A mechanistic model to predict saturated pool boiling Critical Heat Flux (CHF) in a confined gap

Abstract

Boiling enables high rates of heat transfer from a surface made possible at a relatively low thermal resistance motivating the use of two-phase cooling for increasingly compact thermal management solutions. However, extreme geometrical confinement of the liquid above the boiling surfaces is known to have detrimental effects on maximum heat transfer rate by inducing premature onset of film boiling. Moreover, previously developed critical heat flux (CHF) models for confined geometries focused on triggering mechanisms associated with unconfined pool boiling and, thus, are not generalizable. This work proposes a new mechanistic model for predicting CHF during boiling within in narrow gap, specifically developed to account for confinement effects on the triggering mechanism. The model postulates that occurrence of CHF coincides with the irreversible growth of a dry spot on the boiling surface. Three competing forces govern the two-phase interface dynamics, namely vapor momentum, surface tension, and hydrostatic forces. Dryout is triggered when the vapor momentum force due to vaporization at the two-phase interface balances the combined surface tension and hydrostatic forces leading to irreversible growth of the dry spot. The present work offers a predictive confined CHF model that accounts for confined boiling surface shape, size, orientation, confinement gap spacing, and working fluid properties, with a single fluid-specific fitting coefficient that represents the ratio of vapor area to the confinement opening area near CHF conditions. Notably, the developed CHF model is also effective in predicting the threshold gap below which confinement reduces pool boiling CHF. The model is compared to 197 experimentally measured confined CHF data points available from 10 studies in the literature that represent 7 different working fluids and a range of boiling surface inclinations and shapes. The model predicts the confinement-reduced CHF values with a root mean square error of 21%, which is less than half of the error compared to all other available predictive models. This clarification of the triggering mechanism and improved prediction accuracy of CHF, as offered by the current study, will enable broader practical system implementation of compact two-phase cooling technologies.