Diffeo-invariant action for a matter covector field

In summary, the paper discusses the formulation of a diffeo-invariant action principle for a matter covector field, emphasizing the importance of maintaining invariance under diffeomorphisms in theoretical physics. It presents the mathematical framework for constructing the action and analyzes its implications for the dynamics of the covector field, as well as its interactions with other fields in the context of general relativity and field theory. The work highlights the significance of diffeo-invariance in ensuring the consistency and physical relevance of the proposed action.
  • #1
ergospherical
1,072
1,365
I just need a hint to get started, and then I reckon the rest will follow...
We consider a theory where matter is a covector field ##\omega_a## which is described by a diffeomorphism-invariant action ##S_m##. Define:$$E^{a} = \frac{1}{\sqrt{-g}} \frac{\delta S_m}{\delta \omega_a}$$Also, ##T^{ab} = \tfrac{2}{\sqrt{-g}} \tfrac{\delta S_m}{\delta g_{ab}}## is defined as per usual. We would like to derive:$$\nabla_a {T^a}_b = E^a \nabla_b \omega_a - \nabla_a(E^a \omega_b)$$The statement that ##S_m## is diffeomorphism invariant: I guess this means, under ##x\mapsto x - \xi##, and therefore$$\delta g_{\mu \nu} = (L_{\xi} g)_{\mu \nu} = \nabla_{\mu} \xi_{\nu} + \nabla_{\nu} \xi_{\mu}$$that ##S_m## is invariant... how do I use that? I imagine you can write something down in terms of the Lie derivative, and then manipulate the resulting equation into the result. But how to start??
 
Last edited:
Physics news on Phys.org
  • #2
Never mind, I figured it out. For sure, you write out$$\delta S_m = \int d^4 x \left[ \frac{\delta S_m}{\delta \omega_a} \delta \omega_a + \frac{\delta S_m}{\delta g_{ab}} \delta g_{ab} \right]$$and then stick in the expressions ##\delta \omega = L_{\xi} \omega## and ##\delta g = L_{\xi} g## in terms of covariant derivatives, as well as replacing ##\delta S_m / \delta g_{ab}## by the expression involving the energy momentum tensor. Then just integration by parts gives what you want, when you make sure it vanishes under arbitrary ##\xi##.

Please delete or close the thread if you want... (classic rubber-duck example - write something down and then figure it out)
 
Last edited:
  • Like
Likes weirdoguy

Similar threads

Replies
2
Views
868
Replies
4
Views
693
Replies
10
Views
1K
Replies
10
Views
1K
Replies
73
Views
10K
Replies
1
Views
1K
Back
Top