| Abstract Scope |
Although 347H austenitic stainless steel provides good corrosion resistance in high-temperature molten salts, high susceptibility to stress relaxation cracking (SRC) has been reported in the 347H welds made with matching filler in molten salt storage tanks. Alternative filler metals and post-weld heat treatments (PWHT) have been evaluated as mitigation solutions for SRC. However, the subsequent effect of these strategies on weld region corrosion kinetics remains unclear. This study evaluates the localized corrosion rates and reaction layer of alternative 347H weldments across variations in welding process, filler metal chemistry, and PWHT condition, comparing results against an unwelded 347H base metal and matching filler controls. Isothermal immersion testing was performed at 565°C in a binary molten nitrate salt (60% NaNO3 – 40% KNO3) for up to 2000 hours to establish mass-loss kinetics and track oxide layer evolution. Gravimetric analysis quantified degradation rates, while high-resolution scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) characterized the thickness, composition, and stability of the reaction layers across the fusion zone interfaces. Results show that while filler metal selection and welding process parameters alter localized corrosion resistance, the application of PWHT exerts the overriding influence on reaction layer kinetics and oxide stability, especially during early exposure (up to 500 hours). For instance, a maximum corrosion rate of 64.3 µm/yr was observed in the 347 matching filler weld after PWHT at 250 hours of exposure, in comparison to 2.6 µm/yr for the as-welded matching filler control for shielded metal arc welding (SMAW). In comparison, as-welded 16.8.2 filler SMAW exhibited a corrosion rate of 10.6 µm/yr. Overall, the weldment variations exhibited higher corrosion rates compared to both the 347H base metal and matching filler controls. However, these performance differences tend to diminish as exposure time exceeds 1000 hours. |