Civil Engineering Reference
In-Depth Information
Figure 5.71
Machine part with many tiny features.
Table 5.4
Statistics of objects meshed by the ADF method
Object 1 2 3 4
NN 548 413 556 815
NB 1128 822 1112 1626
α
min
0.0358 0.147 0.0445 0.00562
α
0.499 0.57 0.63 0.553
NN* 852 499 608 1075
NE 3443 1643 1965 3836
γ
min
0.00111 0.0234 0.00127 0.00038
γ
0.275 0.363 0.243 0.313
N 24 0 6 25
Note: NN = number of nodes on the domain boundary; NB = num-
ber of triangles on th
e
domain boundary;
α
min
= minimum
α
-quality of
boundary triangles;
α
= geometrical mean
α
-quality of boundary
triangles; NN* = number of nodes in the mesh; NE = number of
tetrahedral eleme
n
ts in the mesh;
γ
min
= minimum
γ
-quality of tetra-
hedral elements;
γ
= geometrical mean
γ
-quality of tetrahedral ele-
ments; and N = number of frontal facets required remedial action.
5.6 DELAUNAY-ADF MESHING
DT is known to be robust and fast for the triangulation of a large system of spatial points,
and the AFT is able to keep the boundary intact and generate well-shaped elements in
compliance with the specified nodal space requirements. As shown in Chapters 3 and 4 on
planar domain and for surface meshing, DT and AFT can be merged into a robust scheme
to generate high-quality isotropic and anisotropic meshes meeting the required size and
