Вы находитесь на странице: 1из 2

Numerical Methods for Differential Equations

and Applications
Camile Fraga Delfino Kunz
13th December 2018

2D Poisson problem
Consider a quantity given by a function u that diffuses on a 2D plane. It can
be a diffusion of a substance, of heat, or any other quantity that behaves like a
diffusion process. We already saw that the diffusion process can be modeled by
the equation
∂u
= D∆u + f (u, x, y, t)
∂t
where ∆ is the Laplace’s operator, D is the diffusion coefficient, f is a source
(f > 0) or sink (f < 0) function, (x, y) are on a spatial domain Ω, the time t is
on a interval (0, T ], and we have boundary conditions for (x, y) ∈ ∂Ω and initial
condition on t = 0.
If we are interested on the steady state, that is, when the concentration (or
temperature) doesn’t change ∂u/∂t = 0, we would have the following boundary
value problem (BVP),
 2 2
 ∂ u + ∂ u + = f (x, y)
∂x 2 ∂y 2
u(x, y) = α , (x, y) ∈ ∂Ω

note that here on this case we have Dirichlet boundary conditions.

1
Exercise
With the discrete grid xi = x0 + ih, i = 1, . . . , Nx , and yj = y0 + jh, j =
1, . . . , Ny , where h = xi − xi−1 = yj − yj−1 , and x and y both on the interval
[0, 1], solve the BVP using 4-point finite differences:
1 1
2
(Ui−1,j − 2Ui,j + Ui+1,j ) + (Ui,j−1 − 2Ui,j + Ui,j+1 ) = fij
h h
where f is nonzero only on a single point of the grid, let us say,

F (f loor((N y ∗ N x)/4), 1) = −0.1;

References
1. Edelstein-Keshet, L. Mathematical models in biology. SIAM, 2005.
2. Leveque, R. J. Finite difference methods for ordinary and partial differen-
tial equations. SIAM, 2007.

Вам также может понравиться