NUMERICAL MODELING AND EXPERIMENTAL ANALYSIS OF RESIDUAL STRESSES AND MICROSTRUCTURAL DEVELOPMENT DURING LASER-BASED MANUFACTURING PROCESSES
This study is focused on the prediction of residual stresses and microstructure development of steel and aluminum alloys during laser-based manufacturing processes by means of multi-physics numerical modeling.
A finite element model is developed to predict solid-state phase transformation, material hardness, and residual stresses produced during laser-based manufacturing processes such as laser hardening and laser additive manufacturing processes based on the predicted temperature and geometry from a free-surface tracking laser deposition model. The solid-state phase transformational model considers heating, cooling, and multiple laser track heating and cooling as well as multiple layer tempering effects. The residual stress model is applied to the laser hardening of 4140 steel and to laser direct deposition of H13 tool steel and includes the effects of thermal strain and solid-state phase transformational strain based on the resultant phase distributions. Predicted results, including material hardness and residual stresses, are validated with measured values.
Two dendrite growth predictive models are also developed to simulate microsegregation and dendrite growth during laser-based manufacturing processes that involve melting and solidification of multicomponent alloys such as laser welding and laser-based additive manufacturing processes. The first model uses the Phase Field method to predict dendrite growth and microsegregation in 2D and 3D. It is validated against simple 2D and 3D cases of single dendrite growth as well as 2D and 3D cases of multiple dendrite growth. It is then applied to laser welding of aluminum alloy Al 6061 and used to predict microstructure within a small domain.The second model uses a novel technique by combining the Cellular Automata method and the Phase Field method to accurately predict solidification on a larger scale with the intent of modeling dendrite growth. The greater computational efficiency of the this model allows for the simulation of entire weld pools in 2D. The model is validated against an analytical model and results in the literature.