Abstract
The method of neural-network analysis of experimental data is applied to the corrosion system for the first time. The neural-network simulation enables one to predict adequately the response of steel 3-chloride-containing solution system to the external effects: the variations in the pH value and in the concentration of chloride ions. Trained neural network uniquely determines the characteristic potentials of corrosion system (the free-corrosion, passivation, and repassivation potentials) from the data on the chemical composition of corrosive medium and, based on these potentials, predicts the type of corrosion: active dissolution, passivity, or passivity with probable pitting.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have