Abstract:
Predicting the mechanical properties and the microstructural features of metals subjected to severe thermo-plastic deformation processes is of supreme importance in designing novel or enhanced materials. The objective of this research is to investigate these properties using a hybrid physically based multiscale modelling approach. A grain fragmentation model is proposed and implemented into the continuum dislocation dynamics model coupled with a crystal plasticity framework to predict the texture, grain size, yield strength, and dislocation densities. The proposed model is also used to understand the deformation mechanisms that influence the distinct mechanical behaviors of metals. In this study, the grain fragmentation approach was based on the grain-grain interaction where the formation and accumulation of the geometrically necessary dislocations at the grain boundaries restrict the free deformation of the grain. A misorientation difference arises between the core of the grain and its boundaries. The grain fragmentation process is triggered when the misorientation reaches a threshold value leading to the formation of new grains. The model was first applied to the face centered cubic metallic structures such as copper and aluminum subjected to equal channel angular pressing process (ECAP). Prior to the prediction of the ECAP and post-ECAP properties, model parameters have to be calibrated by a simple mechanical testing for the as-received material such as tension, compression, or shear. ECAP predictions have shown good reliability and predictability for both materials. The proposed model was then upgraded and developed to include additional deformation mechanisms such as twinning to be able to mimic the behavior of hexagonal closed packed metallic structures such as magnesium. The microstructural features and mechanical properties of the processed materials via ECAP were in good accordance with the experiments. Mechanical properties of the pre-ECAP, during ECAP, and post ECAP are studied and analyzed by the power of the proposed model.
