Macroinstability and Perturbation in Turbulent Stirred Tank Flows

Abstract

Impeller stirred tank reactors (STRs) are commonly used in the chemical processing industries for a variety of mixing and blending technologies. In this research, a numerical study of flow and mixing inside turbulently agitated STRs are carried out. An immersed boundary method (IBM) is utilized to represent moving impeller geometries in the background of multi-block structured curvilinear fluid. The IBM This curvilinear-IBM methodology is further combined with the large eddy simulation (LES) technique to address the issue of modeling unsteady turbulent flows in the STR. Verification of the combined IBM-LES implementation strategy in curvilinear coordinates is done through comparisons with the measurements of laminar and turbulent flows in baffled STRs with pitched blade impellers. Flow structures are studied inside a dished bottom pitched-blade baffled for different impeller rotational speeds in the turbulent regime to observe the formation of trailing edge vortices which are associated with higher levels of turbulent kinetic energy relative to the remaining flowfield. Instabilities occurring at a frequency lower than the frequency of impeller rotation are identified from the time signal of velocity components. The role of these low frequency macro-instabilities (MI) is explored by observing changes in the three-dimensional circulation pattern within the STR. Significant amount of kinetic energy is observed to be associated with the dynamics of the trailing edge vortices during MI cycles. Flow inside an unbaffled Rushton impeller STR is perturbed using time-dependent impeller rotational speeds at a dominant MI frequency. Perturbation increased the mean radial width of the impeller jet-stream and enhanced overall turbulent kinetic energy compared to the constant rotational speed cases. Large-scale periodic velocity fluctuations due to perturbations produced large strain rates favoring higher turbulence production. Fluctuations in power consumptions are shown to correlate with the perturbation amplitude. Study on the mixing of a passive scalar inside STR showed that the growth rate of unmixed tracer is influenced by the MI oscillations. Perturbation of the STR flow resulted into significant reduction of mixing time. The spatio-temporal behavior of the large-scale mixing structures revealed that fast mixing is promoted due to the break-up of unmixed segregated zones during a perturbation cycle.