Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`2013-01-1496
`Published 04/08/2013
`Copyright © 2013 SAE International
`doi:10 4271/2013-01-1496
`saepcmech saejournals org
`
`Refrigerant and Lubricant Distribution in MAC System
`Shenghan Jin and Pega Hrnjak
`Creative Thermal Solutions and ACRC,UIUC
`
`ABSTRACT
`This paper presents experimental results for refrigerant and lubricant mass distribution in a typical automotive A/C
`(MAC) system.
`Experiments were conducted by closing valves located at the inlet and outlet of each component after reaching steady
`state, isolating the refrigerant and lubricant in each component. Refrigerant mass is recovered in a separate vessel using
`liquid nitrogen to reduce refrigerant vapor pressure to near vacuum. The overall weight is determined within ±1% after the
`separation of refrigerant and lubricant. The mass of lubricant is determined by using three different techniques: Remove
`and Weigh, Mix and Sample, and Flushing. The total mass of lubricant in the system is determined with ±2.5% uncertainty
`on average. R134a and R1234yf are used with PAG 46 oil as working fluid at different Oil Circulation Ratio (OCR),
`ranging from 2% to 4%. Experiments are conducted in two standard testing conditions: I35 and L35 (SAE Standard J2765)
`[11].
`R134a and R1234yf exhibit similar results in terms of refrigerant and lubricant retention. Higher mass flow rate does
`not significantly change the refrigerant distribution, but a lower lubricant retention in the evaporator is observed. It is
`found that only about a quarter of the total mass of lubricant resides in the compressor.
`
`CITATION: Jin, S. and Hrnjak, P., "Refrigerant and Lubricant Distribution in MAC System," SAE Int. J. Passeng. Cars -
`Mech. Syst. 6(2):2013, doi:10.4271/2013-01-1496.
`____________________________________
`
`INTRODUCTION
`In a vapor compression refrigeration system oil is crucial
`to provide sealing and lubricating for the compressor. In an
`automotive A/C system where an oil separator is integrated
`into the compressor if at all present, oil is inevitably mixed
`with refrigerant and travels throughout the system due to a
`separation efficiency of less than 1. The circulating oil
`appears in the form of mist, droplets, a thin oil rich film, or a
`fairly homogeneous liquid mixture with refrigerant depending
`on its location and the system configuration. The presence of
`oil can significantly change the properties of the working
`fluid, degrade heat transfer, and increase pressure drop. A
`large amount of oil retention outside the compressor can
`cause compressor failure due to insufficient lubrication. As a
`result, the amount of oil in circulation and oil retention has a
`direct influence on the system performance and compressor
`reliability [14].
`In the literature, experimental techniques used to measure
`a refrigerant and lubricant mass inventory can be grouped by
`whether the method is intrusive or not. By using the intrusive
`method, Hoehne and Hrnjak [1] were able to measure
`refrigerant mass in less than 0.1 g uncertainty in each section
`of a low charge (<150g) hydrocarbon (propane) refrigerant
`
`system. Peuker and Hrnjak [2] measured refrigerant mass
`distribution in an automotive A/C system with orifice tube
`(OT) and low side accumulator in transient and steady state.
`A measurement uncertainty of 0.4%
`regarding
`total
`refrigerant mass was achieved by using liquid nitrogen as the
`recovery coolant. Björk [3] used an expansion tank where
`refrigerant can reach a superheat thermodynamic equilibrium.
`Using the p-v-T relationship and internal volume, refrigerant
`mass can be calculated. According to the comparison, the
`difference between Peuker's and Björk's procedures is small,
`ranging from 1% to 5%. The non-intrusive measuring
`technique, also known as On-Line Measuring Technique,
`weighs a section directly while the system is still running.
`According to Miller's [4] result on a 3-ton R22 split-system
`air-to-air heat pump, the accuracy for the weighing system is
`about 0.05 kg.
`Various authors have reported lubricant retention results
`using intrusive measuring technique. Crompton et al [5]
`employed a Remove and Weigh technique to investigate oil
`retention in smooth and micro-fin tubes. Similar to the
`intrusive method mentioned previously for refrigerant, valves
`at both ends of
`the section of
`interest were closed
`simultaneously. After refrigerant was slowly released, the
`section was removed from the test loop and weighed so that
`
`1013
`
`Page 1 of 8
`
`Arkema Exhibit 1102
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`oil retention is determined. The average uncertainty was
`4.9%. Zoellick and Hrnjak [6] and Sethi and Hrnjak [7]
`employed a similar method to measure oil retention in suction
`lines. The uncertainties reported by the two authors were
`0.5% and 2%, respectively. Except for remove and weigh
`technique, Peuker and Hrnjak [2] developed Flushing
`technique and Mix and Sample technique. Using these three
`methods, Peuker and Hrnjak [2] were able to determine oil
`amount totals within 2%. As for the non-intrusive method,
`Lee [8] used an injection-extraction method to investigate oil
`retention in a vertical suction line. Oil was injected at the
`bottom of a pure refrigerant suction line and was separated at
`the top by using oil separators. The oil retention was the
`difference between mass of oil injected and that of oil
`extracted. This method was later used to measure oil retention
`in each part of a CO2/PAG 46 A/C system by Lee [9] in 2003
`and that of a residential A/C system by Cremaschi [10] in
`2004. Cremaschi [10] estimated 12% relative error.
`EXPERIMENT SETUP
`In the facility the system components are located at the
`same relative heights as they are in the original vehicle. The
`compressor is a piston-cylinder reciprocating compressor
`with fixed displacement. No oil separator is installed. A
`micro-channel condenser with 48 tubes and a plate and fin
`evaporator with 19 plates are used in the facility, as shown in
`Figure 3 and 4. The condenser does not have a separate pass
`for subcooling. An external receiver is installed after the
`condenser, with an internal diameter of 13.08 mm and a
`length of 560 mm. A thermal expansion valve (TXV) is used
`as throttling device. The schematic of the facility and details
`of the system and components is shown in Figure 1 and 2.
`Due to the large size of the environment chambers, longer
`tubes than in the real system are used to connect the
`components. As a result, the experimental system volume is
`larger than original system and the largest difference in
`tubing length is from condenser to evaporator, or liquid line.
`Two methods are used to measure the system volume: one
`is based on combining the volume calculated from tube
`geometry with the internal volumes of components provided
`by manufacturers; the second is to charge a known amount of
`gas and calculate volume based on density, also referred to as
`the isothermal gas method.
`To avoid the diffusion of gas, lubricant is removed from
`the system by flushing with soluble refrigerant before using
`the isothermal gas method. Each section is then evacuated
`and charged with a certain amount of the designated gas, the
`weight of which is measured by the scale. Gas with higher
`density is preferred since the scale will have better accuracy
`if the charged mass is larger. Carbon dioxide, with a stated
`purity of 99.9% as supplied, is selected. Equilibrium
`temperature and pressure are measured to calculate density
`and the saturation temperature is also calculated to make sure
`the gas is superheated.
`
`The internal volumes calculated using two methods agree
`with each other, as shown in Table 1. The isothermal gas
`method has a consistent but slightly larger value than
`geometrical method as expected, due to the “inactive
`volume” such as pressure taps and charge ports. Since the
`inactive volume can hold refrigerant mass as well, the values
`of isothermal gas method are used.
`The operation condition is mainly focused on idle
`condition, i.e. minimum compressor speed. So the I35-dry
`condition from SAE J2765 [11] is investigated as the design
`condition. Charge tests are performed at this condition to
`determine the refrigerant charge for maximized COP for both
`R134a and R1234yf. The same type of lubricant (PAG 46) is
`used in all tests but varying charge amounts are employed to
`change the OCR. Since the lubricant used in this study is
`essentially oil, the term “lubricant” and “oil” are used
`interchangeably. The system is also operated at higher
`compressor speeds (L35) to investigate the effect of higher
`mass flux.
`
`Figure 1. Schematic of test facility
`
`Figure 2. Schematic of TXV-Receiver system
`
`1014
`
`Page 2 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`(1)
`
`(2)
`The enthalpies are calculated based on temperature and
`pressure measurements. However, refrigerant is usually two
`phase at the exit of the TXV so the enthalpy at the inlet of the
`evaporator is calculated assuming an isenthalpic throttling
`process. The air side and refrigerant side heat balance are
`within ±5% error with their average, respectively, as shown
`in Figure 5 and Figure 6.
`
`Figure 3. Micro-channel condenser
`
`Figure 5. Comparison of indoor heat balance
`
`Figure 4. Plate and fin evaporator
`
`Table 1. Internal volume of TXV system
`
`UNCERTAINTY ANALYSIS
`Applicable standards require at least two independent
`energy balances to experimentally determine steady-state
`system performances. The test facility for this system has a
`refrigerant-side energy balance (Qref) and an air-side energy
`balance (Qair) for both indoor and outdoor chambers,
`according to SAE J2765 [11] and ASHRAE 41.2 [12]. The
`two energy balances are measured and calculated as
`following:
`
`Figure 6. Comparison of outdoor heat balance
`
`in measuring oil
`involved
`Two uncertainties are
`distribution. One is the uncertainty from the instrumentation
`as listed in the Appendix. Both the accuracy of property data
`and the mass flow measurement affect the uncertainty of the
`system performance. In addition, the accuracy of the scale
`can directly influence the results of measured mass retention.
`Another uncertainty lies on the time lag of closing the
`valves simultaneously. The valves are operated manually by
`
`1015
`
`Page 3 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`using the chamber lights as the signal. Since each individual
`has a different reaction time, there is inevitably a time lag in
`closing the valves. It was reported that the average reaction
`time of a university student exposed to light stimuli is 0.24
`seconds [13]. So an uncertainty of 0.3 seconds is used to
`estimate the uncertainty of closing valves as following:
`
`Additionally, subcooling at the outlet of condenser and
`superheat at the of evaporator outlet are calculated from
`Equation 8 and 9.
`
`(8)
`
`(9)
`Figure 7 and Figure 8 present air side cooling capacity,
`the COP, the apparent degree of subcooling at condenser
`outlet, and the apparent degree of superheat at evaporator
`outlet for the TXV system. Both COP and cooling capacity
`reach maximums at the refrigerant charge of 1350 g for
`R134a and 1435 g for R1234yf, which is regarded as the
`optimum charge for the TXV-Receiver system. At their
`optimum charge, R1234yf has higher subcooling than R134a,
`which results in the larger amount of charge. During the tests,
`TXV setting remained the same for both refrigerants at 15 °C.
`
`Figure 7. Charge test with R134a
`
`(3)
`
`(4)
`The combined standard uncertainty is calculated from the
`propagation rule as indicated by Equation 5 and results are
`presented in Figure 9 and 10.
`
`(5)
`
`RESULTS AND DISCUSSION
`The refrigerant charge is one of the factors that affect
`system performance. In the current study, all experimental
`data are taken at the refrigerant charge to maximize the
`Coefficient of Performance (COP).
`The charging process starts with a low charge, which is
`indicated by a high superheat at the outlet of the evaporator.
`A small amount of refrigerant is added in steps at the inlet of
`the evaporator. After the system is stabilized, system
`performance is recorded.
`Since the system is undercharged at first, the refrigerant
`flow through the mass flow meter is not single-phase at all
`times. The mass flow measurement is not accurate for two
`phase flow, so the cooling capacity is determined based on air
`side measurement, as indicated by Equation 6.
`
`(6)
`The air mass flow rate is calculated based on calibrated
`nozzle curve and air enthalpies are calculated based on dry-
`bulb temperature, pressure, and dew point temperature. COP
`is calculated as the ratio of cooling capacity and compressor
`shaft power. Other powered equipment, such as air blowers,
`are not included.
`
`(7)
`Compressor work was determined by measuring torque
`and speed directly on the compressor shaft, between the
`compressor and the drive belt pulley.
`
`With the refrigerant charge determined from charge test,
`three amounts of oil charge, 145g, 175g and 205g, are used
`for both R134a and R1234yf. For each of the charges, the
`mass of refrigerant and lubricant distribution are measured at
`
`Figure 8. Charge test with R1234yf
`
`1016
`
`Page 4 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`the I-35 condition in the TXV-receiver system. During the
`measurement, the system is first running at steady state and
`operating parameters are recorded. Temperature, pressure and
`mass flow rate are measured to calculate heat balance
`between refrigerant side and air side. Valves are closed
`simultaneously and the system is shut off. The measuring
`technique developed by Peuker and Hrnjak [2] is employed to
`measure refrigerant and oil mass trapped in each component.
`Since the volume of liquid line in the experimental system is
`6 times larger than the original system and the volume is
`taken by single phase liquid. A factor of 1/6 is taken for both
`refrigerant and oil mass in liquid tube due to the extra
`volume. The results are shown in Figure 9 and 10.
`
`with R134a then with R1234yf, but with both refrigerants, the
`compressor oil mass is less than 25% of the total oil charge.
`In the TXV system, oil distribution in each component
`seems to be in proportion with total oil charge. However, this
`is a different scenario compared with the result in the OT
`(with an internal diameter of 1.823mm) system. As concluded
`by Peuker and Hrnjak [5], in the OT system with the same
`components except for an extra accumulator, increasing
`refrigerant charge while holding total oil charge constant
`results in shifting oil from other components into the
`accumulator. To make comparison between these two
`systems, the total oil and refrigerant charge is normalized as
`the ratio of total oil in the system. Oil mass distribution in
`percentages is listed in Tables 2 and 3. In comparison, as
`charge ratio decreases, oil tends to shift into accumulator in
`OT system; while in the TXV system oil does not exhibit an
`obvious trend of shifting.
`
`Table 2. OT system oil mass distribution in percentage at
`different charge ratio for fixed oil charge
`
`Figure 9. Mass distributions of refrigerant and lubricant
`in TXV system with R134a
`
`Table 3. TXV system oil mass distribution in percentage
`at different charge ratio for fixed refrigerant charge
`
`Figure 10. Mass distributions of refrigerant and
`lubricant in TXV system with R1234yf
`
`According to Figures 9 and 10, both refrigerants exhibit
`similar
`trends
`in
`terms of
`refrigerant and
`lubricant
`distribution. Oil mass in each component increases with total
`oil charge. However, this increasing oil mass has a minor
`effect on refrigerant distribution, as refrigerant mass remains
`nearly the same in each component. The amount of oil in the
`compressor is slightly higher when the system is operating
`
`the comparison between mass
`Figure 11 presents
`distribution in the OT and TXV systems. The components
`compared are identical. The corresponding system operation
`conditions are listed in Table 4. In the TXV system, more
`liquid refrigerant is retained in the condenser than in the OT
`system as indicated by the much higher apparent subcooling
`at the condenser outlet. As a result, more oil is held in the
`condenser in TXV system. Another difference is the much
`higher oil retention in the evaporator in TXV system. This
`might have been caused by the higher superheat which leads
`to a lower refrigerant solubility with oil in the evaporator
`superheated region. The oil return mechanism in this region is
`primarily due to the shear stress of refrigerant vapor [7]. As
`
`1017
`
`Page 5 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`oil concentration becomes greater, the liquid film would be
`much more viscous leading to higher oil retention.
`
`Figure 11. Comparison of mass distribution in the same
`components of different system configurations
`
`Table 4. Systems operation condition
`
`Figure 12. Mass distribution at different compressor
`speed with R1234yf
`
`Figure 12 presents the mass distribution results at low
`compressor speed (I35) and high compressor speed (L35).
`According to the results, at higher compressor speed the
`increased mass flow rate has a very minor effect on
`refrigerant mass distribution; however it significantly affects
`oil retention in the evaporator and suction line. Oil retention
`decreases by 20% in the evaporator and 79% in the suction
`line. A higher mass flow rate may have a strong “wash-out”
`
`2.
`
`effect on the superheated region. As observed by Sethi and
`Hrnjak [7], in the horizontal suction line, higher mass flux
`leads to lower oil retention.
`SUMMARY/CONCLUSIONS
`Refrigerant and lubricant mass distribution in a typical
`automotive A/C (MAC) system is experimentally measured
`at steady state. The two tested refrigerant, R134a and
`R1234yf, exhibit similar results in terms of refrigerant and
`lubricant retention. Slightly higher oil retention is found in
`the compressor with R134a but for both refrigerants, less than
`a quarter of the total mass of lubricant resides in compressor.
`Refrigerant distribution has more significant effect on oil
`distribution than oil on refrigerant. In the condenser, oil
`retention is a strong function of liquid refrigerant charge. In
`the evaporator, where the refrigerant charge is about the
`same, oil retention is much higher when the system operates
`with higher superheat.
`A higher mass flow rate does not significantly change the
`refrigerant distribution, but lower lubricant retention in the
`evaporator is observed.
`REFERENCES
`1. Hoehne M. R. and Hrnjak P. S.. “Charge minimization in systems and
`components using hydrocarbons as a refrigerant”. Technical Report
`ACRC TR-224, University of Illinois at Urbana-Champaign, 2004.
`Peuker S. and Hrnjak P.S.. “Experimental and analytical investigation of
`refrigerant and lubricant”. Technical Report ACRC TR-277, Air
`Conditioning and Refrigeration Center, University of Illinois at Urbana-
`Champaign, Urbana, IL, 2010.
`3. Björk E.. “A simple technique for refrigerant mass measurement”.
`Applied Thermal Engineering, 25(8-9):1115-1125, 2005.
`4. Miller W. A.. “The laboratory evaluation of the heating mode part-load
`operation of an air-to-air heat pump”. ASHRAE Transactions, 91(2B):
`524-536, 1985.
`5. Crompton J. A., Newell T. A., and Chato J. C.. “Experimental
`measurement and modeling of oil holdup”. Technical Report ACRC
`TR-226, University of Illinois at Urbana-Champaign, 2004.
`6. Zoellick K.F., and Hrnjak P.S., “Oil retention and pressure drop in
`horizontal and vertical suction lines with R410A/POE”, Technical
`Report ACRC TR-271, Air Conditioning and Refrigeration Center,
`University of Illinois at Urbana-Champaign, Urbana, IL, 2010.
`Sethi A., and Hrnjak P.S., “Oil retention and pressure drop of R1234yf
`and R134a with POE ISO 32 in suction lines”, Technical Report ACRC-
`TR 281, Air Conditioning and Refrigeration Center, University of
`Illinois at Urbana-Champaign, Urbana, IL, 2011.
`8. Lee J.P., Hwang Y., Radermacher R., and Mehendale S. S.,
`“Experimental investigations on oil accumulation characteristics in a
`vertical suction line”, In Proc. of ASME IMECE 2001, Vol. 3, AES
`23607, New York, NY, 2001.
`9. Lee J.P., “experimental and theoretical investigation of oil retention in
`carbon dioxide air-conditioning system”, Ph.D. Thesis, CEEE,
`University of Maryland, College Park, MD, 2002.
`10. Cremaschi L., “Experimental and theoretical investigation of oil
`retention
`in vapor compression systems”, Ph.D. Thesis, CEEE,
`University of Maryland, College Park, MD, 2004.
`11. SAE International Surface Vehicle Standard, “Procedure for Measuring
`System COP [Coefficient of Performance] of a Mobile Air Conditioning
`System on a Test Bench,” SAE Standard J2765, Issued Oct. 2008.
`12. ANSI/ASHRAE Standard 41.2-1987 (RA92). Standard methods for
`laboratory airflow measurement. American Society of Heating,
`Refrigeration and Air-Conditioning Engineers, 1992.
`13. Gupta S.. “Simple visual reaction time, personality and strength of the
`nervous system: A signal-detection theory approach”. Personality and
`Individual Differences, 6(4):461-469, 1985.
`14. Gordon, T., Eustice, H., and Brooks, W., “Automotive AC System Oil
`Migration HFO-1234yf Vs. R134a,” SAE Technical Paper
`2011-01-1173, 2011, doi:10.4271/2011- 01-1173.
`
`7.
`
`1018
`
`Page 6 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`ACKNOWLEDGMENTS
`The authors thankfully acknowledge the support provided
`by the Air Conditioning and Refrigeration Center at the
`University of Illinois at Urbana-Champaign.
`
`ABBREVIATIONS
`MAC - Automotive A/C system
`Mr - mass flow rate, g/s
`OCR - Oir Circulation Rate
`ρ - Density, kg/m3
`cond - Condenser
`comp - Compressor
`evap - Evaporator
`OT - Orifice tube
`PAG - polyalkelene glycol oil
`TXV - Thermal Expansion Valve
`ref - Refrigerant
`sat - Saturated
`SC - Subcooling, °C
`SH - Superheat, °C
`tot - Total
`
`1019
`
`Page 7 of 8
`
`

`
`Downloaded from SAE International by Bianca Hamilton, Friday, December 18, 2015
`
`Jin et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)
`
`APPENDIX
`
`Instrument specification
`
`1020
`
`Page 8 of 8