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

Scaling an Axial Compressor

  1. Oct 12, 2012 #1
    Let's say I have a precisely designed axial compressor, comprising of stages of rotor and stator wheels, that resembles something like this:


    Such a compressor, driven by a certain torque at a certain RPM, will deliver a certain air mass per time at a certain pressure. Now, what happens if the compressor is scaled down, so that the diameter is half of what it used to be but all other angles, etc. remain the same? The RPM will be such that the speed of the blade tips is the same in both compressors, i.e. the RPM will be higher in the smaller engine.

    My intuition and some quick calculations tell me that the pressure ratio will remain the same, the mass flow will be divided by 4 (since cross sectional area is divided by 4), and the torque will be divided by 8 (since I think power will be divided by 4 and RPM will be doubled). But, again, I'm not sure, and I'd like to be. Can anyone help? Thanks!
    Last edited: Oct 13, 2012
  2. jcsd
  3. Oct 15, 2012 #2
  4. Oct 16, 2012 #3
    In axial and centrifugal compressors, you keep the peripheral speed when scaling. This keeps the gas pressure and as well the mechanical stress.

    angular speed *2
    pressure *1
    mass flow /4
    torque /8
    power /4 (consistent with the mass flow)

    Needless to say, some technological reasons speak against easy scaling, for instance leakage, Reynolds... The combustion chamber is also more difficult to build.
  5. Oct 16, 2012 #4
    That makes sense, thanks. I wasn't sure if it scaled that linearly, and it probably doesn't, but that kind of simplification is probably OK for only cutting the diameter in half. Maybe it would be smart to approximate that it will need about 2% more power than predicted, and have 2% less throughput, due to gap losses etc., too. I looked at the Reynolds number a little, and it doesn't seem like it will matter that much. In the smaller engine air should be a little less turbulent.
  6. Oct 18, 2012 #5
    Happy you if you can predict flows to 2%...
  7. Oct 18, 2012 #6
    I have a chart of gap loss to gap width. A .07 mm width gives a gap loss of around .7%. Relative to the compressor, then, if the gap is around .07 mm, the gap will double relative to the size of the compressor since it's difficult to get the gap smaller at this size. I made a few estimates and quick calculations, and am coming up with around a 1.3% additional loss. I just decided to round up since there will probably be some other losses from building at a smaller scale since the tolerances of manufactured parts won't scale.
  8. Oct 18, 2012 #7
    In general, when the length of the blades gets down to one inch, it is time to switch to a centrifugal stage. At that point the efficiency of the axial stage will be less than the centrifugal stage. Efficiency is not as simple as you say. It is a function of tip clearance/blade length ratio. But tip clearance is pretty constant regardless of length. The blade packing density also gets too high for the smaller disks. That can lead to surge problems.

    If you compare sections of real engines, you will see that the front of the flow path for an axial compre ssor looks very much like the flow path of a centrifugal compressor. That is because it is a hybrid design. That is something the engineers learned fairly recently to combine the physics of both into one for better overall efficiency.
    Last edited: Oct 18, 2012
  9. Oct 18, 2012 #8


    User Avatar
    Science Advisor
    Homework Helper

    Can you give a link to a drawing of that concept?
  10. Oct 19, 2012 #9
    Probably. But not from this phone. I'll get back to my computer tonight.
  11. Oct 19, 2012 #10
    It actually turns out that's not really the case. It's a common misconception of the model jet engine community that axial compressors are not efficient at that size. The very best small jet engines use axial compressors. Also, you can't buy any centrifugal compressor wheels that small, and it's easier/cheaper to make an axial compressor than a centrifugal one.
  12. Oct 19, 2012 #11
    Then it is a misconception commonly held by all the engineers who design engines for real airplanes. I don't think you can name one engine currently in production in significant numbers that I can't in five minutes walk to the desk of at least one engineer who was on the design team. If I told any one of them that, then they would strongly disagree.

    As for engines on small model toys, I yield to your much better understanding. I don't know anything about them at all.
  13. Oct 19, 2012 #12
    As for the misconception, I was talking about the model community. What about Bladon Jets? A smaller company I believe, but their engine puts out 90+ lbs of thrust with a 4.5 in diameter, and there aren't any centrifugal stages. The very best centrifugal model engines put out ~47 lbs of thrust at that size.
  14. Oct 19, 2012 #13
    Aleph Zero:

    Sorry, I could not find anything published to the public domain as to how concepts of centrifugal compressor design is incorporated into axial compressor design, to produce a more efficient hybrid that looks like a regular axial compressor to the person who does not know the difference. I can tell you generally how this is done, but keep in mind that I’m also being careful not to violate any Intellectual Property (IP) agreements.

    First, I’ll mention the two things are done different from conventional axial design. Then I’ll tell you generally how they work together to make the compressor more efficient. If you have a good understanding of the aerodynamic analysis that goes into both an axial and a centrifugal compressor, then you can derive much more from what I say. But if not, you will still understand the basics of what is being done.

    The first of the two things that is done is only done to within the space constraints available. A flight engine may not have much space available because you have to keep the compressor small to permit the bypass flow from the fan. You are also somewhat space limited near the engine centerline because of the need to put a fat fan shaft out the front of the engine. But you don’t have these space constraints in most industrial gas turbines, so you can do this to a greater extent. The idea is to get the flow path at the front of the compressor to look like the flow path at the front of a centrifugal compressor. Basically, the air enters near the center and immediately finds itself directed to a flow path at a much greater radius.

    If you Google images for the PW4000, which was designed about the time these things were learned and by the same engineers who first learned them, you will see where they did their best to make the front part of the flow path look this way, but were seriously limited by these space constraints.

    The second thing they do is they make the bottom of the front blades look like a centrifugal impeller. They even call that the impeller section. The cross section of the blade in that area does not look like the regular air foil section of a compressor blade. The top of the blade looks like a regular airfoil section, and that is where nearly all the compression work is done.

    I said two things, but I guess I meant three. Besides this, the vanes behind the blades are designed to perform a similar function as the diffuser section of a centrifugal compressor. But in my opinion, you can’t tell the difference by looking at them, even though the aero guys say they designed it that way. Mechanical designers like me make the blades to the section defined by the aero guy, and we don’t always understand all the details as to why they look that way.

    As the air enters, it is spun to a great tangential velocity by the lower impeller section of the blades. It moves radically and is compressed as it does so, just like in a centrifugal compressor, until it gets to the section of the blade shaped like a regular air foil, which retains much of the tangential velocity, but also moves the air back to the vanes. The vanes act like a diffuser and straighten out the air flow and distribute it properly to repeat the process in the second stage of impeller/blades.

    If you check out the Ramgen web site, you can see where they brag about a new compressor design that they are trying to develop. It is an idea that many people have tried to make work in the last five decades. All the aerodynamics professors believe that is should work and that it should give us a compressor that is both a whole lot more efficient and less expensive to build. Perhaps Ramgen will be successful. The idea is to take these principles to the max and shape the whole blade like an impeller, which results in supersonic airflow coming off the back of the stage. That much is easy. What is hard is the next step where they process the air to convert all that airspeed into pressure efficiently. If someone can figure out how to make this work, then they will be able to replace five or six normal axial stages with one impulse stage. If they are successful, it may very well make carbon sequestration economically viable, and then we can have a coal plant with zero emissions. How much would that change things?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook