Ask a question from expert

Ask now

Numerical Methods for Partial Differential Equations

5 Pages1165 Words227 Views
   

Added on  2023-04-25

About This Document

This study material covers numerical methods for solving partial differential equations, including Triangular domain, Rothe's method, and 1D convention-diffusion-reaction. It includes detailed derivations and boundary conditions for each method. References are also provided for further reading.

Numerical Methods for Partial Differential Equations

   Added on 2023-04-25

BookmarkShareRelated Documents
3. Triangular domain
Given;
u=f
Let Ω=Ω R+¿2 ¿ amd Ω :=Ω 0< x , y <1
Taking i=0, we discretize as follows;
n
2 u
x2 ( 0 ,1x ) = 1
h2 ( u1 , j2 u0 , j +ui , j ) +O ( h2 ) ... ... ... ... ... (i)
Applying the boundary conditions;
γ0 , j= n u
x ( 0 , 1x )= 1
2 h ( u1 , j u1 , j ) +O ( h2 ) ... ... ... ... ... ... .(ii)
Combining (i) and (ii);
n
2 u
x2 ( t ) = 1
h2 ( 2ui , j2u0 , j ) + 2 γ o , j
h +O(h2)
Whereγ =0;
n
2 u
x2 ( t ) = 1
h2 ( 2ui , j2u0 , j ) +O(h2)
From the solution obtained we observe that the consistency level is 2.
From the solution we obtain that an alternating sequence is formed meaning that north point is
equal to the east point as well as west point being equal to the south point as described in the
figure below (Khasminskii, 2011).
Numerical Methods for Partial Differential Equations_1
From the details in the diagram we come up with a matrix as below;
(
4 1 0
1 4 1
0 1 4
1 0 0
0 1 0
0 0 1
0 0 0
0 0 0
0 0 0
1 0 0
0 1 0
0 0 1
4 1 0
1 4 1
0 1 4
1 0 0
0 1 0
0 0 1
0 0 0
0 0 0
0 0 0
1 0 0
0 1 0
0 0 1
4 1 0
1 4 1
0 1 4
)( u1,1
u2,1
u3,1
u1,2
u2,2
u3,2
u 1,3
u 2,3
u 3,3
) =
( h2 f 1,1u1,0u 0,1
h2 f 2,1u 2,0
h2 f 3,1u 3,0u 4,1
h2 f 1,2u 0,2
h2 f 2,2
h2 f 3,2u 4,2
h2 f 1,3u 0,3u 1,4
h 62 f 2,3u2,4
h2 f 3,3ui 4,3u 3,4
)4. Rothe’s method
uk +1=uk + τ uk+1 ,(x , y)Ω
Rearranging we have;
uk +1uk =τ uk +1
Using the 5-point finite difference formula to discrete the PDE in space we obtain;
dui , j
dt ( x , y ) = D
h2 ( ui+1 , j +ui1 , j +ui , j+1 +ui , j14 ui , j ) +O(h2 )
Therefore,
uk+1 uk
h =ui+1 , j
k+1 + ui1 , j
k+1 4 ui , j
k+1 +ui , j+1
k +1 +ui , j1
k+1
h2 +O (h2)
Multiplying both sides by h;
uk +1uk =ui +1 , j
k+1 +ui1, j
k+1 4 ui , j
k+1 +ui , j+1
k +1 +ui , j1
k +1
h +O(h2)
The convergence order is 2.
To make the linear stable; τ h2
Comparing the 5 point finite difference formula to the implicit method, the implicit method is
inefficient.
Numerical Methods for Partial Differential Equations_2

End of preview

Want to access all the pages? Upload your documents or become a member.

Related Documents
Solving Nonlinear System of Equations and Poisson's Equation with Dirichlet Boundary Conditions
|7
|1012
|295

Multirate Signal Processing: Theory and Applications
|7
|1005
|380

Approximate Derivatives and Differential Equations Solutions
|8
|1368
|198

Partition Function, Entropy, Distinguishability, Planck Distribution, Heat Capacity of Platinum
|10
|2204
|237

Desklib - Online Library for Study Material
|207
|16045
|382

Assignment On The Math Induction
|4
|1438
|9