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 The information provided may be out of date.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.
Value of the force density for a element and not for the nodes of a mesh
Posted 2012年8月6日 GMT-4 04:06 Low-Frequency Electromagnetics, Interfacing, Geometry, Mesh, Structural Mechanics Version 4.2, Version 4.2a, Version 4.3 2 Replies
Please login with a confirmed email address before reporting spam
The element : the nodes composed a element, for exemple 8 nodes composed a simple linear element (a little cubic).
The object : is the geometry, is composed by elements.
When Comsol transfers the Lorent's force from EM analysis to a structural analysis, it transfer the value corresponding of the first mesh(EM) to another mesh (Structural), but it should to adapt the values of the nodes(Nodes are usually placed at the element corners) between the two mesh,right?. The transfered value (ex.force density) is transfered to the nodes of the second mesh(Structural) with the help of the shapes fonctions of the second mesh, because the elements was differents. The nodes beetwen the two mesh (EM and Structural) have differents coordinates and form. In others words I should to export mf.FLtrz (I think is a distrubution on the object, right?) for the area (or volume) of a single element of the mesh (Em or Structural,type of element: cubic,thetraedral,..),it is possible?
For exemple in Ansys(another FEM program) we could apply a force on a single element(for exemple a little cubic,thetraedral,... the same of the mesh) and than ansys compute the distrubution of this force on the corresponding element with tha shape fonctions. It is possible in Comsol? If it is possible is valid the reverse -> read a force density (Surface or volumetric) of a element?
thank s
Please login with a confirmed email address before reporting spam
Please login with a confirmed email address before reporting spam
the thing is that COMSOL does not operate exactly that way.
In COMSOL, your elementary item = ENTITY is the analysed Geometry (not the mesh) = a domain or boundary (possible edge and point, depending on spatial dimensions chosen) . meshing and discretization is the math you apply to "digitize" your Entity and to solve your equations. Then COMSOL maps the result back onto the entity = Domain and/or Boundary).
Therefore, if you want to apply a local force, you should either cut your boundary, or domain to make a small "entity" and apply the force thereon, then ensure that the mesh is fine enough to get several elements around your region. or you apply a boundary load multiplied by some shaping function limiting its spatial extent, provided you get the normalisation correct, and that you ensure enough mesh density to correctly resolve your applied load shape
Applying boundary loads to points/vertices is possible but mathematically its a singularity w.r.t. the FEM approach and you get non-physical stress concentrations, locally
In COMSOL I find we distinguish better the physics and the math with a clear separation between them, the mesh is part of the digitizing to apply the FEM math onto the general physics equations, via the geometrical topology
--
Good luck
Ivar