AG-2026.06-1206·gr-qc
Action based approach to dissipative relativistic fluid systems
Authors
- G. L. Comer
- N. Andersson
- T. Celora
- I. Hawke
Abstract
We develop an action principle for a relativistic two-fluid system with dissipation. The specific constituents of the model - which serves as a proof of principle - are particles and entropy. The linchpin of the action is the assertion that a given flux is dissipative if its covariant divergence is non-zero. For our model, the particle flux is taken to be conservative while the entropy flux is dissipative. This allows for a "top-down" approach where the general question is geometric. Previous work has shown that new terms (the proper time derivative of matter space "metrics") must be included in the Lagrangian in order to produce equations of motion with terms representing bulk and shear viscosity. In addition to including these terms we show that further terms - interpreted as velocities - can be included. The new action-based model recovers known relativistic formulations of the Cattaneo equation, which results in causal heat propagation. We further advance our understanding by exploring the single-fluid limit by locking the entropy four-velocity to that of the matter component. This reduces the system to a single field equation along with a constraint equation. We show that this constraint leads to a dynamical extension of the standard Tolman red-shift condition. Finally, we provide three example actions (of increasing complexity) which demonstrate that the model is able to reproduce (in the single-fluid limit) the anticipated terms from the relativistic Navier-Stokes equations. In the general case, the action based approach allows for a much richer structure, which may be relevant for realistic models of non-linear dissipative relativistic fluid systems.
Submitted
16 June 2026yesterday
Version
v1
License
CC-BY-4.0
DOI
10.48550/arXiv.2606.17686
Chat with this PDF
Ask questions, probe assumptions, request a plain-English summary. Answers cite sections from the preprint itself.
Community
Questions and answers about this paper from other readers. No formal peer review — just a place to think out loud.