Developing Artificial Intelligence Models for Predicting the Tribo-Mechanical Properties of HDPE Nanocomposite Used in Artificial Hip Joints

Abstract

Hip joint fractures are a serious and prevalent health concern, especially among the elderly. The forces that the lower limbs must withstand are quite strong and can exceed many times the person body weight, especially in the femur. The selection of materials for joint replacements depends heavily on mechanical and tribological properties like strength, hardness, and wear rate. HDPE has been effective as an external prosthetic foot, but its mechanical and tribological constraints have made it difficult to employ as a part of an artificial implant. In this study, an Adaptive Neuro-Fuzzy Inference System (ANFIS) models are utilized to forecast HDPE nanocomposites mechanical and tribological characteristics. The model incorporates various weight fractions of graphene and relies on data obtained from experimental tests. These nanocomposites are prepared with various compositions, and their structure and morphology are evaluated using physiochemical characterization. Compared to pure HDPE, adding graphene enhanced the hardness, strength, and elastic modulus of the HDPE. The nanocomposite with a composition of 2 wt.% graphene showed better mechanical properties compared with pure HDPE. In addition, a finite element model is created to simulate the hip joint and assess the HDPE nanocomposite load-bearing capacity under actual loading. The outcomes of the simulation analysis are consistent with the experiment findings. In addition, in comparison to pure HDPE, the nanocomposite containing 0.5 wt.% graphene showed better tribological performances. Finally, the microscopic examination demonstrated how the additives weight fraction affects the HDPE nanocomposites wear mechanism. Eventually, the ANFIS models proved its ability to predict the tribo-mechanical properties of HDPE nanocomposite with an error that didn't exceed 3%.