Parametric Uncertainty Quantification for Nuclear Reactor Depletion Calculations: An Exercise in Computational Sciences
In the technical portion of this talk, I will present an adjoint-based methodology for performing uncertainty quantification and sensitivity analysis in calculations that describe the evolution of materials over the life cycle of a nuclear reactor. I’ll demonstrate that, like so many projects being pursued by the CSGF and DOE communities, this study required a multidisciplinary approach: Mathematics was required to demonstrate existence of a solution; physics was required to produce accurate solutions; computer science was required to address the challenge of solving memory-intensive equations on memory-limited architectures.
During the talk, I also will outline my journey before, during, and since involvement as a CSGF fellow. I’d like to share with the group the value of diversification the Program of Study and practicum experience brought to my graduate school experience. I’ll also try to give some perspectives from industry on the importance of computational sciences in today’s aggressive intellectual property landscape.