microstructure sensitive, is a bridge between computational mechanics and dislocation theory. The results of this model have been very impressive. The third model, the so called combined constraints model, is based on developing a new yield function for a single crystal to overcome the problem arising from the current proposed highly nonlinear yield functions for a single crystal. These models were implemented into ABAQUS/Explicit finite element code and two cases of uniaxial tension and tube hydroforming were simulated using these models. Finally, the combined constraints model was used to investigate the plasticity induced surface roughness in polycrystalline high purity niobium. The results of the newly proposed single crystal model showed a good match with experimental data. Also, compared with other crystal plasticity models such as the singular value decomposition multi-yield surface crystal plasticity model and Gambin model, the proposed crystal plasticity models were more efficient and computationally faster. |
