What is the Integral of Modified Bessel Function using Integral Representation?

  • Thread starter Thread starter qasdc
  • Start date Start date
  • Tags Tags
    Bessel Integral
Click For Summary

Homework Help Overview

The discussion revolves around evaluating the integral of the modified Bessel function \( K_{0}(kr) \) using its integral representation. Participants are exploring the integral \(\int_{0}^{\infty} dk K_{0}(kr)\) and its relation to the cosine integral representation of \( K_{0}(x) \).

Discussion Character

  • Exploratory, Mathematical reasoning, Assumption checking

Approaches and Questions Raised

  • Participants discuss the use of integral representations and convergence techniques for evaluating the integral. There are attempts to clarify the correct form of the exponential terms and the implications of changing variables. Questions arise regarding the steps following the substitution for \(\cos(xt)\) and the handling of Gaussian integrals.

Discussion Status

Some participants have provided guidance on variable changes and the use of the delta function in the context of the integral. There is an ongoing exploration of different approaches, with no explicit consensus reached on the final steps of the evaluation.

Contextual Notes

Participants express a need to work through the integral step by step, indicating a preference for understanding the process rather than relying on computational tools. There are references to specific limits and conditions that may affect the evaluation.

qasdc
Messages
14
Reaction score
0

Homework Statement



I need to evaluate the following integral:
\int_{0}^{\infty} dk K_{0}(kr)
, where K_{0}(x) is the modified Bessel, using the integral representation:
K_{0}(x)=\int_{0}^{\infty} dt \frac{cos (xt)}{ \sqrt{t^2 +1}}

Homework Equations


The Attempt at a Solution

 
Last edited:
Physics news on Phys.org
This is a standard integral that can be found in tables. If you want to explicitly do the integral, you should write

\cos xt = \frac{1}{2} \lim_{\epsilon \rightarrow 0} \left( e^{-\epsilon x^2 +i xt} + e^{-\epsilon x^2 -i xt} \right)

to improve the convergence of the x integral. You will be able to set \epsilon = 0 after making a change of variables for the t integral without encountering any divergences.
 
fzero thank you for your reply.
Yes, I need to do the intergal step by step, otherwise I could get it from mathematica.
The answer I get form mathematica is \frac{\pi}{2r}
In your reply, I suppose that the first term in each exponential is - \epsilon t^2 rather than - \epsilon x^2.
How should I proceed after the substitution for \cos (xt)?
 
qasdc said:
fzero thank you for your reply.
Yes, I need to do the intergal step by step, otherwise I could get it from mathematica.
The answer I get form mathematica is \frac{\pi}{2r}
In your reply, I suppose that the first term in each exponential is - \epsilon t^2 rather than - \epsilon x^2.
How should I proceed after the substitution for \cos (xt)?

I did the x integration first, so I wanted those to be Gaussian integrals, hence you need - \epsilon x^2, not - \epsilon t^2. The x integral is then a Gaussian (you need to complete a square), you write the results down in terms of \epsilon. The t-integral also becomes a Gaussian, but you can simplify it by changing variables and taking \epsilon to zero.
 
So,
\int_{0}^{\infty} dk K_{0}(kr)=\frac{1}{r}\int_{0}^{\infty} dx K_{0}(x)=\frac{1}{r}\int_{0}^{\infty} dx \int_{0}^{\infty} dt \frac{\cos (xt)}{\sqrt{t^2 +1}}=
=\frac{1}{4r}\int_{-\infty}^{\infty} dx \int_{-\infty}^{\infty} dt \frac{\cos (xt)}{\sqrt{t^2 +1}}= \frac{1}{4r}\int_{-\infty}^{\infty} dx \int_{-\infty}^{\infty} dt \frac{1}{\sqrt{t^2 +1}}\frac{1}{2} \lim_{\epsilon \rightarrow 0} \left( e^{-\epsilon x^2 +i xt} + e^{-\epsilon x^2 -i xt} \right)=
=\frac{1}{4r} \lim_{\epsilon \rightarrow 0} \frac{\sqrt{\pi}}{\sqrt{\epsilon}}\int_{-\infty}^{\infty}dt \frac{e^{\frac{-t^2}{4\epsilon}}}{\sqrt{t^2+1}}

Now, at this point I do not know how to proceed. Any help?
 
Change variables to s=t/(2\sqrt{\epsilon}).
 
I found it.
Using the following definition of the delta function:
\lim_{\epsilon \rightarrow 0} \frac{1}{\sqrt{\epsilon}}e^{-\frac{t^2}{4\epsilon}} =2\sqrt{\pi}\delta(t)
we find that,
<br /> =\frac{1}{4r} \lim_{\epsilon \rightarrow 0} \frac{\sqrt{\pi}}{\sqrt{\epsilon}}\int_{-\infty}^{\infty}dt \frac{e^{\frac{-t^2}{4\epsilon}}}{\sqrt{t^2+1}} =\frac{\sqrt{\pi}}{4r} \int_{-\infty}^{\infty}dt \frac{1}{\sqrt{t^2+1}}2\sqrt{\pi}\delta(t) = \frac{\pi}{2r} <br />

Thanks a lot for your help fzero!
 

Similar threads

  • · Replies 9 ·
Replies
9
Views
2K
Replies
1
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 19 ·
Replies
19
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
Replies
6
Views
3K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 14 ·
Replies
14
Views
2K