Artificial Intelligence for Enhanced Abrasive Water Jet Machining Performance with Reinforced Composites

This research has addressed the advanced machining performance of hybrid natural fiber composites reinforced with hemp and Kevlar fiber with filler such as palm shell and coconut shell using Abrasive Water Jet Machining (AWJM). Enhancing the efficiency of machining, it sought to optimize the parameters that included Material Removal Rate (MRR), surface roughness (Ra) and kerf width, while employing a Support Neural Net (SNN) being a hybrid of Support Vector Machine (SVM) and Artificial Neural Networks (ANN). It utilized Pufferfish Optimization Algorithm (POA) to enhance the prediction power of the model. The experiments showing the importance of fillers on the machining processes were able to record significant improvement on MRR, Ra and kerf width during filling sessions. The predictive models developed were able to provide results that were close to reality with some models reaching up to 90% accuracy rating which demonstrates their usefulness in predicting machining operations under some conditions. While the differences were in slight margins due to the fillers added complexity, the study did bring forth aspects that pave the way for more refinement to the models to make them more accurate and suited to the industry’s needs. The results highlight the synergistic benefits of natural and synthetic fiber combinations with bio-based fillers in improving machining efficiency and sustainability, with the predictive models demonstrating their effectiveness in optimizing AWJM parameters.