15
$\begingroup$

What real analysis tools would you recommend me for getting the closed form of the integral below?

$$\int_0^{\infty} \frac{\log(x)}{\cosh(x) \sec(x)- \tan(x)} \ dx$$

$\endgroup$
  • 9
    $\begingroup$ A hammer would suffice. $\endgroup$ – Ali Caglayan Oct 26 '14 at 18:23
  • $\begingroup$ are you sure that a closed form exist? $\endgroup$ – Dr. Sonnhard Graubner Oct 26 '14 at 18:25
  • $\begingroup$ @Dr.SonnhardGraubner $100$%. My brother confirmed that. $\endgroup$ – user 1357113 Oct 26 '14 at 18:25
16
$\begingroup$

The place I would start is the nifty result, proven here, that

$$\frac{\sin{x}}{\cosh{t} - \cos{x}} = 2 \sum_{k=1}^{\infty} e^{-k t} \sin{k x} $$

Of course, the integral actually looks like

$$\int_0^{\infty} dx \frac{\cos{x}}{\cosh{x} - \sin{x}} \log{x} $$

so we need to map $x \mapsto \pi/2 - x$ and we have that the integral is actually

$$2 \sum_{m=0}^{\infty} (-1)^m \int_0^{\infty} dx \, e^{-(2 m+1) x} \cos{(2 m+1) x}\, \log{x} + 2 \sum_{m=1}^{\infty} (-1)^{m+1} \int_0^{\infty} dx \, e^{-2 m x} \sin{2 m x}\, \log{x} $$

We then note that

$$-k \int_0^{\infty} dx \, e^{-(1-i) k x} \log{x} = \frac{\log (k)}{2}+\frac{\gamma }{2}+\frac{\log (2)}{4}+\frac{\pi }{8}+i \left(\frac{\log (k)}{2}+\frac{\gamma }{2}+\frac{\log (2)}{4}-\frac{\pi }{8}\right)$$

So the integral is

$$-\sum_{m=0}^{\infty} (-1)^m \frac{\log{(2 m+1)}}{2 m+1} - \sum_{m=1}^{\infty} (-1)^{m+1} \frac{\log{(2 m)}}{2 m}\\ - \left (\gamma + \frac12 \log{2} + \frac{\pi}{4} \right )\sum_{m=0}^{\infty} \frac{(-1)^m}{2 m+1} - \left (\gamma + \frac12 \log{2} - \frac{\pi}{4} \right )\sum_{m=1}^{\infty} \frac{(-1)^{m+1}}{2 m}$$

The third and fourth sums are well known. The second is less-well known, but may be shown to be

$$\frac{1}{4} \left(3 \log ^2(2)-2 \gamma \log (2)\right)$$

(See, for example, Williams & Hardy, The Red Book of Mathematical Problems, problem 71.)

The first sum, however, is less well known. It may be evaluated by considering the derivative of a generalized zeta function (or a Lerch transcendent). The result is

$$\sum_{m=0}^{\infty} (-1)^m \frac{\log{(2 m+1)}}{2 m+1} = \frac12 \log{2} \left [\psi{\left ( \frac{3}{4} \right )} - \psi{\left ( \frac{1}{4} \right )} \right ] - \frac14 \left [ \gamma_1{\left ( \frac{3}{4} \right )}-\gamma_1{\left ( \frac{1}{4} \right )}\right ]$$

where $\gamma_1(a)$ is a generalized Stieltjes constant.

Put this all together and the result agrees with a numerical evaluation performed with Mathematica ($ \approx -1.35775$).

ADDENUDUM

Apparently I forgot that I had encountered that first sum some time ago here, and the result may be expressed in simpler terms:

$$ \sum_{m=1}^{\infty} (-1)^{m} \frac{\log{(2 m+1)}}{2 m+1} = -\frac{\pi}{4} \gamma - \frac{\pi}{4} \log{\frac{\Gamma{\left ( \frac{3}{4} \right )}^4}{\pi}} $$

so we may now put this all together as follows:

$$\frac{\pi}{4} \gamma + \frac{\pi}{4} \log{\frac{\Gamma{\left ( \frac{3}{4} \right )}^4}{\pi}} - \frac{3}{4} \log^2{2} + \frac12 \gamma \log{2} - \left (\gamma + \frac12 \log{2} + \frac{\pi}{4} \right )\frac{\pi}{4} - \left (\gamma + \frac12 \log{2} - \frac{\pi}{4} \right )\frac12 \log{2}$$

or

$$\int_0^{\infty} dx \frac{\cos{x}}{\cosh{x} - \sin{x}} \log{x} = \frac{\pi}{4} \log{\frac{\Gamma{\left ( \frac{3}{4} \right )}^4}{\pi}} - \log^2{2} - \frac{\pi^2}{16}$$

ADDENDUM II

The methodology outlined above may be applied to other, similar integrals. For example, it is a simpler matter to show that

$$\int_0^{\infty} dx \frac{x \sin{x}}{\cosh{x}-\cos{x}} \log{x} = \frac{1}{6} \pi ^2 \left(-12 \log{A}+1+\frac{\log{2}}{2}+\log{\pi} \right) $$

where $A$ is Glaisher's constant.

$\endgroup$
5
$\begingroup$

$$\int_0^{\infty} \frac{\log(x)}{\cosh(x) \sec(x)- \tan(x)} \ dx=\frac{\pi}{4}\log\left(\displaystyle \frac{4 \pi^3}{\Gamma\left(\displaystyle \frac{1}{4}\right)^4}\right)-\left(\frac{\pi}{4}\right)^2-\log^2(2)$$

$\endgroup$

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.