1 Reply Latest reply on Sep 18, 2013 3:08 AM by Jared Conway

    internal flow parallel to rotating cylinder (not across)

    James Lin

      Hi Everyone,


      I have setup a very simple study about rotating region. I have a smooth cylinder that spinning at 10,000 rpm inside a tube where there is a constant mass flow flowing parallel to the cylinder wall surface. Here is two finding I had.


      1. The part to represent the rotating cylinder must be biger than the cylinder in order to capture it in this case. Other wise my mesh will fail if I make the rotating region the same size as the cylinder because there would be no air flowing through the rotating region if the retating region is same size as the cylinder. (This is not the same case as a propeller where it must be retating region is always bigger since the blades are within the rotating region part).


      2. Second, I have given the rotating cylinder constant volume heat source of 100W. And see how is the mass flow of air cooling down the rotating cylinder. There is a very weird result I got. The solid temperature of the cylinder did get cool down over time as expected. However, the outlet mass flow has a lower temperature than the inlet mass flow. This does not make any sense. The cooling air should have picked up the heat from the rotating cylinder and its temperature should have increased.


      At last, I have another concern about this setup. That is I have to make the rotating region bigger than the rotating cylinder. The rotating region has a bigger OD than the cylinder. So this mean another thing which we manually determine how much of the air is force to rotate (this is wrong, it is the roughness of the cylinder surface that able to drag the air around the cylinder; not what I set it to be). I do not see that I can enter a roughness in my setup.




      I think I am not the only one who setup something like this before. Or I might have a wrong setup in a case like this. If anyone can help here?



        • Re: internal flow parallel to rotating cylinder (not across)
          Jared Conway

          First recommendation, look through the swx kb and the solving engineering problems document. That will tell you about sizing the rotating reiln. You don't mention the gap between the parts but you'll see that this is usually recommended but you need to us stators.


          Second, it sounds like a moving wall would be a better bet.


          Third, I don't know if you can heat something that is a rotating region. Moving wall might work. Check the kb.


          It also sound a like you want to use roughness with rotation, I do not believe that is possible.