I'm doing a workout guide about double integrals and I came across an exercise that I could not resolve for a while.

$$\int_0^2\int_1^2 \frac{x}{\sqrt{1+x^2+y^2}} \,\mathrm dx\,\mathrm dy$$

I guess that the easier order of integration is $dxdy$, because if I try to integrate respect to $y$ first, I would have to deal with the integral of a root of $1+x^2+y^2$, while outside of the root there's no $y$.

I tried Integration by substitution (with $u = 1 + x^2 + y^2$ and $du = 2xdx$) to solve the inner integral, but then I felt I couldn't solve the outside integral.

Any hints you can give me?.

  • 1
    $\begingroup$ Why don't you try substituting $x=r\cos\theta$ and $y=r\sin\theta$?? $\endgroup$ – tattwamasi amrutam Mar 28 '15 at 2:21
  • 1
    $\begingroup$ The region is a rectangle, wouldn't the limits be more difficult? $\endgroup$ – OiciTrap Mar 28 '15 at 2:23
  • $\begingroup$ First do the integration with respect to $x$ as you planned. That is not too hard. The outer integral over $y$ then has terms like $\sqrt{a + y^2}$ for various values of $a$. Look up this integral in your integral table and take it from there. The answer is a bit messy, but there are no essential difficulties. $\endgroup$ – Hans Engler Mar 28 '15 at 2:25

The antiderivative may be deduced, but it takes a little patience. I prefer to integrate by parts rather than do a trig sub right away.

$$\begin{align}\int dx \sqrt{x^2+a^2} &= x \sqrt{x^2+a^2} - \int dx \frac{x^2}{\sqrt{x^2+a^2}}\\ &= x \sqrt{x^2+a^2} - \int dx \sqrt{x^2+a^2} + a^2 \int \frac{dx}{\sqrt{x^2+a^2}} \end{align}$$

Now a trig sub is less work:

$$\implies 2 \int dx \sqrt{x^2+a^2} = x \sqrt{x^2+a^2} + a^2 \int dt \, \sec{t} $$


$$\int dx \sqrt{x^2+a^2} = \frac12 x \sqrt{x^2+a^2} + \frac12 a^2 \log{\left (x+\sqrt{a^2+x^2}\right )} + C$$

  • $\begingroup$ Thanks, IDK why I didn't think about trigonometric substitution xD. $\endgroup$ – OiciTrap Mar 28 '15 at 4:29

$\newcommand{\bbx}[1]{\,\bbox[15px,border:1px groove navy]{\displaystyle{#1}}\,} \newcommand{\braces}[1]{\left\lbrace\,{#1}\,\right\rbrace} \newcommand{\bracks}[1]{\left\lbrack\,{#1}\,\right\rbrack} \newcommand{\dd}{\mathrm{d}} \newcommand{\ds}[1]{\displaystyle{#1}} \newcommand{\expo}[1]{\,\mathrm{e}^{#1}\,} \newcommand{\ic}{\mathrm{i}} \newcommand{\mc}[1]{\mathcal{#1}} \newcommand{\mrm}[1]{\mathrm{#1}} \newcommand{\on}[1]{\operatorname{#1}} \newcommand{\pars}[1]{\left(\,{#1}\,\right)} \newcommand{\partiald}[3][]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} \newcommand{\root}[2][]{\,\sqrt[#1]{\,{#2}\,}\,} \newcommand{\totald}[3][]{\frac{\mathrm{d}^{#1} #2}{\mathrm{d} #3^{#1}}} \newcommand{\verts}[1]{\left\vert\,{#1}\,\right\vert}$ $\ds{\bbox[5px,#ffd]{}}$

\begin{align} &\bbox[5px,#ffd]{\int_{0}^{2}\int_{1}^{2} {x \over \root{1 + x^{2} + y^{2}}}\,\dd x\,\dd y} = \int_{0}^{2}\int_{1}^{2} \nabla\times\pars{\root{1 + x^{2} + y^{2}}\,\hat{y}} \cdot\dd\vec{S} \\[5mm] = &\ \oint_{\partial\bracks{\pars{1,2}\times\pars{0,2}}}\,\,\, \root{1 + x^{2} + y^{2}}\,\hat{y}\cdot\dd\vec{r} \\[5mm] = &\ \int_{0}^{2}\root{1 + 2^{2} + y^{2}}\,\dd y + \int_{2}^{0}\root{1 + 1^{2} + y^{2}}\,\dd y \\[5mm] = &\ \underbrace{\int_{0}^{2}\root{5 + y^{2}}\,\dd y} _{\ds{3 + {5 \over 4}\ln\pars{5}}}\ -\ \underbrace{\int_{0}^{2}\root{2 + y^{2}}\,\dd y} _{\ds{\root{6} + \ln\pars{\root{3} + \root{2}}}} \\[5mm] = &\ \bbx{3 - \root{6} + {5 \over 4}\,\ln\pars{5} - \ln\pars{\root{3} + \root{2}}} \approx 1.4161 \\ & \end{align}

How hard can it be?? I mean once you substitute $x=r\cos\theta$ and $y=r\sin\theta$, all you need to do it find the limits of $r$ and $\theta$.Now $1 \le r\cos\theta \le 2$ and $0 \le r\sin \theta \le 2$. Squaring both sides it is easy to see that $1 \le r \le 2\sqrt{2}$ . Similarly you can find out that $ 0 \le \theta \le tan^{-1}(2)=\theta_{1}$. Then Substituting them in the integral yields $$\int_{0}^{\theta_1} \int_{1}^{2\sqrt{2}}\frac{r\cos\theta}{\sqrt{1+r^2}} rdrd\theta$$

This can be integrated easily.

  • $\begingroup$ You have turned an integral over a rectangle into an integral over a piece of an annulus. That cannot be right. $\endgroup$ – Hans Engler Mar 28 '15 at 2:51
  • $\begingroup$ Yeah, if we are integrating a rectangle is impossible that $1 \le r \le 2\sqrt{2}$. $\endgroup$ – OiciTrap Mar 28 '15 at 3:29

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.