Analysis and prediction of slurry erosion wear response of silicon carbide reinforced Al2124 composite using Taguchi – artificial neural network approach Analyse und Vorhersage der Schlammerosionsverschleißreaktion von siliziumkarbidverstärktem EN AW-2124-Verbundwerkstoff unter Verwendung des Taguchi-Ansatzes und künstliche neuronale Netzwerke

Abstract

Slurry erosion is the prominent failure mechanism in the components exposed to particle entrained slurries. The slurry erosion wear behaviour of powder metallurgically processed Al2124 composite is investigated under slurry conditions with parameters like the impingement angle, impact velocity, slurry concentration, and stand-off distance. Aluminium oxide of 690 μm size is chosen as the erodent and the slurry jet erosion tester is used. The L₁₆ orthogonal array is used for the experimental design and the most influencing parameters were identified using the analysis of variance (ANOVA) results. Among the parameters studied, slurry concentration and impact velocity are observed to be the most influencing parameters on the erosion rate and surface roughness. Further, the experimental results are compared with those predicted by the regression and artificial neural network (ANN) models. The wear profile analysis of eroded samples shows U and W shape profiles for oblique and normal impact angle conditions respectively. Al2124 composite exhibits ductile erosion behaviour. The material removal mechanisms are analysed by scanning electron microscopy.