Investigation of Sliding Wear of SCF-PP Composites and Comparison with Wear Prediction by Machine Learning Methods

In this study, the sliding wear loss of Short Carbon Fibre Reinforced Polypropylene (SCF-PP) composites were investigated. The samples containing various concentrations (3-12 wt.% in steps of 3 wt.%) of SCF-PP were prepared by the injection moulding process. Several experimental trials to estimate sliding wear loss were designed according to the L16 orthogonal array. Further, for all 16 trials of SCF-PP composites, dry-sliding wear loss was examined by experimenting using a pin-on-disc operated as per the American Society for Testing and Materials (ASTM) standard. The wear loss was predicted for the same sliding wear conditions and sample types, applying machine learning algorithms such as Linear Regression (LR), Random Forest, Artificial Neural Network (ANN), and K Nearest Neighbour (KNN) and compared with the experimental results. Using a scanning electron microscope, the worn surface morphology of the selected samples and the sliding wear behaviour of the SCF-PP composites were examined. Major Findings: The effect of SCF concentration on wear loss, Wear loss prediction with Machine Learning (ML), Experimental validation, Worn surface Morphology analysis and Comparative performance of ML models.

View article page

Download PDF