On the role of turbulent large-scale streaks in generating sediment ridges

The role of turbulent large-scale streaks or large-scale motions in forming subaqueous sediment ridges on an initially flat sediment bed is investigated with the aid of particle resolved direct numerical simulations of open channel flow at bulk Reynolds numbers up to 9500. The regular arrangement of quasi-streamwise ridges and troughs at a characteristic spanwise spacing between 1 and 1.5 times the mean fluid height is found to be a consequence of the spanwise organisation of turbulence in large-scale streamwise velocity streaks. Ridges predominantly appear in regions of weaker erosion below large-scale low-speed streaks and vice versa for troughs. The interaction between the dynamics of the large-scale streaks in the bulk flow and the evolution of sediment ridges on the sediment bed is best described as ‘top-down’ process, as the arrangement of the sediment bedforms is seen to adapt to changes in the outer flow with a time delay of several bulk time units. The observed ‘top-down’ interaction between the outer flow and the bed agrees fairly well with the conceptual model on causality in canonical channel flows proposed by Jiménez (J. Fluid Mech., vol. 842, 2018, P1, § 5.6). Mean secondary currents of Prandtl's second kind of comparable intensity and lateral spacing are found over developed sediment ridges and in single-phase smooth-wall channels alike in averages over ${O}(10)$ bulk time units. This indicates that the secondary flow commonly observed together with sediment ridges is the statistical footprint of the regularly organised large-scale streaks.

Surface streaming on nonbreaking wind waves

<p>The energetic, coherent vortical motions in the aqueous surface layer beneath the wind waves dominate the liquid-phase controlled transport processes across the air-water interface. Through interacting with the interface, these coherent vortices manifest themselves by forming quasi-streamwise, high-speed streaks on the wind waves. The density of these streamwise streaks, which can be quantified by the transverse spacing of streaks, thus characterizes the interfacial transfer contributed by the coherent vortices. The formation of surface streaming on the wind waves is geometrically similar to the low-speed streaks observed in the turbulent wall layers. It is generally accepted that the mean spanwise spacing between these low-speed streaks, when scaled by the viscous length, would exhibit a universal value of 100. Observations in wind-wave flumes, however, show that the transverse scale between high-speed streaming on nonbreaking wind waves is narrower than that between low-speed streaks next to no-slip wall. Comparative numerical simulations of shear flow bounded by flat and wavy surfaces are conducted to explain the variation. Analysis of the vorticity transport in the simulated flows bounded by a wavy surface reveals that the presence of surface waves enhances the production of streamwise enstrophy and, consequently, intensifies the generation of quasi-streamwise vortices that form the elongated streaks.<br>This work is supported by the Taiwan Ministry of Science and Technology (107-2611-M-002 -014 -MY3 and 110-2923-M-002 -014 -MY3).</p>

Kinematics and Dynamics of Turbulent Bands at Low Reynolds Numbers in Channel Flow

Channel flow turbulence exhibits interesting spatiotemporal complexities at transitional Reynolds numbers. In this paper, we investigated some aspects of the kinematics and dynamics of fully localized turbulent bands in large flow domains. We discussed the recent advancement in the understanding of the wave-generation at the downstream end of fully localized bands. Based on the discussion, we proposed a possible mechanism for the tilt direction selection. We measured the propagation speed of the downstream end and the advection speed of the low-speed streaks in the bulk of turbulent bands at various Reynolds numbers. Instead of measuring the tilt angle by treating an entire band as a tilted object as in prior studies, we proposed that, from the point of view of the formation and growth of turbulent bands, the tilt angle should be determined by the relative speed between the downstream end and the streaks in the bulk. We obtained a good agreement between our calculation of the tilt angle and the reported results in the literature at relatively low Reynolds numbers.

Spatial and statistical analysis of coherent vortical structures within a stress-driven free-surface turbulent shear flow

<p>The water surface under high wind condition is characterized by elongated high-speed streaks and randomly emerged low-speed streaks, which are attributed to underneath coherent vortical motions. These vortical structures within aqueous turbulent boundary layer plays a critical role in turbulent exchange, their characteristics and statistics are therefore of interest in this study. Direct numerical simulation of an aqueous turbulent flow bounded by a stress-driven flat free surface was performed. Simulation results of cases with high wind condition (surface friction velocity = 1.22 cm/s) as well as weak wind condition (surface friction velocity = 0.71 cm/s) are analyzed. To identify the underlying vortical structures, an indicator of swirling strength derived from local velocity-gradient tensor is adopted. A formal classification scheme, based on the topological geometry of the vortex core, is then applied to classify the identified structures. Surface layers with the two wind conditions reveal similar results in statistics and spatial distribution of vortical structures. Two types of characteristic vortices which induce the surface streaks are extracted, including quasi-streamwise vortex and reversed horseshoe vortex (head pointing upstream), most inclining at about 10 to 20 degrees. Quasi-streamwise vortices are the dominant structure, and both high- and low-speed streaks are fringed with such vortices; they adjoin the surface streaks as counter-rotating arrays in either staggered or side-by-side spatial arrangement. The length of quasi-streamwise vortices, however, are significantly shorter than the corresponding surface streaks, only 10% of the extracted quasi-streamwise vortices are longer than 150 wall units. Reversed horseshoe vortices, associated with downwelling motions and surface convergence, are located beneath the high-speed streaks. In contrast to the turbulent boundary layer next to a flat wall, typical forward horseshoe vortices (head pointing downstream) associated with upwelling motions are barely found within the free-surface turbulent shear flow.</p><p>This work was supported by the Taiwan Ministry of Science and Technology (MOST 107-2611-M-002 -014 -MY3).</p>

Large eddy simulation of bed formation in subaqueous bedload

<p>Large eddy simulation of incompressible turbulent flow over a loose bed of spherical particles are investigated in an open channel. Eulerian and Lagrangian point-particle methods is applied to solve the Navier-Stokes equations and particle motion respectively and the particle-flow interaction is also considered. A new method solving particle-particle collisions is utilized for the first time to reduce the computational time spending on calculating the pairwise distances between particles. A turbulent fluid condition from the experiment of Robert and Uhlman (2001) are chosen of which the corresponding sediment patterns are ‘ripple’. Flow over the formed bed is considered and it is found that double-averaged Reynolds stresses including shear stress and three normal stresses reach their peak values near the bed. However, affected by the movable bed, they decay quickly as the height increases. The flow direction slightly rises over the stoss-side of ripples and falls after the crest and the velocity magnitude of time-averaged flow accelerate and decelerate before and after the crest of ripples as well. Hence, recirculation zones and clockwise vorticity appear at the trough of the bed where kolk boil vortices like hairpins and elongated streamwise vortices is also evident. Coherent structures, in the form of high- and low-speed streaks near the bed are also affected by the bed formation. The near-bed low-speed streaks entrain into the main flow domain over the stoss-side of ripples and the high-speed fluid streaks from the main flow rush toward the bed over the leeside of ripples. The bedload transport rate is well represented by previous empirical formulas. The bed surface elevation changes from upstream to downstream with time and there is a difference in the direction of sediment transport which is obvious three-dimensionality. In addition, it is also found that the bed surface elevation shows a positive correlation with particle streamwise velocity and entrainment rate which means higher bed elevation leads to larger bedload transport rate.</p>