Revisiting the Entropy-Scaling Concept for Shear-Viscosity Estimation from Cubic and SAFT Equations of State: Application to Pure Fluids in Gas, Liquid and Supercritical States

• Published in: Industrial & engineering chemistry research 2021-09, Vol.60 (34), p.12719-12739
• Main Authors: Dehlouz, Aghilas, Privat, Romain, Galliero, Guillaume, Bonnissel, Marc, Jaubert, Jean-Noël
• Format: Article
• Language: English
• Subjects: Chemical and Process Engineering; Engineering Sciences; Thermodynamics, Transport, and Fluid Mechanics
• ISSN: 0888-5885; 1520-5045
• DOI: 10.1021/acs.iecr.1c01386

The entropy scaling concept postulates that reduced transport properties of fluids are related to the residual entropy, a property that reveals intermolecular interactions and can be estimated from equations of state (EoS) in a straightforward way. In that framework, two models for dynamic viscosity are presented in this paper: in both models, similar expressions inspired from Rosenfeld’s seminal idea are used to reduce transport properties and are related to a carefully selected function of the Tv-residual entropy. This latter is estimated from the PC-SAFT EoS for one model or the tc-PR cubic EoS for the other. The two models are able to predict the viscosities in the entire fluid region (liquid, gas, and supercritical states), which is a great advantage, in comparison to most of the correlations available in the open literature that are specific to a physical state. Model parameters were fitted over a large database containing more than 100 000 pure-fluid experimental data associated with 142 chemical species. For each model, different sets of parameters are provided, each of them being likely to be used in specific situations: first, component-specific parameters were estimated for 142 pure compounds; second, chemical-family specific parameters were proposed for describing components not included in our database but belonging to one of the chemical families we considered. Eventually, for compounds present neither in the original database, nor in the considered chemical families, universal parameters (leading to lower accuracy but applicable to any species) are proposed. The accuracy of the models is obviously maximal when using component-specific parameters and minimal with universal parameters. Thus, the entropy-scaling formulation presented in this work can be used for routinely modeling the dynamic viscosity of any pure fluid. As main advantages, it can be applied to any pure species without restriction and is valid for all fluid states, from the dilute gas to the liquid and even the supercritical state.