# Should the universe be modeled by perfect fluid or ideal gas？

In physical cosmology, the content of the Universe is modeled by the stress-energy-momentum tensor of perfect fluid, with energy density $\rho(t)$ and pressure $P(t)$. I'm wondering, why not use ideal gas instead？

• I suppose the difference would be that an ideal gas exhibits heat conduction and viscosity, while a perfect fluid does not. Most cosmological models (or at least the simpler ones that were developed first historically) are isotropic and homogeneous, which means that even if your matter was capable of heat conduction and viscosity, they wouldn't occur. E.g., no temperature gradients, so no heat conduction. – Ben Crowell Sep 24 '17 at 3:08