Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Ploting zero order Bessel function

  1. Sep 30, 2004 #1
    Hello guys, i had a little chat with a teacher of mine and he asked me how can someone plot the zero order Bessel function. Here is what i've done..

    using the integral expresion for [tex]J_{0}(r)[/tex]
    [tex]J_{0}(r)=\frac {1}{\pi}\int_0^\pi \cos(r\cos\theta)d\theta[/tex]
    i can calculate the first order derivative with respect to r
    [tex]\frac {\partial}{\partial r}J_{0}(r)=-\frac {1}{\pi}\int_0^\pi \sin(r\cos\theta)\cos\theta d\theta[/tex]
    wich when evaluated in r=0 is 0. For the second derivative
    [tex]\frac {\partial^2}{\partial r^2}J_{0}(r)=-\frac {1}{\pi}\int_0^\pi \cos(r\cos\theta)\cos^{2}\theta d\theta[/tex]
    wich evaluated in r=0 is equal to -1/2.
    The idea is to construct the taylor series around r=0. And given the fact that
    i can easily bound the error, ie, if i only take two terms of the series
    [tex]|E(r)|\leq \frac{r^4}{4!\pi}[/tex]
    so, if, for instance, i want to know where is the first zero of the function, given the first approximation, i can say that is on 2 with an error of 0.21....
    given the next term
    [tex]|E(r)|\leq \frac{r^6}{6!\pi}[/tex]
    tells me that the zero is in 2^(3/2) with an error of 0.23
    and so on...

    do you guys think this is a correct procedure?

    is there any other way i can construct the plot?

    i really want to impress my teacher, so any help would be well received.

    Last edited: Sep 30, 2004
  2. jcsd
  3. Oct 1, 2004 #2


    User Avatar
    Science Advisor
    Homework Helper

    Have you thought about creating a table of Bessel functions?
  4. Oct 1, 2004 #3
    Looks good, you have derived the small series representation for Jo :

    \mbox{\Huge \[
    J_\nu (z) = \left( {\frac{z}{2}} \right)^\nu \sum\limits_{k = 0}^\infty {\frac{{\left( { - z^2 /4} \right)^k }}{{k!\Gamma (\nu + k + 1)}}}

    Unfortunately, this will only converge up to about 10 on the real axis with 10 signifigant digit calculator. For arguments larger an asymptotic form must be used : Hankel's Simiconvergent Asymptotic Expansion is probably the best.

  5. Oct 1, 2004 #4
    Heh... i know that. That was exactly the point my teacher wanted to make.

    He said to me: "ok, you know how it behaves for small values of r and for large values, but what hapens with regular values of r? how can you see the graph of the function?"

    But yeah, i think ill need more terms to get a nice aproximation of the first zero *at least*.

    Do u guys know any other method that i might consider?

    Thx for the help :)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook