Documente Academic
Documente Profesional
Documente Cultură
Jean-Franois Remacle
Mesh Generation
Curvilinear Meshes
Basic concept
A curvilinear mesh is build using a parametric space
Sponge analogy
Let us consider the real coordonates (x, y) and a parametric space (u, v).
The transformation x = x(u, v) and y = y(u, v) has to be build.
Jean-Franois Remacle
Mesh Generation
Curvilinear Meshes
Jean-Franois Remacle
Mesh Generation
Curvilinear Meshes
Jean-Franois Remacle
Mesh Generation
Curvilinear Meshes
Jean-Franois Remacle
Mesh Generation
Curvilinear Meshes
The transformation transforms a cartesian mesh in the parametric space into
a structured mesh in the real space with the following properties:
Two non intersecting lines in the parametric space do not cross in the
real space,
The mesh adapts to the boundary of on the real space,
The mesh has not to be orthogonal in the real space.
The advantage of curvilinear meshes are numerous. They are usually high
quality/very regular meshes. Yet, the sponge analogy cannot be applied
easily to complex domains.
Jean-Franois Remacle
Mesh Generation
Transfinite Interpolation
Jean-Franois Remacle
Mesh Generation
Transfinite Interpolation
Let us try the following intuitive formula
~S(u, v) = (1 v)~c1 (u) + v~c3 (u) + (1 u)~c4 (v) + u~c2 (v)
Mesh Generation
Transfinite Interpolation
Jean-Franois Remacle
Mesh Generation
Transfinite Interpolation
~S(u, v)
Jean-Franois Remacle
Mesh Generation
Transfinite Interpolation
Jean-Franois Remacle
Mesh Generation
Transfinite Interpolation
Transfinite interpolation exists for 3-sides shapes,
~S(u, v)
Jean-Franois Remacle
Mesh Generation
Elliptic meshes
It is not always simple to devise an explicit formula for the mesh
mapping,
With the transfinite interpolation, mesh lines may cross each others,
leading to a wrong mesh
PDE based mesh have been developped in order to avoid those
difficulties
Two kind of approaches are possible, using either an hyperbolic PDE,
or using an elliptic PDE. This last approach is now studied.
Jean-Franois Remacle
Mesh Generation
"
Jean-Franois Remacle
"
Mesh Generation
Mesh Generation
The maximum principle : any solution of the Laplace equation has its
maximum on the boundaries (no local maxima is allowed),
The solution is unique,
The solution belongs to H 1 , it is smooth,
Iso-contours cannot intersect,
Iso-contours stay inside the domain (obvious for a PDE, not for a
general mapping),
Jean-Franois Remacle
Mesh Generation
The maximum principle : any solution of the Laplace equation has its
maximum on the boundaries (no local maxima is allowed),
The solution is unique,
The solution belongs to H 1 , it is smooth,
Iso-contours cannot intersect,
Iso-contours stay inside the domain (obvious for a PDE, not for a
general mapping),
Jean-Franois Remacle
Mesh Generation
, i = 1, 2.
Boundary
~c1
~c2
~c3
~c4
1
1 (x, y) = 0
n 1 (x, y) = 0
1 (x, y) = 1
n 1 (x, y) = 0
Jean-Franois Remacle
2
n 2 (x, y) = 0
2 (x, y) = 0
n 2 (x, y) = 0
2 (x, y) = 1
Mesh Generation
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
2
=
and
=
xj
j xi
xi xk
j l xk xi
2
j=1
j=1 l=1
In expanded form
2
2
= 2
2
x1
1
2
2
= 2
2
x2
1
1
x1
2
1
x2
2
2
1 2
+2
2
1 2
+2
Jean-Franois Remacle
1 2
x1 x1
1 2
x2 x2
Mesh Generation
2
22
2
22
2
x1
2
2
x2
2
2
21
+
2
22
!
1 2
2
1 2 1 2
1 2
+
+2
+
+
x1
x2
1 2 x1 x1 x2 x2
!
2 2
2 2
+
.
x1
x2
2 x
2 x
2
1
21
2 x2
21
+ 2
x1
1 2
2 x2
1 2
22
2 x2
22
= 0.
with
1 2 1 2
1 2
1 2
2 2
2 2
=
+
, =
+
+
, =
x1
x2
x1 x1 x2 x2
x1
x2
Jean-Franois Remacle
Mesh Generation
"
x1
1
2
+
x1
2
2 #
=
1
x1
2
+
1
x2
2
x1 x2 x1 x2
+
= 1 2 + 1 2,
1 1 2 2
x1 x1 x2 x2
"
#
x2 2
x2 2
2 2
2 2
2
=J
+
+
,
=
1
2
x1
x2
= J 2
with
J = det[j xi ]
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
2f
2
2f
2
2f
fi+1,j fi1,j
2
fi,j+1 fi,j1
2
fi+1,j 2fi,j + fi1,j
2
fi,j+1 2fi,j + fi,j1
2
fi+1,j+1 fi+1,j1 fi1,j+1 + fi1,j1
4
Jean-Franois Remacle
Mesh Generation
2 x1
2 x1
2 x1
2
21
2 + 2
+ 2 12
x2
2 x2
2 x2
2
2
2 2
21 L =
+ 22
+1 2 2
2
= 0.
Their discrete version can be written, for any vertex (i, j) of the mesh in the
(i, j)
parametric space, as
x
Jean-Franois Remacle
#
Mesh Generation
(yi+1,j
(4 )
yi1,j )(yi,j+1 yi,j1 )
(4 )
Jean-Franois Remacle
Mesh Generation
direct methods,
iterative methods.
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
+
"
" xi+1,j 2x
i,j + xi1,j + xi,j+1 2xi,j + xi,j1
!
"
+
2 " xi+1,j+1 xi1,j+1 x+
i+1,j1 + xi1,j1
!
"
!
"
+
+ " yi,j+1 2yi,j
" yi+1,j 2yi,j
+ yi1,j
+ yi,j1
!
"
+
+
+ yi1,j1
2 " yi+1,j+1 yi1,j+1 yi+1,j1
n
y+
Here,
and
are for the values of x and
n +y 1at the previous sweep and x
"
"
"
and y are the values atthe present
sweep. The linearization
x y of the
coefficients consist in using previous values of x and y in 0 , 0 and 0 .
Jean-Franois Remacle
Mesh Generation
(x x)
(y y)
xi,j
= xi,j + Cx i, j/ and y
i,j = yi,j + Cy i, j/.
Jean-Franois Remacle
Mesh Generation
2( + )
Cx i, j
"
"
Rx i, j + " Cx i 1, j
"
jCx i + 1, j 1) + " Cx i, j 1
R
j x+i
"1,
Cxj i1 1,
x i,(C
2(" + " )
Cy i, j
2( + )
Cy i, j
"
=
Ry i," (C
j+
1,1j Cx i + 1, j 1) + " Cx i, j 1
y i j
x iC1,
"
i,
(Cy i "1, j 1 Cy i + 1, j 1) + " Cy i, j 1
R
y j + Cy i 1, j
RR
" [x
2xi,j + xi1,j ] +" " [xi,j+1 2xi,j + xi,j1 ]
x i,i,j j =
= " [yi+1,j
y
i+1,j 2yi,j + yi1,j ] + [yi,j+1 2yi,j + yi,j1 ]
"
yxi1,j+1
xi+1,j1
xi1,j1 ]
2
i+1,j+1
i1,j+1
[yi+1,j+1
yi+1,j1
+ y+
2 " [x
i1,j1 ]
Ry i, j
'
$
%
'
Jean-Franois Remacle
#
$
Mesh Generation
Cx i, j
Cx 1, j
Cx m, j
Cy 1, j
Cy m, j
Jean-Franois Remacle
Mesh Generation
Jean-Franois Remacle
Mesh Generation
j+1
j
valeurs ltape n + 1
j-1
valeurs courantes
valeurs ltape n
i-1
i+1
The resulting system is tridiagonal and ad hoc linear solvers can be used for
accelerating the process.
Jean-Franois Remacle
Mesh Generation
+
!
! x
i+1,j 2xi,j + xi1,j + xi,j+1 2xi,j + xi,j1
!
"
+
+
!
2 xi+1,j+1 xi1,j+1 xi+1,j1 + xi1,j1
!
"
!
"
+
+ ! yi,j+1 2yi,j
! yi+1,j
2yi,j
+ yi1,j
+ yi,j1
!
"
+
+
2 ! yi+1,j+1 yi1,j+1 yi+1,j1 + yi1,j1
Jean-Franois Remacle
Mesh Generation