Minimization of two-phase flow frictional pressure drop in small channel with environmentally friendly refrigerant

Accurate prediction of the friction factor and consequently the pressure drop in small two-phase flow channels are still lacking with large disagreements. In addition, the environmental concerns associated with industrial refrigerants currently used have further posed a challenge to find thermally and hydrodynamically compatible and environmentally friendly alternatives. The goal of this study is to determine the optimal friction factor and frictional pressure drop using singleobjective and multi-objective genetic algorithms. A total of 53 friction factor models/correlations have been reviewed from which eight were utilized to address discrepancies. Then, minimization of the frictional pressure drop by implementing single-objective genetic algorithm (SOGA) was carried out. In the multi-objective genetic algorithm (MOGA), the conflicting objectives of friction factor and pressure drop have been minimized simultaneously. Finally, the analysis was carried out on a small horizontal tube of 7.6 mm inner diameter utilizing experimental data for the refrigerant R-22 and the natural refrigerant R-290. It has been shown that the disagreements occur due to (i) the use of fluid data from different sources, (ii) utilization of different correlations on viscosity, and consequently on predicting the friction factor, and (iii) the applications of different friction factor correlations on predicting the frictional pressure drop. It has been proven that the Blasius friction factor correlation for turbulent flow in smooth pipe can be used by experimental researchers to determine their frictional pressure drop or/and matching of data and predicted values due to the coincidentally good agreement obtained. The optimal outcomes using MOGA are found to be closest to the experimental data. The percentage difference between the predicted and experimental frictional pressure drop is up to 1.93% and 0.25% when the Blasius friction factor equation is used with the McAdams and Dukler viscosity equations for R-22 and R-290 respectively