Investigation of Compatibility of the Heat Transfer Coefficient Correlations for Macro and Mini Channels

Many correlations had been and are being developed to predict the heat transfer coefficient in two-phase flow in a mini channel, particularly for new coolants investigated for their low environmental impacts. However, large discrepancies have been reported between the predicted and experimental heat transfer coefficient. Accurate prediction of the heat transfer coefficient is important to reduce cost and save energy and material associated with over or under design of heat exchanging devices. This study was conducted to examine two types of heat transfer coefficient correlations often used, under a common platform - optimized conditions - using genetic algorithm. One was developed for a macro channel but often used for a mini channel, and two others specific for a mini channel. Optimization was performed on R290, a natural refrigerant recently being investigated for its potential to replace R22, for a 3-mm channel at the saturation temperature of 10°C. The optimised predictions for the highest heat transfer coefficient compared showed that the optimized heat transfer coefficient obtained from the correlation that was developed for a macro channel is close to that obtained from the correlation developed specifically for a mini channel. However, the former is found at the high end of the quality and mass flux region while the latter was obtained at the low end of the quality and mass flux region. The correlations exhibit different behaviour under optimized conditions attributed to the parameters assumed to represent the behaviour of the refrigerants. This study has shown that even under the best expected performance, selection of the appropriate correlation, mini or macro channel, is important for accurate prediction of heat exchanging devices.