Quantitative analysis of tribological performance in polyamide 6/nano--zeolite composites for prosthetic and orthotic applications: Integration of computer vision, numerical modeling, and experimental evaluation

Abstract

This study investigated the effect of nano-zeolite (NZ) integration on the tribological properties of polyamide 6 (PA6) nanocomposites using a hybrid approach combining image segmentation, experimental techniques, and numerical modeling. Automated SEM image analysis with U-Net segmentation quantified nanoparticle distribution, while pin-on-disc tests assessed wear performance. A support vector regression (SVR) model linked nanoparticle dispersion, load, and hardness to friction and wear rates. A combined dispersion index (CDI) further quantified nanoparticle distribution. Results demonstrated significant wear rate reductions (41 % at 60 N and 54 % at 100 N) with 5 wt% NZ. The SVR model highlighted the importance of nanoparticle dispersion and load on performance, with hardness having minimal influence. Morphological analysis confirmed smoother worn surfaces with NZ inclusion.