$$\int_{1}^{e} \frac{\ln(x)}{\left( 1+\ln(x) \right)^2}dx$$

I consider using the u-subs $u=\ln(x)$ or $u=\ln(x)+1$, but I was left with a factor of $e^u$ I couldn't get rid of.

I was also considering usign the series expansion of $\frac{1}{(1-x)^2}$, but I couldn't find a way to easily integrate powers of $\ln(x)$. Any insight or help would be greatly appreciated.

  • $\begingroup$ Try $u=\ln(x)$ and then using the series expansion. The integral is from 0 to 1 so it is valid. $\endgroup$ – Moya Jul 7 '19 at 19:21
  • $\begingroup$ $u=ln(x)+1$ works nicely. The resulting integrand is a sum ($e^u/u$ and $-e^u/u^2$). Apply integration by parts to the $e^u/u^2$ integral and you will see some nice cancellation in the sum :) $\endgroup$ – thomasfermi Jul 7 '19 at 19:41

The square in the denominator should remind you of the quotient rule. looking at the integrand and rewriting it as the derivative of $\frac{u}{\ln x +1}$with the quotient rule you get $$\frac{\ln x}{(1+\ln x)^2}=\frac{u'(\ln x +1)-\frac ux}{(1+\ln x)^2}$$ from which you can guess that $u=x$. This indicates that $$\int_1^e \frac{\ln x}{(1+\ln x)^2} dx = \left [ \frac{x}{1+ \ln x} \right]^e_1=\frac e2 -1$$

  • $\begingroup$ Thanks, that was pretty big brain $\endgroup$ – Anirudh Nov 14 '19 at 6:44

Substituting $t=\ln x$ gives us that the integral $I$ satisfies $$I=\int_0^1\frac{e^tt}{(t+1)^2}dt.$$ Now we integrate by parts with $u=e^tt$ and $dv=\frac{1}{(t+1)^2}dt$. In particular, we get that $du=e^tt+e^t$ and $v=-\frac{1}{t+1}$. This implies that\begin{align*}I&=\int_0^1\frac{e^t(t+1)}{t+1}dt-\frac{e^tt}{t+1}\Bigg|_0^1\\&=-1+e-\frac e2=\boxed{\frac e2-1.}\end{align*}

  • 1
    $\begingroup$ In your first integral the limits should be $0$ and $1$. $\endgroup$ – Anurag A Jul 7 '19 at 19:31
  • $\begingroup$ @AnuragA Thanks! I just fixed it $\endgroup$ – boink Jul 7 '19 at 19:41
  • 1
    $\begingroup$ I believe that you might have a sign error. Wolfram Alpha and my calculation gives $e/2-1$ as a result. $\endgroup$ – thomasfermi Jul 7 '19 at 19:44
  • 2
    $\begingroup$ @thomasfermi In addition to you and Wolfram Alpha, the positivity of the integrand on the interval of integration says that the answer should be positive. $\endgroup$ – Andreas Blass Jul 7 '19 at 19:47
  • $\begingroup$ Thanks—I swapped signs when integrating by parts apparently! $\endgroup$ – boink Jul 7 '19 at 19:48

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.