Abstract :
As part of the task of assessing and qualify the frozen version of PennState University TRAC-BF1/v2001.2 as a BWR LOCA code, in this work, we shall study the extent to which changes in physical models in the code can affect the predicted rod surface temperature histories during separate-effect tests and system calculations. In particular, we shall concentrate our attention to the reflooding phase and after first demonstrating by two examples the necessity of using the code with the PSI updates, we shall implement in the code a dispersed flow interfacial shear model which is consistent with the corresponding interfacial heat transfer model, i.e. it utilizes the interfacial area per unit volume for both closure relations and not only for the interfacial heat transfer. For this, we shall implement in the code the dispersed flow interfacial shear model of RELAP5/MOD3; furthermore, we shall assume that the average droplet diameter is a function of the distance from the bottom quench front. Having done this, we shall analyze one FLECHT bottom flooding test with both code versions and show that the differences between measured and predicted PCTs are for both cases approximately the same. Finally, we shall study the effect of the number of nodes (8, 24 and 48) in the core by analyzing a hypothetical LB-LOCA in a BWR with limited ECC availability both with the frozen and the modified versions of the code and we shall show that when the frozen version of the code is used, the predicted peak clad temperatures (PCTs) are 120 K lower if 48 rather than when 8 nodes are used in the active core.
