diffusion

From Physics wiki

Jump to: navigation, search

Diffusion on a lattice

Consider particles restricted to moving on an d\,-dimensional lattice with spacing a\,, such that during a time interval \Delta t\, a particle may move to any adjacent lattice point with probability p\, \Delta t\, (but not diagonally, for simplicity). Let the number of particles at a point \mathbf{x}\, be given by n(\mathbf{x})\,. In one dimension,

n(x, t+\Delta t) = n(x,t) -2 p \Delta t n(x) +  p \Delta t \left[   n(x+a,t) + n(x-a,t)  \right]\,.

Let us represent n(x,t)\, by a continuous function that takes on the exact values on the lattice points. Then

n(x, t+\Delta t)\,  = n(x,t) + 2 p \Delta t \left[  a\frac{\partial n(x,t)}{\partial x} + \frac{a^2}{2}\frac{\partial^2 n(x,t)}{\partial x^2}  -a\frac{\partial n(x,t)}{\partial x}  + \frac{a^2}{2}\frac{\partial^2 n(x,t)}{\partial x^2}  + O(a^3)\right]\,.

Therefore,

\frac{\partial n(x,t)}{\partial t} = p a^2 \frac{\partial^2 n(x,t)}{\partial x^2} + O(a^3)\,.

The terms of O(a^3)\, vanish as a \to 0\, provided that n(x)\, is sufficiently regular. Generalizing to higher dimensions, then,

\frac{\partial n(\mathbf{x},t)}{\partial t} = p a^2 \nabla^2 n(\mathbf{x},t)\,.
Personal tools