This study addresses one of the major barriers to additive manufacturing: the anisotropic properties in printed parts, primarily attributed to columnar grain structures formed during rapid melting and solidification, which compromise mechanical performance. The objective is to enhance the strength, ductility, and toughness of AM-fabricated parts by reducing anisotropic effects. The approach combines ultrasound-assisted AM, which controls grain structure, with computational modelling. Employing the integrated computational materials engineering (ICME) approach, the study develops a digital twin of the ultrasound-assisted AM process, integrating computational and experimental materials information for simultaneous analysis and prediction of properties and manufacturing process parameters. This method aims to reduce development costs and reduce environmental impact, enabling novel, optimised properties and alignment with topology-optimisation.

The research methodology incorporates finite element analysis, cellular automata, and a genetic algorithm to predict melt pool geometry and grain formation, examining the influence of various manufacturing process parameters on material properties and mechanical performance. Initial results show an alignment between the predicted and experimentally observed melt pool dimensions, underscoring the potential of this approach. The study identifies discrepancies due to variations in Goldak's ellipsoidal variables, highlighting the need for further optimisation of the numerical model. To address this, the non-dominated sorting genetic algorithm II (NSGA-II) is introduced for efficient selection of optimal Goldak ellipsoidal parameters, aiming to reduce computational time and manual model fitting. This study not only advances the field of AM by proposing a method to enhance mechanical properties of printed metals but also sets a precedent for more efficient and sustainable manufacturing processes.