Computational Design of Functionally Graded Materials from Sintered Powders

A new computational method is presented for the efficient design of alloy systems in functionally graded materials (FGMs), optimized for manufacturability (sintering) as well as performance. The design methodology uses a multi-objective genetic algorithm (GA) integrated with computational thermodynamics and physics-based predictive models to optimize the composition of each alloy in the FGM. Thermodynamic modeling, using the CALPHAD method, is used to establish microstructural constraints and calculate the effective diffusivity in each alloy of the FGM. Physics-based predictive models are used to estimate performance properties. The model is verified by comparing results with data from the literature. A design exercise is also presented for an FGM that combines a ferritic and an austenitic stainless steel to demonstrate the capability of the methodology. It is shown that the mismatch in sintering rate between the two alloys, which causes processing defects during co-sintering, can be minimized while the solution hardening and corrosion resistance in the austenitic alloy can be optimized by independently controlling the composition of both alloys, the initial particle sizes and the sintering temperature.