Negative Streamwise Velocity Research Articles

Dynamic Analysis of Flow Field at the End of Combustor Simulator

This study was carried out in order to extend database knowledge about the flow field characteristics and define the various flow field contours inside a combustor simulator. The modern gas turbine industries try to get higher engine efficiencies. Brayton cycle is a key to achieve this purpose. According to this cycle industries should increase the turbine inlet temperature to get more engine efficiency and power. However the turbine inlet temperature increasing creates an extremely harsh environment for the downstream critical components such as turbine vanes. In this research a three-dimensional representation of a true Pratt and Whitney aero-engine which studied before in Virginia University was simulated and analyzed to collect essential data. This combustor simulator combined the interaction of two rows of dilution jets, which were staggered in the stream wise direction and aligned in the span wise direction, with that of filmcooling along the combustor liner walls. The overall findings of the study indicate that three-component velocity measurements showed the dilution jet-mainstream interaction produced shear forces and as a result a counter-rotating vortex pair was created. The highest turbulent kinetic energy was found at the top of recirculating region due to the interaction of the second row of dilution jets and mainstream flow. Furthermore, the centers of the counter-rotating vortex pair were spread relatively far apart due to the opposing dilution jets. Along the dilution jet centerline, negative stream wise velocities were measured indicating the recirculation region just downstream of the jet. Into the combustor exit, the acceleration of the flow increased and thereby the uniformity of the velocity profile enhancement was found as well.

Turbulent decay in the near field of multi-scale and conventional grids

Wind tunnel turbulence generated by a conventional and two multi-scale grids has been investigate. The grids were all designed to produce turbulence with the same integral scale, so that a direct comparison could be made between the flows, both in physical and scaled space. It has been suggested in the literature (e.g. Hurst and Vassilicos, 2007) that for a particular class of multi fractal grids, the turbulence decay depends exponentially on the distance from the grid. After a short distance where the flow is highly dependent of the geometry, it was found that the exponential decay is not unique to a particular geometry, but may be found over the same streamwise distances also behind the multi-scale grids, as well as for the conventional grid.By comparing the probability density functions measured using laser Doppler and hot wire anemometry it is shown that hot wire measurements may contain severe errors if taken too close to the grid. It is shown that negative streamwise velocity components may occasionally be found as far as 10 times the mesh widths downstream of the grid. Since hot wire anemometry is not able to measure the sign of the velocity vector, this leads to a folding of the data which artificially increases the derived mean velocity and, more seriously, reduces the width of the probability distribution. Hence the interpreted turbulent stress is reduced.

Rare backflow and extreme wall-normal velocity fluctuations in near-wall turbulence

Rare negative streamwise velocities and extreme wall-normal velocity fluctuations near the wall are investigated for turbulent channel flow at a series of Reynolds numbers based on friction velocity up to Reτ = 1000. Probability density functions of the wall-shear stress and velocity components are presented as well as joint probability density functions of the velocity components and the pressure. Backflow occurs more often (0.06% at the wall at Reτ = 1000) and further away (up to y+ = 8.5) from the wall for increasing Reynolds number. The regions of backflow are circular with an average diameter, based on ensemble averages, of approximately 20 viscous units independent of Reynolds number. A strong oblique vortex outside the viscous sublayer is found to cause this backflow. Extreme wall-normal velocity events occur also more often for increasing Reynolds number. These extreme fluctuations cause high flatness values near the wall (F(v) = 43 at Reτ = 1000). Positive and negative velocity spikes appear in pairs, located on the two edges of a strong streamwise vortex as documented by Xu et al. [Phys. Fluids 8, 1938 (1996)]10.1063/1.868973 for Reτ = 180. The spikes are elliptical and orientated in streamwise direction with a typical length of 25 and a typical width of 7.5 viscous units at y+ ≈ 1. The negative spike occurs in a high-speed streak indicating a sweeping motion, while the positive spike is located in between a high and low-speed streak. The joint probability density functions of negative streamwise and extreme wall-normal velocity events show that these events are largely uncorrelated. The majority of both type of events can be found lying underneath a large-scale structure in the outer region with positive sign, which can be understood by considering the more intense velocity fluctuations due to amplitude modulation of the inner layer by the outer layer. Simulations performed at different resolutions give only minor differences. Results from experiments and recent turbulent boundary layer simulations show similar results indicating that these rare events are universal for wall-bounded flows. In order to detect these rare events in experiments, measurement techniques have to be specifically tuned.

Negative streamwise velocities and other rare events near the wall in turbulent flows

Negative streamwise velocities, extreme wall-normal velocites and high flatness values for the wall-normal fluctuations near the wall are investigated for turbulent channel flow simulations at a series of Reynolds numbers up to Reτ = 1000 in this paper. Probability density functions of the wall-shear stress and velocity components are presented, as well as joint probability density functions of the velocity components and the pressure. Backflow occurs more often (0.06% at Reτ = 1000) and further away from the wall into the buffer layer for rising Reynolds number. An oblique vortex outside the viscous sublayer is found to cause this backflow. Extreme v events occur also more often for rising Reynolds number. Positive and negative velocity spikes appear in pairs, located on the two edges of a strong streamwise vortex: the negative spike occurring in a high speed streak indicating a sweeping motion, while the positive spike is located between a high and low speed streak. These extreme v events cause high flatness values near the wall (F(v) = 43 at Reτ = 1000).

Image analysis applied to study on frictional-drag reduction by electrolytic microbubbles in a turbulent channel flow

We investigate frictional-drag reduction with electrolytic microbubbles based on image measurement of a turbulent flow in a water channel at Re = 4800 (based on the half channel height). Microbubbles with a diameter ranging 30–200 μm can reduce frictional drag by as much as 30% relative to single-phase flow even at low void fractions (α ≈ 3 × 10−4); however, drag reduction is only effective within a limited downstream distance from an electrode array. Arrangement of the optical system allows us to measure the bubble-production rate by water electrolysis from images near the wall and to trace the motion of bubbles. We also measure velocity fields using particle-tracking velocimetry based on a shallow depth-of-field approach by segregating tracer particles from microbubbles. Vertically oscillating microbubbles likely represent interaction with vortical structures near the wall, and bubbles approaching the wall appear to induce negative streamwise velocity relative to the surrounding fluid. We relate the wall friction with the double integral of the Reynolds-stress profile and show that its variation due to microbubbles decreases the drag on the wall. Microbubbles tend to coalesce downstream resulting in a fewer bubbles but with greater size; accordingly, the oscillatory motion diminishes, and the frictional drag rather increases.

DNS and LES of estimation and control of transition in boundary layers subject to free-stream turbulence

Transition to turbulence occurring in a flat-plate boundary-layer flow subjected to high levels of free-stream turbulence is considered. This scenario, denoted bypass transition, is characterised by the non-modal growth of streamwise elongated disturbances. These so-called streaks are regions of positive and negative streamwise velocity alternating in the spanwise direction inside the boundary layer. When they reach large enough amplitudes, breakdown into turbulent spots occurs via their secondary instability. In this work, the bypass-transition process is simulated using direct numerical simulations (DNS) and large-eddy simulations (LES). The ADM-RT subgrid-scale model turned out to be particularly suited for transitional flows after a thorough validation. Linear feedback control is applied in order to reduce the perturbation energy and consequently delay transition. This case represents therefore an extension of the linear approach (Chevalier, M., Hoepffner, J., Åkervik, E., Henningson, D.S., 2007a. Linear feedback control and estimation applied to instabilities in spatially developing boundary layers. J. Fluid Mech. 588, 163–187, 167–187.) to flows characterised by strong nonlinearities. Control is applied by blowing and suction at the wall and it is both based on the full knowledge of the instantaneous velocity field (i.e. full information control) and on the velocity field estimated from wall measurements. The results show that the control is able to delay the growth of the streaks in the region where it is active, which implies a delay of the whole transition process. The flow field can be estimated from wall measurements alone: The structures occurring in the “real” flow are reproduced correctly in the region where the measurements are taken. Downstream of this region the estimated field gradually diverges from the “real” flow, revealing the importance of the continuous excitation of the boundary layer by the external free-stream turbulence. Control based on estimation, termed compensator, is therefore less effective than full information control.

Evidence of very long meandering features in the logarithmic region of turbulent boundary layers

A regime of very long meandering positive and negative streamwise velocity fluctuations, that we term ‘superstructures’, are found to exist in the log and lower wake regions of turbulent boundary layers. Measurements are made with a spanwise rake of 10 hot-wires in two separate facilities (spanning more than a decade ofReτ) and are compared with existing PIV and DNS results. In all cases, we note evidence of a large-scale stripiness in the streamwise velocity fluctuations. The length of these regions can commonly exceed 20δ. Similar length scales have been previously reported for pipes and DNS channel flows. It is suggested that the true length of these features is masked from single-point statistics (such as autocorrelations and spectra) by a spanwise meandering tendency. Support for this conjecture is offered through the study of a synthetic flow composed only of sinusoidally meandering elongated low- and high-speed regions. From detailed maps of one-dimensional spectra, it is found that the contribution to the streamwise turbulence intensities associated with the superstructures appears to be increasingly significant with Reynolds number, and scales with outer length variables (δ). Importantly, the superstructure maintains a presence or footprint in the near-wall region, seeming to modulate or influence the near-wall cycle. This input of low-wavenumber outer-scaled energy into the near-wall region is consistent with the rise in near-wall streamwise intensities, when scaled with inner variables, that has been noted to occur with increasing Reynolds number. In an attempt to investigate these structures at very high Reynolds numbers, we also report on recent large-scale sonic anemometer rake measurements, made in the neutrally stable atmospheric surface layer. Preliminary results indicate that the superstructure is present in the log region of this atmospheric flow atReτ= 6.6×105, and has a size consistent with outer scaling.

On the Development of Coherent Structure in a Plane Jet

In order to clarify the dynamics of the coherent structure in a turbulent plane jet, the simultaneous measurement of the main streamwise velocity at 21 points in the self-preserving region of a turbulent plane jet has been performed by an array of I-type hot-wire probes. Then the KL (Karhunen-Loeve) expansion was applied to extract the coherent structure in the jet. The total number of eigenfunctions (modes) is N=21, which corresponds to the number of probes. The eigenfunctions (modes) are numbered in order of magnitude of their corresponding eigenvalues. From the investigation of the random coefficients and the eigenfunctions (modes), it is found that the low-numbered (energetic) modes represent the large scale (coherent) structure, the middle-numbered modes represent the finer (small-scale) random structure, and the higher-numbered modes contribute mainly to the intermittent structure in the outer edge region. From the spatio-temporal velocity field reconstructed by the first KL mode, it is found that there exist a pair of fluid lumps with the positive and negative streamwise velocity fluctuation on the opposite sides of the jet centerline, and the signs of velocity fluctuation for those fluid lumps change alternately as time proceeds. These characteristics are consistent with the so called “jet flapping” phenomenon.

On the Development of Coherent Structure in a Plane Jet

In order to clarify the dynamics of the coherent structure in a turbulent plane jet, the simultaneous measurement of the main streamwise velocity at 21 points in the self-preserving region of a turbulent plane jet has been performed by an array of I-type hot-wire probes. Then the KL (Karhunen-Loeve) expansion was applied to extract the coherent structure in the jet. The total number of eigenfunctions (modes) is N=21, which corresponds to the number of probes. The eigenfunctions (modes) are numbered in order of magnitude of their corresponding eigenvalues. From the investigation of the random coefficients and the eigenfunctions (modes), it is found that the low-numbered (energetic) modes represent the large scale (coherent) structure, the middle-numbered modes represent the finer (small-scale) random structure, and the higher-numbered modes contribute mainly to the intermittent structure in the outer edge region. From the spatio-temporal velocity field reconstructed by the first KL mode, it is found that there exist a pair of fluid lumps with the positive and negative streamwise velocity fluctuation on the opposite sides of the jet centerline, and the signs of velocity fluctuation for those fluid lumps change alternately as time proceeds. These characteristics are consistent with the so called “jet flapping” phenomenon.

Nonparallel linear stability analysis of Long’s vortex

A nonparallel linear stability analysis of a family of self-similar vortex cores which includes Long’s vortex as a particular member is performed using parabolized stability equations (PSE). The resulting streamwise variation of both the spatial growth rate and the axial wave number of the different unstable modes is compared with the results from a local spatial stability analysis which also takes into account the effects of viscosity and of the streamwise variation of the basic flow, so that the effect of the history of the disturbances on their stability is quantified. It is shown that this last effect is negligible for high Reynolds numbers, but becomes increasingly important as the Reynolds number decreases, especially for very small growth rates. The marching method used to solve the PSE is computationally much faster than the standard methods for solving the nonlinear eigenvalue problem resulting from the local stability equations. As a new result, the local spatial calculations reveal the existence of unstable counter-rotating spiral modes with negative group velocities for Type II Long’s vortices (that is, vortices with negative streamwise velocity at the axis), thus showing that these flows are subcritical in Benjamin’s sense. This kind of instability does not appear for Type I vortices, which can only sustain non-axisymmetric convective instabilities, and are therefore supercritical. Thus, the spatial stability analysis establishes a fundamental distinction between Type I and Type II Long’s vortices.

The Use of Pattern Recognition and Proper Orthogonal Decomposition in Identifying the Structure of Fully-Developed Free Turbulence

The eigenvectors obtained from proper orthogonal decomposition (POD) are used as the selection criteria to classify the individual events contained in date files of two-component velocity signals recorded in the non-homogeneous (vertical) and homogeneous (horizontal) planes of a fully-developed turbulent cylinder wake. This procedure uses the dominant eigenvectors from POD as initial templates to perform a pattern recognition (PR) analysis of the signals so that the individual coherent events appearing randomly in the signals can be educed more objectively. It is found that the prototype or ensemble average of the group of events classified as the large-scale structure in the vertical plane has a circulatory motion with strong negative streamwise and outward lateral velocity fluctuations. In the horizontal plane, the average structure is a double roller with negative streamwise velocity fluctuations in its centerplane. The class of instantaneous events selected contribute significantly to the variance in the outer intermittent region, but much less in the fully turbulent core. While the first eigenvector from POD captures some of the features shown by this prototype, it is not enough to depict by itself all the characteristics shown by the ensemble average of the repetitive three-dimensional instantaneous events that occur in the fully-developed turbulent wake.

LDV Measurements in Separated Flow on an Elliptic Wing Mounted at an Angle of Attack on a Wall

The flow around a half-ellipsoid with axes ratio 12:6:1 mounted on a plane wall at an angle of attack of 25° and at a Reynolds number of 47,000 (based on the maximum chord) was studied using a three-dimensional LDV system. In this paper the mean velocity distributions in a volume enclosing the separated region and the near wake are described. The flow is shown to be consistent with the findings of Johnson (1991) based on flow visualization and related topological analysis. The flow on most of the pressure side was attached and laminar, while that on the suction side and in the wake was separated and turbulent. The region of separation had mean negative streamwise velocities as large as 25 percent of the freestream velocity, and the reversed-flow region extended up to one chord length behind the body. The measurements clearly reveal the complex vortex structure that arises from the three-dimensional separation.