Comparative Analysis of the MRF and Sliding-Mesh Strategies to Simulate the Fluid Flow in a Rushton Turbine Mixer

Abstract

One of the main concerns about stirred-tank reactors is the mixing time. Experimental measurements of the mixer performance require expensive equipments, therefore computational fluid dynamics simulations are an attractive alternative. Among the simulation strategies, the multiple reference frame (MRF) method has gained popularity due to its simple mesh generation and reduced computational cost when it is compared to sliding-mesh. However, it is found that solutions computed with MRF show a dependence on the size where fluid rotation is considered. In this work it is shown that even though MRF is accurate to compute mixer global indices like power number, the mixing time is overestimated and the flow field is not correct. A reference mixer with a Rushton turbine is used to this end, with MRF, sliding-mesh and experimental results used in the comparison. Both OpenFOAM and StarCCM are employed for the simulations.