Mattingly’s paper “An Evaluation of Monte Carlo Simulations of Neutron Multiplicity Measurements of Plutonium Metal,” co-authored by University of Michigan Nuclear Engineering Ph.D. student Eric Miller, demonstrated that the published value of the mean neutron multiplicity (nu-bar) of induced fission in 239Pu is significantly lower than it’s actual value. Mattingly and Miller made this discovery by comparing Monte Carlo simulations of subcritical neutron multiplicity counting experiments to measurements Mattingly performed at the Nevada Test Site in 2009. Their findings are of potentially substantial impact – they imply that current models of fission chain-reactions in plutonium significantly underpredict neutron multiplication.
Eric Miller graduated with his Ph.D. later in 2012; he now works for the Johns Hopkins University Applied Physics Laboratory.
Mattingly and Miller’s original work has formed the basis for continuing RADIANS research on sensitivity analysis, uncertainty quantification, and parameter estimation conducted by Ph.D. students Sean O’Brien and Alex Clark and postdoctoral scholar Todd Evans. We have developed new methods for nuclear data evaluation using subcritical neutron multiplicity counting experiments, and we are very close to demonstrating, quantitatively and irrefutably, that the published value of nu-bar for 239Pu needs to be reevaluated.