Hi,
am searching for a solution for scripted meshing of 2d multiregion heat conduction.
What I did so far: Read all the tutorials, read the mailing list two years and reviewed the 40 mesh options in 300 geo files of the gmsh sources such as: gmsh/ -name "*.geo"|xargs grep -o --color "Mesh\.[^ ;=]*"|cut -f2 -d":"|sort|uniq|wc
42
-So concerning the meshing strategy I guess as general rule of thumb
1) one can make points (and less lines) "beam" scriptedly, that is use attractor/treshold fields for each (according to some previously measured object size). This looks feasible.
2)And secondly this should be flanked by:
-loading some postprocess-gradient file from the first solving and
-set the element size according to (all) node gradients as feedback from the solver.
These are rough estimations of a newbie.
-Especially on point 2 I found not really something, which makes me wonder, because it seems "first guess" to me; maybe have overread something or something is odd?
-if this is a common, advisable way to go, would be the general question.
Any opinions/tips/warnings/experiences of yours truly welcome
Tia
(using python, freecad, elmer, ubuntu 14.4, gmsh so far)
_______________________________________________ gmsh mailing list gmsh@geuz.org http://www.geuz.org/mailman/listinfo/gmsh
