Question: What is the source of truth for calculating density in GUNNS?

[mc1998]

Does GUNNS use NIST data for these calculations?

Is there a way to overwrite this, or rather choosing a density within GUNNS for a gas?

[joeval920]

Yes, the GUNNS density curves for most fluids are modeled after the NIST Chemistry WebBook. More info here: https://github.com/nasa/gunns/wiki/Fluid_Properties#modeled-properties

You could overwrite the values used in the density table (used to derive a curve fit).
For example, N2:

gunns/ms-utils/properties/FluidProperties.cpp

Line 264 in 584bf2d

const double DefinedFluidProperties::mN2DensityTable[] =

I don't believe we have a way to do this at runtime yet.

[mc1998]

@joeval920 thanks for the response!

That would be my next question then. Is there a way to turn off compressibility for the gases?

For example, I want to use Helium gas but this results in significant errors because the densities don't match up (~35 kg/m^3 from GUNNS vs 32 kg/m^3 from NIST). However, if I use HE_REAL fluid instead, then I do get the correct density (32 kg/m^3), but not the correct mass initialization that is associated with ideal gas laws.

I did this by copying the mDensityHeReal to mDensityHe within FluidProperties.cpp and converting the mDensityHe instance from QuadraticFit -> TsBilinearInterpolator to match HE_REAL.

The idea being that I still want to follow ideal gas laws, but calculate density from NIST interpolation. This however, still brings errors, because the density that I see now is 30 kg/m^3

Or other suggestions I can try out? I also want to try implementing an overwriting density variable, but Im worried this will overwrite the density every time.

[joeval920]

Sorry, circling back to this. I think your idea to redefine mDensityHe as a TsBilinearInterpolator was a good one. We can initialize the interpolator with values focused on your usage range of T&P, to provide a more accurate interpolation. Do you mind sharing what Ts & Ps you are interested in?

[mc1998]

Sorry for the delay, the ranges I'm interested in are as follows:

2900 psi (19994.8 kPa) -> 5850.0 psi (40334.33 kPa)
50 F (283.15 K) -> 110 F ( 316.483 K)

Thanks @joeval920 ! Let me know if there is anything else needed.