Numerical solution of partial differential equation

In summary, the conversation discussed the need to solve a system of equations for n=0,1,2 with given initial and boundary conditions. The possibility of solving the system numerically was also mentioned, with a request for suggestions on the best scheme for accuracy. The challenging coupled boundary conditions were also brought up, with a plea for assistance. The equations and conditions were also listed, including the journal recommendation for further help.
  • #1
Suvadip
74
0
I need to solve the following system of equations for [TEX]n=0,1,2 [/TEX] subject to the given initial and boundary conditions. Is it possible to solve the system numerically. If yes, please give me some idea which scheme I should use for better accuracy and how should I proceed. The coupled boundary conditions are challenging for me. Please help.

[TEX]\frac{\partial C_n}{\partial t}-\frac{\partial^2 C_n}{\partial r^2}-\frac{1}{r}\frac{\partial C_n}{\partial r}=\beta n\, f(r,t)C_{n-1}+n(n-1)C_{n-2}[/TEX]
[TEX]\frac{\partial \zeta_n}{\partial t}-\frac{\partial^2\zeta_n}{\partial r^2}-\frac{1}{r}\frac{\partial \zeta_n}{\partial r}=\beta n \,g(r,t)\zeta_{n-1}+n(n-1)\zeta_{n-2}[/TEX][TEX]C_n(0,r)=1 \quad\mbox{for}\quad n=0[/TEX]
[TEX]=0 \quad\mbox{for}\quad n>0 [/TEX][TEX]\zeta_n(0,r)=1 \quad\mbox{for}\quad n=0[/TEX]
[TEX]\quad\quad\quad=0 \quad\mbox{for}\quad n>0[/TEX][TEX]\frac{\partial C_n}{\partial r}+\gamma C_n=0 \quad\mbox{at}\quad r=a[/TEX]
[TEX]\frac{\partial C_n}{\partial r}=\kappa \frac{\partial \zeta_n}{\partial r} \quad\mbox{at}\quad r=b[/TEX]
[TEX]C_n=\lambda\zeta_n \quad\mbox{at}\quad r=b[/TEX]
[TEX]\frac{\partial \zeta_n}{\partial r}=0 \quad\mbox{at}\quad r=0[/TEX]
 
Last edited:
Mathematics news on Phys.org
  • #2
You should check out the journal on Numerical Methods for Partial Differential Equations. It comes out in monthly in volumes that are the size of a 300 page textbook. I have volume 29 number 6 Nov 2013 and that may not be much of a help to you but there is bound to be a volume of interest.

You can also view the journal online at wilyonlielibrary.com/journal/num
 

FAQ: Numerical solution of partial differential equation

What is a partial differential equation (PDE)?

A PDE is a mathematical equation that involves multiple independent variables and their partial derivatives. It describes a relationship between a function and its rates of change with respect to these variables.

Why do we need numerical solutions for PDEs?

Many PDEs do not have analytical solutions, meaning they cannot be solved using traditional algebraic methods. In such cases, numerical methods provide a way to approximate the solution by dividing the domain into smaller regions and solving the equation at discrete points.

What are the different numerical methods for solving PDEs?

There are several methods for numerically solving PDEs, including finite difference, finite element, and finite volume methods. These methods differ in their approach to discretizing the problem and solving the resulting system of equations.

How do we choose the appropriate numerical method for a specific PDE?

The choice of numerical method depends on several factors, including the type of PDE (e.g. elliptic, parabolic, hyperbolic), the boundary conditions, and the desired accuracy and computational efficiency. It is important to carefully analyze the problem and consider these factors before selecting a numerical method.

What challenges or limitations are associated with numerical solutions of PDEs?

One major challenge is the trade-off between accuracy and computational cost. More accurate solutions often require more computational resources, which can be a limitation for large and complex problems. Additionally, numerical solutions may be affected by round-off errors and stability issues, which must be carefully managed to obtain reliable results.

Similar threads

Back
Top