The major parameters of interest in heat transfer research are the refrigerant charge, pressure drop, and heat transfer capacity. Smaller channels reduce the refrigerant charge with higher heat transfer capability due to the increased in surface area to volume ratio but at the expense of a higher pressure drop. Differences between the predicted and experimental frictional pressure drop of two-phase flow in small tubes have frequently been discussed. Factors that could have contributed to that effect have been attributed to the correlations used to model the flow, some being modified from the originals developed for a macro system. Experimental test-rigs have varied in channel geometry, refrigerant type, and flow conditions. Thousands of data have been collected to find a common point among the differences. This paper reports an investigation of four different two-phase friction factor correlations used in the modeling of the frictional two-phase flow pressure drop of refrigerant R-22. One had been specifically developed for laminar flow in a smooth channel, another was modified from a laminar flow in a smooth pipe to be used for a rough channel, and two correlations are specific for turbulent flow that consider internal pipe surface roughness. Genetic algorithm, an optimization scheme, is used to search for the minimum friction factor and minimum frictional pressure drop under optimized conditions of the mass flux and vapor quality. The results show that a larger pressure drop does come with a smaller channel. A large discrepancy exists between the correlations investigated; between the ones that does not consider surface roughness and that which does, as well as between flow under laminar and turbulent flow conditions.
Read full abstract