Exploring a parallel rheological framework to capture the mechanical behaviour of a thin-strut polymeric...

Computational modelling of bioresorbable scaffolds (BRS) has employed several different material property models, ranging from those based on simple elasto-plastic theory through to anisotropic parallel network models that capture the viscoelastic-plastic behaviour observed in poly-l-lactic acid (PLLA). The increased complexity of higher fidelity material models, particularly in terms of calibration to in-vitro data, can limit their use. Consequently, their suitability for predicting the mechanical response of next-generation BRS is not well understood. Therefore, we have used the Bergstrom-Boyce (BB) parallel network material model, implemented in Abaqus/Explicit (Dassault Systemes), to investigate the mechanical response of a scaffold based upon the ArterioSorb^TM BRS (Arterius Ltd, Leeds, UK). In-silico crimping, balloon expansion and radial crushing were simulated and validated against an analogous in-vitro test. Calibration of the model to uniaxial tensile test data was considered given the model's strain rate dependency and the inability to maintain the natural time period of the simulation when using the explicit solution method in finite element analysis. The isotropic limitations of this model were also explored. The model was also compared to an elasto-plastic model developed by the authors in previous work. Relative to bench-top measurements, prediction of the final diameter and radial strength of the scaffold by the BB model was found to be significantly more accurate than other models, within 2% of the in-vitro result. Additionally, the effect of the crimping strategy and an elevated ambient temperature upon the in-silico prediction of the post-crimping scaffold diameter were investigated. A multi-step crimping process with holding to facilitate stress relaxation and the lower stresses induced by the increased temperature were found to improve the accuracy of the predicted post-crimping scaffold diameter.