Partial differential equation with conditions

Click For Summary
SUMMARY

The discussion centers on solving the partial differential equation (PDE) given by du/dt = 3 d²u/dx² with boundary conditions u(0,t) = -1, u(π,t) = 1, and initial condition u(x,0) = -cos(7x). The suggested approach involves using a steady-state solution to achieve a homogeneous initial condition. The method of separation of variables is employed, leading to the substitution u(x,t) = v(x,t) + Ψ(x) to simplify the boundary conditions. The final solution is expressed as u(x,t) = v(x,t) + Ψ(x), where v(x,t) is solved using standard techniques.

PREREQUISITES
  • Understanding of partial differential equations (PDEs)
  • Familiarity with boundary value problems
  • Knowledge of separation of variables technique
  • Experience with steady-state solutions in PDEs
NEXT STEPS
  • Study the method of separation of variables in detail
  • Learn about steady-state solutions in the context of PDEs
  • Explore boundary value problems and their solutions
  • Investigate the implications of homogeneous and non-homogeneous boundary conditions
USEFUL FOR

Mathematicians, physicists, and engineers working with partial differential equations, particularly those focused on boundary value problems and initial conditions in mathematical modeling.

selzer9
Messages
2
Reaction score
0
I'm not sure how to solve this:

du/dt = 3 [itex]\frac{d^{2}u}{dx^{2}}[/itex]

These are the conditions:
u(0,t)= -1
u(pi,t)= 1
u(x,0) = -cos 7x

Suggestion:
I should use steady state solution to get a homogeneous initial condition.

Starting with separtion of variables

u(x,t) = G(x)H(t)

And du/dt = GH' and d^2u/dx^2 = G"H

So GH'= c^2 G"H

and H'/(c^2)H = G"/G

Then H'/(c^2)H = k = G"/G

thus H' - (c^2)Hk = 0 and G" - Gk = 0.

Once here how do I make sure u(x,t) follows the conditions?
 
Physics news on Phys.org
selzer9 said:
I'm not sure how to solve this:

du/dt = 3 [itex]\frac{d^{2}u}{dx^{2}}[/itex]

These are the conditions:
u(0,t)= -1
u(pi,t)= 1
u(x,0) = -cos 7x

Suggestion:
I should use steady state solution to get a homogeneous initial condition.

Starting with separtion of variables

u(x,t) = G(x)H(t)

And du/dt = GH' and d^2u/dx^2 = G"H

So GH'= c^2 G"H

and H'/(c^2)H = G"/G

Then H'/(c^2)H = k = G"/G

thus H' - (c^2)Hk = 0 and G" - Gk = 0.

Once here how do I make sure u(x,t) follows the conditions?

I wouldn't start it that way. You need to make the boundary conditions homogeneous before you separate the variables. I would start with the substitution$$
u(x,t) = v(x,t) + \Psi(x)$$ Substituting that into your system gives$$
v_t = 3v_{xx}+3\Psi''(x)$$ $$
v(0,t) + \Psi(0)= -1$$ $$
v(\pi, t) +\Psi(\pi) = 1$$ $$
v(x,0) + \Psi(x) = -\cos x$$
Now if you let ##3\Psi''(x) = 0,\, \Psi(0) = -1,\, \Psi(\pi)= 1##, you can solve for ##\Psi(x)## and you are left with the homogeneous system$$
v_t = 3v_{xx}$$ $$
v(0,t) = 0$$ $$
v(\pi, t) = 0$$ $$
v(x,0) = -\cos x-\Psi(x)$$Solve this system in the usual way with separation of variables. Once you know ##v(x,t)##, your solution will be ##u(x,t)=v(x,t)+\Psi(x)##.
 
Last edited:

Similar threads

  • · Replies 3 ·
Replies
3
Views
2K
Replies
4
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K
Replies
2
Views
2K
Replies
5
Views
2K
Replies
5
Views
2K
Replies
6
Views
1K
Replies
6
Views
3K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K