Study the Effect of Seawater Environments and Surface Roughness on Uniform Corrosion Rate of Carbon Steel Using Neural Network Modeling

Abstract

In this research, the effect of seawater environments and surface roughness on uniformcorrosion rate of carbon steel (A516 grade 65) was studied depending on the experimentalwork and artificial neural network modeling. The experimental work involves chemicalcomposition, samples machining, roughness measurements (for carbon steel specimens),conductivity and salinity measurements (for seawater), and uniform corrosion test. Weightloss technique was employed in determining the uniform corrosion rate in carbon steelmaterial. Also, artificial neural network (ANN) model was built to predict the values ofuniform corrosion rate (mpy) at different values of conductivity, salinity for seawater androughness factor for carbon steel depending on the experimental results which were used trainand test the ANN.The results obtained of uniform corrosion rate by ANN predictions are shown to beagreed well against experimental values. i.e. correlation coefficient, R=0.9974