Study On Corrosion Monitoring Technology Based On Electrochemical Noise

As we know corrosion does enormous harm to our society. If the corrosion process ismonitored timely and reliably, the damage and loss which caused by corrosion will beminimized. Electrochemical noise test is one of the common methods to do that. As theworking environment has become more complex than before the use of traditionalelectrochemical noise test is limited. It is urgent for people to improve the electrochemicalnoise test technology. The new test technology must be maneuverable and more adaptable. Atfirst traditional three-electrode system was improved where a very thin platinum wire wasused as micro-cathode to replace the former working electrode made from the same material.Then the corrosion monitoring probe was developed which was built on Ag/AgCl solidreference electrode. The probe had a micro-negative-electrode with platinum filament embedin its center.Two kinds of Ag/AgCl solid-state reference electrodes were made respectively in thispaper, one was by pressing method and the other was by sintering. After that both of theirmain performance was studied such as electrode long stability, Nernst response property,temperature response property and electrochemical impedance. In general，the sinteredAg/AgCl solid reference electrode had a better performance than the pressed one.The pressed Ag/AgCl solid reference electrode was chosen to be made into corrosionmonitoring probe. Both methods of EN and EES were used to test electrochemical noisemade by AA2014aluminum alloy when the alloy was immersed in different kinds of solutionsuch as Ce(NO3)3, NaCl, Ce(NO3)3+NaCl and NaOH. Meanwhile the linear polarization wastested too. Through the analysis of the compare between the noise resistance and the linearpolarization resistance as well as the wavelet and corrosion morphology, the results showed:the corrosion monitoring probe is accurate in the passivation system, pitting system andinhibition system, but inaccurate in uniform corrosion system with large precipitation of H.The corrosion rates in different corrosion system, wavelet analysis and corrosionmorphology analysis were undertaken to study the applicability of the corrosion monitoringprobe. Due to corrosion monitoring probe showed good accuracy in crevice corrosioncondition, flowing corrosion condition and galvanic corrosion system, the corrosionmonitoring probe would have good applicability in all kinds of conditions. At last a model was established by artificial neural network technology. A number ofdata from different corrosion forms collected by the corrosion monitoring probe was took astraining samples of the artificial neural network. After comparing the verified results with theactual measurement results most of the measurement errors were quite small with a maximumof0.01.This illustrated that the artificial neural network had a quite good accuracy and thenetwork could differentiate corrosion forms from others.