Impact of inlet subcooling on flow boiling in microchannels

• Effect of inlet subcooling on flow boiling in silicon-based microchannels has been experimentally studied. • Boiling hysteresis occurs at high inlet subcooling and gradually disappears as inlet subcooling decreases. • A new dimensionless number, boiling utilization, is proposed to evaluate the flow boiling heat transfer performance in microchannels. • Flow reversal is significantly mitigated as inlet subcooling decreases. • A new CHF correlation considering the effect of inlet subcooling on flow boiling in microchannel is proposed. Inlet subcooling is of paramount importance for the flow boiling heat transfer in microchannels. This work carries out the experimental investigation of the impact of inlet subcooling on the deionized water flow boiling in silicon microchannels ( H =200 μm, W =80 μm, L =10 mm). Experiments are conducted with inlet subcooling degrees of 70, 40 and 20 K, mass fluxes of 446-963 kg/m 2 ·s and base heat fluxes of 17.5-407.2 W/cm 2 . It is found that the boiling hysteresis with explosive boiling occurs in the boiling inception at a high subcooling degree of 70 K, and boiling hysteresis gradually gets gentle as inlet subcooling decreases, which has not been reported in the previous works. At a higher subcooling degree, more heat is required for the bubble nucleation but higher critical heat flux (CHF) can be achieved. Further, as inlet subcooling decreases, the flow reversal is well mitigated and thus the flow boiling becomes more stable. Based on the above observations, we put forward a new dimensionless number called the boiling utilization to evaluate the flow boiling characteristics at different inlet subcooling conditions. It is found that a larger degree of subcooling results in a smaller boiling utilization; and the copper-based microchannel has a higher boiling utilization than that of the silicon-based microchannel. A new CHF correlation considering the influence of inlet subcooling is proposed, which shows good agreement with experiments, and the mean absolute error between the experimental values and predicted values by the proposed correlation is 7.3%.