From experiments to DFT simulations : comprehensive overview of thermal scattering for neutron moderator materials

Thumbnail Image
Ramić, Kemal
Issue Date
Electronic thesis
Nuclear engineering
Research Projects
Organizational Units
Journal Issue
Alternative Title
For polyethylene both ARCS/SEQUOIA and VISION method were used to generate the new thermal scattering libraries. Overall VISION method performed better than SEQUOIA method, where the RPI DFT+oClimax library calculates best the double differential scattering and total cross section for polyethylene. The RPI polyethylene library performs better than the old ENDF/B-VII.1 library in calculation of Keff from a set of HEU-THERM benchmarks from the DICE library of benchmarks
Overall, the VISION method, incorporating DFT and oClimax, has shown to be the best method for generating new thermal scattering libraries for solid moderators. The most important advantages of the VISION method are: 1) calculation of partial contribution to GDOS, 2) phonon expansion (calculation of fundamental mode and multiple phonon scattering contributions), 3) scaling of frequencies to match the experimental data.
For quartz, in order to create the new library, changestoLEAPRmoduleofNJOY2016 were made. The phonon spectrum derived from oClimax S(Q,ω) was used to generate a new thermal scattering library. The RPI DFT+oClimax library improves the calculation of DDSCS and total cross section, while at the same time improving performance on the benchmarks from DICE.
Lucite was a unique material to perform the VISION method on due to the fact that lucite, unlike polyethylene, ice-1h and quartz, has no crystalline unit cell to be used with DFT. The mismatch in the locations of peaks, in the experimental S(Q,ω) and DFT-based oClimax calculated one, was fixed, and the new library improved upon the current lucite ENDF/B-VIII library. The RPI DFT+oClimax library calculates DDSCS and total cross section better, while performing significantly better on a set of benchmarks from DICE.
Due to the ravaging effects of global warming, nuclear energy is seen as a viable source of clean energy and as part of the solution; the accurate prediction of neutron thermalization in nuclear systems is of extremely high importance. Nuclear data plays an indispensable role in moving away from building costly and eco-unfriendly prototypes (from bombs to nuclear reactors) to the computer modeling of the nuclear phenomena. The calculation of neutron moderation, which is slowing down of neutrons to thermal energies where there is a greater probability that a neutron will cause a fission, is dependent on improvements in determination (measuring, calculating and validating) of the thermal scattering law of moderator materials. Due to the lack of experimental data for validation of current thermal scattering libraries for different moderators of interest for nuclear criticality safety purposes, the goal of this work was to perform inelastic neutron scattering experiments and use the obtained data as a guide in creation of new thermal scattering libraries. The experiments were performed at Spallation Neutron Source at Oak Ridge National Lab using two different types of spectrometers: indirect geometry (VISION) and direct geometry (ARCS and SEQUOIA) neutron spectrometers.
The data from ARCS and SEQUOIA spectrometers have been used in two manners: 1) the experimental data in form of double differential scattering cross section (DDSCS) measurements at different incident energies and scattering angles has been used for comparison and validation of thermal scattering libraries created with NJOY2016, where the ARCS experimental setup has been simulated using MCNP6.1; 2)the experimental data in form of S(Q,ω) has been transformed to generalized density of states (GDOS), which has been processed with NJOY2016 to create a thermal scattering library (ARCS/SEQUOIA method). The data from VISION spectrometer, because of the design of the spectrometer (“white” beam incident energy spectra, Q-dependence parallel to c-axis of the sample, and the momentum transfer proportionality to the energy transfer), was more suitable for transformation of S(Q,ω) to GDOS than the ARCS data. Additionally, using oClimax the output of the DFT calculations could be directly re-calculated to S(Q,ω) and comparedtotheobservedS(Q,ω)atVISIONspectrometer(VISIONmethod). Furthermore, oClimax offers additional advantages due to the capability of oClimax to calculate partial contributions of each atom to the whole molecule GDOS, while at the same time being capable of calculating the fundamental vibrational mode (n=1) and multiple phonon scatteringcontributions(n<1 overtones). Furthermore,oClimax enables scalingof frequenciesin S(Q,ω)spectra to correct th emismatch created by structural differences of the experimental sample and how the material is simulated using DFT.
For Ice-1h only the VISION method was used to create the thermal scattering library. The frequencies were scaled to properly match the frequencies in the experimental data, and the resulting library, although our group didn’t measure any DDSCS data at ARCS, improved upon DDSCS. The RPI library calculates the total cross section similar to the current ENDF/B-VIII library.
August 2018
School of Engineering
Full Citation
Rensselaer Polytechnic Institute, Troy, NY
PubMed ID