Laplace's equation on an annulus with Nuemann BCs

  • Thread starter Thread starter nathan12343
  • Start date Start date
  • Tags Tags
    Laplace's equation
nathan12343
Messages
26
Reaction score
0

Homework Statement


Solve Laplace's equation inside a circular annulus (a<r<b) subject to the boundary conditions \frac{\partial{u}}{\partial{r}}(a,\theta) = f(\theta)\text{, }\frac{\partial{u}}{\partial{r}}(b,\theta) = g(\theta)


Homework Equations


Assume solutions of the form u(r,\theta) = G(r)\phi(\theta). This leads to an equation for phi and G, both of which can be solved by substitution:

<br /> \frac{d^2\phi}{{d\theta}^2} = -\lambda^2\phi<br />
<br /> \phi = A\cos{\lambda\theta} + B\sin{\lambda\theta}<br />

<br /> r^2\frac{d^2G}{{dr}^2} + r\frac{dG}{dr} - n^2G = 0<br />
<br /> G = c_{1n}r^{-n} + c_{2n}r^n\text{ for } n\ne0<br />
<br /> G = c_{10} + c_{20}\ln(r)\text{ for } n=0<br />

The Attempt at a Solution


Periodic boundary conditions require that u and its derivative with respect to theta be continuous between -pi and pi. Then means \lambda = n for all n greater than or equal to zero. We can write,

<br /> u(r,\theta) = &amp;\, A_{01} + A_{02}\ln(r) + \sum_{n=1}^{\infty}(A_{n1}r^n + A_{n2}r^{-n})\cos{n\theta} + (B_{n1}r^n + B_{n2}r^{-n})\sin{n\theta}<br />
<br /> \frac{\partial{u(r,\theta)}}{\partial{r}} = &amp;\, A_{02}r^{-1} + \sum_{n=1}^{\infty}(nA_{n1}r^{n-1} + -nA_{n2}r^{-n-1})\cos{n\theta} + (nB_{n1}r^{n-1} - nB_{n2}r^{-n-1})\sin{n\theta} <br />

I know that I can set everything in parentheses in front of the cosines and sines above equal to some constant when I set r=a,b to enforce the boundary conditions on the edge of the annulus. This leaves me with two equations in two unknowns for all the A's and B's with n greater than 1. My problem is how to set A_{02}[/tex] such that it will work at both boundaries. It seems like I really need two constants to match to the boundary for when n=0. Am I missing something?<br /> <br /> Thanks for your help!
 
Physics news on Phys.org
Can anyone help?
 
Really? It seems like a simple problem...
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top