Isolating trabecular morphology to study bone damage

Bone structure is particularly complex and characterized by an intricate hierarchical architecture. Consequently, bone damage occurs at the multi-scale. Clinical applications typically analyze bone fractures at the macro-scale, but currently damage modes at lower scales are not fully understood yet. This research focuses on the understanding of meso-scale damage, characterized by a network of trabeculae of different thickness and spatial orientation. In order to isolate this articulated morphology, bone samples from porcine vertebrae are scanned through micro-computed tomography (micro-CT) and replicated by means of selective laser melting technique (SLM), obtaining Ti6A14V specimens. This is particularly useful, because these samples are realized with a uniform material, permitting to isolate morphological features. The SLM samples, after a check of the internal morphology, are mechanically tested under static compression. The load-displacement curve shows a first linear elastic section, followed by a collapse of the structure. This behavior is similar to the one of porcine vertebrae. Starting from micro-CT volume reconstruction, three finite element models are implemented. A global preliminary model of the entire sample is developed and the area with the highest level of strain is identified. In order to understand the distribution of stresses and strains in the critical zone, a sub-region of the original cylinder is considered. The results of the simulations identify a homogeneous distribution of deformations over the entire geometry, with the exception of the region characterized by a thinning of the trabeculae, called the failure band of the sample. By implementing an additional sub-model, the most strained trabecula is identified as the critical location, causing the collapse of the structure. The numerical models are then validated by comparing the numerical and experimental stiffness. This will allow to perform further analyses by varying the trabecular architecture and quantitatively evaluate the effect of morphology.