Process modeling and parameter optimization of surface coatings using artificial neural networks (ANNs): State-of-the-art review

Abstract

Thin-film coatings and surface engineering procedures have a significant role in developing materials with extended mechanical, thermal and tribological properties. Advancement in the surface modification technology has encouraged investigators to develop new deposition techniques for producing high hardness, decent wear resistance, good corrosion resistance, high adhesion strength, and self-lubricating nature coated components. In this context, aiming desired level of such properties is key to successful performance of surface engineered coatings and their deposition methods. Parametric optimization and process modeling is a crucial step in the coating deposition process and for studying the properties attained for the coated components, but unfortunately it is not well understood in the literature. Although general, statistical and conventional approaches have been examined to model and predict the surface coating properties, there are still some challenges in outlining and comprehending the process parameters due to complex and non-linear nature of coating methods. Currently, machine learning techniques alternative to statistical approaches showing some sense of direction to address the quantitative gap between process inputs and outputs and build their relationship effectively. The review of this paper is primarily focused on providing a summary on artificial neural networks (ANNs) role in process modeling and parameter optimization of surface coatings. The present review can endow some knowledge to the researchers involving in this field of research.