Solve a very complex line integral

Mathman23
Messages
248
Reaction score
0

Homework Statement



Solve I = \int_{\gamma} f(z) dz where \gamma(t) = e^{i \cdot t} and 0 \leq t \leq \pi

Homework Equations



Do I use integration by substitution??

The Attempt at a Solution



If I treat this as a line-integral I get:

I = \int_{a}^{b} f(\gamma(t)) \cdot \gamma'(t) dt = \int_{0}^{\pi} cos(e^{i \cdot t}) \cdot (i \cdot e^{i \cdot t}) dt

then If I choose

u = e^{it} and dt = i \cdot e^{i \cdot u} du

where t=0, u=e^{i \cdot 0} = 1 and t=\pi, u=e^{i \cdot 0} = -1

\int_{0}^{\pi} cos(e^{i \cdot t}) \cdot (i \cdot e^{i \cdot t}) dt = \int_{1}^{-1} (i \cdot (cos(u) \cdot u) du = (i \cdot (cos(u) + u \cdot sin(u))) \cdot i \cdot e^{i \cdot u} ]_{1}^{-1} = 2 sin(1) \cdot ((cos(1) + sin(1)) \cdot i

Am I on the right path here??

Best regards
Fred
 
Last edited:
Physics news on Phys.org
Mathman23 said:

Homework Statement



Solve I = \int_{\gamma} f(z) dz where \gamma(t) = e^{i \cdot t} and 0 \leq t \leq \pi

Homework Equations



Do I use integration by substitution??

The Attempt at a Solution



If I treat this as a line-integral I get:

I = \int_{a}^{b} f(\gamma(t)) \cdot \gamma'(t) dt = \int_{0}^{\pi} cos(e^{i \cdot t}) \cdot (i \cdot e^{i \cdot t}) dt
So f(z)= cos(z)? You didn't mention that before! Cosine is an analytic function isn't it?

then If I choose

u = e^{it} and dt = i \cdot e^{i \cdot u} du

where t=0, u=e^{i \cdot 0} = 1 and t=\pi, u=e^{i \cdot 0} = -1

\int_{0}^{\pi} cos(e^{i \cdot t}) \cdot (i \cdot e^{i \cdot t}) dt = \int_{1}^{-1} (i \cdot (cos(u) \cdot u) du = (i \cdot (cos(u) + u \cdot sin(u))) \cdot i \cdot e^{i \cdot u} ]_{1}^{-1} = 2 sin(1) \cdot ((cos(1) + sin(1)) \cdot i

Am I on the right path here??

Best regards
Fred
That might work but looks overly complicated! How about just finding the anti-derivative of cos(z) (I'm sure you know it!) and evaluating it at the two endpoints of the contour?
 
Hi there,

You are right :)

Then I get

$<br /> \int_0^{\pi}Cos[e^{it}]ie^{it}dt

and I let u=e^{it}

then:

du=ie^{it}dt

right?

so that part is already in the integral so now it's just:

$<br /> \int_0^{\pi}Cos[e^{it}]ie^{it}dt=\int_1^{-1}Cos<u>du</u>

right? I then get -2Sin[1]

I got that result in Maple too. Had an idear that it was the right one :)

By following the method above. I get:

1)f(z) = sinh(z) and 2) f(z) = (exp(z))^3 and 3)f(z) = tan(z).

1) Solution

\int_{0}^{\pi} sinh(e^{it}) \cdot i \cdot e^{i \cdot t} dt = \int_{1}^{-1} sinh(u) du = 0

2) Solution

\int_{0}^{\pi} (e^{e^{it}})^3 \cdot i \cdot e^{i \cdot t} dt = \int_{1}^{-1} (e^{u})^3 du = \frac{-2 sinh(3)}{3}

3) Solution

\int_{0}^{\pi} tan(e^{it}) \cdot i \cdot e^{i \cdot t} dt = \int_{1}^{-1} tan(u) du = 0

How do they look? Am I following you guidelines correctly??

Best regards Fred
 
dt = i \cdot e^{i \cdot u} du

This should be du = i \cdot e^{i \cdot t} dt

I think...

notice your integral becomes easier
 
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