Session 1 - Theme: Experimental Techniques & Results (1)

Mixing systems in wastewater treatment plants consume a major share of overall energy, creating a persistent trade-off between process reliability, sufficient mixing performance, and energy efficiency. To address this challenge, a comprehensive methodological framework for the comparative evaluation of different propeller geometries was developed as part of a government funded project called RIOWAR (funding code 03EN2092C). Computational Fluid Dynamics (CFD) and electrical resistance tomography (ERT) were applied under well-defined and reproducible boundary conditions to systematically assess mixing efficiency. A process-tomography test rig incorporating ERT was used to assess mixing time. Mixing time was quantified at a target homogeneity of 90%, using methods well-documented in the literature. Scaling laws based on the similarity theory were applied to ensure dynamically similar flow conditions across the laboratory model and the full-scale propeller. CFD models were validated through flow measured by time-resolved Particle Image Velocimetry (PIV). The study followed an iterative design–simulate–evaluate workflow in which simulation outcomes directly informed the development of improved propeller geometries, trageting an enhanced or similar mixing performance at reduced energy input. While comparing different propeller laboratory models, it was found, that at higher rotation rates the optimized geometries reduced the mixing time by 10 to 20 % compared to the established geometry. In future research the application of the optimized propeller geometry shall be investigated in real-scale environments.

Dual-shaft, turbine–anchor (DSTA) stirred tanks are widely used for mixing complex rheological fluids due to their versatility across a broad range of rheology. Different impeller types can be installed at varying heights and speeds to serve specific functions, such as pumping, shearing/aggregate break-up or dead-zone elimination. Despite widespread industrial use, limited literature exists on DSTA mixer design and performance. In this study, transitional-regime mixing was investigated in a 19 L flat-bottomed DSTA equipped with a top-driven anchor and an eccentric Rushton turbine (RT) or down-pumping pitched blade turbine (PBTd). Acid–base decolourisation, particle image velocimetry (PIV) and torque measurements were used to assess mixing time (θ₉₅), flow patterns, velocity fields and power consumption. Complementary CFD simulations were performed using STAR-CCM+ to support interpretation of the experimental results. The results show that the effect of eccentric impeller type (RT or PBTd) on mixing time depends on the anchor rotation speed, NANC. At maximum NANC, mixing times are independent of impeller type, whereas at lower NANC—including fixed anchor (NANC = 0 rpm)—the PBTd achieves faster mixing (for equivalent total specific power, ̄ε). Correlations were also developed to quantify changes in anchor power draw due to impeller interactions as functions of the Reynolds number ratio, ReECC/ReANC—yielding R2 values of 0.76 and 0.97 for co- and counter-rotation modes, respectively.

When contacting two immiscible liquids in stirred vessels, there exists a minimum impeller speed, NJD, to completely incorporate the dispersed phase. We develop a model based on the energy or force balance approach developed for solid-liquid mixing studies (Baldi et al., 1978; Davies, 1986; Grenville et al., 2015). Similar to solid liquid, the model strikes a balance between the disruptive forces of the turbulent eddies and the dispersed phase buoyancy, however, unlike with solids, the liquid pool breaks up into smaller and small droplets until a balance is struck and so interfacial forces and viscous forces are also considered. To test this model appropriately, the continuous and dispersed phase physical properties (interfacial tension, viscosity, density) are varied using several model fluids and additives. The study is conducted primarily in 305mm diameter vessels with select experiments conducted in 450 and 610mm diameter vessels to evaluate scale-up. The majority of impeller configurations investigated consisted of dual wide blade hydrofoils (MHS), triple counterflow impellers, and dual paddles (SLS). NJD was determined visually in triplicate by incrementally increasing the impeller speed until no remaining floating undispersed organic remained. Impeller reaction torque was measured using an air bearing. The initial model is found to fit the experimental data better when a model is used to estimate a mixture or emulsion viscosity opposed to using the continuous phase. Furthermore, the experiments suggest that the 305mm diameter vessel behaves differently than the larger scale vessels.