Wavelet transform modulus maxima and holder exponents combined with transient detection for the differentiation of pitting corrosion using electrochemical noise

A potentially powerful tool to detect and classify corrosion mechanisms is the analysis of electrochemical noise (EN). Data analysis in the time-frequency domain using, e.g., continuous wavelet transform (CWT) allows the extraction of localized frequency information, providing information on the type of corrosion, i.e., uniform or localized corrosion, from the EN signal. The CWT provides the opportunity to analyze changes in frequency behavior of EN signals over time. In the presence of transients generated by pitting corrosion that occur only during short instants of time, this is an important property. This paper introduces the combination of automated transient detection with wavelet transform modulus maxima (WTMM) and the Holder exponent. WTMM enhances the determination of transient frequencies by indicating the ridges of a CWT spectrum. The Holder exponent, a measure of singularity of an EN signal, provides a single parameter discrimination tool based on WTMM and serves to differentiate between general corrosion and two types of pitting corrosion of stainless steel Type 304 exposed to aqueous HCl solutions of different concentrations and as such at different pH values.