Nuclear Engineering ETDs

Publication Date

Spring 5-16-2026

Abstract

This work details the process of developing several new features into the MCNP6.3 Monte Carlo code for use in uncertainty quantification and reduction. These new capabilities, which include the CLUTCH method of calculating k-eigenvalue cross section sensitivities and the Windowed Multipole method of calculating cross sections, are implemented and verified against previous implementations and existing cross section data, respectively. These capabilities are then combined to produce sensitivities of k-eigenvalue to resonance parameters, which are verified against direct perturbation sensitivity estimates. The resonance parameters are calibrated using linear Bayesian methods and the accuracy of the resulting cross section is evaluated via the prior and posterior accuracy of benchmark simulations. The changes observed through this method were minimal; however, this work has developed a numerical framework that, with an extension to the number of benchmarks considered, can identify and resolve long-standing sources of uncertainty in nuclear data evaluations.

Keywords

Sensitivity analysis, R-Matrix, Windowed Mulitpole, Monte Carlo, Adjoint Methods

Sponsors

Consortium for Monitoring, Technology, and Verification (MTV) and the NNSA

Document Type

Dissertation

Language

English

Degree Name

Nuclear Engineering

Level of Degree

Doctoral

Department Name

Nuclear Engineering

First Committee Member (Chair)

Christopher Perfetti

Second Committee Member

Forrest Brown

Third Committee Member

Michael Rising

Fourth Committee Member

Mark Paris

Fifth Committee Member

Vladimir Sobes

Download

Share

COinS