Simplification of the kinetic theory (Chapman and Enskog theory)
approach neglecting the second-order effects.
This equation has been extensively tested (Amdur and Mason, 1958;
Bromley and Wilke, 1951; Cheung, 1958; Dahler, 1959; Gandhi and Saxena,
1964; Ranz and Brodowsky, 1962; Saxena and Gambhir, 1963a; Strunk, et
al., 1964; Vanderslice, et al. 1962; Wright and Gray, 1962). In most
cases, only nonpolar mixtures were compared, and very good results
obtained. For some systems containing hydrogen as one component, less
satisfactory agreement was noted. Wilke's method predicted mixture
viscosities that were larger than experimental for the H2-N2 system,
but for H2-NH3, it underestimated the viscosities.
Gururaja, et al. (1967) found that this method also overpredicted in
the H2-O2 case but was quite accurate for the H2-CO2 system.
Wilke's approximation has proved reliable even for polar-polar gas
mixtures of aliphatic alcohols (Reid and Belenyessy, 1960). The
principal reservation appears to lie in those cases where Mi>>Mj
and etai>>etaj.
function gasMixtureViscosity extends Modelica.Icons.Function; input MoleFraction[:] yi "Mole fractions"; input MolarMass[size(yi, 1)] M "Mole masses"; input DynamicViscosity[size(yi, 1)] eta "Pure component viscosities"; output DynamicViscosity etam "Viscosity of the mixture"; end gasMixtureViscosity;