2
$\begingroup$

I know that for particles carrying no electric charge, given a Killing vector $K_{\mu}$, we have a conserved quantity $K_{\mu}p^{\mu}$ along geodesics, where $p^{\mu}$ is a tangeant vector. However, if $e \neq 0$, the equation of motion isn't simply $p^{\nu}\nabla_{\nu}p^{\mu} = 0$ anymore and so the above doesn't hold. I thought that the quantity $K_{\mu}(p^{\mu} - e A^\mu$), with $A^\mu$ the vector potential, would be conserved instead, but I'm having trouble showing this. Is this quantity really conserved and, if not, what is the correct quantity?

$\endgroup$
1
  • $\begingroup$ A general procedure would be to start with lagrangian of the system :(free particle+electromagnetic field+interaction $J^{\mu}A_{\mu}$), then evaluate $\delta S=0$ by considering variation of coordinates $\delta x^{\mu}=\alpha K^{\mu}$ and then solve for E-L equations of motion $\endgroup$
    – KP99
    Commented Jul 12, 2021 at 16:48

1 Answer 1

1
$\begingroup$

Hints:

  1. Write down the action $S[x]$ for a relativistic point charge in a background metric $g=g_{\mu\nu}\mathrm{d}x^{\mu}\odot\mathrm{d}x^{\nu}$ and a background gauge field $A=A_{\mu}\mathrm{d}x^{\mu}$.

  2. Assume that the vertical infinitesimal transformation $\delta x^{\mu}=\epsilon K^{\mu}$ is a symmetry, i.e. that the Lie derivatives $$ {\cal L}_Kg~=~0\quad\text{and}\quad{\cal L}_KA~=~0 $$ vanish.

  3. Deduce that the conserved Noether charge is $$ Q ~=~K^{\mu}p_{\mu}, $$ where $$ p_{\mu}~:=~\frac{\partial L}{\partial \dot{x}^{\mu}}~=~ p^{\rm kin}_{\mu} \pm q A_{\mu} $$ is the canonical/conjugate 4-momentum. Here we are using the signature convention $(\mp,\pm,\pm,\pm)$, respectively.

$\endgroup$
3
  • $\begingroup$ Right, so you need the Killing vector to also generate a symmetry of the vector potential for this to hold. Thanks a lot! $\endgroup$
    – johnny
    Commented Jul 14, 2021 at 18:29
  • $\begingroup$ Right. That's the main point. Btw $\odot$ stands for the symmetric tensor product. $\endgroup$
    – Qmechanic
    Commented Jul 14, 2021 at 18:35
  • $\begingroup$ Interesting, never came across that notation before! $\endgroup$
    – johnny
    Commented Jul 14, 2021 at 18:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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