Viscosity of liquids—Enskog-2σ model

• Published in: Fluid phase equilibria Vol. 372; pp. 34 - 42
• Main Authors: Umla, Rudolf, Vesovic, Velisa
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 25-06-2014
• Subjects: Correlation; Enskog theory; High pressure; Liquid; Viscosity; Viscosity; Enskog theory; Liquid; Correlation; High pressure
• ISSN: 0378-3812; 1879-0224
• DOI: 10.1016/j.fluid.2014.03.016

•Enskog-2σ model has been extended to analyze the viscosity of liquids.•The model is based on Enskog hard sphere formulation of kinetic theory.•Two effective diameters are used to represent the geometric and dynamic aspects of molecular interactions.•Universal behaviour of the effective diameter is used to predict the viscosity of a given liquid from the viscosity of reference liquid.•The accuracy of the predicted viscosity is similar to the uncertainty of the current correlations.
The recently developed Enskog-2σ model has been extended to correlate and predict the viscosity of simple liquids. The model is based on the Enskog formulation and makes use of two effective diameters to represent two aspects of molecular interactions in Enskog's treatment; namely, the effective diameter representing the excluded volume of a molecule and the effective diameter accounting for the increased probability of collision in comparison to the dilute gas. We have tested the model by analysing the viscosity of six simple fluids (Ar, N2, O2, CO2, CH4, and C2H6) in the liquid phase. The model is capable of correlating the liquid viscosity of the studied fluids with the AAD in the range of 0.1–0.7% and with maximum deviation not exceeding 4.6%. We made use of the observed universal behaviour of the effective diameters to show that the viscosities of the fluids studied are conformal to that of argon. Based on this finding the Enskog-2σ model can be used in the predictive mode. Under these circumstances it reproduces the viscosity data obtained from the recommended correlations with AAD in the range 0.8–3.2%, depending on the way in which the length scaling parameters are estimated.