Speaker and Affliation:
Prof. Peter Voorhees
Frank C. Engelhart
Professor of Materials Science and Engineering
9th February, 2021 (Tuesday), 08:00 PM (India Standard Time)
The classical method for designing materials to achieve certain performance goals involves a laborious procedure wherein intuition drives the design of a material that is then created and tested. In most cases, the performance goals are not achieved, and this costly procedure is repeated. By integrating data, computations, and artificial intelligence it is possible to break this expensive cycle and bring innovative new materials to the marketplace faster and at a lesser expense. The materials design process requires links between processing conditions and the resulting microstructure. We illustrate an approach to coupling the processing conditions of additive manufacturing (AM) to microstructure. A phase field model has been developed that follows the evolution of thousands of grains in three dimensions as a heat source propagates along a surface at the high rates seen during AM. Through this approach it is possible to determine the effects of the solidification conditions, the weld pool geometry, and multiple passes of the heat source on the resulting grain morphology. We also discuss a new method for determining difficult-to-measure materials parameters (data needed for design), in this case grain boundary mobilities. The mobility of grain boundaries is thought to be determined by the crystallography of the adjacent crystals. Through a direct comparison of a three-dimensional experimental movie to simulations of the evolution of 1500 grains in iron we have determined over 1600 reduced grain boundary mobilities. We find that the reduced mobilities vary by three orders of magnitude and exhibit no correlation with the boundary crystallography. This implies that factors other than crystallography set grain boundary mobilities. This brings new challenges to predicting the kinetics of grain growth in polycrystalline materials.