A Study on the Oil Supply System of a Rotary Compressor

Transcription

1 Purdue University Purdue e-pubs International Compressor Engineering Conference School of Mechanical Engineering 2 A Study on the Oil Supply System of a Rotary Compressor H. J. Kim University of Inchon E. S. Lee University of Inchon S. H. Kwag University of Inchon K. W. Yun United Technologies Carrier Corporation K. K. Min United Technologies Carrier Corporation Follow this and additional works at: Kim, H. J.; Lee, E. S.; Kwag, S. H.; Yun, K. W.; and Min, K. K., "A Study on the Oil Supply System of a Rotary Compressor" (2). International Compressor Engineering Conference. Paper This document has been made available through Purdue e-pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for additional information. Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at Herrick/Events/orderlit.html

2 A STUDY ON THE OIL SUPPLY SYSTEM OF A ROTARY COMPRESSOR H.J. Kim, E.S. Lee, S.H. Kwag University of Inchon, Dept. of Mechanical Engineering 177 Dowha-dong, Nam-gu, Inchon, ) g685@lion.inchon.ac.kr K. W. Yijll, K.K. Min United Technologies Carrier Corporation c/o Daewoo Carrier Corporation, Kwangju Korea ABSTRACT For a rolling piston type of rotary compressor, oil supply system consisting of individual lubrication elements such as pumps, oil passages, and sliding surfaces has been modeled by employing equivalent electric circuit. Numerical solutions of the oil network model include total oil flow rate into the shaft hole, oil flow rates to various sliding surfaces, and oil leaks through roller end clearances. Also, analysis has been made on the oil content in the suction and compression chambers and in discharge refrigerant. Validation of the numerical simulation has been made by measurement of the total oil flow rate into the shaft. NOMENCLATURE A area R resistance c bearing clearance, constant r radius c. discharge coefficient sf Sommerfeld Number D bearing diameter T temperature F correction factor for resistance X oil factor H height a solubility L bearing length P. roller angle m mass flow rate b roller end clearance N rps E bearing eccentricity n specific heat ratio, power index jj viscosity p pressure (j) angular velocity P, pressure ratio q side leak flow coefficient Q volume flow rate s head coefficient INTRODUCTION Lubrication is crucial to good reliability and high performance in a rotary compressor, as in other types of hermetic refrigeration compressors. For vertical rotary compressor which is the main stream of the rotaries, oil pumping power is produced by the shaft rotation, and the lubricating oil is drawn into the oil gallery inside the shaft and the oil is distributed into various sliding surfaces via several radial feeding holes. Some experimental and analytical studies on the individual lubrication elements such as shaft pump, radial pumps and various types of sliding surfaces of vertical rotary compressors have been reported[ 1-5], but very few studies have been carried out on the integrated oil supply system[6]. Lubrication elements whose performance characteristics are determined individually may change their characteristic values when they are integrated as one system, since individual operating conditions for each elements are set only after they are combined together as a system under real compressor operating condition. This study aims to introduce a modeling method of the integrated oil supply system, and to present computer simulation program for predicting the oil distributions inside a vertical rotary compressor. Purdue University, West Lafayette, IN, USA- July 25-28, 2 153

3 MODELING OF OIL SUPPLY SYSTEM A schematic oil supply system of a rotary compressor is shown in Fig. I. Inflow of the oil into the system is made at the oil cap hole of the shaft, and the oil flows out of the system through the main journal bearing, sub journal bearing, and the clearances between the roller end faces and the cylinder flat walls. Between the inlet and outlets, various lubrication elements such as pumps and oil passages and sliding surfaces constitute oil distribution network. To model the individual lubrication elements and the whole oil supply system, analogy between the pipe flow theory and the electrical circuit theory is employed. In the analogy, the pressure differential, volume flow rate and flow resistance of the flow system correspond to the voltage, electric current, and electric resistance, respectively. Shaft Pump As the shaft rotates, the flow inside the shaft is forced to rotate to yield a pumping head. The shaft pump head will be changed when there exist discharge flows out of the oil gallery inside the shaft through the radial discharge holes. The head variation with the flow rate is expressed in equation (1) Here, P; is the theoretical total head, and head coefficient of the shaft, s s~o, and c and n are experimentally determined constants. At the inlet of the shaft, the oil cap is necessary to support the pump head. The flow resistance of the oil cap can be written as (1) Q p 2 (C.Aj (2) Radial Pump The performance curve of the radial feeding holes and the resistance at the pumps are given by equations (3) and (4), respectively AP R - _rp_ - r p c Qn- rp_q_';,rpe rp (3), (4) Here, P; is the theoretical Euler head, and s,p, crp, n are to be determined by experiment. Roller End Clearance The oil flow through the roller end clearance and the associated flow resistance are given by equations (5) and ( 6), respectively. Q = 2Ho 3 Ap P. rb 6pln(R, I RJ 2H R = _Ap_ = _,6 p'---ln-'-( R_,_,_l R--'''-'-.) 27! rb Q 2H3 p ~ v (5), (6) Journal Bearing Oil flow in the journal bearing consists of pressure-driven flow and the side leakage flow due to viscous dragging of the rotating shaft. The pressure-driven flow, QP and the corresponding resistance are given in equation (7), and (8), respectively. 3 Q = 1lfJ.rc ( e') p 6pd ' (7), (8) The side leakage flow, Q., and the corresponding viscous pumping head, P. are given by equations (9) and (1 ), respectively. Purdue University, West Lafayette, IN, USA- July 25-28, 2 154

4 Here side leak coefficient ; is a function of Sommerfeld number and bearing aspect ratio[7]. (9),(1) Spiral Groove Grooves at journal bearings are to improve the oil supply through the bearings. When the groove is inclined, viscous action takes place to drag the flow in addition to the flow due to the pressure differential. The flow resistance in the groove, the flow due to viscous pumping Qvp and the viscous pumping head, P vp are given by equations ( 11 )-( 13 ), respectively. (11), (12),{13) Here, Fg and Cv can be obtained as functions ofreynolds number and groove shape with the aid ofcfd. Oil Supply Network-Electrical Analogue Fig. 2 shows an electrically analogous circuit to the oil flow network of Fig. I. Kirchoff's laws are applied to the joints and the loops to find currents at various passages. From the network, 23 equations can be obtained with 23 unknowns. OIL FLOWS IN SUCTION AND COMPRESSION CHAMBERS Fig.3 shows gas/oil flow transfers at suction and compression chambers. Expression of the individual mass flow rates and oil portions involved in the flows are listed in Table 1. When the main flow is regarded as compressed gas flow, the following general expression is used m = c p AJ 2n RT Jpu.- p +ll " u n-1 u r r (14) Viscous effect along the flow passage can be accounted in flow coefficient cv, In Table I, it is to be noted that the vaporization of the gas immersed in the oil has to be accounted when the oil flow experiences large variations in pressure and temperature during migration. EXPERIMENTAL RESULTS Measurements for lubrication elements' performance characteristics were all carried in water, since the viscosity of water at room temperature is rather close to that of the lubricating oil in real operation condition: The oil viscosity in operating condition is J.l == 1~2 cp, while that of water at room temperature is J.l w"" 1cP. In the following figures which show discharge coefficient, pump head, and oil flow rates, normalization by reference values was made on the ordinate. Fig.4 shows discharge coefficient of the oil cap, which is needed for the resistance calculation of equation (2). Over Reynolds number range tested, Cv is found more or less cons~t. Fig.5 shows variation of head of shaft pump with rotation speed under several different conditions. The presence of discharge holes reduces the head significantly, while the slinger increases the pumping head slightly. Free surface level also has small effects on the head. Head coefficient s s~o is obtained in Fig.6. Discharge flow rate of a radial pump depends on its relative position with respect to its neighboring pumps as in Fig.7: Upper pump has smaller discharge. And the discharge is found to be decreased at higher Reynolds number. Pumping head was not measured, but head coefficient was taken from known data source of other pump models[8] by averaging them. The amount of total oil flow into the oil supply system was measured by using oil flow meter as suggested by Itoh et al [6]. In the present measurement, refrigerant was R22 with synthetic oil as lubricant. And the oil return from lower journal bearing groove was made to not to re-circulate into the oil cap hole, but to be discharged directly into the oil sump as in Fig.8. Total oil flow rate measured at various operating conditions is given in Fig.9. Purdue University, West Lafayette, IN, USA- July 25-28, 2 155

5 CALCULATION RESULTS Computer simulation program for the analysis of the oil supply system has been made based on the previously mentioned oil network modeling. To compare the experimental results with analytical ones, the insertion of oil flow meter and related connection lines in the oil system has to be taken into account for the oil network modeling. These effects are regarded as additional resistances at the inlet line as in Fig.lO. Comparisons between the measurement and the analysis are presented in Table 2. The predicted values only differ from the measured ones by 5.8% at most, except at condition E. These results are plotted against the oil viscosity in Fig.11. It appears that the total oil flow rate almost linearly decreases with increasing the oil viscosity, indicating that some errors might be involved in the measurement at condition E, rather than on the side of analysis. The total oil flow rate in real compressor operations can be estimated by removing the flow meter connections from the oil network model. At operating condition A (ARI condition), the oil flow is found to be increased by 2%, when the flow meter is removed. To see the effect of the groove configuration on the oil flow rates, calculations were also made at six different groove shapes offig.12, and the results are shown in Fig.l3. As the groove shapes become flattened (I ~IV), the viscous pumping increases slightly, but the reduction in the pressure-driven flow is more rapid, and consequently the bearing flows decrease. This reduction of oil flows in the bearing outlets makes the total oil inflow decrease. Groove shapes V and VI have much smaller cross sectional area than the former shapes, and further reduction in the oil flows is induced. These trends are also reflected in the radial feedings at the lower and upper pumps. Oil flow rates and oil fractions into the suction and compression chambers are shown in Fig.14. Averaged value of the oil fraction of compression chamber can be regarded as the fraction of oil contained in the discharge gas flow. CONCLUSIONS The following conclusions are made from the present study on the oil supply system of a vertical rotary compressor. (I) Modeling of individual lubrication elements with experimental and analytical determination of related coefficients have been carried out. (2) The whole oil supply system consisting of various individual lubrication elements has been modeled by adopting analogous electrical circuit. (3) By using computer simulation program based on the oil network modeling, the total oil flow rates have been predicted at several different operating conditions, and good agreements between the predictions and the experiments have been obtained. (4) Effects of design changes in individual lubrication elements on the oil flow distribution in the system can be detected by the simulation program REFERENCES [I] Asanuma, H., Itami,T.,Ishikawa,H.,"An experimental study of the shaft oil supply mechanism of a rotary compressor," Proc. Oflntern. Compr. Eng. Conf. At Purdue, 1984, pp [2] Takebayashi, M., Iwata, H., Sakazume, A., "Discharge characteristics of an oil feeder pump using nozzle type fluidic diodes for a horizontal compressor depend on the driving speed," Proc. Of Intern. Compr. Eng. Conf. At Purdue, 1988, pp.l9-26 [3] Kim, K., Cho, K., "A study on lubricating system of hermetic rotary compressor," Proc. Of Intern. Compr. Eng. Conf. At Purdue, 1988, pp [4] Costa, C., Ferreria, R., Prata, A., "Considerations about the leakage through the minimal clearance in a rolling piston compressor," Proc. Of Intern. Compr. Eng. Conf. At Purdue, 199, p [5] Minami, K., Hattori, H., Hayano, M., "Lubrication analysis of rotary compressors for HFC refrigerants," Proc. Of Intern. Compr. Eng. Conf. At Purdue, 1998, pp [6] Itoh,T., Kobayashi,H., Fujitani,M., Murata,N., "Study on the oil supply system for rotary compressors," Proc. Oflntern. Compr. Eng. Conf. At Purdue, 1992, pp [7] Pinkus,., Sternlicht, B., "Theory ofhydrodynamic lubrication," McGraw-Hill, [8] Private communication with DCC Purdue University, West Lafayette, IN, USA- July 25-28, 2 156

6 Table 1 Mass flow rates in suction and dischan!e chambers Mixed Flow: Passage Oil portion (moil) Eq.(14) if not specified Suction port Sue. chamb. thru. vane side Roller/cylinder gap Vane tip Vane upper & lower faces Discharge port m 21 xcm Roller/cylinder gap mcb xcmcb Comp. chamb. thru. roller face f3v ~ 2n - flv, Eq.(7) Comp. chamb. m = c H(~ + (pd- Pc)& 2 ). [I ( )] thru. vane side ec Po v 2 I2p.L x.m.c - ad - ac Discharge port to suction chamber Suction chamber to suction port '[ a bl e 2 c ompanson o Operating Condition A B c D E F f 1 "I fl ows b etween pre d" IC f IOn an d expenmen t s qnp I q. qpred I qexp I Purdue University, West Lafayette, IN, USA- July 25-28, 2 157

7 7s 7c 7s 7c Fig. 1 Oil supply system of a rotary compressor Fig. 2 Electrically analogous circuit >.9. () > () fficb, ffi11 Fig.3 Oil flows at suction and compression chambers Note: Y -axes for Fig.4,5,6,7,9,11, & 13 are relatively indexed Re Fig.4 Discharge coefficient of the oil cap. X SLINGER FEEING FREE SURF HOLES LEVEL X 18. X A- 18 -T I /./ ~;./ ~~:~./ ~~ ~ ~/ - ~ ---==*~.= /' --== --~ Hz Fig.S Shaft pump head Purdue University, West Lafayette, IN, USA- July 25-28, 2 158

8 , -- iii.8 J?.6 >J'.4 - Jl_ ~ _. o_ u.8 * *, LMB -t/ ' ' Speed[Hz] 5 6 Fig.6 Head coefficient of shaft pump Re Fig.7 Discharge coefficient of radial pumps qexp I qo F D a "rs: bb eo.8 9 c 16 "d..747 p.. B 12 A c Ps [kg/cm 2 G] Fig.8 Total oil flow measurement apparatus Fig.9 Measurement of total oil flow rate , o-.8.6 ~ D F ~c A I~ Pred., Exp. E B ~[cp] Fig. to Resistance of flow meter device Fig.ll Comparison of oil flow rate between experiment and analysis Purdue University, West Lafayette, IN, USA- July 25-28, 2 159

9 ... Fig.l2 Various groove shapes...,qn: ' X ' Crank angle [ deg ] 36 Fig.14 Variation of oil flow and oil fraction at suction and compression chambers Purdue University, West Lafayette, IN, USA- July 25-28, " " cr<".2.1 Clrp2 qrp3 qrp4 Fig.l3 Effects of groove shape on the oil flow 16 ".8. IllillNV~ IITillNV~ IITillNV~ ~\. - - X sp