Validated, high-resolution, non-linear, explicit finite element models for simulating screw - bone interaction

Primary stability evaluation of screw implants through pull-out or push-in experiments is commonly used to investigate the mechanism of screw loosening. Numerical models simulating these testing methods could provide an enhanced understanding of the underlying attachment mechanisms as well as save time and cost in the development of new screws. However, previous numerical models have been limited by compromises between modelling the trabecular structure at high resolution versus incorporating sophisticated mechanical properties and boundary conditions, leading to overestimated mechanical performance. The aim of this study was to overcome these limitations. We developed explicit models incorporating the microstructure of trabecular bone, with frictional contact, and a non-linear material model incorporating damage. One model digitally inserted the screw into the trabecular bone structure using Boolean operations, while another model simulated the screw's rotational insertion. The results showed a strong correlation between numerical and experimental results (R2: 0.54-0.93) for force-displacement response in terms of stiffness and strength. We found that the damage induced by the screw insertion process is an important factor to be considered, as the absence of modelling it led to an overestimated stiffness in previous studies. The study highlights the importance of including frictional contact and also identified screw insertion damage as an important part of the simulating screw-bone interaction. Our findings demonstrate the potential of explicit finite element models for accurately replicating experimental push-in results and optimizing orthopaedic screws. The code is available at https://github.com/zhou436/Bone-Screw-Constructs-eFEM.