AN AUTOMATED CROSS SECTION PARAMETRIZATION METHOD FOR MTR APPLICATION

The advent of more advanced MTR (Material Test Reactor) fuel and reactor designs necessitate a corresponding improvement in the cross-section representation models applied to few group homogenized diffusion parameters. This problem may be decomposed into the construction of an appropriate cross-section representation model (the definition of relevant state parameters and associated basis functions), and the subsequent determination of approximation coefficients. In this work a sparse grid - based global polynomial cross-section parameterization method is presented which addresses both these aspects. A technique known as analysis of variance (ANOVA) is applied to construct the model and an efficient sparse grid integration scheme is implemented to determine the expansion coefficients. The method has a number of advantages, such as: built-in error control; consistent treatment and importance evaluation of all state parameters, including burn-up; and identification and approximation of all cross-section dependencies, including cross-terms. The method is applied to different MTR fuel designs and compared to more traditional MTR parameterization methods. The usefulness of the method as both model buildingand few-group, cross-section reconstruction tool, is illustrated.