Professional Documents
Culture Documents
Lectures On The Finite Element Method: Ph. Ciarlet
Lectures On The Finite Element Method: Ph. Ciarlet
By
Ph. Ciarlet
Lectures on
The Finite Element Method
By
Ph. Ciarlet
Notes by
Printed In India
By
Anil D. Ved At Prabhat Printers, Bombay 400004
And Published By
The Tata Institute of Fundamental Research, Bombay.
Acknowledgements
THE AUTHOR EXPRESSES his deep gratitude to Professor Ramanathan for his very kind invitation to deliver these lectures at Bangalore,
as part of the T.I.F.R.I.I.Sc. Mathematics Programme. Also, the author
tanks the T.I.F.R. for financial support.
Ph. Ciarlet
Preface
OUR BASIC AIM has been to present some of the mathematical aspects of the finite element method, as well as some applications of the
finite element method for solving problems in Elasticity. This is why
important topics, such as curved boundaries, mixed and hybrid methods, time-dependent problems, etc..., are not covered here. No attempt
has been made to give an exhaustive bibliography.
vii
Contents
Acknowledgements
Preface
vii
2 Examples
29
35
53
59
67
8 Numerical Integration
77
95
ix
113
Chapter 1
J(v) =
1
a(v, v) f (v)
2
for all
v V,
(1.4)
2
a(u, v u) f (v u)
for all
v K.
v V a(v, v) 2
(1.6)
f (v) = a( f, v)
for all v V.
J(v) =
1
1
= a(v f, v f ) a( f, f ).
2
2
3
a( f u, v u) 0 for all
v K.
Figure 1.1:
Thus, a( f, v u) a(u, v u) which by virtue of (1.7) is precisely
the relation (1.5). This completes the proof.
We can state the following
Corollary 1.1. (a) If K is a non-empty closed convex cone with vertex at origin 0, then the solution of (P) is characterised by:
a(u, u) = f (u).
a(u, v) = f (v)
for all
v K.
Remark 1.2. The relations (1.5), (1.9) and (1.10) are all called variational formulations of the problem (P).
Proof.
Figure 1.2:
(b) Applying (a) to K, since any subspace is a cone with vertex at 0,
we get (b) immediately. For if v K, then v K and applying
(1.9) both to v and v we get (1.10).
This completes the proof.
Remark 1.3. The solution u of (P) corresponding to f V (for a fixed
a(, )) defines a map V V. Since this solution is the projection of
a(, ) a continuous, bilinear, V-elliptic form, f V . If (P) is the problem: to find u V such that for all v V,
(1.11)
a(u, v) = f (v),
(1.12)
a(v, v) ||v||2 ,
|| Au || = sup
vV
v,0
| Au(v)|
|a(u, v)|
= sup
M||u||.
||v||
||v||
vV
v,0
Au = f.
f (v) = (( f, v)),
(1.16)
7
g(v) = v (Av f )
for
v V.
Then the solution to (P) will be the unique fixed point of this contraction map, which exists by the contraction mapping theorem.
Let v1 , v2 V. Set v = v1 v2 . Then
||g(v1 ) g(v2 )|| = ||(v1 v2 ) A(v1 v2 )||
= ||v Av||.
But,
||v Av||2 = ((v Av, v Av))
= (1 2 + 2 M 2 )||v||2
2
since Av(v) = a(v, v) ||v||2 and ||A|| M. Choosing ]0, 2 [, we
M
get that
(1.18)
1 2 + 2 M 2 < 1
Remark 1.5. The problem (P) of Theorem 1.2 is well-posed. The existence and uniqueness were proved in the theorem. For the continuous
dependence of u on f , we have
(1.19)
Chapter 2
Examples
WE GIVE IN this section several examples of the abstract problem for- 9
mulated in Sec. 1. We interpret the solutions of these problems as solutions of classical boundary value problems which often occur in the
theory of Elasticity.
Before we proceed with the examples, we summarize briefly the
results (without proofs) on Sobolev spaces which will prove to be very
useful in our discussion.
Henceforth Rn will denote an open set (more often will
be a bounded open set with a specific type of boundary which will be
described presently). A multi-index will denote an n-tuple (1 , . . . , n )
of non-negative integers, and we denote
(2.1)
|| = 1 + + n ,
v =
||v
.
x11 . . . xnn
2. Examples
10
where
supp(v) = {x ; v(x) , 0}.
(2.4)
10
||v||m,
21
X Z
=
| v|2 dx ,
v H m ().
21
X Z
=
| v|2 dx ,
v H m ().
||m
|v|m,
||=m
H0m () = D(),
the closure being taken with respect to the topology of H m (). Since
H0m () is a closed subspace of H m (), it is also a Hilbert space under
the restriction of the norm || ||m, . We also have a stronger result:
Theorem 2.1. Assume that is a bounded open set. Then over H0m ()
the semi-norm | |m, is a norm equivalent to the norm || ||m, .
This result is a consequence of the following:
|v|0, C|v|1, .
2. Examples
11
11
||v||L2 () C||v||1, .
By virtue of Theorem 2.3 we get that if v C (), then its restriction to is an element of L2 (). Thus we have a map from the space
C () equipped with the norm || ||1, into the space L2 () which is
continuous. We also have:
Theorem 2.4. The space C () is dense in H 1 (), for domains with
Lipschitz continuous boundaries.
Consequently, the above map may be extended to a continuous map
L2 () which we denote by tr . It is called the trace operator.
An important result on the trace is the characterization:
n
o
(2.11)
H01 () = v H 1 (); tr v = 0 .
H 1 ()
v X v
=
.
i
i=1 xi
n
(2.12)
2. Examples
12
v
L2 (). We now define
xi
xi
v X
v
=
.
i tr
i=1
xi
n
(2.13)
for all 1 i n.
If we assume u H 2 (), we may replace u in (2.15) by
Z
Z
Z X
n
u
u v
dx =
u v dx +
v d,
i=1 xi xi
(2.16)
where =
u
; sumxi
Pn
2
i=1 x2
i
is the Laplacian.
13
for u, v H 2 ().
2. Examples
13
H 2 (),
(2.18)
v
uv dx =
u v dx + u d
(u)
vd.
The formulae (2.15) through (2.18) are all known as Greens formulae in Sobolev spaces.
We derive two results from these formulae. These results will be
useful later.
Lemma 2.1. For all v H02 (),
(2.19)
|v|0, = |v|2, .
Consequently over H02 (), the mapping v 7 |v|0, is a norm equivalent to the norm || ||2, .
Proof. Since D() is dense in H02 (), it suffices to prove (2.19) for
v (). Let v D(). Then
n 2 2
X 2 v 2 v
X
v
2
|v| =
.
2 + 2
xi
x2i x2j
1i< jn
i=1
By Greens formula (2.15),
!2
Z
Z
Z 2 2
2 v
v 3 v
v v
dx =
dx =
dx
(2.20)
2
2
2
xi xi x j
xi x j
xi x j
for, the integrals over vanish for v D(). (cf. (2.11)). Now (2.19)
follows directly from (2.20). This proves the lemma.
Lemma 2.2. Let R2 . Then for u H 3 (), v H 2 (),
Z
2 v
2 u 2 v 2 u 2 v
2 u
2 2 2 2 dx
2
x1 x2 x2 x1
x1 x2 x1 x2
(2.21)
!
Z
2
2
u v
u v
=
2
+
d,
where
14
2. Examples
14
= ( , ),
Proof. Let
1 2 = (1 , 2 ) be the unit vectors along the outer
normal and the tangent respectively. Without loss in generality we may
assume 1 = 2 , 2 = 1 . Also note that 21 + 22 = 1. The second
derivatives occurring in the right hand side are defined by
(2.22)
u 2 u 2
2 u
2 u 2
+
2
1
2
x1 x2
x22 2
2 x21 1
2 u
2 u
2 u
2 u
+
2 2 .
=
(
)
+
2 1
x2 1 1 x1 x2 1 2
x22
1
2 u v
2 u v 2 u v
2 u v
1
2
+
=
x1 x2 x2 x22 x1
2
u v
2 u v
(2.23)
2
+
x1 x2 x1 x21 x2
= X
,
2 u v
2 u v
X
=
x1 x2 x2 x22 x1
(2.24)
2 u v
2 u v
X
=
2
x1 x2 x1 x21 x2
Also note that,
(2.25) div X =
X1 X2
2 v
2 u
2 u 2 v 2 u 2 v
+
=2
2 2 2 2.
x1 x2
x1 x2 x1 x2 x1 x2 x2 x1
15
xi
2. Examples
15
v = (v , . . . , v ),
Hence summing over all i, if
1
n
Z
Z
v dx =
v
d.
(2.26)
div
(This is known as the Gauss Divergence Theorem and also as the Ostrogradskys formula). From (2.23), (2.25) and (2.26) the result follows.
With this background, we proceed to examples of the abstract problem of Sec. 1.
Example 2.1. Let K = V = H01 (). Let a L () such that a 0 a.e.
in . Let f L2 (). Define the bilinear form a(, ) and the functional
f (), by
!
R Pn u v
+ auv dx,
a(u, v) = i=1
xi xi
(2.27)
f (v) =
f v dx.
!2
Z X
n
2
+ av dx
a(v, v) =
i=1 xi
!2
Z X
n
v
dx (since a 0 a.e. in )
i=1 xi
= |v|21, .
16
2. Examples
16
Z X
Z
n
u v
(2.29)
+ auv dx =
f v dx, for all v H01 ().
x
x
i
i
i=1
From the inclusion D() H01 (), we get that u satisfies the equation u + au = f in the sense of distributions.
If we assume that u is sufficiently smooth, for example u H 2 (),
then we may apply Greens formula (2.16), which gives
Z
Z
Z
Z
u
f v dx.
v d =
uv dx +
auv dx
(2.30)
17
Varying v over H01 (), we get that u satisfies the equation u+au =
f in . Further since u H01 (), we get the boundary condition u = 0
on . Thus we may interpret u as the solution of the classical boundary
value problem:
u + au = f in ,
(2.32)
u = 0 on .
This is known as the homogeneous Dirichlet problem for the operator u + au.
A particular case of this equation arises in the theory of Elasticity,
for which R2 and a = 0. Thus u = f in and u = 0 on . This
corresponds to the membrane problem:
Consider an elastic membrane stretched over and kept fixed along
. Let Fdx be the density of force acting on an element dx of . Let
u(x) be the vertical displacement of the point x R2 , measured
2. Examples
17
u = f in
u = 0 on
Figure 2.1:
Remark 2.1. To solve the problem (2.32) by the classical approach, one
needs hard analysis involving Schauders estimates. By the above procedure viz. the variational method, we have got through with it more
easily.
The above problem is a typical example of a second-order problem. 18
Exercise 2.1. The obstacle problem. Let , be as in example (2.1).
Let X be an obstacle in this region. Let X 0 on . Let F dx be the
density of the force acting on a membrane stretched over , fixed along
. The displacement u(x) at x in the vertical direction is the solution of
the following problem:
n
o
If V = H01 (), K = v H01 (); v X a.e. in ,
2. Examples
18
let a(u, v) =
Z
Z X
n
u v
dx, f (v) =
f v dx, f L2 (),
x
x
i
i
i=1
X H 2 (), X 0 on .
Show that K is a closed convex set and hence that this problem admits of a unique solution. Assuming the regularity result u H 2 ()
H01 (), show that this problem solves the classical problem,
u X in ,
u = f when
u = 0
u > X,
( f = F/t)
Figure 2.2:
19
2. Examples
19
u + au = f in .
u = u0 on .
!2
Z X
n
2
+ av dx
a(v, v) =
i=1 xi
min(1, a0 )||v||21, .
for all v H 1 ().
But the left hand side of (2.35) is zero since u Rsatisfies the differential equation as above so that for all v H 1 (), u
v d = 0. Thus
1
As in Example 2.1, the equation u + au = f is always satisfied in the sense of
distributions since D() H 1 ().
2. Examples
20
u
= 0 on , and we may interpret this problem as the classical prob
lem:
u + au = f in
(2.36)
u = 0 on .
Z
Z
f L2 (),
f (v) =
f v dx + gv d,
g L2 ().
21
Show that the abstract problem leads to a solution of the non-homogeneous Neumann problem
u + au = f in
= g on .
Remark 2.2. In these examples one may use the more general bilinear
from defined by
Z X
n
u v
(2.37)
a(u, v) =
+ auv dx,
ai j
x j xi
i, j=1
n
X
i, j=1
ai j i j
n
X
i2
i=1
for all Rn and a.e. in . This is the classical ellipticity condition for
second order partial differential operators. One should check (exercise!)
in this case that the abstract problem leads to a solution of the boundary
value problem
!
n
X
u
ai j
+ au = f in
(2.39)
xi
x j
i, j=1
2. Examples
21
u = 0 on if K = V = H01 (),
n
P
(2.40)
ai j x
i = 0 on if K = V = H 1 ().
j
i, j=1
The latter boundary operator in (2.40) is called the conormal derivative associated with the partial differential operator,
!
n
X
ai j
.
xi
x j
i, j=1
v = (v , v , v ); v H 1 (), 1 i 3 and
v =
K=V=
0 on 0 .
1 2 3
i
Define
(2.41)
22
1 vi v j
i j ( v ) = 2 x + x ,
j
i
v ) + 2 (
v ) = P3 (
i j (
ij
i j v ),
k=1 kk
Z X
3
i, j=1
u (
i j (
i j v )dx
v + 2
( div
3
X
u ) (
i j (
i j v )dx.
i, j=1
g = (g , g , g ), g L2 (), be
Let f = ( f1 , f2 , f3 ), fi L2 (), and
1 2 3
i
given.
2. Examples
22
Define the linear functional f (), by,
Z
Z
g
v d.
(2.43)
f( v ) =
f v dx +
23
Z X
3
i, j=1
v ))2 dx
(i j (
and the fact that the square root of the integral appearing in the right
hand side of the above inequality is a norm over the space V, equivalent
v = (v , v , v ) 7 (P3 ||v ||2 ) 12 . This is a nontrivial fact
to the norm
1 2 3
i=1 i 1,
which uses essentially the fact that 0 has measure > 0 and an inequality
known as Korns inequality. We omit the proof here.
u ,
v ) = f (
v ) admits a unique solution. AsAgain the problem a(
suming sufficient smoothness, we may apply Greens formula:
Z X
3
1
u ) (
i j (
i j v )dx =
2
i, j=1
=
!
vi v j
+
dx
i j ( u )
x j xi
i, j=1
Z X
3
Z X
3
i, j=1
u )
i j (
vi
dx
x j
(i j ( u ))vi dx +
=
i j vi j d.
1 i, j=1
i, j=1 x j
Thus, the abstract problem leads to a solution of
3
x j (i j ( u )) = fi (1 i 3) in
j=1
u =
(2.44)
0 on 0 and
i j ( u ) j = gi on 1 (1 i 3).
j=1
2. Examples
23
u )) =
(i j (
x
x
j
j
j=1
j=1
3
X
j=1
kk ( u )i j + 2i j ( u )
k=1
!!
3
X
ui u j
X uk
i j
2
+
x j
xk
x j
x j xi
j=1
k=1
v ) u .
= ( + )(grad div
i
i
u ( + ) grad div
u =
f in .
Figure 2.3:
24
Z X
3
i, j=1
u ) (
i j (
i j v )dx =
f v dx +
g
v d
24
2. Examples
1
2 a( v , v ) is the strain energy, and the functional f ( v ) is the potential
energy of exterior forces. This example is of fundamental importance in
that the finite element method has been essentially developed for solving
this particular problem or some of its special cases (membranes, plates,
shells, etc.,) and generalizations (nonlinear elasticity, etc. . . ).
2. Examples
25
Remark 2.3. The above problems are all examples of linear problems2 : 25
The map from the right-hand side of the equation and of the boundary
conditions to the solution u is linear. The non-linearity may occur in
three ways:
(i) When K is not a subspace of V. (e.g. Exercises 2.1 and 2.4);
(ii) If in Example 2.3 we have, instead of the first equality in (2.41):
! X
3
vk vk
1 vi v j
.
+
+
i j ( v ) =
2 x j xi
x
x
i
j
k=1
This is the case for instance when one derives the so-called Von
Karmanns equations of a clamped plate;
Show that K is a closed convex cone with vertex 0. Using the results
of Sec. 1 show that the interpretation is
u + au = f in ,
u
u
u 0,
0, u
= 0 on .
uv dx,
a(u, v) =
R
(2.47)
f (v) =
f v dx, f L2 (),
2. Examples
26
26
(by Lemma 2.1. Since | |2, and || ||2, are equivalent on H02 (), the
V-ellipticity follows from (2.48).
Hence there exists a unique function u H02 () such that
Z
Z
f v dx for all v H02 ().
uv dx =
for all v H02 (). Hence by varying v over H02 (), we get that u satisfies
2 u = f in . Since u H02 (), the boundary conditions are given by
(2.14). Thus we interpret this problem as the classical problem
u = f in ,
(2.51)
u =
= 0 on .
27
f (v) = f v dx,
!#
"
R
2 v
2 u 2 v 2 u 2 v
2 u
dx.
u
v
+
(1
)
2
a(u,
v)
=
2
2
22 u 2 v
u v u v
.
(2.53)
uv + (1 ) 2 2 + 2 2 +
x1 x1 x2 x2 x1 x2 x1 x2
2. Examples
27
1
Usually, from physical considerations, 0 < < .
2
Note that
(2.54)
assuming sufficient smoothness of u. Thus again we get the same equation as in (2.51). Notice that the additional term in the definition of a(, )
has contributed nothing towards the differential equation.
This latter problem is known as the clamped plate problem:
Consider a plate of small thickness e lying on the x1 x2 -plane. Let
E be its Youngs modulus and its Poisson coefficient. Let there be a
load F acting on the plate. The displacement u is the solution of (2.51),
where f is given by (cf. Fig. 2.4):
(2.56)
F=
Ee3 f
.
12(1 2 )
Figure 2.4:
We will return to this problem in Sections 10 and 11.
28
28
2. Examples
Show that we may apply the result of Sec. 1 and give an interpretation of this problem.
REFERENCES. For details on Sobolev spaces, see Necas [20] and
Lions and Magenes [17]. For the theory of Elasticity, one may refer to
Duvaut and Lions [10] and Landau and Lipschitz [14].
Chapter 3
29
30
Proof. If wh Vh , then
a(u, wh ) = f (wh ) = a(uh , wh ).
Thus for all wh Vh
a(u uh , wh ) = 0.
(3.2)
= a(u uh , u vh ) + a(u uh , vh uh )
= a(u uh , u vh ) by (3.2)
M||u uh || ||u vh ||.
30
31
We may now describe the finite element method (f.e.m.) in its simplest terms. The method consists in making special choices for the subspaces Vh such that the solutions uh of the problems (Ph ) converge to
u.
We will outline the procedure for obtaining the spaces Vh by consid- 31
ering, for example, a second-order problem.
Let V = H01 () or H 1 (). Let us assume to be a polygonal domain
in Rn . That is, is a polygon in Rn . We then have the following stepby-step procedure:
(i) We first establish a finite triangulation kh of the domain such
S
that =
K. The sets K are called finite elements. If n = 2,
Kkh
Figure 3.1:
(ii) The space Vh is such that for each vh Vh , its restriction vh |K to
each K belongs to some finite-dimensional space Pk of real valued
functions over K which are preassigned. In practice we choose PK
to be a space of polynomials.
(iii) We then need inclusions such as Vh H01 () or H 1 (). We establish a simple criterion to realise this.
32
32
(3.3)
vi dx =
v dx. (1 i n).
xi
v
= vi and hence v H 1 ().
xi
(v|K)
L2 (K) for
However, v|K PK H 1 (K) implies that
xi
1 i n. Let D(). Since the boundary K of any K of the triangulation is Lipschitz continuous, we apply the Greens formula (2.15)
to get
Z
Z
Z
(v|K)
dx = (v|K) dx +
(v|K)i,K dK ,
(3.4)
xi
K
K
K xi
Then it will follow that
= ( , . . . , ) is the outer
where dK is the measure on K and
K
1,K
n,K
normal on K. Summing over all the finite elements K, we get
Z
X Z (v|K)
vi dx =
dx
xi
Kkh K
Z
(3.5)
XZ
v dx +
(v|K)i,K dK ,
=
x
i
K
Kk
h
(v|K)
.
xi
The summation on the right-hand side of the above equation is zero
for the following reasons:
On the boundary , since D(), the integral corresponding to
K is zero. So the problem, if any, is only on the other portions
of the boundary of each K. However, these always occur as common
boundaries of adjacent finite elements. The value of v|K on the common
boundary of two adjacent finite elements is the same (Vh C 0 ()). But
where vi is the function whose restriction to each K is
33
33
the outer normals are equal and opposite from orientation considerations. (See Fig. 3.2).
Figure 3.2:
Hence the contributions from each K along the common boundaries
cancel one another. Thus the summation yields only zero. Hence vi
satisfies (3.3) for 1 i n, and clearly vi L2 (). The last part of the
theorem follows the characterization (2.11).
Exercise 3.1. If for all K kh , PK H 2 (K) and Vh C 1 (), then
v
show that Vh H 2 (). Also if v =
= 0 on , for all v Vh , then
Vh H02 ().
We finally describe the system of linear equations associated with
the space Vh . Suppose {w j ; 1 j M} is a basis for Vh . Let uh be the
solution of (Ph ). If uh is given by
(3.6)
uh =
M
X
u jw j.
j=1
M
X
j=1
a(w j , wi )u j = f (wi ),
1 i M.
34
Note that the symmetry of a(, ) implies the symmetry of the matrix
and the V-ellipticity says that the matrix is positive definite. In practical
computations these observations are important.
Since we have to handle the matrix of the system, it would be ideal
of course to have a diagonal matrix. We could in principle achieve this
through a Gram-Schmidt orthogonalisation procedure applied to the basis functions. However such a process is not feasible since it is highly
numerically unstable. So the best we may hope for is a matrix with a
lot of zeros in it - what is known as a sparse matrix.
For example in the problem given by
u + au = f in
u = 0 on
the (i, j)-coefficient of the matrix is
Z X
n w j wi
+ aw j wi dx.
(3.8)
a(w j , wi ) =
k=1 xk xk
The matrix will be sparse if the supports of the basis functions are
as small as possible so that their inner-products will be most often
zero. We will study subsequently methods to achieve this. This trivial
criterion extends, of course, to all types of problems.
Chapter 4
36
36
1
1 ...
1
then det A , 0.
The above definition generalises the notion of a triangle to n dimensions. Geometrically the condition det A , 0 simply means that the
points {a j }n+1
j=1 do not lie in the same hyperplane. For det A is equal to,
by elementary column operations, the determinant of the matrix
..
..
.
.
and that this is non-zero means that (a1 an+1 ), . . . , (an an+1 ) are
linearly independent vectors in Rn , which is the same as saying that
a1 , . . . , an+1 do not lie in the same hyperplane.
n
Definition 4.2. Let {a j }n+1
j=1 be (n + 1)-points in R satisfying the conditions of definition 4.1. The barycentric coordinates of any x Rn with
respect to these points are numbers { j }n+1
j=1 such that
n+1
P
x=
ja j,
j=1
(4.2)
n+1
j.
1 =
j=1
The barycentric coordinates exist because they are merely the com
A =
where x = ...
1
xn
37
n+1
n+1
X
X
K=
x
=
a
;
0
1;
=
1
.
j j
j
j
j=1
j=1
P
Let K be an n-simplex. Let PK = P1 . We define a set K =
{p(ai ); 1 i n + 1} of degrees of freedom for p PK , where {ai }n+1
i=1
P
are the vertices of K: The set K determines every polynomial p
PK uniquely. For, note that dim PK = dim P1 = n + 1. Consider
1 , . . . , n+1 P1 , the barycentric coordinate functions. These are linP
early independent since k k = 0 implies that its value at each vertex
is zero. Since k (a j ) = k j we get that j = 0 for all j. Thus these
functions form a basis for P1 . Let us write
p=
n+1
X
i i .
i=1
Then
p(a j ) =
n+1
X
i i (a j ) =
i=1
n+1
X
i=1
Thus,
(4.4)
p=
n+1
X
i=1
p(ai )i .
i i j = j .
38
Figure 4.1:
P
Let PK = P2 . We define for p P2 , the set K = {p(ai ), 1 i
P
n + 1; p(ai j ), 1 i < j n + 1} of degrees of freedom. Again
n+2 K
determines p P2 completely. To see this note that dim PK = 2 and
there are as many functions in the set {i (2i 1), 1 i n+ 1; i j , 1
i j n + 1}. There are all functions in P2 . Further since
2 if i = k or j,
i (a j ) = i j , i (ak j ) =
0 otherwise,
we see again that these are linearly independent in P2 . Let us write
p=
n+1
X
i=1
i i (2i 1) +
Then
p(ak ) =
n+1
X
i=1
i j i j .
1i< jn+1
i ik (2ik 1) = k .
Further,
p(akl ) =
n
X
i=1
39
i j i (akl ) j (akl ).
1i< jn+1
n
P
is zero. Further,
39
i=1
0 otherwise.
p=
n+1
X
i=1
i (2i 1)p(ai ) +
4i j p(ai j ).
1i< jn+1
2ai + a j
,
3
Figure 4.2:
Set PK = P3 . Define the set of degrees of freedom
X n
= p(ai ), 1 i n + 1; p(aii j ), 1 i , j n + 1;
K
o
p(ai jk ), 1 i < j < k a + 1 .
40
2
1
Note that i (aii j ) = ; j (aii j ) = ; 1 (aii j ) = 0 if 1 , i, 1 , j;
3
3
1
if 1 = i, j or k and 0 otherwise, etc. Using these, one
1 (ai jk ) =
3
checks the linear independence of the functions
n
i (3i 1)(3i 2), 1 i n + 1; i j (3i 1), 1 i , j n + 1;
o
i j k 1 i < j < k n + 1 .
These then form a basis for P3 , for there are as many functions in the
above collection as dim PK . Using the values of i at the special points
described above, we get
n+1
X
i (3i 1)(3i 2)
p(ai )
p=
2
i=1
X 9
+
i j (3i 1)p(aii j )
2
1i, jn+1
X
27i j k p(ai jk ).
(4.6)
1i< j<kn+1
40
Thus
completely determines p P3 .
P
The points of K at which the polynomials are evaluated to get K
P
are known as the nodes of the finite element. The set K is the set of
degrees of freedom of the finite element.
Exercise 4.1. Generalize these ideas and describe the n-simplex of type
(k) for any integer k 1.
We now show how these finite elements may be used to define the
space Vh .
First of all we show the inclusion Vh C 0 (). Consider for instance
a triangulation by n-simplices of type (1). Number all the nodes of the
P
triangulation by {b j }. Let us define h = {p(b j ); b j is a node.}: This is
the set of degrees of freedom of the space Vh : A function v in the space
Vh is, by definition, determined over each K kh by the values v(b j ) for
those nodes b j which belong to K.
41
Figure 4.3:
41
42
i1
in
Qk =
p;
p(x)
=
a
x
.
.
.
x
.
i1 ,...in 1
n
0i j k
1 jn
42
Figure 4.4:
P
To show that K indeed determines p Q1 uniquely we adopt a
different method now. (There are essentially two methods to show that
P
K completely determines P K ; the first was used in the previous examples where we exhibited a basis for PK such that the corresponding
P
coefficients in the expansion of p in terms of this basis came from K ;
the second is illustrated now).
P
Observe first that dim PK = card K = 2n . To determine a polynoP
mial completely in terms of the elements of K we must solve 2n linear
43
Figure 4.5:
Here again one can prove the unisolvency as above. Now let p Q2
be given such that p(ai ) = 0 for all 1 i 9; then p = 0 on the four
1
It is not necessary to restrict ourselves to a square. Any rectangle with sides parallel
to the coordinate axes would do.
44
44
sides and on the two central (dotted, in Fig. 4.5) lines. Now take lines
parallel to one of the axis and p vanishes on each of these. Thus p 0
P
on K and we get that K uniquely determines p Q2 .
Exercise 4.3. Describe the rectangle of type (3) and generalize to hyperrectangles of type (k).
4
8
X
X
PK =
p
Q
;
4p(a
)
+
p(a
)
2
p(a
)
=
0
.
2
9
i
i
i=1
i=5
and that P2 PK .
3
X
(23i + 32i 71 2 3 )p(ai ) + 271 2 3 p(a123 )
i=1
3 X
X
i=1 j=1
j,i
P
Thus, K = {p(ai ), 1 i 3; Dp(ai )(a j ai ), 1 i , j 3; p(a123 )}
is the corresponding set of degrees of freedom.
45
X
K
)
p
p
= p(ai ), 1 i 3; p(a123 );
(ai ),
(ai ), 1 i 3 .
x1
x2
Figure 4.6:
Exercise 4.6. The Hermite Triangle of Type (3 ).
This is also known
( as the Zienkiewicz triangle in) Engineering litP
P
p
p
erature. Set K = p(ai ),
(ai ),
(ai ), 1 i 3 . Show that K
x1
x2
uniquely determines a function in the space
3
3
X
X
.
p
P
;
6p(a
)
2
p(a
)
+
Dp(a
)(a
a
)
=
0
PK =
3
123
i
i
i
123
i=1
i=1
46
p
2 p
p(a
),
(a
),
.
.
.
,
(ai ), 1 i 3;
=
i
i
x1
x22
K
)
p
(ai j ), 1 i j 3 .
Figure 4.7:
47
p
is indicated by a line
The knowledge of the normal derivative
47
!
b + b
dr
dr
2 v
v
v
2 v
2 v
v
(b
),
(b
),
(b
),
(b
),
(b
),
(c
)
=
v(b
),
j
j
j
j
j
k
j
2
2
x1
x2
x1 x2
k
x
x
h
48
agreeing in their values with first derivatives at end points and agree at
(v|K1 ) (v|K2 )
the mid-point in their values. Thus
=
on K . Similarly,
(v|K1 ) (v|K2 )
=
on K and hence v C 1 (). Thus Vh C 1 ().
t
t
Figure 4.8:
Exercise 4.7. The 18-Degree-of-Freedom-Triangle
Let K be a triangle in R2 . Let PK consist of those polynomials of
degree 5 for which, along each side of K, the normal derivative is
a polynomial of degree 3, in one variable of course. Show that a
polynomial in PK is uniquely determined by the following set of degrees
of freedom:
2 p
p
(a
),
.
.
.
,
(a
),
1
3
=
p(a
),
i
i
i
.
x
x
K
49
49
Figure 4.9:
Show that
uniquely determines p PK .
p1
p2
p3
(a) =
(a) =
(a) for i = 1, 2, gives 4 more conditions.
xi
xi
x1
50
50
(iv)
p2
p1
=
along a1 a and two more similar conditions give 3
conditions.
Figure 4.10:
Let PK = Q3 , the degrees of freedom being given by
)
X (
p
2 p
p
(ai ),
(ai ),
(ai ), 1 i 4 .
= p(ai ),
x1
x2
x1 x2
K
P
Show that K determines uniquely a polynomial p Q3 (a double
dotted arrow indicates that the mixed second derivative is a degree of
freedom). Show also that in this case Vh C 1 ().
51
51
support of basis
function associated
with node
support of basis
function associated
with node
Figure 4.11:
with the degree of freedom.
Chapter 5
i (p) = i , 1 i N.
54
p P we may write
(5.2)
p=
N
X
i (p)pi .
i=1
53
P
P
Instead of (K, , P) one writes at times (K, K , PK ) for the finite
element.
In the various examples we cited in Sec. 4 our set of degrees of freedom for a finite element K (which was an n-simplex or hyper-rectangle)
had elements of the following type:
Type 1: 0i given by p 7 p(a0i ). The points {a0i } were the vertices, the
mid-points of sides, etc...
1 ). For instance, in the Hermite
Type 2: 1i,k given by p 7 Dp(a1i )(i,k
1 = a a ,
triangle of type (3) (cf. Example 4.6), we had a1i = ai , i,k
i
k
where a1 , a2 , a3 were the vertices.
2 , 2 ). For example, in the 18Type 3: 2i,kl given by p 7 D2 p(a2i )(i,k
i,l
2 = e = 2 , the unit vecdegree-of-freedom triangle, a2i = ai , i,k
1
i,1
2 p
(ai )
tor in the x1 -direction so that we have D2 p(ai )(e1 , e1 ) =
x21
as a degree of freedom. (cf. Exercise 4.7).
In all these cases the points {ais } for s = 0, 1 and 2, are points of K
and are called the nodes of the finite element.
Definition 5.2. A finite element is called a Lagrange finite element if its
degrees of freedom are only of Type 1. Otherwise it is called a Hermite
finite element. (cf. Remark 4.2)
Let (K, , P) be a finite element and v : K R be a smooth
function on K. Then by virtue of the P-unisolvency of , there exists a
unique element, say, v P such that i (v) = i (v) for all 1 i N,
N . The function v is called the P-interpolate function
where = {i }i=1
of v and the operator : C (K) P is called the P-interpolation
55
(5.3)
N
X
i ()pi .
i=1
(5.4)
3
X
v(ai )i .
i=1
P)
,
as fixed in the
In as much as we consider the finite element (K,
sequel, it will be called the reference finite element of the family.
Given any finite element K with vertices a1 , a2 , a3 in this family,
there exists a unique invertible affine transformation of R2 i.e. of the
56
55
and
PK = {p : K R; p = p F K1 , p P}.
(ii) K = F(K),
(iii) ais = F(ais ), s = 0, 1, 2,
1 = B 1 , 2 = B 2 , 2 = B 2
(iv) i,k
i,k i,k
i,k i,l
i,l
and
57
Let us see why the relations given by (iv) must be precisely of that
form. We have by (v), p(x) = p(
x). This must be valid when we use the
basis functions as well. We have:
X
X
1
) p 1i,k ( x)
p ( x) =
p (a0i ) p 0i ( x) +
D p(
a1i )(i,k
i
i,k
2 2
, i,l ) p 2i,kl ( x).
p (a2i )(i,k
i,k,l
by (iv).
i,k
i,k,l
i (v)
b = i (v) = i (v), 1 i N. i (v ) = i (v) = i (v), 1 i N.
Thus, i (v ) = i (v)
b for 1 i N. Hence v = v
b by uniqueness
of the function v .
57
Let us consider a polygonal domain with a triangulation th . Suppose to each K th is associated a finite element, (K, K , PK ), K being
the set of degrees of freedom, and PK the finite dimensional space such
that K is PK -unisolvent. Then we have defined the interpolation operator K . All these make sense locally i.e. at a particular finite element K.
We now define the global counterparts of these terms. The comparison
is given in the following table.
58
Table 5.1.
Local definition
1.
Finite element K.
1.
Global definition
S
The set =
K
K=tn
2.
3.
4.
5.
6.
7.
58
The boundary of K, K.
The space PK of functions
K R, which is finitedimensional.
P
N of
The set K = {i,K }i=1
degrees of freedom of K.
2.
3.
5.
K is the PK -interpolation
operator,
defined
by
i,K (K v) = i,K (v), for all
P
i,K K .
(h v)|K = K (v|K)
4.
6.
The boundary of , = .
The space Vh of functions
R, which is also finitedimensional.
The set of degrees of freeP
N , where
dom h = {i }i=1
i (p|K) = i,K (p|K).
Basis function of Vh are {w j }.
The nodes of th are by defS
{Nodes of K} =
inition,
Kth
7.
{b j } say.
The Vh interpolation operator h is defined by h v
Vh such that, i,K (h v|K) =
P
i,K (v|K) for all i,K K .
for all
K th .
Chapter 6
Interpolation Theory in
Sobolev Spaces
WE OUTLINED THE internal approximation method in Sec. 3. We are 59
naturally interested in the convergence of the solutions uh Vh to the
global solution u V. As a key step in this analysis we obtained the
error estimate (cf. Theorem 3.1):
(6.1)
u + au = f in ,
(6.2)
u = 0 on .
Let th be a triangulation of by triangles of type (1), (2) or (3).
Then uh Vh H01 () and (6.1) reads as
(6.3)
We know a priori that u H01 (). Let us assume for the moment
that u C 0 (). (Such assumptions are made possible by the various
59
60
X
2
||u h u||1,K
= C
Kth
12
X
= C
||u K u||21,K
Kth
60
|| m}.
| v| dx
(6.4)
||v||m,p, =
||m
(6.5)
|v|m,p, =
||=m|
1/p
| v| p dx
61
and
61
|v|k+1,p, = |v|k+1,p, .
(6.7)
pPk
62
63
Proof. Let {e1 , . . . , en } be the standard basis for Rn . Let be a multiindex with || = m. By choosing a suitable collection {e1 , . . . , em }
with appropriate number of repetitions from the basis, we may write,
( v )( x) = (Dm v )( x)(e1 , . . . , em ),
where Dm v is the mth order Frechet derivative of v and Dm v ( x) is consequently an m-linear form on Rn . Thus,
| v ( x)| ||Dm v ( x)|| = sup |Dm v ( x)(1 , . . . , m )|.
||i ||=1
1im
(6.13)
||D v ( x)|| d x
!1/p
C2 |v|m,p, .
This gives
64
(6.15)
64
=
p
C1 ||B||mp | det B|1
||Dm v(x)|| p dx
and
||B1 || h/.
||||=
such that = y z. Then
Let Rn with |||| = .
Choose y , z
B = By Bz = y z, where F(y) = y, F(z) = z. But y, z and hence
||y z|| h. Thus ||B|| h. Hence ||B|| h/,
which completes the
proof.
We conclude this section with an important, often used, result.
P)
,
be a finite element. Let s(= 0, 1 or 2) be
Theorem 6.6. Let (K,
Assume that:
the maximal order of derivatives occurring in .
65
C s (K)
P)
,
such that for all affine
Then there exists a constant C = C(K,
equivalent finite elements (K, , P) we have
(6.16)
|v K v|m,q,K C(meas K) q p
hk+1
K
|v|k+1,p,K
m
K
65
(6.17)
i,k
X
2 2
= + (D2 v (a2i )(i,k
, i,l )) p 2i,k,l ,
i,k,l
all these sums being finite (the second and third may or may not be
W m,q (K)).
(6.18)
X
i,k
X
i,k,l
1
|Dv(a1i )(i,k
)| || p 1ik ||m,q,K
2 2
|D2 v (a2i )(i,k
, i,l )| || p 2ikl ||m,q,K
C s (K)
and all the numbers v (a0 ), etc. . . , are
Since W k+l,p (K)
i
Hence the claim is valid. Now by virtue of (ii) and also our observation on preservation of polynomials, we may apply theorem 6.3 to .
P)
,
such that
Hence there exists C = C(K,
|v v |m,q,K C|v|k+1,p,K
for
v W k+1,p (K).
Notice that v = d
v = v d
K v. Thus
K v by Theorem 5.1. Thus v
= K where F K ( x) = BK x + bK , we get
if F K (K)
(6.19)
m
1/q
|v K v|m,q,K C1 ||B1
|v v |m,q,K ,
K || | det B K |
66
and ||BK ||
| det BK | =
, ||B1
K || h/K by theorem 6.5.
meas K
,
Since h,
meas K are constants, combining (6.19), (6.20) and the preceding observation we complete the proof of the theorem.
References: See Bramble and Hilbert [28], Bramble and Zlqmal [2],
sek
Ciarlet and Raviart [7], Ciarlet and Wagschal [8], Strang [21], Zeni
[23, 24] and Zlamal [25, 32].
Chapter 7
Applications to
Second-Order Problems
Over Polygonal Domains
WE APPLY THE results of the preceding section in studying the conver- 67
gence of the finite element method, i.e. the convergence of the solutions
uh of (Ph ) to the solution u of a problem (P) which corresponds to the
choice V = H 1 () or H01 (), which we saw in Sec. 2 led to second-order
problems.
Let be a polygonal domain throughout.
Definition 7.1. A family (th ) of triangulations of is regular is
(i) for all th and for each K th , the finite elements (K, , P) are
P)
,
called the
all affine equivalent to a single finite element, (K,
reference finite element of the family;
(ii) there exists a constant such that for all th and for each K th
we have
(7.1)
hK
68
(7.2)
Kth
then h 0.
Remark 7.1. The condition (ii) in definition 7.1 assures us that as h 0
the triangles do not become flat; cf. Exercise 7.1.
68
Exercise 7.1. If n = 2 and the sets K are triangles, show that condition
(ii) of definition 7.1 is valid if and only if there exists 0 > 0 such that
for all th and K th , K 0 > 0, K being the smallest angle in K.
Exercise 7.2. Consider the space Vh associate with th . Since Vh is finite
dimensional all norms are equivalent and hence
|vh |0,, Ch |vh |0,
for all
vh Vh ,
|vh |0,,
C
|vh |0,
hn/2
for
vh Vh .
|vh |1,
C
|vh |0,
h
for all
vh Vh .
We now obtain an estimate for the error ||u uh ||1, when the family
of triangulations is regular, which also gives convergence.
Theorem 7.1. Let (th ) be a regular family of triangulations on of
P).
,
We
class C 0 (i.e. Vh C 0 ()) with reference finite element (K,
assume that there exists an integer k 1 such that
(i) PK P H 1 (K)
C s (K)
where s(= 0, 1, or 2) is the maximal order of
(ii) H k+1 (K)
P
derivatives in .
69
(7.5)
H k+1 (K)
H 1 (K)
(k 1),
H k+1 (K)
Pk P H 1 (K),
and we may apply Theorem 6.6 with p = q = 2, m = 1 to get
(7.6)
|u K u|1,K C|u|k+1,K
hk+1
K
K
C|u|k+1,K hkK
(Since
hK
).
K
Now since hK h,
||u h u||1,
21
X
2
||u K u||1,K
=
Kth
21
X
|u|2k+1,K
C hk
Kth
= C h |u|k+1, .
This completes the proof.
70
(7.9)
h0
C(meas K) 2 h|v|2,, .
Summing over K, we get
||v h v||1,
12
X
meas K
C h|v|2,,
Kth
= C h|v|2,, ,
since
(7.10)
Kth
h0
71
71
(7.10) choose h0 such that for all h h0 , ||v0 h v0 ||1, /2. Now, by
(6.1)
||u uh ||1, C||u h v0 ||1,
C, for h h0 .
(i) V H,
(ii) V = H.
and
(P)
a(u, v) = f (v)
(Ph )
and that the assumptions on (P) are as in the Lax-Milgram lemma. Then
we have the following theorem.
72
Theorem 7.3. Let the spaces H and V satisfy (7.11). Then with our
above mentioned notations,
)
(
1
inf || h || ,
(7.12)
|u uh | M||u uh || sup
gH |g| h Vh
a(v, ) = (g, v)
for all v V,
(7.13)
gH
g,0
|(g, u uh )|
.
|g|
For a given g H,
(7.14)
73
(g, u uh ) = a(u uh , )
Also if h Vh we have,
(7.15)
a(u uh , h ) = 0.
(g, u uh ) = a(u uh , h ),
which gives us
|(g, u uh )| M||u uh || || h ||,
73
!
|| h ||
|u uh | M||u uh | sup
.
|g|
gH
g,0
by (7.13). Since this is true for any h Vh we may take infimum over
Vh to get (7.12), which completes the proof.
For dimensions 3 and Lagrange finite elements we now show that
|u uh |0, = 0(hk+1 ). For this we need one more definition.
Definition 7.2. Let V = H 1 () or H01 (), H = L2 (). The adjoint
problem is said to be regular if the following hold:
(i) for all g L2 (), the solution of the adjoint problem for g
belongs to H 2 () V;
(ii) there exists a constant C such that for all g L2 ()
||||2, C|g|0, ,
(7.17)
h Vh
74
By (7.12) and (7.19).
gL2 ()
!
1
Ch||2, .
|g|0,
By the regularity of (P ),
||2, ||||2,
constant.
|g|0,
|g|0,
(7.20)
0<
hK
, f rm[o],
h
being a constant.
L (K).
If (P ) is regular,
Let u H 2 () and P1 P H 1 (K)
then there exists a constant C independent of h such that
|u uh |0,, C h|u|2, ; if n = 2
(7.22)
|u uh |0,, C h|u|2, if n = 3.
|uh h u|0,,
C
|uh h u|0, .
h
Thus,
|uh h u|0,,
C
|uh u|0, + |u h u|0,
h
75
i
Ch
C1 h2 |u|2, + C2 h2 |u|2,
h
C h|u|2, (by Theorem 7.4 and Theorem 6.6).
Thus,
Also H 2 () C 0 (); H 2 () L () and P1 P L (K).
by Theorem 6.6 with k = 1, p = 2, m = 0, q = ,
1
C2 2
C 2
h
h
2 K
2
76
For n = 3, the only variation in the proof occurs in the fact that
|uh h u|0,,
C
|uh h u|0,
h3/2
C
C3 3
3
h .
K
3
3
References: One may refer to Ciarlet and Raviart [6] for 0(h) convergence in the norm | |0,, for any n. See also Bramble and Thomee
[1].
Chapter 8
Numerical Integration
LET US START with a specific problem. Let be a polygonal domain 77
in Rn . Consider the problem
(8.1)
n
P u
xi ai j x j = f in ,
i, j=1
u = 0 on = .
where the (ai j ) and f are functions over which are smooth enough.
Let us further assume that there exists > 0 such that, for all Rn ,
(8.2)
n
X
i, j=1
ai j i j
n
X
i2 .
i=1
It has been seen earlier (cf. Remark 2.2) that the above problem (8.1)
is obtained from a problem (P) with a(, ) and f () being defined by
(8.3)
R Pn
u v
a(u, v) = i, j=1 ai j x j xi dx
R
f (v) =
f v dx
8. Numerical Integration
78
P),
,
we get the problems (Ph ), i.e., to find uh Vh such
element (K,
that
(8.4)
for all vh Vh .
a(uh , vh ) = f (vh ),
(8.5)
uh =
M
X
uk wk .
k=1
M
X
k=1
78
Notice that
a(wk , wm ) =
(8.7)
f (wm ) =
n
XZ X
Kth
K i, j=1
Kth
XZ
ai j
wk wm
dx
x j xi
f wm dx.
the functions and being in the usual correspondence. We then replace the expression in the right-hand side by the following:
(8.9)
b
(
x)dx
L
X
1=1
1 (
b 1 ).
8. Numerical Integration
79
In this section will denote the right-hand side replacing the expression in the left hand side in similar relations). In (8.9) the quantities
1 are called the weights and the points b 1 are called the nodes
of the quadrature scheme. While in general we may assume
1 R
and b 1 Rn , we will restrict ourselves to the most common case where
1 l L.
1 > 0 and b 1 K,
We may now define the error functional E by
(8.10)
E ()
=
(
x)d x
L
X
1 (
b 1 ).
1=1
79
1,
1,K = (det BK )
(8.11)
b1,k = F K (b 1 ),
(x)dx
L
X
1,K (b1,K ).
1=1
l=1
EK () = (det BK )E ().
8. Numerical Integration
80
(
ai j ).
(
x)d x (meas K)
3
K
1i< j3
80
Z
3
X
X
1
.
3
(
a
)
+
8
(
a
)
+
27
(
a
)
(
x)d x (meas K)
i
i
j
60
K
i=1
1i< j3
Figure 8.1:
Exercise 8.3. Show that E ()
= 0 for P3 , in Example 8.3.
Let us now review the whole situation. We had the original approximation problem (Ph ): To find uh Vh such that a(uh , vh ) = f (vh )
for all vh Vh .
This led to the solution of the linear system (8.6). By virtue of the
quadrature scheme we arrive at a solution of a modified approximation
problem (Ph ): To solve the linear system
(8.15)
M
X
k=1
8. Numerical Integration
where
(8.16)
81
n
P PL
P
wk wm
ah (wk , wm ) =
ai j x j xi (b1,K )
l=1 1,K
Kth
i, j=1 !
L
P P
1,K ( f wm )(b1,K ) .
fh (wm ) =
Kth l=1
(8.17)
M
X
uk wk .
k=1
Thus the problem (Ph ) (not to be confused with any adjoint problem!) consists in finding uh Vh such that, for all wh Vh ,
81
ah (uh , wh ) = fh (wh )
(8.18)
where
(8.19)
L
n
P P
P
vh wh
ah (vh , wh ) =
ai j x j xi (b1,K )
1,K
Kth l=1
i, j=1
L
P P
1,K ( f vh )(b1,K ),
fh (vh ) =
Kth l=1
for vh , wh Vh .
8. Numerical Integration
82
(i) What are sufficient conditions such that (Ph ) have unique solutions?
(ii) Can we find an abstract error estimate for ||u uh ||?
(iii) If ||u uh || = 0(hk ), i.e., without numerical quadrature, under what
conditions is this order of convergence preserved, i.e. when can
we say ||u uh || = 0(hk )?
82
(8.20)
(8.21)
C inf
vh Vh
||u vh || + sup
wh Vh
)
!
| f (wh ) fh (wh )|
|a(vh , wh ) ah (vh , wh )|
+ sup
.
||wh ||
||wh ||
wh Vh
8. Numerical Integration
83
a(u, uh vh ) = f (uh vh ),
83
(8.22)
+ |a(vh , uh vh ) ah (vh , uh vh )|
Thus,
e
||uh
|a(vh , uh vh ) ah (vh , uh vh )|
vh || M||u vh || +
||uh vh ||
| fh (uh vh ) f (uh vh )|
+
||uh vh ||
|a(vh , wh ) ah (vh , wh )|
M||u vh || + sup
||wh ||
wh Vh
| f (wh ) fh (wh )|
+ sup
||wh ||
wh Vh
e
wh Vh
||wh ||
| f (wh ) fh (wh )|
1
.
+ sup
e
wh Vh
||wh ||
Varying vh over Vh and taking the infimum, and replacing (1+ M/e
), 84
(1/e
) by a larger constant C, we get (8.21), which completes the proof.
8. Numerical Integration
84
The following theorem tells us when the uniform Vh -ellipticity assumption of Theorem 8.1 is satisfied in the example we started with.
Theorem 8.2. Let ah (, ) and fh () be as in (8.19). Assume further that
L
S
(i)
1 > 0, 1 l L, (ii) P Pk , and (iii) {b l } contains a Pk 1
l=1
n
L
X
XX
v
v
h
h
(b )
ai j
ah (vh , vh ) =
1,K
l,K
x
x
j
i
i, j=1
Kth l=1
(8.23)
n
!2
L
XX
X pK
l,K
(bl.K )
xi
Kt l=1
i=1
h
n
X
b jl xl + bK, j .
l=1
Then
p K X pK (x) x j X pK (x)
=
=
b ji .
xi
x j xi
x j
j=1
j=1
n
Thus is
p K ( x)
p K ( x)
,...,
D =
x1
xn
and
D=
!
pK (x)
pK (x)
,...,
,
x1
xn
8. Numerical Integration
85
85
(8.25)
l=1
!2
n
X
p
(b1 ) = 0.
1
xi
i=1
p
(bl ) = 0 for all 1 l L and 1 i
xi
p
p
Pk 1 and hence
= 0 by the Pk 1 n. Since P Pk , we have
xi
xi
0
unisolvency. Thus p P0 , and on the finite dimensional space P/P
we have a norm defined by the
(in practice we always have P0 P)
square-root of the left hand side of (8.25). By the finite dimensionality
this is equivalent to the norm defined by the | |1,K norm on P. Hence we
have a constant > 0 such that
Then since
1 > 0, we have
(8.26)
L
X
l=1
!2
n
X
p
p|
(bl ) |
2l,K .
i
i=1
2
1
2
| p K |2l,K C||B1
K || (det B K ) |pK |l,K ,
ah (vh , vh )
(det BK )||BK ||2 C||B
K || (det B K ) |pK |l,K
Kth
= C
Kth
2
2
(||BK || ||B1
K ||) |pK |l,K
Kth
|pK |2l,K
= C|v
h |2l,
h |1, = |v
2 by Theorem 6.5. This proves the theorem.
since (||BK || ||B1
K ||)
86
8. Numerical Integration
Let us now review our Examples 8.1 through 8.3 to see if the condi- 86
tions of Theorem 8.2 are satisfied.
Let n = 2 and consider Example 8.1. Clearly
= meas K > 0. Also
P
We now turn our attention to the evaluation of the bound for ||u uh ||
given by (8.21). For second-order problems, for which the norm is ||
||1, , we will take as usual for vh Vh the element h u Vh so that we
now get the bound
"
|a(h u, wh ) ah (h u, wh )|
8. Numerical Integration
(8.29)
87
(8.30)
|a(h u, wh ) ah (h u, wh )|
C(u)hk ,
sup
||wh ||1,
wh Vh
| f (wh ) fh (wh )|
C( f )hk ,
wsup
||w
||
h
1,
V
h
h
and these will in turn be obtained from local estimates. (cf. Theorem 8.4 and Exercise 8.5).
As a preliminary step, we need two results which we prove now.
The first of these is a historically important result in the interpolation
theory in Sobolev spaces.
Theorem 8.3 (BRAMBLE-HILBERT LEMMA; cf. Bramble and
Hilbert [27]). Let Rn be open with Lipschitz continuous boundary . Let f W k+1,p ()) which vanishes over Pk . Then there exists a
constant C = C() such that, for all v W k+1,p (),
(8.31)
88
pPk
8. Numerical Integration
88
Lemma 8.1. Let W m,q (), w W m, (). Then w W m,q (), and
there exists a numerical constant C, independent of and w such that
(8.32)
|w|m,q, C
M
X
j=0
m X
X
C, () (w)
j=0 ||= j
We may now apply Lemma 8.1 and Theorem 8.3 to get the estimates
(8.30). We do this in two stages (Theorem 8.4 and Exercise 8.5) in
which, for the sake of simplicity, we present our results for the special
case PK = P2 .
Theorem 8.4. Let PK = P2 and consider a quadrature scheme such that
)
for all P2 , (
= 0. Then there exists a constant C, independent of
K, such that for all ai j W 2, (K) and for all p, p PK we have
#
"
p
p
p p
||1,K |
|0,K .
| C h2K ||ai j ||2,,K ||
(8.33)
|EK (ai j )
x j xi
x j
xi
p p
,
P1 , it suffices to find a suitable estix j xi
mate for EK (avw), for a W 2, (K), v, w P1 . Further, since
Proof. Since we have
(8.34)
89
and v ,
we will first find an estimate for E (av w),
with a W 2, (K)
w P1 . Let 0 w be the orthogonal projection of w onto the subspace P0
Then we may write
in the sense of L2 (K).
(8.35)
E (av w)
= E (av 0 w)
+ E (av (w 0 w)).
8. Numerical Integration
89
R defined by
Consider the functional E : W 2, (K)
=
7 E ()
x)d x =
(
L
X
b 1 ).
1 (
l=1
|
||
C|
0,,K C||
2,,K . Thus E is a continuous linear func|E ()|
2,
Hence by Theorem 8.3, since E vanishes on P1 (
tional on W (K).
1
P2 ) , we have a constant C such that
(8.36)
|
C|
2,,K .
|E ()|
Thus
av 0 w|
|E (av 0 w)|
C|
2,,K
av |2,,K |0 w|
C|
0,,K
since 0 w P0 is a constant function. By Lemma 8.1 (recall that v
P1 ),
h
i
0 w|
|E (av 0 w)|
C|
0,,K |a|1,,K |v|1,,K + |a|2,,K |v|0,,K .
By the equivalence of the L2 and L norms on P0 , and since the
projection has norm less than that of the vector itself in any Hilbert
space we have the chain of inequalities
0 w|
w|
|0 w|
0,,K C|
0,K C|
0,K .
90
1
In this estimate, we do not use the full polynomial invariance of the quadrature
scheme.
8. Numerical Integration
90
(ii) Estimate for E (av (w 0 w)).
Then
Let w P1 be fixed and let W 2, (K).
(
|E ((
w 0 w))|
C|
w 0 w)|
0,,K
C||
0,,K |w 0 w|
0,,K
w 0 w|
C|
0,,K ||||
2,,K .
defined by 7 E ((
Thus the functional on W 2, (K)
w 0 w))
is
w
continuous, linear with norm C|
0 w|
0,,K . Since for P1 , (
w
0 w)
P2 , we have that the functional vanishes on P1 . By Theorem 8.3,
w 0 w|
|E ((
w 0 w))|
C|
0,,K ||
2,,K .
Set = a v . Now,
a|2,,K |v|0,,K + |a|1,,K |v|1,,K ).
|av |2,,K C(|
Again we may use the equivalence between the L -norms of v and
w 0 w and the L2 -norms of the same functions as in (i) since they belong to the finite dimensional space P1 . Also, by the triangle inequality,
w|
|w 0 w|
0,K C|
0,K .
Thus we get
(8.38)
91
(iii) We can now complete the proof. Recall that EK (avw) = (det BK )
E (av w).
Also,
|a|m,,K C hm
K |a|m,,K
(8.39)
2
|v|2m,K C h2m
K (det B K ) |v|2m,K .
1
|w|
0,K C(det BK ) 2 |w|0,K ,
8. Numerical Integration
91
Setting a = Ai j , v =
the proof.
p
p
,w =
we obtain (8.33), thus completing
x j
xi
8. Numerical Integration
92
n
XX
Kth i, j=1
n
XX
Kth i, j=1
2
n
X
i, j=1
|EK (ai j
(h u|K )
(wh |K)
||1,K |
|0,K
x j
xi
12
X (h u|K)
2
||1,K
||ai j ||2,,
||
x j
Kth
21
X (wh |K)
||
||0,K
xi
Kth
(8.41)
Similarly, we have
| f (wh ) fh (wh )|
93
Since q 2,
X
Kth
1
2
Kth
Kth
|EK ( f wh |K )|
1
1
0 and by the general Holders inequality,
q
1
21 1q
X
meas K
Kth
q1
X
||
f
||
2,q,K
Kth
21
X
||w
||
h
1,K
Kth
8. Numerical Integration
93
| f (wh ) fh (wh )|
Ch2 || f ||2,q, .
||wh ||1,
Combining (8.28), (8.29), (8.41) and (8.42) we get (8.40), thus completing the proof.
Remark 8.5. The condition n 5 (needed for the continuous inclusion
H 3 () C 0 ()) was already necessary for the definition of h u.
References: For a survey on numerical quadrature in general one may
refer to Habers survey article [13]. For application to the finite element
method, the Sec. 4.2 of the book by Strang and Fix [22] or Chapter 2 of
the forthcoming book of Ciarlet and Raviart [5].
Chapter 9
R P
2 u v
dx
a(u, v) =
(9.1)
i=1 xi xi
f (v) =
f v dx, f L2 ()
for u, v V = H01 (), where f = F/t, t being the tension. The subset K
is given by
K = {v H01 (); v a.e. in}.
96
Figure 9.1:
95
a(u, v u) f (v u),
for all
v K.
u in ,
(9.4)
u = f where u > ,
u = 0 on .
We have a few regularity results which are listed below:
97
Our aim in this section is to use the finite element method to approximate this problem and obtain error estimates. We list our assumptions
now:
Let be a convex polygon, f L2 (), H 2 () and let u
2
H () H01 ().
Remark 9.1. We cannot assume any more smoothness on u other than
H 2 (). For instance in the 1-dimensional case if f = 0, and even if
the function is very smooth the points of contact of u with will have
discontinuous second derivatives in general; cf. Fig. 9.2.
Figure 9.2:
With the above assumptions we proceed to the approximate problems, first in the abstract setting, as usual.
We have the problems (Ph ) associated with the subspaces Vh V =
1
H0 (). We now choose closed convex subsets Kh Vh . One has to bear
in mind that, in general, Kh 1 K (we will see that this is the case in our
approach, sub-sequently).
We find uh Kh such that for all vh Kh ,
(9.5)
a(uh , vh uh ) f (vh uh ).
98
a(u, v) = (Au)(v),
"
# 21
||u uh || C inf (||u vh || + |u vh |) + inf |uh v| .
2
vh Kh
vK
(9.8)
98
99
||u uh ||
M
||u uh || ||u vh ||
M
||u vh ||)2 0, we have
1
M
||u uh ||2 + ||u vh ||2 ,
2 M
and hence
||u uh ||2 C[|u vh | + |uh v|] +
M2
||u uh ||2 +
||u vh ||2
2
2
Or,
||u uh ||2 C[(|u vh | + ||u vh ||2 ) + |uh v|].
Varying vh Kh and v K and extracting the square root after
taking the infima we get (9.7). This completes the proof.
Remark 9.2. If we have a linear problem then f = Au gives the solution
and we get the original bound (3.1).
Remark 9.3. From (9.8) we see that this estimate holds even if a(, ) is
not symmetric.
We now apply this to the specific membrane problem. Maintaining
our assumptions on , let th be a triangulation by triangles of type (1),
and let Vh be the corresponding subspace of V = H01 ().
Remark 9.4. It is of no practical use if we go to more sophisticated finite
elements, unlike the linear problem. Since u H 2 () is the maximum
smoothness, we may atmost use our abstract estimate theorems only on
the spaces P1 .
One may be tempted to try for Kh those vh which are a.e. in .
However this is not of value from numerical and computational points
of view for we do not easily know where exactly our piecewise linear 99
solution functions would touch . We set instead
(9.10)
100
Remark 9.5.
Figure 9.3:
As seen in Fig. 9.3, though for nodes b, vh (b) (b), it does not
guarantee that vh a.e. Thus we see that Kh 1 K. Now the relation
(9.10) is very easy to implement using the computer.
We now have our main result on the error bound.
Theorem 9.2 (FALK). There exists a constant C depending on || f ||0,
and ||||2, such that for a regular family of triangulations (th ) as above
we have
||u uh ||1, C h.
(9.11)
Remark 9.6. The order of convergence is therefore the same as that for
the linear problems when we use piecewise linear approximations.
Proof. By Theorem 9.1,
# 21
"
2
||u uh ||1, C inf ||u vh ||1, + |u vh |0, + inf |uh v|0, .
vh Kh
100
vK
101
We have
|uh
v1 |20,
|( h )(x)|
21
| h |2 dx = | h |=0, C h2 ||2, .
Hence
Chapter 10
for all
v V,
where
(10.2)
K = V = H02 (), R2 ,
(
)!
2
R
2 v
2 u 2 v 2 u 2 v
a(u, v) = u v + (1 ) 2 u
dx,
u = f in ,
u =
= 0 on = .
103
104
Remark 10.1. It was commented in Sec. 2 that the second term of the
integrand in the definition of a(, ) does not contribute to the differential
equation. Our method here will be equally applicable to both the cases.
viz. with the second term present (the plate problem) or with that term
absent (as it happens in Hydro-dynamics). In our next section, on nonconforming methods, we will see that the second term is essential in
order that we may apply that method.
We assume that is a polygonal domain in R2 . We saw in Sec. 3
that for fourth-order problems we need the inclusion Vh C 1 (). (cf.
Exercise 3.1). Thus we need to use finite elements of class C 1 , such as
the Argyris triangle, the Bogner-Fox-Schmidt rectangle and so on (cf.
Sec. 4).
103
(10.4)
X
K
)
(
p
2 p
p
(ai ),
(ai ),
(ai ); 1 i 4 .
= p(ai ),
x1
x2
x1 x2
105
Figure 10.1:
Equivalently, one may also use the PK -unisolvent set
(10.5)
X
K
= p(ai ), Dp(ai )(ai+1 ai ), Dp(ai )(ai1 ai ),
D2 p(ai )(ai+1 ai , ai1 ai ); 1 i 4
1 = BK 1 , . . . , 2 = BK 2
i,k
i,k
i,1
i,1
for then d
v which is essentially what we need for the abstract
Kv =
error analysis.
P
In K note that
(10.7)
106
105
x2
v
v
(b) =
(c) = 0
is a polynomial in x1 of degree 3, we need to set
x2
x2
2
2
v
v
(b) =
(c) = 0. Thus the degrees of freedom on all
and
x1 x2
x, x2
boundary nodes must be zero. The only free or unknown parameters
are the degrees of freedom at the interior nodes. This takes care of the
boundary conditions.
Let us now turn to the Argyris triangle (cf. Example 4.7).
Figure 10.2:
P
We recall that PK = P5 , dim PK = 21, and K is given by (cf.
Fig. 10.2)
(10.9)
p
3 p
p
x1
x
K
107
We may replace the first and second derivative values at the vertices
by Dp(ai )(ai+1 ai ), Dp(ai )(ai1 ai ), D2 p(ai )(ai+1 ai , ai+1 ai ),
D2 p(ai )(ai+1 ai , ai1 ai ), D2 p(ai )(ai1 ai , ai1 ai ) in order to get
degrees of freedom for which the relations of the type (10.6) may be
p
(ai j ),
satisfied. However, one cannot replace the normal derivatives
o
Dp(ai j )(ak ai j ), 1 i < j 3, k , 1, , j .
Figure 10.3:
This element can be put in an affine family as is readily seen. If K
106
108
|v K v|m,K
h6K
C m |v|6,K , 0 m 6,
K
for v H 6 (K).
Remark 10.2. Though the Hermite triangle of type (5) yields an affine
family, one cannot use it since Vh C 0 () but, in general, Vh 1 C 1 ()
as is necessary for fourth-order problems. This is so because the adjacent triangles will not patch up, in general, in their derivatives along the
medians; cf. 10.4.
Figure 10.4:
107
v
v
2 v
on we need to set, v(b) = v(b ) = 0,
(b) =
(b ) = 0, 2 (b) =
v
v
v
v
2 v
= 0 on we set,
(b) =
(b ) =
(c) = 0,
(b ) = 0. For
2
2 v
2 v
(b) =
(b ) = 0. Thus the only free or unknown parameters
2 v
are 2 at vertices on and the degrees of freedom at all interior nodes.
109
We now get an error estimate when we have triangulations of Argyris triangles. We use our usual terminology more loosely here1 . By a
regular family of triangulations made up of Argyris triangles we mean
hK
, a
that all th consist only of Argyris triangles and that for all K,
K
constant. We also assume that if h = max hK , then h 0.
Kth
Figure 10.5:
Let K be the interpolation operator for the Argyris triangle K and
let K be that for the corresponding Hermite triangle of type (5).
1
Because we have to drop the assumption that all the finite elements are affine equivalent to a reference finite element.
110
(ai j ) =
(K v K v)(ai j ) =
(v K v)(ai j ).
K
K
K
Also since K v = K v along any side of K (since the values of
these polynomials of degree 5 as well as those of their first and second
= 0.
derivatives agree at the end-points), we have
K
Since
D(ai j )(ak ai j ) =
i + (a )ha a ,
(ai j )hak ai j ,
K
ij
k
i j K i,
K
K
D(ai j )(ak ai j ) =
(10.12)
109
and
K
K
i
(v K v)(ai j )hak ai j ,
K
(10.13)
ip .
(v K v)(ai j )hak ai j ,
K i jk
K
1i< j3
k,i,k, j
Now
i| ||a a || ||
|| h ,
|hak ai j ,
K
k
ij
K
K
(10.14)
and
|
(v K v)(ai j )| |v K v|1,,K
K
6
1 h
C(meas K) 2 K |v|6,K .
K
111
(10.15)
h6
(v K v)(ai j )| C 2K |v|6,K .
K
K
(10.16)
hK
| p i jk |m,K .
m
K
(10.17)
1i< j3
k,i,k, j
h8K
m+2
K
i| |p |
(v K v)(ai j )| |hak ai j ,
K
i jk m,K
K
|v|6,K ,
and hence,
|v K v|m,K |v K v|m,K + ||m,K
h6K
h2K
112
Chapter 11
Non-Conforming Methods
for the Plate Problem
WE START WITH a brief classification of finite element methods. The 111
first class of methods are called conforming methods, which we have
described upto now, except when we considered numerical integration.
The second class consists of methods other than conforming. In the
latter class we have the Non-conforming methods included:
Given the abstract problem, the conforming methods deal with the
finding of subspaces Vh V and solving the problems
(Ph )
ah (uh , vh ) = fh (vh ),
for all
vh Vh ,
114
112
115
Figure 11.1:
)
p
p
(ai ),
(ai ), 1 i 4 .
= p(ai ),
x1
x2
(
113
Of course this element can be used only for plates with sides parallel
to the coordinate axes, such as rectangular plates.
P
Exercise 11.1. Show that in Example 11.1, K is PK -unisolvent and
that Adinis rectangle is a finite element of class C 0 , and, in general, not
of class C 1 .
Thus we get a priori that Vh H 1 (). For the boundary condition,
we set all degrees of freedom on the boundary nodes as zero. This gives
us that Vh H01 (). Thus the only unknown or free parameters are
vh
is zero only
the degrees of freedom at the interior nodes. Note that
116
(11.1)
2
R
2 u 2 v 2 u 2 v
2 v
uv + (1 ) 2 u
dx,
a(u,
v)
=
The second integral is defined over Vh as well. Thus for the discrete
problem (Ph ) we may set fh = f . However while for u, v Vh the first
integral is defined over each K th , we cannot define it over , since
we get Dirac measure-like terms along the boundary. To get over this,
we now define
!#
2 uh 2 uh 2 vh
2 uh 2 vh
dx
uh vh + (1 ) 2
ah (uh , vh ) =
x1 x2 x1 x2
x21
x22 x21
Kth
K
(11.2)X Z "
!#
2 uh 2 vh 2 uh 2 vh
2 uh 2 vh
dx,
uh vh + (1 )
=
+
+
2
x1 x2 x1 x2
x21 x21
x22 x22
Kt
XZ "
h K
and we have the discrete problem (Ph ): To find uh Vh such that for all
vh Vh
(11.3)
ah (uh , vh ) = f (vh ).
114
21
X
|vh |22,K |22,K .
||vh ||h =
Kth
Notice that this may be defined over V = H02 () as well and for
v V, ||v||h = |v|2, . In the same way for u, v V, ah (u, v) = a(u, v).
We now show that the seminorm || ||h is indeed a norm on Vh . Let
vh
vh Vh with ||vh ||h = 0. This gives that
= constant over any K. But
x1
vh
at the common vertices
given adjacent finite elements the value of
x1
vh
is constant over . But this is zero on the
coincide and hence
x1
117
vh
vh
= 0 on . Similarly
= 0 on . Since
x1
x2
Vh C 0 () and vh = 0 on , the above conditions give that vh = 0 over
. Thus (11.4) defines a norm on Vh .
To show the existence and uniqueness of the solution of (Ph ), we
show that ah (, ) is Vh -elliptic. In fact we do more than this. We show
that the ah (, ) are Vh -elliptic uniformly with respect to h.
Recall that from physical considerations, 0 < < 12 (see Sec. 2).
Now
XZ
a(vh , vh ) =
(vh )2 dx + (1 )||vh ||2h
Kth K
(11.5)
boundary nodes. Hence
(1 )||vh ||2h .
Remark 11.1. It was mentioned in passing in Sec. 10 that in order to
apply non-conforming methods one needed the second term involving
in the integral defining a(, ). The uniform Vh -ellipticity could not be
got in the Hydrodynamical case where this term is absent.
We now proceed with the abstract error analysis.
Theorem 11.1 (STRANG). Let ah (, ) be Vh -elliptic uniformly with respect to h with e
> 0 so that for all vh Vh
(11.6)
ah (vh , vh ) e
||vh ||h .
(11.7)
115
118
(11.9)
= ah (u vh , uh vh ) + f (uh vh ) ah (u, uh vh )
e vh ||h ||uh vh ||h + | f (uh vh ) ah (u, uh vh )|,
M||u
e
M
||u vh ||h +
e
e
M
||u vh ||h +
e
||uh vh ||h
116
1 | f (uh vh ) ah (u, uh vh )|
e
||uh vh ||h
1
| f (wh ) ah (u, wh )|
sup
.
e
wh Vh
||wh ||h
for all
u, v Vh V.
119
1
e uh ||h sup
inf || h ||h
|u uh | M||u
gH |g| h Vh
#
"
1
inf (Eh (u, h ) + Eh (, u uh ))
+ sup
gH |g| h Vh
where for all v V, a(v, ) = g(v).
We now go on with the error analysis and study the order of convergence. We assume that u H 3 () H02 () which is quite realistic from
the regularity results.
Now, since h u Vh , we have that
inf ||u vh ||h ||u h u||h .
vh Vh
117
Again applying error bounds for each K and then summing over all
K we get
||u h u||h Ch|u|3, .
Thus
inf ||u vh ||h Ch|u|3, .
(11.10)
vh Vh
Our aim is to get a similar estimate for the second term in (11.8). In
fact we show that
(11.11)
sup
wh Vh
| f (wh ) ah (u, wh )|
Ch||u||3, .
||wh ||h
(11.12)
f wnh dx =
Z "
uwnh + (1 ) 2
2 u 2 wnh
2 u 2 wnh 2 u 2 wnh
2
2
x1 x2 x1 x2 x1 x22
x2 x21
!#
dx
120
=
f (wh ) =
f wh dx =
XZ
Kth
"
#!
2 u 2 wh 2 u 2 wh
2 u 2 wh
2
uwh + (1 ) 2
dx
x1 x2 x1 x2 x1 x22
x22 x21
X Z
grad u grad wh dx
=
Kth
(11.14)
wK
u h d + (1 )
K
118
!
2 u wh
2 u wh
d ,
2
+
K K
K
K
K
by Greens formula (2.15) and Lemma 2.2 again. Notice however that
by standard orientation arguments and the continuity of wh over ,
(11.15)
XI
Kth
2 u wh
d = 0.
K K K
(11.16)
X I
2 u wh
u + (1 )
Eh (u, wh ) =
d.
2
K
K
Kt
h K
121
Figure 11.2:
Splitting the boundary into four parts as in Fig. 11.2, we get
(11.17)
Eh (u, wh ) =
Kth
1,K
!
!!
wh
wh
u,
+ 2,K u,
,
x1
x2
! Z
wh
=
u,
x j
K j
!!
2 u wh
wh
u + (1 )
d
K
x j
2K x j
!!
Z
wh
2 u wh
K
d,
u + (1 ) 2
x j
K x j
Kj
K being the Q1 -interpolation operator associated with the values at the 119
four vertices.
Note that (11.17) is the same as (11.15). This is evident provided
we show that the contribution of the terms involving K is zero. This is
so because
! wh = 0 on the boundary and on the common boundaries,
wh
is linear and equal in value for both adjacent finite elements
K
x j
since it agrees at the vertices, but occurs with opposite signs as is obvi
122
Figure 11.3:
We also record that
(11.19)
wh
j PK where
x j K
(
)
p
j PK =
: K R; p PK .
x j
120
Theorem 11.2 (BILINEAR LEMMA). Let Rn be open with Lipschitz continuous boundary . Let W be a subspace of W l+1,q () such
that P1 W. Let b be a continuous bilinear form over W k+1,p () W
such that
(11.21)
Proof. For a given w W, b(, w) : v 7 b(v, w) is a continuous linear form on W k+1,p () vanishing on Pk . Hence by the Bramble-Hilbert
lemma,
(11.22)
123
2 u
=
u
+
(1
)
H 1 (K), since u H 3 (),
wh
1 P K .
v =
x1
Define
(11.24)
1,K (, v) =
K1
(v K v)d
K1
(v K V)d,
1,K (, v) = h2 1,K (,
v ),
121
124
(11.27)
Hence,
Z
(
v K v )d =
Z1
0
(11.28)K1
=
K1
122
Thus 1,K (,
v ) = 0 for P0 , v d
1 P.
Note further that the bilinear form 1,K is continuous, for
|1,K (,
v )| C||||
L2 (K)
v||L2 (K)
||
.
C||||
1,K ||v||1,K , by the Trace theorem (cf. Th. 2.3).
Thus we may apply the bilinear lemma to the bilinear form 1,K with
l = k = 0 to get
(11.29)
|1,K (,
v )| C||
1,K |v|1,K .
125
2 u
= u + (1 ) 2K ,
v = wh ,
x1
and similarly,
(11.32)
126
Example 11.2. The Zienkiewicz triangle (cf. Exercise 4.6) cf. Fig. 11.4.
Figure 11.4:
Figure 11.5:
124
127
Figure 11.6:
Example 11.3. Morleys Triangle (cf. Fig. 11.7).
Figure 11.7:
Here PK = P2 . We always get convergence for regular families, of
course. In fact if u H 4 (), then
(11.34)
128
and
X
K
=
d,
1
3
p(a
),
1
3;
p(a
),
1
i
<
j
3;
.
i
ij
K
i
125
Figure 11.8:
Here also the finite element is not of class C 0 in general, but the
method always yields convergence.
References: For general reference on non-conforming methods, see
Strang and Fix [22], for the bilinear lemma, see Ciarlet [29]. For a detailed study of the Zienkiewicz triangle, Moreleys triangle and Fraeijs
de Veubeke triangle, see Lascaux and Lesaint [15]. For a nonconforming method with penalty, see Babuska and Zlamal [26].
Bibliography
V.
[1] BRAMBLE, J.H: THOMEE,
Interior maximum norm estimates for some simple finite element methods. Revue FranC
aise dAutomatique, Informatique,
Recherche Operationelle (RAIRO), R-2, (1974), pp.-5-18.
126
BIBLIOGRAPHY
130
[7] CIARLET, P.G; RAVIART, P.-A.
General Lagrange and Hermite interpolation in Rn with applications to finite element methods, Arch. Rational Mech. Anal. 46,
(1972), pp. 177-199.
[8] CIARLET, P.G; WAGSCHAL, C.
Multipoint Taylor formulas and applications to the finite element
method, Numer. Math., 17, (1971), pp. 84-100.
[9] CIAVALDINI, J.F; NEDELEC, J.C.
Sur lelement de Fraeijs de Veubeke et Sander, RAIRO, R-2,
(1974), pp. 29-45.
[10] DUVAUT, G; LIONS, J.-L.
Les Inequations en Mecanique et en Physique, Dunod, Paris
(1972).
[11] FALK, R.S.
Error estimates for the approximation of a class of variational inequalities, Math. Comput. (to appear).
127
BIBLIOGRAPHY
131
J.
[20] NECAS,
Les Methodes Directes en Theorie des Equations Elliptiques, Masson, Paris, (1967).
[21] STRANG, G.
Approximation in the finite element method, Number. Math.,
19(1972), pp. 81-98.
[22] STRANG, G; FIX, G.J.
An Analysis of the Finite Element Method, Prentice-Hall, Englewood Cliffs, (1973).
[23] ZENI
SEK,
A.
Interpolation polynomials on the triangle, Numer. Math., 15,
(1970), pp. 283-296.
[24] ZENI
SEK,
A.
Polynomial approximation on tetrahedrons in the finite element
method, J. Approximation Theory, 7, (1973), pp. 334-351.
BIBLIOGRAPHY
132
[25] ZLAMAL,
M.
On the finite element method, Numer. Math., 12, (1968), pp. 394409.
[26] BABUSKA,
I.; ZLAMAL,
M.
Nonconforming elements in the finite element method with
penalty, SIAM J. Numer. Anal., 10(1973), pp. 863-875.
[27] BRAMBLE, J.H.; HILBERT, S.R.
Estimation of linear functionals on Sobolev spaces with application to Fourier transforms and spline interpolation, SIAM J. Numer.
Anal., 7, (1970), pp. 113-124.
[28] BRAMBLE, J.H.; HILBERT, S.R.
Bounds for a class of linear functionals with applications to Hermite interpolation, Numer. Math., 16, (1971), pp. 362-369.
[29] CIARLET, P.G.
Conforming and nonconforming finite element methods for solving the plate problem, in Conference on the Numerical Solution
of Differential Equations (G.A. WATSON, Editor), pp. 21-31,
Lecture Notes in Mathematics, Vol. 363, Springer-Verlag, Berlin
(1974).
[30] CIARLET, P.G.; RAVIART, P.-A.
Interpolation theory over curved elements, with applications to finite element methods, Comp. Methods in Appl. Mech. and Engineering, 1, (1972), pp. 217-249.
[31] CIARLET, P.G.; RAVIART, P.-A.
The combined effect of curved boundaries and numerical integration in isoparametric finite element methods, in The Mathematical Foundations of the Finite Element Method with Applications to
Partial Differential Equations (A.K. AZIZ, Editor), pp. 409-474,
Academic Press, New York, (1972).
128
BIBLIOGRAPHY
133
[32] ZLAMAL,
M.
A finite element procedure of the second order of accuracy, Numer.
Math., 16, (1970), pp. 394-402.
BIBLIOGRAPHY
134
129
Additional References
Books (Additional References)
BABUSKA, I; AZIZ, A.K. First part of The Mathematical Foundations of the Finite Element Method with applications to Partial
Differential Equatios (Symposium, University of Maryland, Baltimore, June-26-30, 1972) (Edited by A.K. AZIZ) Academic
Press, New York, (1972).
NORRIE, D.H; DE VRIES, G. The Finite Element Method - Fundamentals and Applications, Academic Press, New York, (1973).
ODEN, J.T. Finite Elements of Nonlinear Continue, McGraw Hill,
New York, (1972).
ZIENKIEWICZ, O.C. The finite Element Method in Engineering Science, McGraw Hill, London, (1971).
130
Journals (where articles dealing with the finite element method may
be regularly found).
Revue Francaise dAutomatique, Informatique, Recherche Operationelle (RAIRO), Red Series
Computer Methods in Applied Mechanics and Engineering.
International Journal for Numerical Methods in Engineering.
Mathematics of Computation.
SIAM Journal on Numerical Analysis
Numerische Mathematik.
Proceedings of Conferences totally or partially devoted to the finite
element method:
The Mathematics of Finite Elements and Applications, (Conference,
Brunel University, 18-20 April, 1972), Edited by J.R. WHITEMAN,
Academic Press, London, 1973.
Conference on the Numerical Solution of Differential Equations,
(Conference, Dundee, 03-06 July 1973), Edited by G.A. WATSON, Lecture Notes in Mathematics, Vol. 363, Springer-Verlag, New York, 1974.
The Mathematical Foundations of the Finite Element Method with
BIBLIOGRAPHY
135
Applications to Partial Differential Equations, (Symposium, University of Maryland, Baltimore, 26-30 June. 1972), Edited by A.K. AZIA,
Academic Press. New York, 1972.
Computing Methods in Applied Science and Engineering, Vol. 1
and 2 (International Symposium IRIA LABORIA, Versailles, 17-21 December, 1973) Edited by R. GLOWINSKI and J.L. LIONS, Lecture notes
in Computer Science, Vol., 10 and 11, Springer-Verlag, New York, 1974.
Topics in Numerical Analysis (Royal Irish Academy Conference on
Numerical Analysis, Dublin, 1972), Edited by J.J.H. MILLER, Academic Press, New York, 1973.
Mathematical Aspects of Finite Elements in Partial Differential
Equations (Symposium, 01-03 April, 1974, Mathematics Research Center, The University of Wisconsin, Madison), Edited by C. de BOOR, Academic Press, New York, 1974.
Proceedings of the Second Conference on the Mathematics of Finite
Elements and Applications, (Brunel University, April 07-10, 1975, to
appear.)