American Nuclear Society
Home

Home / Publications / Journals / Nuclear Technology / Volume 206 / Number 10

Coupling of the Smoothed Particle Hydrodynamic Code Neutrino and the Risk Analysis Virtual Environment for Particle Spacing Optimization

Emerald D. Ryan, Chad L. Pope

Nuclear Technology / Volume 206 / Number 10 / October 2020 / Pages 1506-1516

Technical Paper / dx.doi.org/10.1080/00295450.2019.1704576

Received:July 24, 2019
Accepted:December 10, 2019
Published:October 16, 2020

Flooding is a hazard for nuclear power plants (NPPs) and has caused extensive damage and economic impact. Improved NPP flooding risk characterization starts with improving scenario realism by using physics-based flooding simulations. Smoothed particle hydrodynamics (SPH) is one method for modeling fluid flow and is being investigated for NPP flooding simulation. While still in its infancy as a fluid simulation tool, SPH offers enticing features especially in three-dimensional modeling. However, when conducting SPH simulations, users must establish, inter alia, the appropriate particle spacing, which can be a tedious and time-consuming process. This paper describes the coupling of the SPH code Neutrino and the Idaho National Laboratory developed Risk Analysis Virtual Environment (RAVEN). By coupling Neutrino and RAVEN, the RAVEN optimization capabilities can now be applied to the particle spacing selection problem. A brief description of SPH, the overall capabilities of RAVEN, and the protocol used to couple the codes are provided. Additionally, the paper details a hypothetical problem and demonstrates the ability of automating the particle spacing selection and performing an example particle spacing optimization using RAVEN. With the Neutrino/RAVEN coupling established, a wide range of capabilities can now be utilized including optimization, reduced order model training and analysis, uncertainty quantification, sensitivity analysis, etc. Previously, these capabilities would require extensive work and time from the Neutrino user. Now, these capabilities are readily available and require only the creation of a RAVEN input file.