3
$\begingroup$

How to Prove: $$\int_{0}^{\infty} \log\biggl(x+\frac{1}{x}\biggr) \cdot \frac{1}{1+x^{2}} \ dx = \pi \: \log{2}$$

$\endgroup$
  • 1
    $\begingroup$ Please avoid titles that are entirely in $\TeX$. $\endgroup$ – J. M. is a poor mathematician Apr 20 '12 at 17:35
  • 2
    $\begingroup$ @J.M. I couldn't think of a better title. $\endgroup$ – peter Apr 20 '12 at 17:45
6
$\begingroup$

Put $x = \tan{\theta}$. Then you have

\begin{align*} I &= \int_{0}^{\pi/2} \log(\tan\theta + \cot\theta) \ d\theta \\\ &= \int_{0}^{\pi/2} \log\frac{2}{\sin{2\theta}} \ d\theta \\\ &= - \int_{0}^{\pi/2} (\log{\sin\theta} +\log{\cos\theta}) d\theta \\\ &= -2 \int\limits_{0}^{\pi/2} \log\: {\sin\theta} \ d\theta = \pi \: \log{2} \end{align*}

$\endgroup$
  • 1
    $\begingroup$ With regards to the log sine integral, see this article. $\endgroup$ – J. M. is a poor mathematician Apr 20 '12 at 17:50
  • $\begingroup$ @J.M. Yeah i think $\log{\sin{\theta}}$ was asked here sometime before. $\endgroup$ – user9413 Apr 20 '12 at 17:53
  • 1
    $\begingroup$ @Chandrasekhar : You might consider up-voting the question. $\endgroup$ – Michael Hardy Apr 20 '12 at 19:10
  • $\begingroup$ @Hardy: Why do you say like that $\endgroup$ – user9413 Apr 20 '12 at 20:00
5
$\begingroup$

Rewrite it as $$ I =\int_0^\infty \frac{\log\left(x+\frac{1}{x}\right)}{x+\frac{1}{x}} \frac{\mathrm{d} x}{x} = \left.\frac{\mathrm{d}}{\mathrm{d}s} \mathcal{I}(s)\right|_{s=-1} $$ where $$ \mathcal{I}(s) = \int_0^\infty \left( x+\frac{1}{x}\right)^s \frac{\mathrm{d} x}{x} \stackrel{x=\sqrt{\frac{u}{1-u}}}{=} \frac{1}{2} \int_0^{1} (u(1-u))^{-s/2-1} \mathrm{d} x $$ Thus $$ \mathcal{I}(s) = \frac{1}{2} \operatorname{B}\left(-\frac{s}{2},-\frac{s}{2} \right) = \frac{1}{2} \frac{\Gamma^2\left(-\frac{s}{2}\right)}{\Gamma(-s)} $$ It remains to evaluate the derivative at $s=-1$: $$ I = \left.\frac{\mathrm{d}}{\mathrm{d}s} \mathcal{I}(s)\right|_{s=-1} = \mathcal{I}(-1)\left( \psi(1) - \psi\left(\frac{1}{2}\right) \right) = \pi \log(2) $$ The result follows from $\mathcal{I}(-1) = \frac{1}{2} B(1/2,1/2) = \frac{\pi}{2}$, and from duplication identity for digamma, evaluated as $s=1$: $$ \psi(s) = \log(2) + \frac{1}{2} \left( \psi\left( \frac{s}{2} \right) + \psi\left( \frac{s+1}{2} \right) \right) \stackrel{s=1}{\implies } \psi(1) - \psi(1/2) = 2 \log(2) $$

$\endgroup$
0
$\begingroup$

Here I give another way to solve the problem. We rewrite the integral as \begin{eqnarray} I&=&\int_0^\infty\log(x+\frac{1}{x})\frac{1}{1+x^2}dx\\ &=&\int_0^\infty\frac{\log(1+x^2)}{1+x^2}dx-\int_0^\infty\frac{\log x}{1+x^2}dx \end{eqnarray} Note $$ \int_0^\infty\frac{\log(1+x^2)}{1+x^2}dx=\pi\log 2$$ from Evaluating $\int_0^{\infty}\frac{\ln(x^2+1)}{x^2+1}dx$ and $$ \int_0^\infty\frac{\log x}{1+x^2}dx=0. $$ Thus we have $$ I=\pi \log 2.$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.