You are on page 1of 12

Point Collocation Method

FEM - Introduction - Methods of Solving Differential Equations

Suddhasheel Ghosh, PhD


Department of Civil Engineering
Jawaharlal Nehru Engineering College
N-6 CIDCO, 431003

Advanced Numerical Methods Series

shudh (JNEC)

PCM

MEStru2k1617

1 / 12

DiffEq1

Introduction to terminology
Given a differential equation
!
d2 y dy

, , y, x = 0,
dx2 dx

(1)

and the initial conditions,


F1

!
dy
, y, x = a = 0
dx

F2

!
dy
, y, x = b = 0
dx

So, given the points x1 = a, x2 , x3 , . . . , xi , . . . , xn , xn+1 = b, it is desired to


find the solution of the differential equation at the points
xj , j = 2, . . . , n. The points xj , j = 2, . . . , n are known as the points of
collocation.
shudh (JNEC)

PCM

MEStru2k1617

2 / 12

DiffEq1

A second-order Boundary Value Problem

A boundary value problem is given as follows:


d2 y
dx
+ P(x) + Q(x)y = R(x)
2
dy
dx
along with the conditions
y(x1 ) = A,

shudh (JNEC)

y(xn+1 ) = B

PCM

MEStru2k1617

3 / 12

Collocation Method

Point collocation Method


Derivative calculation

Assume that
y=

n
X

i xi .

i=0

Therefore, we will have


n

dy X
=
i i xi1 ,
dx
i=0

and

d2 y X
=
i(i 1)i xi2
dx2
i=0

shudh (JNEC)

PCM

MEStru2k1617

4 / 12

Collocation Method

Point collocation Method I


Substitution and formulation

Substituting these in the differential equation, we have


n
X

i(i 1)i xi2 + P(x)

i=0

n
X

i i xi1 + Q(x)

i=0

n
X

i xi = R(x).

i=0

Thus giving,
n
X

h
i
i i(i 1)xi2 + ixi1 P(x) + xi Q(x) = R(x)

i=0

The aim of the interpolation method is to agree at the node points,


and therefore, we shall have:
n
X

h
i
i i(i 1)xji2 + ixji1 P(xj ) + xij Q(xj ) = R(xj ), j = 2, . . . , n

i=0
shudh (JNEC)

PCM

MEStru2k1617

5 / 12

Collocation Method

Point collocation Method II


Substitution and formulation

For the nodes x1 and xn+1 , we have the following conditions:


n
X

i xi1 = A

i=0
n
X

i xin+1 = B

i=0

shudh (JNEC)

PCM

MEStru2k1617

6 / 12

Collocation Method

Point collocation Method


Matrix formulation of the problem

x1
x21
1

Q(x2 ) P(x2 ) (2 + 2x2 P(x2 ) + x22 Q(x2 ))

..
..
.
...
.

Q(xn ) P(xn ) (2 + 2xn P(xn ) + x2n Q(xn ))

1
xn+1
x2n+1

...
...
...
...
...

xn1
0

n
n1
n2

[n(n 1)x2 + nP(x2 )x2 + x2 Q(x2 )] 1


.
..
..
.

n
n1
n2
[n(n 1)xn + nP(xn )xn + xn Q(xn )] n1

n
n
xn+1

R(x2 )
.
= ..

R(xn )

The solution can then be achieved by any of the standard methods like Gauss-Siedel,
Gaussian Elimination or Gauss-Jordan Elimination.

shudh (JNEC)

PCM

MEStru2k1617

7 / 12

Collocation Method

Point collocation method I


Example

Use the point collocation method to solve the following differential


equation:
d2 y
y=x
dx2
Use the boundary conditions y(x = 0) = 0 and y(x = 1) = 0. Choose
x = 0.25 and x = 0.5 as collocation points. (Desai, Eldho, Shah)
Solution: There are four points where we are considering the solution
for, x = 0, 0.25, 0.5, 1. We label them as x1 , x2 , x3 , x4 . Since there are four
points, we will consider a cubic polynomial.
y = 0 + 1 x + 2 x2 + 3 x3

shudh (JNEC)

PCM

MEStru2k1617

8 / 12

Collocation Method

Point collocation method II


Example

We have
dy
= 1 + 22 x + 33 x2
dx
d2 y
= 22 + 63 x
dx2
Substituting these in the given differential equation, we have
22 + 63 x 0 1 x 2 x2 3 x3 =

0 1 x + (2 x2 )2 + (6x x3 )3 =

From the first boundary condition y(x = 0) = 0, we have


0 + 1 (0) + 2 (02 ) + 3 (03 ) = 0 = 0 = 0
shudh (JNEC)

PCM

(2)
MEStru2k1617

9 / 12

Collocation Method

Point collocation method III


Example

From the second boundary condition y(x = 1) = 0, we have


0 + 1 (1) + 2 (12 ) + 3 (13 ) = 0 = 1 + 2 + 3 = 0

(3)

At the collocation points, we have the following equations:


For x = 0.25, we have
1 (0.25) + (2 (0.25)2 )2 + (6 0.25 (0.25)3 ) =

0.25

0.251 + 1.93752 + 1.48443 =

0.25

(4)

For x = 0.5, we have


0.51 + (2 0.52 )2 + (6 0.5 0.53 ) =

0.5

0.51 + 1.752 + 2.8753 =

0.5

shudh (JNEC)

PCM

MEStru2k1617

(5)
10 / 12

Collocation Method

Point collocation method IV


Example

Using the equations above, we have the following matrix based


arrangement

1
1 1 0
1

0.25 1.9375 1.4844 2 = 0.25

0.5
1.75
2.875 3
0.5

(6)

which gives on the inverse operation,


1 = 0.1459, 2 = 0.006738, 3 = 0.1526
Therefore the polynomial approximation for y is
y = 0.1459x 0.006738x2 + 0.1526x3

shudh (JNEC)

PCM

(7)

MEStru2k1617

11 / 12

Collocation Method

Thank you!

shudh (JNEC)

PCM

MEStru2k1617

12 / 12

You might also like