Generalized functions (distributions) problem - Mathematical physics

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
3 replies · 3K views
Sigurdsson
Messages
24
Reaction score
1

Homework Statement


Find a distribution [tex]g_n[/tex] which satisfies
[tex]g'_n(x) = \delta(x - n) - \delta(x + n)[/tex]
and use it to prove
[tex]\lim_{n \to \infty} \frac{\sin{nx}}{\pi x} = \delta(x)[/tex]


Homework Equations


Nothing relevant comes up at the moment.


The Attempt at a Solution


Well the first part is pretty easy I think. The distribution would be
[tex]g_n(x) = \theta(x - n) - \theta(x + n) = \left\{ \begin{array}{l l}<br /> -1 & \quad |x| < n \\<br /> 0 & \quad |x| \geq n \\<br /> \end{array} \right.[/tex]


The limit will indeed resemble a delta function when [tex]n[/tex] goes to infinity and π is probably just a normalization constant. But applying the two Heaviside functions to solve this has got me stumped.

P.S. Gotta catch some sleep, I will be back in 7 hours hopefully with some ideas to solve this.

Cheers
 
Physics news on Phys.org
Thanks Vela! That did the trick, I'll post the full solution later. :)
 
Right guys so here it is, me and my buddies came up with this solution to the problem.

So we have established that our distribution should be
[tex]g_n(x) = \theta(x - n) - \theta(x + n)[/tex]
Like Vela said, the Fourier transform of this sine resembles our distribution differentiated. So that means we have (using Euler equation for sine)
[tex]\frac{d}{d\xi} \mathcal{F} \left\{ \frac{\sin{(nx)}}{\pi x} \right\}(\xi) = \frac{d}{d \xi} \int \frac{1}{\pi x} e^{i \xi x} \left( \frac{e^{inx} - e^{-inx} }{2i} \right) dx = \int \frac{i x}{2 i \pi x} e^{i\xi x} (e^{inx} - e^{-inx}) dx[/tex]
[tex]= \frac{1}{2\pi} \int e^{ix(\xi +n)} - e^{ix(\xi - n)} dx = \frac{2\pi}{2\pi} \left( \delta(\xi + n) - \delta(\xi - n) \right) = \frac{d}{d\xi} \left( \theta(\xi + n) - \theta(\xi - n) \right) = \frac{d}{d\xi} (- g_n(\xi))[/tex]
If we look at the limit of our distribution. You can easily see that it will become unity. (You can plot those heaviside functions and see for yourself).
[tex]\lim_{n \to \infty} (- g_n(\xi)) = 1[/tex]
Using inverse Fourier-Transform we have
[tex]\mathcal{F}^{-1} \mathcal{F} \left\{ \frac{\sin{(nx)}}{\pi x} \right\}(\xi) = \mathcal{F}^{-1} \{1\}(x) = \delta(x)[/tex]
Which ultimately proves what was supposed to prove
[tex]\lim_{n \to \infty} \frac{\sin{(n x)}}{\pi x} = \delta(x)[/tex]


gg