WA gives $\pi^2/9$

  • $\begingroup$ Use the fact that $\dfrac1{x^2-x}=\dfrac1{x(x-1)}=\dfrac1{x-1}-\dfrac1x$ to break it up into two convergent integrals. Then notice that $x^2-x+1=\dfrac{x^3+1}{x+1}$ , and use the fact that $\ln\dfrac ab=\ln a-\ln b$. Then employ the well-known Taylor series for the natural logarithm. $\endgroup$ – Lucian Feb 26 '14 at 9:54

Mhenni has struck first with the approach I have taken, but I would like to elaborate. Again, the integrand may be Taylor expanded:

$$\begin{align}-\int_0^1 dx \frac{\log{[1-(x-x^2)]}}{x-x^2} &= \sum_{n=0}^{\infty} \frac1{n+1} \int_0^1 dx \, x^n (1-x)^n\\ &= \sum_{n=0}^{\infty} \frac1{n+1} \frac{n!^2}{(2 n+1)!}\\ &=2 \sum_{n=0}^{\infty} \frac1{(2 n+2) (2 n+1) \binom{2 n}{n}} \end{align}$$

It turns out that

$$\frac{\arcsin{x}}{\sqrt{1-x^2}} = \sum_{n=0}^{\infty} \frac{2^{2 n} x^{2 n+1}}{(2 n+1) \binom{2 n}{n}} $$

So the sum in question is simply

$$4 \int_0^1 dx \frac{\arcsin{(x/2)}}{\sqrt{1-x^2/4}} = 8 \int_0^{\pi/6} d\theta \, \theta = \frac{\pi^2}{9}$$

as was to be shown.

  • 2
    $\begingroup$ While you have every right to downvote this answer, as well as Mhenni's, it would be nice to know how either of these answers constitutes something not useful. $\endgroup$ – Ron Gordon Feb 26 '14 at 19:19

$\newcommand{\+}{^{\dagger}} \newcommand{\angles}[1]{\left\langle\, #1 \,\right\rangle} \newcommand{\braces}[1]{\left\lbrace\, #1 \,\right\rbrace} \newcommand{\bracks}[1]{\left\lbrack\, #1 \,\right\rbrack} \newcommand{\ceil}[1]{\,\left\lceil\, #1 \,\right\rceil\,} \newcommand{\dd}{{\rm d}} \newcommand{\down}{\downarrow} \newcommand{\ds}[1]{\displaystyle{#1}} \newcommand{\expo}[1]{\,{\rm e}^{#1}\,} \newcommand{\fermi}{\,{\rm f}} \newcommand{\floor}[1]{\,\left\lfloor #1 \right\rfloor\,} \newcommand{\half}{{1 \over 2}} \newcommand{\ic}{{\rm i}} \newcommand{\iff}{\Longleftrightarrow} \newcommand{\imp}{\Longrightarrow} \newcommand{\isdiv}{\,\left.\right\vert\,} \newcommand{\ket}[1]{\left\vert #1\right\rangle} \newcommand{\ol}[1]{\overline{#1}} \newcommand{\pars}[1]{\left(\, #1 \,\right)} \newcommand{\partiald}[3][]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} \newcommand{\pp}{{\cal P}} \newcommand{\root}[2][]{\,\sqrt[#1]{\vphantom{\large A}\,#2\,}\,} \newcommand{\sech}{\,{\rm sech}} \newcommand{\sgn}{\,{\rm sgn}} \newcommand{\totald}[3][]{\frac{{\rm d}^{#1} #2}{{\rm d} #3^{#1}}} \newcommand{\ul}[1]{\underline{#1}} \newcommand{\verts}[1]{\left\vert\, #1 \,\right\vert} \newcommand{\wt}[1]{\widetilde{#1}}$ $\ds{\int_{0}^{1}{\ln\pars{x^{2} - x + 1} \over x^{2} - x}\,\dd x:\ {\large ?}}$

Roots of $\ds{x^{2} - x + 1 = 0}$ are given by $\ds{\quad x_{\rm r} \equiv {1 + \root{3}\ic \over 2}\quad}$ and $\ds{\quad{1 \over x_{\rm r}} = x_{\rm r}^{*}}$.

\begin{align}&\color{#c00000}{% \int_{0}^{1}{\ln\pars{x^{2} - x + 1} \over x^{2} - x}\,\dd x} =-\int_{0}^{1}\ln\pars{x^{2} - x + 1}\pars{{1 \over x} + {1 \over 1 - x}}\,\dd x \\[3mm]&=-2\int_{0}^{1}{\ln\pars{x^{2} - x + 1} \over x}\,\dd x =-2\bracks{\int_{0}^{1}{\ln\pars{1 - x/x_{\rm r}} \over x}\,\dd x+ \int_{0}^{1}{\ln\pars{1 - x/x_{\rm r}^{*}} \over x}\,\dd x} \\[3mm]&=-2\bracks{\int_{0}^{1/x_{\rm r}}{\ln\pars{1 - x} \over x}\,\dd x+ \int_{0}^{1/x_{\rm r}*}{\ln\pars{1 - x} \over x}\,\dd x} =2\bracks{{\rm Li}_{2}\pars{1 \over x_{\rm r}} + {\rm Li}_{2}\pars{x_{\rm r}}} \end{align} where $\ds{{\rm Li}_{2}\pars{z}}$ is the Dilogarithm Function.

Note that $\ds{{\rm Li}_{1}\pars{z} = -\ln\pars{1 - z}}$ and $\ds{{\rm Li}_{2}\pars{z} = \int_{0}^{z}{{\rm Li}_{1}\pars{t} \over t}\,\dd t}$.

By using the Dilogarithm Inversion Formula $\ds{{\rm Li}_{2}\pars{z} + {\rm Li}_{2}\pars{1 \over z} =-\,{\pi^{2} \over 6} - \half\,\ln^{2}\pars{-z}}$ where $\ds{z \not\in \left[\vphantom{\Large A}0, 1\right)}$: \begin{align}&\color{#c00000}{% \int_{0}^{1}{\ln\pars{x^{2} - x + 1} \over x^{2} - x}\,\dd x} =2\bracks{-\,{\pi^{2} \over 6} - \half\,\ln^{2}\pars{-1 - \root{3}\ic \over 2}} \\[3mm]&=2\bracks{-\,{\pi^{2} \over 6} - \half\,\pars{-\,{2\pi \over 3}}^{2}} \end{align}

$$ \color{#66f}{\large% \int_{0}^{1}{\ln\pars{x^{2} - x + 1} \over x^{2} - x}\,\dd x = {\pi^{2} \over 9}} $$


Here is an approach which is based on the Taylor series and the beta function

$$ I = \sum_{k=1}^{\infty}\frac{(-1)^{k+1}}{k} \int_{0}^{1}\frac{(x^2-x)^k}{(x^2-x)}dx = \sum_{k=0}^{\infty}\frac{1}{k} \int_{0}^{1}{x^{k-1}(1-x)^{k-1}}dx $$

$$ = \sum_{k=0}^{\infty}\frac{1}{k} \beta(k,k) = \sum_{k=0}^{\infty}\frac{1}{k} \frac{\Gamma(k)\Gamma(k)}{\Gamma(2k)} = 4(\sin^{-1}(1/2))^2 \sim 1.096622711. $$

  • 1
    $\begingroup$ How do you go from $\sum_{k=0}^{\infty}\frac{1}{k} \frac{\Gamma(k)\Gamma(k)}{\Gamma(2k)}$ to $4(\sin^{-1}(1/2))^2 $? $\endgroup$ – user85798 Feb 26 '14 at 16:20

Let me present you the simplest approach I could even think about. At start, I have to say I am using a method suggested earlier in comments by @Lucian . So it begins by computation of this kind of integrals, by expanding the integral in power series :

$$I_n = \int_0^1 \frac{\ln(1-x^n)}{x} \;\mathrm{d}x= -\int_0^1 \sum_{k=1}^{\infty} \frac{x^{nk-1}}{k} \;\mathrm{d}x = -\sum_{k=1}^{\infty} \frac{1}{nk^2}=-\frac{\zeta{(2)}}{n} $$

Next, using the fact that,


And $x^2-x+1$ is symmetric inder transformation $x \to 1-x$ we get for our original integral :

$$I=-2\int_0^1\frac{\ln\left(x^2-x+1\right)}{x}\;\mathrm{d}x=-2\int_0^1\frac{\ln\left(\frac{1-x^6}{1-x^3}\cdot\frac{1-x}{1-x^2}\right)}{x}\;\mathrm{d}x = -2\left(I_6-I_3-I_2+I_1\right)=2\zeta(2)\left(\frac16-\frac13-\frac12+1\right)=\frac23\zeta(2)=\frac{\pi^2}{9}$$


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.