Results of velocity measurements of liquid and gas bubbles in a tank with a self-aspirating disk impeller are analysed. Studies were carried out using a fluorescent dye tracer in the measuring system with two cameras (simultaneous phase velocity measurement) and with one camera (sequential measurement of phase velocity). Based on a comparative analysis of the acquired data it was found that when differences in the phase velocities were small the simultaneous velocity measurement gave good results. However, sequential measurement gives greater possibilities for setting the measuring system and if the analysis of instantaneous velocities is not necessary, it seems to be a better solution.

Geometric parameters of a ribbon impeller were optimized on the basis of numerical calculations obtained from the solution of our own 3D/2D hybrid model. The optimization was made taking into account mixing power and homogenization time for ribbon impellers with a different number of ribbons and width operating in a laminar motion for Newtonian fluid. Due to minimum mixing energy required to stir a unit volume of liquid the most efficient impeller appeared to be that with one ribbon of width equal to 0.1 to 0.15 of the mixing vessel diameter. Impellers with more than one ribbon needed much higher mixing power but did not increase significantly secondary circulation in the vessel. These impellers increased first of all primary circulation, i.e. they increased only circular motion of liquid in the vessel.

In the paper, the mixing power and distributions of velocity and velocity pulsations in a baffled stirred tank with a flat blade turbine impeller placed at different distances from the bottom were determined. It was found that the mixing power reaches minimum values when the relative clearance of the impeller is C/D = 0.6÷0.7. The investigations of velocity distributions using the PIV method showed the axial flow of the liquid through the impeller. This results in deviations from the typical radial-circumferential flow and changes in mixing power vs. impeller clearance versus a Rushton impeller. With a clearance corresponding to the minimum power, the flow is axial-circumferential with one circulation loop. For a flat blade turbine impeller, good mixing conditions are obtained for a clearance of 0.8 < C/D < 0.9.

The paper presents a photographic analysis of the break-up of gas bubbles flowing out of the outlets of a self-aspirating disk impeller. It was found that bubbles detached from the interfacial surface most often disintegrate to form several daughter bubbles. Further in the work, the population balance model was verified for several formulas describing the bubble break-up rate. It has been found that a good fit to the experimental data is provided by the formula given by Laakkonen for 5 daughter bubbles. The possibility of using the Monte Carlo method to model the bubble break-up processwas also determined. For this method, a good agreement of results was achieved for the division into a maximum of 10 daughter bubbles. In the case of this method it was also found necessary to use the function of break-up frequency at a higher rate for smaller bubbles.

In the study a new proposal of convective velocity determination necessary for eddy size determination from the dissipative range in a turbulent flow in a mixer was made. The proposed quantity depends on all the mean and fluctuating velocity components. By applying convective velocity one may determine the distribution of time and linear Taylor microscale in a stirred vessel.

Feasibility of a model of gas bubble break-up and coalescence in an air-lift column enabling determination of bubble size distributions in a mixer with a self-aspirating impeller has been attempted in this paper. According to velocity measurements made by the PIV method with a self-aspirating impeller and Smagorinski’s model, the spatial distribution of turbulent energy dissipation rate close to the impeller was determined. This allowed to positively verify the dependence of gas bubble velocity used in the model, in relation to turbulent energy dissipation rate. Furthermore, the range of the eddy sizes capable of breaking up the gas bubbles was determined. The verified model was found to be greatly useful, but because of the simplifying assumptions some discrepancies of experimental and model results were observed.

This study presents the results of tests on the mixing power and distribution of three velocity components in the mixing tank for an FBT impeller during tank emptying with an operating impeller. A laser PIV system was used to determine speed distributions. It was found that for the relative liquid height in the tank H* = H/H0 ≈ 0.65 and H* ≈ 0.45, the liquid circulation in the impeller zone changed from radial to axial and vice versa. These changes were accompanied by changes in the mixing power which even reached 40%. In the theoretical part, a method of calculating the mixing power using the classical model of the central vortex and distribution of the tangential speed in the impeller zone was proposed. Although the method turned out to be inaccurate, it was useful for determining the relative power.