Identifying corrosion forms on synthetic electrochemical noise signals by the Hilbert–Huang transform method

The analysis of electrochemical noise (EN), oriented to detect and to identify corrosion damages early, is a relevant task pursued by the researchers with the aim to obtain a reliable non-intrusive monitoring technique. Many analytical tools can be considered for this purpose: some of which are widely used (time analysis or frequency analysis), and others have been introduced only recently and are considered quite innovative, such as the Hilbert–Huang Transform (HHT). In this work, the efficiency and effectiveness of the HHT method to identify corrosion events on the basis of EN signals were evaluated. To achieve this aim, some synthetic signals structured to simulate real EN signals emitted during typical corrosion processes were generated. Uniform corrosion, passivation and pitting mechanisms were used as a reference for EN signal models. Furthermore, the signals have been combined in different ways and eventually the HHT model was applied to the so-obtained signals.