On the Use of Large Eddy Simulation and Direct Numerical Simulation to Improve Near-Wall Modeling in Porous Media Models of Pebble Bed Reactors
Restricted (Penn State Only)
- Author:
- Reger, David
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 22, 2024
- Committee Members:
- Jon Schwantes, Program Head/Chair
Elia Merzari, Chair & Dissertation Advisor
Paolo Balestra, Special Member
Saya Lee, Major Field Member
Bladimir Ramos Alvarado, Outside Unit & Field Member
William Walters, Special Member
Xing Wang, Major Field Member - Keywords:
- Pebble Bed Reactor
Wall Chanelling
Porous Media
Computational Fluid Dynamics
Large Eddy Simulation
Direct Numerical Simulation
Correlation Devleopment
Pressure Drop
Interphase Heat Transfer - Abstract:
- This work aims to improve capabilities for modeling localized effects in porous media models of Pebble Bed Reactors (PBRs). The wall-channel effect is the primary local phenomena of interest in a PBR, where the presence of the reflector wall disrupts the pebble packing, causing the pebbles near the wall to pack less efficiently and creating large void regions. Accurate modeling of the near-wall region is important as it will affect core bypass flow and temperature predictions. Failing to properly model this region therefore leads to increased uncertainty in the prediction of maximum fuel temperatures and thus larger safety margins are necessary. Porous media models are commonly used for design scoping and plant-level simulations of PBRs. Although these models have some capabilities to model the near-wall region, the correlations that are available in porous media codes are often inaccurate when a multi-region model is used to discretize the near-wall region. A high-to-low methodology is therefore developed in this work to improve local modeling capabilities in porous media codes. NekRS, a spectral-element computational fluid dynamics (CFD) code, is used to generate a large high-fidelity flow dataset with Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS). An analysis of the Turbulent Kinetic Energy (TKE) budget is performed, revealing notable regions of negative TKE production near pebble contact points. Further investigation reveals that the amount of negative production in a given region can be linked to the local porosity of that region. This investigation of the flow physics helps to provide a better understanding of the differences in flow behavior across the various regions of the bed. Porous media models equivalent to the high-fidelity models are created and simulated with Idaho National Laboratory's Pronghorn porous media code. The KTA correlation is used for the pressure drop calculation in Pronghorn and is the correlation that is targeted for improvement. Results between NekRS and Pronghorn are compared to determine areas of discrepancy between the two codes. A method is then developed to extract the form loss coefficients from the high-fidelity dataset where the bed is split into several concentric ring subdomains and flow data is averaged in each region. An improved pressure drop correlation is determined with the extracted form coefficients and is implemented in Pronghorn, reducing the error in the near-wall velocity prediction from greater than 30\% to less than 5\%. A similar high-to-low investigation is performed on the interphase heat transfer closure. Several correlations are compared to the high-fidelity results where it is found that the KTA heat transfer correlation is capable of accurately predicting the local Nusselt numbers that were determined in the high-fidelity simulation. Comparison of the radial solid temperature profiles, however, reveal discrepancies between NekRS and Pronghorn. It is discovered that the implementation of the interphase heat transfer coefficient that exists in many current porous media codes is not valid when local porosities are modeled. Instead, it is suggested that the interphase heat transfer coefficient should be dependent on the local porosity, the Nusselt number, and the local solid surface-to-volume ratio. Implementation of this change produces improvement in the agreement between NekRS and Pronghorn while using the KTA heat transfer correlation.