Artificial Neural Network Modelling of Corrosion Inhibition of Mild Steel in Marine Environment Using Epoxy-Nickel Oxide Nanocomposite Coatings

The corrosion inhibition of mild steel in 3.5 wt. % NaCl in the absence and presence of epoxy coatings containing NiO nanoparticles with concentrations of 1.0, 2.0, 3.0 and 5.0 wt. % respectively was studied using the gravimetric technique for a duration of 60 days at room temperature and varying temperatures ranging from 30 to 60°C for 5 hours. The Nickel oxide nanoparticles with average particle size was 23 nm were synthesized by the chemical precipitation technique followed by calcination in a muffle furnace for 3 hours at a temperature of 300°C. Results from the study reveal that epoxy-Nickel oxide nanocomposite coatings are effective green corrosion inhibitors for mild steel in 3.5 wt. % NaCl under different operating conditions and at temperatures within the range of 30 to 60°C. A predictive model based on the Artificial Neural Network (ANN) was developed to study the relationship between the input variables (exposure time, inhibitor concentration and Temperature) and output variables (Corrosion Rate and Inhibition Efficiency). The ANN model was based on the Multilayer Perceptron algorithm with input layer comprising of 3 factors and 23 units. Hyperbolic tangent was used as the activation function for the hidden layer which was made up of 3 units. The output layer with two dependent variables was made up of 2 units. Corrosion test data obtained from 80 experimental runs were successfully modelled using ANN with minimal errors. 56 cases corresponding to 70% of test data were used for training the network and 24 cases corresponding to 30% of test data was used for testing the efficacy of the network. The model had sum of squares error of 0.981, average overall relative error of 0.018 for the training component and values of 3.190 and 0.043 for the sum of squares error and average overall relative error respectively for the testing component.