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

Help me! Stress tensor

  1. Nov 30, 2008 #1
    Well lately i have in mess for this. The problem is about the stress energy tensor. Well we know that
    T_mn = r0 U^m U^n
    where r0 is mass density and U is proper velocity. Ok now consider the local observer. For him except for U^0 other U^m will be jero. So for local observer.
    T_00 = r0 c^2
    other component of this tensor will be zero. Surely for other observer almost all the component maybe nonzero. And T_00 component for other observer will be
    T_00 = r0 c^2 (dt/dTou)^2
    here tou is proper time. And it will be greater than T_00 of local observer. So how can the T tensor remain invariant. Sure its trace can't be the same. For other observer T_00 it self is bigger let alone the other nonzero component. Please help me here. Its bugging me a lot. If u dnt understand something about my writing than tell me.
     
  2. jcsd
  3. Nov 30, 2008 #2

    Mentz114

    User Avatar
    Gold Member

    If you boost a 4-vector, the vector changes, but the scalar length is unchanged.
    So it is with Tmn, the components change but scalars made from it do not.
     
  4. Nov 30, 2008 #3
    Yes i know it. But thats the problem. Local observer has only one nonzero component while other observer's same tensor have lot of nonzero component. some of those component may be bigger than the local observer's only componet. So will the scalar or length will be same? Other observer's T tensor's magnitude sure seems much bigger. So how will the mixed tensor's diagonal terms sum will be same. Or is it that the metric tensor will be such that when converting those contravariant tensor to mix tensor the other observer's component will be such drastically reduced so that their sum will be the same as the local observer's only component.
     
  5. Nov 30, 2008 #4

    Mentz114

    User Avatar
    Gold Member

    I was overlooking the fact that energy is not Lorentz invariant. If you boost T then the kinetic energy will increase and the gravitational effect should be different.
     
  6. Nov 30, 2008 #5
    Then r u saying that T_mn will not remain invariant under cordinate transformation. If so then how come it became tensor?
     
  7. Dec 1, 2008 #6

    Mentz114

    User Avatar
    Gold Member

    Consider the EM force tensor Fmn. Under Lorentz boost the components of the tensor change, but F2=-(E2-B2) remains the same. Only scalars formed from tensors are invariant. I don't know what the invariants of T are ( or if it has any ).
     
  8. Dec 1, 2008 #7

    Ich

    User Avatar
    Science Advisor

    Tensors are covariant, not invariant. Their components change under coordinate transformations.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?