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
4 feb 2010, 00:37 GMT-5
Hi
from my knowledge there is no direct way (see a couple of other discussion threads). In anycase you must define a segment length, thesselation or finally a "meshing" density to represent you analytical function.
The only workaround I know about is using the ALE to deform a surface. Lets say you start from a cube in 3D, you would want to change the height Z of one face for each X,Y, then you first mesh the system to the desires local element size, you define your function Z=f(x,y) and you impose a deformation of the top surface based on this function. Once solved you save your meshed geometry, and you go on with your physical model.
One point, selecting the mesh finesse: too small it imposes a high mesh density on your problem, too coarse it limits the accuracy of your analytical function, but yoo can always make the mesh finer for your physics.
Hope this helps
Good luck
Ivar
Hi
from my knowledge there is no direct way (see a couple of other discussion threads). In anycase you must define a segment length, thesselation or finally a "meshing" density to represent you analytical function.
The only workaround I know about is using the ALE to deform a surface. Lets say you start from a cube in 3D, you would want to change the height Z of one face for each X,Y, then you first mesh the system to the desires local element size, you define your function Z=f(x,y) and you impose a deformation of the top surface based on this function. Once solved you save your meshed geometry, and you go on with your physical model.
One point, selecting the mesh finesse: too small it imposes a high mesh density on your problem, too coarse it limits the accuracy of your analytical function, but yoo can always make the mesh finer for your physics.
Hope this helps
Good luck
Ivar