Numerical study of vertical subcooled boiling flow with new wall nucleation models

Correct modeling of subcooled boiling flow has generated significant interest considering its use in a large number of industrial applications, including nuclear power plants. Due to the decrease in computational cost and the limitations of experimental studies, CFD simulation is a proven, powerful tool in multiphase modeling. Understanding and predicting the process of wall nucleation are of particular significance in numerical simulation of subcooled boiling flow, since wall nucleation parameters, such as bubble departure diameter and bubble departure frequency, partially constitute the boundary conditions of gas phase. However, in previous work, sensitivity study of wall nucleation models and validation of predicted wall nucleation parameters were not sufficient. In this work, Eulerian-Eulerian multiphase model and RPI wall boiling model were adopted to simulate the upward forced convective subcooled boiling flow in a vertical annulus using the commercial CFD code, Fluent 18.2. The flow behaviors were quantitatively predicted at three different pressures and the results were compared with available experimental data. Newly proposed wall nucleation models were used to replace the built-in models of Fluent 18.2. At each pressure, the impacts of bubble size modeling, inlet liquid temperature and interfacial heat transfer modeling were studied. It was found that significant under-predictions of void fraction resulted from implementing new wall nucleation models. Reasonable prediction of gas velocity could only be obtained at Port 3 at high pressures. With a specific combination of built-in models, reasonable results of void fraction distribution could be obtained. However, to provide physical predictions of void fraction profiles, unphysically large wall nucleation parameters must be assumed. Hence, it could be deduced that the contribution of wall nucleation to the volume fraction of gas was overemphasized in simulation. It was also found that new wall nucleation models performed better at low pressure. The other parameters impacted the results similarly at different pressures: a) bubble size impacted the void fraction distribution mainly by affecting the volumetric heat transfer rate and thus volumetric mass transfer rate; b) change of inlet liquid temperature only led to growth or drop in the magnitude of void fraction profiles; c) interfacial heat transfer was not differentiated between evaporation and condensation, so various leading coefficients of the heat transfer model affected liquid-to-gas and gas-to-liquid mass transfers simultaneously. Considering the requirement of unphysically large evaporation heat flux to obtain reasonable predictions, inaccurate modeling of bubble growth might be an important cause of under-prediction of void fraction. A mesh refinement study was attempted to identify if the discrepancy was due to a coarse mesh. Higher mesh resolution did help to enhance the prediction, but the improvement was limited, still leaving a large gap between simulations and experimental results. Further improvements of bubble size modeling, heat transfer modeling and near-wall treatment are suggested to increase the accuracy of prediction.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Nuclear, Plasma, Radiolgc Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2019

Author dc:creator

• Wang, Longcong

Contributors dc:contributor

• Brooks, Caleb S.
• Kozlowski, Tomasz

• CFD
• subcooled boiling
• wall nucleation
• annlus

Statement dc:rights

• Copyright 2018 Longcong WANG

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/102834

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/102834