Revaluation of Constant Strain Test Method for Stress Corrosion Cracking

Abstract

Among three well known test methods for SCC, i. e., constant total strain, constant load and constant strain rate ones, the first is most populary used as a preliminary searching method because of cheapness and convenience. But it only supplies a rough qualitative measure of susceptibility to SCC. Neverthless, if we focus our attention on the so-called “time to crack initiation” which is one of the most meaningfull factor in a practical engineering design, the method will give full play to its effectiveness because the stresses remain constant untill crack initiation. From this viewpoint, the authors took up some problems on this test method in order to obtain more reliable and reproducible data than ever. Tests were carried out at several temperature ranging from the boiling point to the solubility limit of corrodents (Fig. 10), on two kinds of stainless steels, i. e., SUS 304 and N. T. K. C-1 (Tables 1 and 2), using four different loading devices for bent-beam specimens (Fig. 1). The corrodents were MgCl₂ solution of several concentrations (35-42%). The results obtained are summarized as follows: 1) Exact calculation of the initial stresses in a specimen is of primary importance for an evaluation of SCC data. In this connection, a new method was presented considering the non-elastic properties of a ductile metal represented by 18/8 stainless steel (Fig. 6), and also the edge-to-center stress/strain variation along the width of specimen. At the central point on a specimen surface both state of plane stress and plane strain exist, therefore it holds that m=ε₂/ε₁=0 and n=σ₂/σ₁=0.5 (Fig. 9). Using these relationships and/or the following equation,
M=∫^b/2_-b/2∫^t/2_-t/2σ_1z·z dzdy
we can easily find the maximum initial fiber stress σ₁ under a given bending moment, which is taken as the ordinate in Fig. 12. As a result, it was found that the threshold stress for SCC depended on the type of specimen holder, or properly speaking, on the ratio of Q to M/l, where M is the maximum bending moment, Q the shear force at the central cross-section and l the span of specimen (Fig. 13). 2) Effects of temperature on the time to crack specimen in the corrodent of constant concentration, are represented as an Arrhenius' equation with a parameter of initial stress (Fig. 18). Crack pattern appeared on the surface of specimen depends on the temperature and concentration of corrodents (Fig. 19). It also depends on the type of specimen holder (Fig. 14) and the initial stress (Fig. 19).