Because of its practical importance and various industrial applications, the process of subcooled flow boiling has attracted a lot of attention in the research community in the past. However, the existing models are primarily phenomenological and are based on correlating experimental data rather than on a first-principle analysis of the governing physical phenomena. Even though the mechanisms leading to critical heat flux (CHF) are very complex, the recent progress in the understanding of local phenomena of multiphase flow and heat transfer, combined with the development of mathematical models and advanced Computational Fluid Dynamics (CFD) methods, makes analytical predictions of CHF quite feasible. Various mechanisms leading to CHF in subcooled boiling have been investigated. A new model for the predictions of the onset of CHF has been developed. This new model has been coupled with the overall boiling channel model, numerically implemented in the CFX 4 computer code, tested and validated against the experimental data of Hino and Ueda. The predicted critical heat flux for various channel operating conditions shows good agreement with the measurements using the aforementioned closure laws for the various local phenomena governing nucleation and bubble departure from the wall. The observed differences are consistent with typical uncertainties associated with CHF data.

Because of the complexity of phenomena governing boiling heat transfer, the approach to solve practical problems has traditionally been based on experimental correlations rather than mechanistic models. The recent progress in computational fluid dynamics (CFD), combined with improved experimental techniques in two-phase flow and heat transfer, makes the use of rigorous physically-based models a realistic alternative to the current simplistic phenomenological approach. The objective of this paper is to present a new CFD model for critical heat flux (CHF) in low quality (in particular, in subcooled boiling) forced-convection flows in heated channels.

Thermal performance of heat flux controlled boiling heat exchangers are usually limited by the Critical Heat Flux (CHF) above which the heat transfer degrades quickly, possibly leading to heater overheating and destruction.

In the context of computational fluid dynamics (CFD), modelling low-pressure subcooled boiling flow is of particular significance. A review is provided in this book of the various numerical modelling approaches that have been adopted to handle subcooled boiling flow. The main focus in the analysis of such a challenging problem can be broadly classified according into two important categories: (i) Heat transfer and wall heat flux partitioning during subcooled boiling flow at the heated wall and (ii) Two-phase flow and bubble behaviours in the bulk subcooled flow away from the heated wall. For the first category, details of both empirical and mechanistic models that have been proposed in the literature are given. The enhancement in heat transfer during forced convective boiling attributed by the presence of both sliding and stationary bubbles, force balance model for bubble departure and bubble lift-off as well as the evaluation of bubble frequency based on fundamental theory depict the many improvements that have been introduced to the current mechanistic model of heat transfer and wall heat flux partitioning. For the second category, details of applications of various empirical relationships and mechanistic model such as population balance model to determine the local bubble diameter in the bulk subcooled liquid that have been employed in the literature are also given. A comparison of the predictions with experimental data is demonstrated. For the local case, the model considering population balance and improved wall heat partition shows good agreement with the experimentally measured radial distributions of the Sauter mean bubble diameter, void fraction, interfacial area concentration and liquid velocity profiles. Significant weakness prevails however over the vapor velocity distribution. For the axial case, good agreement is also achieved for the axial distributions of the Sauter mean bubble diameter, void fraction and interfacial area concentration profiles. The present model correctly represents the plateau at the initial boiling stages at upstream, typically found in low-pressure subcooled boiling flows, followed by the significant rise of the void fraction at downstream.