Performance Analysis of a Twin Rotary Compressor

Purdue University Purdue e-pubs International Compressor Engineering Conference School of Mechanical Engineering 2004 Performance Analysis of a Twin Rotary Compressor Hyun Jin Kim University of Incheon Jong Min Ahn University of Incheon Seon Woong Hwang LG Electronics Myung Kyun Kiem LG Electronics Follow this and additional works at: http://docslibpurdueedu/icec Kim, Hyun Jin; Ahn, Jong Min; Hwang, Seon Woong; and Kiem, Myung Kyun, "Performance Analysis of a Twin Rotary Compressor" (2004) International Compressor Engineering Conference Paper 1682 http://docslibpurdueedu/icec/1682 This document has been made available through Purdue e-pubs, a service of the Purdue University Libraries Please contact epubs@purdueedu for additional information Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W Herrick Laboratories at https://engineeringpurdueedu/ Herrick/Events/orderlithtml

C038, Page 1 PERFORMANCE ANALYSIS OF A TWIN ROTARY COMPRESSOR Hyun Jin Kim 1, Jong Min Ahn 2, Seon Woong Hwang 3, Myungkyun K Kiem 4 1,2 Department of Mechanical Engineering, University of Incheon, Incheon, Korea 3,4 Digital Appliance Research Lab, LG Electronics, Seoul, Korea 1 Corresponding author, E-mail) kimhj@incheonackr ABSTRACT For a twin rotary compressor, a computer simulation program has been developed With R22, predicted cooling capacity, compressor input and EER were found to be in good agreement with experimental results Calculated P-V diagram was also compared well to that of measurement With the simulation program, parametric study has been performed to investigate effects of cylinder dimensions on R410A twin rotary compressor performance Over the ranges of cylinder dimensions studied, some variations were found for volumetric, adiabatic and mechanical efficiencies, but overall compressor performance did not change much Discharge port configuration was found to have large effect on adiabatic efficiency, but not much on volumetric efficiency 1 INTRODUCTION While scroll compressors are widely used for air conditioners, there is a certain range where rolling piston rotary compressors exhibit better performance With R22, the capacity range favorable for rolling piston rotary compressors has been known up to small capacity residential air conditioners It has been shown, however, that the range in which rotary compressor operates more efficiently than scroll compressor expands up to considerably larger capacity when R22 is replaced by R410A(Kato et al, 1996) But, there are some limitations on increasing capacity by increasing either the cylinder size or operation speed for a single cylinder rotary compressor, because there occurs significant increase in vibration in accordance with capacity or speed increase For rolling piston type rotary compressors, there are two major sources of vibration: shaft whirling and torque load variation Adoption of two cylinder configuration is a good way of improving vibration characteristics For twin rotary compressors, mass unbalance due to shaft eccentric does not take place so that only relatively small counterweights are used, resulting in a great suppression of rotor top whirling(hattori and Kawashima, 1990 ; Okoma et al, 1990) Twin rotary also has good advantage over single rotary in reducing torque load variation Since gas compression processes in two cylinders occur with 180 o phase difference to each other, torque fluctuation becomes quite small for twin rotary compressors(kageyama et al, 2002) Some papers have been reported on performance of twin rotary compressors(saitoh et al, 1992 ; Hayano et al, 1996 ; Fujita,1998) In this paper, computer simulation program for a twin rotary compressor with its validation will be introduced And effects of design parameters on the compressor performance will be investigated by the computer simulation program to obtain optimum design for a two-cylinder rotary compressor 2 COMPUTER SIMULATION PROGRAM AND ITS VALIDATION A computer simulation program has been developed based on thermodynamic equations describing the refrigerant states in various control volumes and equations of motion for moving elements such as rollers, vanes, and crankshaft Gas pressure in a control volume is given by Equation (1) p M g s p = ρ s Vg M o / ρo n (1)

C038, Page 2 Masses of gas and oil in a control volume are calculated by considering mass flow rates between adjacent control volumes as in Equations (2a) and (2b) M g = M g ( 0) + Σ m & gdt, M o = M o ( 0) + Σm& o dt (2a),(2b) Various leakage paths are shown in Fig 1 Detailed descriptions on the individual leakage flows are found in Kim (2003) Flushing effect of refrigerant gas resolved in oil due to pressure and temperature changes across a leakage path is also considered For solubility change α across a leakage path, the rate of refrigerant gassing resolved in the oil flow rate m& is given by Equation (3) o m &, = m& o α (3) g flash Reaction forces and associated mechanical losses at various sliding surfaces are obtained by solving equations of motion for moving elements Dynamics of moving elements such as roller, vane, and crankshaft for a single or twin rotary compressor have been well documented in previous works(yanagisawa and Chu, 1982 ; Jun, 2002) Simulation program coding has been made by using visual basic Input data for the program include compressor operating conditions, type of refrigerant with mated oil, compressor dimensions, discharge port details, and clearances at various locations Calculation results are presented in tabular forma or in graphical modes Validation of the program has been made by testing of a twin rotary compressor model in a compressor calorimeter Fig 2 shows a schematic of a twin rotary compressor model instrumented with sensors, and raw sensor signals are illustrated in Fig3 Table 1 shows comparison of simulation results with testing data Excellent agreement between them has been obtained Simulated and measured P-V diagrams are also compared in Fig 4 3 R410A TWIN COMPRESSOR DESIGN BY USING THE SIMULATION PROGRAM For optimum cylinder design of a R410A twin compressor, cylinder bore diameter D, cylinder height H, and roller eccentricity e are taken as three design parameters These three parameters are related for displacement V c as in Equation (4) Vc H = (4) π e( D e) Top Flange II meb mvt mec mvb m54 m Top Flange I Middle Flange II mcb ṁrb ṁrc Middle Flange I Dyanmic Sensor 1 & 2 Bottom Flange II Static Sensor Adapter Bottom Flange I SIGNAL ADAPTER SIGNAL FRAME Fig 1 Leakage paths in the cylinder Fig 2 Twin rotary compressor test model

C038, Page 3 Table 1 Comparison of simulation with experiment for R22 twin rotary Description Basic model Instrumented model Simulation m&, Flow rate 1 1003 1008 W, Comp Input 1 1012 1007 c Q c, Cooling capacity 1 1001 1003 EER, EER 1 991 996 Voltage [V] 12 8 4 0-4 -8-12 Dynamic sensor(suction chamber) Dynamic sensor(compression chamber) Static sensor(lower tube) Static sensor(upper tube) Gap sensor(valve motion) Gap sensor(vane motion) Pressure [MPa] 25 20 15 05 :Simulation :Measurement P ad 000 002 004 006 008 010 time [sec] 0 5 10 15 20 25 Volume [cc] Fig 3 Raw signals of various sensors from test compressor model Fig 4 Comparison of P-V diagram between measurement and simulation (a) v (b) ad H 08 06 0974 0984 e= 8 [mm] e= 7 [mm] e= 6 [mm] e= 5 [mm] 0985 0993 0993 0991 0998 0995 08 06 15 10 14 04 09 16 05 11 17 084 088 092 096 0 (c) mech 084 088 092 096 0 (d) v ad mech H 08 06 07 05 02 03 01 0998 0999 0997 08 06 0996 0998 02 0999 04 05 05 06 0993 05 084 088 092 096 0 084 088 092 096 0 D Fig 5 Effects of cylinder dimensions on efficiencies of R410A twin rotary D

C038, Page 4 0004 0033 0003 m& 42 mdot [kg/s] 0002 0001 0000 m& cb m& vb -0001 m& vt m& rc -0002 0 90 180 270 360 450 Crank angle [deg] Fig 6 Various leakage flows form compression chamber Range of cylinder bore diameter is limited by boundary conditions Since displacement volume of R410A compressor is about 72% of that of R22 compressor to give the same capacity at ASHRAE-T condition, cylinder bore of a R22 twin rotary of the same capacity can be taken as an upper limit for R410A cylinder bore Lower limit for the cylinder bore is set from the crankshaft diameter as in Equation (5) D / 2 R + 2e + (5) sh t r Over certain ranges of cylinder design parameters for the same displacement volume, compressor efficiencies have been calculated by using the simulation program, and presented in Fig 5 Cylinder height and bore diameter are normalized by those of R22 twin rotary compressor, respectively From Fig 5(a), volumetric efficiency v does not increase with decreasing cylinder height Rather it is more or less constant Leakage mass flows in the upper compression chamber are shown in Fig 6 Leakage through the flank clearance between roller and cylinder m& cb is larger than any other leakages until the sealing line reaches discharge port, after which leakage from clearance volume of discharge port to the following suction chamber m& 42 becomes dominant With decreasing cylinder height, m& cb proportionally decreases, but m& 42 increases due to increasing leakage area between discharge port and the following suction chamber It is because roller radius becomes smaller for the same displacement with constant Cylinder bore As a result, amount of overall leakage does not vary much with changing cylinder height Pressure [MPa] 40 35 30 25 20 15 13 12 11 I l IV l Case D H 1 085 1 056 IV I I IV 0 90 180 270 360 Crank angle [deg] 350 360 370 380 390 400 410 Fig 7 Pressure rise at the beginning of compression process

Relative mechanical loss 12 08 06 04 02 00 L mech L s L c L v L j 055 060 065 070 075 080 085 H C038, Page 5 Table 2 Performance comparison between R410A and R22 twin rotary compressors R410A/R22 Notation 50Hz 60Hz v 958 966 ad 999 1040 mech 999 996 c 998 1036 EER 932 977 From Fig 5(b), adiabatic efficiency ad increases with decreasing cylinder height As mentioned above, m& 42 Fig 8 Mechanical loss vs cylinder height increases with decreasing H, and less gas of high pressure remains in the clearance volume of discharge port, resulting in smaller pressure add-up at the beginning of the compression process in the following cycle This phenomenon can be seen in Fig 7 where variations of cylinder pressure, pc and pressure at the clearance volume, p cl are shown at two different cylinder heights From Fig 5(c), mechanical efficiency mech decreases with decreasing H As shown in Fig 8, with decreasing H, roller bearing loss L c and journal bearing loss L j decrease due to reduction in projection area of gas pressure, and friction loss at vane slot Ls increases because of increased vane speed with larger e As a result, overall mechanical loss increases slightly with decreasing H From Fig 5(d), overall compressor efficiency represented by the product of v ad mech doest not seem to depend much on the variation of cylinder height and bore diameter for the ranges under consideration The difference between the maximum and minimum values is only about 077% Discharge port size and porting angle also affect the compressor performance Effects of discharge port size and 2 2 Relative efficiency 1 0 099 098 ad v Relative efficiency 1 0 099 098 097 ad v 097 085 090 095 0 D p 096 30 45 60 θ p [deg] Fig 9 Effects of discharge port configuration on compressor efficiencies

C038, Page 6 porting angle on v and ad have been calculated and presented in Fig 9 With increasing port diameter, ad increases by 31%, and v decreases very slightly Discharge port diameter is limited by roller thickness Increasing porting angle θ p also brings similar effect, since effective discharging area becomes larger with increasing θ p In Table 2, compressor performances are compared between optimized R410A and R22 twin rotary compressors At the operation speed of 50Hz, compressor performance of R410A twin rotary is about 932% of that of R22 twin rotary, but it increases to 977% at 60Hz 4 CONCLUSIONS In the study on the performance of a twin rotary compressor, (1) A computer simulation program has been developed, and its validation has been achieved by comparing simulation results and experimental data for a R22 twin rotary compressor model (2) With this simulation program, parametric study has been performed to investigate effects of cylinder dimensions on R410A compressor performance (3) Over the range of 085 D 1 0 and 056 H 1 0, the compressor performance represented by the product of v ad mech did not change much Maximum variation observed was abut 077% (4) Effects of discharge port on ad are rather large, while v is not much affected : ad was increased by about 31% by increasing either port diameter or porting angle over the given ranges considered REFERENCES Fujita, S, 1998, Performance development of 2-cylinder rotary compressor for R410A, Proc Intern Symp on R22 and R502 Alternative refrigerants, p67-171, Kobe, Japan Hattori, H, Kawashima, N, 1990, Dynamic analysis of a rotor-journal system for twin rotary compressors, Proc Intern Comp Eng Conf at Purdue, p750-760 Hayano, M, Fukuta, T, Yajima, T, and Fujita, S, 1996, Performance evaluation of 2-cylinder rotary compressor for R410A, Proc Intern Symp on R22 and R502 Alternative refrigerants, p169-173, Kobe, Japan Jun, Y, 2002, Mechanical loss analysis of inverter controlled two cylinder type rotary compressor, Proc Intern Comp Eng Conf at Purdue, Paper No C5-6, Kageyama, K et al, 2002, Development of high-efficiency low-vibration rotary compressor for HFC-410A air conditioner, Proc Intern Comp Eng Conf at Purdue, Paper No C5-1 Kato, T, Shirafuji, Y, and Kawaguchi, S, 1996, Comparison of compressor efficiency between rotary and scroll type with alternative refrigerants for R22, Proc Intern Comp Eng Conf at Purdue, p69-75, Kim, H J, 2003, Development of computer simulation program for performance analysis of a twin rotary compressor, Univ of Incheon Technical Report Okoma, K, Tahata, M, and Tsuchiyama, H, 1990, Study of twin rotary compressor for air-conditioner with inverter system, Proc Intern Comp Eng Conf at Purdue, p541-547, Saitoh, K, Hagiwara, S, and Fujimoto, S, 1992, Development of high efficiency dual cylinder type rotary compressor, Proc Intern Comp Eng Conf at Purdue, p373-382, Yanagisawa, T, Chu, I, 1982, Motion analysis of rolling piston in rotary compressor, Proc Intern Comp Eng Conf at Purdue, p185-192