Finite Elements

Basic formulation
Basis functions
Stiffness matrix
Poissons equation
Regular grid
Boundary conditions
Irregular grid

Numerical Examples
Scope: Understand the basic concept of the finite element
method with the simple-most equation.

Finite element method

Formulation
Let us start with a simple linear system of equations

Ax b

|*y

and observe that we can generally multiply both sides of this

equation with y without changing its solution. Note that x,y
and b are vectors and A is a matrix.

yAx yb

y n

We first look at Poissons equation

u ( x) f ( x)
where u is a scalar field, f is a source term and in 1-D
2

2 2
x
Finite element method

Poissons equation
We now multiply this equation with an arbitrary function v(x),
(dropping the explicit space dependence)

uv fv
... and integrate this equation over the whole domain. For
reasons of simplicity we define our physical domain D in the
interval [0, 1].

uv fv
D

D
1

uvdx fvdx
0

... why are we doing this? ... be patient ...

Finite element method

Das Reh springt hoch,

das Reh springt weit,
warum auch nicht,
es hat ja Zeit.

Discretization
As we are aiming to find a numerical solution to our problem it is
clear we have to discretize the problem somehow. In FE problems
similar to FD the functional values are known at a discrete set
of points.
... regular grid ...
... irregular grid ...

Domain D
The key idea in FE analysis is to approximate all functions in
terms of basis functions , so that
N

u u~ ci i
i 1

Finite element method

Basis function
N

~
u u
ci i
i 1

where N is the number nodes in our physical domain and ci are real
constants.
With an appropriate choice of basis functions i, the coefficients ci
are equivalent to the actual function values at node point i. This of
course means, that i=1 at node i and 0 at all other nodes ...
Doesnt that ring a bell?
Before we look at the basis functions, let us ...

Finite element method

Partial Integration
... partially integrate the left-hand-side of our equation ...
1

uvdx fvdx
1

( u )vdx uv 0 vudx
1

we assume for now that the derivatives of u at the boundaries vanish

so that for our particular problem
1

( u )vdx vudx
Finite element method


... so that we arrive at ...
1

uvdx fvdx
... with u being the unknown. This is also true for our
approximate numerical system
1

~vdx fvdx

... where ...

N

u~ ci i
i 1

was our choice of approximating u using basis functions.

Finite element method

Partial integration
1

~vdx fvdx

... remember that v was an arbitrary real function ...

if this is true for an arbitrary function it is also true if

v j
... so any of the basis functions previously defined ...

... now lets put everything together ...

Finite element method

The discrete system

N

The ingredients:

v k

u~ ci i
i 1

~vdx fvdx

c dx f dx
0

i 1

... leading to ...

Finite element method

The discrete system

... the coefficients ck are constants so that for one particular
function k this system looks like ...
1

c dx f dx
i 1

... probably not to your surprise this can be written in matrix form

bi Aik g k
A b gk
T
ik i

Finite element method

The solution
... with the even less surprising solution

T 1
ik

bi A

gk

remember that while the bis are really the coefficients of the basis
functions these are the actual function values at node points i as well
because of our particular choice of basis functions.
This become clear further on ...

Finite element method

Basis functions
we are looking for functions i
with the following property

... otherwise we are

free to choose any
function ...

10

The simplest choice

are of course linear
functions:

9
8

6
5

+ grid nodes
blue lines basis
functions i

4
3
2
1

Finite element method

1
i ( x)
0

for x xi
for x x j , j i

Basis functions - gradient

To assemble the stiffness matrix we need the gradient (red) of the basis
functions (blue)
10
9
8
7
6
5
4
3
2
1

Finite element method

Stiffness matrix
Knowing the particular form of the basis functions we can now
calculate the elements of matrix Aij and vector gi
1

c dx f dx
i 1

bi Aik g k
1

Aik i k dx
0

g k f k dx
0

Note that i are continuous functions defined in the interval [0,1], e.g.
x xi 1
x x
i i 1
xi 1 x

i ( x)

xi 1 xi
0

Finite element method

for

xi 1 x xi

for

xi x xi 1
elsewhere

Let us for now assume a

regular grid ... then

Stiffness matrix regular grid

x xi 1
x x
i i 1
xi 1 x

i ( x)

xi 1 xi
0

for

xi 1 x xi

for

xi x xi 1
elsewhere

x
~
dx 1
~

x
~
i ( x ) 1
dx

for

dx ~
x 0

for

0 ~
x dx
elsewhere

... where we have used ...

~
x x xi
dx xi xi 1
xi
dx
Finite element method

Regular grid - gradient

1 / dx

~
i ( x ) 1 / dx

for
for

dx ~
x 0
0~
x dx
elsewhere

i
1/dx

xi
dx
-1/dx
Finite element method

~
x x xi
dx xi xi 1

Stifness matrix - elements

10

9
8
7
6

Aik i k dx

5
4
3

2
1

... we have to distinguish various cases ... e.g. ...

1

A11 11dx

x1 dx

dx
1

x1

x2

A22 2 2 dx
0

1
2
dx
Finite element method

x1 dx

x1

x2 dx

dx dx
2

x2 dx

1
dx

dx dx 2

1 1
1
dx 2
dx dx
dx

x2

dx

dx

2
dx

dx

1
dx
dx

Stiffness matrix
10

9
8
7
6

Aik i k dx

5
4
3

2
1

......and
and......
1

A12 1 2 dx
0

1
2
dx

dx

dx

x1 dx

x1 dx

dx
1

x1

x1

1 1
dx
dx dx

1
dx

A21 A12
... so that finally the stiffness matrix looks like ...
Finite element method

Stiffness matrix
10

9
8
7
6

Aik i k dx

5
4
3

2
1

Aij

1
dx

1 1

1 2 1
1

2 1
1 1

... so far we have ignored sources and boundary conditions ...

Finite element method

Boundary conditions - sources

... let us start restating the problem ...

u ( x) f ( x)
... which we turned into the following formulation ...
1

c dx f dx
i 1

... assuming ...

N 1

u~ ci i
i 1

with b.c.

u~ ci i u (0)1 u (1) N
i 2

where u(0) and u(1) are the values at the boundaries of the domain [0,1].
How is this incorporated into the algorithm?

Finite element method

Boundary conditions
1

u ( x) f ( x)

c dx f dx
i 1

... which we turned into the following formulation ...

1

n 1

c dx f dx u (0) dx u (1) dx
i 2

... in pictorial form ...

T

= g

boundary condition

source heterogeneity (f)

=
boundary condition

... the system feels the boundary conditions through the (modified) source term
Finite element method

Numerical Example
u ( x) f ( x)
Domain: [0,1]; nx=100;
dx=1/(nx-1);f(x)=d(1/2)
Boundary conditions:
u(0)=u(1)=0

Matlab FEM code

% source term
s=(1:nx)*0;s(nx/2)=1.;
% boundary left u_1 int{ nabla phi_1 nabla phij }
u1=0;
s(1) =0;
% boundary right u_nx int{ nabla phi_nx nabla phij }
unx=0; s(nx)=0;
% assemble matrix Aij

Matlab FD code
f(nx/2)=1/dx;

for it = 1:nit,
uold=u;
du=(csh(u,1)+csh(u,-1));
u=.5*( f*dx^2 + du );
u(1)=0;
u(nx)=0;

end
Finite element method

A=zeros(nx);
for i=2:nx-1,
for j=2:nx-1,
if i==j,
A(i,j)=2/dx;
elseif j==i+1
A(i,j)=-1/dx;
elseif j==i-1
A(i,j)=-1/dx;
else
A(i,j)=0;
end
end
end
fem(2:nx-1)=inv(A(2:nx-1,2:nx-1))*s(2:nx-1)';
fem(1)=u1;
fem(nx)=unx;

Regular grid
u ( x) f ( x)
Domain: [0,1]; nx=100;
dx=1/(nx-1);f(x)=d(1/2)
Boundary conditions:
u(0)=u(1)=0

FD (red) - FEM (blue)

0.25

0.2

Matlab FD code (red)

u(x)

0.15

0.1

Matlab FEM code (blue)

0.05

Finite element method

0.1

0.2

0.3

0.4

0.5
x

0.6

0.7

0.8

0.9

Regular grid - non zero b.c.

u ( x) f ( x)
Domain: [0,1]; nx=100;
dx=1/(nx-1);f(x)=d(1/2)
Boundary conditions:
u(0)=0.15
u(1)=0.05

FD (red) - FEM (blue) -> Regular grid

0.4
0.35
0.3

% Quelle
0.25

s=(1:nx)*0;s(nx/2)=1.;
% Randwert
links
0.2
phij }
u(x)

Matlab FD code (red)

u1=0.15;
0.15

nabla

u_nx int{ nabla phi_nx nabla

unx=0.05; s(nx-1)=unx/dx;
0.05
0

Finite element method

int{ nabla phi_1

s(2) =u1/dx;

% Randwert links
phij0.1
}

Matlab FEM code (blue)

u_1

0.1

0.2

0.3

0.4

0.5
x

0.6

0.7

0.8

0.9

Stiffness irregular grid

10
9

Aik i k dx

7
6

5
4
3
2
1

x1 h1

x1 h1

x1

A12 1 2 dx

1 2 dx

x1

1 1
dx
h1 h1

2
+

3
+

1 1
1
2 dx
A21
h1 0
h1
Aii
Finite element method

1
1

hi 1 hi

i=1
+

h1 h2

4
+
h3

5
6 7
+ + +
h4 h5 h6

Example
Stiffness matrix A

u ( x) f ( x)

for i=2:nx-1,
for j=2:nx-1,
if i==j,
A(i,j)=1/h(i-1)+1/h(i);

Domain: [0,1]; nx=100;

dx=1/(nx-1);f(x)=d(1/2)
Boundary conditions:
u(0)=u0; u(1)=u1

elseif i==j+1
A(i,j)=-1/h(i-1);
elseif i+1==j
A(i,j)=-1/h(i);
else
A(i,j)=0;

i=1 2
3
4 5
6
+ +
+
+ +
+
h1 h2 h3 h4 h5 h6

Finite element method

7
+

end
end
end

Irregular grid non zero b.c.

u ( x) f ( x)

FEM on Chebyshev grid

Domain: [0,1]; nx=100;

dx=1/(nx-1);f(x)=d(1/2)
Boundary conditions:
u(0)=0.15
u(1)=0.05

FD (red) - FEM (blue)

0.4
0.35
0.3
0.25

u(x)

Matlab FD code (red)

0.2
0.15

Matlab FEM code (blue)

0.1
0.05

+ FEM grid points

Finite element method

0.1

0.2

0.3

0.4

0.5
x

0.6

0.7

0.8

0.9

Summary

In finite element analysis we approximate a function defined in

a Domain D with a set of orthogonal basis functions with
coefficients corresponding to the functional values at some
node points.
The solution for the values at the nodes for some partial
differential equations can be obtained by solving a linear
system of equations involving the inversion of (sometimes
sparse) matrices.
Boundary conditions are inherently satisfied with this
formulation which is one of the advantages compared to finite
differences.

Finite element method