Predicting the behavior of magnetorheological elastomer parameters on cutting performance during boring of AISI4340 steel using ANN

A novel approach has been devised to address the issue during the boring process, which is caused by the extended length of the tool. The approach involves the use of magnetorheological elastomer (MRE) parameters. By implementing MRE with varying iron particle sizes (10, 30, and 60 microns), different MRE curing times (12, 24, and 36 h), and various MRE composition ratios of iron particles (68%, 78%, 88%), silicon oil (30%, 20%, 10%), and additives (2%), the aim is to enhance cutting parameters when working on hardened AISI4340 steel. To evaluate the impact of these MRE parameters, an Artificial Neural Network (ANN) was employed to predict their effects on cutting force, tool wear, and tool vibration. The ANN model's results were then compared to experimental outcomes using the mean squared error, average absolute deviation, mean absolute percentage error, and R values as metrics. Remarkably, the experimental results and the ANN model's predictions aligned closely with each other, validating the effectiveness of the approach in mitigating, cutting force, tool vibration, and tool wear during the boring process.