Advanced mechanistic model for critical heat flux prediction in subcooled flow boiling

Abstract

Through the extensive reviews on the prediction of critical heat flux (CHF), which is characterized as an apparent rise of the wall temperature due to a sharp reduction of the local heat transfer coefficient, it is recognized that the fundamental understanding for the CHF phenomenon at low qualities (or subcooled flow boiling) is unsatisfactory compared with that at high qualities. Therefore the mechanistic modeling of the former is naturually required to identify the crucial mechanism leading to CHF and to comprehensively predict the CHF at low qualities. For this reason a mechanistic model based on the fundamental mechanism devised from already established visual observations is developed. The fundamental mechanism is expressed by a physical criterion of CHF occurrence and a process limiting the thermal transport between a stagnant bubbly layer and bulk stream. Based on this mechanism a CHF formula can be mathematically formulated by coupling the equation of limiting mixing-mass-flux, which is derived from momentum balance equations in two regions, with local mass and energy balance equations on the bubbly layer. The resulting form of the model is represented by a general and straight forward CHF formula involving two empirical constants related to the void fraction and the thickness of the bubbly layer. The preliminary validation is performed to verify the validity of the model and to adequately determine two empirical constants over an extensive CHF data of water in uniformly heated tubes. Various sensitivity studies are also performed to investigate the influence of various constitutive models as well as the acceleration and buoyancy terms in force balance on the CHF formula. From the preliminary validation and the sensitivity study of the model, it is proved that the prediction based on the model is successful and that the effects of the acceleration and buoyancy terms are negligible. It is also recognized that some model refinement for constitutive model...