Computationally efficient micromechanical process-structure-property modelling for biomedical materials

Additive manufacturing (AM) offers significant potential for producing complex, patient-specific biomedical implants using materials such as 316L stainless steel and bioresorbable magnesium alloys. However, AM introduces uncertainties, including microstructural anisotropy, porosity, and corrosion-fatigue interactions, that can compromise mechanical reliability and implant performance. This thesis develops a computationally efficient micromechanical framework for process-structure-property-performance (PSPP) modelling of these biomedical materials, fabricated via conventional and AM methods, with a primary focus on the latter. Motivated by the need to address these uncertainties, the work creates pragmatic simulation tools to predict mechanical behaviours such as tensile anisotropy, ductility, and corrosion-fatigue, while minimising computational demands through optimised modelling strategies. For AM 316L stainless steel, a geometrically based process-structure (GBPS) methodology is introduced to simulate columnar microstructures in struts, integrated with crystal plasticity finite element (CPFE) analysis. This approach evaluates tensile and bending responses across different build orientations, revealing orientation-dependent anisotropy. Specifically, horizontal builds exhibit higher yield strength, but reduced ductility compared to vertical ones, in agreement with experimental observations. As a physics-based alternative, cellular automata (CA) models are implemented for grain growth simulation during solidification, optimised for meshing efficiency to facilitate integration with CPFE for tensile property prediction. This CA-CPFE linkage is extended to three-dimensional (3D) space, providing a foundation for simulating complex AM microstructures. Furthermore, an efficient 3D extension of the GBPS methodology captures experimentally observed anisotropy in horizontal and vertical laser-based powder bed fusion (LB-PBF) 316L tensile samples. It enables scalable modelling of bulk samples under varying scan strategies, achieving substantial reductions in computational time through element shape optimisation, adaptive scaling, and convergence analysis. For bioresorbable magnesium alloys, the impact of pitting pre-corrosion on fatigue and fretting cycles to crack initiation is examined both experimentally and computationally. Fatigue testing of pre-corroded samples is conducted, complemented by a microstructure-informed micromechanical model that incorporates stochastic pitting corrosion and CPFE, with RVE size convergence validated to ensure computational reliability. The model demonstrates that even a single corrosion pit, corresponding to less than 0.5% mass loss, significantly accelerates crack initiation and reduces cycles to failure, aligning with experimental fractography and fatigue results. Preliminary LB-PBF experiments on WE43 magnesium alloy reveal achievable relative densities >97% but highlight persistent porosity and melt pool irregularities, underscoring the need for optimised AM parameters to enhance implant reliability. The integrated modelling approaches and experimental insights, establish PSPP relationships that provide insights into the effects of AM parameters on material behaviour. These frameworks offer tools for predicting microstructural influences on mechanical properties, supporting more informed design decisions for medical implants.

Funder

University of Galway
Research Ireland

Publisher

University of Galway

Rights

CC BY-NC-ND

cbn

Collections

University of Galway Theses (PhD Theses)