Issue #2719 modified Eucken R in J/kg.K

[thorade]

This closes #2719.

data.R = Modelica.Constant.R/data.MM

@finback-at: This should be identical to the solution you suggested, could you verify it gives the same results?

[finback-at]

Thank you for this change!
Cp*data.MM/Modelica.Constants.R and Cp/data.R give same results.
Both data.MM and data.R are provided from Media.IdealGases.Common.SingleGasesData.
Consistencies of those depend on this data table.
I checked it and the calculation results of thermal conductivity using these 2 form of equations for
the cases of Ar, CH4, CO, CO2, H2, N2, H2O, O2, He and C3H8.
Off course, the results are same.
And those shows better agreements with NIST data than previous.
I think it's alright!

[finback-at]

My first suggested equation (#2719) is using Cp*dataMM/Modelica.Constants.R,
and this correction using Cp/data.R. In principle, these are identical.
Difference between them depend on SingleGasesData (Nasa Glenn data table).
I checked for the cases of Ar, CH4, CO, CO2, H2, N2, H2O, O2, He, C3H8.
The order of differences are about 1e-6 - 1e-5.
This correction gives better agreements of thermal conductivity with the data of
NIST chemistry webbook for all cases.