Detecting pitting corrosion and its severity using wavelet entropy in electrochemical noise measurement

Abstract

Entropy as a measure of uncertainty was used to represent the results of the wavelet technique in electrochemical noise analysis. The experimental signals were obtained by recording the electrochemical potential and current noise of 7075 aluminum alloy in 3.5% NaCl solution. The electrochemical potential and current noise were decomposed into 16 levels using Daubechies wavelets. Wavelet output etropy was calculated using a kernel density estimator and the Shanon equation. It was demonstrated that low frequency scales had the highest entropy magnitude for passivation of aluminum surfaces in all experiments. The entropy contribution to high frequency scales decreased with the appearance of pits and as the pits increased and enlarged on the electrode surfaces (detected by scanning electron microscopy). The amount of reduction in entropy magnitude was attributed to the severity of pitting corrosion. It appears that a wavelet-entropy plot is a promising method to differentiate between different types and the severity of corrosion.