CFD simulation of a Rushton turbine stirred-tank using open-source software with critical evaluation of MRF-based rotation modeling

TL;DR

This study critically evaluates the influence of the MRF technique on steady RANS simulations of Rushton turbine stirred-tanks using OpenFOAM, highlighting how MRF zone size affects turbulence predictions and mixing time estimations.

Contribution

It provides a systematic analysis of MRF zone size effects on turbulence and mixing predictions in stirred-tank simulations, emphasizing the need for careful MRF region selection.

Findings

Limited differences in velocity profiles across MRF sizes.

Significant variations in turbulence intensity predictions.

Mixing time predictions vary up to a factor of three.

Abstract

A critical evaluation of the impact of the Multiple Reference Frame (MRF) technique on steady RANS simulations of a Rushton turbine stirred-tanks is presented. The analysis, based on the open source software OpenFOAM, is focused on the choice of the diameter and thickness of the MRF region and on their effect on the predicted velocity field and mixing times in the tank. Five diameters of the MRF region are compared for the same operating conditions of the turbine, showing limited differences in velocity profiles, which are found in general good agreement with available experimental data. Significant differences are nonetheless found in the predicted levels of turbulence intensity within the tank, with a considerable amount of artificially generated turbulence at the boundary of the MRF region for the largest diameters. The impact of the different predictions of the turbulent field on the modeling of the mixing process in the tank is evaluated by simulating the release of a passive scalar, using the frozen-flow field hypothesis. The results show changes in mixing times up to a factor of three when comparing MRF regions of different size. Thus, the present investigation highlights the importance of assessing the effect of the MRF zone size on numerical results as a standard practice in RANS based simulations of stirred-tanks.