My question is, how to solve an inhomogeneous heat equation on finite interval. The problem is $u_{t} - ku_{xx} = f(x, t) $ on finite interval with homogeneous Dirichlet boundary condition $ u(0, t) = u(L, t) = 0$ and initial condition $ u(x, 0) = \phi(x) $ in $ 0 < x < L$. My approach is using extended function that, solving $ u_x - ku_{xx} = f_{ext}(x, t)$ with homogeneous DC and initial condition $ u(x, 0) = \phi_{ext}(x)$ where $$ f_{ext} = \begin{cases} f(x, t) & \mbox{if } 0<x<L \\ -f(-x, t) &\mbox{if } -L < x< 0\\ \mbox{extended to be of period 2L} \end{cases}$$ and $\phi_{ext}$ defined likely. I think that the solution is, $$ u (x, t) = \int_{-\infty}^{\infty} S(x-y, t) \phi_{ext}(y) \, dy \, + \int_0^t \!\! \int_{-\infty}^{\infty} S(x-y, t-s) f_{ext}(y, s) \, dy \,ds.$$ $S(x, t) = \frac{1}{\sqrt{4\pi kt}} \exp(-\frac{x^2}{4kt})$ is heat kernel. I think that this solution satisfies boundary condition, since two extended functions are odd-symmetric to $x = 0$ and $x = L$. I wonder that whether this argument is valid. Further, can this method be applied for Neumann boundary and Robin boundary condition?


  • How to solve 1-D inhomogeneous heat equation in finite interval?
  • Is the method above valid?
  • If it is valid, can it be applied for other boundary conditions?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.