Abstract
The linear blending rule (LBR) is the simplest mixing rule linking physical properties with mixture composition. It is just a composition-weighted arithmetic mean over the pure component property values. Unfortunately, the composition dependence of most mixture properties shows nonlinear deviations from the LBR. The conventional approach is to use higher-order Scheffé polynomials. However, this introduces cross-parameters that characterize nonlinear blending effects and that require extensive mixture data for their evaluation. Instead, we propose new parameter-sparse mixing rules that feature pure component parameters only. The new mixing rules were constructed by (i) employing composition-weighted power means, (ii) transforming the composition variables via Wohl’s q-fraction concept, and (iii) combinations of these two approaches.
