[nina1234]

Hello,

I am trying to model a rotating propeller. I watched several youtube tutorials. The idea is to create a static domain (box), and a rotating domain (cylinder) inside, which contains the propeller.
I created booleans in order to separate the cylinder from the box and the propeller from the cylinder.
I checked my mesh and everything looks fine.
I checked that Ansys created interfaces between both domains (rotating and static) but when I look at the results it behaves as if the cylinder would be a solid...

If someone knows how to make my cylinder "transparent" so that the air flow could interact with the propeller directly, thank you!

[AlexanderZ]

picture 3

the issue is: you have 2 parts 2 bodies
solution: select your bodies -> right mouse click -> make a new part
to be: 1 part 2 bodies

remesh

[nina1234]

Thank you very much, really helpful.
The velocity contour has the right shape now, but when I look at the streamlines around the propeller, they still get stopped...
I displayed them on a plane for better clarity. I would expect the flow to rotate a bit and then continue until the output face.
Also, when computing the thrust force value, it appears to be half of what I expect, and that is my main concern.

[nina1234]

I tried with a 2D model for better efficiency in the computation time and I got the same behavior. I made everything to one part as you told me, the contour is fine but the streamlines are still considering the rotating domain as a border.
Here, the triangle is not moving, no rotation. Only the flow goes from left to right.

[AlexanderZ]

from these pictures I can see, that it is cfd-post plotting issue

play with settings to plot stream line in rotation area as well.
other option: you may plot pictures inside fluent

[nina1234]

Alright, thank you very much Alexander !
Last question; any idea on why the thrust appears to be half of what I expected?
I checked the mesh quality and I tried several parameters on the time steps, doesn't change my result.

[zeldaa]

Hello
Iam also modeling a rotating propeller. I watched youtube tutorials. And I created a static domain and and a rotating domain and both of them are cylinder.
I created booleans in order to subtract the rotating domain from the static domain and the propeller from the rotating domain
I checked my mesh and everything looks fine.
Ansys created interfaces between both domains
I run the solution and my results for thrust and torque (the parameters that I need) were really good compare to the experiments

But my velocity contour is like what nina1234 said.

Is this a serious problem?

[AlexanderZ]

make a picture of veloity destribution in your domain in fluent (not cfd-post)

[zeldaa]

hello
Thanks, AlexanderZ for your reply.
I made a picture in fluent and it seems okay now, I posted a picture from the x-y plane for y velocity and its looks like what I've seen in one of my references.
but I have another question if it is okay
fluent makes interfaces automatically when I create a boolean in the design modeler, and with this auto-generated interface everything seems fine but I have some worries... how could we understand that these interfaces are correct?
what are the signs of incorrect setup for the interface?

Thank you!

[AlexanderZ]

interface must have coupled boundary condition (in most cases)
you may check it if you open interface tab

on the other hand, if it diesn't most likely you will get diverged solution which leads to error or you will see, that flow/heat is not propageted from one zone to another

[zeldaa]

dear AlexanderZ, thank you for your reply,
I've just checked what you said
I couldn't find the coupled option in this version(2018) but I've seen it in version 2021R1 and I've never used coupled option, and all my results are based on this.
so when I checked the mesh I can see the mesh for connected regions but there are also two walls that I can't display on the mesh tab
and also I've never got a divergence problem and everything seems fine
and based on the contour that I've shared before it seems that flow passes through the zones, am I wrong? (which contour in which plane should I check? could you please tell me)
of course, I have a problem with my streamlines like nina1234 but I've seen many people have this problem and we have to do sth in CFD-post (i don't remember right now)

about my zones: they are both fluids(air) and I use the MRF method.

so because I've never got a divergence problem and my contour is like that and also my results are close to the experiment can I trust these auto-generated interfaces?


thank you so much for your replies and help

[AlexanderZ]

flow passes zones correctly, dob;t seem any problems here, no problems with streamlines as well (problem was with the way you made a picture in cfd-post)
you can trust this interface

[zeldaa]

dear AlexanderZ,
yes I've seen others have problems too and the solution is here I think (I've shared the link in case someone need it)
https://www.cfd-online.com/Wiki/Ansy...f_reference.3F

Thank you so much for your help AlexanderZ