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

Homework Help: Line Integrals (Complex Variables)

  1. Oct 16, 2008 #1
    1. The problem statement, all variables and given/known data
    i) Define a path [itex]\gamma[/itex] whose image is the ellipse [tex]\frac{x^2}{a^2}[/tex] + [tex]\frac{y^2}{b^2}[/tex] = 1 traced counterclockwise.
    ii) Show that [itex]\int[/itex] [itex]\frac{1}{z}[/itex] dz = [itex]\int[/itex] [itex]\frac{1}{z}[/itex] dz for a suitable circle [itex]\beta[/itex]
    iii) Use the above result (part (ii)) to show that [itex]\int[/itex] [itex]\frac{1}{a^2 cos^2(t) + b^2 sin^2(t)}[/itex] dt = [itex]\frac{2\pi}{ab}[/itex] for a>0, b>0.

    2. Relevant equations
    Green's Theorem, Cauchy's Theorem... Cauchy's formula too maybe?

    3. The attempt at a solution
    Ok, so first of all I parametrized the ellipse, I got the following:

    [itex]\gamma[/itex] (t) = t - i(b/a) [itex]\sqrt{a^2 - t^2}[/itex] for -a[itex]\leq[/itex] t [itex]\leq[/itex] a
    [itex]\gamma[/itex] (t) = (2a-t) + i(b/a) [itex]\sqrt{a^2 - (2a-t)^2}[/itex] for a[itex]\leq[/itex] t [itex]\leq[/itex] 3a

    I've checked and I'm pretty sure that is a correct parametrization for the ellipse given.
    Now, my real problem is with ii) and iii). For ii) I'm thinking about using Cauchy's Theorem maybe... because the function 1/z is analytic everywhere except at the origin, so a line integral over an ellipse centered at the origin and a circle that's also centered at the origin would be the same (since the function is analytic everywhere in the space between the two paths). However, I'm not that sure on how to "show" that, mathematically. The corrector checks the homeworks very strictly so I need to show it very clearly. Any ideas? As for iii) well I need ii) first...

    Any suggestions are greatly appreciated. Cheers.
  2. jcsd
  3. Oct 17, 2008 #2


    User Avatar
    Science Advisor

    I wouldn't have used that parameterization. I would have used x= a cos(t), y= b sin(t) so that [itex]\gamma(t)= a cos(t)+ i b sin(t)[/itex]. You should find that simpler. The [itex]dz/z[/itex] becomes [itex](-a sin(t)+ i b cos(t))dt/(a cos(t)+ i b sin(t))[/itex]. Try multiplying both numerator and denominator by the complex conjugate of the denominator.
  4. Oct 17, 2008 #3
    Oh, thanks Halls, that was great help.
    How did you get that parametrization, are they simlpy obtained from the the elliptic coordinate transformations?
    I found this on wikipedia:
    "The most common definition of elliptic coordinates (μ,ν) is
    x = a cosh [itex]\mu[/itex] cos v
    y = a sinh [itex]\mu[/itex] sin v "
    Can you show me how to get to your parametrization?

    Thanks again.
    Last edited: Oct 17, 2008
  5. Oct 17, 2008 #4


    User Avatar
    Science Advisor

    Those are ellipitic coordinates in a plane, not a parameterization of an ellipse. What does Wikipedia say [itex]\mu[/itex] and [itex]\nu[/itex] represent there?

    My derivation starts from cos2(t)+ sin2(t)= 1 Since the equation of the ellipse is x2/a2+ y2/b2= 1, letting x/a= cos(t) and y/b= sin(t) should be obvious and that gives x= a cos(t), y= b sin(t).
    Last edited by a moderator: Oct 17, 2008
  6. Nov 3, 2010 #5
    How can I evaluate integral 1/z over the ellipse?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook