Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
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.
Result is not symmetric in y-dir while geometry and boundary conditions are symmetric.
Posted Mar 31, 2012, 2:26 p.m. EDT Fluid & Heat Version 4.2a 3 Replies
Please login with a confirmed email address before reporting spam
I've tryed to see how much reaction forces exerted on the structure within the water flow tunnel.
Instead of modelling half of the structure in y-coordinate, I just model all in 3D.
Current is flowing in x-direction, and geometry of structure and tunnel is symmetric in y-direction.
Regardless of any changes in boundary conditions, symmetric boundaries, flowing direction, mesh size, turbulent or larminar, the reaction force on the structure and the pressure on the structure by flow current are not symmetric with respect to y direction.
for example, in the atteched file, I have reaction force in y-dir on the structure is -1887.67 N with the setting that + x_dir flowing. It is fairly small value compared to x direction force, -9.24e6 N.
Now, if I just change the flow direction to - x_dir(minus), I can get similar values of reaction force in x,z direction. However in y-direction, I get -22926.8N.
the reaction forces in y-direction are varying a lot by every changes and also it always comes up in (-),minus, value.
How do I treat these unexpected results? Every things in y direction are symmetric, but I can't get rid of the unbalanced numbers in y-direction.
Can anyone explain this results from COMSOL?
Thanks in advance,
p.s. I used Volume integration in Derived Values node in Results to get the reaction forces
Attachments:
Please login with a confirmed email address before reporting spam
I would suspect first of all non symmetric mesh effecting the numeric results, have you tried to increase the mesh density, and to mesh one half and mirror the mesh along the symmetry boundary ?
Also if you have some singularities (sharp corners, you can easily get rather large errors on force calculations if you integrate over such singularities
But it might still be something else, if you are in a turbulent regime, do not expect turbulence to respect absolutely the symmetry
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Thank you for quick response.
I also tried larminar and turbulet. Since I know I cant expect symmetric result from turbulent so I'm using larminar flow to find out what makes differences in y-direction.
Anyway, I also tried finer and extream fine mesh for this case, but sometime I can get larger number of unsymmetric force in y-dir from the finer mesh.
If it is mesh problem, Shouldn't I get opposite number if I change to the opposite inlet condition?? I'm wondering how i get the consistant (-) reaction force in y-dir for any scenarios...
Please login with a confirmed email address before reporting spam
Meshing is one possible issue there are probably others, as Im not behind my WS I havent seen your model yet, not back before a couple of days though, and there are others out here with much more experiene influidics than me ;)
--
Good luck
Ivar