Chance, Christopher (2014-12). Partial Core Blockage Simulation Using COBRA-TF. Master's Thesis.
COBRA--TF (CTF) was used to simulate the cooling capabilities for two reactor types that undergo different blockage scenarios. One case considered is a pool-type TRIGA reactor in which instrumentation tubes were inserted into existing coolant channels. The inclusion of the instrumentation tubes reduces the available coolant flow area, requiring an investigation of fuel coolability. The second case considered a pressurized water reactor (PWR) that has experienced a loss of coolant accident. Debris not stopped by the sump strainer can be deposited at the inlet of the core, creating blockage at the inlet of select assemblies. The blockage at the inlets could create a lack of cooling capability for the blocked assemblies and could result in fuel failure. CTF has not been widely used for testing these scenarios and the results should be compared alongside other simulations performed in STAR-CCM+ and RELAP5-3D.
. For the TRIGA reactor simulations, limitations of CTF were discovered such as a lack of natural convection correlations and setbacks with transient simulations. Therefore, steady-state simulations were created based on input data from the Safety Analysis Report for the reactor and data obtained by STAR-CCM+ models of the core. Single pin analyses were conducted on 3 different fuel rods in order to determine if the fuel rods were at risk of overheating from the proposed modifications with the instrumentation tubes. Larger simulations that modeled two different 4x4 arrays of fuel rods within the TRIGA core were used to determine how well the coolant traversed through these areas with the instrumentation tubes inserted. The results for all simulations were compared to STAR-COM+ results from similar models and it was determined that the insertion of the instrumentation tubes in the TRIGA reactor would not be a concern for the safety of the reactor under normal operating conditions.
The PWR simulations revealed more limitations of CTF and did not prove fruitful in providing reliable results. The original goal was to employ CTF to model a PWR core in full detail in order to compare the results to the RELAP5-3D model that used a lumped approach. However, we encountered difficulties with CTF, mostly due to issues with transient simulations and parallel processing. Significant improvements were made to CTF over the course of this research and have resulted in a more robust and bug-free version of the code. However, CTF is still being developed and has some remaining hurdles to overcome before it can be used to reliably model the simulations that were planned. This research still continues today and CTF is being used and improved upon to become a more versatile sub-channel analysis code.