In Selective Laser Melting (SLM), microstructure highly influences the mechanical properties of the alloys, such as strength, hardness, ductility, creep resistance and fracture toughness. Modeling the microstructure evolution in the solidification process is of great importance to optimize the ultimate mechanical properties. In this dissertation, a multiscale integrated model is built to simulate microstructure evolution of stainless steel in selective laser melting. This model couples three sub-models: transient thermal analysis model, nucleation model and grain growth model. The temperature field obtained in the macroscale finite element thermal analysis model built in ANSYS is transferred into the microscale nucleation model and grain growth model built in MATLAB. Nucleation model in...[ Read more ]

In Selective Laser Melting (SLM), microstructure highly influences the mechanical properties of the alloys, such as strength, hardness, ductility, creep resistance and fracture toughness. Modeling the microstructure evolution in the solidification process is of great importance to optimize the ultimate mechanical properties. In this dissertation, a multiscale integrated model is built to simulate microstructure evolution of stainless steel in selective laser melting. This model couples three sub-models: transient thermal analysis model, nucleation model and grain growth model. The temperature field obtained in the macroscale finite element thermal analysis model built in ANSYS is transferred into the microscale nucleation model and grain growth model built in MATLAB. Nucleation model includes the columnar-to-equiaxed (CET) effect, which is experimentally observed in selective laser melting. Grain growth is modeled by phase field method. Parallel computing embedded in MATLAB is applied to accelerate the simulation. The simulation is performed on 304L stainless steel to reveal the liquid to austenite solid state phase transformation process. The effectiveness of this model is verified by showing close agreement to the morphology of the samples produced by SLM experiments. This model could have further application on dynamic control of the cooling condition to optimize the mechanical properties without post-processing.