Super-resolution reconstruction of turbulent flows using deep learning has gained significant attention, yet challenges remain in accurately capturing physical small-scale structures. This study introduces the Conditional Enhanced Super-Resolution Generative Adversarial Network (CESRGAN) for reconstructing high-resolution turbulent velocity fields from low-resolution inputs. CESRGAN consists of a conditional discriminator and a conditional generator, the latter being called CoGEN. CoGEN incorporates subgrid-scale (SGS) turbulence kinetic energy as conditional information, improving the recovery of small-scale turbulent structures with the desired level of energy. By being aware of SGS turbulence kinetic energy, CoGEN is relatively insensitive to the degree of detail in the input. As shown in the paper, its advantages become more pronounced when the model is applied to heavily filtered input. We evaluate the model using direct numerical simulation (DNS) data of forced homogeneous isotropic turbulence. The analysis of Q-criterion isosurfaces, energy spectra, and probability density functions shows that the proposed CoGEN reconstructs fine-scale vortical structures more precisely and captures turbulent intermittency better compared to the traditional generator. Particle-pair dispersion simulations validate the physical fidelity of CoGEN-reconstructed fields, closely matching DNS results across various Stokes numbers and filtering levels. This paper demonstrates how incorporating available physical information enhances super-resolution models for turbulent flows.

In the present work a detailed numerical study of the parameters influencing the droplet formation in a flow-focusing microfluidic device are made. First, an extensive verification of the simulations with data from the literature is carried out. Influence of parameters like viscosity and inflow velocity are compared with the results from literature showing a good agreement. Some differences are attributed to the different numerical techniques used: in the present work a pure volume-of-fluid method is used, while in the reference study this method is combined with the level-set method. As a second step of the verification of the present model, a comparison with experimental data from the literature was carried out which shows a very good agreement. After the verification was completed, eight new simulations are carried out covering a wide range of velocities of the continuous phase uc . In these simulations the velocity of the discrete phase ud remains unchanged. The variation of the continuous phase velocity reveals that with increasing the value of uc , respectively the value of the capillary number Ca, the droplet length reaches a point of saturation, i.e. a point where the droplet length does not decrease any more. For the present setup this saturation occurs for Ca > 0,03. On the other hand, when the velocity of the continuous phase goes towards very low values (Ca < 0,01 for the present setup), the droplet size increases significantly. Further, it was found that for increasing capillary numbers Ca above a value around 0,015 for water/oil and above 0,025 for water + 40% glycerol / oil systems, a transmission from the dripping towards the jetting regimes of droplet formation occurs. It was shown that when the viscosity of the continuous phase increases, higher total pressure jumps in the droplet occur, also leading to the formation of smaller droplets.

Tree belts are commonly applied over agricultural terrain where seeds of wheat and other vegetation are planted in the ground in order to prevent the seeds from being blown by the wind. The tree belt comprises a long and thin (10–20 m thick) section of trees, which spans in a direction normal to the prevailing wind direction. While serving its agricultural goal, the belt does inevitably modify the boundary layer profile of the wind. This, on its part, is likely to affect the operation of small-scale wind turbines installed in the vicinity of the belt. The goal of this study is to determine the span and range at which this effect manifests itself. It was found that in the near vicinity downstream and slightly above the tree belt, the wind velocity actually increased due to the mass conservation. The flow became independent on the tree belt drag coefficient when its value was higher than 0.2 1/m. The turbulence introduced by the belt was restricted to a height of 1.5–2 tree belts.

The paper presents the theoretical basis for the extension of the V $\pi$ LES method, originally developed in recent works of the authors for incompressible flows, to flows with variable density and transport properties but without chemical reactions. The method is based on the combination of grid based and grid free computational particle techniques. Large scale motions are modelled on the grid whereas the fine scale ones are modelled by particles. The particles represent the fine scale vorticity, and scalar quantities like e.g., temperature, mass fractions of species, density and mixture fraction. Coupled system of equations is derived for large and fine scales transport.