Application of artificial neural networks to predict corrosion behavior of Ni-SiC composite coatings deposited by ultrasonic electrodeposition

A feed-forward, multilayer perceptron artificial neural network (ANN) model with eight hidden layers and 12 neurons was used to predict the corrosion behavior of Ni-SiC composite coatings deposited by ultrasonic electrodeposition. The effect of process parameters, namely, ultrasonic power, SiC particle concentration, and current density, on the weight losses of Ni-SiC composite coatings was investigated. The grain sizes of Ni and SiC were determined by using X-ray diffraction (XRD) and scanning probe microscopy (SPM). Results indicate that ultrasonic power, SiC particle concentration, and current density have significant effects on the weight losses of Ni-SiC composite coatings. The ANN model, which has a mean square error of approximately 3.35%, can effectively predict the corrosion behavior of Ni-SiC composite coatings. The following optimum conditions for depositing Ni-SiC composite coatings were determined on the basis of the lowest weight loss of Ni-SiC deposits: ultrasonic power of 250 W, SiC particle concentration of 8 g/l, and current density of 4 A/dm2. XRD and SPM results demonstrate that the average grain sizes of Ni and SiC in the Ni-SiC composite coating are 90 and 70 nm, respectively.