Investigation of deterioration process of organic coating using 1-dimention som network combined with EIS

In this paper, the EIS characteristic of organic coatings tested under cyclic wet-dry conditions was analyzed. The changing rate of impedance was selected as the training sample for 1-Dimention SOM (Self-Organizing Feature Map) neural network. Each sample corresponds to a certain activation level of the neuron in the output layer. According to the speciality of 1 -D SOM network, the changing trend of the activation level can generally reflected the coating degradation process. Combined with the characteristic of EIS plots,the entire deterioration process can be divided into three stages i. e. medium penetration into coatings, corrosion initiation under coatings and the corrosion extension which causes coating delamina-tion.