Influence of Nozzle Divergent Part Length and Throat Diameter on Vortex Control of Initially Subcooled Flashing Flow.
Ejectors which can recover the kinetic energy released during the expansion process are known to be beneficial to vapor compression cycle performance (Elbel and Hrnjak ; Lawrence and Elbel ). Figure 1 shows the layout and pressure-specific enthalpy diagram of a two-phase ejector cooling cycle first proposed by Gay . In this cycle, high pressure motive flow leaving the condenser enters the ejector through the motive inlet. The motive flow is expanded in the motive nozzle and creates a low pressure zone at the nozzle outlet, which entrains the suction flow from the evaporator. The two streams are mixed in the mixing chamber and kinetic energy is transferred from the motive flow to the suction flow. The mixed fluids leave the ejector through the diffuser. The fluid velocity is reduced in the diffuser which results in recompression of the mixed fluids by converting velocity energy back into pressure energy. Therefore, the ejector diffuser outlet pressure is higher than the suction flow pressure (that is, the evaporator pressure). The two-phase flow then gets separated in the separator. Saturated vapor enters the compressor while saturated liquid gets throttled and is fed into the evaporator via a metering valve. That way, kinetic energy released during expansion is utilized to compress the fluid from the evaporator. As a result, some compressor work is saved while the cooling capacity is increased if the heat rejection capacity remains constant.
Transcritical R744 cycles usually have larger expansion losses caused by throttling process than subcritical cycles under common working conditions. It is very beneficial to apply an ejector to transcritical R744 cycles due to the large recovery potential. COP improvements of 20% over baseline cycle using a conventional expansion valve has been achieved by using an ejector (Ozaki et al. ).
Ejector cooling cycles using low-pressure refrigerants, such as R134a or R1234yf, can also have noticeable performance improvements. Lawrence and Elbel  experimentally investigated the performance of an alternate two-phase ejector cycle in which the pressure lift provided by the ejector was utilized in order to provide multiple evaporation temperatures. Low-pressure fluids R134a and R1234yf were used. The ejector cycle showed maximum COP improvements of 12% with R1234yf and 8% with R134a when compared to a two evaporation temperature expansion valve cycle. When compared to a single evaporation temperature expansion valve cycle, the ejector cycle showed maximum COP improvements of 6% with R1234yf and 5% with R134a.
However, ejector cycle performance is usually sensitive to working condition changes which are common in many applications, including automotive AC systems. Different working conditions require different ejector geometries to achieve maximum performance. Slightly different geometries may result in substantially different COPs under the same conditions. Therefore, it is desirable to introduce an adjustable feature to the ejector so that ejector cycle performance can be optimized under different working conditions, which could make ejector technology more suitable for real world applications (Sumeru et al. ).
The ejector motive nozzle throat diameter (restrictiveness) is one of the key factors that affect ejector cycle COP. It has a direct impact on motive mass flow rate. One way to adjust the motive nozzle throat diameter in order to optimize ejector cycle performance according to the working conditions is by using a needle which moves in the axial direction so that the nozzle throat diameter can be varied, as illustrated in Figure 2.
Elbel and Hrnjak  were the first researchers to publish experimental results of introducing a variable two-phase ejector to a transcritical R744 system by installing a needle in the motive nozzle to control the motive nozzle throat diameter. The needle mechanism allowed control of gas cooler high-side pressure, which is an important task for a transcritical cycle to get optimum performance. However, nozzle and ejector efficiencies were impaired because of the additional frictional losses introduced by the needle. It was found that the benefits of high-side pressure control offset the losses in nozzle and ejector efficiencies.
A variable geometry ejector with adjustable needle in the motive nozzle can optimize ejector cycle performance under different conditions, but this design is complicated and costly, and more frictional losses are incurred because of the additional surface area introduced which results in lower nozzle and ejector efficiencies. This provides motivation to develop a new technology to control the motive nozzle restrictiveness.
Zhu and Elbel  were the first to introduce vortex control to ejector for the control of ejector cooling cycles. A vortex ejector which employs the vortex control to adjust motive nozzle restrictiveness differs from a conventional ejector in that an adjustable vortex is generated at the ejector motive inlet, as shown in Figure 3. The motive inlet vortex can be created by injecting part of the motive flow tangentially. After injection the tangential flow is mixed with the axial motive flow. The total mass flow rate passing through the vortex nozzle is equal to the sum of mass flow rates entering through the nozzle's axial and tangential flow inlets. The ejector cooling cycle using a vortex ejector, as shown in Figure 4, is almost the same as the conventional ejector cooling cycle. The only difference is that the flow at the condenser outlet of the vortex ejector cooling cycle is separated into two streams. One stream enters the vortex ejector through the motive flow tangential inlet and another enters through the motive flow axial inlet. In such a way, a vortex is created at the ejector motive inlet. The ratio of mass flow rates through the two inlets can be adjusted by a valve installed at the motive flow tangential inlet, thereby changing the vortex strength. The pressure drop across the control valve is usually small. It can be assumed that the thermodynamic state at the motive nozzle inlet after the vortex is introduced (downstream of the tangential inlet valve) is the same as the refrigerant state at the condenser outlet.
Zhu and Elbel's experiments  on vortex nozzle with initially subcooled R134a show that the strength of the nozzle inlet vortex can change the restrictiveness of the two-phase nozzle without the need of changing the nozzle geometry. The nozzle becomes more restrictive as the strength of the vortex increases. The mass flow rate can be reduced by 36% with vortex control under the same inlet and outlet conditions. The control range of inlet pressures and mass flow rates that can be achieved by vortex control appears to be large enough to be applicable for real world applications.
This paper is a continuation of the previous experimental investigation of vortex control by Zhu and Elbel . In the following sections, the influence of nozzle divergent part length and nozzle throat diameter on the vortex control of initially subcooled flashing flow will be presented. Visualization of initially subcooled flashing flow in the nozzle will also be provided.
EXPERIMENTAL FACILITY AND METHODS
Several transparent nozzles with controllable vortex at the nozzle inlet have been designed and manufactured for experiments, as shown in Figure 5. Important dimensions of the tested vortex nozzles have been summarized in Table 1 and these dimensions are shown with corresponding letters in Figure 5. Nozzles 1 and 2 are different in throat diameter, while nozzles 2, 3, and 4 have different divergent part lengths but the same convergent part, throat diameter and divergent angle (1 degree full divergent angle). Nozzle 4 is a nozzle that has almost no divergent part. It should be noted that the nozzle is only part of a vortex ejector. Experimental investigation of vortex ejectors in ejector cooling systems will be conducted in the future. The vortex nozzle is composed of three components: a tee-shaped part made of brass, a sleeve and a nozzle, as shown in Figure 6, both made of an optically clear resin called Waterclear Ultra 10122 from SOMOS and manufactured with a Stereo Lithography Apparatus (SLA) from 3D SYSTEMS.
The layout of the experimental facility for the vortex nozzle tests is shown in Figure 7. A pumped-refrigerant-loop was used for adjustment of test conditions to investigate the influence of nozzle divergent part length and nozzle throat diameter on the vortex control. The working fluid was R134a. A visualization chamber was built from clear PVC pipe. The temperature readings were all obtained from ungrounded Type-T immersion thermocouples. The measured temperatures are regarded as total temperatures. Absolute pressures were read by piezo-electric pressure transducers. Pressures and temperatures at the axial and tangential inlets of the nozzle were measured. The differences between the vortex nozzle axial inlet pressures and tangential inlet pressures were within 10 kPa when the tangential inlet mass flow rate is less than 13.0 g/s. The pressure difference increases to 28 kPa when the tangential inlet mass flow rate is 20.5 g/s. This pressure difference is caused by the relatively small inner diameter of the tangential inlet. The axial inlet pressure is assumed to be the nozzle inlet pressure [P.sub.in]. The pressure at the nozzle outlet [P.sub.out], was measured as well. The total mass flow rate [m.sub.total] and the nozzle axial inlet mass flow rate [m.sub.axial], were measured by Coriolis-type mass flow meters. The nozzle's tangential inlet mass flow rate [m.sub.tangential] can be calculated by subtracting the nozzle axial inlet mass flow rate from the total mass flow rate. The ratio of the nozzle tangential inlet mass flow rate to the total mass flow rate was adjusted by two valves. The larger the ratio is, the large the vortex strength is for the same total mass flow rate. In this paper, the vortex strength is defined as the ratio of the nozzle tangential inlet mass flow rate to the total mass flow rate, which can be expressed as shown in Equation 1:
It should be noted that in order to use the full range of vortex control from zero vortex to maximum vortex, two valves were installed in the test rig at both the nozzle's axial and tangential inlets. However, in actual applications, one valve should be sufficient to achieve nozzle restrictiveness control over a suitable range.
Different nozzle inlet pressures were achieved by adjusting the heating water temperature and pump speed which determine the saturation pressure of the refrigerant in the heater. The nozzle outlet pressure can be adjusted by a valve installed downstream of the nozzle. For all experimental results shown in this paper, the liquid flow at the nozzle inlet was subcooled by approximately 0.3 [degrees]C. Admittedly, the calculated subcooling is close to the uncertainty of thermocouple reading and it is imprudent to claim the inlet is subcooled solely based on the thermocouple readings. The sight glass installed at the nozzle inlet and the transparent body of the nozzle allows for visual confirmation that no bubbles are present at the nozzle inlet, which provides ultimate confirmation for inlet subcooling. Different nozzle inlet states with different levels of subcooling or vapor quality will be the subject of future work.
Videos and pictures of the two-phase flow in the nozzle are captured using a high-speed camera PHANTOM v4.2 from Vision Research.
VORTEX NOZZLE EXPERIMENTAL AND VISUALIZATION RESULTS
Figure 8 shows a comparison of the influence of nozzle inlet vortex strength on the total choked mass flow rate through nozzles 1 and 2 with throat diameters of 1.00 mm and 1.03 mm at different constant inlet conditions. The mass flow rate going through nozzle 1 is smaller than that of nozzle 2 under the same inlet and outlet conditions due to the smaller throat diameter and, therefore, larger nozzle restrictiveness. The control range of vortex control is also reduced as the nozzle throat diameter is reduced. The variation of total mass flow rate through nozzle 2(1.03 mm throat diameter) from zero to maximum inlet vortex strength is 6.0 g/s (32% of maximum total mass flow rate at zero vortex strength) while that of nozzle 1 (1.00 mm throat diameter) is only 3.3 g/s (21% of maximum total mass flow rate at zero vortex strength). At large inlet vortex strength, the restrictiveness of the two nozzles with different throat diameters becomes close.
Figure 9 shows a comparison of the influence of vortex nozzle outlet pressure on the total mass flow rate through nozzle 2 and 3 with different divergent part lengths. The nozzle inlet conditions are [P.sub.in] = 917 kPa, [T.sub.in] = 36.0 [degrees]C. The mass flow rates through each nozzle with zero or maximum vortex strength are shown at different nozzle outlet pressures. It can be observed that when the divergent part of the nozzle is shortened to 2.1 mm, the nozzle is not choked even when the nozzle outlet pressure is as low as 475 kPa. With long divergent part (40.0 mm), the nozzle is choked when the outlet pressure is approximately 600 kPa. Therefore, it can be concluded that the divergent part of a convergent-divergent nozzle contributes to the choking of initially subcooled flashing flow. At low nozzle outlet pressures, more mass flow rate can be driven through nozzle 3 (2.1 mm long divergent part) than nozzle 2 (40.0 mm long divergent part) under the same inlet and outlet conditions. When the nozzle outlet pressure is close to the inlet pressure, more mass flow rate can be driven through nozzle 2 with longer divergent part and larger outlet diameter than nozzle 3. The vortex control range of the nozzle 3 is smaller than that of nozzle 2. When the nozzle outlet pressure is approximately 480 kPa, the difference in mass flow rates through nozzle 3 with zero and maximum vortex strength is 4.1 g/s (19% of maximum total mass flow rate at zero vortex strength). The difference in choked mass flow rates through nozzle 2 with zero and maximum vortex strength is 6.0 g/s (32% of maximum total mass flow rate at zero vortex strength).
Visualization results of nozzle 3 (no inlet vortex) have been shown in Figure 10. From the visualization of initially subcooled flashing flow in the nozzle with short divergent part (2.1 mm long), it can be observed that when the nozzle outlet pressure is very close to the inlet pressure, as shown in Figure 10(a). the flow throughout the nozzle is transparent which means that the flow is single-phase. As the outlet pressure is further lowered, the flow starts to become bubbly in the downstream of the nozzle throat (Figure 10fb) and (c)). Eventually, when the outlet pressure is as low as 784 kPa or lower (Figure 10(d), (e), and (f)). the flow becomes bubbly immediately after the throat while remaining single-phase in the region upstream of the throat. These phenomena are very similar to what have been observed in Zhu and Elbel's  experiments with long divergent part nozzle.
Figure 11 displays a comparison of the influence of vortex nozzle outlet pressure on the total mass flow rate through nozzles 3 (2.1 mm long divergent part) and 4 (0.3 mm long divergent part). The nozzle inlet conditions are [P.sub.in] = 920 kPa and [T.sub.in] = 36.0 [degrees]C. The mass flow rates through each nozzle with zero or maximum vortex strength are shown at different nozzle outlet pressures. It can be observed that the vortex control effect disappears as the divergent part of the nozzle is almost fully removed. There is almost no difference in the mass flow rates through nozzle 4 with or without inlet vortex. Therefore, the divergent part of the nozzle is crucial in vortex control of initially subcooled flashing flow. Even with a divergent part as short as 2.1 mm (nozzle 3) can create a significant variation of mass flow rate with vortex control. The total mass flow rate curve (with/without inlet vortex) of nozzle 4 with 0.3 mm long divergent part almost overlaps with that of nozzle with 2.1 mm long divergent part with no vortex.
Figure 12 shows the influence of nozzle outlet pressure on the total mass flow rate through nozzle 4 with 0.3 mm long divergent part at different inlet conditions. For all the three inlet conditions, the inlet vortex does not create any significant variation of mass flow rate through the nozzle.
In order to see the influence of divergent part length on the vortex control more clearly, a comparison of nozzles 2, 3, and 4 with 40.0 mm, 2.1 mm, and 0.3 mm long divergent part, respectively, at different inlet vortex strengths is shown in Figure 13. It can be concluded that for the same convergent part, throat diameter and divergent angle, the longer the divergent part is, the larger the vortex control range is.
In this study, the influence of nozzle divergent part length and nozzle throat diameter on the vortex control of initially subcooled flashing flow has been experimentally investigated. Visualization results of initially subcooled flashing flow in the nozzle are also provided. According to the experimental results, it has been shown that the control range of vortex control is reduced as the nozzle throat diameter is reduced. At large inlet vortex strength, the restrictiveness of the two nozzles with different throat diameters becomes close. The divergent part of a convergent-divergent nozzle contributes to the choking of initially subcooled flashing flow. With almost zero divergent part length, the inlet vortex strength does not affect the total mass flow rate through the nozzle. For the same convergent part, throat diameter and divergent angle, the longer the divergent part of the nozzle is, the larger the vortex control range is. Therefore, the divergent part of a nozzle seems to be crucial in the choking and vortex control of initially subcooled flashing flow Visualization results show that when the nozzle outlet pressure is sufficiently low, the initially subcooled flow becomes bubbly immediately after the nozzle throat while remains single-phase in the upstream of the throat. For future work, more nozzle geometries will be tested to provide more insight into the vortex control mechanism. Experimental investigation of vortex ejector cooling cycle under different working conditions will also be conducted and the cycle performance will be compared with that of cycles using other control mechanisms or with no control.
[1.] Elbel, S. and Hrnjak, P., "Experimental validation of a prototype ejector designed to reduce throttling losses encountered in transcritical R744 system operation," International Journal of Refrigeration, 31(3): 411-422, 2008.
[2.] Lawrence, N. and Elbel S., "Experimental investigation of a two-phase ejector cycle suitable for use with low-pressure refrigerants R134a and R1234yf," International Journal of Refrigeration, 38: 310-322, 2014.
[3.] Gay, N. H., "Refrigerating System," U.S. Patent 1,836,318, 1931.
[4.] Ozaki, Y., Takeuchi, H., and Hirata, T., "Regeneration of expansion energy by ejector in C[O.sub.2] cycle," 6th IIR Gustav Lorentzen Conference on Natural Working Fluid, Glasgow, UK, 11-20, 2004.
[5.] Sumeru, K., Nasution, H., and Ani, F. N, "Areview on two-phase ejector as an expansion device in vapor compression refrigeration cycle," Renewable and Sustainable Energy Reviews, 16(7): 4927-4937, 2012.
[6.] Zhu, J. and Elbel, S., "A New Control Mechanism for Two-Phase Ejector in Vapor Compression Cycles for Automotive Applications Using Adjustable Motive Nozzle Inlet Swirl," SAE Int. J. Passeng. Cars - Mech Syst.9(1):44-51, 2016, doi:10.4271/2016-01-0243.
The authors would like to thank the member companies of the Air Conditioning and Refrigeration Center at the University of Illinois at Urbana-Champaign for their support.
COP - coefficient of performance [-]
h - specific enthalpy [kJ/kg]
m - mass flow rate [g/s]
P - pressure [kPa]
T - temperature [[degrees]C]
axial - axial inlet
in - inlet
out - outlet
tangential - tangential inlet
total - total mass flow rate through the nozzle
Jingwei Zhu and Stefan Elbel
University of Illinois
Table 1. Geometric parameters of tested vortex nozzles. Nozzle # 1 2 3 4 (a) Nozzle inlet diameter 15.0 15.0 15.0 15.0 (mm) (b) Nozzle throat 1.00 1.03 1.03 1.03 diameter (mm) (c) Nozzle outlet diameter 1.74 1.77 1.07 1.03 (mm) (d) Nozzle convergent 9.9 9.9 9.9 9.9 part length (mm) (e) Nozzle divergent part 40.0 40.0 2.1 0.3 length (mm) (f) Tangential inlet inner 2.0 2.0 2.0 2.0 diameter (mm) (g) Vortex decay distance 168.0 168.0 168.0 168.0 (mm)
|Printer friendly Cite/link Email Feedback|
|Author:||Zhu, Jingwei; Elbel, Stefan|
|Publication:||SAE International Journal of Passenger Cars - Mechanical Systems|
|Date:||Apr 1, 2017|
|Previous Article:||Condition Monitoring of Multi-Bolted Joints with Combination of Preload Measurement and Extended Evaluation from FEA.|
|Next Article:||Experimental and Numerical Investigation of Vehicle Drive and Thermal Soak Conditions in a Simplifed Engine Bay.|