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 a thin plate with a hole

Please login with a confirmed email address before reporting spam

Dear all,

I met a problem when I try to mesh a model with a thin plate inside as in the attached figure.
The size of outer block is 100*100*200 [m], but the thickness of the thin plate is 0.2[m].
Actually, the thin plate is attached with a hemisphere as an entire object.
When I try to use the traditional mapped+swept mesh method, there comes the problem.
In the manual, the first requirement for swept mesh is that each subdomain must be bounded by one shell (no hole there), but the interface between plate and hemisphere leads to a hole of the top boundary of thin plate.
Such trying will lead to Error:4125. Invalid topology of subdomain.
I am still on it, and it will be appreciated if there are some suggestions.

Regards.

YAO


3 Replies Last Post 26.12.2009, 07:25 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.12.2009, 16:51 GMT-5
Hi - not sure this works, but if I understand you right, you should be able to define a 2D mesh on the internal boundary between the hemisphere and the plate, respectively, the face of the thin plate with the hole in (using Free Mesher/Boundary). Now you should be able to mesh the thin plate with the swep mesh. The Free Mesher should now work on the hemisphere.
-- kurt
Hi - not sure this works, but if I understand you right, you should be able to define a 2D mesh on the internal boundary between the hemisphere and the plate, respectively, the face of the thin plate with the hole in (using Free Mesher/Boundary). Now you should be able to mesh the thin plate with the swep mesh. The Free Mesher should now work on the hemisphere. -- kurt

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.12.2009, 09:09 GMT-5
Hi, K.P.
Thank you for your reply.
As you mentioned, it is true that I can mesh the interface between two objects, but, the interface belongs to the thin plate will definitely has a hole, and it did not work when I try to use the "swept mesh".

Currently, my solution comes from a poster here named "mesh a thin plate" or something similar.
I divided the thin plate into at least two pieces-- a plate with a central hole and a central part whose surface is exactly equal to the bottom of the hemisphere. In this case , I can use the Swept mesh with no problem.

But, I made the central part by extruding the 2D circle, and I found the edge of circle is not consistent with the edge of 3D shpere well, or I should say rather rough.

These un-matched interface will lead to the inverted mesh element at their edge and a warning accured after all the calculation.

Does anyone meet the rough 2D problem before?
Hi, K.P. Thank you for your reply. As you mentioned, it is true that I can mesh the interface between two objects, but, the interface belongs to the thin plate will definitely has a hole, and it did not work when I try to use the "swept mesh". Currently, my solution comes from a poster here named "mesh a thin plate" or something similar. I divided the thin plate into at least two pieces-- a plate with a central hole and a central part whose surface is exactly equal to the bottom of the hemisphere. In this case , I can use the Swept mesh with no problem. But, I made the central part by extruding the 2D circle, and I found the edge of circle is not consistent with the edge of 3D shpere well, or I should say rather rough. These un-matched interface will lead to the inverted mesh element at their edge and a warning accured after all the calculation. Does anyone meet the rough 2D problem before?

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.12.2009, 07:25 GMT-5
Hi

When I see your jpg, my first question is cannt you transform these cubes into cylinders?, so that you can use 2D-axi and get rif on one DoF.

If not, then I would split the parts into 4 symmetric items, using the split lines of the hemisphere. there are then 2 ways out.

Either the case is symmetric so one can work with symmetric/anti-symmetric boundaries and mesh only 1/4 of the problem.

Or, to work in full 3D. By having split the volumes, it is easier to apply a sweep mesh.

By the way a sweep mesh can be defined manually from ONE or SEVERAL starting surface(s), but then ending on only ONE opposed surfaces, allowing a simple projetion. I agree there are severe limitations in the topology to get it meshing by the sweep function, some excercices are often required.

Good luck
Ivar
Hi When I see your jpg, my first question is cannt you transform these cubes into cylinders?, so that you can use 2D-axi and get rif on one DoF. If not, then I would split the parts into 4 symmetric items, using the split lines of the hemisphere. there are then 2 ways out. Either the case is symmetric so one can work with symmetric/anti-symmetric boundaries and mesh only 1/4 of the problem. Or, to work in full 3D. By having split the volumes, it is easier to apply a sweep mesh. By the way a sweep mesh can be defined manually from ONE or SEVERAL starting surface(s), but then ending on only ONE opposed surfaces, allowing a simple projetion. I agree there are severe limitations in the topology to get it meshing by the sweep function, some excercices are often required. Good luck Ivar

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.