SUMMARY
When a new computational transport code is developed, an important first step is for that code to be thoroughly tested in a variety of scenarios in order to verify the computational methods upon which it is constructed. The Coarse-Mesh Transport Method (COMET) code has been benchmarked and proven for traditional thermal systems, but has yet to be tested with a fast neutron design. In order to properly test a fast neutron spectrum, a large three dimensional full core model that is highly heterogeneous is required in order to represent a realistic reactor.A full-core benchmark problem based on Argonne National Lab's conceptual Advanced Burner Test Reactor (ABTR) has been developed, and then used to validate COMET against the Monte Carlo code MCNP5 (MCNP). This reactor is a sodium cooled fast reactor designed to burn recycled transuranics with the goal of generating energy while transmuting long term radioactive waste. In this stylized description of the ABTR, a certain degree of geometric simplifications were necessary to decrease the complexity of the core. All such alterations were made with care to preserve the physics as much as possible. Additionally, 15 group material cross sections for this core were developed using MoCSGen so that COMET and MCNP could be directly compared with identical cross sections. Three configurations of the benchmark core were developed to test different scenarios: All Rods Out (ARO), All Rods In (ARI), and Near Critical (NC).