t17.py

# ------------------------------------------------------------------------------
#
#  Gmsh Python tutorial 17
#
#  Anisotropic background mesh
#
# ------------------------------------------------------------------------------

# As seen in `t7.py', mesh sizes can be specified very accurately by providing a
# background mesh, i.e., a post-processing view that contains the target mesh
# sizes.

# Here, the background mesh is represented as a metric tensor field defined on a
# square. One should use bamg as 2d mesh generator to enable anisotropic meshes
# in 2D.

import gmsh
import math
import os
import sys

gmsh.initialize()

gmsh.model.add("t17")

# Create a square
gmsh.model.occ.addRectangle(-2, -2, 0, 4, 4)
gmsh.model.occ.synchronize()

# Merge a post-processing view containing the target anisotropic mesh sizes
path = os.path.dirname(os.path.abspath(__file__))
gmsh.merge(os.path.join(path, os.pardir, 't17_bgmesh.pos'))

# Apply the view as the current background mesh
bg_field = gmsh.model.mesh.field.add("PostView")
gmsh.model.mesh.field.setNumber(bg_field, "ViewIndex", 0)
gmsh.model.mesh.field.setAsBackgroundMesh(bg_field)

# Use bamg
gmsh.option.setNumber("Mesh.SmoothRatio", 3)
gmsh.option.setNumber("Mesh.AnisoMax", 1000)
gmsh.option.setNumber("Mesh.Algorithm", 7)

gmsh.model.mesh.generate(2)
gmsh.write("t17.msh")

# Launch the GUI to see the results:
if '-nopopup' not in sys.argv:
    gmsh.fltk.run()

gmsh.finalize()