June 8th, 2019, 08:27 AM  #1 
Newbie Joined: Jun 2019 From: Ukraine Posts: 6 Thanks: 0  Integral
how to solve the integral with residues? If I can solve it like that 
November 1st, 2019, 03:41 PM  #2 
Newbie Joined: Jun 2016 From: UK Posts: 8 Thanks: 8 Math Focus: Anything but stats :) 
In order to evaluate $I \displaystyle = \int_{\infty}^{\infty} \dfrac{dx}{x^2+4}$ using residues, we first need to consider the complex contour integration given by $I_C = \displaystyle \int_C \dfrac{dz}{z^2 + 4}$ where $C$ is a CLOSED contour representing a semicircle lying in the upperhalf of the complex plane, centred at the origin and of radius $R>0$. This contour integral can be decomposed into two simpler ones: (a) A contour integral along the real line from $R$ to $R$. This is represented as $\displaystyle \int_{R}^R \dfrac{dz}{z^2 + 4}$. Note that on the real line, we have $z=x$ therefore this is literally just the same as $\displaystyle \int_{R}^R \dfrac{dx}{x^2 + 4}$. and (b) A contour integral along the semicircle in the upper half plane. This is represented by $\displaystyle \int_{\Gamma} \dfrac{dz}{z^2 + 4}$ where $\Gamma : z=Re^{i\theta}$ where $\theta \in (0,\pi)$. Therefore, we have that $\displaystyle I_C = \int_{R}^R \dfrac{dx}{x^2 + 4} + \int_\Gamma \dfrac{dz}{z^2 + 4}$. Now consider both sides in the limit $R \to \infty$. We get that $\boxed{ \displaystyle \lim_{R \to \infty} I_C = \int_{\infty}^{\infty} \dfrac{dx}{x^2 + 4} + \lim_{R \to \infty}\int_\Gamma \dfrac{dz}{z^2 + 4} }$ and as you can see, the integral we want is literally just sitting there. Luckily, we can evaluate the other two terms with some introductory theory. Firstly, note that in the second integral, we have $z = Re^{i \theta}$ where $\theta \in (0,\pi)$. Also note that $z^2 + 4 \geq z^2 + 4 = R^2 + 4$ therefore $\dfrac{1}{z^2 + 4} \leq \dfrac{1}{R^2 + 4}$. This is important because we can use something known as Estimation Lemma (or ML inequality) to get an upper bound on our integral. Therefore, notice that $\displaystyle \left \int_\Gamma \dfrac{dz}{z^2 + 4} \right \leq \dfrac{1}{R^2 + 4}L$ where $L$ is the length of our contour. In our case, since the contour is a semicircle, $L = \pi R$. Therefore $\displaystyle \left \int_\Gamma \dfrac{dz}{z^2 + 4} \right \leq \dfrac{\pi R}{R^2 + 4}$ and as we take the limit $R \to \infty$, the upper bound goes to zero therefore this entire integral vanishes, i.e. $\displaystyle \lim_{R \to \infty} \int_{\Gamma} \dfrac{dz}{z^2+4} = 0$. Secondly, to work out $\displaystyle \lim_{R \to \infty} I_C$ we note that $I_C$ is a closed contour and so we use what's known as Residue Theorem whereby the value of this term is precisely the $2\pi i$ multiplied by the sum of all residues of $\dfrac{1}{z^2 + 4}$. There is only one residue for this function in our contour, and this is precisely $z_0 = 2i$. Therefore, we have that $\displaystyle \lim_{R \to \infty} I_C = 2\pi i \mathrm{Res}_{z = 2i} \dfrac{1}{z^2 + 4} = 2\pi i \lim_{z \to 2i} (z2i)\dfrac{1}{z^2 + 4} = 2\pi i \lim_{z \to 2i}\dfrac{1}{z+2i} = 2\pi i \dfrac{1}{4i} = \dfrac{\pi}{2}$ Therefore, our boxed expression reduces (in the limit) down to $\dfrac{\pi}{2} = \displaystyle \int_{\infty}^{\infty} \dfrac{dx}{x^2 + 4}$ And so coming back to your problem, your integral is precisely half this, therefore $\dfrac{\pi}{8}$. 
November 4th, 2019, 11:10 AM  #3 
Global Moderator Joined: Dec 2006 Posts: 21,114 Thanks: 2329 
$\displaystyle \frac12\left(\frac{\pi}{2}\right) = \frac{\pi}{4} \text{, not }\frac{\pi}{8}$.


Tags 
integral 
Thread Tools  
Display Modes  

Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Double integral, repeated integral and the FTC  Jhenrique  Calculus  5  June 30th, 2015 04:45 PM 
a new discovery in integral calculus??for integral pros only  gen_shao  Calculus  2  July 31st, 2013 10:54 PM 
integral of double integral in a region E  maximus101  Calculus  0  March 4th, 2011 02:31 AM 
Prove Lower Integral <= 0 <= Upper Integral  xsw001  Real Analysis  1  October 29th, 2010 08:27 PM 
integral of double integral in a region E  maximus101  Algebra  0  December 31st, 1969 04:00 PM 