Nuclear Science and Engineering / Volume 177 / Number 2 / June 2014 / Pages 141-155
Technical Paper / dx.doi.org/10.13182/NSE13-4
Articles are hosted by Taylor and Francis Online.
Grid spacers within nuclear fuel assemblies play a critical role in fuel performance and contribute to safety margins by enhancing the margin to the critical heat flux. The Organisation for Economic Co-operation and Development/Nuclear Energy Agency has organized a computational benchmark wherein the prediction of flows and turbulence downstream of a mixing-type grid spacer are examined. Studies performed by McMaster University using STAR-CCM+ for the final submission to this MATiS-H blind benchmark exercise related to inter-subchannel mixing and turbulence are presented in this paper. The rationale behind the choice of the computational scheme along with comparisons of the submitted results to the experiments is reported. The goal at the outset of the study was to obtain a reasonably accurate solution with a minimum number of nodes and appropriate turbulence models such that the results would be relevant for engineering applications that include property variations and heat transfer. As such, advanced modeling methods such as large eddy simulation and unsteady Reynolds-averaged Navier-Stokes (URANS) were not included within the scope of the models tested. However, URANS was used to study some specific separate-effect flow features within the grid spacer, and these tests were compared to their steady counterparts.
A comprehensive separate-effect study was performed first in order to finalize the computational scheme for the submission. Several partial geometries were studied for steady and unsteady behavior as well as for mesh sensitivity, turbulence, and wall modeling effects. A series of successively more complex simulations, sometimes involving unsteady modeling, was performed up to and including a study of similar 5 × 5 rod bundle geometry reported in the literature. The final submission results are presented in the paper and are compared with the benchmark data that have recently been released.