Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

My question is: How can I integrate $\sqrt{1+\frac{1}{3x}} \, dx$ ? (I don't mind about negative Xs).

I'm aware that there's a theorem which you can use that integrates the function using the points where the function can't be solved or something like that(improper integrals?) but I haven't learned anything but parts and substitution yet. I know that this integral can be solved using just those and that's what I am searching for =).

Unfortunately, this is no homework (and my calculus teacher has no idea how to solve this... I asked =( ). I just saw this integral by accident. I've been struggling with it for like a week now. I just want the answer and preferably the steps. Possible starting points:

The best approach I could find is:

$$u^2=3x \implies 2u\ du = 3\ dx$$

$$\frac{2}{3}\int \sqrt{ 1+\frac{1}{u^2} } \, \, u\ du$$

$$\frac{2}{3}\int \sqrt{ u^2+1 } \, \, du$$

Now I know that I can use sinh but I have no clue how.

That was one approach that a guy told me. If you don't like it you could start with the more traditional way:


$$ \frac{1}{3} \int\sqrt{1+\frac{1}{u}}\,du.$$ Now $$w^2=1+\dfrac{1}{u}$$ $$2w\,dw=-\frac{du}{u^2}=-(w^2-1)^2 \,du.$$ $$\frac{1}{3}\int-\frac{2w^2\,dw}{(w^2-1)^2}.$$

After that I've been told that you could integrate by parts but it's just too hard for me. I simply get nowhere. I won't type everything cause it's a waste of time. If anyone wants me to type in more work I'd be happy to if it helps in any way.

I'm in a stage where I just want to see how this super-complex (at least for me) integral is solved. I've used wolframalpha but it's nowhere near a human approach. So well... thanks a lot for any help guys! And sorry for the long post! =)!

share|cite|improve this question
To call this primitive weird is uninformative and... well, weird. – Did Oct 29 '12 at 17:52
@did well it's weird for me.... if you want I can change the title.... – Damieh Oct 29 '12 at 18:10
up vote 4 down vote accepted

Method 1: Let's try integration by parts \begin{eqnarray*} \int\sqrt{u^2+1}\,du&=&\int1\cdot\sqrt{u^2+1}\,du=u\sqrt{u^2+1}-\int u\frac{2u}{2\sqrt{u^2+1}}\,du=\\ &=&u\sqrt{u^2+1}-\int \frac{u^2+1-1}{\sqrt{u^2+1}}\,du= u\sqrt{u^2+1}-\int\left(\sqrt{u^2+1}-\frac{1}{\sqrt{u^2+1}}\right)\,du=\\ &=&u\sqrt{u^2+1}-\underbrace{\int\sqrt{u^2+1}\,du}_{\text{same}}+\int\frac{1}{\sqrt{u^2+1}}\,du. \end{eqnarray*} The unknown integral can be solved now as $$ 2\int\sqrt{u^2+1}\,du=u\sqrt{u^2+1}+\int\frac{1}{\sqrt{u^2+1}}\,du $$ which gives you the answer if you know how to calculate the last integral (quite standard). If you don't then you can try

Method 2: The standard substitution for "square root of the square plus constant" $$ t=u+\sqrt{u^2+1} $$ that gives after some algebra $$ u=\frac{t^2-1}{2t}=\frac{t}{2}-\frac{1}{2t},\quad du=\frac{t^2+1}{2t^2}\,dt,\quad \sqrt{u^2+1}=t-u=\frac{t^2+1}{2t} $$ $$ \int\sqrt{u^2+1}\,du=\int\frac{(t^2+1)^2}{4t^3}\,dt=\int\frac{t^4+2t^2+1}{4t^3}\,dt=\frac{1}{4}\int\left(t+\frac{2}{t}+\frac{1}{t^3}\right)\,dt $$ which I believe you can manage yourself.

P.S. When doing the final substitution back to $u$ it is beneficial to note that $$ \frac{1}{t}=\frac{1}{\sqrt{u^2+1}+u}=\frac{1}{\sqrt{u^2+1}+u}\cdot\frac{\sqrt{u^2+1}-u}{\sqrt{u^2+1}-u}=\frac{\sqrt{u^2+1}-u}{u^2+1-u^2}=\sqrt{u^2+1}-u. $$

P.P.S. Can you now calculate the last integral in the method 1 by the method 2?

share|cite|improve this answer

\begin{equation} \int\sqrt{1+\frac{1}{3x}}\ dx \end{equation} Let \begin{align*} x&=\frac{\cot^2\theta}{3},\\ \frac{\rm{d}x}{\rm{d}\theta}&=\frac{2}{3}(\cot\theta)(-\csc^2\theta) \end{align*} Use the above substitution in (1) \begin{equation}\begin{split} -\frac{2}{3}\int\sqrt{1+\tan^2\theta}\cot\theta\csc^2\theta \ d\theta &=-\frac{2}{3}\int\sec\theta\cot\theta\csc^2\theta \ d\theta\\ &=-\frac{2}{3}\int\frac{\cot\theta}{\cos\theta\sin^2\theta}\ \rm{d}\theta\\ &=-\frac{2}{3}\int\csc^3\theta \ d\theta \end{split}\end{equation} From here, we need integration by parts \begin{align*} u&=\csc\theta,\quad &dv=\csc^2\theta \ d\theta,\\ \rm{d}u&=-\cot\theta\csc\theta \quad &v=-\cot\theta \end{align*} We now have \begin{equation*} \begin{split} -\frac{2}{3}\int\csc^3\theta \ \rm{d}\theta&=-\frac{2}{3}\left[-csc\theta\cot\theta-\int\cot^2\theta\csc\theta \ d\theta\right]\\ &=-\frac{2}{3}\left[-csc\theta\cot\theta-\int(\csc^2\theta-1)\csc\theta \ d\theta\right]\\ &=-\frac{2}{3}\left[-csc\theta\cot\theta-\int\csc^3\theta\ d\theta +\int\csc\theta)\ \rm{d}\theta\right]\\ &=-\frac{2}{3}\left[-csc\theta\cot\theta-\int\csc^3\theta\ d\theta +\int\csc\theta\left(\frac{\csc\theta-\cot\theta}{\csc\theta-\cot\theta}\right)\ d\theta\right]\\ &=-\frac{2}{3}\left[-csc\theta\cot\theta-\int\csc^3\theta\ d\theta +\int\frac{\csc^2\theta-\csc\theta\cot\theta}{\csc\theta-\cot\theta}\ d\theta\right] \end{split} \end{equation*} Now, in the rightmost integral we make the substitution $$u=\csc\theta-\cot\theta, \quad\rm{d}u=(\csc^2\theta-\csc\theta\cot\theta)\ d\theta$$ and we end up with \begin{equation*} -\frac{2}{3}\int\csc^3\theta\ d\theta=\frac{2}{3}\csc\theta\cot\theta+\frac{2}{3}\int\csc^3\theta\ \rm{d}\theta-\frac{2}{3}\ln(\csc\theta-\cot\theta) \end{equation*} Finally collecting similar terms we get \begin{equation*} \begin{split} -\frac{4}{3}\int\csc^3\theta\ d\theta&=\frac{2}{3}\csc\theta\cot\theta-\frac{2}{3}\ln(\csc\theta-\cot\theta),\\ \int\csc^3\theta\ d\theta&=-\frac{1}{2}\csc\theta\cot\theta+\frac{1}{2}\ln(\csc\theta-\cot\theta), \end{split} \end{equation*} So the final result is $$\int\sqrt{1+\frac{1}{3x}}\ \rm{d}x=\frac{1}{2}\left[\ln(\csc\theta-\cot\theta)-\csc\theta\cot\theta\right]$$

share|cite|improve this answer

This is a nice candidate for the “integral of the inverse” technique.
The idea is to use the integration by parts formula in a weird way: $$\int {y dx} = xy - \int{x dy}$$ Let $ y= \sqrt{1+ \frac{1}{3x} } $.
So $x=\frac{1}{3} \frac{1}{y^2-1} $

Plugging this into the integration by parts formula: $$ \int{\sqrt{1+\frac{1}{3x}}dx} = x\sqrt{1+\frac{1}{3x}} - \frac{1}{3}\int {\frac{dy}{y^2-1}} $$ which is much easier to integrate. Afterwards, substitute $ y= \sqrt{1+ \frac{1}{3x} } $ for any remaining y’s.

share|cite|improve this answer

For your first approach, $\int \sqrt{u^2+1} du$, you could try trig substitution, i.e. you may let $u = \tan(\theta)$, and use the identity $\tan^2(\theta)+1 = \sec^2(\theta)$, the integral after the substitution is $\int \sec^3 \theta d\theta$, to do this one you can find it on wikipedia:

share|cite|improve this answer
Okay after 20 minutes I realized that I am in a dead end again. After solving the integrals I end with different trigonometric functions (I used wikipedia as you said.. yay!). What should I do to replace them in form of u? Should I find the trigonometric identity between the functions I end with and the tangent? Thanks a lot! =) – Damieh Oct 29 '12 at 19:02
I believe you find something like $\int \sec^3 xdx=\frac{1}{2}\sec x \tan x+ \frac{1}{2} \log|\sec x + \tan x|$ as in the Wikipedia article, now you know that $u = \tan \theta$, so the matter is to figure out how to represent $\sec(\theta)$ in terms of $u$, but again by the identity we have $1+u^2=1+\tan^2 (\theta) = \sec^2 \theta $ – TTY Oct 29 '12 at 21:09
This was the way to go. By the time I read your comment I had already "solved it". I watched a khanacademy video on trigonometric substitutions and I understood how to do it. I honestly can't believe that I (well.. actually YOU) solved it. THANKS A LOT!!! – Damieh Oct 29 '12 at 21:50

I think the following substitution is easier to work with:

$$u^2=1+\frac{1}{3x}\Longrightarrow 2u\,du=-\frac{dx}{3x^2}\Longrightarrow dx=-6u\frac{1}{9(1-u^2)^2}\,du\Longrightarrow$$

$$\Longrightarrow \int\sqrt{1+\frac{1}{3x}}dx=-\frac{2}{3}\int\frac{u^2}{(1-u^2)^2}du$$

which is already a rational integral (partial fractions and etc.).

Added $\;\;\;$ Partial fractions:



In the last polynomial identity assign values to u (recommended: $\,u=0\,,\,\pm1\,$) and compare powers of the variable (say, of $\,u^3\,$) in order to get the RHS coefficients. If I didn't make a mistake ( and I wouldn't waige on this!), one gets


share|cite|improve this answer
Wait... is that even possible? I'm at work so I'm doing calculations on my head... Wow.. just... wow. So how can I go on from here? Should I integrate by parts? (one part $u^2$ another part $(1-u^2)^2$). Or should I try to find some substitution after cancelling the powers? Thanks a lot I'll try as soon as I go home!! =) – Damieh Oct 29 '12 at 17:59
Read my added stuff in my answer – DonAntonio Oct 29 '12 at 18:07
WTF? I'm guessing I'll need to learn partial fractions then. I'll read about partial fractions and watch a video or two. After that can I come back to you? =) – Damieh Oct 29 '12 at 18:17
Ya me parecia conocido tu nick! Sos el mexicano que me ayudo cuando estaba aprendiendo limites! Jajaja muchas gracias =)!! – Damieh Oct 29 '12 at 18:18
¿Tú sabes (o en lunfardo: ¿vos sabés?) a cuántos miles de personas he encontrado a lo largo de los años en estos sitios? Igual me da gusto el poder haberte ayudado – DonAntonio Oct 29 '12 at 18:36

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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