I've been learning NetGen (an open source mesh generator) and NgSolve (finite element libraries closely linked to Netgen).There is still a lot to cover, I've just started. So far I've gotten them to work in a VC++ project. Well, ngsolve is acting weird, spewing out garbage characters on screen. But at leats it doesn't crash :)
Anyway, I got an idea.
Both Netgen and NGSolve can be compiled as libraries. A DLL wrapper can be created which could be called from a GrassHopper programming module (.NET)
In Grasshopper, the geometry would be generated as a triangulation which would be fed to Netgen to generate a volumetric mesh.
Boundary conditions would be defined in GH and written as a text file. NgSolve would access the geometry or the mesh, the boundary conditions and the PDEs to solve.
The results could be fed back into GH as colored triangles and a stress legend somehow would be created...no idea about that part. Like I said, I didn't get too far. But I think it would be a neat idea to have a completely free FEM solver for GH.
Idea for FEA using Grasshopper and Open Source Tools
by Mihai Pruna
Jan 6, 2014
I've been learning NetGen (an open source mesh generator) and NgSolve (finite element libraries closely linked to Netgen).There is still a lot to cover, I've just started. So far I've gotten them to work in a VC++ project. Well, ngsolve is acting weird, spewing out garbage characters on screen. But at leats it doesn't crash :)
Anyway, I got an idea.
Both Netgen and NGSolve can be compiled as libraries. A DLL wrapper can be created which could be called from a GrassHopper programming module (.NET)
In Grasshopper, the geometry would be generated as a triangulation which would be fed to Netgen to generate a volumetric mesh.
Boundary conditions would be defined in GH and written as a text file. NgSolve would access the geometry or the mesh, the boundary conditions and the PDEs to solve.
The results could be fed back into GH as colored triangles and a stress legend somehow would be created...no idea about that part. Like I said, I didn't get too far. But I think it would be a neat idea to have a completely free FEM solver for GH.
Best regards
Mihai Pruna