next up previous
Next: The wave equation Up: The diffusion equation Previous: An example 2-d solution

3-d problems

The techniques discussed above for solving the diffusion equation in two dimensions with a restricted class of boundary conditions can easily be generalized to three dimensions. In the 3-d case, it is necessary to Fourier transform in two directions (the $y$ and $z$ directions, say) in order to reduce the problem to a system of uncoupled tridiagonal matrix equations. These equations can be inverted in the usual manner, and the solution can then be reconstructed via a double inverse Fourier transform.

Richard Fitzpatrick 2006-03-29