Rotating Circle in a parrallel plate
 

Hi friends

I want to rotate a circle(solid) inside a rectangular domain(fluid). what type of mesh have to create. I'm using gambit for meshing and solver fluent. Please help.:)

is it 2d?
will your circle rotate (moving mesh) or is it steady state problem?

My problem is 2D and transient. Initially i want to rotate the circle then i have to implement translation motion also. Now i'm concentrating only on rotation.

*create your rectangle
*create your disk
*place your disk on its right position.
*split the rectangle with your disk
*copy the disk anywhere with a translation vector
*delete the initial disk (check with shaded mode that your rectangle has an hollow in place of the disk)
*define the circle edges (from rectangle) as interface
*define the circle edges (from copied disk) as interface
*move the copied disk with opposite translation vector back
*define the disk domain as fluid continuum
*define your other BC
*mesh
*export
That's it for Gambit side

Thank you..
I'll follow the procedure.

you told that "define the disk domain as fluid continuum".
Actually my rotating body is solid so i should be defined the disk domain as solid and the rectangular domain to be fluid?? is it correct??

Can i use any type of meshing (structured or unstructured) for circular and rectangular domain ?

you can choose any kind of mesh, is it only your choice.
Ok then is your disk is solid, you are rightm you disk will appear as hollow since no NS-equations will be solved.
Than you can create a circle with greater radius than solid-disk, and split the rectangle.
And you do the same as previously explained but with an annulus

Sorry..
i didn't understand what you have told may be because I'm new to moving mesh and FSI idea.

*create your rectangle
*create your disk
*place your disk on its right position.
*split the rectangle with your disk
*delete the initial disk (check with shaded mode that your rectangle has an hollow in place of the disk)
*create a circle with radius greater than solid-disk (create and move at right place)
*split the previous domain with the new circle (it will create an annulus around your solid disk which is hollowed)
*copy the annulus anywhere with a translation vector
*delete the annulus surface
*define the circle edges (from rectangle) as interface
*define the outer-circle edges (from copied annulus) as interface
*move the copied annulus with opposite translation vector back
*define the annulus domain as fluid continuum
*define your other BC
*mesh
*export

Thank you so much.. :-)
i'll follow the procedure..

Can we use boundary layer meshing instead of creating annulus region?? i have seen one tutorial of concentric cylinder rotating. In that they used boundary layer meshing. But i didnt get the complete tutorial. So i dont know for what purpose they have used.

you can also attach a BL on the inner-circle from annulus (this is par of meshing)

If you don't want to move you mesh (sliding mesh), then forget all points after "*delete the initial disk ".
And then you will have to set the circle as wall BUT with a velocity,
And off course you won't see you grid moving

thank you very much for clearing the doubts and helping.. :-)

GAMBIT Geometry
 

Dear MAx,

Hello Sir/Ma'am.
I want to make the kind of geometry as shown in the image (attached) in the gambit software. I believe that you can help me.

Awaiting your reply soon.

Thank you.

In my case the clearance between circle and walls of the rectangle is very less (.1 to .3mm). In this case how can i mesh??

Please open a new thread with your request (I will reply) ;)

Post a picture

Please find the attachment.

decrease circle's radius, else you will have skewness problem

My geometry is like that. The clearance is very less

the circle represents your rotating solid?

Yes my circle is rotating and solid

This is typical reason why you need to post picture of your problem at the beginning.
Ok then your annulus will be very thin. Let's say, the outer radius from annulus will split the gap between solid and rectangle into 2 parts

Sorry, next time I'll post the picture. But all your replies is useful for me.
Now you are telling to split the geometry into two at the annulus region right??

If i split geometry in gambit , is it possible to combine the geometry in fluent??

no you create your second circle so that the gap between your solid-circle and rectangle-edge will be cutted in half

sorry can you pleas explain?

Now i just want to rotate the solid body (no sliding mesh) with this current geometry.

Next step I'll make sliding mesh.

no sliding mesh = no rotation
or you can use moving mesh or MRF (check points 10 and 11: http://aerojet.engr.ucdavis.edu/flue...g/main_pre.htm)

Attachment 16335

Attachment 16336

Thank you.. :-)

Do you have any simple tutorials for rotation or sliding meshing??

check on the web site I mentionned earlier, you can found the tutorials

I need meshing tutorial. The available tutorials are problems in Fluent solver. Any tutorial is available for meshing in gambit related to moving mesh/rotation?

i meshed the geometry in gambit. Initially I'm doing MRF ( by giving angular velocity to Circle keeping circle as wall). Can you check the meshing is correct? I split the faces and did meshing.

there is a tool for checking mesh quality in gambit:
http://aerojet.engr.ucdavis.edu/gamb....htm#gc_exmesh

Thank you.. Actually there will be no persons for helping. That why asking doubts like this. I'm new to all the cfd modelings. Now i got some idea about (theory) SRF,MRF,sliding mesh and dynamics mesh. Once again thank you.. :-)

Hi
I rotated circle inside the rectangle domain by giving circle as wall with a angular velocity. But I'm not getting required result. As you told earlier giving rotation to the annulus region (fluid) by considering concentric circle is also not possible because of less clearance between circle and rectangle. Is both the method obtain same result??

So can i consider a rectangular region around the circle and giving rotation to the region(shown in fig) between smaller rectangle and circle? Is it possible??

you need to set rotation speed on annulus domain, but no rotation to inner circle (solid), since it already belongs to the annulus domain (and consequently has rotation speed)
your last method won't work.
The easiest way would be to set only the inner circle with velocity (but without moving grid)

But the clearance is very less then how can i make annulus region and meshing it??