ASHRAE OR-16-C041-2016 A Non-Dimensional Mapping of a Dual-Port Vapor Injected Scroll Compressor.pdf

《ASHRAE OR-16-C041-2016 A Non-Dimensional Mapping of a Dual-Port Vapor Injected Scroll Compressor.pdf》由会员分享，可在线阅读，更多相关《ASHRAE OR-16-C041-2016 A Non-Dimensional Mapping of a Dual-Port Vapor Injected Scroll Compressor.pdf（8页珍藏版）》请在麦多课文档分享上搜索。

1、 Christian K. Bach is an assistant professor, Mechanical and Aerospace Engineering, Oklahoma State University, Stillwater, OK. Eckhard A. Groll and James E. Braun are professors, and Travis W. Horton is an assistant professor at the School of Mechanical Engineering, Purdue University, West Lafayette

2、, IN. A Non-Dimensional Mapping of a Dual-Port Vapor Injected Scroll Compressor Christian K. Bach Eckhard A. Groll Associate Member Fellow ASHRAE James E. Braun Travis W. Horton Fellow ASHRAE Member ABSTRACT The application of vapor injection to the compression process leads to a decrease in dischar

3、ge temperature, extending the operating envelope to lower suction pressures. Additionally, it increases the coefficient of performance as well as the heating capacity under these conditions. Vapor injected compressors are therefore ideally suited for cold climate heat pump applications. This paper i

4、ntroduces a PI-type mapping of a dual port vapor injected compressors performance data trained with data from both, in-system testing as well as test-stand testing. The in-system testing was conducted in a prototype cold climate heat pump, where injection mass flowrates and suction superheat were a

5、result of the operating conditions. In contrast, suction superheat of the test-stand data was fixed while injection flowrates were dictated by the test plan. These differences result in limitations of the mappings if the model is trained with only one of these sets, as shown in the companion paper.

6、INTRODUCTION AND LITERATURE REVIEW Operation of compressors at high pressure ratios, such as in heat pump applications, leads to high discharge temperatures, limiting the operating envelope of the equipment. One method to overcome this issue is a staged compression process using two separate compres

7、sors, combined with an intercooler or economizer. However, this increases system complexity and cost compared to single compressor systems and can lead to compressor failure due to oil migration (Caskey et al., 2012) that can be overcome by adding additional controls (Hutzel and Groll, 2013). Vapor

8、injected compression does not require additional controls for the compressor oil level and only requires a single rather than multiple compressors. Vapor injection is achieved using dedicated injection ports within the compression path in the scroll compressor. Adding an additional injection pressur

9、e level is now reduced to adding two injection ports (one in each compression path) rather than having to add an additional compressor. Mathison et al. (2014) found in a theoretical study that the achievable incremental increase in COP decreased with each additional injection pressure level. In part

10、icular, they found that 2 injection pressure levels lead to 67% of the maximum COP improvement of 12% for an R410A cycle operating between 5C (41F) and 40C (140F) saturation temperature. Experimental research on dual pressure level vapor injection is limited. Most papers in open literature investiga

11、te single port vapor injection. Xu et al. (2011) conducted a detailed literature review on refrigerant injection for heat pump and AC systems and found that the internal heat exchanger and flash tank cycle are the most commonly investigated system configurations. Navarro et al. (2013) conducted a pa

12、rametric study using a single port vapor injected compressor. They developed a correlation that used two coefficients as well as the pressure ratio at the injection and the suction mass flowrate. Their correlation resulted in a correlation factor of greater than 0.99. Song et al. (2014) developed a

13、mapping for a dual port vapor injected compressor operated at a fixed condensing temperature of 43.3C (110F). They found that the COP improvement of a dual vapor injection heat pump system increased with increasing pressure ratio up to 19% at a pressure ratio of 8. Ramaraj (2012) investigated both o

14、il flooded and vapor injected compressors. She developed a system model based on inter/extrapolation from experimental data and found that oil flooded compression leads to a larger COP improvement than vapor injection while the model predicted that there is no increase in low ambient temperature cap

15、acity for either system. Bach et al. (2014) investigated the effects of dual port vapor injection and other parameters onto seasonal heating performance (HSPF) using a bin-type method based on experimental data of a 5-ton (17.6 kW) prototype heat pump. Low temperature cutout was found to reduce the

16、HSPF of the single stage baseline system by only 1% for a cold climate (Minneapolis, Minnesota). The main benefit of the vapor injected compression was an increase in low temperature capacity, leading to a reduction in auxiliary electric heat, equivalent to more than 6% of annual single stage power

17、consumption. Song (2013), Ramaraj (2012) and Bach (2014) used compressors that were manufactured identically. The difference between the available data in their work lies within the range of operating conditions. Most significantly, the injection pressures, injection mass flowrates as well as the ra

18、nge of discharge pressures are quite different. The motivation of this paper is to provide a compressor mapping for dual port vapor injected compressors that is more generally applicable than the ones provided by Song et al. (2014) and Ramaraj (2012) by considering all available data as training dat

19、a for the mapping. This mapping can then be used with greater confidence for future studies of the dual port vapor injected scroll compressor. DATA SOURCES Data used in this paper comes from three different test setups: a modified hot gas bypass test stand, a calorimeter, and a cold climate heat pum

20、p with vapor injection. The data for the different test setups is shown in Figures 1a and b. The hot gas bypass test stand and calorimeter used the nominal power supply frequency, while it was varied from 40 to 70 Hz for the heat pump (“In.-sys.” and additional “Map. Pts.”). Injection mass flowrates

21、 varied between the different setups. Hot Gas Bypass (HGB) Test Stand Data. The HGB test stand is shown in Figure 2. Vapor for injection was generated using two heat exchangers for cooling the discharge gas followed by expansion valves, allowing free adjustment of the injection pressures. Suction su

22、perheat and discharge discharge dew temperature had fixed target values, while the suction pressure was varied according to the test plan. More details on the hot gas bypass test stand can be found in Song (2013). Calorimeter (CM) Data. The modified calorimeter is shown in Figure 3. Vapor for inject

23、ion was generated using a single desuperheater with the cooling water flowrate adjusted to obtain a fixed superheat. The lower injection pressure and superheat was a result of a second expansion process from that condition. Discharge pressure had a fixed target value and suction superheat was the sa

24、me for all but 1 value. More details on the calorimeter testing can be found in Ramaraj (2012), p. 64 ff. a) b) Figure 1 (a) Discharge and suction pressure of different data sets and (b) high and low pressure (HP, LP) vapor injection (VI) mass flowrates and overall pressure ratio of different data s

25、ets. Data “in-sys.” and “Map. Pts” are both from tests conducted with the cold climate heat pump. Degree of superheat 5.5 to 19.2 K (10 to 35 F) Figure 2 Simplified schematic of hot gas bypass test stand used by Song (2013) Figure 3 Simplified schematic of modified calorimeter used to obtain the dat

26、a for Ramaraj (2012) Cold Climate Heat Pump (CCHP) Data. The heat pump, a schematic is shown in Figure 4, was tested according to a test plan that specified the outdoor and indoor conditions. Discharge and suction pressure were therefore a result of operating condition, compressor speed, and overall

27、 system behavior. Additional mapping points were added to the test plan to further extend the available data beyond what was available from the test stands, as shown in Figure 1a. Vapor for injection was generated by a three stage expansion process with vapor separators. This lead to a coupling betw

28、een expansion process and vapor intake of the compressor, i.e. the injection mass flowrates were a result of the operating condition. Injection mass flowrates for the in-system testing were generally in the lower range of the ones seen for the two test stands. In particular, the high pressure inject

29、ion mass flowrate extends to values below what is available from the test stands. More details on the CCHP can be found in Bach (2014). Figure 4 Simplified schematic of cold climate heat pump used by Bach (2014) MAPPING APPROACH For simplicity, a non-dimensional -group mapping approach was chosen. T

30、he basic principle behind this approach is to express a dimensionless output -group () by suitable input -groups, e.g. = 0 11 22 33 1111, (1) where 1 through 11 are the input groups. The coefficients c0 through c11 need to be determind for each output group using either a minimization procedure for

31、the residuals or ordinary least squares (as shown in more detail in the companion paper, Bach and Cheung, 2016). For this paper, the selection of suitable input groups resulted in = 0 ( )1 ( , )2 ( ,)3 (,)4 (,)5 (,)6(,)7 ()8 (,)9 ()10 ()11, (2) where the superheat enthalpy differences are calculated

32、 as = (,) (, = 0), (3) which in case of the chosen data represents the superheat but allows to update the mapping coefficients in case of two phase inlet conditions due to the continuity of the expression. The output group is either the (overall) isentropic efficiency (), the volumetric efficiency (

33、), the normalized injection mass flowrates or the normalized discharge temperature (,). Overall isentropic efficiency is defined as = (, )+ , (,(,) ,) + , (,(,) ,), , (4) where is the enthalpy at the compressor suction, , is the discharge enthalpy assuming isentropic compression, and is the suction

34、mass flowrate. Similarly, , is the enthalpy at the low pressure injection port, ,(,) is the corresponding discharge enthalpy assuming isentropic compression, and , is the low pressure injection mass flowrate. The index hp in the last part of the nominator denotes high pressure injection. , is the me

35、asured electric power consumption of the compressor. Overall volumetric efficiency is defined as , (5) where is the volumetric flowrate of the compressor, is the olumetric displacement of the compressor, and is the power supply frequency to the asynchroneous motor of the compressor. The compressor r

36、otor slip was found to be approximately 2% based on data available from the calorimeter dataset. The (normalized) injection mass flowrates are defined as , = ,/ and , = ,/ . (6) The normalized discharge temperature of the compressor is defined as , = , (7) where is the measured discharge temperature

37、 and is the critical temperature of the refrigerant. MAPPING RESULTS Figure 5 shows the mapping result for the isentropic efficiency. All but one value are predicted within 5% of the measured value. Most predictions appear randomly scattered around the parity line. However, if there appears to be a

38、trend towards over prediction of isentropic efficiency if injection mass flowrate is increased above 100% of the suction mass flowrate. Figure 6 shows the mapping results for the volumetric efficiency. All data points are predicted within 5% of the measured value. Figure 7 show the mapping results f

39、or the discharge temperature. 94 of the 98 data points (= 96%) are predicted within 5 K (9 F) of the measured value. Figure 8 shows the mapping results of the high pressure injection mass flowrates. The map appears to over-predict the injection mass flowrate at normalized flowrates below 15%. All po

40、ints above 10% normalized injection mass flowrate are predicted within 5% of the measured value. For the low pressure injection mass flowrates, all points above 15% normalized flowrate were predicted within 5% of the measured value, while data from the HGB testing exceeded these limits if the flowra

41、te was smaller than 15%. Figure 5 Mapping results of overall isentropic efficiency (A) Figure 6 Mapping results of volumetric efficiency Figure 7 Mapping results of normalized discharge temperature Figure 8 Mapping results of high pressure injection flowrate Table 1 shows the mapping coefficients fo

42、r isentropic efficiency, volumetric efficiency, discharge temperature, and injection flowrates. The coefficient of determination for each corresponding group for the different data sets is given. It was found that the injection mass flowrates are very well predicted if only the corresponding injecti

43、on pressures are considered, leading to an R2 of 94% for all data. Therefore the injection mass flowrates were not used for the other mappings, since they can be represented by the injection pressures. Isentropic efficiency was mapped with and without consideration of injection superheat, leading to

44、 an R2 of 84%. Consideration of injection superheat did not affect the R2 if all data was considered but slightly changed how well individual datasets were predicted. Differences between the two maps were not noticeable when the data was plotted. The prediction of the calorimeter data resulted in ne

45、gative R2, since the range of measured isentropic efficiency values was small. Low pressure injection superheat appears to be not very important (small absolute value of coefficient), while the high pressure vapor injection group has a larger value with positive sign, e.g. isentropic efficiency incr

46、eases with superheat. Potential reasons for this lie in internal heat transfer between port and injection point within the scroll as well as in the chosen test matrices. Volumetric efficiency was predicted with an R2 of 77% for all data. The smaller value of R2 for the volumetric efficiency with the

47、 same 5% deviation limits as for the isentropic efficiency is a result of a smaller spread of the majority of the volumetric efficiency data than for the isentropic efficiency data. Referring to Figures 5 and 6, it can be seen that most volumetric efficiency data is within a 10% range (e.g. 82 to 92%), while the isentropic efficiency data spans a wider ran