Integral of the reciprocal of a quadratic over real line

wotanub
Messages
230
Reaction score
8

Homework Statement


This is from Cahill's Physical Mathematics. Exercise 5.23.

For a \gt 0 and b^{2} – 4ac \lt 0, use a ghost contour to do the integral

\int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x

Homework Equations



Use contour integration and the residue theorem.

The Attempt at a Solution


Mathematica is giving a different result than I got.
It gives \int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x = \frac{2\pi}{\sqrt{4ac-b^{2}}}

My solution:

The roots of the polynomial are

x = \frac{-b \pm \sqrt{b^{2}-4ac}}{2a}

For convenience, write

x = \alpha \pm \mathrm{i}\beta

where \alpha \equiv \frac{-b}{2a}, \beta \equiv \frac{-\mathrm{i}}{2a}\sqrt{b^{2}-4ac} \gt 0

My contour \mathcal{C} will be the real axis, and a large CCW semicircle that encloses the entire UHP. If z = R \mathrm{e}^{\mathrm{i}\theta} then the integral along the arc trivially vanishes as R→\infty

\int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x

= \oint_{\mathcal{C}} \frac{1}{az^{2}+bz+c} \mathrm{d}z

= \oint_{\mathcal{C}} \frac{1}{(z-(\alpha - \mathrm{i}\beta))(z-(\alpha + \mathrm{i}\beta))} \mathrm{d}z

= \mathrm{Res}(\frac{1}{(z-(\alpha - \mathrm{i}\beta))(z-(\alpha + \mathrm{i}\beta))}, z = \alpha + \mathrm{i}\beta), since this is the only pole in the UHP

= 2\pi\mathrm{i}\frac{1}{\alpha + \mathrm{i}\beta - (\alpha - \mathrm{i}\beta)}

= 2 \pi\mathrm{i}\frac{1}{2\mathrm{i}\beta}

= \frac{\pi}{\beta}

= \frac{2a\mathrm{i}\pi}{\sqrt{b^{2}-4ac}}

= \frac{2\pi a}{\sqrt{4ac-b^{2}}}

So I've got an extra factor of a here? Mistakes?
My Mathematica code is
Code:
Integrate[1/(a*x^2 + b*x + c), {x, -Infinity, Infinity}, Assumptions -> {a > 0, 4*a*c > b^2}]
 
Physics news on Phys.org
wotanub said:

Homework Statement


This is from Cahill's Physical Mathematics. Exercise 5.23.

For a \gt 0 and b^{2} – 4ac \lt 0, use a ghost contour to do the integral

\int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x

Homework Equations



Use contour integration and the residue theorem.

The Attempt at a Solution


Mathematica is giving a different result than I got.
It gives \int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x = \frac{2\pi}{\sqrt{4ac-b^{2}}}

My solution:

The roots of the polynomial are

x = \frac{-b \pm \sqrt{b^{2}-4ac}}{2a}

For convenience, write

x = \alpha \pm \mathrm{i}\beta

where \alpha \equiv \frac{-b}{2a}, \beta \equiv \frac{-\mathrm{i}}{2a}\sqrt{b^{2}-4ac} \gt 0

My contour \mathcal{C} will be the real axis, and a large CCW semicircle that encloses the entire UHP. If z = R \mathrm{e}^{\mathrm{i}\theta} then the integral along the arc trivially vanishes as R→\infty

\int^{\infty}_{-\infty} \frac{1}{ax^{2}+bx+c} \mathrm{d}x

= \oint_{\mathcal{C}} \frac{1}{az^{2}+bz+c} \mathrm{d}z

= \oint_{\mathcal{C}} \frac{1}{(z-(\alpha - \mathrm{i}\beta))(z-(\alpha + \mathrm{i}\beta))} \mathrm{d}z

= \mathrm{Res}(\frac{1}{(z-(\alpha - \mathrm{i}\beta))(z-(\alpha + \mathrm{i}\beta))}, z = \alpha + \mathrm{i}\beta), since this is the only pole in the UHP

= 2\pi\mathrm{i}\frac{1}{\alpha + \mathrm{i}\beta - (\alpha - \mathrm{i}\beta)}

= 2 \pi\mathrm{i}\frac{1}{2\mathrm{i}\beta}

= \frac{\pi}{\beta}

= \frac{2a\mathrm{i}\pi}{\sqrt{b^{2}-4ac}}

= \frac{2\pi a}{\sqrt{4ac-b^{2}}}

So I've got an extra factor of a here? Mistakes?
My Mathematica code is
Code:
Integrate[1/(a*x^2 + b*x + c), {x, -Infinity, Infinity}, Assumptions -> {a > 0, 4*a*c > b^2}]

If ##r_1, r_2## are the roots of the denominator, you need to write ##ax^2 + bx + c = a(x-r_1)(x-r_2)##, not just ##(x-r_1)(x-r)2)## as you wrote.
 
Oh cool. I know how to contour integrate, I just need to practice multiplying.
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top