Solving Definite Integral with a, b Parameters

zeebek
Messages
27
Reaction score
0
This is not really a homework.

Homework Statement



I am trying to solve one definite integral

Homework Equations



\int_0^{2 \pi} \frac{\sin^2{t}}{\sqrt{a\cos{t} + b}} dt

where a, b are some positive numbers.

The Attempt at a Solution



I tried integrate by parts, also differentiate by parameter a. Does not help because of the square root. In this case I got some diff.equation for the integral I want, but I need to solve even more nasty integrals:

either

\int_0^{2 \pi} \sqrt{a\cos{t} + b} \cos{t} dt

or

\int_0^{2 \pi} \frac{1}{\sqrt{a\cos{t} + b}} dt
 
Physics news on Phys.org
You joking, right? Even by breaking this up into two separate integrals, the first one 1/sqrt(a*cos(t)+b), t=0..2pi and second one -cos^2(t)/sqrt(a*cos(t)+b), t=0..2pi.. This gives some complete elliptic integral of first one for first part, and series expansion for second part. At best you can get a complete series expansion of this with http://www.wolframalpha.com/input/?i=Expand+(1-cos^2(t))/sqrt(a*cos(t)+b)" and numerically evaluate it if you know a and b.
 
Last edited by a moderator:
cronxeh said:
You joking, right? Even by breaking this up into two separate integrals, the first one 1/sqrt(a*cos(t)+b), t=0..2pi and second one -cos^2(t)/sqrt(a*cos(t)+b), t=0..2pi.. This gives some complete elliptic integral of first one for first part, and series expansion for second part. At best you can get a complete series expansion of this with http://www.wolframalpha.com/input/?i=Expand+(1-cos^2(t))/sqrt(a*cos(t)+b)" and numerically evaluate it if you know a and b.

Thank you for your reply. As I said this is not a homework problem. In fact if somebody could help me represent this intergral via some generalized functions this would do. Sorry for the misconception.
 
Last edited by a moderator:
zeebek said:
Thank you for your reply. As I said this is not a homework problem. In fact if somebody could help me represent this intergral via some generalized functions this would do. Sorry for the misconception.

I can't see how this could be done. a and b make this impossible
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top