WIP: Update cell density based on cell-avg'ed temperature

[RonRahaman]

In this PR, the CoupledDriver computes cell density from cell-averaged temperature and the heat/fluid solver's pressure BC using the IAPWS correlation.

I'm marking this as "WIP" because, even though it's reasonable for Nek, it might not be appropriate for the heat surrogate. The SurrogateDriver solves for enthalpy H in fluid; then it computes T from pressure P and H using the IAPWS correlation. Seems odd to compute rho using a second correlation with T and P. Seems like it would be better to directly calculate rho based on T and H. But currently, the base class HeatFluids driver doesn't expose H to the CoupledDriver.

Given that we eventually want to do enthalpy coupling for all the head/fluids solvers, we might not want to merge this PR. I'm just creating the PR in case @emerzari and @hapfang want to use this for testing.