Performance of the two-phase flow in a minichannel had in the past been measured by the pressure drop or/and heat transfer coefficient. The desired low pressure drop across a small channel follows a low heat transfer coefficient. Optimization of the two-phase flow system is generally achieved either experimentally through discrete variations of each of the parameters involved while holding the rest constant, or numerically which is also possible through a parametric study. The objective of this study was to investigate the thermodynamic performance in terms of entropy generation minimization (EGM) of two-phase flow of ammonia, R22, and R134A in a 3-mm minichannel using a random search technique, genetic algorithm. The EGM performance and the optimization approach have never been attempted before. R22 has been identified as a hazardous refrigerant and alternatives are being investigated with performance as good if not better. In this study, under the optimization of the mass flux and vapour quality at the saturation temperature of 10C, simultaneous minimization of the entropy generation and maximization of the heat transfer coefficient showed that between 250 and 450 kg/m 2 s, ammonia has a much higher heat transfer coefficient than R22 and R134A, and at a lower quality but with very high entropy generation. Furthermore, ammonia has many sets of optimal solutions, several combinations of entropy generation and heat transfer coefficient under optimized heat flux operation and vapour quality. R22 and R134A have their optimized heat transfer coefficients over a limited range and which occurred beyond the quality of 0.8. The study has shown that ammonia could be the replacement refrigerant to R22 and R134A in terms of heat transfer but at the expense of a higher entropy generation rate.

Accurate prediction of the friction factor and consequently the pressure drop of two-phase flow in small channels is still an issue. Many correlations exist for the determination of the viscosity and the friction factor that appear in the frictional pressure drop and their combination often determined the degree of disagreements between the experimental data and predicted outcomes. Demands for environmentally friendly refrigerants have further posed a challenge to find compatible alternatives with as good a performance as the current coolants. Despite the many available correlations developed to date, many more are studied in effort to reduce the discrepancies. This paper presents the outcomes of a study comparing the optimized conditions when three different viscosity equations are paired with eight different friction factor correlations to minimize the frictional pressure drop. The approach used multi-objective genetic algorithm (MOGA) to assist in selecting the best pairing. Comparison is then completed with available experimental data. The study showed that the Blasius friction factor paired with the Dukler viscosity produced the least percentage difference for R22, while when paired with the McAdams viscosity produced a lower difference for R290, an environmentally friendly refrigerant being considered to replace R22.

No abstract

No abstract