Two-phase flow of HCFC-22 and HFC-134a through short tube orifices

Abstract

HCFC-22 and HFC-134a were used with short tube orifices having length to diameter ratios ranging from 5 to 20 in a 9.53 mm (3/8 in.) refrigerant line to investigate both two-phase and subcooled liquid flow entering the short tubes. Flow temperature and pressure conditions were those typically found in air-conditioner and heat pump applications. Effects of each operating parameters and short tube geometry on the flow rate were discussed and included in the modeling. Both an analytical and semi-empirical model were developed. These models allowed better prediction of mass flow rate and investigation of transport properties inside the short tube. Generally, the flow trends for both refrigerants were similar even though mass flow rate range for HFC-134a was 30 % lower than that for HCFC-22 due to lower operating pressure conditions. Within normal heat pump or air-conditioner operating conditions, second-stage choking (choking at the outlet) was only approximated because the flow showed a small dependency on downstream pressure. For the conditions investigated, the mass flow rate was directly proportional to upstream pressure and upstream subcooling, but it showed only a small dependency on downstream pressure. For two-phase flow entering the short tube, the flow rate decreased with increasing quality. The mass flow rate was found to be dependent on cross sectional area, chamfer depth, and short tube length. For subcooled liquid entering the short tube, the delay of flashing, i.e. metastable liquid flow, was observed at the inlet section of the short tube. It was observed that the effects of oil concentration on mass flow rate varied as a function of short tube geometry and upstream subcooling or quality. The semi-empirical flow model and flow charts for both single and two-phase flow were developed by empirically correcting the modified single-phase orifice equation to satisfy the short tube orifice flow. For both HCFC-22 and HFC-134a, the maximum difference between measured data and the model's prediction was within [plus or minus]10%...