JOHNSON–COOK MODEL COMBINED WITH COWPER–SYMONDS MODEL FOR BONE CUTTING SIMULATION WITH EXPERIMENTAL VALIDATION

Abstract

Constitutive models are widely used to predict the mechanical behavior of different kinds of materials. Although the Johnson–Cook model for bovine bone and Cowper–Symonds model for human thoracic rib and tibia was developed, the predictability of these models was found good only at particular strain rates. This study addresses this lack of information by investigating the Cowper–Symonds model, Johnson–Cook model, and Johnson–Cook model combined with Cowper–Symonds model at different strain rates to utilize in the bone cutting simulation. Specimens prepared using two rear femurs harvested from a 3.50-year-old bovine were investigated at different strain rates (0.00001–1/s). A comparative study made among the stresses predicted from these models showed 29.41%, 10.91%, and 11.11% mean absolute percentage errors using Cowper–Symonds model, and 2.03%, 7.19%, and 3.62% mean absolute percentage errors using Johnson–Cook model, respectively, at 0.0001, 0.001 and 1/s strain rates. However, the Johnson–Cook model combined with the Cowper–Symonds model predicted the stress with a maximum of only 2.03% mean absolute percentage error. The potential of each model to utilize in the orthogonal bone cutting was also evaluated using Ansys® and found that the combined model predicted the cutting force close to experimental cutting force with minimal error (5.20%). The outcomes of this study can be used in the surgical practice and osteotomy procedure before commencing actual surgery.