| description abstract | A method for evaluating crack growth by repeatedly modeling and analyzing the transitional crack shapes is developed for a general computing environment in which a commercial finite element preprocessor and analysis code are used. The proposed method calculates stress intensity factors (SIFs) by finite element analysis (FEA) by directly distributing estimated weld residual stress obtained from noncracked components on the crack surface on the basis of the superposition principle. In present case, to specify a nonuniform residual stress distribution, a subroutine for a commercial FEA code (ABAQUS) was developed. Arbitrary crack shapes during the crack propagation were expressed by applying the submodeling technique which allowed arbitrary crack shapes to be meshed. The sequence of steps in the proposed method was designed to make it possible to consider complicated stress distributions, such as weld residual stress, and to express arbitrary crack shapes. The applicability of the proposed FEA based method was verified by comparing the result of a stress corrosion cracking (SCC) growth analysis results of a flat plate obtained with the proposed method and with the ASME code procedure. As an application example, the SCC growth behavior of a crack at the bottom of a nuclear reactor pressure vessel (RPV) involving a dissimilar metal weld and a unique geometry was evaluated by the proposed method. The evaluation results were compared with results obtained using a conventional method, i.e., the influence function method (IFM). Since both sets of results were in reasonable agreement, it was concluded that IFM can be applied to this case. Previously, it was difficult to assess the applicability of conventional methods, such as the code procedure and IFM, to a complicated problem because of the existence of complicated residual stress fields, dissimilar metals, and the complicated crack shapes involved. The proposed method using FEA allows the applicability of conventional methods to complicated crack growth evaluations to be assessed. | |
