Confusion about Separation of Variable with Periodic Boundary Conditions

71 Views Asked by At

I'm terribly confused about the separation of variable method for PDE (heat eq particularly). The given conditions are:

$u_t(x,t)=u_{xx}(x,t), u(0,t)=u(L,t), u_x(0,t)=u_x(L,t), u(x,0)=f(x)$

Since I know the general cases, the general $u(x,t)$ should look like

$u(x,t)=\frac{a_0}{2}+\sum_{n=1}^{\infty} \left\{A_n\cos{\frac{2n{\pi}x}{L}+B_n}\sin{\frac{2n{\pi}x}{L}} \right\}e^{-(2n\pi)^2t}$

But I cannot reach this conclusion, and even worse, I have no idea about this part:

$X(0)=X(L), X'(0)=X'(L)$ follows

$A=X(0)=X(L)=A\cos{\sqrt{\lambda}L}+B\sin{\sqrt{\lambda}L}$

and

$\sqrt{\lambda}B=X'(0)=X'(L)=-\sqrt{\lambda}A\sin{\sqrt{\lambda}L}+\sqrt{\lambda}B\cos{\sqrt{\lambda}L}$

How can I conclude $\sqrt{\lambda}=2n\pi$ from these conditions above? It is clear when $-L\leq x\leq L$ since I can use the property of trig functions and simplify the expressions. But I have no idea when $0\leq x \leq L$.

1

There are 1 best solutions below

0
On

So basically, when we plug in the boundary conditions we want a non trivial solution, such that either $A\neq0$ or $B\neq0$. If we look at the equations we can write them as $$ \underbrace{\begin{pmatrix} \cos(\sqrt{\lambda}L)-1 & \sin(\sqrt{\lambda}L)\\ -\sqrt{\lambda}\sin(\sqrt{\lambda}L)&\sqrt{\lambda}\cos(\sqrt{\lambda}L)-\sqrt{\lambda} \end{pmatrix}}_{=:M} \begin{pmatrix} A\\B \end{pmatrix}=0 $$

If we demand a nontrivial solution the determinant of $M$ shall be zero: $$ \det(M)=-2 \lambda (\cos(\lambda L)-1)\\ \Rightarrow \lambda=0\quad\lor\quad \lambda L=2\pi n, n\in\mathbb{N} $$

However for such $\lambda$ we see that is $M$ already the null matrix and therefore all $A,B$ lead to valid solutions.