0

So, if I'm working in spherical coordinates, how would I evaluate the following integral? I know that I'm supposed to use contour integration and Jordan's lemma, but the fact that the singularity is located on the real axis is really throwing me off. Any advice for solving this improper integral via contour integration only?

$$\int_0^\infty{\frac{r \sin{(r\rho)}}{r^2-\alpha^2}} dr$$ where alpha is a real number

Incognito
  • 435
  • This doesn't converge. – Potato Apr 27 '14 at 22:15
  • @Potato I think the Cauchy principal value exists, though – Ben Grossmann Apr 27 '14 at 22:17
  • @Potato thanks for the response. It should converge, as my resource is asserting that it is equivalent to a specific value (what I'm trying to figure out). Could I do a contour with 2 indentations around alpha and -alpha? – Incognito Apr 27 '14 at 22:19
  • @Incognito It does not converge! Around its singularity, it looks like $1/(r-\alpha)$, which is not integrable if $\alpha$ is real. – Potato Apr 27 '14 at 22:20
  • As @Omnomnomnom insinuated, you need to take the principal value. If you do that, you get a good value. – Daniel Fischer Apr 27 '14 at 22:21
  • @DanielFischer What exactly do you mean by the principal value? There are no branches here, correct? – Incognito Apr 27 '14 at 22:23
  • @Incognito see this link. For example, we would not say, strictly speaking, that the integral $$ \int_{-1}^1 \frac 1x,dx $$ exists. However, its Cauchy principal value is $0$. – Ben Grossmann Apr 27 '14 at 22:25
  • @Omnomnomnom uhuh, yes. So that means that we need 2 indentations around the singularities, right? – Incognito Apr 27 '14 at 22:28
  • @Incognito: Yes, that's right. You may find that it's tricky to show that the integral along the semicircular contour goes to zero, though – Ben Grossmann Apr 27 '14 at 22:29
  • @Omnomnomnom Alright, I understand what's going on now. Though, honestly, I think it will be more challenging to evaluate the integrals around these indentations. Since there are no residues, the final integral will have to be equal to these 2 indentation integrals (they are not easy to solve). – Incognito Apr 27 '14 at 22:33
  • The indentations are just $1/2$ the full integral around, so you just end up with $\pi i$ times the residue at the point. – Ben Grossmann Apr 27 '14 at 22:35
  • @Omnomnomnom Ahhhhh. So, that trick still works for integrals on the real axis even if it is away from the origin. Thank you. This isn't as bad as I initially thought it would be. – Incognito Apr 27 '14 at 22:38

2 Answers2

2

Hint: begin by noting that $$ \int_0^\infty{\frac{r \sin{(r\rho)}}{r^2-\alpha^2}}\,dr = \Im\left\{ \int_0^\infty{\frac{r e^{(ir\rho)}}{r^2-\alpha^2}}\,dr \right\} $$

For the indentation contour around $\alpha$: let $\gamma_\epsilon$ be a semicircle of radius $\epsilon$ that goes counterclockwise around $\alpha$. We wish to evaluate $$ \int_{\gamma_\epsilon} \frac{z e^{(iz\rho)}}{z^2-\alpha^2}\,dz = \int_{\gamma_\epsilon} \frac{1}{z - \alpha} \cdot \frac{z e^{(iz\rho)}}{z+\alpha}\,dz $$ Noting that $g(z) = \frac{z e^{(iz\rho)}}{z+\alpha}$ is continuous around $z = \alpha$, we can take $\epsilon$ sufficiently small so that $g(z) \approx g(\alpha)$ when $|z- \alpha| < \epsilon$. We conclude $$ \lim_{\epsilon \to 0^+} \int_{\gamma_\epsilon} \frac{g(z)}{z-\alpha}\,dz = \int_{\gamma_\epsilon} \frac{g(\alpha)}{z-\alpha}\,dz = \pi i g(\alpha) $$

Ben Grossmann
  • 225,327
0

$\newcommand{\+}{^{\dagger}} \newcommand{\angles}[1]{\left\langle\, #1 \,\right\rangle} \newcommand{\braces}[1]{\left\lbrace\, #1 \,\right\rbrace} \newcommand{\bracks}[1]{\left\lbrack\, #1 \,\right\rbrack} \newcommand{\ceil}[1]{\,\left\lceil\, #1 \,\right\rceil\,} \newcommand{\dd}{{\rm d}} \newcommand{\down}{\downarrow} \newcommand{\ds}[1]{\displaystyle{#1}} \newcommand{\expo}[1]{\,{\rm e}^{#1}\,} \newcommand{\fermi}{\,{\rm f}} \newcommand{\floor}[1]{\,\left\lfloor #1 \right\rfloor\,} \newcommand{\half}{{1 \over 2}} \newcommand{\ic}{{\rm i}} \newcommand{\iff}{\Longleftrightarrow} \newcommand{\imp}{\Longrightarrow} \newcommand{\isdiv}{\,\left.\right\vert\,} \newcommand{\ket}[1]{\left\vert #1\right\rangle} \newcommand{\ol}[1]{\overline{#1}} \newcommand{\pars}[1]{\left(\, #1 \,\right)} \newcommand{\partiald}[3][]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} \newcommand{\pp}{{\cal P}} \newcommand{\root}[2][]{\,\sqrt[#1]{\vphantom{\large A}\,#2\,}\,} \newcommand{\sech}{\,{\rm sech}} \newcommand{\sgn}{\,{\rm sgn}} \newcommand{\totald}[3][]{\frac{{\rm d}^{#1} #2}{{\rm d} #3^{#1}}} \newcommand{\ul}[1]{\underline{#1}} \newcommand{\verts}[1]{\left\vert\, #1 \,\right\vert} \newcommand{\wt}[1]{\widetilde{#1}}$ $\ds{\int_{0}^{\infty}{r\sin\pars{r\rho} \over r^{2} - \alpha^{2}}\,\dd r:\ {\large ?}.\quad}$ These integrals are very common in Quantum Mechanics as related to scattering theory. In that cases, $\ds{\alpha}$ has a 'small' imaginary part which indicates whether a wave is going to a scattering center or is leaving the scattering center.

The 'history' is more or less like this: \begin{align} &\color{#c00000}{\int_{0}^{\infty}{r\sin\pars{r\rho}\over r^{2} - \pars{\verts{\alpha} \pm \ic 0^{+}}^{2}}\,\dd r} \\[3mm]&=\int_{0}^{\infty}r\sin\pars{r\rho}\, {1 \over 2\pars{\verts{\alpha} \pm \ic 0^{+}}}\bracks{% {1 \over r - \pars{\verts{\alpha} \pm \ic 0^{+}}} -{1 \over r + \pars{\verts{\alpha} \pm \ic 0^{+}}}}\,\dd r \\[3mm]&={1 \over 2\pars{\verts{\alpha} \pm \ic 0^{+}}}\,2\ic\, \Im\int_{0}^{\infty}{r\sin\pars{r\rho} \over r - \verts{\alpha} \mp \ic 0^{+}}\,\dd r \\[3mm]&={\ic \over \verts{\alpha} \pm \ic 0^{+}} \int_{0}^{\infty}r\sin\pars{r\rho} \bracks{\pm\,\pi\,\delta\pars{r - \verts{\alpha}}}\,\dd r ={\ic \over \verts{\alpha} \pm \ic 0^{+}} \,\bracks{\pm\pi\verts{\alpha}\sin\pars{\verts{\alpha}\rho}} \\[3mm]&=\pm\ic\pi\verts{\alpha}\sin\pars{\verts{\alpha}\rho}\, \bracks{\pp{1 \over \verts{\alpha}} \mp \ic\pi\delta\pars{\verts{\alpha}}} \end{align} where $\ds{\delta\pars{x}}$ is the Dirac Delta 'Function' and we used the identities ( 'under' the integral sign ) $\ds{{1 \over x \pm \ic 0^{+}} = \pp{1 \over x} \mp \ic\pi\delta\pars{x}}$. $\ds{\pp}$ denotes the Principal Value.

$$ \color{#00f}{\large\int_{0}^{\infty} {r\sin\pars{r\rho} \over r^{2} - \pars{\verts{\alpha} \pm \ic 0^{+}}^{2}}\,\dd r =\pm\,\ic\pi\sin\pars{\verts{\alpha}\rho}} $$

Felix Marin
  • 89,464