Hello all
I hope you are fine.
I am trying to do a simple simulation of an impellor design.
I have placed my impellor in a tube, with length 5 diameters of the impellor. Boundary conditions are only 0 pressure in both end of the tube.
My rotating region is big enough to have the whole impellor inside it but at the same time it does not touch the pipe, I would say that is in the middle of the distance almost.
First question is the following. When I am importing my model, CFD creates two extra parts, the volume of the air inside the tube, which is normal, and then a volume inside the rotation region. So I have my rotating region, inside another volume and inside there my impellor. Which one I assing as rotating region? The one that I design or the volume created inside of it that actually touches the impellor?
If I put the outside (the one that I design ) the inside should I just assing it as air??
Thank you very much in advance.
Kind regards
Dimitris
Hello Jon
Thank you very much, one more time.
I could guess that OD and ID are outside and inside diameter (just to be on the safe side)
Also, you mentioned to make them longer. Do you mean to make the caps longer or if better to make the pipe longer and keep the caps really thin?
Kind regards
Dimitris
Yes, correct.
Make the caps longer, otherwise you are splitting the mesh right before the boundary condtion. Or don't have caps at all and use the CFD void fill function (under Geometry), either is fine.
Thanks,
Jon
Hello Jon
I hope you are fine.
I have put the latest support file in the Dropbox folder. I have followed all your suggestions, hopefully I did it right.
Please take a look and tell me what is wrong with it. The solver stops after it just starts with a message that the solver stopped unexpectedly.
Thank you very much.
Kind regards
Dimitris
Hi Dimitris,
I would suggest you might have too many elements.
Did you make the rotating region (RR) and impeller touch the wall? It looks like there is still a very small gap there, is that critical to you?
I am not sure why there is an air volume around the RR.
Check your units too, is this really 150m in diameter?
I would not worry about the Gap Refinement in the mesh, just close up that gap so there is no need for it 🙂
Hope that helps,
Jon
Hello Jon
Thank you very much for the reply.
The rotating region touches the wall completely. I can see that since when the model loads it creates two volumes, one in front and one in the back of the region. Of course I assign them both as air. I left the impellor at the same size since I want to keep it realistic.
When you create a rotating region in you 3D software, CREO in my case and then you imported to the CFD it creates a volume inside the rotating region, in order to "filll the gap", since the rotating region is a hollow part. I apply the rotation region in the volume that touches the impellor and the outside volume as air.
The units are correct.
In the surface refinement I will try to make it a bit more coarse to see if it can run. I will not use Gap refinment.
Thanks
Dimitris
Hi Dimitris,
I think then that the RR needs to be a solid part, not hollow, so there is no internal air volume.
I do still think we should close that gap, or you will need 3-4 elements inbetween the impeller tips and the wall (so that air volume might prove useful). Knowing the scale of this model though, I would start with it closed, this is going to need substantial meshing to capture the flow regime.
Then the RR can be the part that touches the wall with no thin air volume present.
It might also be well worth your time to test the advanced turbulence setup talked through in this webinar: Turbulence Models in Simulation CFD.
Kind regards,
Jon