The paper presents research upon the gas distribution in a physical model and the computer simulation of dust separation in a horizontal electrostatic precipitator (ESP) with a flat inlet diffuser. The research of a gas flow was carried out using the visualization method and the velocity measurement in cross sections of a model chamber. By selecting suitable choking diffusion screens and deflecting vanes in a diffuser the oblique profiles of a gas velocity were obtained for different obliqueness degree. It was assumed that the velocity profiles obtained should guarantee higher performance of an ESP than those uniform profiles as used so far. Those assumptions were proved by the results of computer simulation obtained using a program SYMULA-X. The results of experiments and computer simulation arc presented in a graphical form.

The velocity field around the standard Rushton turbine was investigated by the Laser Doppler Anemometry (LDA) and Particle Image Velocimetry (PIV) measurements. The mean ensembleaveraged velocity profiles and root mean square values of fluctuations were evaluated at two different regions. The first one was in the discharge stream in the radial direction from the impeller where the radial flow is dominant and it is commonly modelled as a swirling turbulent jet. The validity range of the turbulent jet model was studied. The second evaluated region is under the impeller where flow seems to be at first sight rather rigorous but obtained results show nonnegligible values of fluctuation velocity.

For underground mine workings, the shape of the computational domain may be difficult to define. Historically, the geometry models of mine drifts were not accurate representations of the object but rather a simplified approximation. To fully understand a phenomenon and save time on computations, simplification is often required. Nevertheless, in some situations, a detailed depiction of the geometry of the object may be necessary to obtain adequate simulation results. Laser Scanning enables the generation of 3D digital models with precision beyond the needs of applicable CFD models. Images composed of millions of points must be processed to obtain geometry suitable for computational mesh generation. A section of an underground mine excavation has been selected as an example of such transformation. Defining appropriate boundary conditions, especially the inlet velocity profile, is a challenging issue. Difficult environmental conditions in underground workings exclude the application of the most efficient and precise methods of velocity field measurements. Two attempts to define the inlet velocity profile have been compared. The first one used a sequence of simulations starting from a flat profile of a magnitude equal to the average velocity. The second one was based on the sixteen-point simultaneous velocity measurement, which gave consistency with measurement results within the range of applied velocity measurement method uncertainty. The article introduces a novel methodology that allows for more accurate replication of the mine excavation under study and the attainment of an appropriate inlet velocity profile, validated by a satisfactory correspondence between simulation outcomes and field measurements. The method involves analysing laser-scanned data of a mine excavation, conducting multi-point velocity measurements at specific cross-sections of the excavation that are unique to mining conditions, and utilising the k-ω SST turbulence model that has been validated for similar ventilation problems in mines.