1、International Journal of Heating,Ventilating, Air-conditioning and Refrigerating Research I ! Volume 1, Number 2, April 1995 1894-95 1 1994-95 International Journal of Heating, Ventilating, Air-conditioning and Refrigerating Research Editor Raymond Cohen, Ph.D P.E., Professor of Mechanical Engineeri
2、ng and Herrick Professor of Engineering, Purdue University, U.S.A. Associate Editors Arthur E. Bergles, Ph.D P.E., John A. Clark and Edward T. Crossan Professor of Engineering. Department of Mechanical Engineering, Aeronautical Engineering and Mechanics, Rensselaer Polytechnic Institute. U.S.A. Univ
3、ersity of Oxford, United Kingdom Fire Research Laboratory, National Institute of Standards and Technology, U.S.A. Arthur L. Dexter, D.Phi1 C.Eng University Lecturer, Department of Engineering Science, David A Didion, D.Eng., P.E., Leader, Thermal Machinery Group, Building and Ralph Goldman, Ph.D., S
4、enior Consultant, Arthur D. Little, Inc., U.S.A. Hugo Hens. Dr.Ir., Professor, Department of Civil Engineering, Laboratory of Building Physics, Katholieke Universiteit. Belgium Kea-IChi Kimura, Dr. Eng., Professor. Department of Architecture, Waseda University and President, Society of Heating, Air-
5、conditioning and Sanitary Engineers of Japan, Japan Universitt Hannover, Germany Universit de Liege, Belgium University of Wisconsin-Madison, U.S.A. University of California, Santa Barbara, U.S.A. Horst Kruse, Dr.-Ing., Professor, Institut fr Kltetechnik und Angewandte Wrmetechnik. Jean J. Lebrun, P
6、h.D., Professor, Laboratoire de Thermodynamique, John W. Mitchell. Ph.D., P.E., Professor, Mechanical Engineering. Dale E. kborg, Ph.D Professor, Chemical Engineering, Policy Committee Ronald J. Kessner. chair Frank M. Coda Eugene Stamper Fritz W. Steimle W. Stephen Comstock Raymond Cohen Editorial
7、Assistant Jenny OtletJakovljevic PubWher Rank M. Coda Publishing Director W. Stephen Comstock AS- Editorial and Publishing Services Staff Robert A. Parsons, Handbook Editor Nancy F. Thysell Adele J. Brandstrom Christina D. Tate Susan Boughadou 81995 by the American Society of Heating, Refrigerating
8、and Alr-Conditioning Engineers. Inc 1791 Tuille Circle, Atlanta. Georgia 30329. All rights resewed. Second class postage paid at Atlanta. Georgla. and additional mailing offices. WAC by which we do not mean to encourage mummified museum pieces-quite the contrary. But we do want to empha- 97 I intere
9、sted to get some feedback from our readers concerning their own research. ASHRAE TITLEUIJHVAC 1-2 95 0759650 0515458 663 98 HVACBrR RESmcn size that papers should be written to the strictest standards, not only scientifically sig- nificant and accurate, and lucid and logically expressed, but also fu
10、lly documented to ensure that the material presented will be accessible and coherent for anyone wishing to continue the research. Thus, many papers intended primarily to stimulate discus- sion and debate at meetings, or which are purely state-of-the-art reports, or single case studies would not qual
11、ify, even though they may be appropriate for publications in other journals, such as ASHRAE Journal and ASHRAE Transactions. For an idea about topics of interest, I refer you to the editorial in the first issue which noted a number of “hot topics.” In terms of subject matter, certainly ALL research
12、topics of interest to ASHRAE are appropriate. For additional information about the various ASHRAE publications, I refer you to the one page flier “Publications Overview for Authors Submitting Manuscripts,” which is available from ASHRAE headquarters. It states that the monthly ASHRAE Journal is inte
13、nded for a broad cross-section of readers. In general, therefore, research papers intended for those researchers working in great depth on a specific topic would not qualify for the ASHRAE Journal. Of course, The interruionul Journal of HVAC Ayub, Bush, and Haller 1992; Tojo et al. 1992). refrigeran
14、t liquid is even deliberately injected into the gas pockets to decrease the discharge gas tempera- ture under high compression ratios and thereby to extend the operating range of such compressors. Liquid refrigerant entering compressor cylinders has been a concern for a long time. However, not enoug
15、h has been achieved from the viewpoint of mathematical modeling. An extensive literature review was done in this area, but no satisfactory thermodynamic model for simulating compressor cylinder processes was found. Simpson and Lis (1988) measured cylinder pressures during compressor liquid ingestion
16、 in reciprocating com- pressors and presented an empirical approach for predicting the maximum cylinder pressure in terms of compressor power, bore size, and stroke length. Singh. Nieter, and Prater (1986) applied the polytropic equation to simulate compressor cylinder pressure, but the polytropic i
17、ndex in their model is considered as a variable and can be chosen arbitrarily. It is necessary to develop a mathematical model to predict the instanta- neous values of the refrigerant when liquid enters the cylinder. The effects on compres- sor performance made by liquid entering the cylinder can be
18、 studied with this model by simulating the thermodynamic process of compressing the liquid and vapor mixture. Slugging problems are detrimental to compressors. Under extreme conditions, the compressor cylinder pressures could be as high as ten times the peak pressures under normal operations (Singh,
19、 Nieter. and Prater 1986). due to the incompressibility of Zheji Liu is a senior analysis engineer in the Department of Test and Analysis for Waker Manufacturing, Grass Lake, Michigan. Werner Soedel is a professor of mechanical engineering at Purdue University, West Lafayette, Indiana. 99 HVAC ) 1 d
20、V V( hg-hidmc +- -p-+- m,da mC p-ug-uI - da The derivative of the total internal energy with respect to rotation angle is 113) 1151 dT Substituting Equation 116) into the energy balance Equation (11, and solving for gives: where and V c x=- - u -u1 u -u1 9 9 (191 Equations (2) and (17) can be integr
21、ated to determine the temperature and the mass of the refrigerant inside the cylinder control volume. All the other terms in those two equations either are known or can be determined whenever Tand m, are obtained. Computer Simulation As mentioned earlier, ail the variables with subscripts g or 1 are
22、 only functions of temperature. In order to numerically solve the highly nonlinear differential Equations ASHRAE TITLE*IJHVAC 1-2 95 W 0757650 05L5Yb3 T20 VOLUME 1, NUMBER 2. APRIL 1995 103 (221 and (171, we need to know those functions. Different refrigerants have different func- tions of temperatu
23、re for the variables. The following relations are obtained from a numerical interpolation (sometimes called curve-fitting) of the published data from ASHRAE (1985) for saturated liquid and saturated vapor of R-22, which are valid for refrigerant R-22 in the state of saturation from T = 45F to T = +
24、195F. where z stands for each of those mentioned variables. The results of the numerical inter- polations corresponding to each property are listed in the appendix. Similarly, the results are shown in the appendix for curve fitting processing of the data obtained from numerical differentiation with
25、respect to Tof the same data published in ASHRAE (1985) for the properties. This produced the results dp/dT, dug /dT, dvl /dT. d% /dT, and dh, /dT These results for the derivatives can also be obtained in the form of Equation (20). After the two-phase refrigerant mixture enters the compressor cyiind
26、er, there are two possibilities: the refrigerant changes from the two-phase state into superheated gas during the compression process, or it is compressed into subcooled liquid. When the first case occurs, the thermodynamic model derived in this paper should be imple- mented together with the thermo
27、dynamic model for superheated vapor, as described in Liu 1993, to simulate the properties of the refrigerant. However, when the refrigerant becomes pure liquid, there is no thermodynamic model to predict the refrigerant prop- erties. Under liquid compression, the cylinder pressure can reach a very h
28、igh value. This condition is called slugging. NUMERiCAL RESULTS AND DISCUSSIONS The thermodynamic model was applied to simulate a variable speed rolling piston rotary compressor, as depicted in Figure 1, with a speed range from 1 100 rpm to 9000 rpm. The effect of increasing the liquid portion in th
29、e refrigerant mixture can be seen from the plots of Figures 2 through 5, which were computed under the same conditions of nominal suction pressure ps = 444.7 kPa (64.5 psia), nominal discharge pressure pd = 1670.6 kPa (242.3 psia), and compressor speed Q = 7200 rpm. These plots show the instantaneou
30、s values of four variables (cylinder pressure, temperature, refrigerant discharge processes. As the quality of the refrigerant entering the cylinder decreased but did not cause the refrigerant to change to pure liquid during the cycle, the cylinder pressure did not change dramatically except during
31、the discharge process due to the fact that the denser refrigerant needs more pressure to be pushed out of the cylinder. Reducing the quality in the suction side increased the capacity of the compressor since more refrigerant was drawn into the cylinder, as shown in Figure 4, and decreased the cylind
32、er gas temperature, as shown in Figure 3. The initial quality of the refrigerant entering the cylinder also affects the state of the refrigerant during the compression process. When the suction refrigerant quality is large, the cylinder refrigerant mixture becomes superheated vapor sometime during t
33、he process, as represented by the dashed line in Figure 5. However, the opposite can also happen when a larger portion of saturated liquid flows into the compressor, as shown in Figure 6. It is important to prevent the latter case, since the catastrophic cylinder pressure buildup incurred is destruc
34、tive to the compressor. The fact that the two-phase, saturated liquid and vapor mixture can be compressed to pure liquid seems to be contrary to expectations. However, this can be proven from the pressure-enthalpy diagram, as shown in Figure 7. As the two-phase liquid and vapor mixture is compressed
35、 along the isentropic line AB of Figure 7. its value of quality becomes smaller and smaller, and finally the two-phase mixture changes to pure liquid. I mass, and quality), as a function of crank angle during the suction, compression, and i ASHRAE TITLESIJHVAC 1-2 i5 = 0757650 0515464 7b7 104 Suctio
36、n Compression chamber Figure 1. Rolling piston rotary compressor mechanism 90, 1-1 80 c -, 1 I .I_ (1 I I Q- - -YI- fa- =a- 2 m- g 10 - 0- Olmm4m5mm7mm Crank Angle (degrec) Figure 3. Refrigerant temperature during cylinder processes ips = 444.7 Wa (64.5 psla). pd = 1670.6 kPa (242.3 psia), and W = 7
37、200 rpml I 1 HVAC and thus the correct heat-transfer coefficients will be larger than those reported at the heat fluxes cited. For example, at a typical heat flux of 10 kW/m2 (3170 Btu/h-ft2.0F), for a plain copper tube the reported local heat-transfer Coefficient would be in the range of 2500 W/(m2
38、. (440 Btu/h.ft.“F), such that the wall superheat (Tu- Gat) is 4.0 K (7.2“F). If (Tbub- Td) is equal to 0.2 K (0.36“F), then the correct heat-transfer coefficient will actually be 5.3% higher, i.e. 10000/4.0 - 0.2) = 2632 W/(m2.K) (463.5 Btu/h.ft2.“F). For the widely-used microfin type of copper tub
39、es, the reported flow boiling heat-transfer coefficient will be approximately twice that of the plain tube, or 5000 W/(m2.K) (880.6 Btu/h.ft2.“F). giving an apparent wall superheat of 2.0 K (3.6F) and a correct value of 1.8 K (3.24“F). The cor- rect heat-transfer coefficient will then be 5556 W/m2.K
40、) (978.4 Btu/h.ft2a0F) or 11.1% larger. With large local oil concentrations, which occur at high local vapor qualities in these tubes, the effect is much more dramatic. For instance, evaluating the effect of using (T, - Tsat) equal to 0.5 K (0.9“F) in the above example, the correct heat-transfer coe
41、ffi- cients for the plain and microfmed tubes will be higher by 14.3% and 33.3%, respec- tively. Hence, applying the thermodynamic approach in future tests (or reevaluating prior data) provides a substantial improvement in the accuracy and the integrity of experimen- tal test data. Another important
42、 aspect to consider is the manner in which published refrigerant- oil flow boiling data have been presented to date. All local coefficients are cited for the nominal inlet oil concentration winkt, say 1, 2 or 5% y mass) oil. Using Equation (5), it can be seen that the local oil concentration whd inc
43、reases as a function of the inlet oil concentration and the local vapor quality. Thus, comparing local heat-transfer coeffi- cients at a local vapor quality of 0.50 really means a comparison of data at local oil con- centrations at 2, 4 and 10% by mass) rather for w,kt values of 1, 2 and 5% (by mass
44、); similarly at a vapor quality of 0.80 the comparison is for whal of 5, 10 and 25% (by mass) oil in the liquid phase, respectively. As a consequence, in the future it would be helpful if refrigerant-oil flow boiling data were presented showing the local oil con- centration in addition to the nomina
45、l value at the tube inlet. ASHRAE TITLErIJHVAC 1-2 95 W 0759650 0515480 OOT W 120 HVAC it is normally determined at both the inlet and outlet vapor qualities of the zone, O. 15 and 0.23 for the present example, and then the average coef- ficient is determined from these two values. Now (U, 1 for the
46、 zone can be calculated. The calculation iterates with assumption of a new value of A,I1 until the heat duty pre- dicted by the right-hand side of Equation (10) matches the heat duty of the zone 81. The thermal design process then passes onto the next zone j = 21 and the procedure is repeated. The h
47、eat absorbed in the second zone is (31.28 - 15.64) = 15.64 kJ/kg, the inlet and outlet vapor qualities are 0.23 and 0.31, and the inlet and outlet bubble point temperatures are 4.5 16C and 4.525“C, respectively. After analyzing all the individual zones, the size of the evaporator is known from summa
48、tion of the tube lengths for each zone. ASHRAE TITLE*IJHVAC 1-2 95 0759650 05L54L T4b VOLUME 1. NUMBER 2. APRIL 1995 121 Physical Properties Besides the heat release curve, the thermodynamic and physical properties of a refrigerant-oil mixture are required for thermal design. These properties are de
49、termined at the vapor quality values corresponding to the heat release curve. The liquid phase properties must be estimated for each local vapor quality using the local oil concentra- tion and the local bubble point temperature. The liquid phase properties required are density, viscosity, specific heat, thermal conductivity and surface tension. The vapor phase properties are for the pure refrigerant since no oil is present in the vapor. The vapor physical properties usually required are density, viscosity, specific heat, thermal conductivity and latent
