Investigation of erosive wear behaviour of tungsten carbide cobalt coated metal matrix composites using ANN

An artificial neural network (ANN) model was employed on erosive wear result of aluminium 7034 T6 reinforced with different weight percentage of Al2O3 and SiC with tungsten carbide cobalt (WC-Co) coated composites. WC-Co in the form of powder was pretreated before coating, to identify the inflence of pretreatment on the substrate. The WC-Co coating was deposited with high velocity oxy fuel (HVOF) sparay method. The microstructural examination was carried through scanning electron microscope (SEM) and to identify the elemental composition of coated specimens, energy dispersive X-ray analysis (EDX) was used. These material charecterisation shows that pretreated WC-Co coated Al 7034 T6 composites at 500 mu m size exhibit thick closely packed, low porosity and good adhesion of membrane, while specimen at 400 mu m size showed properties opposite to the 500 mu m size speciemen. To predict the erosive wear behaviour of WC-Co coated Al7034 T6 composite, neural network model uses the parameters such as impact velocity, stand-off-distance, erodent temperature, impingement angle, the weight percentage of reinforcement and coating thickness. It is found that predicted values from the artificial neural network can be compared with the experimental result values.