Solving Heat Equation w/ Neumann BCs Different Domain

Click For Summary
SUMMARY

The discussion focuses on solving the heat equation \(\frac{\partial u}{\partial t} = \frac{\partial^2 u}{\partial x^2}\) with initial condition \(u(x,0) = a \exp(-\frac{x^2}{2c^2})\) and Neumann boundary conditions \(\frac{\partial u}{\partial x}(-1,t) = \frac{\partial u}{\partial x}(1,t) = 0\). The user initially attempts a separation of variables approach, leading to the equations \(f''(x) + \lambda f(x) = 0\) and conditions involving \(A\) and \(B\). A suggestion to utilize a Laplace transform for the time variable is made, indicating that a transform method may yield a more effective solution than separation of variables.

PREREQUISITES
  • Understanding of partial differential equations (PDEs)
  • Familiarity with Neumann boundary conditions
  • Knowledge of separation of variables technique
  • Basic concepts of Laplace transforms
NEXT STEPS
  • Research the application of Laplace transforms in solving PDEs
  • Study the implications of Neumann boundary conditions in heat equations
  • Explore advanced techniques in separation of variables for non-standard domains
  • Investigate the uniqueness and existence of solutions for heat equations
USEFUL FOR

Mathematicians, physicists, and engineers working with heat transfer problems, particularly those dealing with partial differential equations and boundary value problems.

maxtor101
Messages
22
Reaction score
0
Hi guys!

I'm to find the solution to

\frac{\partial u}{\partial t} = \frac{\partial^2 u}{\partial x^2}

Subject to an initial condition

u(x,0) = u_0(x) = a \exp(- \frac{x^2}{2c^2})

And Neumann boundary conditions

\frac{\partial u}{\partial x} (-1,t) = \frac{\partial u}{\partial x} (1,t) = 0

I can usually do this no problem assuming the domain is for instance [0,L], but I get stuck with this one :

Using separation of variables :

u(x,t) = f(x)g(t)

This yields:

\frac{1}{g} \frac{dg}{dt} = \frac{1}{f} \frac{d^2f}{dx^2} = -\lambda


Spatial Part:

\frac{1}{f} \frac{d^2f}{dx^2} = -\lambda

\frac{d^2f}{dx^2} + \lambda f = 0

Therefore :

f(x) = A \cos(\sqrt{\lambda} x) + B \sin(\sqrt{\lambda} x)

And since I'm considering Neumann Boundary conditions I get the derivative of this

f \prime (x) = -A \sqrt{\lambda}\sin(\sqrt{\lambda} x) + B \sqrt{\lambda} \cos(\sqrt{\lambda} x)


So, f \prime (-1) = 0

This gives:

A \sin(\sqrt{\lambda}) + B \cos(\sqrt{\lambda}) = 0

And for f \prime (1) = 0

I get :

-A \sin(\sqrt{\lambda}) + B \cos(\sqrt{\lambda}) = 0

So from these two equations I can conclude that:

Firstly by just adding the two equations

B \cos(\sqrt{\lambda}) = 0

So either B = 0 or \cos(\sqrt{\lambda}) = 0

Now substituting B \cos(\sqrt{\lambda}) = 0 back into A \sin(\sqrt{\lambda}) + B \cos(\sqrt{\lambda}) = 0

I also get
A \sin(\sqrt{\lambda}) = 0

So either A = 0 or \sin(\sqrt{\lambda}) = 0

Obviously \sin(\sqrt{\lambda}) and \cos(\sqrt{\lambda}) can't both equal zero, so how do I approach this...

Apologies if this is a stupid question..
Any help would be greatly appreciated
Max
 
Physics news on Phys.org
Have you tried a transform method? I might be tempted to try a Laplace transform on the t variable. I think that this problem requires a transform solution rather than a separation of variable solution.
 
sin(\sqrt{\lambda}) and cos(\sqrt{\lambda}) don't both have to be 0: one of A and B can be 0 and still give you a non-trivial solution.
 
Ah ok I see! Thanks for your help.

So I can choose for example in case 1: A=0 , B=B . case2: B=0 , A=A
 

Similar threads

  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 3 ·
Replies
3
Views
4K
  • · Replies 28 ·
Replies
28
Views
4K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K
Replies
6
Views
2K
  • · Replies 4 ·
Replies
4
Views
4K