Fluids

Fluids

Topics related to Fluent, CFX, Turbogrid and more.

Should fluid be rotating or wall be rotating

    • tracyxu
      Subscriber

      I am solving the problem in the FLUENT tutorial "Using a Single Rotating Reference Frame" to check some doubts I have. I got different results if I set the "frame motion" of "fluid zone" to be rotational as in the tutorial, OR if I set the wall boundary condition to be rotational. I use the rotational speed 400 rpm instead of the value in the tutorial to see the difference.


      In the case of setting up the "frame motion" of "fluid zone" to be 400 rpm, the pressure is from -25.9 pa to 4.7 pa (left). However, in the case of setting up the rotational wall boundary condition to be 400 rpm, the pressure is from -15.1 pa to 4.6 pa (right). I put the two pressure contour plots side by side with the same color map as a comparison.


      For the velocity, the velocity magnitude in the core area from "frame motion of fluid zone" (left) is larger than that from "rotational wall boundary condition" (right), although the maximal velocities are the same.


      Can anyone tell me why I got different results? And which setting would be the correct one? In my opinion, the wall is the rotating body in the physics world. So the results by setting up the rotational wall boundary condition should make more sense, right?


      Thank you very much!



    • Amine Ben Hadj Ali
      Ansys Employee

       When reference frame is activated Fluent solve four additional terms in the momentum equations ( Coriolis acceleration and the centripetal acceleration). When these terms are very small then moving wall approach and single moving reference frame would result in similar result.


      Are both cases deeply converged?

    • tracyxu
      Subscriber

      Hello Amine,


      Thank you for your explanation. Since Coriolis acceleration and centripetal acceleration are solved when reference frame is activated, then the results from the single moving reference frame should describe the physics better than the moving wall approach, right? However, the results I am showing below confuse me.


      The results I showed about have the convergence criteria 1e-3. I tried to use 1e-6 for the criteria after I saw your message. However, the case of single moving reference frame converges only to 1.5e-4. The velocity and pressure contours are as below. Notice that the pressure is in the range of -15.6 pa to 4.6 pa.



      For the case of moving wall approach, the solutions are able to converge to the criteria 1e-6. the velocity and pressure are as below. What we can see here is that when I decreased the criteria from 1e-3 to 1e-6, the results of the moving wall approach don't change much. However, the results of the single moving reference frame converged to the results of the moving wall approach. So, again, my question is, which approach describes the physics better? Thanks again.


      Tracy


Viewing 2 reply threads
  • The topic ‘Should fluid be rotating or wall be rotating’ is closed to new replies.