Tribo-Performance Analysis of Fly Ash–Aluminum Coatings Using Experimental Design and ANN

Abstract

This article proposes the application of an artificial neural network (ANN) to a Taguchi orthogonal experiment to develop a robust and efficient method of analyzing and predicting the solid particle erosion wear response of a new class of metal–ceramic coatings. An ANN model based on data obtained from experiments performs self-learning by updating weightings and repeated learning epochs. In this work, plasma-sprayed coatings of fly ash premixed with aluminum powder in different weight proportions are deposited on aluminum substrates at various input power levels of the plasma torch. Erosion wear characteristics of these coatings are investigated following a plan of experiments based on the Taguchi technique, which is used to acquire the erosion test data in a controlled way. The study reveals that the impact velocity is the most significant among various factors influencing the wear rate of these coatings. An ANN approach is then implemented taking into account training and test procedure to predict the tribo-performance under different erosive wear conditions. This technique helps in saving time and resources for a large number of experimental trials and successfully predicts the wear rate of the coatings both within and beyond the experimental domain.

KEY WORDS:

• Erosion Wear
• Metal–Ceramic Coatings
• Fly Ash
• Taguchi Technique
• ANN

Review led by David Burris