The Question:

(i) Find the Laplace Transform of the Bessel Function $J_0(x)$

(ii) Hence, show that if $f(x)$ satisfies the differential equation

$$f''(x)+f(x)=J_0(x) \qquad f(0)=f'(0)=0$$

then $f$ is given by

$$f(x) = \int_0^x tJ_0(t)dt$$

(i) I found $\overline{J_0}(p) = \dfrac{1}{\sqrt{1+p^2}}$ which is definitely correct

(ii) I Laplace Transformed both sides of the equation and applied the boundary conditions to get

$$\bar f(p) = \frac{\overline{J_0}(p)}{1+p^2} = \big(1+p^2\big)^{-3/2}$$

It is not immediately clear how to Laplace Invert $(1+p^2)^{-3/2}$ and it probably won't give the required integral in terms of $J_0$ anyway.

Instead, I know that the Laplace Transform of $\sin(x)$ is precisely $(1+p^2)^{-1}$, i.e. we have

$$\bar f(p) = \overline {J_0} (p) \overline{\sin}(p)$$

So by the convolution theorem,

$$f(x) = \big(J_0 * \sin \big)(x) = \int_0^x \sin (x-t) J_0(t)dt$$

and I don't see how in the world you can make the $\sin(x-t)$ a $t$ as required.

Have I done something wrong, or is there a Bessel function identity I am unaware of?

  • $\begingroup$ Ping me When answered. +1. $\endgroup$ – mick Jun 2 '18 at 23:47
  • 1
    $\begingroup$ *ping ping ping* $\endgroup$ – glowstonetrees Jun 3 '18 at 12:20

The calculations are correct. In addition, you mat recognize that \begin{equation} \frac{1}{\left( p^2+1 \right)^{3/2}}=-\frac{1}{p}\frac{d}{dp}\frac{1}{\sqrt{p^2+1}} \end{equation} Then, using successively the frequency domain derivative and the time domain integration formulae for the Laplace transform, you find the expected result. In fact, you derived the general result \begin{equation} \int_0^x \sin (x-t) J_0(t)\,dt= \int_0^x tJ_0(t)\,dt \end{equation} Moreover, using the properties of the Bessel functions, it can be shown that $f(x)=xJ_1(x)$.


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.