Best estimate thermo-mechanical fuel modelling in transient analyses

The paper documents the second phase of systematic evaluation of FRAPTRAN+SUSA software package application on IFA650.3 LOCA experiment. A short comment to the first set of calculations has been maid and temperature dependent uncertain input data (boundary conditions) as well as time dependent results for all performed calculations and results of SUSA uncertainty analysis with two-sided tolerance at 95% probability and 95% confidence have been presented. The basic results of the set of the calculations are as follows: 1) uncertainty of thermal boundary conditions ±0.5% of nominal values produced maximum spread in clad hoop strain [from ∼0.064 through ∼0.092(nom. data) to ∼0.099], i.e. -53% to +107%, and clad effective elasto-plastic strain [from ∼0.060 through ∼0.113(nom. data) to ∼0.121], i.e. -70% to +108%; 2) these results are appropriate to last 15 - 20 sec. of the process before the cladding burst. It has been found that: 1) The fill gas pressure is the second most important parameter; 2) Temperature range of uncertainties has growing influence; 3) The BE method for LB LOCA modelling is applicable; 4) The processes with very high dynamics / instabilities create difficulties in frame of BE. The further goal is to perform numerical experiment for real PWR LB LOCA accident fuel rod response to test applicability of the developed methodological procedure