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

Motivation behind vacuum Einstein equation

  1. Dec 2, 2012 #1
    Hi,

    I am using Hartle to study GR and at one point, there is a leap that I don't understand. He finds the result for the geodesic deviation equation and introduces the Riemann curvature tensor.

    Then, we are told that there is an object called the Ricci curvature tensor which is a contracted version of the Riemann curvature tensor and that in a vacuum, the components of the Ricci tensor go to zero.

    Can anyone tell me what the motivation behind this is? Hartle's explanation is not very clear at all. After getting the Riemann curvature tensor, how would one logically go to the Ricci tensor and equate it to 0 for vacuum?

    Thank you!
     
  2. jcsd
  3. Dec 2, 2012 #2

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Let's work through this.

    First, what is Einstein's equation?
     
  4. Dec 2, 2012 #3

    zonde

    User Avatar
    Gold Member

    Isn't the statement that Ricci tensor goes to zero equivalent to Einstein's equivalence principle?
    It means that locally spacetime is Euclidean, right?

    If that would be so then we just have to sort out why Ricci tensor would be non zero in presence of mass/energy.
     
  5. Dec 3, 2012 #4

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    In general, it is not a good idea to use the the equivalence principle as a substitute for a calculation. At any event in spacetime, it is always possible to find an orthonormal basis, even in the presence of matter. Also, even if the Ricci tensor is zero, the curvature tensor does not have to be zero.

    The way to proceed is to start with Einstein's equation in its usual form, do a little calculation that transforms Einstein's equation to a slightly less usual form in which is obvious that Ricci is zero for a vacuum.
     
  6. Dec 3, 2012 #5
    We haven't got to the Einstein equation yet. It was introduced first using [itex]R_{\alpha\beta}=0[/itex] in a vacuum. This is section 21.4 of Hartle, in case anyone has the book too.

    Hartle justifies this as a generalization of the Newtonian version [itex]\nabla^{2}\phi=0[/itex] where [itex]\phi[/itex] is the graviational potential. His line of reasoning, if I understand correctly is as follows. Let's call [itex]\chi[/itex] the separation vector which measures the separation between two nearby particles as they fall freely. The Newtonian calculation yields
    [tex]\frac{d^{2}\chi_{i}}{dt^{2}}=-\frac{\partial^{2}\phi}{\partial x_{i}\partial x_{k}}\chi_{k}[/tex]

    We also have the relation [itex]\nabla^{2}\phi=4\pi G\mu[/itex] where [itex]\mu[/itex] is the mass distribution. In a vacuum, this is just [itex]\nabla^{2}\phi=0[/itex].

    Then, there is the separation vector for nearby geodesics which obeys a similar relation as the Newtonian one but this time, the right hand side has the Riemann curvature tensor instead of the partial derivative of [itex]\phi[/itex]. The relationship is
    [tex](\nabla_{\mu}\nabla_{\mu}\chi)^{\alpha}=-R^{\alpha}_{\beta\theta\delta}u^{\beta}u^{\delta} \chi^{\theta}[/tex]

    From here, he somehow makes the conclusion that we must have the Ricci curvature zero to get the analog of the Newtonian case. But, I'm not sure how it came about. How does one even think of the contracting the Riemann tensor in such a way to get the Ricci curvature tensor?

    I understand this is the sort of thing that can't be derived but I am looking for some sort of insight. I feel like I am missing something obvious because Hartle generally does proceed very logically. Thank you for your help!
     
    Last edited: Dec 3, 2012
  7. Dec 3, 2012 #6

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    When, I first responded, I did not have Hartle at hand; now I do.

    Compare equation (21.27) to (21.13).
     
  8. Dec 4, 2012 #7
    Oh I see it now! Must have missed that earlier.

    Thank you, George Jones!
     
  9. Dec 4, 2012 #8

    Bill_K

    User Avatar
    Science Advisor

    Could someone who has the book reveal the answer to those of us who don't? Thanks!
     
  10. Dec 4, 2012 #9
    Sure, let me try to explain. I'm only sort of sure so if someone finds something wrong, please do correct it. Hartle takes the following equation in a local inertial frame.
    [tex](\nabla_{\mu}\nabla_{\mu}\chi)^{\alpha}=-R^{\alpha}_{\beta\theta\delta}u^{\beta}u^{\delta} \chi^{\theta}[/tex]

    This yields,
    [tex]\frac{d^{2}\chi^{\alpha}}{d\tau^{2}}=R^{\alpha}_{ \beta \theta\delta}\chi^{\beta}[/tex]

    The R here is the Riemann curvature tensor in the LIF.

    Now, he uses the static weak field metric to work out this Riemann tensor and contracts it to get the Ricci tensor. This yields (approximately, since it is a weak perturbation of the flat space metric)

    [tex]\frac{d^{2}\chi^{i}}{dt^{2}}=R^{i}_{ t j t}\chi^{j}[/tex]

    [tex]R_{ij}=\frac{\partial^{2}\phi}{\partial x^{i}\partial x^{j}}[/tex]

    Comparing with the Newtonian case, the Ricci tensor must be zero in a vacuum.
     
    Last edited: Dec 4, 2012
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook