Prediction of the Pool Boiling Critical Heat Flux Using Artificial Neural Network

This study deals with artificial neural network (ANN) based prediction of the pool boiling critical heat flux (CHF) for dielectric liquids under a variety of operating conditions. For this purpose, first, the effects of pressure and subcooling of the fluid on nucleate pool boiling and CHF on the surface of an integrated circuit (IC) immersed in a fluorocarbon FC-72 liquid were investigated experimentally. Then, using the experimental data, an ANN model was developed for the prediction of CHF. The input parameters of the ANN were pressure, temperature, and saturation temperature of the fluid, heater temperature and surface temperature of the chip, while the output of the ANN was the CHF. Finally, the ANN predictions were compared with the experimental results. Furthermore, the effects of the input parameters on the CHF were also investigated. It was found that the ANN yielded a satisfactory correlation between the experimental and predicted CHFs with a correlation coefficient of 0.996, a mean relative error of 8.81%, and a root mean square error of 0.97Wcm^-2. This study shows that the pool boiling CHF can alternatively be predicted using ANN within a very high degree of accuracy