Prediction of grinding parameters for additively manufactured Ti-6Al-4V alloy using machine learning techniques

Abstract

In this study advanced machine learning (ML) techniques - Random Forest, Gradient Boosting, XGBoost, and Neural Network have been used to develop predictive models for grinding parameters in additively manufactured Ti-6Al-4V alloy. It examines the key process variables, including table speed, depth of cut, and environmental conditions (dry, wet, and cryo) on grinding forces, temperature, and surface roughness. The entire dataset, comprising 99 experimental runs is evaluated using the K-fold cross-validation technique to ensure comprehensive model training and validation. The performance of each algorithm is compared using metrics like R² and Mean Square Error, providing a comparison of predictive accuracy. This finding offers the role of cooling conditions and the potential of ML for enhancing predictive accuracy in advanced manufacturing processes.