Machine Learning-Based Predictions of Porosity during Cold Spray Deposition of High Entropy Alloy Coatings

Abstract

Porosity poses a challenge to the mechanical properties of cold sprayed coatings, especially when it is open or surface-connected, limiting the coatings’ capabilities to act as a barrier. The porosity formation is dependent on the feedstock powder characteristics and the cold spray process parameters. We present a machine learning-based approach to predict porosity based on the above-mentioned factors. Nine different machine learning models based on linear regression (LR), decision trees, random forests, gradient boosting, support vector machine (SVM), and neural networks were explored. Considering the excellent properties of high entropy alloys, Cantor alloy was taken as the consumable. Our dataset, derived from the literature and experiments, identified SVM with a linear kernel and LR as the top-performing models based on the Pearson correlation coefficient (PCC) and root mean square error, where the PCC values exceeded 0.8. The SHapley Additive exPlanations method helped in identifying that the type of gas and powder are the top two factors in pore formation.