1,720,960 research outputs found
Structure and dynamics of turbulent flows over highly permeable walls
No full textHighly porous materials are found in various industrial applications and environmental flows. In previous studies it was found that a turbulent flow along a highly porous wall experiences a higher skin friction as compared to a solid wall with similar surface roughness when the so-called permeability Reynolds number (Re_K) is larger than O(1). The main objective of the present study was to gain understanding of the characteristic structures and auto-generation mechanisms of turbulence for Re_K >> 1. To this purpose the Volume-Averaged Navier-Stokes (VANS) equations were solved in a Direct Numerical Simulation (DNS) of a turbulent flow through a plane channel with an upper solid wall and a lower porous wall at Re_K = 5.91. The DNS results are in good agreement with available Particle Image Velocimetry (PIV) data for the same flow geometry. A linear stochastic estimation technique was used to capture the structure associated with the characteristic ejection event that contributes most to the Reynolds shear stress near the porous wall. This structure is similar to a horseshoe vortex. Contrary to the conventional hairpin vortex found near solid walls, this horseshoe vortex has a significantly higher inclination angle with the wall and its legs are much shorter. The latter is consistent with the observed absence of low and high-speed streaks near highly permeable walls. Next, the auto-generation mechanisms of the horseshoe vortex were studied in another DNS in which the horseshoe vortex was released in the Reynolds-averaged flow field obtained from the former DNS. Two distinct auto-generation mechanisms were observed: (1) the generation of new structures at the upstream end of the horseshoe vortex, which evolve rapidly into a turbulent spot with an arrowhead shape, and (2) the interaction of the horseshoe vortex with spanwise oriented Kelvin-Helmholtz vortex rollers originating from the inflexion point in the mean velocity profile near the porous wall
Predicting growth rates of interfaces and internal layers in a turbulent boundary layer using a first order jump model
No full textExperimental research is presented on the characteristics of interfaces and internal layers that are present in a turbulent boundary layer (TBL). Both the turbulent non-turbulent interface (T/NT) and internal shear layers are detected in snapshots of the stereo-PIV data. It turns out that the internal layers exhibit similar characteristics compared to the T/NT interface. A theoretical approximation of the large scale boundary layer growth indicates that the correct boundary layer growth can be obtained by employing a modified first order jump model on the conditional statistics. Employing the same framework to the internal shear layers indicates that shear layers tend to move slower in close proximity to the wall, whereas they accelerate when moving away from the wall. Based on previous research it is believed that these internal layers separate large regions of approximately uniform momentum. Hence, boundary entrainment velocities may be interpreted as growth rates of large scale motions in a TBL
Scale interaction in a mixing layer: The role of the large-scale gradients
No full textThe interaction between scales is investigated in a turbulent mixing layer. The large-scale amplitude modulation of the small scales already observed in other works depends on the crosswise location. Large-scale positive fluctuations correlate with a stronger activity of the small scales on the low speed-side of the mixing layer, and a reduced activity on the high speed-side. However, from physical considerations we would expect the scales to interact in a qualitatively similar way within the flow and across different turbulent flows. Therefore, instead of the large-scale fluctuations, the large-scale gradients modulation of the small scales has been additionally investigated
Universal aspects of small-scale motions in non-equilibrium turbulent channel flow
No full textDirect numerical simulations of transient turbulent channel flow were conducted in order to study and characterize the large-small scale interaction in this type of flows. This was achieved by analyzing some well-known universal aspects of turbulence. Additionally, the so called strain-rate-eigenframe analysis was applied to study the local flow topology during the transient conditions of the flow evolution.Solid and Fluid MechanicsProcess and EnergyMechanical, Maritime and Materials Engineerin
Complete removal of ghost particles in Tomographic-PIV
No full textThis paper discusses and compares several methods, which aim to remove spurious peaks, i.e. ghost particles, from the volume intensity reconstruction in Tomographic-PIV. The assessment is based on numerical simulations of timeresolved tomographic-PIV experiments in linear shear flows. Within the reconstructed volumes intensity peaks are detected and tracked over time. These peaks are associated to particles (either ghosts or actual particles) and are characterized by their peak intensity, their size and their track length. Peak intensity and track length are found to be effective in discriminating between most ghosts and the actual particles, although not all ghosts can be detected using only a single threshold. The size of the reconstructed particles does not reveal an important difference between ghosts and actual particles. Simultaneous plotting of peak intensity and track length however does, under certain conditions, allow a complete separation of ghosts and actual particles. The ghosts can have either a high intensity or a long track length, but not both combined like all the actual particles. Finally removing the detected ghosts from the reconstructed volume and performing additional MART iterations can decrease the particle position error at low to moderate seeding densities, but increases the position error and tracking errors at higher densities.Process EnergyMechanical, Maritime and Materials Engineerin
Tomographic particle image velocimetry and its application to turbulent boundary layers
No full textTomographic Particle Image Velocimetry is a new experimental method developed to study three-dimensional motion in turbulent flows. The technique is an extension of standard PIV and makes use of several simultaneous views of illuminated tracer particles and their three-dimensional reconstruction as a light intensity distribution by means of tomography. The reconstructed tomogram pair is then analyzed by means of 3D cross-correlation returning the three-component velocity vector distribution over the measurement volume. The principles and details of the tomographic algorithm are discussed and a parametric study is carried out by to identify the most important parameters governing the experimental setup and to show their effect on the reconstruction accuracy. The capability of the technique in real experimental conditions is assessed with the measurement of the turbulent flow in the near wake of a circular cylinder. Next, this new technique has been applied to study the three-dimensional coherent structures in turbulent boundary layers. Quantitative visualizations of the individual (hairpin) vortices as well as the large-scale structures in both a low speed turbulent boundary layer and a high Reynolds number supersonic boundary layer have been obtained. The high Reynolds number data also suggests a very-large-scale flow organization exists not only in streamwise direction but also in spanwise direction. These very-large scale motions appear to consist of large-scale hairpins, which display a preferential alignment in streamwise direction and in the spanwise direction along the 45 degrees diagonal with the respect to the streamwise direction. Moreover, the time evolution of the flow structures is visualized in an experiment, in which the Tomographic-PIV technique is applied to nearly time-resolved image sequences recorded at 1.5 kHz.Aerospace Engineerin
Self-Sustaining Mechanisms in Wall Bounded Turbulence: Merging & auto-generation of vortices
No full textFor channel flow, we explore how a hairpin eddy may reach a threshold strength required to produce additional hairpins by means of auto-generation. This is done by studying the evolution of two eddies with different initial strengths (but both below the threshold strength), initial sizes and initial stream-wise spacing between them. The numerical procedure followed is similar to Zhou et al [1999]. The two eddies were found to merge into a single stronger eddy in case of a larger upstream and a smaller downstream eddy placed within a certain initial stream-wise separation distance. Subsequently, the resulting stronger eddy was observed to auto-generate new eddies. Merging of eddies thus is a viable explanation for the creation of the threshold strength eddies.Solid and Fluid MechanicsProcess and EnergyMechanical, Maritime and Materials Engineerin
Investigating the 3D Flow Structure Downstream of a Fence
No full textIn this research the experiments were performed in the water tunnel at the Laboratory for Aero and Hydrodynamics to investigate the 3D flow structure over roughness elements as a model for the flow over urban environment. It represents the first phase in an extended study on the dispersion of pollutants in an urban environment, which process is governed by the mentioned flow structure. This is an urgent problem as the air quality in cities is decreasing due to increasing pollution from cars and industrial activity. If we can improve the understanding of the dispersion of these gases, we can possibly improve this situation or at least better predict these problems. The principal experimental technique that will be employed is Tomographic-PIV , which is a state-of-the-art experimental technique to measure the flow velocity in a 3D volume. So far it has been applied mainly to investigate the flow structures in some classic turbulence cases: a cylinder wake and turbulent boundary layer over a flat plate. Its applicability to more complex geometries, like the rough wall considered here, is unexplored yet and needed to be assessed as a part of this project.Mechanical Engineering (Solid and Fluid Mechanics)Laboratory for Aero and HydrodynamicsMechanical, Maritime and Materials Engineerin
Pair dispersion statistics and coherent structures
No full textPair dispersion is studied to model scalar transport in many natural and industrial applications. The link between the particle pair dispersion and coherent flow structures is explored in this work. This was done by kinematically simulating tracer particles in an ideal flow structure [4] extracted from an isotropic turbulent flow. It was found that the variation of the mean and the mean square separation lengths with time were qualitative similar to the results in actual turbulent flows. It was also observed that the quantitative results matched till 4-5 Kolmogrov time units. Ideal structure with two vortices and a shear layer was able to emulate the qualitative results. Is the combination of shear layer and one/two vortices is sufficient or necessary to emulate pair dispersion statistics needs to be studied in the future.Process and EnergyMechanical, Maritime and Materials Engineerin
- …
