Mechanical strength estimation of ultrafine-grained magnesium implant by neural-based predictive machine learning

Abstract

The relation between severe plastic deformation (SPD) and the mechanical behavior of the biodegradable magnesium (Mg) implants is not clearly understood yet. Thus, the present study aims to provide, for the first time, a framework for modeling the mechanical features of the ultrafine-grained (UFG) biodegradable Mg-based implant. First, an adaptive neuro-fuzzy inference system (ANFIS) and support vector machine (SVM) were employed to determine relationships between SPD parameters, including the kind of metal forming process, the number of the pass, and temperature of the procedure based on the restricted training dataset. Second, gene expression programming (GEP) and genetic programming (GP) were then used to further verify the estimation capability of neural-based predictive machine learning techniques. Comparison of estimation results with real data confirmed that both ANFIS and SVM-based models had high accuracy for predicting the mechanical behavior of UFG Mg alloys for fracture fixation and orthopedic implants.