Abstract: [Background] The rod bundle is the most common geometric structure of nuclear fuel assembly which is the core component of nuclear reactor. Computational fluid dynamics (CFD), as an auxiliary tool, is widely used in the flow analysis of rod bundles, but there is no unanimous CFD application method, nor the best practices guideline. As a result, the CFD simulation credibility is controversial. Meanwhile, CFD analysis is based on the certain geometry and boundary conditions, while large uncertainty exists in the physical conditions, which will challenge the credibility of the simulation results. [Purpose] This study aims to find a new method to do the uncertainty quantification of the CFD simulation for the flow analysis of rod bundles. [Methods] Considering the fact that traditional uncertain analysis methods, such as Monte Carlo sampling, are not applicable to the uncertainty analysis of CFD because of its huge amount of calculation, the first step was to sum up the best practice method of CFD on the single-phase rod bundle flow, based on the analysis of grid, turbulence model and wall treatment. Thereafter, the sensitivity coefficient method was used to analyze the influence of wall function uncertainty in the prediction of pressure drop. [Results] The simulation results show that the forecast uncertainty for pressure drop is 3.2%. [Conclusion] The best practice method of CFD combined with the sensitivity coefficient method can be used to analyze the uncertainty quantification of the CFD simulation the flow analysis of rod bundles.

Key words: Rod bundle flow, CFD, Uncertainty quantification