21/05/2026
This work focuses on the intersection of composite structures, orthopaedic biomechanics, finite element analysis, mechanoregulation, and fracture-healing simulation.
In many computational fracture-healing studies, the callus geometry is predefined or assumed. However, the size and shape of the callus are strongly influenced by the mechanical environment, especially the stiffness of the fixation implant. This assumption can affect the accuracy of mechanoregulation-based healing predictions and implant-comparison results.
In this study, we developed a Python-Abaqus-based biphasic stimulus-driven callus geometry estimation algorithm. The algorithm estimates the mechanically active callus envelope by iteratively analysing a candidate callus region, calculating a biphasic mechanical stimulus from deviatoric strain and interstitial fluid flow, and removing mechanically inactive regions. The predicted callus geometries were benchmarked using objective comparison metrics, including the Dice similarity coefficient and projected area error.
After benchmarking, the workflow was applied to a simplified tibial diaphysis fracture model stabilised with three intramedullary nail materials: glass/polypropylene composite, titanium alloy, and stainless steel. The results showed that reduced implant stiffness increased the predicted mechanically active callus envelope and improved the simulated mean callus stiffness under the assumed loading conditions. The glass/polypropylene composite nail produced the largest predicted callus volume and the highest mean callus Young’s modulus at the end of the 112-day healing simulation.
This study provides a benchmarked computational workflow that links callus geometry prediction with mechanoregulation-based healing simulation, enabling a more consistent evaluation of stiffness-tailored composite implants compared with conventional metallic fixation devices.
