Three-Dimensional Probabilistic Finite Element Model of a Cemented Hip Implant for Brittle Fracture

Author

Daniel J. Truong, Cal Poly at San Luis ObispoFollow

DOI: https://doi.org/10.15368/theses.2022.19
Available at: https://digitalcommons.calpoly.edu/theses/2390

Date of Award

3-2022

Degree Name

MS in Biomedical Engineering

Department/Program

Biomedical Engineering

College

College of Engineering

Advisor

Lanny Griffin

Advisor Department

Biomedical Engineering

Advisor College

College of Engineering

Abstract

Hip implant fractures are rare yet severe complications that follow total hip arthroplasty. These fractures often occur at the implant stem for cemented hip implants due to proximal debonding of the bone cement. While deterministic models offer some predictive power of implant durability, they often are based on average properties, failing to capture the variability in properties that can be drastically different from the mean. This study developed a probabilistic finite element model of a debonded total hip replacement to better account for this variability to determine the most likely failure mode in a particular case. The hypothesis was that fatigue was the more likely cause of failure than overloading. The model used Monte Carlo sampling to create 1000 possible simulations varying in material stiffness and loading magnitude. The analysis found virtually no possibility of failure due to overloading. In contrast, failure due to fatigue had a 99.36% chance of occurring by comparing the maximum stresses recorded at the stem for these simulations to distributions of failure criteria based on literature. Fracture mechanics predicted that the implant in this loading configuration would fail between 0.15-2.2 years with a mean life of 0.6 years. The model had good agreement with the case study outcomes, particularly at predicting the failure site and durability of the implant. Moreover, altering assumptions only moderately changed the probabilistic analysis results compared to a deterministic analysis. Future work could be spent adjusting the model to match the cement implant interface or applying the probabilistic approach to other stages of the implant’s degradation.