Related but not necessary to know: here

Looking at the temperature distribution in an infinitely long cylinder of metal with insulated sides and initial temperature distribution $f(x)= \left\{\begin{align}0,\quad|x|\lt L \\ C,\quad|x| \gt L \end{align} \right.$

$C$ is constant.

Now I want to workout $B_n$ for the fourier series, and I thought that I would want:

$$B_n = \frac2L\int_L^\infty C\sin\left(\frac{n\pi x}{L}\right) dx$$

But perhaps I haven't setup the integral correctly. Thank you for listening.

I based my choice off of the general form of the solution to the heat equation: $u(x,t)=\sum \limits_{n=1}^\infty B_n e^{({-n\pi C/L})^2} \sin\left(\frac{n\pi x}{L}\right)$ With $B_n = \frac2L \int_0^L \sin\left(\frac{n\pi x}{L}\right) f(x) dx$

  • 1
    $\begingroup$ It's not a periodic function, so why are you trying to compute a Fourier series to begin with? $\endgroup$ – Hans Lundmark Oct 19 '14 at 12:24
  • 1
    $\begingroup$ @HansLundmark My logic is added above $\endgroup$ – Analysis Oct 19 '14 at 12:30
  • 1
    $\begingroup$ That's the solution to the heat equation on an interval $[0,L]$. It's not going to help you if the interval is unbounded, as in your case. You know that a Fourier series is something that's associated to a periodic function, don't you? It just doesn't make sense to try to compute Fourier coefficients for a function that's not periodic. $\endgroup$ – Hans Lundmark Oct 19 '14 at 12:35
  • 1
    $\begingroup$ @HansLundmark That is a good point. I'll have to find some other way to find $u(x,t)$ thank you $\endgroup$ – Analysis Oct 19 '14 at 12:37
  • 2
    $\begingroup$ The Fourier transform should be more useful. But make the change of variables $v(x,t)=u(x,t)-C$ first, so that your initial data are zero for $|x| > L$, otherwise you're going to run into convergence problems again. $\endgroup$ – Hans Lundmark Oct 19 '14 at 12:48

It would seem you have a number of concepts confused, and I would personally suggest you refer to a textbook on this matter.

You want to use $$\frac{\partial u}{\partial t} = c^2 \frac{\partial^2 u}{\partial x^2}$$

which helps you obtain your standard $F''+p^2 F =0$ and $\dot{G} + c^2p^2 G=0$, if you can't get this you are in trouble.

$f(x)= \left\{\begin{align}0,\quad|x|\lt L \\ C,\quad|x| \gt L \end{align} \right.$

$C, L\lt x$ and $x\lt -L$

$u(x,t) = \frac{C}{2c\sqrt{\pi t}}\int_L^\infty$



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.