I have given my friend an integral that I genuinely don't know how to calculate: $$\int_{0}^{\frac{\pi}{2}} \sinh^{-1}(\sin(x))dx$$ He tried differentiating under the integral and complex integration, neither of which worked.

I would love to see a way of calculating this. I know for a fact that it equals Catalan's Constant ~$.915$.

Thanks in advance!

Edit: If anyone is curious where I found this, I found it under the article for Catalan's Constant in Wolfram Alpha: http://mathworld.wolfram.com/CatalansConstant.html

  • $\begingroup$ Is this a fun question? $\endgroup$ – Friedrich Philipp Nov 14 '17 at 0:11
  • $\begingroup$ Quick: what is $\arcsin(\sin(x))$? No calculus, please. $\endgroup$ – Lubin Nov 14 '17 at 0:11
  • 1
    $\begingroup$ @TomHimler You gave your friend a problem that you can't solve!? That's a dirty trick! >:D $\endgroup$ – Franklin Pezzuti Dyer Nov 14 '17 at 0:19
  • 3
    $\begingroup$ @TobyMak But... that's an indefinite integral! There are all kinds of tricks for definite integrals that can't be applied to indefinite ones. $\endgroup$ – Franklin Pezzuti Dyer Nov 14 '17 at 0:22
  • 1
    $\begingroup$ There's an explanation of sorts given at the source that is cited in that article. cs.cmu.edu/~adamchik/articles/catalan/catalan.htm $\endgroup$ – G Tony Jacobs Nov 14 '17 at 0:34

For any $a\in(0,1)$ we have $$ \int_{0}^{\pi/2}\frac{\sin(x)\,dx}{\sqrt{1+a^2 \sin^2 x}}=\frac{\arctan a}{a} $$ hence by applying $\int_{0}^{1}(\ldots)\,da $ to both sides we get $$\int_{0}^{\pi/2}\operatorname{arcsinh}(\sin x)\,dx=\int_{0}^{1}\sum_{n\geq 0}\frac{(-1)^n a^{2n}}{(2n+1)}\,da = \sum_{n\geq 0}\frac{(-1)^n}{(2n+1)^2}=G. $$ Differentiation under the integral sign does work.

An alternative way is to notice that the Taylor series of the (hyperbolic) arcsine function and the integrals $\int_{0}^{\pi/2}(\sin x)^{2n+1}\,dx$ interact by simplifying each other: $$\int_{0}^{\pi/2}\operatorname{arcsinh}(\sin x)\,dx=\sum_{n\geq 0}\frac{\binom{2n}{n}(-1)^n}{(2n+1)4^n}\int_{0}^{\pi/2}(\sin x)^{2n+1}\,dx=\sum_{n\geq 0}\frac{(-1)^n}{(2n+1)^2}. $$ This is reminiscent of one of Euler's proofs of the Basel problem.


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.