Limitations on the usage of Fuego law - Low temperatures

[SRkumar97]

Hello,
So far, I have been using the Fuego gas law to run simulations in PeleC. The kind of flows which I am studying are simple, non-reacting viscous flows - essentially compressible aerodynamics. Now, in one of the cases the freestream temperature is pretty low (145 K) which translates to (-128.15 C)
The issue I'm facing is that the values for specific and total internal energy i.e. (rho_e and rho_E respectively) are completely negative, for cold inflow/freestream temperatures. As an example, I'm attaching a simple boundary layer flow, in which flow transits from a laminar flow at the inlet, gradually to a turbulent boundary layer. The freestream temperature is 145 K and pressure is 4 kPa, I'm using the Fuego law, with Chemistry as air and Transport model as Sutherland.
In the attached contours, though temperature as well as density are positive, the internal energies are completely negative, which is practically impossible.
The simulation snapshots are 2D planar slices, at a time of 2e-4 sec., at the 750th time step.

I was wondering whether this is due to a limitation in the applicability of this Fuego law to low temperatures such as this case. Since the mentioned NASA paper in the PelePhysics documentation discusses about data in the range 200 to 20,000 K. And this same issue occurs of negative rho_e and rho_E using Fuego law for air, is occurring in inviscid simulations too at such low temperatures. Requesting for help on this.

Thanks in advance

[baperry2]

With Fuego EOS, the energy variables include both chemical and thermal energy. Chemical energy can have large negative values so negative values for the energy variables are not an issue.

[SRkumar97]

I'd like to clarify once - The NASA paper mentions the temperature range as 200 K to 20,000 K; but here you seem to mention that PeleC does work below 200 K for Fuego. But is there any limiting, lowest temperature below which PeleC would fail to give correct results?

With Fuego EOS, the energy variables include both chemical and thermal energy. Chemical energy can have large negative values so negative values for the energy variables are not an issue.

Also, for the thermal diffusion component, is there any real input control parameter which has to be used to implement the correction flux for SUM(F_m)=0? The PeleC documentation site mentions about this under diffusion for ideal gas, mass conservation enforcement part.

[baperry2]

The range of validity is dependent on the Thermo data for each species in the mechanism you are using. You can see this as the temperature-ranges entries in the mechanism.yaml file for the mechanism you are using. Going below the lower bound is not strictly correct but often works out ok. Going above typically leads to invalid data that causes the simulation to crash.

All aspects of the diffusion correction flux are always on and cannot be toggled from the input file.

[SRkumar97]

@baperry2 thanks! I checked the file, the temperature ranges in the file. Most of the gases appear to have in the range [300, 1000, 5000], especially for both N2 and O2. I believe the first and last values of this square bracket are the lower and upper limits of the range; is the middle value 1000 simply for an increment frequency like we have in for loops?

[drummerdoc]

There are usually two polynomial fits and the middle value is the temperature where the fit switches between the two.

[SRkumar97]

Ohh, okay. Thanks for clarifying.