Integral Calculation with Complex Analysis - Can Residue Theorem Help?

asi123
Messages
254
Reaction score
0
Hey guys.
I need to calculate this integral so I was thinking about using the residue theorem.
The thing is that the point 0 is not enclosed within the curve that I'm about to build, it's on it.
Can I still use the theorem?

Thanks a lot.
 

Attachments

  • 1.jpg
    1.jpg
    1.5 KB · Views: 384
Physics news on Phys.org
So use Cauchy Principal value to bypass the singularity at 0.
 
You have a bigger problem then that, the residue at x=0 is zero (removeable singularity). This is a standard example and is often done as follows
consider
To fix the residue=0 problem add in an odd function the usual one is to use
exp(x i)/x=cos(x)/x+i sin(x)/x
the cos(x)/x diverges as an improper integral, but as lqg states cauch principle sense may be used
now consider ther contour formed by theese pieces
(-R,r) straight line
(-r,r) semicircle arc about z=0
(r, R) straight line
(R,r) semicircle arc about x=0
 
lurflurf said:
You have a bigger problem then that, the residue at x=0 is zero (removeable singularity). This is a standard example and is often done as follows
consider
To fix the residue=0 problem add in an odd function the usual one is to use
exp(x i)/x=cos(x)/x+i sin(x)/x
the cos(x)/x diverges as an improper integral, but as lqg states cauch principle sense may be used
now consider ther contour formed by theese pieces
(-R,r) straight line
(-r,r) semicircle arc about z=0
(r, R) straight line
(R,r) semicircle arc about x=0

I didn't quite understand the contour you described there.
What's R and What's r?

Thanks a lot.
 
asi123 said:
I didn't quite understand the contour you described there.
What's R and What's r?

Thanks a lot.

Take limits R-> +infinity,r->+0
If you draw a pictuis an upper semicircle that in the limit is large with a tiny semicircle at the origin then you get for the various integrals

(-R,r) straight line
-infinity+(Integral you want)/2
(-r,r) semicircle arc about z=0
{+,-}[+ if it was upper - if it was lower] pi*i (residue theorem)
(r, R) straight line
-infinity+(integral you want)/2
(R,r) semicircle arc about x=0
0
 
lurflurf said:
Take limits R-> +infinity,r->+0
If you draw a pictuis an upper semicircle that in the limit is large with a tiny semicircle at the origin then you get for the various integrals

(-R,r) straight line
-infinity+(Integral you want)/2
(-r,r) semicircle arc about z=0
{+,-}[+ if it was upper - if it was lower] pi*i (residue theorem)
(r, R) straight line
-infinity+(integral you want)/2
(R,r) semicircle arc about x=0
0

Is it something like that?

Thanks again.
 

Attachments

  • scan0011.jpg
    scan0011.jpg
    9.7 KB · Views: 359
asi123 said:
Is it something like that?

Thanks again.

That is it. I guess that is one reasone why some people do not like old math books with no pictures. Were you able to finish?
 
lurflurf said:
That is it. I guess that is one reasone why some people do not like old math books with no pictures. Were you able to finish?

I'll try, if I'll have some troubles, I'll be back :smile:.
Thanks a lot.
 
Back
Top