Session: 3.4 - CFD for Nuclear Thermal Hydraulics

Paper Number: 157123

157123 - Multiscale Overlapping Domain Coupling for Thermal Hydraulics Simulations Within the BlueCRAB Code Suite 

Abstract: 

Enabling multiple approaches for modeling and simulating fluid flow and heat transfer for analyzing nuclear systems has been a research topic of great interest over the past decades. Such an effort is motivated by the broad spectrum of scales existing in both time and space that manifest in reactor core analysis, for which different numerical methods can have their own set of advantages. System Thermal-Hydraulic (STH) codes are a legacy within the community for evaluating safety-relevant transients in nuclear systems. These codes afford fast, whole-system analysis capabilities by employing a simplified description of the fluid flow based on correlations. However, multidimensional effects such as complex three-dimensional (3D) flow patterns, thermal stratification, and mixing in enclosures may violate the limits of most models. Currently, extending the STH codes for application involving these effects is an actual demand in the nuclear industry. The next generation of nuclear reactors typically consider non-conventional fluids, including, gas, sodium, and liquid metals, for which those effects are even more pronounced.

Alternatively, computational fluid dynamic (CFD) had proven to accurately predict coolant flow under these conditions. The present work introduces a novel coupling method between STH and CFD as a pathway to enable thermal hydraulics analysis of advanced reactors. We implement and validate the coupling method within the Comprehensive Reactor Analysis Bundle (BlueCRAB), a code suite in active development at Idaho National Laboratory for tailored for multi-physics analysis of advanced reactors.

Different from previous approaches, BlueCRAB implements an agnostic interface between STH – CFD while its coupling formulation is tailored for arbitrary geometry with varying inlets/outlets in coupled components. The coupling is established by overlapping STH – CFD simulation domains while explicitly exchanging data across their interface boundaries and within the interior of the STH components. Picard iterations are performed at each timestep to preserve continuity, momentum, and transport of scalars between coupled simulations. Here, we detail the implementation for incompressible flows. Results of an experiment that involves mixing between two systems connected by a double tee component demonstrate the accuracy of our approach. This setup is particularly interesting to validate the methodology while the 3D flow pattern in the connecting component dictates the split of scalars between the systems. Further, the model includes multiple inlets and outlets in the coupled section, which imposes non-trivial constraints to achieve stability while also conserving different physical quantities across the simulations. Ultimately, this work is the first to successfully couple both the Navier-Stokes and scalar transport equations in this problem class.

Presenting Author: Victor Coppo Leite Idaho National Laboratory

Presenting Author Biography: Dr. Victor Coppo Leite is a researcher at Idaho National Laboratory with primary interest in Computational Thermal-Hydraulics. Specifically, Leite's research activities include multi-scale coupling between different simulation scales and incorporating Machine Learning techniques into nuclear thermal-hydraulics.