Prediction of Boiling and Critical Heat Flux Using an Eulerian Multiphase Boiling Model

Abstract

The present paper concerns the development and validation of an Eulerian multiphase boiling model to predict boiling and critical heat flux within the general-purpose computational fluid dynamics (CFD) solver FLUENT. The governing equations solved are generalized phase continuity, momentum and energy equations. Turbulence effects are accounted for using mixture, dispersed or per-phase multiphase turbulence models. Wall boiling phenomena are modeled using the baseline mechanistic nucleate boiling model, developed in Rensselaer Polytechnic Institute (RPI). Modifications have been introduced to the quenching heat flux model to achieve mesh-independent solutions. The influences of boiling model parameters have also been systematically investigated. To model non-equilibrium boiling and critical heat flux, the PRI model is extended to the departure from nucleate boiling (DNB) by partitioning wall heat flux to both liquid and vapor phases and considering the existence of thin liquid wall film. Topological functions are introduced to consider the wall boiling regime transition from the nucleate boiling to critical heat flux (CHF), and the corresponding flow regime change from bubbly flows to mist flows. A range of sub-models are implemented to model the interfacial momentum, mass and heat transfer and turbulence-bubble interactions. To validate the Eulerian multiphase boiling model, it has been used to predict nucleating boiling and critical heat flux in a range of 2D and 3D boiling flows. The examples presented in the paper include: (1). Nucleate boiling of sub-cooled water in an upward heated pipe; (2) R113 liquid flows through a vertical annulus with internal heated walls; (3). 3D boiling flows in a rectangular-sectioned duct; and (4). Critical heat flux and post dryout in vertical pipes. The results demonstrate that the model is able to predict reasonably well the distributions of wall temperature, the bulk fluid sub-cooling temperature and cross-sectional averaged vapor volume fraction in the vertical pipe. The computed profiles of the vapor volume fraction, liquid temperature, and the liquid and vapor velocity profiles are generally in good agreement with available experiments in the 2D annular case. In the 3D rectangular duct, the cross-sectional averaged vapor volume fractions are well captured in all the ten cases under investigation. In the case of critical heat flux and post dryout, the model is also able to predict reasonably well the location and the temperature rise under critical heat flux conditions. The computed wall temperature distributions along the pipes are in overall good agreement with available experiments.