Simulation-driven microstructure engineering: linking process, structure, and performance

Tailoring microstructure for enhanced performance is a central objective in the advancement of metal additive manufacturing (AM). Achieving this goal requires a predictive understanding of how process parameters influence grain morphology, texture, and local heterogeneities, the key features that govern material behaviour. To support microstructure design in AM, we have developed an integrated computational framework that links process conditions to microstructural outcomes and mechanical properties across multiple length scales. This framework combines melt-pool-scale thermal simulations based on the finite difference method, cellular automata-based simulations for microstructure evolution, and micromechanical simulations using crystal plasticity finite element method. An analytical method is also included for rapid estimation of effective elastic properties at the mesoscale. These tools have been validated experimentally for laser powder bed fusion systems and extended to accommodate directed energy deposition. A distinguishing aspect of this work is its capacity to simulate grain structure evolution not only at the multi-layer, but also at the part scale. Using high-performance computing resources we achieved efficient cellular automata simulations of parts exceeding 1,000 mm³ in volume. This capability enables targeted exploration of processing routes to design grain size, shape, and crystallographic orientation distributions aligned with desired performance outcomes. By enabling direct links between AM process parameters, evolving microstructures, and resultant properties, this computational approach provides a foundation for the digital design of microstructure. It advances beyond descriptive modelling to support rational, simulation-driven microstructural engineering in AM.