scholarly journals Improvement of Intergranular Corrosion and Intergranular Stress Corrosion Cracking Resistance of Nickel-Base Weld Metals by Stabilization Parameter Control

1986 ◽  
Vol 35 (11) ◽  
pp. 605-615 ◽  
Author(s):  
Kiyoshi Yamauchi ◽  
Ikuhisa Hamada ◽  
Tomoaki Okazaki ◽  
Tomomi Yokono
CORROSION ◽  
1981 ◽  
Vol 37 (2) ◽  
pp. 104-115 ◽  
Author(s):  
T. M. Devine ◽  
B. J. Drummond

Abstract The ability of various intergranular corrosion and pitting corrosion tests to detect sensitization and intergranular stress corrosion cracking susceptibility in high temperature water of duplex 308 stainless steel was evaluated. The duplex alloy contained 10 v/o ferrite and had a specific austenite-ferrite interfacial area of 930 cm2/cm3. The intergranular corrosion resistance of the alloy following 30 different aging treatments was determined in boiling acidified CU-CUSO4 solution (A262E), boiling 65% HNO3 (A262C), and by an electrolytic etch test in 10% oxalic acid (A262A). The pitting corrosion resistance was evaluated by potentiodynamic and galvanostatic tests in 0.1 N HCl at room temperature. The results of all five tests were compared to the results of slow extension rate stress corrosion cracking tests conducted at 550 F in high purity water containing 8 ppm dissolved oxygen. The microstructural features responsible for the corrosion in each test were determined. The A262E test was found to attack the chromium depleted zones in the material. The A262C test preferentially corroded the chromium depleted zones but also dissolved, albeit to a much less extent, the chromium rich phases such as M23C6. The A262A test primarily attacked chromium rich phases. Pitting corrosion was confined to the chromium depleted zones. The results of the A262E test and the galvanostatic pitting test agreed quite closely with the results of the stress corrosion cracking test. There was considerable scatter in the weight loss measurements obtained in the A262C test. The latter could detect the presence of a severely sensitized microstructure but was incapable of distinguishing between mildly sensitized and unsensitized materials. Although pitting was largely confined to the chromium depleted boundaries, unless the sample was severely sensitized, there was considerable scatter in the measured values of the pitting potential. This precluded the use of the latter as a parameter to characterize the degree of sensitization. The A262A test results exhibited the least correlation with the results of the stress corrosion cracking tests. Since the A262A test attacks chromium rich phases such as M23C6, samples which were aged long enough to intergranularly precipitate M23C6 but not long enough to produce chromium depleted zones of sufficient width or contiguity, nevertheless were severely corroded in the A262A test. And samples which were aged long enough to replenish the chromium depleted zones still failed the A262A test because of the continued presence of the chromium rich grain boundary carbides. These tests were also conducted on samples of sensitized and unsensitized austenitic 308 stainless steel. The results were consistent with the findings of the duplex alloy tests.


CORROSION ◽  
1986 ◽  
Vol 42 (2) ◽  
pp. 99-105 ◽  
Author(s):  
Y. S. Garud ◽  
A. R. McIlree

Abstract A logical approach to quantitative modeling of intergranular stress corrosion cracking (IGSCC) is presented. The approach is based on the supposition (supported partly by experimental and field observations, and by a related plausible underlying mechanism) that strain rate is a key variable. The approach is illustrated for the specific case of NiCrFe Alloy 600 in high-purity water. Model parameters are determined based on the constant stress IGSCC data (between 290 and 365 C) assuming a power law relation between the damage and the nominal strain rate. The model may be interpreted in terms of a film rupture mechanism of the corrosion process. The related mechanistic considerations are examined for the specific case. Resulting calculations and stress as well as temperature dependence are shown to be in good agreement with the data. More data are needed for further verification under specific conditions of interest.


Sign in / Sign up

Export Citation Format

Share Document