The Woodruff School

SUMMARY

The increase in designing high temperature gas reactors, large-scale and microreactors, led to a boost in demand for validation of current tools for analyzing these types of reactors. There are many challenges related to the multiphysics nature of the modeling and simulation challenges. From a neutronic perspective, the double-heterogeneity of tri-structural isotropic particle (TRISO) fuel and particle-based burnable poisons provides challenges in modelling these types of reactors. Special treatment must be taken in deterministic codes and the large number of particles can cause a large computational cost in Monte Carlo codes. Considerable effort has been made in benchmarking codes on high temperature gas reactors. Ultra Safe Nuclear Corporation (USNC) has made a tool called RAPTOR that is meant to act as a nuclear reactor design and analysis framework, and currently is used in generating inputs to the Serpent Monte Carlo code and also handles running the input and extracting the output information. This thesis investigated modeling a representative supercell of the MHTGR with RAPTOR as specified in the MHTGR 350 MWth benchmark and compared the results to published results from the benchmark. The full MHTGR 350 MW core was also modeled with RAPTOR according to the core design in the NEA MHTGR 350 MWth benchmark. These results were then compared to a handmade Serpent model of the full core. Cross section generation was also explored for this reactor type by generating cross sections on individual fuel blocks and supercells. Different methodologies were explored in the generation of assembly discontinuity factors. These cross sections were then used in a diffusion code and the results were compared to the full reference Serpent case.

SUBJECT:	M.S. Thesis Presentation

BY:	Christopher Wells

TIME:	Thursday, April 18, 2024, 9:00 a.m.

PLACE:	Boggs, 3-47

TITLE:	Validation of RAPTOR and Analysis of Cross Section Generation Techniques with the MHTGR 350MWth Core

COMMITTEE:	Dr. Dan Kotlyar, Chair (ME/NRE) Dr. Steven Biegalski (ME/NRE) Dr. Andrew Johnson (USNC)

SUMMARY The increase in designing high temperature gas reactors, large-scale and microreactors, led to a boost in demand for validation of current tools for analyzing these types of reactors. There are many challenges related to the multiphysics nature of the modeling and simulation challenges. From a neutronic perspective, the double-heterogeneity of tri-structural isotropic particle (TRISO) fuel and particle-based burnable poisons provides challenges in modelling these types of reactors. Special treatment must be taken in deterministic codes and the large number of particles can cause a large computational cost in Monte Carlo codes. Considerable effort has been made in benchmarking codes on high temperature gas reactors. Ultra Safe Nuclear Corporation (USNC) has made a tool called RAPTOR that is meant to act as a nuclear reactor design and analysis framework, and currently is used in generating inputs to the Serpent Monte Carlo code and also handles running the input and extracting the output information. This thesis investigated modeling a representative supercell of the MHTGR with RAPTOR as specified in the MHTGR 350 MWth benchmark and compared the results to published results from the benchmark. The full MHTGR 350 MW core was also modeled with RAPTOR according to the core design in the NEA MHTGR 350 MWth benchmark. These results were then compared to a handmade Serpent model of the full core. Cross section generation was also explored for this reactor type by generating cross sections on individual fuel blocks and supercells. Different methodologies were explored in the generation of assembly discontinuity factors. These cross sections were then used in a diffusion code and the results were compared to the full reference Serpent case.