Assessment of damage models of additively manufactured Ti–6Al–4V lattice structures under compression

Abstract The design of lattice structures for additive manufacturing (AM) involves computer-aided simulations that require accommodating process-induced irregularities into their 3D models, coupled with proper material definition. For simulating lattices of materials with limited ductility like Ti–6Al–4V, the use of damage models is essential to accurately predict the post-yielding behavior and fracture path. Although additively manufactured lattices have been extensively studied, and numerous models have been developed, the literature lacks guidelines for selecting a suitable failure criterion for a given lattice type. The current study attempts to address this limitation by examining two damage models, Johnson-Cook (JC) and Gurson-Tvergaard-Needleman (GTN), over three Ti–6Al–4V lattice types: BCC (bending-dominated), octet (stretching-dominated), and TPMS gyroid. Different relative densities were obtained by changing cell size and strut diameter/wall thickness. Experimental results showed that all samples failed at a 45° angle, regardless of their relative density. Moreover, stress-strain behavior exhibited the post-yielding softening/collapse typically associated with Ti-based lattices. The finite element simulation results for different damage models were compared against the experimental results of compression tests of different AMed lattice structures. Results of the numerical models based on actual structural dimensions measured by scanning electron microscopy (SEM) were found to agree with the experimental findings up to the peak stress value, beyond which the models were sensitive to the damage criterion applied. It was concluded that the GTN model was best for strut-based lattices as they fail by microvoid nucleation and growth near the nodes, while the JC model was best for TPMS gyroid as they fail by progressive plastic deformation, given their continuous surfaces.