Benchmarking Thermal Neutron Scattering in Graphite

Abstract

The Very High Temperature Reactor (VHTR), one of the Generation IV reactor concepts, is a helium-cooled, graphite-moderated nuclear reactor with a core temperature reaching 1000°C. It can provide high quality process heat for hydrogen production beside power generation and will become deployable around 2030. At such temperature, graphite is an appropriate neutron moderator material due to its high sublimation temperature and high temperature strength. Furthermore, graphite has a large heat capacity and stable structure due to its large thermal inertia. However, the current thermal neutron cross-section libraries of graphite are based on models and data developed in the 1950s and 1960s. Significant discrepancies between measurements and the computational predictions of these libraries were observed. As a result, a study was performed in this dissertation to benchmark modern and traditional thermal neutron scattering libraries of graphite.

In this work, a Slowing-Down-Time experiment was designed and performed at the Oak Ridge National Laboratory (ORNL) by using the Oak Ridge Electron Linear Accelerator (ORELA) as a neutron source to study the neutron thermalization in graphite at room and higher temperatures. The MCNP5 code was utilized to simulate the detector responses and help optimize the experimental design including the size of the graphite assembly, furnace, shielding system and detector position. To facilitate such calculation, MCNP5 version 1.30 was modified to enable perturbation calculation using point detector type tallies. By using the modified MCNP5 code, the sensitivity of the experimental models to the graphite total thermal neutron cross-sections was studied to optimize the design of the experiment. Measurements of slowing-down-time spectrum in graphite were performed at room temperature for a 70x70x70 cm graphite pile by using a Li-6 scintillator and a U-235 fission counter at different locations. The measurements were directly compared to the Monte Carlo simulations that use different graphite thermal neutron scattering cross-section libraries. Simulations based on the ENDF/B-VI graphite library were found to have a 30%-40% disagreement with the measurements.

In addition to the graphite SDT experiment, which provided the data in the energy region above the graphite Bragg-cutoff energy, transmission experiments were performed for different types of graphite samples using the NIST 8.9 Å beam (located at NG-6) to investigating the energy region below the Bragg-cutoff energy. Measurements confirmed that reactor grade graphite, which is a two phase material (crystalline graphite and amorphous carbon), has different thermal neutron scattering cross section from pyrolytic graphite (crystalline graphite).

The experiments presented in this work compliment each other and provide an experimental data set which can be used to benchmark graphite thermal neutron scattering cross section libraries that are generated using different methodologies. Further investigation is necessary.