Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Mesh visible in solution
Posted May 25, 2010, 11:09 a.m. EDT 1 Reply
Please login with a confirmed email address before reporting spam
Hi all!
I'm simulating a coupled 3D problem including stokes flow and solid mechanics (there are only small displacements therefore I do not use an ALE) the get the response of a vibrating beam in a fluid. If I plot the velocity field there are slight "discontinuties" which seem to come from the mesh (see attached file, rectangle: solid, rest fluid). My question: Is this a graphical problem? I guess not, but where does it come from? I've seen this effect also in other simulations.
Does anyone have suggestions, how to solve the problem? I would prefer not to use more degrees of freedom (a denser mesh oder higher order elements), since my hardware is rather at the limit.
Thanks for any suggestions.
Philipp
I'm simulating a coupled 3D problem including stokes flow and solid mechanics (there are only small displacements therefore I do not use an ALE) the get the response of a vibrating beam in a fluid. If I plot the velocity field there are slight "discontinuties" which seem to come from the mesh (see attached file, rectangle: solid, rest fluid). My question: Is this a graphical problem? I guess not, but where does it come from? I've seen this effect also in other simulations.
Does anyone have suggestions, how to solve the problem? I would prefer not to use more degrees of freedom (a denser mesh oder higher order elements), since my hardware is rather at the limit.
Thanks for any suggestions.
Philipp
Attachments:
1 Reply Last Post May 26, 2010, 3:58 p.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
my first reaction is that I find your mesh rather coarse, you should at least run once with a refinement of 1 just to evaluate the sensitivity to the mesh, but you could also reduce the order, as this would avoid any "ripples" that could come from the higher order polynomial on your rather steep step functions.
Finally does the physics allow ripples / oscillations ? or do you expet smooth transitions ? Finally FEM is there to give you some good clues of what is going on, not necesarily to model reality to the last digit.
Have fun Comsoling
Ivar
my first reaction is that I find your mesh rather coarse, you should at least run once with a refinement of 1 just to evaluate the sensitivity to the mesh, but you could also reduce the order, as this would avoid any "ripples" that could come from the higher order polynomial on your rather steep step functions.
Finally does the physics allow ripples / oscillations ? or do you expet smooth transitions ? Finally FEM is there to give you some good clues of what is going on, not necesarily to model reality to the last digit.
Have fun Comsoling
Ivar