The entire liquid-film dryout process in vertical annular two-phase boiling flow is characterized experimentally, from inception to completion, in a newly defined flow regime between annular two-phase flow and mist flow, called ‘intermittent dryout’. Experiments are conducted using saturated R245fa flowing vertically through a heated transparent channel. Liquid-film thickness measurements are taken during dryout events. The state (wet or dry) of the heated surface is determined using a laser reflectance measurement, and the signal is used to calculate the time-averaged dry fraction, fdry. The local heat transfer coefficient (HTC) is characterized as a function of fdry. For all the investigated mass fluxes, optimum boiling conditions (OBC), corresponding to the heat flux that results in maximum HTC, consistently occur when the time-averaged dry fraction is fdry≈0.05. Subsequently, the critical boiling transition (CBT), corresponding to the heat flux that results in a significantly lower HTC, can occur for dry fractions of fdry>0.1. Further insight into the liquid film behavior within the intermittent dryout regime is obtained by combining analyses of high-speed videos, time-resolved liquid-film thickness signals, and statistics about the duration of and time between dryout events. In general, the dryout mechanism is dominated by disturbance waves.Mechanistic model frameworks are presented for the prediction of two-phase HTC and heat flux associated with dryout. The model inputs are local flow quantities (pressure, mass flux, vapor quality, and liquid-film thickness) and output a prediction for local heat transfer coefficient. The model framework accurately predicts the local two-phase HTC for a range of mass flux and vapor quality in the annular flow regime. The model framework captures the observed initial increase in the local HTC and subsequent decrease that is associated with liquid-film dryout. The mechanistic approach provides predictions for time-averaged HTC and OBC heat flux that both agree well with experimentally measured values.
Read full abstract