Abstract
Electrochemical noise measurements were performed on electrodes made from a sample of AISI 304 (UNS S30400) stainless steel. Potential and current noise were measured under open-circuit conditions in a three-electrode system, one of the electrodes being maintained under load and acting as a working electrode. Two types of electrode set-ups were used, which differed according to the method of load application. In the first case, the common electrode was a U-bend test specimen, which was subjected to a static, large, but undefined load. In the second case, it was a tensile test specimen, which was subjected to a gradually increasing load (the slow rate load test [SRLT]). U-bend tests were performed in a concentrated solution of magnesium chloride (MgCl2) at an elevated temperature, test specimens being dismantled and examined daily through a microscope. SRLT were conducted in a dilute solution of sodium thiocyanate (NaSCN) at room temperature, with simultaneous electrochemical noise (ECN) and load/elongation measurements. Toward the end of both types of tests, significant simultaneous spikes of voltage and current noise were observed. The spikes detected during SRLT correlated with drops in the applied load and sudden increases in elongation. After the tests were completed, several characteristic ECN time series, referring to both types of test specimen, were subjected to spectral and chaos analysis. Attempts to distinguish between the active cracking and nonactive cracking periods on the basis of the aforementioned types of analysis were unsuccessful since, in the case of the U-bend tests, the events that are typical for stress corrosion cracking (SCC) processes were too rare. In the case of the SRLT, the ECN time series were not stationary, which is a necessary condition for a successful outcome of such analyses.
