Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Gas ejector experiment

  1. Oct 19, 2006 #1
    Hello

    Im currently working on a CFD project looking a gas ejector design, however I need a comparison of the results found from the CFD software. For this Ive managed to get hold of a small gas ejector and Im now trying to design a suitable experiment to let me measure the mass flow at the entrainment section and the exit of the ejector. My first thoughts was to connect the entrainment section up to a small manometer and calculate the difference in pressure. However my tutor tells me that this would not provide accurate results as its a "dead pressure"?

    Im really scratching my head now of how to go about this and would really appreciate any ideas or thoughts on this subject that anyone may have

    Kind regards

    James
     
  2. jcsd
  3. Oct 20, 2006 #2

    FredGarvin

    User Avatar
    Science Advisor

    Can you provide a cross section of the test article?
     
  4. Oct 22, 2006 #3
    Gas ejector

    Hi

    I have attached a JPEG of the ejector - the dimensions are rough (not a million miles away though!!) apart from the orrifice, thats bang on as I have been able to measure that with a pin gauge. I can't get the exact dimensions of the internal geometry as I can't section the component until I have tested it - catch 22!!

    The ejector is being fed from a 5 bar gas canister, although for the test I can hook it up to a compressor which will allow me to control the mass or volumetric flow rate more accurately.

    Th ejector design is from Green-clean, heres a link to see the component and it's application.

    http://www.fotosense.co.uk/Green_Clean_CCD_Sensor_Cleaning_System_Kit.asp?productID=2754

    I hope this helps - still scratching my head!

    James
     

    Attached Files:

  5. Oct 22, 2006 #4

    FredGarvin

    User Avatar
    Science Advisor

    I'm not quite sure what your tutor means by a dead pressure. The entrainment section should have a steady flow to it once your compressor's flow has been established. The first two methods that come to mind to do what you need would be an orifice and a hot wire anemometer. Do you have a limitation on the size of the entrained air's entrance? Every time I have ever used an ejector, it was, essentially, an infinite source. If that were your case, I see no reason why you couldn't set up a flow measurement a good distance upstream of the ejector itself.
     
  6. Oct 22, 2006 #5

    Danger

    User Avatar
    Gold Member

    Sorry for 'dumming-down' the thread, but... what's a gas ejector? :confused:
     
  7. Oct 22, 2006 #6

    FredGarvin

    User Avatar
    Science Advisor

    Think of an air brush set up and how the paint gets out of the can attached to it. We occasionally use them to simulate altitude flowing conditions for certain tests.
     
  8. Oct 22, 2006 #7

    Danger

    User Avatar
    Gold Member

    Oh, gotcha... venturi effect. Thanks.
     
  9. Oct 22, 2006 #8
    i think he meant if the the pressure was to be measured using a simple manometer to see the difference in head, the flow would be impeded through the entrainment section. If the ejector doesnt entrain then its not an ejector any more as they are dynamic devices. Atleast I think this is what he meant by a "dead pressure".

    I did think about using an orifice to calculate the flow, but they need machining up and then calibrating? I only have another two weeks to test the ejector If im to stay on schedule for completing the project.

    One idea I did have was to conect tubing up to the ejector then place the end of the tubing into a beaker with water in it upside down in a container of water. As the the air is blown in to the beaker the bubbles will displace the water, hopefuly from there i will be able find a volumetric flow rate? Ovbiously the entrainment section will also be inserted into a test tube except this time it will be pulling the test tube down instead of displacing the water.

    What do you think? Im just after a really simple way of getting the flow rate do you no of another method I could try out?

    Heres a link Danger to show you what an ejector is

    http://www.cheresources.com/ejectors.shtml
     
  10. Oct 22, 2006 #9

    Danger

    User Avatar
    Gold Member

    Thanks... that helps a lot. So I take it then that the purpose of these things is to create a vacuum elsewhere in the system? So the 'open-ended' incomplete pipes in the linked diagram (top in W, Y & Z stage, and left side in X stage) lead to vacuum vessels?
     
  11. Oct 22, 2006 #10
    Not sure about them leading to vacuum vessels? The way I seen it in the diagram thats where the motive fluid is being injected. The easiest way to think about ejectors atleast for me is just to think of them as pumps, except they have no moving parts.
     
  12. Oct 22, 2006 #11

    Danger

    User Avatar
    Gold Member

    Oh. I was going by the text in that link, which seemed to indicate that the scavenging was used to create vacuum for other purposes (ie: pumping air out). So if there's some other fluid incoming from those pipes, how is it that the steam doesn't corrupt it?
     
  13. Oct 22, 2006 #12

    FredGarvin

    User Avatar
    Science Advisor

    Ejectors are used as mixing devices mostly. They are used (as I have used the,) to lower the pressure in other parts of a system. Here is another link that might help. As a note, Larry Fox and his folks at Fox valves are a great asset to have in this area:

    http://www.foxvalve.com/air_ejectors/index.html

    As far as your idea of volumetric measurement, I think that is good idea and is quick and dirty. You are correct in that an orifice does need proper sizing and calibration to determine it's discharge coefficient if you desire an inaccuracy below, say 5%. The other drawback is that if your measurement range is rather broad, then you need multiple orifices to cover the range.
     
  14. Oct 22, 2006 #13

    Danger

    User Avatar
    Gold Member

    Thanks for that link, Fred. I got caught up in the steam-powered thing from the other link, and forgot that the initial mention here was about compressed air.
     
  15. Oct 22, 2006 #14
    Thanks for the input guys, I'll have a go with the test this week and let you both no how it went.
     
  16. Oct 24, 2006 #15

    Q_Goest

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Hi jimdean. I'd like add to what Fred and Danger have already said, and to respond specifically to the question:
    Note that the gas ejector you're refering to is also called an eductor or a jet pump. They're found all over the place, there's one in your toilet that uses water. There are some references on how to design them, you don't need CFD to do that. Look in your library for "Pump Handbook". Chapter 4 has a pretty fair amount of information on design of eductors, what Cunningham calls "Jet Pumps".

    Note that as you vary one of the 3 fluid parameters, the others will change along with it. In your case, you've fixed the outlet pressure (atmospheric pressure) and you've essentially fixed the "primary fluid" pressure (inlet pressure of the eductor from the can).

    The secondary fluid (entrained fluid) will vary in flow depending on pressure at the point it is entrained. In your case, the secondary fluid pressure entering the entrained section can be Patm at the most. The lower this pressure is, the lower the flow rate will be. So the highest flow rate will be when the secondary fluid pressure is equal to Patm. Flow would of course be higher if you elevated this pressure above atmospheric, but the device you're refering to is essentially a vacuum cleaner, so that's not applicable. In the case where you have some lowest possible secondary pressure, the flow will drop to zero. That's what your tutor mentioned, and why simply putting a manometer on that line won't give you what you're looking for - a pressure/flow curve for the device.

    There's a very easy way to create that pressure/flow curve for your secondary fluid. Simply put a rotometer on the line and vary the restriction through it. You should also consider putting a vacuum gage on this line (manometer would work) to measure the pressure. Rotometers are very inexpensive, though you'll need the right one to get the range of flows that you'll want. You might get in the ball park using the CFD analysis you did. If that showed a flow of X, consider how that flow will increase or decrease given a change in pressure at the secondary fluid inlet.

    Cheap rotometers can be purchased at industrial supply stores for about $30 US. I'm guessing you're not in the US though because of the reference you have to the product. Not sure what stores are used overseas for such things, but in the US there are McMaster Carr, Grainger, MSC Industrial Supply and others.

    You might also consider what rotometers there are lying around the college you attend. Ask your tutor/instructors and they might even be able to locate one you can borrow.

    One last thought. The picture of the "gas ejector" you provided shows something that is very inefficient for a large number of reasons. It's a horrible design. If part of your project is to improve it, there's a lot that could be done. If you want help with that, let me know or look at the Pump Handbook I mentioned above.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Gas ejector experiment
  1. Cantilever experiment (Replies: 2)

  2. Ethanol gas (Replies: 3)

  3. Cantilever Experiment (Replies: 0)

Loading...