0
$\begingroup$

I want to find the inverse Laplace transform of $$f(s)={\frac{1}{s^{3/2}}}$$ Refer to the Laplace transform table, and I found that the result is $$F(t)=2\sqrt{\frac{t}{\pi}}$$ But I do not know how to get this result. I tried to use the Bromwich integral $$F(t)=\frac{1}{2\pi i}\int_{c-i\infty}^{c+i\infty}\frac{1}{s^{3/2}}e^{st}\,ds$$ My progress so far has been stunted by the fact that we have a branch point at s=0. The contour should be like this, but I don't know how to perform the integration.

Any help is appreciated.

enter image description here

$\endgroup$
3
  • $\begingroup$ $s=0$ is a branch point of $f(s)$ not an isolated singularity (thus not an essential singularity). For $t > 0$ use the change of variable $u = st$ to express $F(t)$ in term of $F(1)$ $\endgroup$
    – reuns
    May 16 '19 at 14:21
  • $\begingroup$ thanks for the comment. I revised my question. $\endgroup$ May 17 '19 at 9:04
  • $\begingroup$ See this. math.stackexchange.com/questions/1343764/laplace-inverse/… $\endgroup$
    – Ron Gordon
    Jul 5 '19 at 16:10
0
$\begingroup$

$s=0$ is not an essential singularity. It is a branch point. Choose a branch to calculate your integral, for example, choose branch $-\pi<\arg z<\pi$ and integrate along the contour that made up of:

  1. straight line from $c-iR$ to $c+iR$
  2. along large quarter circle to approximately $(c-R)+\varepsilon i$
  3. straight line to $\varepsilon i$
  4. right half of circle $\lvert z\rvert=\varepsilon$, to $-\varepsilon i$
  5. straight line to approximately $(c-R)-\varepsilon i$
  6. another large quarter circle to $c-iR$

Can you finish from here? Be careful when you take square-root.

$\endgroup$
3
  • $\begingroup$ I don't see what you are doing. With the principal branch of $s^{-3/2}$ and $c > 0$ for $t > 0, \frac{1}{2\pi i}\int_{c-i\infty}^{c+i\infty}s^{-3/2}e^{st}\,ds=\frac{1}{2\pi i}\int_{ct-i\infty}^{ct+i\infty}(u/t)^{-3/2}e^{u}\,d(u/t)= t^{1/2} F(1)$. To evaluate $F(1)$ shift the contour to the one enclosing $(-\infty,0]$ $\endgroup$
    – reuns
    May 16 '19 at 16:10
  • $\begingroup$ This is designed to change from the Bromwich integral to one encircling the nonpositive real axis, $\endgroup$ May 16 '19 at 16:28
  • $\begingroup$ Thanks for the reply. I revised my question. So the contour should be like this, but I don't know how to perform the integration. Would you like to help me further? $\endgroup$ May 17 '19 at 9:09
0
$\begingroup$

Haven't studied about the Bromwich integral yet but wouldn't it be sufficient to use the identity:

$L(t^n) = \dfrac{\Gamma(n+1)}{s^{n+1}}$

with $n = \dfrac{1}{2}$ and deduce hence?

$\endgroup$

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.