Dry sliding wear behaviour of HVOF-sprayed cermet coatings (CrC-NiCr, WC-Co and WC-Co-Cr) using statistical analysis and ANN models

Abstract

Thermal spray is a cost-effective technique for achieving better surface characteristics. It is widely applied in automobiles, agriculture, aerospace, marine and chemical refineries. There is enough space to extend the service life of new components or repair and overhaul worn or damaged components. In this work, three kinds of coatings, namely CrC-NiCr, WC-Co and WC-Co-Cr on a base substrate (AISI 4340 steel) through the High-Velocity Oxy Fuel (HVOF) method. A pin-on-disc wear test rig was used to study the wear rate of the coated substrates. The experimentation was planned in accordance with Taguchi’s L27 orthogonal array. Scanning Electron Microscopy (SEM) of coated samples was used for microstructural characterization. The main effects of plots and ANOVA technique were adopted to investigate the significance of wear parameters and their influence on wear rate. The WC-Co-Cr exhibited better wear resistance associated with CrC-NiCr coating. Experimental results are analysed by developing Artificial Neural Network (ANN) models and regression models. The developed models’ predictions show good agreement with experimental values. Furthermore, ANN shows better prediction performance characteristics compared to regression. The R-Sq. (Adj.), R-Sq and R-Sq. (Pred.) of regression models and the R-value of ANN models depicted satisfactory adequacy and feasibility.