Improved Discrimination of Electrochemical Noise Signals Using Wavelet Analysis

Wavelet transform methods were applied to electrochemical noise voltage vs time data to investigate whether these transforms offer an improved methodology for discriminating among electrochemical noise signals arising from different types of localized corrosion. The ultimate goal of this effort is to provide a framework by which electrochemical noise data might offer more reliable, real-time predictions of corrosion. A number of alloy-environment combinations known to cause pitting and crevice corrosion were used. The signals were analyzed by using conventional signal processing techniques in the time and frequency domains and by using wavelet techniques in the time-frequency phase plane. A method was proposed to identify and visualize the corrosion intensity from the phase plane data. The predictions are compared with microscopic visual examination of the corroded specimens. The agreement provides evidence that wavelet transforms can offer an enhanced ability to categorize different types of electrochemical noise responses.