axisymmetric boundary condition

The graphical user interface of Elmer

axisymmetric boundary condition

Postby guybrush » 09 Oct 2009, 18:16

Hello,

I would like to model an axisymmetric object. I have drawn and meshed the 2D surface of rotation with Gmesh and have imported the resulting mesh in ElmerGUI. My problem now is that I cannot find how I can define one of the lines of my model as being the axis of rotation. Under the menu "Boundary Condition", there is a tab for periodic boundary conditions, but I haven't found how to use this. I mean I have to set the coordinates system to "Axisymmetric" under setup, but how is the axis, around which my section is spun, defined?

Any help would be greatly appreciated.

Many thanks in advance

Guybrush
guybrush
 
Posts: 2
Joined: 09 Oct 2009, 18:09

Re: axisymmetric boundary condition

Postby mal » 09 Oct 2009, 19:00

Hi,

The axis of revolution is y (that is, x=0).

Mikko
mal
Site Admin
 
Posts: 54
Joined: 21 Aug 2009, 14:21

Re: axisymmetric boundary condition

Postby hazelsct » 10 Oct 2009, 14:04

There is no automatic axis boundary condition, you need to define it by hand, e.g. x-velocity = 0. But the natural zero-flux boundary condition should work for many field variables.

This is a possible feature request though: it would be convenient if Elmer had an axis or symmetry boundary condition setting all of the normal fluxes, normal velocity, normal E and B field components, to zero.
hazelsct
 
Posts: 153
Joined: 05 Oct 2009, 17:02
Location: Boston, MA, USA

Re: axisymmetric boundary condition

Postby raback » 10 Oct 2009, 21:57

hazelsct wrote:This is a possible feature request though: it would be convenient if Elmer had an axis or symmetry boundary condition setting all of the normal fluxes, normal velocity, normal E and B field components, to zero.

This could indeed be implemented. Limiting to cases where x=0 is the symmetry axis would be rather straight-forward. However, the symmetry axis may also be some other axis. This would require some additional intelligence from the code.

The implementation could perhaps be put to the Solver_init code of each solver. If the symmetry flag would be present an appropriate Dirichlet condition would be added to the boundary list structure. And in GUI just a flag in BoundaryCondition / General as for the periodic stuff.

-Peter
raback
Site Admin
 
Posts: 1943
Joined: 22 Aug 2009, 11:57
Location: Espoo, Finland

Re: axisymmetric boundary condition

Postby guybrush » 12 Oct 2009, 08:55

Thank you for your replies.
So if I understand it correctly, I have to:
1) draw the model so that the axis of rotation corresponds to the y axis (i.e. x=0)
2) Choose axisymmetric coordinates
3) Define the boundary conditions on the axis accordingly. I am doing a static structural simulation, so setting x-displacement=0 on the axes should do the trick.

Thanks,

guybrush
guybrush
 
Posts: 2
Joined: 09 Oct 2009, 18:09


Return to ElmerGUI

Who is online

Users browsing this forum: No registered users and 2 guests