Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Strange geodesic in Schwartzschild metric

  1. Feb 22, 2010 #1
    The following curve is geodesic in Schwardschild metric:
    [tex] \tau \mapsto [(1-2m/r_0)^{-1/2}\tau,r_0,0,0][/tex].
    The tangent vector is: [tex] [(1-2m/r_0)^{-1/2},0,0,0] [/tex], its length is 1 and its
    product with killing vector [tex]\partial_t [/tex] is equal: [tex] (1-2m/r_0)^{1/2} = \textrm{const}[/tex]. So the body lays at rest in gravitational field - why it's possible??
    In newtonian limit it's impossible - the body which does not rotate around a star cannot
    have constant radious.
  2. jcsd
  3. Feb 22, 2010 #2


    User Avatar
    Science Advisor

    Constant energy is necessary for a geodesic, not sufficient. Plug this tangent vector in the equation of motion, you'll get [itex]\ddot r \neq 0[/itex].
    Btw., it's Schwarzschild.
  4. Feb 22, 2010 #3

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    This not a geodesic. If

    [tex]\mathbf{u} = \left( u^t , u^r, u^\theta, u^\phi, \right) = \left( \left( 1 - \frac{2m}{r_0} \right)^{-\frac{1}{2}}, 0, 0, 0 \right),

    then the 4-acceleration is given by

    \mathbf{a} &= \nabla_{\mathbf{u}} \mathbf{u} \\
    &= u^\alpha \nabla_{\partial_\alpha} \left( u^\beta \partial_\beta \right) \\
    &= u^\alpha \left( \nabla_{\partial_\alpha} \left( u^\beta \right) \partial_\beta + u^\beta \nabla_{\partial_\alpha} \left( \partial_\beta \right) \right) \\
    &= \left( u^t \right)^2 \Gamma^\mu {}_{tt} \partial_\mu

    which is non-zero.
    Last edited: Feb 22, 2010
  5. Feb 23, 2010 #4

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I was waiting for comments before finishing this off.


    [tex]0 = \Gamma^t {}_{tt} = \Gamma^\theta {}_{tt} = \Gamma^\phi {}_{tt}[/tex]


    [tex]\Gamma^r {}_{tt} = \left( 1 - \frac{2m}{r_0} \right) \frac{m}{r_0^2}[/tex]


    [tex]\mathbf{a} = \left( 0, \frac{m}{r_0^2}, 0, 0 \right)[/tex]

    with magnitude

    [tex]a = \left( 1 - \frac{2m}{r_0} \right)^{-\frac{1}{2}} \frac{m}{r_0^2}[/tex]

    Taking [itex]r_0[/itex] to be much larger that the Schwarzschild radius, and restoring [itex]c[/itex] and [itex]G[/itex] gives

    [tex]a = \frac{Gm}{r_0^2}[/tex].

    Consequently, such a hovering observer experiences normal Newtonian weight.
  6. Feb 23, 2010 #5


    User Avatar
    Gold Member

    George, that's very instructive, thank you. I'm still reading Lee's book and I've bookmarked this thread.
  7. Feb 23, 2010 #6
    A very confusing thing here is the use of [tex]m[/tex] to denote both half of the Schwarzschild redius and the mass of gravitating body! I think in textbooks whose authers prefer using the notation [tex]2m[/tex] instead of [tex]r_s[/tex] to symbolize the Schwarzschild redius, they later use


    where [tex]M[/tex] is the mass of mass of gravitating body. But I respect George's style and accept it as another alternative.:wink:

  8. Feb 23, 2010 #7


    User Avatar
    Science Advisor
    Gold Member

    There is a convention that many authors of advanced texts use, as well as choosing units of distance and time such that c=1, they also choose units of mass such that G=1. It can cause confusion to persons unfamiliar with it.
  9. Feb 24, 2010 #8
    But you didn't notice that George put a G in the last equation which means the convention that I probably seem to have forgotten leads to


Share this great discussion with others via Reddit, Google+, Twitter, or Facebook