Wikipedia:Reference desk/Archives/Mathematics/2012 September 29

= September 29 =

Solutions to Laplace's equation
My electromagnetism text derives a formula for azimuthally-symmetric solutions to Laplace's equation in spherical coordinates by separation of variables ie it first finds solutions of the form $$V(r,\theta)=R(r)\Theta(\theta)$$. Plugging this into Laplace's equation, it gets two ODEs: $$\frac{d}{dr}\left(r^2\frac{dR}{dr}\right)=kR$$ and $$\frac{1}{\sin\theta}\frac{d}{d\theta}\left(\sin\theta\frac{d\Theta}{d\theta}\right)=-k\Theta$$.

Now at this point, my text asserts that k is positive, without qualification or justification. Can anyone explain why k has to be positive? 74.15.136.9 (talk) 19:08, 29 September 2012 (UTC)
 * You could greatly help yourself specifying in which domain do you solve Laplace's equation, and with which boundary conditions. It is not easy to extract this information telepathically. Incnis Mrsi (talk) 19:44, 29 September 2012 (UTC)
 * Domain and boundary conditions are not being specified. The text is deriving a general formula for any azimuthally symmetric potential in spherical coordinates. The formula it gets is $$V(r,\theta)=\sum_{l=0}^\infty \left(A_lr^l+\frac{B_l}{r^{l+1}}\right)P_l(\cos\theta)$$, where $$P_l$$ is the lth Legendre polynomial. — Preceding unsigned comment added by 74.15.136.9 (talk) 23:48, 29 September 2012 (UTC)

k is positive because the second equation needs to have periodic solutions for &Theta;. Sławomir Biały (talk) 00:49, 30 September 2012 (UTC)
 * Could you explain that in a little more detail please? 74.15.136.9 (talk) 01:56, 30 September 2012 (UTC)
 * Multiply both sides by $$\Theta\sin^2\theta$$ and integrate from zero to 2&pi;. Integrate by parts.  Periodicity implies the boundary terms cancel.  Enforcing that both sides have the same sign forces k>0.  Sławomir Biały  (talk) 00:46, 1 October 2012 (UTC)

Legendre Polynomials
A Legendre polynomial $$P_n(x)$$ is supposed to satisfy the equation $${d \over dx} \left[ (1-x^2) {d \over dx} P_n(x) \right] + n(n+1)P_n(x) = 0$$. Substituting x = 1 into the above equation leads to the conclusion that $$P_n(1)=0$$. But $$P_n(x)=1$$. So what gives? 74.15.136.9 (talk) 19:44, 29 September 2012 (UTC)
 * Is this what you are doing, for $$x=x$$ at $$x=1$$:
 * $$x=1$$
 * Taking the derivative of each side,
 * $$\frac{d}{dx}x = \frac{d}{dx}1$$
 * Therefore
 * $$1=0.$$
 * Q.E.D.
 * The derivative is the slope of the function not a property of its value at a point. Dmcq (talk) 20:16, 29 September 2012 (UTC)
 * Yikes, what a stupid mistake. Thank you. 74.15.136.9 (talk) 23:44, 29 September 2012 (UTC)