UKA Component Fatigue Test Development Using DOE and FEA

Abstract

Unicompartmental knee arthroplasty (UKA) is a clinical option for knee osteoarthritis continuum of care, providing a soft tissue sparing and less invasive approach compared to total knee arthroplasty (TKA). Compared to TKA, a unicompartmental procedure involves reduced blood loss, faster recovery, and fewer clinical complications. When designing new UKA implants, clinical risks must be assessed and mitigated. Implant fracture, while rare clinically, is an important consideration in design, and while design for minimal strength risk can be supported by simulation techniques and physical testing, published guidelines for such evaluations are not currently available. This publication presents a methodology developed in the course of verifying a new UKA tibial baseplate design and combines finite element analysis (FEA), design of experiments (DOE) and bench-top fatigue testing. The FE model and fatigue test setup have a three-point bend configuration which creates peak stress in the region where clinical fractures have occurred. Both FE model and fatigue test leveraged a draft ASTM standard for three-point bend testing of UKA baseplates. FEA and DOE were used in concert to identify a position for applied loading that produces worst-case stress in the implant. Subsequently, fatigue testing was performed on the implant that was analyzed by FEA and it was confirmed that the simulation accurately predicted the fatigue crack initiation site in implants tested to fracture. In recent years, the use of computational modeling to predict clinical performance has drawn the attention of regulatory agencies worldwide, and there is a growing expectation that such modeling must meet certain standards for verification and validation for it to be credible and also applicable to the performance questions it is used to answer. In this case, DOE explores the design environment by using not one but a range of load positions to find a worst-case scenario and to illustrate the sensitivity to model inputs. The bench-top test serves as a validation comparator, showing that the simulation is predictive of the product's physical behavior.