Comparison of RSM, ANN and PSO Algorithms to Predict the Porosity Level of Twin Wire Arc Sprayed Aluminium Coatings on AZ31B Magnesium Alloy

Abstract

The twin wire arc spray method (TWAS) is one among the efficient methods that is widely recognised and utilised for protecting components in surface coating technology due to its economic viability and adaptability. The quality of coating obtained in TWAS is excellent and the porosity is very low in comparison to other conventional methods. The TWAS spray parameters were optimized by single response optimization to obtain the minimum porosity utilizing statistical tools such as DOE and response surface methodology (RSM). Investigations focused on the influence of power, gas pressure and standoff distance. The analysis was done by three factors and five stages of analysis in the central composite design (CCD). The lowest porosity (1.21%) was obtained at optimum conditions of 5 kW power, 3.5 bar air pressure and 200mm stand-off distance, respectively, for TWAS coatings. The RSM, ANN and PSO is utilized to optimise the constructed model. The optimized value obtained through PSO and ANN were compared with the RSM. ANN gives a superior outcome when comparing PSO and RSM. It is validated by the response and contour plots in order to assist in selecting the spray parameters. The obtained results indicate that the coating porosity depends mostly on the arc current.