- Email: [email protected]
Formation, distribution, and movement of oil droplets in the compressor plenum
Accepted Manuscript
Formation, distribution, and movement of oil droplets in the compressor plenum Jiu Xu , Pega Hrnjak PII: DOI: Reference:
S0140-7007(18)30226-3 10.1016/j.ijrefrig.2018.06.020 JIJR 4029
To appear in:
International Journal of Refrigeration
Received date: Revised date: Accepted date:
13 March 2018 11 June 2018 25 June 2018
Please cite this article as: Jiu Xu , Pega Hrnjak , Formation, distribution, and movement of oil droplets in the compressor plenum, International Journal of Refrigeration (2018), doi: 10.1016/j.ijrefrig.2018.06.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
CR IP T AN US M ED PT CE AC
Highlights: Demonstrate the formation oil droplets near a compressor discharge valve. Quantify and explain the oil mist distribution in a scroll compressor plenum. Simulate the movement of oil droplets by CFD and verified by visualization results. Provide upstream information for the design of oil separation structures.
1
ACCEPTED MANUSCRIPT
Formation, distribution, and movement of oil droplets in the compressor plenum Jiu Xu a, Pega Hrnjak a,b,* an Air Conditioning and Refrigeration Center, University of Illinois at Urbana-Champaign 1206 West Green Street, Urbana, IL 61801, USA b Creative Thermal Solutions, 2209 Willow Rd., Urbana, IL, USA * Corresponding author: [email protected] +1-217-390-5278 ABSTRACT
1.
CR IP T
Oil is important to compressors, but it brings some side effects on the performance of refrigeration systems. Discharge valve or orifice in the compressor plenum is the gateway for the oil to leave the compressor and circulate in the system. The space of the compressor plenum has the potential to separate the oil mist from the refrigerant so that oil circulation ratio (OCR) is reduced. In this study, the formation, distribution, and movement of oil droplets in the plenum are investigated through visualization and CFD simulation. Videos of oil atomization are captured by the high-speed camera to describe how oil mist is formed. The distribution of droplet size and velocity is quantified by video processing techniques. Discrete phase model gives the numerical approach to study the oil droplet movement inside compressor plenum. The numerical simulation is verified by comparing visualization results with simulation results. The oil flow phenomenon and its quantification provide important information for the design of separation structure incorporated in the compressor. Also, a set of tools are developed and verified to study the oil droplets behavior in different types of compressor plenums.
Introduction
AN US
1.1. Background
Zimmermann and Hrnjak (2014, 2015) showed that oil droplets are formed by the break-up of oil film between the valve and the valve seat during the opening and closing in a scroll compressor. The visualization of the reed valve in the scroll compressor shows the valve periodical movement together with the oil atomization. Xu and Hrnjak (2016, 2017) measured the OCR with a non-invasive way based on the oil annular mist flow visualization in the compressor discharge pipe. Effect of focal depth is considered during the video processing to provide a good estimation of the number of oil droplets to provide the amount of oil in that form in a specific control volume. Although flow visualization provides the image of the oil mist when part of the compressor is made transparent, this approach is restricted by narrow view angle and limited light exposure. Videos can provide the droplet flow velocity in the two-dimensional case, but it is much more difficult to estimate the velocity in the third dimension. Also, the velocity field of the vapor phase and slippery velocity between the droplet and vapor phase are hard to obtain through videos. Therefore, computational fluid dynamics is used as an additional numerical approach to simulate the complicated two-phase flow. Bhatia et al. (2002) presented a CFD case that analyzed the oil flow in the centrifugal type oil separator. Design suggestions were given to achieve minimum pressure drop and maximum separation efficiency for an oil separator integrated into a swashplate reciprocating compressor. Yokoyama et al. (2012) analyzed the oil flow in a CO2 rotary compressor shell by CFD. Numerical simulations showed that flattening the lower balance weight or installing a rotating disk over upper balance weight can reduce the OCR and the results are verified by system experimental results. Xu and Hrnjak (2016b, 2017b) validated the CFD simulation model using visualization results for oil droplet flow inside the scroll compressor plenum. CFD is a favorable approach when flow details are needed under the realistic geometries and working conditions are. The verification and validation of the CFD simulation is the key to ensure the reliability of calculation results. Overall, previous researchers have developed some strong tools to study the in-compressor oil flow. Oil mist has been observed in the compressor plenum under different working conditions of different types of compressors. However, a combination of visualization and CFD is still uncommon. This justifies further study of oil mist flow in the compressor using both approaches.
CE
PT
ED
M
Oil is necessary for cooling, lubrication, and sealing of conventional compressors but oil circulating in the refrigeration cycle typically has adverse effects on heat transfer and pressure drop (DeAngelis and Hrnjak, 2005; Li and Hrnjak, 2014). Besides the traditional vapor compression cycle, oil circulating in the ejector cycle will also bring some side effects (Zhu et al., 2018). Even application of magnetic bearings and new material makes oil-free compressors possible, their number is still very small. Therefore, engineering attempts in the design of compressors are made to keep the oil in the compressor and reduce the oil circulation ratio (OCR) in the system. Separating the oil mist from the high-pressure refrigerant vapor and draining it back to the compressor is a feasible way to achieve this goal. How oil droplets and film flow with the refrigerant vapor is the key information for oil separator design. Important characteristics of oil flow include oil droplet size distribution, oil droplet velocity distribution, oil droplet trajectories and refrigerant vapor velocity field. Due to desired compactness, it is preferred to integrate the oil separation structure into the compressor rather than having an external separator between the compressor and the condenser. These factors justify a closer look at the oil flow inside the compressor plenum, especially the formation, distribution and movement of oil droplets. 1.2. Literature review
AC
Oil flow inside the compressor plenum and discharge tube has been studied mainly by two approaches: flow visualization and computational fluid dynamics (CFD) simulation. As one of strong tool to study misty flow, flow visualization is the most intuitive approach to understand how oil droplets flow inside the compressor shell. Toyama et al. (2006) described methods to quantify oil droplet behavior using high-speed photography in the scroll compressor shell via sight glasses. The size distribution of oil droplets is described by the video. It was concluded that the mean diameter of oil droplets decreases as the flow speed of the refrigerant vapor increases. Wujek and Hrnjak (2011) developed visualization techniques to quantify the annular mist flow in the compressor discharge pipe. The oil film thickness is measured by the laser reflection method. The size and velocity of oil droplets captured by the camera were determined by video processing program. Using a swash-plate compressor with R134a and PAG 46 oil, it showed the developing annular-mist flow along the discharge pipe.
1.3. The focus of this study
2
ACCEPTED MANUSCRIPT
2.
CR IP T
The objective of this paper is to analyze the oil droplet flow in the compressor plenum by both visualization and CFD simulation. By modifying a scroll compressor plenum, a transparent prototype is made for visualization. A high-speed camera captures the flow videos at different locations in the compressor plenum. The videos are then processed to provide flow details of the oil mist. Discrete phase model is used to simulate the oil droplet trajectory based on the vapor flow field and droplet size distribution near the reed valve. The simulation results are then verified with the droplet size distribution given by video processing. The paper aims to provide a comprehensive and detailed understanding of the formation, distribution, and movement of oil droplets in the compressor plenum. It provides the design guidelines for oil separation structures integrated into the compressor plenum. The verified method for studying oil mist can also be applied in further research.
Experimental Methods
2.1. System facilities and test conditions
AC
CE
PT
ED
M
AN US
A full vapor compression cycle is built to test and visualize the compressor under different working conditions. The diagram of experimental facilities is shown in Fig. 1. The working condition is controlled by adjusting the capacity of condenser or evaporator, and by adjusting the opening of the expansion valve. Compressor suction pressure and temperature, compressor discharge pressure and temperature are measured by the pressure transducers and thermocouples. OCR sampling and mass flow rate measurement takes place in the liquid line between the condenser and the expansion valve. OCR is determined by sampling method according to ASHRAE standard 41.4. A Coriolis type mass flow meter measures the mass flow rate of the system. The compressor used in this study is a 3-ton scroll compressor with a low-pressure shell arrangement. It was designed for operation at 50 Hz, however, in this work, a variable frequency drive is used to set the compressor operating frequency to the desired level. The compressor was originally designed to use R-410A as the refrigerant for residential air conditioning application. To ensure the reliability of transparent material for visualization, R-134a is used as the refrigerant to provide relatively lower compressor discharge pressure. Most of the experiments are carried out with PVE 32 oil but POE 120 was also used to study the effect of oil properties.
Fig. 1 – System facilities for experiments The testing conditions were selected to provide the most realistic operation under air conditioning conditions while assuring the safety of the experiment and operators by keeping the discharge pressure and temperature levels low enough to avoid failure of the transparent part. The compressor operating frequency was varied from 30Hz to 60 Hz while the pressure ratio and pressure levels were kept relatively close, so the only change should be the refrigerant mass flow rate and the OCR as a result of more or less oil pumping in the crankcase. The detailed test conditions are listed in Table 1. OCR determined by sampling methods has the maximum absolute uncertainty of ±0.2%. Temperature measurements carry an uncertainty of ±0.2°C, and pressure measurements are accurate to ±0.25%. Mass flow rate carries uncertainty of ±0.5%.
Table 1 – Detailed test conditions Test case No. Compressor frequency [Hz] System mass flow rate [g·s-1] Suction pressure [kPa] Discharge pressure [kPa] Superheat [K] Discharge temperature [°C] Oil circulation ratio [%] Oil mass flow rate [g·s-1]
1 30 16.7 337.9 906.8 21.3 78.0 0.15 0.025
2 40 20.3 309.7 954.4 23.0 77.1 0.6 0.124
3 50 21.9 312.7 961.0 24.3 81.1 0.98 0.214
4 60 23.5 315.3 987.2 24.6 88.7 1.17 0.275
2.2. Modification of the compressor The oil flow inside the compressor plenum was visualized by highspeed CCD camera. Some modifications were made to the upper part of the scroll compressor plenum to improve imaging of the oil droplets inside the plenum. Firstly, two flanges were added between the top cover and the compressor shell to provide visibility to the discharge valves. Initially mounted four sight glasses were not sufficient to bring necessary light for the camera. For that reason and better visual access, the transparent cylindrical ring was added between the flanges (Fig. 2).
3
ACCEPTED MANUSCRIPT clear image of droplets, but the additional light had to be brought from the front to capture the movement of the reed valve. Videos of the oil flow were taken at three locations: center reed valve, valve front (the vicinity area in front of the valve) and discharge pipe. The locations of video frames are marked in red and the corresponding camera view angles are marked in green in Fig. 4. Videos of the center reed valve and videos of the valve front used the same camera position at a different focal distance. By varying the lens, the frame was focused on either the valve or the injected droplets.
AN US
CR IP T
Fig. 2 – Outer shape of the modified plenum of a scroll compressor
Fig. 4 – Locations of video taken in the modified plenum
Fig. 3 – Opened plenum of a scroll compressor
AC
CE
PT
ED
M
As shown in Fig. 3, a transparent screen was placed between the transparent flange wall and the valve to remove the interference of fine oil mist. In this way, a clear vision can be achieved at a plane about 5 mm from the tip of the valve. The compressor used in this project had three discharge valves, two of which were intentionally forced to be closed to reduce the complexity of oil flow. Modification mentioned above added around 30% of total volume of the original geometry to the original geometry of the scroll compressor plenum and made the relative position of discharge pipe higher than that in realistic design. It must be mentioned that such modification provided the access to visualization at the cost of some changes to the original vapor and oil droplets flow. The flow details obtained in the modified compressor differs from that in the original compressor, but authors believe it is acceptable as validation of the CFD simulation is done for the same geometry after modification. Furthermore, the geometry of plenum varies with different types of compressor. The prototype made in this study is just one example how the visualization is realized. Though the difference of the plenum shape can be significant, the same method can be applied to investigate of oil flow in the compressor plenum. As an approach to studying the oil droplets inside the compressor, visualization also has its limitation. A part of compressor shell must be transparent before the internal flow can be seen, so some modification of the original geometry is inevitable. Also, transparent material like plastic or glass is typically much more fragile than metal so that the compressor discharge pressure and temperature must be maintained at a relatively low level to prevent damage. 2.3. Visualization set-up High-speed camera Phantom 6.0 was used with different lenses. The frame rate was kept higher than 1000 fps to capture the fast movement of oil droplets. The light was provided from the back for a
When taking the videos of the oil droplets in the valve front and those in the discharge pipe, backlighting was used since it is the preferred mode for achieving good contrast and crisp images of the droplets. On the case of the center reed valve, backlighting did not help to make the image better, so front lighting was used. When using front lighting, instead of concentrating and absorbing the light, the droplets and edge of the reed valve appear as a reflection, which poses some challenge to the image processing procedure. Fig. 5 (a) shows a sample frame from a video. The focus is on the tip of the reed valve from about 45o. Zimmermann and Hrnjak (2014b) have shown that the break-up pattern of the oil film between valve and valve seat. A cloud of oil droplets injected by the high-pressure refrigerant vapor as the valve opens. Video processing method is used to quantify the initial velocity of oil droplets injected into the plenum during the oil film break-up. More details about the formation of oil droplets will be discussed in section 3. Fig. 5 (b) shows a sample frame from a video taken at the valve vicinity when focusing on the transparent screen in front of the valve. It is processed to estimate the size distribution of the oil droplets injected in the valve opening. In the plenum chamber, the oil droplets are moving in random directions at a relatively low velocity (0.5-2.0 m·s-1). The droplet size can be better estimated than the droplet velocity because of the difficulties to assess three-dimensional vector with one camera. Therefore, oil droplet size distribution is the key parameter that is used to compare visualization and simulation results. Fig. 5(c) shows the frame taken in transparent discharge pipe located around 20 centimeters from the compressor discharge port. The view in a raw form presents droplets (dark spots) behind the film with waves. During quantification and analysis, several filtering procedures were conducted to better capture and measure the size, number, and velocity of droplets. Since the flow is strongly one dimensional, droplet velocities determined are closer to the actual values. When estimating the total number of droplets in the tube authors have first determined the depth of view, then the number of droplets in the view volume and their diameter and its distribution and then the total mass in the droplet form. 4
ACCEPTED MANUSCRIPT Details of the method were presented in recently published papers (Xu and Hrnjak, 2017a).
3.
Formation of oil droplets
3.1. Oil atomization at the start-up
AN US
CR IP T
The gateway for the oil to leave the compression chamber is the valve opening gap or the discharge orifice. Therefore, in our prototype, the reed valve is the main research objective to study the formation of oil droplets in the compressor plenum. Start-up is usually the condition when the compressor experiences its highest mass flow rate (under equalized pressures) and consequently the highest lift on the valve because the pressure difference is the greatest over the first few operating cycles. To improve the video quality, the authors tried to capture the video of oil atomization at the start-up, avoid foggy foreground in steady-state operation. In the beginning, the oil mist inside the discharge plenum originated mostly from liquid film accumulated in between the reed valve and its seat. After a few opening cycles of the valve, the discharge plenum is inundated with liquid coming from the system to the compressor suction due to the sudden pressure change. Fig. 6 shows the starting process sequence of the compressor after a long off-period with R134a and PVE 32 oil. The video is taken by the camera with a 55 mm lens at the resolution of 256*256 and frame rate of 4600 fps. The pictures show the evolution in time from the moment the compressor is turned on until there is a first full opening of the valve. When the valve hits the stop ((f)), a large amount of liquid is pushed through the discharge orifice. On the intermediary steps, the interaction between the small amount of vapor coming out of the discharge port and any oil that was present in the plenum near the valve is seen, and large droplets (>1mm) are being generated.
Fig. 5 - Sample video frames taken in different locations inside the compressor: (a) center reed valve; (b) valve front; (c) discharge pipe 2.4. Video processing
AC
CE
PT
ED
M
For the videos that used backlighting, the image processing procedure used in this work is the same as that used by Wujek and Hrnjak (2011). The procedure consists of first eliminating any noise by averaging the pixel values at each location on the image over the entire period of the video and subtracting it from each frame. After the noise elimination, the processing code searches for peaks and valleys between frames and applies a cross-correlation to determine where the peaks and valleys coincide. This is used to remove the waviness of any liquid film that might be present on the walls. Once this clean image is present, a thresholding process is applied to the in which the same is transformed into a binary matrix with a cut-off level. With the help of built-in functions in Matlab, the code then applies a maximum derivative approach to identify the droplets in the raw image. The algorithm goes back to the original clean image and looks for the slope of the sharpest change in pixel values along the horizontal and vertical axis of the identified droplet. A minimum slope criterion is then applied to discard droplets that might not be in focus. Other criteria such as aspect ratio of the major and minor axis, the region being darker than the average surrounding pixels are also applied to discard possible false droplet or droplets which dimensions cannot be identified with the desired accuracy. The uncertainty associated with these methods can be found well documented in (Hay et al., 1998) and the limiting factor was cited to be the pixel width of the camera sensor for these types of flows. In their work Hay et al., 1998, estimated the resolution in their method to be the width of half a pixel. The pixel width is a function of the magnification and the size of the pixel array of the camera. In this work the sensor array consists of 512 x 512 pixels and the resulting pixel widths were 51.5µm for the valve videos and 21.47 µm for the plenum videos, resulting in ±25.75 µm and ±10.73 µm. Considering that the minimum drop size identifiable was set to two pixels in width we can determine that the maximum uncertainty for the histograms is ±25% for the smallest droplet.
Fig. 6 - Discharge reed valve opening at the compressor start-up A shorter and more zoomed view is taken to capture the details of oil film break-up process. The video frames shown in Fig. 7 were taken with the same lens as Fig. 6 and the resolution of 128*128 at the frame rate of 19,000 fps. Also, for better visualization result, the light source and direction is changed from front-light to top and side light through sight glasses. Fig. 7 (a) to Fig. 7(f) show the time evolution of the startup procedure. Fig. 7 (b) shows that on the left side close to the tip of the valve, there is film breakup present and it also indicates that the flow around the valve is not symmetrical. Later in Fig. 7(c) and Fig. 7(d), the unstable structure of oil film cause discontinuities in the film, making certain portions of the film to protrude outwards. The deformed oil film then ruptures into smaller ligaments and atomizes into droplets. Fig. 7(e) and Fig. 7(f) show the later stages the camera view becomes blurry when a large number of droplets are generated. It needs to be mentioned here the time span of videos shown in Fig. 7 is much shorter than that in 5
ACCEPTED MANUSCRIPT Fig. 6 because the oil film break-up details can only be observed at the beginning of the compressor operation cycle.
from underneath the valve when the gap between the reed valve and the valve base is large enough to break the oil film (Fig. 8(b) and Fig. 8(c)). The front of the oil droplet cloud propagates further and then invisible in a few microseconds. The amount of oil injected by the atomization during the steady-state operation seems less than that during the start-up. Another piece of information that can be extracted from the videos at steady state operation is the valve cycling times. By tracking the bright spot of on the edge of the reed valve, the cycling period of valve opening, and closing can be calculated. The overall cycle time showed very good agreement with the period calculated by the frequency set on the variable frequency drive. 3.3. Discussion
CR IP T
The observation of oil droplet near the valve at the start-up and steady-state helps to understand the mechanism of oil droplet formation. One hypothesis of the formation is that oil is either entrained in the vapor refrigerant or scraped from the clearances between moving parts in the compression chamber. The other hypothesis is that most of the oil is dispersed in the plenum comes from the atomization of oil film between valve and seat. Though it is difficult to find direct evidence from the videos, some estimation can still be made to verify the hypothesis mentioned above. The oil volume being discharged by the valve opening in each cycle can be calculated based on the compressor frequency, mass flow rate, and oil circulation ratio. The volume of the gap between the valve and seat is estimated to be 20 mm3 while the discharged oil volume for each cycle varies from 2 mm3 to 5 mm3 depending on the compressor frequency. This means the gap can hold an order of magnitude more oil than the amount of oil actually discharged. Therefore, for this particular compressor design, the second hypothesis is more credible. When the reed valve closes, the gap between valve and base gets replenished with some oil and holds it until the valve opens again. When the reed valve opens, the continuous oil film is firstly stretched is firstly stretched and then atomized to form the oil droplet cloud. The oil mist is then entrained by the fast-moving refrigerant vapor and occupies the plenum.
AN US
Fig. 7 – A zoom-in view of discharge reed valve opening at the compressor start-up 3.2. Oil atomization at steady state operation
AC
CE
PT
ED
M
Oil atomization is then visualized during the steady state operation of the compressor. This is even more important than that at the start-up because compressor operates at a steady state for most of the time. Fig. 8 shows samples of images of one cycle of opening and closing of the discharge valve with the compressor operating at 40Hz with PVE 32 oil. The phenomena are similar to what has been observed during the compressor start-up, but the image is more obscure because of blockage caused by the oil mist between the video object and the camera. Therefore, filters were applied to distinguish the oil cloud boundary and measure the distance in the video frames. The filter chosen in the case of processing these videos was a sharpening filter which enhances the edges and sharpens transitions, so it becomes easier to pinpoint features such as the edge of the valve or the edge of a droplet cloud in movement.
Fig. 8 – Discharge reed valve opening and closing at the steady state of a compressor Though the image is relatively blurry, it is still possible to watch the video frame by frame and determine when oil is atomized by valve opening. The video clearly showed a cloud of oil droplets are injected
4.
Size and velocity distribution of oil droplets
As shown by visualization, oil droplets are formed by the periodic movement of the reed valve and then entrained by the refrigerant vapor. Not satisfied with the just qualitative description of oil atomization by visualization, we made more progress to quantify the oil mist in the plenum and calculated the droplet size distribution. Using the video processing techniques mentioned in section 2.3, oil droplets are identified, and the size distribution is calculated. Distributions can be weighed by either number or mass. As shown in Fig. 9(a), normalized distribution weighed by number is straightforward, which represents a number of droplets in a particular diameter range captured by video. Normalized distribution weighed by mass, however, is more practical for oil mist study because the mass of the oil is more of interest. Shown in Fig. 9(b), the peak of the mass distribution histogram shifts right from that of the number distribution because larger droplets contribute more mass to the total mass. For one specific working condition, normalized distribution histogram is good enough to provide an overview of oil droplet size. However, more information is needed when we want to compare the oil flow under different working conditions. In this study, compressor frequency and mass flow rate are the main variable between different test conditions. So mass flow rate contributed by a group of droplets with the same diameter range is multiplied by the normalized mass distribution. In this way, a distribution weighed by mass flow rate is shown in Fig. 9(c). The histogram shape of mass flow rate distribution is same as that of normalized mass distribution. The only difference is that the y-axis of mass flow rate represents the mass flow rate per 6
ACCEPTED MANUSCRIPT distribution weighed by mass, multiplied by corresponding mass flow rate. 4.1. Droplet size distribution at the valve front With the mass flow rate distribution mentioned above, the oil droplet size distribution at the valve front and discharge pipe for the same compressor at different motor speed are shown in Fig. 10. The majority of oil droplets are in the range of diameter from 50 μm to 500 μm. If comparing the droplet size distribution under different compressor speeds (from 30 Hz to 60 Hz), it concludes that more oil droplets appear in the valve front, with increasing compressor speed. The trend is consistent with the rule of thumb that higher compressor speed gives higher OCR. As the compressor frequency increases, so do the mass flow rate of the system and correspondingly the vapor velocity at the discharge orifice. Another way to characterize a misty flow is by computing several different mean diameters that can characterize a certain histogram. Commonly used means can be the number mean, which only considers how many drops are grouped at a certain diameter, or also more meaningful means such as the volume mean which directly correlates to where the mass of the distribution can be found. As shown in Fig. 11, both the volume mean diameter of the oil mist at the valve front and that in the discharge pipe decrease as the compressor frequency increases. This is reasonable because higher kinetic energy atomizes the oil droplets into a finer dispersion. The droplet size distribution in the compressor plenum obtained in this work can be comparable with the results published by (Toyama et al., 2006). The mean oil droplet diameter of their high-pressure shell scroll compressor is around 100 μm while the volume mean diameter of our case varies from 100 to 300 μm depending on the location and operating conditions.
ED
M
AN US
CR IP T
droplet diameter interval. The area under the distribution curve represents the total oil mass flow rate captured the video frame. Here, it is assumed that droplets in different sizes move in the same bulk velocity.
AC
CE
PT
Fig. 9 – Histogram sample of oil droplet size distribution: (a) distribution weighed by number; (b) distribution weighed by mass; (c)
7
AN US
CR IP T
ACCEPTED MANUSCRIPT
4.2. Droplet size distribution in the discharge pipe
M
Fig. 10 - Oil size distribution from the video at the valve front and at discharge tube 20 cm downstream of the discharge under different compressor speeds
PT
ED
The droplet size distribution in the discharge pipe is also captured by the video under different test conditions. As shown in Fig. 10, more oil droplets appear in the discharge pipe with higher compressor speed. If comparing the droplet size distribution in the valve front to that in the discharge tube, it indicates that bigger droplets convert into smaller droplets and smaller droplets tend to flow out of the plenum with the refrigerant vapor. The volume mean diameter of the oil mist is also plotted in Fig. 11.
CE
4.3. Droplet velocity distribution in the discharge pipe
AC
Due to the tri-dimensional nature of the flow at the vicinity of the valve and in the plenum, only a 2-D flow field can be determined at the locations where videos were taken. The velocities at the vicinity of the valve were already explored by Zimmermann and Hrnjak (2014a). Fig. 11 - Volume mean diameter of oil mist at the valve front and in the discharge pipe
8
ACCEPTED MANUSCRIPT
CR IP T
moving in all directions in the discharge plenum. Some oil droplets deposit on the wall and form a layer of the oil film. Smaller oil droplets may be generated after the interaction between the droplet and oil film, which is called “re-entrainment”. From the observation in the discharge pipe, the average size of the oil droplets become smaller when they move from the valve opening to the compressor discharge pipe. With the limitation of exposure strength and view angle, it is difficult to depict the complete flow regime in the compressor plenum by visualization only. For a better understanding of oil droplets movement in the plenum, CFD simulation is a preferable approach.
AN US
Fig. 13 Qualitative description of oil droplet formation and movement in the discharge plenum
ED
M
5.2. CFD set-up
PT
Fig. 12 - Droplet speed distribution in the plenum at different compressor speeds
AC
CE
Fig. 12 shows the velocity distributions at the vicinity of the valve. The velocity distribution of working condition at 30 Hz is not shown here because the videos at relatively low compressor speed contain too few droplets to provide a reasonable estimation of droplet velocity. As the compressor speed increases, there is a shift in the peak speed distribution towards higher values. Also, by comparing the two charts we see that for the higher viscosity of the oil the peak in speed distribution is at a lower value. It is also worth saying that the bulk direction of the flow observed in the videos is mainly downwards in a mix of motion dominated by settlement of larger droplets and the bulk movement of the vapor being downwards since the inlet to the discharge tube going to the condenser is at a lower level than where the video was captured.
5.
Movement of oil droplets
5.1. General description of oil droplet flow Based on the oil flow video, the formation and movement of oil droplets can be described qualitatively in Fig. 13. The break-up of oil film during the opening of the reed valve generates oil droplet cloud. The oil droplets are entrained by the compressed refrigerant vapor and
Discrete phase model is commonly used to simulate one fluid or solid particle dispersed in another continuous fluid phase in a Lagrangian frame of reference. An important preliminary assumption made in the discrete phase model is that the dispersed phase occupies a low volume fraction. In this research, oil flow inside compressor discharge chamber is misty and oil volume ratio is estimated less than 5%, which satisfies the assumption of the discrete phase model. In this research, we use the Euler-Lagrange approach. The fluid phase is treated as a continuum by solving the time-averaged NavierStokes equations, while the dispersed phase is solved by tracking many droplets through the calculated flow field. The dispersed phase can exchange momentum with the fluid phase. Mesh independence is also carried out to achieve a balance between discretion error and calculation time. Six different cases with different element numbers, from 0.1 million to 1.5 million, are tested and the average velocity in the discharge pipe is used as a reference to check the mesh independence. In the final simulation, the mesh number is around 1 million and the mesh type is tetrahedron with acceptable skewness and mesh quality. For discrete phase model, discrete phase is introduced in the simulation by defining an injection. As shown by visualization, oil droplets are generated at the valve when it opens. So, the inlet of the flow region is set as the surface between the center reed valve and the valve base. “droplet pulsating injection” instead of dynamic mesh is applied to simulate the actual oil injection because the plenum space is the main research object and the simplification of the opening of the valve can save significant calculation resources. Droplet size distribution of the injection is obtained from the video processing results at the valve vicinity. Droplet injection direction and vapor phase velocity is gotten from simplified reed valve model from Zimmermann and Hrnjak (2015). The outlet of the flow region is the cross-section of the discharge pipe. The boundary condition is simplified to four different cases: trap, escape, reflect and wall film. Trap condition means that the droplet will be absorbed by the wall while escape condition allows the droplet to pass through when they hit the boundary. A reflect wall will give an opposite velocity vector to the incoming droplet based on momentum balance. Wall-film model is the most realistic case, which allows a single component liquid droplet to impinge upon a boundary surface and form a thin film. For the CFD simulation carried out in this study, 9
ACCEPTED MANUSCRIPT the discrete phase boundary condition of all the inner wall of the compressor plenum is set as wall film. Inlet is set as the discrete phase injection while the outlet of the discharge pipe is set as an escape. The effect of turbulence is considered using k-ε model. The simulation takes SIMPLE as the scheme and uses second order for pressure discretization and second-order upwind for momentum discretization. The time step for transient calculation should be as small as possible but here it is set as 0.001s for reasonable computational time.
Fig. 15 – Contour of the droplet number density shown in middle plane (working condition: f = 40 Hz, Pcpro = 954 kPa, Tcpro = 77.1 °C, ṁ =20.7 g·s-1, OCR = 0.6%)
ED
M
AN US
The trajectories of oil droplets are simulated by CFD method mentioned in the previous section using ANSYS Fluent. Simulation cases with the realistic 3D geometry of modified compressor under four different conditions are carried out. The following results are shown in the middle plane cutting through the centerline of the central reed valve and the discharge pipe. Vapor flow field is the first concern because the movement of oil droplets are mainly driven by the gravity and the drag caused by vapor phase flow. Fig. 14 shows the refrigerant vapor path lines, colored by velocity magnitude. The vapor comes from the opening of the reed valve and slows down dramatically because of the large volume of the plenum comparing to the compression chamber. The flow finally heads to the discharge pipe where the refrigerant vapor accelerates again.
CR IP T
5.3. Simulation results
PT
Fig. 14 – Refrigerant vapor flow path lines colored by velocity magnitude (working condition: f = 40 Hz, Pcpro = 954 kPa, Tcpro = 77.1 °C, ṁ =20.7 g·s-1, OCR = 0.6%)
AC
CE
With the known vapor flow field, discrete phase model updates the oil droplet trajectory every 10 iterations. Fig. 15 shows the contour of the droplet number density shown in the middle plane. Droplet number density is defined as how many droplets in one cubic centimeter. It can be told from the colormap that most of the droplets fled out of the plenum through the discharge pipe but some of them are stuck in the lower part. Droplet number density gives a general idea where the oil droplets go but the more specific trajectory of oil droplets in different sizes is needed to explain the difference between the droplet size distribution in the discharge tube and that in the valve front.
Fig. 16 – Droplet trajectories of three groups of droplets with different droplet diameter (60 μm, 240 μm and 600 μm) When the oil droplets hit the boundary, they will either bounce back or splash into more droplets or be absorbed by the wall (presenting absorption in the film). Some of the oil droplets can and will leave the plenum and are recorded by the monitor at the compressor discharge tube. Fig. 16 shows the trajectories in the plenum space of three groups of tracked oil droplets, colored by droplet diameter. Larger droplets tend to settle down in the film at the bottom of the compressor while smaller droplets tend to flow along with the vapor stream. Also, larger droplets may break up into smaller ones but only droplets in the same size are tracked in Fig. 16. The difference of trajectories of small, medium and large droplets can explain the difference between the droplet size distribution in the front of the valve and that in the discharge pipe. 5.4. 10
Validation of simulation results
ACCEPTED MANUSCRIPT
AN US
CR IP T
Most CFD simulation research needs verification and validation to ensure its reliability. Oil droplet size distribution is chosen as the key parameter to compare visualization results and CFD results. This is mainly because droplet size distribution is the most reliable and statistically representative parameter that can be obtained from both approaches and it is important for oil separator design. The droplet size distribution captured by visualization is compared to the result simulated by CFD under the same geometry and working conditions. The logic and process of comparison and validation are shown graphically in Fig. 17. Video taken near the valve is used to give them as the input parameters for injection, like oil droplet size and velocity. The vapor velocity is given by the pulsating flow model. On the other side, the video taken at the discharge tube is processed to give a droplet size distribution. These two distributions are compared to see if the CFD results can agree with the video results.
Fig. 17 – The process of validation of CFD results based on droplet size distribution from visualization
AC
CE
PT
ED
M
Both experiments and simulation are carried out at various compressor speeds in steady state. Fig. 18 shows the comparison of droplet size distribution between CFD results and video results. The curve represents the injected oil droplet size distribution, which is set as the input information of the injection in the simulation. The intention is to compare the distribution curve by CFD (dash line) and the distribution curve by video (solid line) in each plot. By comparison, it shows the CFD results can match well with video results at 30, 40 and 50 Hz. The discrepancy between the two results at 60 Hz may be explained by high OCR and high mass flow rate of oil injected by valve frequent opening will lead to a high possibility of collision between droplets. More droplets collision will make smaller droplets coalesce into bigger ones. Since the interaction between droplets is not considered in the simulation, it is possible that the simulation underestimates the number of large droplets and overestimate the number of smaller droplets at the compressor discharge. This is one of the limitations of CFD simulation as an approach to studying the oil mist inside the compressor. The results show that CFD simulation can give a qualitative prediction of the droplet size distribution at the compressor outlet. From the comparison between droplet size distribution of the injection and that of the discharge, large droplets are more likely to hit the wall and splash on the wall film to generate more small droplets. Smaller droplets are less affected by gravity and inertia so that they can catch up with the vapor flow to escape the compressor and form the annular-mist flow in the discharge tube.
Fig. 18 – Comparison between video results and CFD results of droplet size distribution under different compressor speed
6.
Conclusions
This paper illustrates and analyzes the formation, distribution, and movement of oil droplets inside the discharge plenum of a compressor by visualization and simulation. Two main approaches are developed and presented: CFD simulation and visualization based on the high-speed video. Visualization of the reed valve shows the mechanism of oil droplet formation. Processed oil 11
ACCEPTED MANUSCRIPT
Acknowledgment
Abbreviations and variables Droplet diameter [μ The increment of droplet diameter [μm] Mass flow rate [g·s-1] Oil circulation ratio Pressure [kPa] Polyalkylene glycol Polyolester Polyvinylether Temperature [oC]
Subscripts cpro
Compressor refrigerant outlet
CR IP T
CE
PT
ED
d di ṁ OCR P PAG POE PVE T
M
The authors would like to acknowledge the companies that sponsoring the Air Conditioning and Refrigeration Center for their technical and financial support to this work. The authors would like to thank for Augusto Zimmermann who made the compressor prototype and built a solid foundation of the visualization work. The authors also thank Scott Wujek for developing the video processing program.
References
Quantify Behavior of Oil Droplets in a Scroll Compressor, in: International Compressor Engineering Conference. Paper 1788. Wujek, S., Hrnjak, P., 2011. Mist to annular flow development quantified by novel video analysis. ACRC Report TR285. Urbana. Xu, J., Hrnjak, P., 2017a. Quantification of flow and retention of oil in compressor discharge pipe. Int. J. Refrig. 80, 252–263. https://doi.org/10.1016/j.ijrefrig.2017.05.004 Xu, J., Hrnjak, P., 2017b. Oil flow at the scroll compressor discharge: Visualization and CFD simulation. IOP Conf. Ser. Mater. Sci. Eng. 232. https://doi.org/10.1088/1757-899X/232/1/012051 Xu, J., Hrnjak, P., 2016a. Refrigerant-Oil Flow at the Compressor Discharge, in: SAE Technical Paper 2016-01-0247. https://doi.org/10.4271/2016-01-0247 Xu, J., Hrnjak, P., 2016b. Visualization and Simulation of Oil Flow in a Scroll Compressor Plenum Corresponding author, in: International Compressor Engineering Conference. Paper 2507. Yokoyama, T., Shingu, K., Sekiya, S., Maeyama, H., Nishimura, N., 2012. CFD Analysis inside a CO2 Rotary Compressor Shell to Improve Oil Separation and Reduce the Shell Size, in: International Compressor Engineering Conference. Paper 2074. Zhu, J., Botticella, F., Elbel, S., 2018. Experimental Investigation and Theoretical Analysis of Oil Circulation Rates in Ejector Cooling Cycles. Energy. Zimmermann, A., Hrnjak, P., 2015. Oil flow at discharge valve in a scroll compressor. Proc. Inst. Mech. Eng. Part E J. Process Mech. Eng. 229, 104–113. https://doi.org/10.1177/0954408914559315 Zimmermann, A., Hrnjak, P., 2014a. Visualization of the Opening Process of a Discharge Reed Valve in the Presence of Oil, in: International Compressor Engineering Conference. Paper 2369. Zimmermann, A., Hrnjak, P., 2014b. Source Identification and In Situ Quantification of Oil-Refrigerant Mist Generation by Discharge Valve Opening Process, in: International Compressor Engineering Conference. Paper 2367.
AN US
mist videos provide the droplet size and velocity distributions under different working conditions. CFD simulation uses part of the information from the video as the input of discrete phase model. The simulation results are validated by comparing of oil drop size distribution at the compressor discharge tube obtained by CFD and that obtained by video processing. With these two approaches, a complete set of tools has been established to study similar multiphase problem inside the compressor. The phenomenon of oil flow is clearly captured by high-speed videos. Oil droplets are formed from the break-up of the oil film between the reed valve and the valve base. The oil droplets are then entrained by the compressed refrigerant vapor. Some of them deposit into the oil film on the inner wall while some of them escaped the compressor plenum and form mist-annular flow in the discharge pipe. In general, higher compressor speed introduces more oil droplets with slightly smaller size. Larger droplets tend to be retained in the plenum while smaller droplets are more likely to follow the refrigerant vapor flow to escape. This can explain why higher compressor speed brings higher OCR. The discharge plenum space of a vertical scroll compressor has the potential to separate large oil droplets. One cannot design the separator structure without knowing the characteristics of the oil mist. The illustration and analysis of oil droplet behavior contribute significantly to the design the geometry of compressor plenum with a goal of minimizing the oil discharge rate.
AC
Bhatia, K., Pitla, S., Khetarpal, V., 2002. Computational Analysis Of An Oil Separator, in: International Compressor Engineering Conference. Paper 606. DeAngelis, J., Hrnjak, P., 2005. Experimental study of system performance improvements in transcritical R744 systems with applications to bottle coolers. Air Conditioning and Refrigeration Center. College of Engineering. University of Illinois at UrbanaChampaign. Hay, K.J., Liu, Z.-C., Hanratty, T.J., 1998. A backlighted imaging technique for particle size measurements in two-phase flows. Exp. Fluids 25, 226–232. https://doi.org/10.1007/s003480050225 Li, H., Hrnjak, P., 2014. Lubricant effect on performance of R134a MAC microchannel evaporators. SAE Technical Paper. Toyama, T., Matsuura, H., Yoshida, Y., 2006. Visual Techniques to 12
Formation, distribution, and movement of oil droplets in the compressor plenum Jiu Xu , Pega Hrnjak PII: DOI: Reference:
S0140-7007(18)30226-3 10.1016/j.ijrefrig.2018.06.020 JIJR 4029
To appear in:
International Journal of Refrigeration
Received date: Revised date: Accepted date:
13 March 2018 11 June 2018 25 June 2018
Please cite this article as: Jiu Xu , Pega Hrnjak , Formation, distribution, and movement of oil droplets in the compressor plenum, International Journal of Refrigeration (2018), doi: 10.1016/j.ijrefrig.2018.06.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
CR IP T AN US M ED PT CE AC
Highlights: Demonstrate the formation oil droplets near a compressor discharge valve. Quantify and explain the oil mist distribution in a scroll compressor plenum. Simulate the movement of oil droplets by CFD and verified by visualization results. Provide upstream information for the design of oil separation structures.
1
ACCEPTED MANUSCRIPT
Formation, distribution, and movement of oil droplets in the compressor plenum Jiu Xu a, Pega Hrnjak a,b,* an Air Conditioning and Refrigeration Center, University of Illinois at Urbana-Champaign 1206 West Green Street, Urbana, IL 61801, USA b Creative Thermal Solutions, 2209 Willow Rd., Urbana, IL, USA * Corresponding author: [email protected] +1-217-390-5278 ABSTRACT
1.
CR IP T
Oil is important to compressors, but it brings some side effects on the performance of refrigeration systems. Discharge valve or orifice in the compressor plenum is the gateway for the oil to leave the compressor and circulate in the system. The space of the compressor plenum has the potential to separate the oil mist from the refrigerant so that oil circulation ratio (OCR) is reduced. In this study, the formation, distribution, and movement of oil droplets in the plenum are investigated through visualization and CFD simulation. Videos of oil atomization are captured by the high-speed camera to describe how oil mist is formed. The distribution of droplet size and velocity is quantified by video processing techniques. Discrete phase model gives the numerical approach to study the oil droplet movement inside compressor plenum. The numerical simulation is verified by comparing visualization results with simulation results. The oil flow phenomenon and its quantification provide important information for the design of separation structure incorporated in the compressor. Also, a set of tools are developed and verified to study the oil droplets behavior in different types of compressor plenums.
Introduction
AN US
1.1. Background
Zimmermann and Hrnjak (2014, 2015) showed that oil droplets are formed by the break-up of oil film between the valve and the valve seat during the opening and closing in a scroll compressor. The visualization of the reed valve in the scroll compressor shows the valve periodical movement together with the oil atomization. Xu and Hrnjak (2016, 2017) measured the OCR with a non-invasive way based on the oil annular mist flow visualization in the compressor discharge pipe. Effect of focal depth is considered during the video processing to provide a good estimation of the number of oil droplets to provide the amount of oil in that form in a specific control volume. Although flow visualization provides the image of the oil mist when part of the compressor is made transparent, this approach is restricted by narrow view angle and limited light exposure. Videos can provide the droplet flow velocity in the two-dimensional case, but it is much more difficult to estimate the velocity in the third dimension. Also, the velocity field of the vapor phase and slippery velocity between the droplet and vapor phase are hard to obtain through videos. Therefore, computational fluid dynamics is used as an additional numerical approach to simulate the complicated two-phase flow. Bhatia et al. (2002) presented a CFD case that analyzed the oil flow in the centrifugal type oil separator. Design suggestions were given to achieve minimum pressure drop and maximum separation efficiency for an oil separator integrated into a swashplate reciprocating compressor. Yokoyama et al. (2012) analyzed the oil flow in a CO2 rotary compressor shell by CFD. Numerical simulations showed that flattening the lower balance weight or installing a rotating disk over upper balance weight can reduce the OCR and the results are verified by system experimental results. Xu and Hrnjak (2016b, 2017b) validated the CFD simulation model using visualization results for oil droplet flow inside the scroll compressor plenum. CFD is a favorable approach when flow details are needed under the realistic geometries and working conditions are. The verification and validation of the CFD simulation is the key to ensure the reliability of calculation results. Overall, previous researchers have developed some strong tools to study the in-compressor oil flow. Oil mist has been observed in the compressor plenum under different working conditions of different types of compressors. However, a combination of visualization and CFD is still uncommon. This justifies further study of oil mist flow in the compressor using both approaches.
CE
PT
ED
M
Oil is necessary for cooling, lubrication, and sealing of conventional compressors but oil circulating in the refrigeration cycle typically has adverse effects on heat transfer and pressure drop (DeAngelis and Hrnjak, 2005; Li and Hrnjak, 2014). Besides the traditional vapor compression cycle, oil circulating in the ejector cycle will also bring some side effects (Zhu et al., 2018). Even application of magnetic bearings and new material makes oil-free compressors possible, their number is still very small. Therefore, engineering attempts in the design of compressors are made to keep the oil in the compressor and reduce the oil circulation ratio (OCR) in the system. Separating the oil mist from the high-pressure refrigerant vapor and draining it back to the compressor is a feasible way to achieve this goal. How oil droplets and film flow with the refrigerant vapor is the key information for oil separator design. Important characteristics of oil flow include oil droplet size distribution, oil droplet velocity distribution, oil droplet trajectories and refrigerant vapor velocity field. Due to desired compactness, it is preferred to integrate the oil separation structure into the compressor rather than having an external separator between the compressor and the condenser. These factors justify a closer look at the oil flow inside the compressor plenum, especially the formation, distribution and movement of oil droplets. 1.2. Literature review
AC
Oil flow inside the compressor plenum and discharge tube has been studied mainly by two approaches: flow visualization and computational fluid dynamics (CFD) simulation. As one of strong tool to study misty flow, flow visualization is the most intuitive approach to understand how oil droplets flow inside the compressor shell. Toyama et al. (2006) described methods to quantify oil droplet behavior using high-speed photography in the scroll compressor shell via sight glasses. The size distribution of oil droplets is described by the video. It was concluded that the mean diameter of oil droplets decreases as the flow speed of the refrigerant vapor increases. Wujek and Hrnjak (2011) developed visualization techniques to quantify the annular mist flow in the compressor discharge pipe. The oil film thickness is measured by the laser reflection method. The size and velocity of oil droplets captured by the camera were determined by video processing program. Using a swash-plate compressor with R134a and PAG 46 oil, it showed the developing annular-mist flow along the discharge pipe.
1.3. The focus of this study
2
ACCEPTED MANUSCRIPT
2.
CR IP T
The objective of this paper is to analyze the oil droplet flow in the compressor plenum by both visualization and CFD simulation. By modifying a scroll compressor plenum, a transparent prototype is made for visualization. A high-speed camera captures the flow videos at different locations in the compressor plenum. The videos are then processed to provide flow details of the oil mist. Discrete phase model is used to simulate the oil droplet trajectory based on the vapor flow field and droplet size distribution near the reed valve. The simulation results are then verified with the droplet size distribution given by video processing. The paper aims to provide a comprehensive and detailed understanding of the formation, distribution, and movement of oil droplets in the compressor plenum. It provides the design guidelines for oil separation structures integrated into the compressor plenum. The verified method for studying oil mist can also be applied in further research.
Experimental Methods
2.1. System facilities and test conditions
AC
CE
PT
ED
M
AN US
A full vapor compression cycle is built to test and visualize the compressor under different working conditions. The diagram of experimental facilities is shown in Fig. 1. The working condition is controlled by adjusting the capacity of condenser or evaporator, and by adjusting the opening of the expansion valve. Compressor suction pressure and temperature, compressor discharge pressure and temperature are measured by the pressure transducers and thermocouples. OCR sampling and mass flow rate measurement takes place in the liquid line between the condenser and the expansion valve. OCR is determined by sampling method according to ASHRAE standard 41.4. A Coriolis type mass flow meter measures the mass flow rate of the system. The compressor used in this study is a 3-ton scroll compressor with a low-pressure shell arrangement. It was designed for operation at 50 Hz, however, in this work, a variable frequency drive is used to set the compressor operating frequency to the desired level. The compressor was originally designed to use R-410A as the refrigerant for residential air conditioning application. To ensure the reliability of transparent material for visualization, R-134a is used as the refrigerant to provide relatively lower compressor discharge pressure. Most of the experiments are carried out with PVE 32 oil but POE 120 was also used to study the effect of oil properties.
Fig. 1 – System facilities for experiments The testing conditions were selected to provide the most realistic operation under air conditioning conditions while assuring the safety of the experiment and operators by keeping the discharge pressure and temperature levels low enough to avoid failure of the transparent part. The compressor operating frequency was varied from 30Hz to 60 Hz while the pressure ratio and pressure levels were kept relatively close, so the only change should be the refrigerant mass flow rate and the OCR as a result of more or less oil pumping in the crankcase. The detailed test conditions are listed in Table 1. OCR determined by sampling methods has the maximum absolute uncertainty of ±0.2%. Temperature measurements carry an uncertainty of ±0.2°C, and pressure measurements are accurate to ±0.25%. Mass flow rate carries uncertainty of ±0.5%.
Table 1 – Detailed test conditions Test case No. Compressor frequency [Hz] System mass flow rate [g·s-1] Suction pressure [kPa] Discharge pressure [kPa] Superheat [K] Discharge temperature [°C] Oil circulation ratio [%] Oil mass flow rate [g·s-1]
1 30 16.7 337.9 906.8 21.3 78.0 0.15 0.025
2 40 20.3 309.7 954.4 23.0 77.1 0.6 0.124
3 50 21.9 312.7 961.0 24.3 81.1 0.98 0.214
4 60 23.5 315.3 987.2 24.6 88.7 1.17 0.275
2.2. Modification of the compressor The oil flow inside the compressor plenum was visualized by highspeed CCD camera. Some modifications were made to the upper part of the scroll compressor plenum to improve imaging of the oil droplets inside the plenum. Firstly, two flanges were added between the top cover and the compressor shell to provide visibility to the discharge valves. Initially mounted four sight glasses were not sufficient to bring necessary light for the camera. For that reason and better visual access, the transparent cylindrical ring was added between the flanges (Fig. 2).
3
ACCEPTED MANUSCRIPT clear image of droplets, but the additional light had to be brought from the front to capture the movement of the reed valve. Videos of the oil flow were taken at three locations: center reed valve, valve front (the vicinity area in front of the valve) and discharge pipe. The locations of video frames are marked in red and the corresponding camera view angles are marked in green in Fig. 4. Videos of the center reed valve and videos of the valve front used the same camera position at a different focal distance. By varying the lens, the frame was focused on either the valve or the injected droplets.
AN US
CR IP T
Fig. 2 – Outer shape of the modified plenum of a scroll compressor
Fig. 4 – Locations of video taken in the modified plenum
Fig. 3 – Opened plenum of a scroll compressor
AC
CE
PT
ED
M
As shown in Fig. 3, a transparent screen was placed between the transparent flange wall and the valve to remove the interference of fine oil mist. In this way, a clear vision can be achieved at a plane about 5 mm from the tip of the valve. The compressor used in this project had three discharge valves, two of which were intentionally forced to be closed to reduce the complexity of oil flow. Modification mentioned above added around 30% of total volume of the original geometry to the original geometry of the scroll compressor plenum and made the relative position of discharge pipe higher than that in realistic design. It must be mentioned that such modification provided the access to visualization at the cost of some changes to the original vapor and oil droplets flow. The flow details obtained in the modified compressor differs from that in the original compressor, but authors believe it is acceptable as validation of the CFD simulation is done for the same geometry after modification. Furthermore, the geometry of plenum varies with different types of compressor. The prototype made in this study is just one example how the visualization is realized. Though the difference of the plenum shape can be significant, the same method can be applied to investigate of oil flow in the compressor plenum. As an approach to studying the oil droplets inside the compressor, visualization also has its limitation. A part of compressor shell must be transparent before the internal flow can be seen, so some modification of the original geometry is inevitable. Also, transparent material like plastic or glass is typically much more fragile than metal so that the compressor discharge pressure and temperature must be maintained at a relatively low level to prevent damage. 2.3. Visualization set-up High-speed camera Phantom 6.0 was used with different lenses. The frame rate was kept higher than 1000 fps to capture the fast movement of oil droplets. The light was provided from the back for a
When taking the videos of the oil droplets in the valve front and those in the discharge pipe, backlighting was used since it is the preferred mode for achieving good contrast and crisp images of the droplets. On the case of the center reed valve, backlighting did not help to make the image better, so front lighting was used. When using front lighting, instead of concentrating and absorbing the light, the droplets and edge of the reed valve appear as a reflection, which poses some challenge to the image processing procedure. Fig. 5 (a) shows a sample frame from a video. The focus is on the tip of the reed valve from about 45o. Zimmermann and Hrnjak (2014b) have shown that the break-up pattern of the oil film between valve and valve seat. A cloud of oil droplets injected by the high-pressure refrigerant vapor as the valve opens. Video processing method is used to quantify the initial velocity of oil droplets injected into the plenum during the oil film break-up. More details about the formation of oil droplets will be discussed in section 3. Fig. 5 (b) shows a sample frame from a video taken at the valve vicinity when focusing on the transparent screen in front of the valve. It is processed to estimate the size distribution of the oil droplets injected in the valve opening. In the plenum chamber, the oil droplets are moving in random directions at a relatively low velocity (0.5-2.0 m·s-1). The droplet size can be better estimated than the droplet velocity because of the difficulties to assess three-dimensional vector with one camera. Therefore, oil droplet size distribution is the key parameter that is used to compare visualization and simulation results. Fig. 5(c) shows the frame taken in transparent discharge pipe located around 20 centimeters from the compressor discharge port. The view in a raw form presents droplets (dark spots) behind the film with waves. During quantification and analysis, several filtering procedures were conducted to better capture and measure the size, number, and velocity of droplets. Since the flow is strongly one dimensional, droplet velocities determined are closer to the actual values. When estimating the total number of droplets in the tube authors have first determined the depth of view, then the number of droplets in the view volume and their diameter and its distribution and then the total mass in the droplet form. 4
ACCEPTED MANUSCRIPT Details of the method were presented in recently published papers (Xu and Hrnjak, 2017a).
3.
Formation of oil droplets
3.1. Oil atomization at the start-up
AN US
CR IP T
The gateway for the oil to leave the compression chamber is the valve opening gap or the discharge orifice. Therefore, in our prototype, the reed valve is the main research objective to study the formation of oil droplets in the compressor plenum. Start-up is usually the condition when the compressor experiences its highest mass flow rate (under equalized pressures) and consequently the highest lift on the valve because the pressure difference is the greatest over the first few operating cycles. To improve the video quality, the authors tried to capture the video of oil atomization at the start-up, avoid foggy foreground in steady-state operation. In the beginning, the oil mist inside the discharge plenum originated mostly from liquid film accumulated in between the reed valve and its seat. After a few opening cycles of the valve, the discharge plenum is inundated with liquid coming from the system to the compressor suction due to the sudden pressure change. Fig. 6 shows the starting process sequence of the compressor after a long off-period with R134a and PVE 32 oil. The video is taken by the camera with a 55 mm lens at the resolution of 256*256 and frame rate of 4600 fps. The pictures show the evolution in time from the moment the compressor is turned on until there is a first full opening of the valve. When the valve hits the stop ((f)), a large amount of liquid is pushed through the discharge orifice. On the intermediary steps, the interaction between the small amount of vapor coming out of the discharge port and any oil that was present in the plenum near the valve is seen, and large droplets (>1mm) are being generated.
Fig. 5 - Sample video frames taken in different locations inside the compressor: (a) center reed valve; (b) valve front; (c) discharge pipe 2.4. Video processing
AC
CE
PT
ED
M
For the videos that used backlighting, the image processing procedure used in this work is the same as that used by Wujek and Hrnjak (2011). The procedure consists of first eliminating any noise by averaging the pixel values at each location on the image over the entire period of the video and subtracting it from each frame. After the noise elimination, the processing code searches for peaks and valleys between frames and applies a cross-correlation to determine where the peaks and valleys coincide. This is used to remove the waviness of any liquid film that might be present on the walls. Once this clean image is present, a thresholding process is applied to the in which the same is transformed into a binary matrix with a cut-off level. With the help of built-in functions in Matlab, the code then applies a maximum derivative approach to identify the droplets in the raw image. The algorithm goes back to the original clean image and looks for the slope of the sharpest change in pixel values along the horizontal and vertical axis of the identified droplet. A minimum slope criterion is then applied to discard droplets that might not be in focus. Other criteria such as aspect ratio of the major and minor axis, the region being darker than the average surrounding pixels are also applied to discard possible false droplet or droplets which dimensions cannot be identified with the desired accuracy. The uncertainty associated with these methods can be found well documented in (Hay et al., 1998) and the limiting factor was cited to be the pixel width of the camera sensor for these types of flows. In their work Hay et al., 1998, estimated the resolution in their method to be the width of half a pixel. The pixel width is a function of the magnification and the size of the pixel array of the camera. In this work the sensor array consists of 512 x 512 pixels and the resulting pixel widths were 51.5µm for the valve videos and 21.47 µm for the plenum videos, resulting in ±25.75 µm and ±10.73 µm. Considering that the minimum drop size identifiable was set to two pixels in width we can determine that the maximum uncertainty for the histograms is ±25% for the smallest droplet.
Fig. 6 - Discharge reed valve opening at the compressor start-up A shorter and more zoomed view is taken to capture the details of oil film break-up process. The video frames shown in Fig. 7 were taken with the same lens as Fig. 6 and the resolution of 128*128 at the frame rate of 19,000 fps. Also, for better visualization result, the light source and direction is changed from front-light to top and side light through sight glasses. Fig. 7 (a) to Fig. 7(f) show the time evolution of the startup procedure. Fig. 7 (b) shows that on the left side close to the tip of the valve, there is film breakup present and it also indicates that the flow around the valve is not symmetrical. Later in Fig. 7(c) and Fig. 7(d), the unstable structure of oil film cause discontinuities in the film, making certain portions of the film to protrude outwards. The deformed oil film then ruptures into smaller ligaments and atomizes into droplets. Fig. 7(e) and Fig. 7(f) show the later stages the camera view becomes blurry when a large number of droplets are generated. It needs to be mentioned here the time span of videos shown in Fig. 7 is much shorter than that in 5
ACCEPTED MANUSCRIPT Fig. 6 because the oil film break-up details can only be observed at the beginning of the compressor operation cycle.
from underneath the valve when the gap between the reed valve and the valve base is large enough to break the oil film (Fig. 8(b) and Fig. 8(c)). The front of the oil droplet cloud propagates further and then invisible in a few microseconds. The amount of oil injected by the atomization during the steady-state operation seems less than that during the start-up. Another piece of information that can be extracted from the videos at steady state operation is the valve cycling times. By tracking the bright spot of on the edge of the reed valve, the cycling period of valve opening, and closing can be calculated. The overall cycle time showed very good agreement with the period calculated by the frequency set on the variable frequency drive. 3.3. Discussion
CR IP T
The observation of oil droplet near the valve at the start-up and steady-state helps to understand the mechanism of oil droplet formation. One hypothesis of the formation is that oil is either entrained in the vapor refrigerant or scraped from the clearances between moving parts in the compression chamber. The other hypothesis is that most of the oil is dispersed in the plenum comes from the atomization of oil film between valve and seat. Though it is difficult to find direct evidence from the videos, some estimation can still be made to verify the hypothesis mentioned above. The oil volume being discharged by the valve opening in each cycle can be calculated based on the compressor frequency, mass flow rate, and oil circulation ratio. The volume of the gap between the valve and seat is estimated to be 20 mm3 while the discharged oil volume for each cycle varies from 2 mm3 to 5 mm3 depending on the compressor frequency. This means the gap can hold an order of magnitude more oil than the amount of oil actually discharged. Therefore, for this particular compressor design, the second hypothesis is more credible. When the reed valve closes, the gap between valve and base gets replenished with some oil and holds it until the valve opens again. When the reed valve opens, the continuous oil film is firstly stretched is firstly stretched and then atomized to form the oil droplet cloud. The oil mist is then entrained by the fast-moving refrigerant vapor and occupies the plenum.
AN US
Fig. 7 – A zoom-in view of discharge reed valve opening at the compressor start-up 3.2. Oil atomization at steady state operation
AC
CE
PT
ED
M
Oil atomization is then visualized during the steady state operation of the compressor. This is even more important than that at the start-up because compressor operates at a steady state for most of the time. Fig. 8 shows samples of images of one cycle of opening and closing of the discharge valve with the compressor operating at 40Hz with PVE 32 oil. The phenomena are similar to what has been observed during the compressor start-up, but the image is more obscure because of blockage caused by the oil mist between the video object and the camera. Therefore, filters were applied to distinguish the oil cloud boundary and measure the distance in the video frames. The filter chosen in the case of processing these videos was a sharpening filter which enhances the edges and sharpens transitions, so it becomes easier to pinpoint features such as the edge of the valve or the edge of a droplet cloud in movement.
Fig. 8 – Discharge reed valve opening and closing at the steady state of a compressor Though the image is relatively blurry, it is still possible to watch the video frame by frame and determine when oil is atomized by valve opening. The video clearly showed a cloud of oil droplets are injected
4.
Size and velocity distribution of oil droplets
As shown by visualization, oil droplets are formed by the periodic movement of the reed valve and then entrained by the refrigerant vapor. Not satisfied with the just qualitative description of oil atomization by visualization, we made more progress to quantify the oil mist in the plenum and calculated the droplet size distribution. Using the video processing techniques mentioned in section 2.3, oil droplets are identified, and the size distribution is calculated. Distributions can be weighed by either number or mass. As shown in Fig. 9(a), normalized distribution weighed by number is straightforward, which represents a number of droplets in a particular diameter range captured by video. Normalized distribution weighed by mass, however, is more practical for oil mist study because the mass of the oil is more of interest. Shown in Fig. 9(b), the peak of the mass distribution histogram shifts right from that of the number distribution because larger droplets contribute more mass to the total mass. For one specific working condition, normalized distribution histogram is good enough to provide an overview of oil droplet size. However, more information is needed when we want to compare the oil flow under different working conditions. In this study, compressor frequency and mass flow rate are the main variable between different test conditions. So mass flow rate contributed by a group of droplets with the same diameter range is multiplied by the normalized mass distribution. In this way, a distribution weighed by mass flow rate is shown in Fig. 9(c). The histogram shape of mass flow rate distribution is same as that of normalized mass distribution. The only difference is that the y-axis of mass flow rate represents the mass flow rate per 6
ACCEPTED MANUSCRIPT distribution weighed by mass, multiplied by corresponding mass flow rate. 4.1. Droplet size distribution at the valve front With the mass flow rate distribution mentioned above, the oil droplet size distribution at the valve front and discharge pipe for the same compressor at different motor speed are shown in Fig. 10. The majority of oil droplets are in the range of diameter from 50 μm to 500 μm. If comparing the droplet size distribution under different compressor speeds (from 30 Hz to 60 Hz), it concludes that more oil droplets appear in the valve front, with increasing compressor speed. The trend is consistent with the rule of thumb that higher compressor speed gives higher OCR. As the compressor frequency increases, so do the mass flow rate of the system and correspondingly the vapor velocity at the discharge orifice. Another way to characterize a misty flow is by computing several different mean diameters that can characterize a certain histogram. Commonly used means can be the number mean, which only considers how many drops are grouped at a certain diameter, or also more meaningful means such as the volume mean which directly correlates to where the mass of the distribution can be found. As shown in Fig. 11, both the volume mean diameter of the oil mist at the valve front and that in the discharge pipe decrease as the compressor frequency increases. This is reasonable because higher kinetic energy atomizes the oil droplets into a finer dispersion. The droplet size distribution in the compressor plenum obtained in this work can be comparable with the results published by (Toyama et al., 2006). The mean oil droplet diameter of their high-pressure shell scroll compressor is around 100 μm while the volume mean diameter of our case varies from 100 to 300 μm depending on the location and operating conditions.
ED
M
AN US
CR IP T
droplet diameter interval. The area under the distribution curve represents the total oil mass flow rate captured the video frame. Here, it is assumed that droplets in different sizes move in the same bulk velocity.
AC
CE
PT
Fig. 9 – Histogram sample of oil droplet size distribution: (a) distribution weighed by number; (b) distribution weighed by mass; (c)
7
AN US
CR IP T
ACCEPTED MANUSCRIPT
4.2. Droplet size distribution in the discharge pipe
M
Fig. 10 - Oil size distribution from the video at the valve front and at discharge tube 20 cm downstream of the discharge under different compressor speeds
PT
ED
The droplet size distribution in the discharge pipe is also captured by the video under different test conditions. As shown in Fig. 10, more oil droplets appear in the discharge pipe with higher compressor speed. If comparing the droplet size distribution in the valve front to that in the discharge tube, it indicates that bigger droplets convert into smaller droplets and smaller droplets tend to flow out of the plenum with the refrigerant vapor. The volume mean diameter of the oil mist is also plotted in Fig. 11.
CE
4.3. Droplet velocity distribution in the discharge pipe
AC
Due to the tri-dimensional nature of the flow at the vicinity of the valve and in the plenum, only a 2-D flow field can be determined at the locations where videos were taken. The velocities at the vicinity of the valve were already explored by Zimmermann and Hrnjak (2014a). Fig. 11 - Volume mean diameter of oil mist at the valve front and in the discharge pipe
8
ACCEPTED MANUSCRIPT
CR IP T
moving in all directions in the discharge plenum. Some oil droplets deposit on the wall and form a layer of the oil film. Smaller oil droplets may be generated after the interaction between the droplet and oil film, which is called “re-entrainment”. From the observation in the discharge pipe, the average size of the oil droplets become smaller when they move from the valve opening to the compressor discharge pipe. With the limitation of exposure strength and view angle, it is difficult to depict the complete flow regime in the compressor plenum by visualization only. For a better understanding of oil droplets movement in the plenum, CFD simulation is a preferable approach.
AN US
Fig. 13 Qualitative description of oil droplet formation and movement in the discharge plenum
ED
M
5.2. CFD set-up
PT
Fig. 12 - Droplet speed distribution in the plenum at different compressor speeds
AC
CE
Fig. 12 shows the velocity distributions at the vicinity of the valve. The velocity distribution of working condition at 30 Hz is not shown here because the videos at relatively low compressor speed contain too few droplets to provide a reasonable estimation of droplet velocity. As the compressor speed increases, there is a shift in the peak speed distribution towards higher values. Also, by comparing the two charts we see that for the higher viscosity of the oil the peak in speed distribution is at a lower value. It is also worth saying that the bulk direction of the flow observed in the videos is mainly downwards in a mix of motion dominated by settlement of larger droplets and the bulk movement of the vapor being downwards since the inlet to the discharge tube going to the condenser is at a lower level than where the video was captured.
5.
Movement of oil droplets
5.1. General description of oil droplet flow Based on the oil flow video, the formation and movement of oil droplets can be described qualitatively in Fig. 13. The break-up of oil film during the opening of the reed valve generates oil droplet cloud. The oil droplets are entrained by the compressed refrigerant vapor and
Discrete phase model is commonly used to simulate one fluid or solid particle dispersed in another continuous fluid phase in a Lagrangian frame of reference. An important preliminary assumption made in the discrete phase model is that the dispersed phase occupies a low volume fraction. In this research, oil flow inside compressor discharge chamber is misty and oil volume ratio is estimated less than 5%, which satisfies the assumption of the discrete phase model. In this research, we use the Euler-Lagrange approach. The fluid phase is treated as a continuum by solving the time-averaged NavierStokes equations, while the dispersed phase is solved by tracking many droplets through the calculated flow field. The dispersed phase can exchange momentum with the fluid phase. Mesh independence is also carried out to achieve a balance between discretion error and calculation time. Six different cases with different element numbers, from 0.1 million to 1.5 million, are tested and the average velocity in the discharge pipe is used as a reference to check the mesh independence. In the final simulation, the mesh number is around 1 million and the mesh type is tetrahedron with acceptable skewness and mesh quality. For discrete phase model, discrete phase is introduced in the simulation by defining an injection. As shown by visualization, oil droplets are generated at the valve when it opens. So, the inlet of the flow region is set as the surface between the center reed valve and the valve base. “droplet pulsating injection” instead of dynamic mesh is applied to simulate the actual oil injection because the plenum space is the main research object and the simplification of the opening of the valve can save significant calculation resources. Droplet size distribution of the injection is obtained from the video processing results at the valve vicinity. Droplet injection direction and vapor phase velocity is gotten from simplified reed valve model from Zimmermann and Hrnjak (2015). The outlet of the flow region is the cross-section of the discharge pipe. The boundary condition is simplified to four different cases: trap, escape, reflect and wall film. Trap condition means that the droplet will be absorbed by the wall while escape condition allows the droplet to pass through when they hit the boundary. A reflect wall will give an opposite velocity vector to the incoming droplet based on momentum balance. Wall-film model is the most realistic case, which allows a single component liquid droplet to impinge upon a boundary surface and form a thin film. For the CFD simulation carried out in this study, 9
ACCEPTED MANUSCRIPT the discrete phase boundary condition of all the inner wall of the compressor plenum is set as wall film. Inlet is set as the discrete phase injection while the outlet of the discharge pipe is set as an escape. The effect of turbulence is considered using k-ε model. The simulation takes SIMPLE as the scheme and uses second order for pressure discretization and second-order upwind for momentum discretization. The time step for transient calculation should be as small as possible but here it is set as 0.001s for reasonable computational time.
Fig. 15 – Contour of the droplet number density shown in middle plane (working condition: f = 40 Hz, Pcpro = 954 kPa, Tcpro = 77.1 °C, ṁ =20.7 g·s-1, OCR = 0.6%)
ED
M
AN US
The trajectories of oil droplets are simulated by CFD method mentioned in the previous section using ANSYS Fluent. Simulation cases with the realistic 3D geometry of modified compressor under four different conditions are carried out. The following results are shown in the middle plane cutting through the centerline of the central reed valve and the discharge pipe. Vapor flow field is the first concern because the movement of oil droplets are mainly driven by the gravity and the drag caused by vapor phase flow. Fig. 14 shows the refrigerant vapor path lines, colored by velocity magnitude. The vapor comes from the opening of the reed valve and slows down dramatically because of the large volume of the plenum comparing to the compression chamber. The flow finally heads to the discharge pipe where the refrigerant vapor accelerates again.
CR IP T
5.3. Simulation results
PT
Fig. 14 – Refrigerant vapor flow path lines colored by velocity magnitude (working condition: f = 40 Hz, Pcpro = 954 kPa, Tcpro = 77.1 °C, ṁ =20.7 g·s-1, OCR = 0.6%)
AC
CE
With the known vapor flow field, discrete phase model updates the oil droplet trajectory every 10 iterations. Fig. 15 shows the contour of the droplet number density shown in the middle plane. Droplet number density is defined as how many droplets in one cubic centimeter. It can be told from the colormap that most of the droplets fled out of the plenum through the discharge pipe but some of them are stuck in the lower part. Droplet number density gives a general idea where the oil droplets go but the more specific trajectory of oil droplets in different sizes is needed to explain the difference between the droplet size distribution in the discharge tube and that in the valve front.
Fig. 16 – Droplet trajectories of three groups of droplets with different droplet diameter (60 μm, 240 μm and 600 μm) When the oil droplets hit the boundary, they will either bounce back or splash into more droplets or be absorbed by the wall (presenting absorption in the film). Some of the oil droplets can and will leave the plenum and are recorded by the monitor at the compressor discharge tube. Fig. 16 shows the trajectories in the plenum space of three groups of tracked oil droplets, colored by droplet diameter. Larger droplets tend to settle down in the film at the bottom of the compressor while smaller droplets tend to flow along with the vapor stream. Also, larger droplets may break up into smaller ones but only droplets in the same size are tracked in Fig. 16. The difference of trajectories of small, medium and large droplets can explain the difference between the droplet size distribution in the front of the valve and that in the discharge pipe. 5.4. 10
Validation of simulation results
ACCEPTED MANUSCRIPT
AN US
CR IP T
Most CFD simulation research needs verification and validation to ensure its reliability. Oil droplet size distribution is chosen as the key parameter to compare visualization results and CFD results. This is mainly because droplet size distribution is the most reliable and statistically representative parameter that can be obtained from both approaches and it is important for oil separator design. The droplet size distribution captured by visualization is compared to the result simulated by CFD under the same geometry and working conditions. The logic and process of comparison and validation are shown graphically in Fig. 17. Video taken near the valve is used to give them as the input parameters for injection, like oil droplet size and velocity. The vapor velocity is given by the pulsating flow model. On the other side, the video taken at the discharge tube is processed to give a droplet size distribution. These two distributions are compared to see if the CFD results can agree with the video results.
Fig. 17 – The process of validation of CFD results based on droplet size distribution from visualization
AC
CE
PT
ED
M
Both experiments and simulation are carried out at various compressor speeds in steady state. Fig. 18 shows the comparison of droplet size distribution between CFD results and video results. The curve represents the injected oil droplet size distribution, which is set as the input information of the injection in the simulation. The intention is to compare the distribution curve by CFD (dash line) and the distribution curve by video (solid line) in each plot. By comparison, it shows the CFD results can match well with video results at 30, 40 and 50 Hz. The discrepancy between the two results at 60 Hz may be explained by high OCR and high mass flow rate of oil injected by valve frequent opening will lead to a high possibility of collision between droplets. More droplets collision will make smaller droplets coalesce into bigger ones. Since the interaction between droplets is not considered in the simulation, it is possible that the simulation underestimates the number of large droplets and overestimate the number of smaller droplets at the compressor discharge. This is one of the limitations of CFD simulation as an approach to studying the oil mist inside the compressor. The results show that CFD simulation can give a qualitative prediction of the droplet size distribution at the compressor outlet. From the comparison between droplet size distribution of the injection and that of the discharge, large droplets are more likely to hit the wall and splash on the wall film to generate more small droplets. Smaller droplets are less affected by gravity and inertia so that they can catch up with the vapor flow to escape the compressor and form the annular-mist flow in the discharge tube.
Fig. 18 – Comparison between video results and CFD results of droplet size distribution under different compressor speed
6.
Conclusions
This paper illustrates and analyzes the formation, distribution, and movement of oil droplets inside the discharge plenum of a compressor by visualization and simulation. Two main approaches are developed and presented: CFD simulation and visualization based on the high-speed video. Visualization of the reed valve shows the mechanism of oil droplet formation. Processed oil 11
ACCEPTED MANUSCRIPT
Acknowledgment
Abbreviations and variables Droplet diameter [μ The increment of droplet diameter [μm] Mass flow rate [g·s-1] Oil circulation ratio Pressure [kPa] Polyalkylene glycol Polyolester Polyvinylether Temperature [oC]
Subscripts cpro
Compressor refrigerant outlet
CR IP T
CE
PT
ED
d di ṁ OCR P PAG POE PVE T
M
The authors would like to acknowledge the companies that sponsoring the Air Conditioning and Refrigeration Center for their technical and financial support to this work. The authors would like to thank for Augusto Zimmermann who made the compressor prototype and built a solid foundation of the visualization work. The authors also thank Scott Wujek for developing the video processing program.
References
Quantify Behavior of Oil Droplets in a Scroll Compressor, in: International Compressor Engineering Conference. Paper 1788. Wujek, S., Hrnjak, P., 2011. Mist to annular flow development quantified by novel video analysis. ACRC Report TR285. Urbana. Xu, J., Hrnjak, P., 2017a. Quantification of flow and retention of oil in compressor discharge pipe. Int. J. Refrig. 80, 252–263. https://doi.org/10.1016/j.ijrefrig.2017.05.004 Xu, J., Hrnjak, P., 2017b. Oil flow at the scroll compressor discharge: Visualization and CFD simulation. IOP Conf. Ser. Mater. Sci. Eng. 232. https://doi.org/10.1088/1757-899X/232/1/012051 Xu, J., Hrnjak, P., 2016a. Refrigerant-Oil Flow at the Compressor Discharge, in: SAE Technical Paper 2016-01-0247. https://doi.org/10.4271/2016-01-0247 Xu, J., Hrnjak, P., 2016b. Visualization and Simulation of Oil Flow in a Scroll Compressor Plenum Corresponding author, in: International Compressor Engineering Conference. Paper 2507. Yokoyama, T., Shingu, K., Sekiya, S., Maeyama, H., Nishimura, N., 2012. CFD Analysis inside a CO2 Rotary Compressor Shell to Improve Oil Separation and Reduce the Shell Size, in: International Compressor Engineering Conference. Paper 2074. Zhu, J., Botticella, F., Elbel, S., 2018. Experimental Investigation and Theoretical Analysis of Oil Circulation Rates in Ejector Cooling Cycles. Energy. Zimmermann, A., Hrnjak, P., 2015. Oil flow at discharge valve in a scroll compressor. Proc. Inst. Mech. Eng. Part E J. Process Mech. Eng. 229, 104–113. https://doi.org/10.1177/0954408914559315 Zimmermann, A., Hrnjak, P., 2014a. Visualization of the Opening Process of a Discharge Reed Valve in the Presence of Oil, in: International Compressor Engineering Conference. Paper 2369. Zimmermann, A., Hrnjak, P., 2014b. Source Identification and In Situ Quantification of Oil-Refrigerant Mist Generation by Discharge Valve Opening Process, in: International Compressor Engineering Conference. Paper 2367.
AN US
mist videos provide the droplet size and velocity distributions under different working conditions. CFD simulation uses part of the information from the video as the input of discrete phase model. The simulation results are validated by comparing of oil drop size distribution at the compressor discharge tube obtained by CFD and that obtained by video processing. With these two approaches, a complete set of tools has been established to study similar multiphase problem inside the compressor. The phenomenon of oil flow is clearly captured by high-speed videos. Oil droplets are formed from the break-up of the oil film between the reed valve and the valve base. The oil droplets are then entrained by the compressed refrigerant vapor. Some of them deposit into the oil film on the inner wall while some of them escaped the compressor plenum and form mist-annular flow in the discharge pipe. In general, higher compressor speed introduces more oil droplets with slightly smaller size. Larger droplets tend to be retained in the plenum while smaller droplets are more likely to follow the refrigerant vapor flow to escape. This can explain why higher compressor speed brings higher OCR. The discharge plenum space of a vertical scroll compressor has the potential to separate large oil droplets. One cannot design the separator structure without knowing the characteristics of the oil mist. The illustration and analysis of oil droplet behavior contribute significantly to the design the geometry of compressor plenum with a goal of minimizing the oil discharge rate.
AC
Bhatia, K., Pitla, S., Khetarpal, V., 2002. Computational Analysis Of An Oil Separator, in: International Compressor Engineering Conference. Paper 606. DeAngelis, J., Hrnjak, P., 2005. Experimental study of system performance improvements in transcritical R744 systems with applications to bottle coolers. Air Conditioning and Refrigeration Center. College of Engineering. University of Illinois at UrbanaChampaign. Hay, K.J., Liu, Z.-C., Hanratty, T.J., 1998. A backlighted imaging technique for particle size measurements in two-phase flows. Exp. Fluids 25, 226–232. https://doi.org/10.1007/s003480050225 Li, H., Hrnjak, P., 2014. Lubricant effect on performance of R134a MAC microchannel evaporators. SAE Technical Paper. Toyama, T., Matsuura, H., Yoshida, Y., 2006. Visual Techniques to 12