Coupling of Wall Boiling with Discrete Population Balance Model

Abstract

A coupling between a polydisperse population balance method (Multiple Size Group Model – MUSIG) and the RPI wall boiling model for nucleate subcooled boiling has been implemented in ANSYS CFX. It allows more accurate prediction of the interfacial area density for mass, momentum and energy transfer between phases in comparison to the usual local-monodisperse bubble size assumption and underlying bulk bubble diameter correlations as they are commonly used in boiling flow applications like e.g. the prediction of subcooled nucleate boiling in rod bundles and fuel assemblies of PWR. The paper outlines the methodology of the coupled CFD model, which automatically avoids possible inconsistencies in the model formulation for the heated wall, when the generated steam bubbles on the heater surface are injected exactly in the bubble size class corresponding to the predicted bubble departure diameter. The coupling of the RPI wall boiling model and the MUSIG model has been implemented for both homogenous/inhomogeneous variants of the MUSIG model. The paper presents the validation of the coupled modeling approach for the well known test case of nucleate subcooled boiling of R113 refrigerant in a circular annulus with inner heated rod based on the experiments of Roy et al. [1]. ANSYS CFX results with the newly implemented approach as well as comparison to data and locally-monodisperse simulations are provided.