ABSTRACT

The current and potential fluctuations (Electrochemical Noise) between two nominally identical carbon steel electrodes were recorded in a solution of Na3P04 at different concentrations. After the passivation was reached, pitting was induced by the addition of chlorides. The data were collected with different sampling rates. The Linear Polarization Resistances (LPR) were also measured. The electrochemical signals were studied in the time and frequency domains. The statistical parameters coming from the signals in the time and in the frequency domains were compared with the corrosion rates through LPR. The resistance noise was also calculated. A good correlation between data coming from the different techniques was found. A study on how sampling rate can affect the electrochemical noise data collection is also reported.

INTRODUCTION

All corrosion processes, such as general corrosion, pitting attack, crevice corrosion and stress corrosion cracking, as well as passive film build up, cause spontaneous fluctuations of the free corrosion potential of the electrodes. These fluctuations are termed Electrochemical Noise (EN) and their analysis is widely used for the examination of different corrosion processes [1-10]. Although Electrochemical Noise Analysis (ENA), has become of increasing interest in corrosion studies over the past 15 years, quite often some doubt still remains on the correlation between the occurring phenomenon and the signal generated. From a qualitative point of view, it has been reported that the values of statistical parameters of the EN, such as standard deviation (o) and the root mean squared (rms) values of the potential, can represent the corrosion status of the metal and permit the identification of the nature of the attack Some tentative correlations between standard deviation of EN and the corrosion parameters (corrosion current and polarization resistance) were proposed, but their application seems not to be so general. In the last years, a growing interest in the current noise has been shown. The study of the current fluctuations between corroding electrodes seems to be more significant than the potential studies. More recently, the use of the resistance noise as calculated from potential and current noise data was proposed to be equivalent to the polarization resistance Much discussion can be found in literature on this topic and a deeper analysis is required.

A powerful analysis of EN acquisitions can be performed by transposing data in the frequency domain. A traditional tool that provides this transformation is the Fast Fourier Transform (FFT) . FFT performs a spectral analysis of the random transient of the EN signal in a frequency range dependent upon the sampling time and the length of the data recording. The results of the spectral analysis are commonly given as dB of rms current or potential vs. frequency on a logarithmic scale, so obtaining the Power Spectral Density (PSD) plot. Some authors found an 1/f relationship between the measured noise power and the frequency. This seems to be an over-simplification of the problem, since every phenomenon occurring at the electrodes can have the effect of an additional slope on the PSD plot. The presence of a capacitance due, for example, to a film formation or to a diffusion element can increase the slope in the PSD plot of 20-30 dB/dec . Generally speaking, according to our experience, an f-á dependence would be more realistic, where á is different from unity. Moreover, the power level of the signal as calculated from the PSD in the frequency independent region, seems to be related to the corrosion rate. According to some authors, other transformation analysis such as Maximum Entropy Method (MEM) gives the

This content is only available via PDF.
You can access this article if you purchase or spend a download.