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
07.04.2012, 06:17 GMT-4
Hi
in 3.5 you have a rotate mesh function (if I remember right, its not yet in v4 unfortunately) so you can mesh an edge and rotate the edge to get a mapped 2D or quad 3D mesh. Not that in 2Daxi the axis cannot contain a quad/mapped mesh, in V4.2 yes but you will get a region along the axis where you loose the results, this comesfrom a mathematical singualrity 0/0 on the symmetry axis, use tri or thet mesh in contact with the rotation r=0 axis
--
Good luck
Ivar
Hi
in 3.5 you have a rotate mesh function (if I remember right, its not yet in v4 unfortunately) so you can mesh an edge and rotate the edge to get a mapped 2D or quad 3D mesh. Not that in 2Daxi the axis cannot contain a quad/mapped mesh, in V4.2 yes but you will get a region along the axis where you loose the results, this comesfrom a mathematical singualrity 0/0 on the symmetry axis, use tri or thet mesh in contact with the rotation r=0 axis
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
07.04.2012, 07:56 GMT-4
Thanks for ur reply!
I looked into the options. It seems the revolve mesh function cannot generate 2D mesh.
Thanks for ur reply!
I looked into the options. It seems the revolve mesh function cannot generate 2D mesh.
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
07.04.2012, 14:45 GMT-4
Hi
You are probably right (I havent used 3.5 for quite some time), but then you should be able to use a mapped mesh, by first meshing the edges with N elements per rotation, then a 2D mapped mesh
--
Good luck
Ivar
Hi
You are probably right (I havent used 3.5 for quite some time), but then you should be able to use a mapped mesh, by first meshing the edges with N elements per rotation, then a 2D mapped mesh
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
07.04.2012, 15:45 GMT-4
Yes, thanks Ivar.
I've gotten the mesh by mapped mesh.
Yes, thanks Ivar.
I've gotten the mesh by mapped mesh.
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
19.04.2012, 10:09 GMT-4
Hi
I have just learned that in fact the "sweep mesh" of V4.2 can also generate a revolve mesh geometry.
I hadnet tested it out efore, and not thought of it, havnt found it the doc either, but it works (I got the hint from our local COMSOL rep). Just try it out, make a workplane, add an rectangle make a revolve tube, hen mesh one section and sweep it around.
One limitation, but it has a topology expanation, is that a mapped mesh cannot be rotated if the axis of rotation is part of the sweep surface. As a mapped mesh rotated generate a quad and a quad with a common axis means one of the boundaries is singular (the one on the axis).
If the axis is part of your domain you must make an internalsection along the axis, mesh the external part mapped, revolve it, and mesh the internal, on axis domain with a thet, (convert the internal tube boundary first.
--
Good luck
Ivar
Hi
I have just learned that in fact the "sweep mesh" of V4.2 can also generate a revolve mesh geometry.
I hadnet tested it out efore, and not thought of it, havnt found it the doc either, but it works (I got the hint from our local COMSOL rep). Just try it out, make a workplane, add an rectangle make a revolve tube, hen mesh one section and sweep it around.
One limitation, but it has a topology expanation, is that a mapped mesh cannot be rotated if the axis of rotation is part of the sweep surface. As a mapped mesh rotated generate a quad and a quad with a common axis means one of the boundaries is singular (the one on the axis).
If the axis is part of your domain you must make an internalsection along the axis, mesh the external part mapped, revolve it, and mesh the internal, on axis domain with a thet, (convert the internal tube boundary first.
--
Good luck
Ivar