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.
how to refer to another point or boundary?
Posted May 5, 2011, 11:24 a.m. EDT 1 Reply
Please login with a confirmed email address before reporting spam
Hi,
I would like to define the boundary condition of a point (corner) on geometry such that to be dependent on the value of another point. I am using Structural Mechanics application, and to my knowledge, this cannot be done in COMSOL's GUI, because the constraints for the nodes should be defined as displacements (Rx & Ry in 2D) and there is no point index to refer to.
So I decided to create the m file and use it in MATLAB. My m file has about 33000 lines, although my geometry is a 4-corner plane solid. I could not figure out how index of a point (e.g. pnt.ind=[1,2,2,1]) can be addressed and used somewhere else in the file. Also, I do not know why some of these command lines are repeated in the m file.
Any hint is appreciated.
Thanks,
Manuel.
I would like to define the boundary condition of a point (corner) on geometry such that to be dependent on the value of another point. I am using Structural Mechanics application, and to my knowledge, this cannot be done in COMSOL's GUI, because the constraints for the nodes should be defined as displacements (Rx & Ry in 2D) and there is no point index to refer to.
So I decided to create the m file and use it in MATLAB. My m file has about 33000 lines, although my geometry is a 4-corner plane solid. I could not figure out how index of a point (e.g. pnt.ind=[1,2,2,1]) can be addressed and used somewhere else in the file. Also, I do not know why some of these command lines are repeated in the m file.
Any hint is appreciated.
Thanks,
Manuel.
1 Reply Last Post May 5, 2011, 3:15 p.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
for me, what I understand is that you are making it rather complicated for you. I do not really understand why you have issus to impose something on a point. At least from the moment the point is defined and part of a geometry object (attached to an edge, belonging to a surface, belonging to a volume (in 3D) it should work.
And do not forget that you can always add a point, and an edges/surface as interiour boundaries, without any great loss of computing time (to many boundaries are etter than too feew ;)
on the other side you an extract data from an arbitrary "point" via the extrusion coupling variables.
Only thing to consider: a point load is a singularity and the FEM is giving wrong results in the direct viciniy of a restricted point (or edge in 3D), eve if often the global solution is valid in its "globality", so long you do not look at the details
--
Good luck
Ivar
for me, what I understand is that you are making it rather complicated for you. I do not really understand why you have issus to impose something on a point. At least from the moment the point is defined and part of a geometry object (attached to an edge, belonging to a surface, belonging to a volume (in 3D) it should work.
And do not forget that you can always add a point, and an edges/surface as interiour boundaries, without any great loss of computing time (to many boundaries are etter than too feew ;)
on the other side you an extract data from an arbitrary "point" via the extrusion coupling variables.
Only thing to consider: a point load is a singularity and the FEM is giving wrong results in the direct viciniy of a restricted point (or edge in 3D), eve if often the global solution is valid in its "globality", so long you do not look at the details
--
Good luck
Ivar