A numerical study on the mixing time prediction of miscible liquids with high viscosity ratios in turbulently stirred vessels
Sorosh Mirfasihi, Wrichik Basu, Philip A. Martin, Adam Kowalski, Cláudio P. Fonte, Amir Keshmiri
Abstract
Mixing processes are crucial in industrial applications, including food, pharmaceutical, and chemical manufacturing, to ensure product homogeneity and quality control. Effective control of high-viscosity fluid mixing is essential due to intricate mixing dynamics involved. This study evaluates blending time predictions from two Computational Fluid Dynamics (CFD) methodologies for simulating the mixing of two miscible liquids with high contrasting viscosities. The investigation employed a scalar transport model coupled with a Reynolds-Averaged Navier-Stokes (RANS) Finite Volume Method (FVM) solver and a Lattice Boltzmann Large Eddy Simulation (LB-LES) solver to assess flow parameters against experimental data. Blending times were validated against Electrical Resistance Tomography (ERT) based measurements in a 2.6-litre baffled vessel agitated by a Rushton turbine under turbulent conditions. Results indicated both models align closely with experimental trends of dimensionless blending time relative to fluid properties; however, accuracy reduced as viscosity ratios exceeded a critical Reynolds number threshold.