Journal of Supercritical Fluids, Vol.145, 228-242, 2019
The use of polynomial models to determine thermodynamic properties of turbulent supercritical mixture in SAS process: A statistical analysis
The system studied was the precipitation of beta-carotene in carbon dioxide as antisolvent, using dichloromethane as the solvent in the supercritical antisolvent process and employed turbulence models k-epsilon Realizable and Standard. The main difference between both models is that the Realizable model is showing error margins between 15 and 30% relative to the experimental data, while the Standard model ranges between 30 and 60%. Thus, so that the model k-epsilon is always achievable, two changes from the standard model were incorporated. The physical properties were also evaluated using polynomials adjusted for thermodynamic properties. Simulations performed relative to the cases where the adjusted polynomials were used on a 90 bar pressure and a temperature of 308 K showed results where there were no marked differences when the dependent properties of p, T and composition were used, indicating a good strategy to use the methodology of the polynomials adjusted rather than thermodynamic properties.