Award Date


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Mechanical Engineering

First Committee Member

William Culbreth

Second Committee Member

Alexis Barsilov

Third Committee Member

Robert Boehm

Fourth Committee Member

Ralf Sudowe

Fifth Committee Member

Gary Cerefice

Number of Pages



There has been a reduction in funding for theoretical and applied research for improving the nation's database of continuous neutron cross-sections at BNL-NNDC. From 1940 through the late 1980s, research and applied development produced volumes of reliable neutron continuous cross-sections for many isotopes. Currently, the cross-section work has been mainly computational. The focus of this research is mainly centered on the requirements for improving thermal cross-sections to support reactor operations and fuel storage. The research efforts will also helpfully aid in the fast fission spectrum in order to support fast reactor designs for improving safety analysis and feedback coefficients.

This previous level of effort produced well-defined sets of neutron cross-sections for a few applications, but only incidental information for outside of the desired range of the energy spectrum. Isotope cross-sections are generally characterized from the isotope's fission energy to the thermal absorption energy, but not in as much detail outside of this roughly 0-3 MeV range. The good news is that each version of ENDF has improved the data points between the resolved resonance and hard sphere ranges.

The amount of characterization for an isotope’s cross-section is determined by the number of experiments conducted. The sensitivity is then assigned to an isotope is after evaluation of the series of experiments are conducted and analyzed. If two isotopes were in a given experiment, then the cross-sections from the two isotopes will need to be separated or de-convoluted. Obtaining well-defined neutron cross-sections means that the sensitivity should be on the same order of magnitude through the entire range of the energy spectrum. This will require pure isotope foils for many of the experiments that will require de-convoluting the multiple cross-sections. Many observations will be required to uniformly reduce the sensitivity across the energy spectrum for each isotope.

The scope of this dissertation is to ascertain the quality of the uncertainties in the NNDC database for U-234, U-235, U-236, and U-238. This will aid in determining where the uncertainties are large and require additional future research. A comparison will be conducted to analyze the sensitivities through computational methodology, based on Monte Carlo particle transport code (MCNP6) with modern continuous-energy neutron data libraries (ENDF/B-VII.1). The Godiva HMF-001 and Godiver the water-reflected sphere HMF-004 criticality safety evaluation benchmark models will be used in the analysis.

A created Fortran program will modify the ENDF/B-VII cross-sections, and then standard Type I ACE cross-sections files will be generated using NJOY2012.50. By applying the perturbation to the cross-sections, in increasing increments, the change in Keff will be calculated in MCNP6 and MCNPX, and thus the amount of sensitivity can be determined when Keff varies from the baseline. The difference in Keff can be determined by modeling the two criticality benchmark experiments, through computationally-based analysis.

The perturbations applied to the two benchmarks will demonstrate the effect of the change in sensitivity on the overall bias of the safety analysis. In order to develop safe and reliable designs for fast fission reactors, transmutation and accelerator driven systems, additional research work will need to be conducted to improve the sensitivity for a larger range of energies. This will be coupled with work in uncertainty and feedback coefficients for NRC safety requirements.

This dissertation is to aid in showing where there are gaps in the sensitivities in the isotopes, as well as any correlations between the sensitivity and the change in Keff. The long-term objective is that the developed data will allow quantifying models in a more rigorous manner, for the neutron cross-section related sensitivities in the calculated effective multiplication factor. This in turn should make a drive for an overall strategy for independent quality checks in criticality and reactivity analysis. This will improve the scientific basis for criticality safety analysis and increase the quality of the predictive capabilities, a necessary requirement in the possibility of establishing less conservative but more reliable regulatory safety criteria.


benchmark; criticality; cross-sections; MCNP; NJOY; perturb


Computer Engineering | Electrical and Computer Engineering | Nuclear | Nuclear Engineering