1、4746 (RP-1253) Measurement of Solubility, Viscosity, and Density of R-507A RefrigerantlLu bricant Mixtures Richard C. Cavestri, PhD John R. Thuermer Donna Seeger-Clevenger ABSTRACT The refrigerant vapor/liquid lubricant equilibrium viscos- ity reduction of two polyolesters (POE), viscosity grades 32
2、 and 68 IS0 VG, and one polyalblene glycol (PAG), viscosity grade 68 IS0 VG, were measured with R-507A from -5C (23F) to 125C (257F) up to 500psia. All three lubricants have a miscibility curve starting at 2% and in increments up to 60% lubricant and refrigerant concentration determined from -40C (-
3、40F) to 60C (1 40F). Using validated analyt- ical methods, the viscosi, density, and composition of the solubilized gas mixture in solution with the lubricant are obtained. This rejkigerant has a very small temperature glide, but it is still considered a blend. Fractionation data were obtained from
4、individual isother- mal measurements. The isotherm for each temperature details the composition of the equilibrium of gas fractionation of R- 125 andR143a in the lubricant, mixed vaporpressure, concen- tration of the total mixed blend (percent by weight) in the lubri- cant, and viscosity in centipoi
5、ses (cP) and centistokes (cSt). INTRODUCTION This ASHRAE research is quite similar to the ARTI MCLR refrigerant blend viscosity measurement program (Cavestri 1995) and ASHRAE RP-779 (R-407C) (Cavestri and Falconi 1999) and RP-928 (R-410A) (Cavestri and Scha- fer 2000). Earlier studies indicated that
6、 the viscosity reduc- tions of polyolester fluids by single and blended gases at isothermal conditions are slightly different, but they are simi- lar when based on concentration and the molecular weight of the gases (Cavestri 1995). For the purposes of an analytical approach, it is assumed that the
7、refngerant blend is always in excess to the lubricant. The analytical direction of this project followed that of the ARTI MCLR, ASHRAE RP-779, and ASHRAE RP-928 research in maintaining the composition of the gas vapor in the head space of the viscometer above the liquid to be identical (1%) to the l
8、iquid refrigerant composi- tion (R-507A in this study) injected into the viscometer (Cavestri 1994; Cavestri and Falconi 1999; Cavestri and Scha- fer 2000). When studying a refrigerant blend there are two very important fluid property changes for understanding the refrig- eration cycle: (i) the cons
9、tant vapor compositions and the correlation of liquidliquid miscibility to viscosity reduction at low evaporator temperatures and (2) the refrigerant composi- tion in the lubricant when studying the tribology of lubrication for compressors with various metal couples on bench testers. Sufficient expe
10、rimental evidence exists that illustrates frac- tionation of blended gases occurs at various levels with specific refrigerants and lubricants at various pressures and temperatures. Compressor lubrication is therefore dependent on these fluid properties (Cavestri 1994, 1995). The information gleaned
11、from this study provides the HVAC industry with an evaluation of a 68 IS0 VG polyalky- lene glycol (methoxy terminated propylene oxide polyether) and 32 and 68 IS0 VG polyolester lubricants of different miscilibities with R-507A. The finished study represents viscosity, density, and gas solubility o
12、f R-507A vapor in the lubricant and evaluates the ratio of gases dissolved in the lubri- cant. When liquidliquid immiscibility occurs at low tempera- tures, the equilibrium of refrigerant solubility shifts and the lubricants expel the HFC, reducing the concentration in the lubricant and increasing t
13、he viscosity levels. The velocity of cold, liquid refrigerant in the system piping can then only provide immiscible lubricant return to the compressor. Richard C. Cavestri is president and director, John R. Thuermer is a research associate, and Donna Seeger-Clevenger is a research associate and tech
14、nical writer at Imagination Resources, Inc., Dublin, Ohio. 02005 ASHFIAE. 53 Refrigerant fractionation occurs instantly when a refrig- erant blend is added to any lubricant and is most pronounced at low refrigerant pressures and concentrations. Having a fixed vapor composition over the lubricant at
15、isothermal conditions provides a stable point of analysis and measurement. Reex- amination of physical property data shows this method provides an accurate representation of viscosity and lubricant composition in a hermetic system. Individual isothermal plots show lubricant and refrigerant compositi
16、on. When these data are combined, smoothed curves of viscosity and pressure at constant concentrations are plotted to temperature. Our viscosity reduction program measures the fraction- ation of HFC refrigerant gas mixtures in either mineral oil or synthetic lubricants and is the same method as used
17、 in previous work (Cavestri 1993-2000). The approach is equilibrium gas vapor solubility over liquid lubricant and is similar to the method described by Parmelee (l964), Albright and Mendel- baum (1956), Albright and Lawyer (1959), and Little (1952). While these earlier investigations were temperatu
18、re and pres- sure limited, our present limitations are -40C (-40F) to 125C (257F) and 1000 psia (6895 kPa). A salient feature that has emerged from our research is that viscosity reduction by gas solubility equilibrium condi- tions is not dramatically affected by low liquid/liquid misci- bility lubr
19、icants. It appears that as long as there is a constant composition gas phase present over the liquid, the refrigerant gases will reach solubility equilibrium with the lubricant, demonstrating a viscosity reduction dependent on vapor composition and pressure. The amount of viscosity reduction at the
20、upper end ofthe temperature and pressure scale is nearly identical with lubricants of differing miscibility. However, at lower temperatures and pressures, the gas solubility can be favored by one gas over the other when (1) IiquidAiquid misci- bility of the two lubricant fluids is significant or (2)
21、 a binary or ternary refrigerant blend is used (Cavestri 1994). A signif- icant amount of analytical work with binary and ternary HFC refrigerant blend and synthetic lubricants has been previously reported (Cavestri 1994; Cavestri and Falconi 1999). Yet, with this being said, low-temperature refrige
22、rant miscibility can have viscosity reduction not by vapor but by liquid refrigerant. In this case, the pressure is at saturation with a flat pressure line but increasing refrigerant soluble in the lubricant and decreasing viscosity at lower temperatures. This happens to a point when low temperature
23、 and saturation equi- librium take over. To minimize liquidliquid viscosity reduc- tion, viscometer pressure is kept slightly below saturated conditions. METHODS Viscosity Determination Viscosity and density are accurately determined by a viscometer/densimeter whose design and construction were patt
24、erned after published information (White and Solomon 1965; Solomon and White 1968; Nissen 1980; Nissen and MacMillan 1983; Nieuwoudt and Shankland 1991). All viscometer actuation, controls, and calculations are done by microprocessor. This viscometer consists of an oscillating body device enclosed i
25、n a low-volume pressure tube (Cavestri 1993b). The system is contained within a stainless steel pipe. A highly polished, solid stainless steel, oscillating cylindrical bob is suspended from a spring, which is kept at a constant temperature to maintain its restorative properties. An internal steel tu
26、be core is mechanically connected to the cylindrical bob by a very fine wire. An external electromagnet causes the bob to oscillate until it gradually stops. This decaying sinuso- idal motion is termed the decrement. Position and movement of the bob are determined by a linear variable differential t
27、ransformer (LVDT). The viscosity of the solution is measured by calculating the logarithmic decrement of the bobs oscilla- tions after it is put into motion. Rapid decay indicates high viscosity; slow decay indicates low viscosity. This viscometer allows a wide range (0.50 to 150 cP) of viscosity de
28、terminations with a single oscillating bob. Addi- tionally, the vertical position changes of the bob can determine density to within 0.0005 g/ml. Kinematic viscosity (cSt) is determined from density (g/ml) and absolute viscosity (cP). The viscometer was calibrated for viscosity and density by using
29、known, pure dry fluids. This provides a straight calibra- tion line for density at specified temperatures. The precision of this determination is 0.3%. The viscosity and density standards used were water and certified hydrocarbon standard test fluids that are NIST trace- able. These standards were s
30、upplied calibrated from -25C to 125C (-13F to 257“F), reported in both CP and cSt values, and obtained from Cannon Laboratories as certified test fluids. The density certification was an internal check for viscometer calibrations. The viscometer is readable to 0.06 cP. Calibration meth- ods provide
31、viscosity results with a significant overlap that serves as an internal standard and self-check. Once the viscometer is fully calibrated over its intended measurement range, the error for low viscosity solutions (0.4 to 68 cSt) is 0.5%. This error increases to 1.5% for high viscosity solutions (500
32、cSt). Viscosity, density, and gas solubility measurements are taken when the lubricant and refrigerant gas (vapor) system has reached equilibrium. The lubricant and refrigerant solu- tion mixture is pumped from the bottom of the viscometer and sprayed into the refrigerant vapor space at specific pre
33、ssures and temperatures. Density is monitored for equilibrium condi- tions, the decrement is measured, and the viscosity is calcu- lated. Before the fluid is sampled, it is visually examined several times to ensure that a true solution exists. One data point is an average of six separate density and
34、 viscosity decre- ment determinations made before a liquid sample is taken to determine refrigerant concentration in the lubricant. The fluid sample is drawn through a very low volume capillary line (380 pl) into a deeply evacuated, lightweight glass sampling bulb where a total charge of a refrigera
35、ndlubricant sample is 54 ASHRAE Transactions: Research retained. The ratio of gas to liquid oil (percentage by weight) is then measured. The concentration measurement is repro- ducible to within 0.5% by weight at a given isothermal pres- sure test point in the viscometer. The viscometer is equipped
36、with three sight glasses. In the fluid measurement portion of the viscometer, two sight glasses are adjacent to the suspended, solid stainless steel, cylindrical bob. These two sight glasses are continually used to monitor the solution for the formation of any immiscible layers of lubricant and refr
37、igerant. The operator can also ensure that the viscometer is charged with enough fluid to cover the bob completely. The lowest of these two sight glasses is adjacent to the gas introduction and oil sampling ports. The third sight glass is located above the vapor space and adjacent to the pump exit.
38、This allows observation of the foaming qualities of the lubricant and a view of the liquid level to ensure that the contents are below the top level of the constant temperature circulating fluid. When refrigerant blends are used, the composition of the gas in the vapor space of the viscometer will d
39、etermine the equilibrium concentration (fractionation) of refrigerant solu- bilized by the lubricant. Therefore, it is necessary to maintain a constant gas composition of the subject refrigerant through- out the fluid property measurements. This is achieved by purg- ing the gaseous blend through the
40、 lubricant to maintain equality (*2%) with the pure liquid refrigerant blend intro- duced into the viscometer under pressure. Therefore, the composition of the solubilized gases in the lubricant is at equi- librium with a specified gas mixture at temperature and pres- sure. The fractionation of the
41、mixed gases in solution in the lubricant is measured by gas chromatography at every temper- ature and pressure test point by sampling the mixture and determining the total percent refrigerant by weight in the lubri- cant (Cavestri 1995). Density, viscosity, and vapor pressure are measured under isot
42、hermal conditions. The viscometer temperature is main- tained by a circulating, constant temperature glycol bath using a microprocessor controller equipped with a platinum resis- tance temperature device (RTD) (O. 1 “C/*0.2“F) mounted at the surface of the viscometer tube inside the liquid bath. The
43、 other temperature zones are controlled by electric heaters, using a microprocessor controller (O. 1 “C/*0.2“F) with type “J“ thermocouples. Equilibrium refrigerant vapor pressure is measured using a Heise, double-helix Bourdon tube gauge, model #CMM-110733. This temperature-compensated gauge is cal
44、i- brated with both gas and liquid, is accurate to *0.2 psia (0.001 3 MPa), and is traceable to NIST standards. Because the gas content of the fluid is measured, the amount of gas contained in the Bourdon tube or other parts of the viscometer is irrelevant. The lubricanthefrigerant mixture is pumped
45、 through the viscometer by a magnetically coupled impeller located in the base of the instrument comprising the pump body. To ensure equilibrium, the lubricanthefrigerant mixture is sprayed into a soluble gas vapor space at the top of the viscometer. For equi- librium and thermodynamic reasons, it i
46、s important that viscosity and concentration measurements be taken when the gas (vapor) composition is at equilibrium and at isothermal and isobaric conditions. Therefore, the pressure and tempera- ture in the instrument is varied to simulate the lubricanthefrig- erant pair conditions that exist in
47、operating compressor systems. Prior to use, the test fluid is degassed at 60C (140F) at 20 millitorr (3 Pa) for 24 hours and dried to 150 ppm water. Similarly, the viscometer is then evacuated to 20 millitorr (3 Pa) for several hours, purged with the refrigerant gas several times, and then evacuated
48、. The test lubricant is then drawn through the oil charging valve and reevacuated to 20 millitorr (3 Pa). The lubricadrefigerant mixture is purged with refrig- erant gas and again evacuated. Noncondensable gas content is not allowed to exceed 10 ppm. Refrigerant Blend Sampling The composition of the
49、 solubilized refrigerant in the lubricant is determined by gas chromatography. Stated briefly, the refrigerant is separated from the lubricant by careful heat- ing and samples are taken after equilibrium is established in the gas chamber. Known GC calibrated response curves estab- lished for the individual refrigerants enable determination of the gas composition. A full description of the apparatus and method are reported elsewhere (Cavestri 1995; Cavestri and Schafer 2000). Viscosity Reduction Viscosity reduction is caused by solubilized refrigerant gas in the lubricant a
