Fortran code for projecting a grid on to a surface

AI Thread Summary
The discussion revolves around projecting a grid onto a surface using Fortran code, specifically for triangle and quad meshes. The equation of a sphere is highlighted as a straightforward starting point, with suggestions to create equal interval divisions along latitude and longitude for mesh generation. Element splitting is mentioned as a common method to convert quads into tris while maintaining geometric constraints. Several resources for mesh/grid generation in Fortran are provided, including Geompack, which supports triangle meshes. Additionally, alternatives in MATLAB are suggested for those who may not require a Fortran implementation. There is also a query about Fortran's capabilities for 3D display, with some uncertainty about its integration with 3D graphics.
laminatedevildoll
Messages
211
Reaction score
0
Does anyone know the fortran code for projecting a grid (cartesian or not) onto a surface?

Thnx
 
Computer science news on Phys.org
huh?...do you mean to make your own grid of a finite mesh?
 
yes, so that I can project a triangle mesh on to it.
 
quad mesh is easier..but i don't have it in fortan.
 
think might have somewhere or could find a reference or extract from some code, you want it for linears, quads ... ?
 
PerennialII said:
think might have somewhere or could find a reference or extract from some code, you want it for linears, quads ... ?

yeah, that'd be great.
 
if your willing to do it your self...the equation of a sphere is easy...then you do equal interval divisions along the latitude and longitude. The connecting points for the mesh is an easy pattern for either the quadmesh or trimesh(being alittle complicated then the former).
 
laminatedevildoll said:
yeah, that'd be great.

Ok, I'll look something up.

neurocomp2003 said:
if your willing to do it your self...the equation of a sphere is easy...then you do equal interval divisions along the latitude and longitude. The connecting points for the mesh is an easy pattern for either the quadmesh or trimesh(being alittle complicated then the former).

yeah, it's actually pretty simple and when not making it as general as possible even more so. One way many routines approach it is element splitting, first make quads and split them to tris while preserving/fulfilling geometry etc. restraints (to simplify it further that is).
 
You can find a number of mesh/grid generators from :

http://www-users.informatik.rwth-aachen.de/%7Eroberts/software.html
http://www.andrew.cmu.edu/user/sowen/mesh.html

... for example the Geompack is written in Fortran, source is available and supports triangle meshes :

http://members.allstream.net/~bjoe/

If you don't necessary need it in Fortran there are packages written to Matlab and so, which may be a bit more straightforward.
 
Last edited by a moderator:
  • #10
does fortran now have a 3D display or do you have to convert still to MATLAB or c?
 
  • #11
Been under the impression that Fortran and 3D don't really mix ... although I think for 90/95 there is/are some 3D sets which ougth to enable development using solely fortran :confused: .
 
  • #12
Thank you so much!
 
Back
Top