1. PF Contest - Win "Conquering the Physics GRE" book! Click Here to Enter
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Integral substitution?

  1. Apr 5, 2014 #1
    1. The problem statement, all variables and given/known data

    sorry if question is unclear can't draw the integal sign out

    Show that
    Integral infinity-0 dz/((e^2z) - 1)^1/2 = integral 1- 0 dx/(1-x^2)^1/2 = pi/2

    3. The attempt at a solution

    I can get from the second integral to pi/2, as the second integral is sin^1(1) = pi/2

    However, I do not understand how to go between these two integrals

    infinity-0 dz/((e^2z) - 1)^1/2 = 1- 0 dx/(1-x^2)^1/2

    I tried substituing things like x = z, but doesn't work. Can't see how you would change the limits from infinity to 0, to 1 to 0?

    Maths is not my strong point so this could well be quite simple

    any help much appreciated
  2. jcsd
  3. Apr 5, 2014 #2


    User Avatar
    Homework Helper

    Hint: the denominator can be expressed as ##\displaystyle (e^z + 1)^{\frac{1}{2}}.(e^z - 1)^{\frac{1}{2}} = e^z.(1 + e^{-z})^{\frac{1}{2}}.(1 - e^{-z})^{\frac{1}{2}}##

    Does that help?
  4. Apr 5, 2014 #3
    Thank you for the help!

    I've subsituted x = e^-z

    so dz = dx/-e^-z

    integral becomes

    1/e^z(1+x)^1/2 (1-x^1/2) * dx/-e^-z

    e^z*-e^-z = 1

    so 1/(1+x)^1/2 (1-x^1/2) dx = 1/(1-x^2)^1/2

    and then e^-z = x e^-infinity = 1 hence new limits 1 to 0

    I think that works?
  5. Apr 5, 2014 #4


    User Avatar
    Homework Helper

    Very difficult to read your post without LaTex.

    But this: "e^z*-e^-z = 1" is an error, because the result should be negative one.

    And that negative sign is important when you transform the bounds.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted