Are plasticity models required to predict relative risk of lag screw cut-out in finite element models of trochanteric fracture fixation?

Abstract

Using a finite element model of unstable trochanteric fracture stabilized with a sliding hip screw, the benefits of two plasticity-based formulations, Drucker-Prager and crushable foam, were evaluated and compared to the commonly used linear elastic model of trabecular bone in order to predict the relative risk of lag screw cut-out for five distinct load cases. The crushable foam plasticity formulation leads to a much greater strain localization, in comparison to the other two models, with large plastic strains in a localized region. The plastic zone predicted with Drucker-Prager is relatively more diffuse. Linear elasticity associated with a minimum principal strain criterion provides the smallest volume of elements susceptible to yielding for all loading modes. The region likely to undergo plastic deformation, as predicted by the linear elastic model, is similar to that obtained from plasticity-based formulations, which indicates that this simple criterion provides an adequate estimate of the risk of cut-out.