ELECTROCHEMICAL IMPEDANCE SPECTROSCOPY ANALYSIS OF COATING DETERIORATION PROCESS WITH KOHONEN NEURAL NETWORKS

The Electrochemical impedance spectroscopy (EIS)of the simulated specimens with different coatings was studied.The slope of log impedance modulus | Z | versus log f was put forward as characteristic parameter k(f) .The deterioration process was divided into five sub- processes by using Kohonen neural networks.The input of the networks is k(f) of different samples for different immersed time.The coating state can be defined by the output of the networks.Therefore the output can be considered as the quality index of the coating to some extent.The result indicates that the degradation process varies with different coatings and preparation techniques.The network proves to be useful to evaluate the coating performance.