“drop your thesis!” 2010 final...

Education

Drop Your Thesis! 2010 – Final Report 1

“Drop your Thesis!” 2010

Final Report

1. Executive Summary A study of bubbly jet impingement in microgravity conditions is presented in this project. The impinging jets configuration brings two jets owing along the same axis in opposite direction into collision. As a result, a narrow zone (the impingement zone) in which coalescence events between bubbles can be enhanced, is created. Experiments have been carried out at ZARM Drop Tower through the ESA Education Drop Your Thesis! programme. Filtered air bubbles in distilled water as a carrier liquid are generated using a capillary T-junction (d = 1 mm inner diameter), in which a regular slug-flow is created prior to injection. Within this method, the size and velocity of the bubbles can be controlled via the gas and liquid flow rates. Variations of the separation between jets have been also considered, ranging from s = 50 mm up to s = 100 mm. Results on the global structure of the impinging jets and the individual behavior of bubbles are presented. In particular, the velocity field, bubble size distribution, and the spatial distribution of coalescence events are presented. Good qualitative agreement between CFD simulations and experiments have been obtained. 2. Student Team Description Francesc Suñol graduated in Physics Sciences at Universitat de Barcelona (UB) and is currently realizing his PhD thesis in the field of two-phase flows in microgravity. He is specially interested in the dynamics of bubbles and drops in a low gravity environment. He participated in the 9th ESA's Student Parabolic Flight Campaign (SPFC) in September 2006, and he also flew in the 52nd ESA’s Parabolic Flight Campaign (PFC) in May 2010. He have participated in some International Congresses and published some papers in Scientific Journals. He was also involved in an experiment which studied the formation of water droplets in micogravity, using the I.N.T.A. Drop Tower facility in march 2010. In the current project, he has focused on the design of the mechanical parts and hydraulic circuits of the setup, and he has analyzed the results obtained in the ZARM Drop Tower facility. Oscar Maldonado obtained a degree in Aeronautical Engineering at Universitat Politècnica de Catalunya (UPC) in 2006. Since then, he have been working in some aeronautical industries while being in touch with the Microgravity Laboratory in our University, studying two-phase flows in microgravity conditions. In relation with microgravity experiments, he participated in the 9th ESA's SPFC in 2006 and in the ESA's REXUS Campaign in 2009. He was also involved in the I.N.T.A.’s drop tower experiment in March 2010 together with Francesc Suñol. In the present experiment, he has focused on the acquisition systems, control, and automatization of the experiment. Anna García obtained a 3-year bachelor degree in Aeronautic Engineering with specialization in Air Navigation Systems and Technologies in 2009. She is currently studying a Master in Aerospace Science and Technology (MAST). She participated in the Fly Your Thesis! Contest in 2009 with her final degree project “Multibubble sonoluminescence with liquid and bubble jets under microgravity”, getting the opportunity to test the experiment in a drop tower facility. In this project, she has focused on the control and acquisition devices.

Education


Ricard González is professor at the Applied Physics Department at UPC. He was Deputy Head of Resources and Deputy Head of Aeronautical studies at EPSC (UPC). He has been involved in the direction of microgravity experiments for students such as the 9th SPFC 2006, 47th PFC 2008, REXUS 2009, or the project held at I.N.T.A. Drop Tower in March 2010. He is currently the coordinator of the Master in Space Sciences and Technology (MAST), in which he teaches two subjects related to Microgravity Science.

Figure 1.1: Team picture. Antecedents of our team Four years ago our team started to get in touch with the research group in microgravity in our University, UPC (Barcelona, Spain). The group was interested in the dynamics of a single bubble jet in low gravity, and the characterization of a bubble injector [11, 12]. A few months later, we were proposed to participate in the study of the dynamics of two colliding bubble jets in a parabolic flight. Since then we have been working at the Microgravity Laboratory in UPC, designing and testing the experimental setup that was selected by ESA to participate in the 9th ESA Student Parabolic Flight Campaign. The results obtained in those flights justified the necessity to obtain a better quality of microgravity, leading to focus the experiment to another platform. A drop tower is the most appropriate facility, for the microgravity time requirements of our experiment and for the good quality of microgravity that this kind of platform offers. The technical requirements of the drop tower forces us to change the whole experimental system, so we have designed the new setup that fits the conditions imposed by the infrastructure. We have carried out numerous on ground tests, which can be compared with microgravity experiments. In relation with ESA Education projects related to microgravity, or other facilities such as I.N.T.A.’s Drop Tower, the members of the team have participated in the following campaigns:

9th ESA’s SPFC, September 2006 (Bordeaux, France).

47th ESA’s PFC, December 2007 (Bordeaux, France).

I.N.T.A. Drop Tower, July 2008 (Madrid, Spain).

5/6 REXUS Campaign, May 2009 (Kiruna, Sweden).

I.N.T.A. Drop Tower, March 2010 (Madrid, Spain).

Education


3. Background and Scientific Objectives 3.1 Scientific background The study of two-phase flows for microgravity applications has become an active field of research due to the advantages presented by multiphase systems over the single-phase ones. Small weight and more efficient thermal control are characteristics of two-phase systems which make them appropriate for several space technologies [1, 2]. Multiphase flows based technologies can be developed for applications in thermal management, propulsion systems, solid waste management, water recovery, and many other Environmental Control and Life Support Systems (ECLSS). Extensive on ground and in microgravity platforms studies are still needed to provide a wide knowledge of the behavior of multiphase flows which can facilitate the replacement of single-phase systems. The opposed-jet configuration have attracted special attention in the last decades due to its simple geometry and physical complexity. Opposed jets have been used extensively for studying turbulent properties of fluids [3-6] and the rich behavior of the flow concerning the structure of vortex interactions [7-9]. Many industrial applications have to deal with the improvement of fluid mixing efficiency, and some of them require a flexible control according to operation conditions. As investigated by Tsujimoto et al. [10], such flexibility in the mixing processes can be achieved by changing the impact angle between the colliding jets: reducing the impact angle increases significantly the mixing efficiency. In this sense, the opposed-jet configuration with changeable orientation becomes an attractive method for enhancing mixing systems at low cost while maintaining high-efficiency and direct control. On the other hand, knowledge of jet dynamics obtained after the interaction of two jets emerging from different angles, can give insight into the behavior of two-phase flows meeting at pipe junctions of space devices. A good understanding of bubble jets behavior in microgravity will enhance the development of space technologies based on two-phase systems. The presence of small gas bubble suspensions is a common problem to avoid in many applications. These bubbles can be created by phase change in heating devices, dynamic air entrainment or by cavitation effects. Since the dynamics of those bubbles should be controlled and well understood, studies on two-phase flows become imperative. However, the presence of bubbles can be advantageous for specific purposes like aeration control, fluid mixing or heat and mass transfer. When the density difference between the gas bubbles and the surrounding liquid is large, buoyancy plays an important role since it governs the dynamics of the mean flow. In space, where gravity can be neglected and no buoyant forces are present, many kind of gas-liquid flows are still poorly understood. The applications of two-phase systems in space are increasing dramatically, replacing single-phase systems due to its high efficiency in heat and mass transfer, mixing control and homogeneous oxygenization in bioreactors. Further studies have to be done in order to provide better knowledge of the operation of such systems in space. 3.1.1 Past research on this subject To our knowledge, the impingement of two liquid jets has been studied extensively in normal gravity, but this is not the case of bubble jets. There are only a few studies concerning single bubble jets in normal gravity [14-16] and we recently submitted two papers for publication using impinging bubble jets in 1g [13,17].

Education


Single bubble jets were also studied in microgravity conditions with an experiment realized at ZARM Drop Tower in 2003 by Carrera et al [11]. As far as we know, the impingement of bubble jets had never been studied in a microgravity environment, before this project. 3.1.2 Current research on this subject S. Arias et al. from Universitat Politècnica de Catalunya are characterizing the performance of the bubble injector that we used in the proposed experiment [12]. The work realized in this project lies in the research activities of the PhD. of Francesc Suñol. This project compares the results obtained in low gravity with those obtained in 1g [13,16,17]. Moreover, the results concerning the impinging jets with changeable distance between injectors, can be compared with the behavior of a single bubble jet. We expect that as the separation distance between impinging jets increases, the global behavior approaches a single bubble jet. 3.2.1 Relevance of microgravity Since in normal gravity the buoyancy force is the main responsible for governing the flow field when using bubbles, a microgravity environment is necessary in order to see the structure of two impinging bubble jets without buoyancy. We want to focus on the opening angle of the jets, the coalescence rates, bubble velocities, and we can only do that in low gravity, since bubbles rise with buoyancy and the whole structure becomes modified. The ZARM Drop Tower is the most appropriate platform for our experiment for many reasons:

Very good microgravity level.

Appropriate microgravity duration (we need 4 seconds of microgravity approximately).

Feasible assembly of the experiment in the platform (dimensions, mass).

Flexibilty: possibility to realize 2 or 3 drops per day, with the need to change some parameters between drops (as explained in the Experimental Procedure section).

All the other microgravity platforms present some inconveniences:

Parabolic Flights: Not enough good microgravity level for this experiment (as demonstrated in the 9th SPFC).

Neutral Buoyancy, Clinostats and Magnetic levitation: Not applicable in our case.

Sounding Rocket: Too small dimensions to fit our expeirment. We do not need the long microgravity time provided.

3.2 Objectives of the project The objectives of this project are divided in two main streamlines. On one hand, the main purpose of this project is to study the interactions between two bubble jets in microgravity. The designed experimental setup is based on an injection device that can control bubble generation frequency, size and velocity, injection angle and separation between jets [11,12]. We have obtained on ground results on both individual as well as collective behavior of bubbles. In order to get insight into the role played by gravity on bubble jet structure, the obtained results can be compared with zero-gravity experiments. In particular, coalescence of bubbles and jet structure in microgravity can be studied at different

Education


velocity regimes inside two overlapping jets with different separation between nozzles. In Figure 3.1, a snapshot (left) and an average of 200 frames (right) of impinging bubble jets in normal gravity is presented [13]. In the current project, since we had the opportunity to drop our experiment up to 5 times, we maintained a constant impact angle, φ=0º (i.e. frontal collision), and we focused on the study of the effects of different separation between nozzles and different flow rates on the final structure of the resulting jets. On the other hand, we are interested in the modelling of such phenomena, exploring and trying to go deeper into the theories that should be able to predict the behavior and the mechanisms that control the system dynamics. The obtained experimental data of the most relevant phenomena can be compared to numerical simulations performed at the Microgravity Laboratory at UPC (Barcelona, Spain), and at ICMCB (Bordeaux, France).

Figure 3.1: Left: Snapshot of impinging bubble jets in normal gravity. Right: Average of 200 frames of impinging bubble jets in normal gravity [13].

4. Experimental Set-Up The operation of each subsystem of the experiment (liquid line, gas line, data acquisition system, etc.) is described in section 4.1. In 4.2, each component of the experiment is presented. In section 4.3, the choice of the capsule and drop modes are explained and in section 4.4 the whole experimental setup is described.

4.1. System Description Liquid line: A liquid pump (Ismatec MCP-Z Standard) pumps distilled water from a liquid reservoir (Grifols Nutribag Enteral 500 ml, initially full of water) into the injectors (methacrylate T-junction capillary) creating a slug flow (along with the gas line). Then, some amount of water (depending on the liquid flow rate used) enters inside the test tank (which is initially full of water), and due to incompressibility of the liquid, some water is driven to the outlet of the test-tank which is connected to a residual tank (Grifols Nutribag Enteral 500 ml, initially empty). Gas line: A pressurized CO2 bottle (Genuine innovations, 16 g) is connected to a pressure controller (Bronkhorst EL-PRESS P-702CV), which is controlled via LabView, and governs when the gas line is open or closed. The pressure controller is connected to a gas flow meter (Bronkhorst EL-FLOW F-201CV) also controlled via LabView, and the gas is directed to the injectors, creating therefore a slugflow (along with the liquid line). Then, some amount of gas (depending on the gas flow rate used) enters inside the test tank

Education


(which is initially full of water) in form of small bubbles. In 1g, the gas goes to the outlet of the test-tank which is connected to a residual tank (Grifols Nutribag Enteral 500 ml, initially empty).

Data acquisition: On one hand, the main objective of the data acquisition is to obtain images of the opposed jets. In order to do it, a high-speed camera (Photron FastCam MC2, provided by ZARM) have been used at 1000 fps. There is the need of an illumination system for such framerate, and a matrix of 140 LEDs have been used, together with a diffusive sheet to homogenize the background illumination. On the other hand the liquid flow rate is directly controlled by the liquid pump via LabView, and the gas pressure and flow rate are measured (using LabView) using the previously mentioned pressure controller and gas flow meter. Printed boards and most of the electric connections are placed inside a rigid box, which we call it “Electric box”. Additional notes: The liquid reservoir and residual tanks, are flexible (though very resistant) transparent bags. This simplifies the setup, since they avoid any overpressure in the whole gas and liquid lines (which could occur if they were solid tanks). The liquid reservoir is a bag initially filled with water, and the residual tank is a bag initially empty. Both liquid reservoir and residual tanks are placed inside a water-tight box (which we call it “Tank box”). In 1g conditions, the bubbly jets are generated in the same way as in 0g, but as they enter the test tank, the bubbles go to the top of the test tank due to buoyancy. In order to have no air inside the test tank when microgravity conditions are reached, we have designed the test tank in a way that the roof of the test tank has non-zero slope, and the buoyant bubbles are directed to the residual tank.

Education


4.2. Experiment components

Test tank: Aluminum-made, with two methacrylate windows. Water-tightness is ensured by using two toroidal seal joints. Dimensions: 150x90x100 mm3 (support dimensions: 194x134x102 mm3). Weight: 1 Kg approximately.

Figure 4.1: Test tank. There are 4 holes to connect the injectors, but only two of them (in the middle of the walls) have been used in this campaign. The other two (near the bottom of the walls) are used to change the impact angle between injectors in a future, but in the current campaign they have been closed.

Education


Injectors: Methacrylate T-junction capillary, attached to the test tank by 5 screws. Dimensions: 120x35x40 mm3 (in the image below), although different lengths (120, 107.5 and 95 mm) have been used in order to change the separation between jets (100 mm, 75 mm and 50 mm). Weight: 75 g approximately.

Figure 4.2: Injectors.

Liquid pump: Ismatec MCP-Z Standard. The flow rate is controlled remotely using LabView, provided the “ON/OFF” button is positioned to “ON”. Dimensions: 220 x 155 x 260 mm. Weight: 6.4 kg.

Figure 4.3: Liquid pump.

Education

Drop Your Thesis! 2010 – Final Report 9 t 9

The datasheet of the pump can be downloaded from the Ismatec website, http://www.ismatec.com/images/pdf/manuals/MCP_Z_Standard.pdfThe datasheet of the pump can be downloaded from the Ismatec website, http://www.ismatec.com/images/pdf/manuals/MCP_Z_Standard.pdf Since the liquid pump needs and AC input voltage, we need a DC (provided by the batteries) to AC converter, which is the next item. DC to AC power inverter (from 24 V-DC to 200 V-AC): Dimensions: 100x70x70mm3 approximately. Weigth: 350 g approximately.

Figure 4.4: DC/AC Power inverter. CO2 pressurized bottle: Tefal 16 g of carbon dioxide at a pressure of 50 bar approximately. However, we reduced the pressure to 5 bars (using a manual manorreductor) since the tubbing supports up to 10 bar.

Figure 4.5: Gas bottle. Pressure Controller: Bronkhorst EL-PRESS P-702CV. Controlled via LabView.

Figure 4.6: Pressure controller.

Brochure: http://www.bronkhorst.com/files/downloads/brochures/folder-el-press.pdf Datasheet: http://www.bronkhorst.com/files/downloads/manuals_english/917001manual_mass_flow_pressure_meters_controllers.pdf

http://www.ismatec.com/images/pdf/manuals/MCP_Z_Standard.pdf

http://www.ismatec.com/images/pdf/manuals/MCP_Z_Standard.pdf

http://www.bronkhorst.com/files/downloads/brochures/folder-el-press.pdf

http://www.bronkhorst.com/files/downloads/brochures/folder-el-press.pdf

http://www.bronkhorst.com/files/downloads/manuals_english/917001manual_mass_flow_pressure_meters_controllers.pdf


Education


Gas flow meter: Bronkhorst EL-FLOW F-201CV. Controlled via LabView.

Figure 4.7: Gas flow meter.

Brochure: http://www.bronkhorst.com/files/downloads/brochures/folder-el-flow.pdf Datasheet: http://www.bronkhorst.com/files/downloads/manuals_english/917001manual_mass_flow_pressure_meters_controllers.pdf Residual tank and liquid reservoir: Grifols Nutribag Enteral 500 ml

Figure 4.8: Residual tank and liquid reservoir.

http://www.bronkhorst.com/files/downloads/brochures/folder-el-flow.pdf



Education


LED matrix: LED matrix: Matrix of 140 ultrabright LEDs (10,000 mcd each), 10 rows x 14 columns. Matrix of 140 ultrabright LEDs (10,000 mcd each), 10 rows x 14 columns. Dimensions: 150x90x20 mm3 (including support). Dimensions: 150x90x20 mm3 (including support). Wheight: 300 g approximately (including support). Wheight: 300 g approximately (including support). Power Consumption: 24 V DC, 350 mA. Power Consumption: 24 V DC, 350 mA.

Figure 4.9: LED Matrix. Electric box and Tank box: Plastic box. Dimensions: 250x150x90 mm3 approximately. Weigth: 150 g approximately.

Figure 4.10: Electric box and Tank box.

Education


High-speed camera: Photron FastCam MC2 (provided by ZARM).

Figure 4.11: High-speed camera.

Data sheet: http://www.photron.com/datasheet.php?ds=Fastcam_MC2.pdf&id=8

4.3. Choice of the Drop Mode and Capsule The time constraints in our experiment allowed us to use the Standard drop mode, since we needed at least 4 seconds of microgravity time. The catapult mode was not appropriate for our experiment since the initial hypergravity phase (when launching the capsule) can generate undesirable inertial effects (flows inside the test tank), which can affect the dynamics of the bubbles and therefore the final results. However, the standard drop mode fitted well for us since we had an initial steady state (1g) before the transition to microgravity. Our experimental setup was assembled into the long version of the capsule, and was in agreement with all the technical and safety requirements stated at the ZARM’s Drop Tower User Manual.

4.4. Capsule Schematic The capsule is divided in three floors. The weight and dimensions of the first and second floors are presented in Appendix A.2.1.

0th floor: Batteries, National Instruments’ PXI, … (Not included in the schematics).

1st floor: liquid pump, pressure controller, electric and tank boxes, DC to AC converter, gas bottle.

2nd floor: Illumination system, test tank, high-speed camera, gas flow meter, valves.

The schematics of the first floor of the capsule is presented in Figure 4.12. The second floor is presented in Figure 4.13, and the whole capsule (excluding the 0th floor) is presented in Figure 4.14.

http://www.photron.com/datasheet.php?ds=Fastcam_MC2.pdf&id=8

Education


Figure 4.12: Schematics of the 1st floor of the capsule.

Education


Figure 4.13: Schematics of the 2nd floor of the capsule. Figure 4.13: Schematics of the 2nd floor of the capsule.

Figure 4.14: Schematics of the 1st and 2nd floors of the capsule.

4.5. Capsule Pictures In Section 4.4, the schematics of each individual platform have been presented. However, due to unexpected last-time issues, the final setup design has been slightly modified. 4 valves to regulate the flow with high precision have been added to the 2nd floor of the capsule. In addition, the gas flow meter is now in the 2nd floor (note that in the 2nd floor, there is some empty space in front of the test tank where the high-speed camera has to be placed). However, the schematics presented in Section 4.4 are still a good description of the experimental setup and we have decided to keep these sketches in the document, since the final changes we had made are very small. In the present section, pictures of the experimental setup are presented (not numbered Figures):

1st floor, top view: 2nd floor, top view:

Education


1st floor, front view: 2nd floor, front view:

1st floor, rear view: 2nd floor, rear view:

1st floor, right view: 2nd floor, right view:

Education


1st floor, left view: 2nd floor, left view: 1st floor, left view: 2nd floor, left view:

1st and 2nd floors, top view: 1st and 2nd floors, top view:

Education


Figure 4.15: Pictures of the rack and the capsule at ZARM Drop Tower.

5. Experimental Procedure In the current campaign, we had the possibility to drop our experiment 5 times. The effects of two parameters have been studied: The effect of the jet flow rate (controlled by the gas and liquid flow rates, Qg and Ql respectively) and the effect of separation between jets (s). We planned to use three different values of the separation between jets (s = 100 mm, s = 75 mm and s = 50 mm) and different values of the flow rates. Hence, we planned to distribute the drops in the following manner:

Drop

s (mm) Valve aperture (p = 5 bars)

Ql (ml/min)

1 75 5% 10 2 5 2.5% 30 3 10 3% 60 4 10 3% 100 5 10 4.5% 70

Between drops 1-2, and 2-3, we changed the separation between jets, and we checked that the pressure of the gas bottle was at 5 bars, and that the liquid reservoir was not empty and the residual tank was not completely full.

Education


In order to change the separation between jets (s) we needed to access the rack (outside the capsule) and remove some screws, so this took some minutes. The change in flow rates is automatically controlled via LabView. Between drops 3-4-5, we checked that the pressure of the gas bottle was at 5 bars, and that the liquid reservoir was not empty and the residual tank was not completely full. The liquid pump button must be at “ON” position and the same for the AC/DC Converter (no matter if the alimentation through the batteries is activated or not), in order to control the liquid flow rate via LabView. The gas bottle must be at the right pressure (5 bars) before connecting it to the gas line. We can control the pressure inside the gas bottle using an adapted manometer. The high-speed camera must be at optimum operation parameters: 1000 fps at the maximum resolution available (512x512 pixels), with the zoom properly adapted to obtain the highest quality images possible. The experimental procedure at each drop is explained in the next paragraphs. Once that we had the setup properly assembled in the rack, and the rack is placed inside the capsule, we needed to do the following operations at each drop:

Turn on the alimentation from the batteries. Activate the Illumination system, liquid pump, controllers and camera. Run the LabView program:

o Activate the liquid pump at the proper flow rate. o Activate the gas flow rate sensor. o Open the pressure controller valve in order to let the gas circulate through the

gas line. Choose a proper aperture value (between 2.5% and 5%). Wait a few seconds in order to stabilize the system. Trigger the high-speed camera to start recording. Drop the capsule. Stop the LabView program. Turn off the alimentation from the batteries. Recover the recorded high-speed video and check the results of all the acquisition

devices.

6. Drop Tower Campaign

6.1. Before the drops

During the two-week preparation, the BubJet Team had a busy time integrating the experimental setup to the drop capsule. Once the experiment was completely arranged, numerous on ground tests were carried out to ensure a proper operation of the whole experiment. In addition, some changes to the setup were implemented, as requested by the ZARM staff: we changed the gas bottle to a FESTO’s 2 L pressurized bottle (at 5 bars). This ensured a constant gas flow rate during all the microgravity time. Moreover, mechanical supports to protect the injectors from breaking, due to the high gravity levels reached at the deceleration zone, were fixed. An electrovalve was also placed at the exit of the gas tank to ensure that no gas was flowing through the tubes when not needed.

6.2. Between the drops

We had two major problems during the Campaign. Only one problem could be solved, the solution to the other one will be implemented in future experiments.

Education


The problem that we could solve was the following: between drops 1 and 2, an injector broke. Fortunately, we had 4 injectors available, and we could place a new injector in time, despite the difficulty to remove the injectors from the test tank, place the new injector, and screw the test tank again to the rack.

Figure 6.1: Picture of the broken injector, disassembled from the test tank. The problem that we were not able to solve, is explained in the following paragraphs: We designed a system to remove the residual bubbles from the test tank, to ensure that in microgravity only the bubbles of the jet were present. This system consists in direct the gas into the residual tank in 1g, and we did it by designing the top of the test tank with an angle different from 0º, then the bubbles rising reached the top and they were redirected to the residual tank. However, due to the fact that negative gravity values are reached when dropping the capsule (up to g = 0.7g0, see Figure 6.3), some of the gas that is present in the top of the test tank is directed to the bulk of the test tank, not staying at the top anymore. If a bubble reaches the jet zone (as the bubble on the left), it is pushed by the liquid flow field to the central zone. Once the bubble reaches the central zone of the test tank, it is difficult to escape from this region due to the higher pressure generated by the opposed jets. In Figure 6.2, the presence of big bubbles in the jet zone is presented. The presence of big bubbles coming from the top of the tank occurred in all the 5 drops we had available at the ZARM Drop Tower. These bubbles affect the flow field significantly, and the results reported herein are taken in a way that tries to minimize those perturbations: the measurement of the jet shape or bubble velocities was carried out when the big bubbles were located far from the jet region. In addition, in the measurements of the bubble sizes, the diameters of these big bubbles (and the smaller bubbles surrounding the big ones in some cases) have been neglected.

Figure 6.2: Snapshots of impinging bubbly jets in microgravity at different times. (a) t = 325 ms, (b) t = 700 ms, (c) t = 1100 ms and (d) t = 1200 ms after capsule release.

Education


7. Scientific Results

7.1. Collected Data

During a whole drop, a large amount of data was collected automatically. We designed a LabView application with the aim to save the data in plain text files. As part of our expeirment, we collected:

Liquid flow rate

Gas flow rate

Gas pressure

High-speed movie The first three points were saved as ASCII files. The high-speed movie was saved as an uncompressed AVI file. In addition, there were various sensors placed inside the capsule by the ZARM staff, that provided measurements such as:

Acceleration level (for the three axis, x,y and z)

Temperature inside the capsule

Pressure inside the capsule We didn’t measure the gravity level, since ZARM provided us with those measurements. As an example, in Figure 6.3(a), the normalized gravity level as a function of time for a whole drop is presented. After the capsule is dropped, the gravity level undergoes high frequency damped oscillations, reaching negative values of g up to g = 0.7g0. After 0.7 s approximately, these oscillations can be neglected and the gravity level stabilizes with some significant g-jitter. Figure 6.3(b), is the same as Figure 6.3(a) with a different scaling factor. The reason for including Figure 6.3(b) is the fact that during the capsule deceleration, high gravity levels are reached, up to g = 40g0 approximately, and using this scale factor the decelerating gravity level can be appreciated in more detail.

Figure 6.3: Normalized gravity level as a function of time. (a) Oscillation of gravity level (with

negative values of gravity) after capsule release. (b) Up to g/g0 ≈ 40 when the capsule is decelerating.

Education


7.2. Data Processing and Analysis Once the data is collected, the processing and subsequent analysis has to be carried out. All the data except the high-speed movies were saved as plain text files, which facilitates the manipulation of the data. However, the high-speed movies are much more difficult to process: We saved the movies as uncompressed AVI files, so we preserved all the quality. Nevertheless, each movie is very large in size. For this reason, an appropriate codec (Cinepac codec by Radius) was chosen and the movies were also saved as compressed files. Once we had those files, an extraction of each frame from the movie, together with a crop in the region of interest was made. In addition, a spatial calibration of the images was carried out. For most of the measurements realized on the movies, an adjustment of the image equalization (brightness, contrast and gamma level) and the application of various filters were carried out. The first thing to do was to syncronize the plain text data files with the high-speed movie. This point requires two steps: adjustment of the correct time scale, and identification of the capsule release instant (t = 0). Both for the text files, and for the movie.

Movies: we recorded the videos at 1000 fps, each frame of the movie was separated 1ms from the next frame. This gives us directly the time scale. In order to identify the capsule release instant, we configured the camera trigger to start recording 2.75 seconds before the capsule release, and record 8 seconds continuously (we were able to record up to 8000 frames at 1000 fps with an image resolution of 512x512 pixels). With this, we had: 2.75 s before the drop, 4.75 s during drop, and 0.5 s after drop, approximately. Then, the frame 2750 should correspond to the t = 0 instant. However, due to uncontrollable delays on the trigger signal, a skip of various frames could occur, and this could be corrected by watching directly the movies recorded.

Text files: the timescale from the text files is extracted directly from the writing

interval. The offset to identify the t = 0 point was done by analyzing the vertical (z direction) gravity level and writing the time in which an abrupt drop of the g level was measured.

Once the data files and the movies were syncronized, an analysis of the movies were carried out, in order to obtain the measurements presented in the next Section.

7.3. Results

According to Schlichting [18], the momentum flux J can be considered as the main parameter that characterizes the structure of a single-phase jet,

where ρ is the density of the fluid, vl is the velocity in the direction of injection, and cylindrical coordinates (r,θ,l) are used. Taking into account the effect of both gas and liquid phases, the momentum flux can be computed as

where the liquid velocity is approximated by v(L) ≈ QL /A, and the gas velocity is approximated by v(G) ≈ QG /A, being A the area of the capillary. QL and QG correspond to the liquid and gas flow rates, respectively. Hence,

Education


The momentum flux J is indicative of the jet strength, and most of the results presented in this work are based on this parameter.

7.3.1. Jet Structure

Assuming that in microgravity the bubbles can be considered as passive tracers of the carrier jet, the position of the bubbles inside the jet give valuable information of the global jet structure. By examining visually the recorded movies, one can note that the paths of the bubbles were confined at certain regions (the majority of the bubble positions lie inside a nearly conical shape with its edge located at the nozzle exit), being the velocity of the bubbles higher at the jet centerline. By averaging a certain number of frames of every movie, a delimitation of the conical jet structure was pointed out. The vast majority of the bubble paths are confined inside a region which describe the jet conical shape. Only a small number of bubbles are located outside this shape. With the aim to characterize this shape (which can not be described by a cone opening angle –i.e. a straight line– due to the presence of the opposed jet), a parameter ζ is defined as the position of the delimiting zone of the jet shape, measured directly from the averaged images. A plot of ζ as a function of the injection axis position is presented in Figure 6.4. In this particular case, the separation between jets is s = 100 mm, and the momentum flux is J = 86 gcm/s2. The values of ζ have been obtained by averaging the first 400 ms (squares), 650 ms (circles) and 1760 ms (triangles) respectively and drawing visually the delimitation zone. In the case of t = 400 ms, the bubbles can be located inside two regions with a nearly conical shape emerging from the nozzles. The bubbles coming from the opposed jets are still in its way to the central zone (which correspond to 40 < l < 60 mm in this case approximately). At t = 650 ms, the bubbles coming from one jet are interacting with bubbles coming from the opposed jet in the central zone, reducing its velocity in the jet centerline direction while increasing its velocity in the direction perpendicular to the injection direction. Finally, at t = 1760 ms, the interaction between the opposed jets is highly significant in the central zone, fact that is reflected for the bubble paths: near the central zone, the velocity of the bubbles in the direction of injection becomes reduced. However, the velocity in the direction perpendicular to injection increases. This results in a cross-like shape of the interacting jets, demonstrating that the assumption of the straight conical shape (assumption taken with a single jet) is no longer valid, being only applicable in the region close to the nozzle.

Figure 6.4: Delimitation zone as a function of the jet centerline distance.

Education


7.3.2. Bubble velocities Following the procedure used by Carrera et al. [11] to study the velocity of a bubbly jet in microgravity, we consider the Schlichting solution [18] for a single phase turbulent jet, where the x component of the velocity reads

where νT is the turbulent kinematic viscosity, and

To avoid the divergence at x = 0, a parameter x0 is introduced in order to take into account the finite size of the nozzle. In the jet centerline, y = 0, the modified equation becomes:

In order to determine experimentally the bubble velocity distribution at the jet centerline, velocities of representative bubbles have been measured at different momentum fluxes and separation distances between opposed jets. In Figure 6.5(a), the component of the velocity in the direction of injection is plotted as a function of the distance from the nozzle, for a fixed separation between nozzles of s = 10 cm. It can be observed that for x > 30 mm, the measured velocities are lower than the theoretical prediction (fitted solid lines). This prediction is for a single jet, and the presence of the opposed clearly perturbs the velocity field at the central zone (for s = 10 cm, the central zone can be considered 30 < x < 70 cm approximately). Contrarily to a single jet, the velocity field of opposed jets have a cross-like shape, with a high reduction of vx in the central zone, and with an increment of vy in the collision plane (at x = s/2). The bubbles enter the collision zone with values of the velocity component in direction of injection (vx) much higher than the perpendicular component (vy). After the collision with the incoming jet, a circular “sheet” of bubbles grows radially from the center of the collision zone, being in this region vy much higher than vx. The velocities for different separations between jets are plotted in Figure 6.5(b). In this figure it can be also observed the same behavior: measured velocities are lower in the central zone than the theoretical prediction. Thus we conclude that the Schlichting prediction can be only applied near the nozzles of the opposed jets, due to the perturbation of the velocity field created by the incoming jet.

Education


Figure 6.5: Bubble velocities at visual jet centerline. (a) s = 10 cm; (b) J = 127 gcm/s2 for s = 10 cm, J = 3 gcm/s2 for s = 7.5 cm and J = 32 gcm/s2 for s = 5 cm. Fitted solid lines

correspond to the Schlichting solution. The values of the fitting parameters from Figures 4.15(a) and 4.15(b) are presented in Table 6.1.

Table 6.1: Values of the fitting parameters for opposed bubbly jets in microgravity.

According to Schlichting [18], the structure of the turbulent jet solution is independent of the momentum flux J and consequently all the measurements of Figures 6.5(a) and 6.5(b) should collapse on a single curve. In Figure 6.6, a plot of this collapse is presented. Good fit of the Schlichting solution to the measured velocities is obtained near the nozzle. Far from the nozzle, the measured velocities are lower than the prediction due to the presence of the opposed jet. Hence, we conclude that the Schlichting solution for a single jet can be only applied near the nozzles, where the perturbation of the incoming jet can still be neglected.

Education


Figure 6.6: Collapse of bubble velocity measurements in microgravity, for different momentum fluxes and separation distances.

7.3.3. Bubble sizes and coalescence events Unlike with bubbly jets in normal gravity, the bubbly jets in microgravity evolve continuously with time and no steady states are reached. In normal gravity, all the bubbles that the injector generates are quickly directed to the top of the tank due to buoyancy forces. However, in microgravity, an injected bubble could stay at the same position inside the tank for an indefinite period of time, being able to collide and coalesce with another bubble. Hence, the number of bubbles of bubbly jets in microgravity grows with time, increasing therefore the number of coalescence events. For this reason, the bubble size distribution of the bubbly jets in microgravity also evolves in time, and can not be considered steady anymore. The sizes of the bubbles for different momentum fluxes have been measured from a snapshot of the movies at t = 300 ms after capsule release, and the obtained bubble size distribution is presented in Figure 6.7(a). The number of measurements is of the order of 100 bubbles for each snapshot. The presence of big bubbles coming from the top of the test tank (see Figure 6.2) have been neglected in the measurement of the bubble size distribution. Since no bubble breakup have been observed, the presence of bubbles with diameters smaller than capillary diameter dB < dC is due to the injector performance, or to the depth of field effect. On the other hand, the presence of bubbles with diameters dB > dC is due to the injector performance and the coalescence events. First, it is important to point out that the injector generates bubbles with a certain bubble size distribution, being some bubbles slightly smaller than the capillary diameter and some bubbles slightly larger. Secondly, by observing the recorded movies we deduce that the width of the bubble size distribution from the injector is small, so the presence of bubbles with diameters dB/dC > 1.5, is due uniquely to coalescence events. The mean diameter is higher for the lower value of the momentum flux, due to the fact that higher values of the momentum flux are obtained by high values of the liquid flow rate. Since the bubble size distribution of bubbly jets in microgravity evolves in time, bubble diameters have been measured for a fixed value of momentum flux and separation distance at three different times. In Figure 6.7(b), the obtained bubble size distribution is presented. The number of measurements is approximately 100 bubbles for t = 300 ms and 200 for t = 1100 ms. It can be observed that the bubble mean diameter grows in time, since a lot of coalescence events were observed with no bubble breakups. This growth in the bubble diameter is due uniquely to the growth of the tail of the distribution, not in an offset of the whole distribution. This offset could be created by a change in the injector performance, which was observed to be steady, so the evolution in bubbles sizes is due exclusively to coalescence events. In both Figures 6.7(a) and 6.7(b), the tail of the bubble size distribution is a direct measure of the coalescences, since no bubbles with dB/dC > 1.5 were observed coming out from the nozzle.

Education


Figure 6.7: Distribution of bubble diameters for: (a) different momentum fluxes at a fixed time;

(b) different times at a fixed momentum flux. With the aim to know in which regions of the opposed jet configuration the coalescence events take place, a manual measurement of the (x, y) coordinates of the coalescences occurring in Δt = 200 ms have been carried out. The position of coalescence events have been measured manually form the movies recorded by skipping frame by frame and annotating the position when a coalescence was observed. The obtained results for s = 10 cm and s = 5 cm are plotted in Figures 6.8(a) and 6.8(b) respectively. From these plots, we observe that the coalescences can occur in the whole width of the jet, but most of the coalescence events tanking place near the nozzles or in the central zone.

Figure 6.8: Position of coalescence events in Δt = 200 ms for (a) s = 10 cm and (b) s = 5 cm.

It should be pointed out that the measurements of Figures 6.8(a) and 6.8(b) have been carried out considering a coalescence a collision of two bubbles creating a single larger daughter bubble, without taking into account the number of coalescences that the colliding bubbles may have suffered before. As an illustrative example, the coalescence of three bubbles into a larger daughter bubble is shown in Figure 6.9. In this process, we have considered two coalescence events (marked with double-arrows) since they occur at different times. No simultaneous coalescence of three or more bubbles into a single daughter bubble have been observed from the movies recorded.

Education


Figure 6.9: Sequence of snapshots showing the coalescence process of three bubbles into a single daughter bubble. The two coalescence events are marked with double-arrows.

7.3.4. Numerical simulations Some numerical simulations have been carried out in order to compare the experimental results to the available models. At this stage, the evolution of the velocity field and the distribution of the liquid and gas phases have been computed. The equation governing the dynamics can be written as

Where the following dimensionless numbers are introduced,

The turbulence model implemented in the simulations is the Chen and Kim modified k-ε turbulence modelization. Within this model, the turbulent viscosity reads

where Cμ = 0.09 is a constant of the model, k is the turbulent kinetic energy and ε is the turbulent dissipation. The equation for the turbulent kinetic energy k can be written as

where σk = 0.75 is a constant of the model and Sij is the rate-of-strain tensor

Education


The equation for the turbulent dissipation ε can be written as

where σε = 1.15, C1ε = 1.15 and C2ε = 1.9 are constants of the model. The last term on the right-hand side of the equation for the turbulent dissipation contains the production timescale and marks the difference between the Chen and Kim model and the standard k-ε model.

Figure 6.10: Time evolution of gas volume fraction for h/d = 40, Re = 1000, for the zero gravity case.

In Figure 6.10, the evolution of the gas phase in the initial instants of the microgravity conditions is plotted at different times (dimensionless time). The spatial distribution of the gas

Education


phase is in accordance with the experimental measurements plotted in Figure 6.4, and good qualitative agreement is obtained between the experiments and the numerical simulations. Nevertheless, additional simulations have to be done and compared to the experiments for a good enclosure of the problem. We are currently working on these simulations.

8. Conclusions The main objective of this work was to gain more fundamental knowledge on two-phase flow dynamics, by obtaining experimental data in microgravity conditions of opposed bubbly jets. In this research, several experiments and computational simulations have been carried out in an attempt to clarify some of the fundamental aspects of the complex flow created by the impinging two-phase jets. In particular, the study of the structure of the impinging jets, the velocity field and behavior of bubbles in the collision region of two impinging jets have been investigated. The most relevant contributions are summarized below:

Bubble velocity in the jet centerline within the opposed-jet arrangement is obtained experimentally and compared with the theoretical solution of a single jet. Near the injectors, the structure of the velocity field is similar to that of a single jet. However, in the collision zone, the velocity field is highly perturbed by the opposing jet. Comparing to the velocity field of a single jet, a decrease in the velocity in the direction of injection is observed in the central zone of the opposed-jet configuration. The velocity in the vertical direction is highly increased in the central zone, due to the formation of a growing disc with the stagnation point in its center.

An analysis on the bubble size distribution at different flow rates indicate that the

bubble mean size decrease as the momentum flux is increased.

Coalescence events occur mainly near the nozzle and in the collision zone. A higher number of coalescences have been observed in microgravity than in the normal gravity case. A possible explanation for this is that in microgravity the probability of collision between two (or more) bubbles is much higher than in normal gravity, since in reduced gravity the bubbles follow passively the liquid flow, with no effects of buoyancy forces. As a result, a numerous bubbles remain quiescent at certain zones of the impinging jets, increasing the probability of collision.

Numerical simulations revealed that the structure of bubbly jets adopts the form a

cross-like shape. Qualitative agreement between numerical simulations of opposed bubbly jets and the experimental results have been obtained.

Education


References

1. K. Gabriel. Microgravity two-phase flow and heat transfer. Springer (2007). 2. McQuillen, J., Colin, C. and Fabre, J. Ground-based gas-liquid flow research in

microgravity conditions: state of knowledge, Space Forum, Vol. 3, 165-203 (1998).

3. Eckestein, J.-Y., Chen, J., Chou, C.-P., Janicka, J. Modeling of turbulent mixing in opposed jet configuration: one-dimensional Monte Carlo probability density function simulation. Proceedings of the Combustion Institute, Vol. 28, 141-148 (2000).

4. Chou, C.-P., Chen, J.-Y., Janicka, J., Mastorakos, E. Modeling of turbulent opposed-

jet mixing flows with k-e model and second order closure. International Journal of Heat and Mass Transfer 47, 1023-1035 (2004).

5. Eren, H. Numerical study of unsteady interacting compressible jet flows using finite

difference method for small Reynolds numbers. Applied Mathematics and Computation 172, 876-891 (2006).

6. Weifeng, L., Zhigang, S., Haifeng, L., Fuchen, W., Zunhong, Y. Experimental and

numerical study on stagnation point offset of turbulent opposed jets. Chemical Engineering Journal Vol. 138, Issues 1-3, 283-294 (2008).

7. Voropayev, S. I., Afanasyev, Y. D. Two-dimensional vortex-dipole interactions in a

stratified fluid. J. Fluid. Mech. Vol. 236, 665-689 (1992).

8. Afanasyev, Y. D., Voropayev, S. I., Potylitsin, P. G., Filippov, I. A. Interaction of vortex dipoles: The theory and laboratory experiment. Atmospheric and Oceanic Physics, Vol. 30, No. 5, 665-671 (1995).

9. Voropayev, S. I., Afanasyev, Y. D., Korabel, V. N., Filippov, I. A. On the frontal

collision of two round jets in water. Phys. Fluids Vol. 15, No. 11, 3429-3433 (2003).

10. Tsujimoto, K., Shakouchi, T., Sasazaki, S., Toshitake, A. Direct numerical simulation of jet mixing control using combined jets. JSME International Journal, Series B, Vol. 49, No. 4, 966-973 (2006).

11. Carrera, J., Ruiz, X., Ramírez-Piscina, L., Casademunt, J., Dreyer, M.. Generation of

a monodisperse microbubble jet in microgravity. AIAA Journal Vol. 46, No. 8 2010-2019 (2008).

12. Arias, S., Ruiz, X., Casademunt, J., Ramírez-Piscina, L.,and González-Cinca, R.

Experimental study of a microchannel bubble injector for microgravity applications. Microgravity Science and Technology 21:107-111(2007).

13. Suñol, F., and González-Cinca, R.. Opposed bubbly jets at different impact angles: jet

structure and bubble properties. Accepted for publication in International Journal of Multiphase Flow (2009).

14. Lima Neto, I. E., Zhu, D. Z., Rajaratnam, N. Bubbly jets in stagnant water.

International Journal of Multiphase Flow 34, 1130-1141 (2008).

15. Lima Neto, I. E., Zhu, D. Z., Rajaratnam. Horizontal injection of gas-liquid mixtures in a water tank. Journal of Hydraulic Engineering Vol. 134, No. 12, 1722-1731 (2008).

Education


16. Suñol, F., Maldonado, O., Pino, R., González-Cinca, R. Design o fan experiment for the study of bubble jet interactions in microgravity. Microgravity Science and Technology 21:95-99 (2009).

17. Suñol, F., and González-Cinca, R. Bubbly jet impingement in different liquids.

Submitted to Microgravity Sceince and Technology (2009).

18. Schlichting, H. Boundary-Layer Theory. McGraw-Hill Classic TextBook Reissue, Seventh Edition, 1979.

Education


A. Appendices

A.1. Technical Description An experimental setup with very similar components and arrangements in the capsule, was tested in the I.N.T.A.’s Drop Tower (Madrid, Spain), which provides 2.2 s of microgravity. All the equipment operated as expected during the whole drop procedure and resisted the high acceleration peaks without breaking. This was a first indicative of safety and reliability of the current experimental setup.

A.1.1. Components Description

All the weights presented in the following table are approximate.

Component Weight (kg)

Test tank (filled with water) 2.5 Illumination system 0.4

Injectors 0.2 High-speed camera 1 AC/DC Converter 0.4

Liquid pump 6.4 Gas bottle 0.2

Pressure controller 1 Gas flow meter 1

Electric box 0.5 Tank box 0.7

Tubbing, wires, screws and fixations

4

2 Platforms 30.4 Total 48.7

A.1.2. Products Used for the Experiment

Product Quantity (kg)

Distilled water 1.5 Carbon dioxide 0.016

Distilled water: the test tank is filled with 1 kg of distilled water approximately, the remaining 0.5 L are kept in the liquid reservoir (in a flexible bag which is placed inside the tank box). Carbon dioxide: is stored in a pressurized bottle.

Education


After bubble creation, a mixture of distilled water and carbon dioxide is directed to the residual tank, which is placed inside the tank box.

A.1.3. Electrical System

Figure A.1: Schematics of the electrical system.

A.1.4. Electronic System

Figure A.2: Schematics of the electronic system.

Education


“drop your thesis!” 2010 final...

Documents