Artificial intelligence approaches for enhanced coating performance

Cold spray (CS) is an innovative manufacturing technology designed to produce metallic layers on diverse materials. This process involves propelling metallic particles at supersonic speeds using pressurized gas, causing them to impact the target surface and achieve adhesion through mechanical interlocking between the powders and the substrate. Integrating Artificial Intelligence (AI) techniques can enhance the understanding and quality of this additive manufacturing process. This work focuses on predicting the characteristics of particle deformation upon collision by exploring multiple Machine Learning (ML) and Deep Learning (DL) techniques with the aim of identifying the most suitable approach. The used dataset is mixed data, composed of experimental data and FEM data, generated by Finite Element models (FEM). The input parameters for the model are categorized into three macro-categories: process, powder, and substrate. The research aims to forecast particle behavior through this multidimensional approach and contribute valuable insights for optimizing the cold spray manufacturing process by applying DL methodologies.