基于 preCICE 的多尺度耦合不确定性分析方法研究

Background]Current research on multi-scale coupling primarily focuses on code development and verification, while systematic studies addressing uncertainties in coupled codes remain scarce. [Purpose] This study constructs a multi-scale coupled uncertainty analysis code based on the open-source coupling framework preCICE, integrating the CFD code FLUENT, the subchannel code SUBCHANFLOW, and the uncertainty quantification module DAKOTA for uncertainty quantification in multi-physics coupled simulations. [Methods] This study conducts numerical verification under steady-state and transient conditions through establishing a 33 rod bundle model, and implements uncertainty quantification and sensitivity analysis for multi-physics coupling systems. [Results] Experimental data demonstrate that: (1) The axial temperature distribution of the coupled system under steady-state conditions closely matches the computational results from single-code simulations; (2) In transient verification, outlet mass flow rate fluctuations exhibit perfect synchronization in both period and phase under sinusoidal perturbation of inlet mass flow rate conditions; (3) Uncertainty quantification reveals that key parameters including coolant temperature and peak cladding temperature follow a normal distribution pattern; (4) Sensitivity analysis identifies inlet mass flow rate, outlet pressure, inlet temperature, and fuel rod heat flux as dominant factors governing system responses.[Conclusions] The findings validate the response reliability of the multi-scale coupled system under dynamic operational conditions, providing critical technical support for the application of multi-physics coupling methodologies in nuclear safety analysis.