Effect of refrigerant oil additive on R134a and R123 boiling heat transfer performance

This paper investigates the effect that an additive had on the boiling performance of an R134a/polyolester lubricant (POE) mixture and an R123/naphthenic mineral oil mixture on a roughened, horizontal flat surface. Both pool boiling heat transfer data and lubricant excess surface density data are given for the R134a/POE (98% mass fraction/2% mass fraction) mixture before and after use of the additive. A spectrofluorometer was used to measure the lubricant excess density that was established by the boiling of the R134a/POE lubricant mixture before and after use of the additive. The measurements obtained from the spectrofluorometer suggest that the additive increases the total mass of lubricant on the boiling surface. The heat transfer data show that the additive caused an average and a maximum enhancement of the R134a/POE heat flux between 5 kW m⁻² and 22 kW m⁻² of approximately 73% and 95%, respectively. Conversely, for nearly the same heat flux range, the additive caused essentially no change in the pool boiling heat flux of an R123/mineral oil mixture. The lubricant excess surface density and interfacial surface tension measurements of this study were used to form the basis of a hypothesis for predicting when additives will enhance or degrade refrigerant/lubricant pool boiling.     

Effect of bulk lubricant concentration on the excess surface density during R123 pool boiling

This paper investigates the effect that bulk lubricant concentration has on the non-adiabatic lubricant excess surface density on a roughened, horizontal flat (plain) pool-boiling surface. Both pool boiling heat transfer data and lubricant excess surface density data are given for pure R123 and three different mixtures of R123 and a naphthenic mineral oil. A spectrofluorometer was used to measure the lubricant excess density that was established by the boiling of a R123/lubricant mixture on a test surface. The fluorescent technique was used to measure the effect of bulk lubricant concentration on the lubricant excess layer during refrigerant/lubricant mixture boiling. The refrigerant preferentially boils, thus, concentrating and accumulating the lubricant on the surface in excess of the bulk concentration. The excess lubricant resides in a very thin layer on the surface and influences the boiling performance. Accordingly, the ability to measure the effect of bulk lubricant composition on the lubricant excess density and in turn the effect on the heat transfer would lead to a fundamental understanding of the mechanism by which lubricants can degrade or improve boiling performance. In support of this effort, heat transfer data are provided for pure R123 and three R123/lubricant mixtures at 277.6 K. For heat fluxes between approximately 25 to 45 kW/m², an average enhancement of the heat flux of 9 and 5% was achieved for the 0.5 and 1% lubricant mass fractions, respectively, and an average degradation of 5% in the heat flux was obtained for the 1.8% lubricant mass fraction mixture.

Résumé

This paper investigates the effect that bulk lubricant concentration has on the non-adiabatic lubricant excess surface density on a roughened, horizontal flat (plain) pool-boiling surface. Both pool boiling heat transfer data and lubricant excess surface density data are given for pure R123 and three different mixtures of R123 and a naphthenic mineral oil. A spectrofluorometer was used to measure the lubricant excess density that was established by the boiling of a R123/lubricant mixture on a test surface. The fluorescent technique was used to measure the effect of bulk lubricant concentration on the lubricant excess layer during refrigerant/lubricant mixture boiling. The refrigerant preferentially boils, thus, concentrating and accumulating the lubricant on the surface in excess of the bulk concentration. The excess lubricant resides in a very thin layer on the surface and influences the boiling performance. Accordingly, the ability to measure the effect of bulk lubricant composition on the lubricant excess density and in turn the effect on the heat transfer would lead to a fundamental understanding of the mechanism by which lubricants can degrade or improve boiling performance. In support of this effort, heat transfer data are provided for pure R123 and three R123/lubricant mixtures at 277.6 K. For heat fluxes between approximately 25 kW/m² to 45 kW/m², an average enhancement of the heat flux of 9% and 5% was achieved for the 0.5% and 1% lubricant mass fractions, respectively, and an average degradation of 5% in the heat flux was obtained for the 1.8% lubricant mass fraction mixture.     

A semi-theoretical model for predicting refrigerant/lubricant mixture pool boiling heat transfer

This paper outlines the framework of a semi-theoretical model for predicting the pool boiling heat transfer of refrigerant/lubricant mixtures on a roughened, horizontal, flat pool-boiling surface. The predictive model is based on the mechanisms involved in the formation of the lubricant excess layer that exists on the heat transfer surface. The lubricant accumulates on the surface in excess of the bulk concentration via preferential evaporation of the refrigerant from the bulk refrigerant/lubricant mixture. As a result, excess lubricant resides in a thin layer on the surface and influences the boiling performance, giving either an enhancement or degradation in heat transfer. A dimensionless excess layer parameter and a thermal boundary layer constant were derived and fitted to data in an attempt to generalize the model to other refrigerant/lubricant mixtures. The model inputs include transport and thermodynamic refrigerant properties and the lubricant composition, viscosity, and critical solution temperature with the refrigerant. The model predicts the boiling heat transfer coefficient of three different mixtures of R123 and lubricant to within ±10%. Comparisons of heat transfer predictions to measurements for 13 different refrigerant/lubricant mixtures were made, including two different refrigerants and three different lubricants.     

Use of fluorescence to measure the lubricant excess surface density during pool boiling

This paper presents what are believed to be the first measurements of the non-adiabatic lubricant excess surface density on a roughened, flat, plain horizontal pool-boiling surface. Pool boiling heat transfer data is given for pure R123 and a R123/lubricant mixture. Lubricant excess surface density data are given for the boiling R123/lubricant mixture. A spectrofluorometer was used to measure the lubricant excess density that was established by the boiling of a R123/lubricant mixture on a test surface. The fluorescent measurement technique was used to confirm the existence of the lubricant excess layer during refrigerant/lubricant mixture boiling. The refrigerant preferentially boils, thus, concentrating and accumulating the lubricant on the surface in excess of the bulk concentration. The excess lubricant resides in a very thin layer on the surface and influences the boiling performance. Accordingly, the ability to measure the lubricant excess density on the heat transfer surface would lead to a fundamental understanding of the mechanism by which lubricants can degrade or improve boiling performance. In support of this effort, heat transfer data are provided for both pure R123 and an R123/lubricant (1.8% lubricant mass fraction) mixture at 277.6 K. The heat transfer data shows that the lubricant excess causes an average degradation of 12% in the heat flux for a given superheat.

Résumé

This paper presents what are believed to be the first measurements of the non-adiabatic lubricant excess surface density on a roughened, flat, plain horizontal pool-boiling surface. Pool boiling heat transfer data is given for pure R123 and a R123/lubricant mixture. Lubricant excess surface density data are given for the boiling R123/lubricant mixture. A spectrofluorometer was used to measure the lubricant excess density that was established by the boiling of a R123/lubricant mixture on a test surface. The fluorescent measurement technique was used to confirm the existence of the lubricant excess layer during refrigerant/lubricant mixture boiling. The refrigerant preferentially boils, thus, concentrating and accumulating the lubricant on the surface in excess of the bulk concentration. The excess lubricant resides in a very thin layer on the surface and influences the boiling performance. Accordingly, the ability to measure the lubricant excess density on the heat transfer surface would lead to a fundamental understanding of the mechanism by which lubricants can degrade or improve boiling performance. In support of this effort, heat transfer data are provided for both pure R123 and an R123/lubricant (1.8% lubricant mass fraction) mixture at 277.6 K. The heat transfer data shows that the lubricant excess causes an average degradation of 12% in the heat flux for a given superheat.     

In-tube condensation of mixture of R134a and ester oil: empirical correlations

This paper presents a comparative study of the condensation heat transfer coefficients in a smooth tube when operating with pure refrigerant R134a and its mixture with lubricant Castrol “icematic sw”. The lubricant is synthetic polyol ester based oil commonly used in lubricating the compressors. Two concentrations of R134a-oil mixtures of 2% and 5% oil (by mass) were analysed for a range of saturation temperatures of refrigerant R134a between 35 °C and 45 °C. The mass flow rate of the refrigerant and the mixtures was carefully maintained at 1 g/s, with a vapour quality varying between 1.0 and 0. The effects of vapour quality, flow rate, saturation temperature and temperature difference between saturation and tube wall on the heat transfer coefficient are investigated by analysing the experimental data. The experimental results were then compared with predictions from earlier models [Int J Heat Mass Transfer (1979), 185; 6th Int Heat Transfer Congress 3 (1974) 309; Int J Refrig 18 (1995) 524; Trans ASME 120 (1998) 193]. Finally two new empirical models were developed to predict the two-phase condensation heat transfer coefficient for pure refrigerant R134a and a mixture of refrigerant R134a with Castrol “icematic sw”.     

The effect of lubricant concentration, miscibility, and viscosity on R134a pool boiling

This paper presents pool boiling heat transfer data for 12 different R134a/lubricant mixtures and pure R134a on a Turbo-BII™-HP surface. The mixtures were designed to examine the effects of lubricant mass fraction, viscosity, and miscibility on the heat transfer performance of R134a. The magnitude of the effect of each parameter on the heat transfer was quantified with a regression analysis. The mechanistic cause of each effect was given based on new theoretical interpretation and/or one from the literature. The model illustrates that large improvements over pure R134a heat transfer can be obtained for R134a/lubricant mixtures with small lubricant mass fraction, high lubricant viscosity, and a large critical solution temperature (CST). The ratio of the heat flux of the R134a/lubricant mixture to that of the pure R134a for fixed wall superheat was given as a function of pure R134a heat flux for all 12 mixtures. The lubricant that had the largest CST with R134a exhibited the greatest heat transfer: 100%±20% greater than that of pure R134a. By contrast, the heat transfer of the mixture with the lubricant that had the smallest viscosity and the smallest CST with R134a was 55%±9% less than that of pure R134a. High-speed films of the pure and mixture pool boiling were taken to observe the effect of the lubricant on the nucleate boiling.     

Effect of Al2O3 nanolubricant on R134a pool boiling heat transfer

This paper quantifies the influence of Al₂O₃ nanoparticles on the pool-boiling performance of R134a/polyolester mixtures on a roughened, horizontal, flat surface. The nanoparticles enhanced the boiling heat transfer relative to that for R134a/polyolester mixtures without nanoparticles for the three lubricant mass fractions that were tested. For the 0.5% nanolubricant mass fraction, the nanoparticles caused a heat transfer enhancement relative to the heat transfer of pure R134a/polyolester (99.5/0.5) as large as 400% for the lowest heat flux. The average heat flux improvement for heat fluxes less than 40 kW m⁻² was approximately 105%, 49%, and 155% for the 0.5%, the 1%, and the 2% mass fractions, respectively. A semi-empirical model was developed to predict the boiling enhancement as caused by the interaction of the nanoparticles with the bubbles. The model suggests that small particle size and large nanoparticle volume fraction improve boiling enhancement.     

Effect of CuO nanolubricant on R134a pool boiling heat transfer

This paper quantifies the influence of CuO nanoparticles on the boiling performance of R134a/polyolester mixtures on a roughened, horizontal, flat surface. A lubricant based nanofluid (nanolubricant) was made with a synthetic ester and CuO particles. For the 0.5% nanolubricant mass fraction, the nanoparticles caused a heat transfer enhancement relative to the heat transfer of pure R134a/polyolester (99.5/0.5) of between 50% and 275%. A smaller enhancement was observed for the R134a/nanolubricant (99/1) mixture, which had a heat flux that was on average 19% larger than that of the R134a/polyolester (99/1) mixture. Further increase in the nanolubricant mass fraction to 2% resulted in a still smaller boiling heat transfer improvement of approximately 12% on average. Thermal conductivity measurements and a refrigerant/lubricant mixture pool-boiling model were used to suggest that increased thermal conductivity is responsible for only a small portion of the heat transfer enhancement due to nanoparticles.     

Reducing running costs in air conditioningRéduire les coûts de fonctionnement en conditionnement d'air

Previous studies using wide-boiling refrigerant mixtures were reported in the literature as a method of reducing power consumption in air conditioning. A novel air conditioning test unit incorporating this concept was tested at the Department of Chemical Engineering at Leeds University. Experimental tests with a refrigerant mixture, R407C, produced COPs in excess of 4. Also, a simulation program was used to model the behaviour of other mixtures. It yielded a COP of about 5 for mixture R32/R134a. A financial study was made to calculate the total annual costs of a unit using R32/R134a. It showed that annual savings of up to 41 per cent were possible compared to a unit using R134a.     

The surface tension of HFC refrigerants and mixtures

The surface tension of the refrigerants R32, R125, R134a, R143a and R152a, as well as the binary refrigerant mixtures R32-R125, R32-R134a, R125-R134a, R125-R143a, R125- R152a, R143a-R134a and R134a-R152a, and the commercially available ternary mixtures R404A and R407C was measured across the temperature range from −50 to 60°C using a measuring unit based on the capillary rise method. Different formulations for calculation of the surface tension of the binary and ternary mixtures on the basis of the surface tension of the pure refrigerants were tested. With an approach based on mass proportions in the mixture, a good correspondence between the measured and calculated values was achieved.     

Flow-boiling of R22, R134a, R507, R404A and R410A inside a smooth horizontal tube

Flow boiling heat transfer coefficients of R22, R134a, R507, R404A and R410A inside a smooth horizontal tube (6 mm I.D., 6 m length) were measured at a refrigerant mass flux of about 360 kg/m² s varying the evaporating pressure within the range 3–12 bar, with heat fluxes within the range 11–21 kW/m². The experimental data are discussed in terms of the heat transfer coefficients as a function of the vapour quality. The experimental results clearly show that the heat transfer coefficients of R134a are always higher than those pertaining to R22 (from a minimum of +6 to a maximum of +45%).     

Experimental transient performance of a heat pump equipped with a distillation column

A series of experiments were conducted on a heat pump equipped with a distillation column. The system was operated with R32 and with a 30/70% by mass mixture of R32/134a to examine the difference between the transient performance trends with a pure fluid (R32), and those with a zeotropic mixture (R32/134a). Additionally, the effects of varying heat transfer fluid mass flow, compressor speed, and accumulator sump heat input were examined. Each test was 1 h in duration. The heat pump capacities did not generally achieve steady state during the R32/134a tests. Steady state was generally achieved during the R32 tests. As a percentage of the final (end-of-test) capacity, the rate of capacity increase was greater during the R32/134a tests than during those conducted with the pure fluid. The R32/134a tests exhibited capacity oscillations early in each transient that were not present during the R32 tests. The results show that circulating refrigerant mass and composition are the primary controlling factors with regard to transient capacity.     

Performance characteristics of a small-capacity directly cooled refrigerator using R290/R600a (55/45)

In this study, the performance of a small-capacity directly cooled refrigerator was evaluated by using the mixture of R290 and R600a with mass fraction of 55:45 as an alternative to R134a. The compressor displacement volume of the alternative system with R290/R600a (55/45) was modified from that of the original system with R134a to match the refrigeration capacity. Both systems with R290/R600a (55/45) and R134a were tested, and then optimized by varying the refrigerant charge and capillary tube length under experimental conditions for both the pull-down test and the power consumption test. The refrigerant charge of the optimized R290/R600a system was approximately 50% of that of the optimized R134a system. The capillary tube lengths for each evaporator in the optimized R290/R600a system were 500 mm longer than those in the optimized R134a system. The power consumption of the optimized R134a system was 12.3% higher than that of the optimized R290/R600a system. The cooling speed of the optimized R290/R600a (55/45) system at the in-case setting temperature of −15 °C was improved by 28.8% over that of the optimized R134a system.     

Experimental and theoretical study on flow condensation with non-azeotropic refrigerant mixtures of R32/R134a

Experiments on flow condensation have been conducted with both pure R32, R134a and their mixtures inside a tube (10 m long, 6 mm ID), with a mass flux of 131–369 kg m⁻²s⁻¹ and average condensation temperature of 23–40°C. The experimental heat transfer coefficients are compared with those predicted from correlations. The maximum mean heat transfer coefficient reduction (from a linear interpolation of the single component values) occurs at a concentration of roughly 30% R32 for the same mass flux basis, and is approximately 20% at Gr = 190 kg m⁻²s⁻¹, 16% at Gr = 300 kg m⁻²s⁻¹. Non-ideal properties of the mixture have a certain, but relatively small, influence on the degradation. Among others, temperature and concentration gradients, slip, etc. are also causes of heat transfer degradation.     

Experimental study on forced convective boiling heat transfer of pure refrigerants and refrigerant mixtures in a horizontal tube

Convective boiling heat transfer coefficients of pure refrigerants (R22, R32, R134A, R290, and R600a) and refrigerant mixtures (R32/R134a, R290/R600a, and R32/R125) are measured experimentally and compared with Gungor and Winterton correlation. The test section is made of a seamless stainless steel tube with an inner diameter of 7.7 mm and is uniformly heated by applying electric current directly to the tube. The exit temperature of the test section was kept at 12°C ± 0.5°C for all refrigerants in this study. Heat fluxes are varied from 10 to 30 kW m⁻² and mass fluxes are set to the discrete values in the range of 424–742 kg m⁻² s⁻¹ for R22, R32, R134a, R32/R134a, and R32/R125; 265–583 kg m⁻² s⁻¹ for R290, R600a, and R290/R600a. Heat transfer coefficients depend strongly on heat flux at a low quality region and become independent as quality increases. The Gungor and Winterton correlation for pure substances and the Thome-Shakil modification of this correlation for refrigerant mixtures overpredicts the heat transfer coefficients measured in this study.     

Effect of concentration on R134a/Al2O3 nanolubricant mixture boiling on a reentrant cavity surface

This paper quantifies the influence of Al₂O₃ nanoparticles on the pool boiling performance of R134a/polyolester mixtures on a Turbo-BII-HP boiling surface. Nanolubricants with 10 nm diameter Al₂O₃ nanoparticles of various volume fractions (1.6%, 2.3%, and 5.1%) in the base polyolester lubricant were mixed with R134a at two different mass fractions (0.5% and 1%). The study showed that nanolubricants can improve R134a boiling on a reentrant cavity surface as long as the nanoparticles remain well dispersed in the lubricant and are at sufficiently large concentration. For example, three of the refrigerant/nanolubricant mixtures with the smallest nanoparticle mass fraction exhibited average enhancements over the entire heat flux range of approximately 10%. However, when the nanoparticle mass fraction was increased to a point that likely encouraged agglomeration, an average heat transfer degradation of approximately 14% resulted. An existing model was used to predict the boiling heat transfer.     

Enhancement of R123 pool boiling by the addition of hydrocarbonsAmélioration de l'ébullition libre du R123 par l'ajout d'hydrocarbures

This paper presents pool boiling heat transfer data for 10 different R123/hydrocarbon mixtures. The data consisted of pool boiling performance of a GEWA-T^™ surface for pure R123 and for 10 dilute solutions of five different hydrocarbons: (1) pentane, (2) isopentane, (3) hexane, (4) cyclohexane, and (5) heptane with R123. The heat flux and the wall superheat were measured for each fluid at 277.6 K. A maximum (19±3.5)% increase over the pure R123 heat flux was achieved with the addition of 0.5% mass isopentane to R123. Other mixtures of isopentane, pentane, hexane, and cyclohexane with R123 exhibited smaller maximums than that of the R123/isopentane (99.5/0.5) mixture. Presumably, a layer enriched in hydrocarbon at the heat transfer surface caused the heat transfer enhancement. Conversely, an R123/heptane (99.5/0.5) mixture and an R123/cylcohexane (99.5/0.5) mixture exhibited only degradations with respect to the pure component performance for all test conditions. Several characteristics of the hydrocarbons were examined to determine their influence on the boiling heat transfer performance: molecular weight, molecular structure, composition, surface tension, and vapor pressure.     

Effects of synthetic oil in a compression refrigeration system using R410A. Part II: quality of heat transfer and pressure losses within the heat exchangers

The consequences of the oil rejected by the compressor of a vapour-compression refrigeration system on the operation of the evaporator and condenser are analysed. The modelled prototype uses the mixture of HFC R410A and a synthetic polyolester (POE) oil. The rise of the amount of lubricant circulating in the system leads to a progressive change in the behaviour of the mixture of refrigerant and oil that, for the higher oil mass fraction, evolves like a zeotropic mixture. One also observes that the presence of lubricant is generally associated with a fall of the performances of the heat exchangers, except however in the evaporator where an optimum is observed when the quantity of oil is equal to 0.1% of the total mass of the mixture. Some conclusions are drawn about the choice of correlations for the calculation of the refrigerant side heat transfer coefficient in a plate evaporator.     

Refrigerant R134a vaporisation heat transfer and pressure drop inside a small brazed plate heat exchanger

This paper presents the experimental heat transfer coefficients and pressure drop measured during refrigerant R134a vaporisation inside a small brazed plate heat exchanger (BPHE): the effects of heat flux, refrigerant mass flux, saturation temperature and outlet conditions are investigated. The BPHE tested consists of 10 plates, 72 mm in width and 310 mm in length, which present a macro-scale herringbone corrugation with an inclination angle of 65° and corrugation amplitude of 2 mm.The experimental results are reported in terms of refrigerant side heat transfer coefficients and frictional pressure drop. The heat transfer coefficients show great sensitivity both to heat flux and outlet conditions and weak sensitivity to saturation temperature. The frictional pressure drop shows a linear dependence on the kinetic energy per unit volume of the refrigerant flow.The experimental heat transfer coefficients are also compared with two well-known correlations for nucleate pool boiling and a correlation for frictional pressure drop is proposed.     

Flow condensation heat transfer coefficients of pure refrigerants

Flow condensation heat transfer coefficients (HTCs) of R12, R22, R32, R123, R125, R134a, and R142b were measured experimentally on a horizontal plain tube. The experimental apparatus was composed of three main parts; a refrigerant loop, a water loop and a water-glycol loop. The test section in the refrigerant loop was made of a copper tube with an outside diameter of 9.52 mm and 1 m length. The refrigerant was cooled by cold water passing through an annulus surrounding the test section. All tests were performed at a fixed refrigerant saturation temperature of 40 °C with mass fluxes of 100, 200, 300 kg m⁻² s⁻¹ and heat flux of 7.3–7.7 kW m⁻². Experimental results showed that flow condensation HTCs increase as the quality and mass flux increase. At the same mass flux, the HTCs of R142b and R32 are higher than those of R22 by 8–34% while HTCs of R134a and R123 are similar to those of R22. On the other hand, HTCs of R12 and R125 are lower than those of R22 by 24–30%. Previous correlations predicted the present data satisfactorily with mean deviations of less than 20% substantiating indirectly the reliability of the present data. Finally, a new correlation was developed by modifying Dobson and Chato's correlation with an introduction of a heat and mass flux ratio combined with latent heat of condensation. The correlation showed a mean deviation of 10.7% for all pure halogenated refrigerants' data obtained in this study.