Description: With the elimination of Yucca Mountain as the long-term storage facility for spent nuclear fuel in the United States, a number of other storage options are being explored. Currently, used fuel is stored in dry-storage cask systems constructed of steel and concrete. It is likely that used fuel will continue to be stored at existing open-air storage sites for up to 100 years. This raises the possibility that the storage casks will be exposed to a salt-containing environment for the duration of their time in interim storage. Austenitic stainless steels, which are used to construct the canisters, are susceptible to stress corrosion cracking (SCC) in chloride-containing environments if a continuous aqueous film can be maintained on the surface and the material is under stress. Because steel sensitization in the canister welds is typically avoided by avoiding post-weld heat treatments, high residual stresses are present in the welds. While the environment history will play a key role in establishing the chemical conditions for cracking, weld residual stresses will have a strong influence on both crack initiation and propagation. It is often assumed for modeling purposes that weld residual stresses are tensile, high and constant through the weld. However, due to the strong dependence of crack growth rate on stress, this assumption may be overly conservative. In particular, the residual stresses become negative (compressive) at certain points in the weld. The ultimate goal of this research project is to develop a probabilistic model with quantified uncertainties for SCC failure in the dry storage casks. In this paper, the results of a study of the residual stresses, and their postulated effects on SCC behavior, in actual canister welds are presented. Progress on the development of the model is reported.
Date: August 1, 2013
Creator: Ballinger, Ronald G.; Ferry, Sara E.; Black, Bradley P. & Teysseyre, Sebastien P.
Item Type: Refine your search to only Article
Partner: UNT Libraries Government Documents Department