1
$\begingroup$

I have a question about infinite cylidner. I wanted to calculate a gravitational potential that it creates, but I've stumbled across some difficulties.

From Gauss's Law we know, that force on an object with mass m at distance x due to infinte cylinder with density d and radius R equals:

$$F = \frac{2G\pi R^{2}md}{x}$$

So pluging this into equation for work yields:

$$W = \int_{R}^{\infty} F(x)\cdot \text dx= -2G\pi R^{2}md\int_{R}^{\infty}\frac{1}{x}\,\text dx= -2G\pi R^{2}md \Big(\ln(\infty) - \ln(R)\Big) $$

However, $ln(\infty)=\infty$ and that leaves me a little confused. Any help would be appreciated.

$\endgroup$
6
  • $\begingroup$ I'm not sure I understand why it is confusing that moving an object an infinite distance with a force that doesn't decay fast enough will give infinite work. The same is true for constant forces as well. What about this confuses you specifically? $\endgroup$ Commented Apr 16, 2021 at 12:40
  • $\begingroup$ It simply means infinite work is done. $\endgroup$
    – Notwen
    Commented Apr 16, 2021 at 22:08
  • $\begingroup$ @BioPhysicist if the force was constant, then okay, infinite work is understandable. But given that at infinity 1/x = 0 I cannot fully understand, why this work is also infinite. $\endgroup$
    – CodeForFun
    Commented Apr 17, 2021 at 14:43
  • $\begingroup$ $1/x$ just doesn't go to $0$ fast enough. Integrals do not only depend on the value of the integrand at an end point. $\endgroup$ Commented Apr 17, 2021 at 14:45
  • $\begingroup$ @BioPhysicist So what steps woud you recommend to find an expression for potential? $\endgroup$
    – CodeForFun
    Commented Apr 17, 2021 at 14:53

1 Answer 1

1
$\begingroup$

For a potential, you may choose an arbitrary reference point $\vec r_0$ $$ V(\vec r) - V(\vec r_0) = -\int_{\vec r_0}^{\vec r} \vec F(\vec r') \cdot d\vec \ell'. $$

For a cylindrical mass source, the force $1/x$ leads to potential diverges at both $r=0$ and $r=\infty$. Thereforem these two places are not suitable for reference.

I suggest choose the cylidircal shell at $r=1$ as the reference surface:

$$ V(r) -V(r=1) \equiv -\int_1^r F_r \, dr $$

This will avoid to deal with divergent reference.

\begin{align} V(r) - V(1)=& -\int_1^r \frac{2G\pi R^2md}{r'} \, dr'\\ =& -2G\pi R^2md \ln r'\Big\vert_1^r\\ =& -2G\pi R^2md \{ \ln r - \ln 1 \}\\ =& -2G\pi R^2md \ln r \end{align}

Since the potential can arbitrary choose the zero potential position, we then choose $V(r=1) = 0$. This will render a simple form for potential $$ V(r) = -2G\pi R^2md \ln r. $$

$\endgroup$
6
  • $\begingroup$ If I did something wrong? $\endgroup$
    – ytlu
    Commented Apr 16, 2021 at 18:03
  • $\begingroup$ Could you elaborate a little this idea? I understand, that by cylindrical shell, you mean Gaussian one, yes? $\endgroup$
    – CodeForFun
    Commented Apr 17, 2021 at 14:53
  • $\begingroup$ No. It is just the reference position, where you define $V=0$. Nothing to do with the Gaussian surface. We use to pick $r=0$ or $r=\infty$ as the reference position, but for this problem, both position are divergent, not good to served as reference. $\endgroup$
    – ytlu
    Commented Apr 17, 2021 at 19:17
  • 1
    $\begingroup$ (1) The reference of potential is always an arbitrary chosen for convinence purpose. For example in solving $V = mgh$, we will choose an appropriate $h$ as the reference $V=0$ ,just for convinence to perform computation. (2) Practically, you will fail if you insist using $r=0$ or $r=\infty$. $\endgroup$
    – ytlu
    Commented Apr 17, 2021 at 20:28
  • 1
    $\begingroup$ Okay, I think I understand it better now. Thank you very much. $\endgroup$
    – CodeForFun
    Commented Apr 20, 2021 at 7:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.