Secondary Flow Cells Research Articles

Analysis of the role of turbulence in curved open-channel flow at different water depths by means of experiments, LES and RANS

In order to unravel the main flow and secondary flow characteristics and the role of turbulence in a curved single-bend open-channel flow, large-eddy simulations (LES) and Reynolds-averaged numerical simulations (RANS) were carried out and compared with experiments of the flow through a strongly bent laboratory flume. Turbulence was found to play an important role with respect to processes that are important in natural rivers. The strength of the curvature-induced secondary flow in the core of the flow domain, which is the most typical feature of curved open-channel flow, depends on the turbulence. Turbulence is especially important in the flow regions near the banks. Only the LES model is able to resolve accurately the boundary layer detachment and the formation of an internal shear layer at the inner bank as well as the outer-bank cell of secondary flow, whereas the RANS model is unable to reproduce these processes. Turbulence also conditions the magnitude of the bed shear stress, as indicated by the considerable overestimations of the bed shear stress by the RANS model as compared with the LES model. As a result, only LES properly reproduces the experimentally measured overall friction losses over the bend, whereas RANS overestimates these losses. The large scales of turbulence play an essential role in generating these flow processes by means of their interaction with the mean secondary flow structures, whereas small-scale turbulence is merely dissipative and does not play a dynamic role with respect to the time-averaged flow pattern. This implies that the simulated flow field is rather insensitive to the sub-grid model in the LES computation.

Large-eddy simulation of a mildly curved open-channel flow

After validation with experimental data, large-eddy simulation (LES) is used to study in detail the open-channel flow through a curved flume. Based on the LES results, the present paper addresses four issues. Firstly, features of the complex bicellular pattern of the secondary flow, occurring in curved open-channel flows, and its origin are investigated. Secondly, the turbulence characteristics of the flow are studied in detail, incorporating the anisotropy of the turbulence stresses, as well as the distribution of the kinetic energy and the turbulent kinetic energy. Moreover, the implications of the pattern of the production of turbulent kinetic energy is discussed within this context. Thirdly, the distribution of the wall shear stresses at the bottom and sidewalls is computed. Fourthly, the effects of changes in the subgrid-scale model and the boundary conditions are investigated. It turns out that the counter-rotating secondary flow cell near the outer bank is a result of the complex interaction between the spatial distribution of turbulence stresses and centrifugal effects. Moreover, it is found that this outer bank cell forms a region of a local increase of turbulent kinetic energy and of its production. Furthermore, it is shown that the bed shear stresses are amplified in the bend. The distribution of the wall shear stresses is deformed throughout the bend due to curvature. Finally, it is shown that changes in the subgrid-scale model, as well as changes in the boundary conditions, have no strong effect on the results.

Morphology, flow structure, and suspended bed sediment transport at two large braid‐bar confluences

Mid‐channel bars and their associated confluences are key morphodynamic nodes within braided rivers, with past studies having investigated the morphodynamics of small natural channels or laboratory models with relatively low width/depth (W/D) ratios, typically at <10. This paper investigates the morphology, suspended bed sediment distribution, and flow structure at two large braid bar confluences in the Río Paraná (Argentina), wherein W/D ratios are much higher (approaching 100) than in smaller channels. The results highlight the significant control of the cross‐sectional distribution of downstream flow velocity on confluence flow, suspended bed sediment concentration, and morphodynamics and indicate that this factor may become progressively more significant with increasing channel scale and W/D ratio, particularly when simple discharge (or momentum) ratios between the incoming flows are used to explain the flow dynamics. Additionally, secondary flow cells, often proposed to occupy a large part of the channel width in small river channel confluences, are only identified in relatively small portions of the channel width at these larger spatial scales. Such a restriction seems related to the generative mechanisms of secondary flows at these higher W/D ratios, which are likely to be dominated by turbulence generated along the mixing layer between the two flows and topographic influences that limit the spatial extent of these effects. This paper highlights the importance of these findings with respect to the flow and sediment dynamics of large channel confluences.

The effect of floodplain roughness on flow structures, bedforms and sediment transport rates in meandering channels with overbank flows: Part I

Studies were carried out to understand the effect of floodplain roughness on flow structures, sediment transport rates and bedforms in a mobile meandering channel with overbank flows. Three floodplain roughnesses were examined in this study. Flow structures and bedforms were measured using a three component laser Doppler anemometer (LDA) system and digital photogrammetry, respectively. Comparisons of flow structures and bedforms between different floodplain rougheness are made. Considerable changes in the flow structure and bed form were observed. In particular, multiple secondary flow cells along the meandering channel occur at deeper water depths as the floodplain roughness increases. These cells also cause a series of wavy bedforms along the meandering channel. This paper is separated into two parts. Part I concentrates, in detail, on the flow structure with bedforms in the meandering channel for overbank flow as to change of floodplain roughness. The sister paper, Part II, concentrates on bed form formation during flood, sediment transport rates and flow resistance in the main channel and floodplain.

Secondary flow and roll cells interaction in high-aspect-ratio rotating turbulent duct flows

End-wall effects for high aspect ratio (AR) turbulent duct flows under moderate spanwise rotation are investigated using Reynolds-Averaged Navier–Stokes (RANS) calculations with a Reynolds stress turbulence closure model. It is shown that despite an important uniformisation of the mean streamwise flow compared to the non-rotating case, the channel flow solution (AR = ∞) is not recovered in practical high AR ducts used in experiments. The unavoidable end-wall generated secondary flow causes transverse advection which is capable of altering the mean velocity profile, even for AR as high as 22. In addition, for Re = 40,000 and Ro = 0.22, persistent longitudinal roll cells are found in the RANS solutions. The results suggest that their interaction with the secondary flow may challenge the prospect of formally reaching a steady, streamwise invariant regime in actual rotating duct experiments.

Comparison of Fixed- and Moving-Vessel Flow Measurements with an aDp in a Large River

A comparison of three-dimensional flow velocity measurements, made with an acoustic Doppler profiler (aDp) from fixed and moving vessels at cross sections of the Paraná River, Argentina, was performed. The purpose was to design a rapid and reliable procedure for quantifying the velocity field, and related parameters such as bed shear velocity and the identification of secondary circulations, in large rivers using an aDp. The fixed-vessel measurements were performed over a period of 10min at three vertical profiles along two of the sections. These data were then compared with the results of ten moving-vessel repeat transects made at each of the sections, and which intersected the fixed-vessel sampling locations, using a number of different aDp setup configurations. From the velocity profiles obtained with both fixed- and moving-vessel measurements, total bed shear velocity values were computed by applying the law-of-the-wall. The results indicate there can be significant differences between velocities obtained using the moving-vessel method and fixed-vessel measurements averaged over 10min. These differences in horizontal velocity can be significantly reduced by averaging five, or more, moving-vessel transects, with corresponding shear velocities calculated from five-transect averages showing differences ranging between 10 and 15%, dependent on the aDp configuration. Location of the at-a-point vertical velocity profile in relation to large-scale bed roughness may also be an influential factor, and ideally the bed morphology should be quantified together with the aDp-derived velocities. When using the aDp to identify secondary flow cells, it was found that although one cross-section transect can provide a reasonable overall picture, an average of five cross sections is necessary to resolve the finer details of flow. The implications for applications that use moving-vessel techniques for measurement and analysis of three-dimensional flow structures, including secondary flows, are highlighted.

Estimation of boundary shear stress distribution in open channels using flownet

An approach using flownet to calculate the boundary shear stress distribution along the wetted perimeter in a two dimensional flow is proposed. This approach is based on the assumption that the energy production balances the energy dissipation within the sub-area between two adjacent potential lines. The energy exchange induced by the large scale secondary flow cells between the sub-areas is neglected. The free water surface is used as the imaginary boundary to which the potential lines are perpendicular. The wetted cross section is mapped by orthogonal stream and potential lines. The wetted perimeter and cross section is divided into a number of segments and sub-area by the potential lines. The boundary shear stress over each boundary segment is the gravity force of the corresponding sub-area in the streamwise direction. The flownet method is applied to circular, rectangular, trapezoidal, and compound channels. The comparisons with experimental data and other analytical methods show that the flownet method estimates well the boundary shear stress distribution along the wetted perimeter, especially near the water margins and corners of the channel boundary where most of the existing analytical methods fail

Flow processes and sedimentation in submarine channel bends

Turbidity currents in sinuous submarine channels are an important mechanism for transporting terrestrial sediments to deep water, and their deposits are of increasing importance as hydrocarbon exploration targets. Despite this, the architecture and dynamics of submarine channel systems are not well understood. Analogies are often drawn with fluvial systems due to similarities between their planform shapes even though differences in channel evolution and hydrodynamics have been noted. A key question is the nature of deposition within submarine channel bends; in particular at inner bends where point bars form in alluvial meandering rivers. Recent experimental and numerical work has demonstrated that the fluid dynamics of submarine channel bend flow are markedly different from rivers. Notably, a reversal in the orientation of secondary (helical) flow at bend apices occurs in submarine channels. The potential influence of these differences in fluid dynamics on deposition within submarine channel bends is investigated herein. We report the results of a series of physical experiments in which solute-driven gravity currents were run through pre-formed sinuous channels containing mobile beds. These experiments reveal sedimentation patterns characterised by accumulation zones downstream of bend apices and erosion zones at outer bends. These patterns are broadly analogous to the point bars and outer-bank pools observed in meandering rivers, demonstrating that the longitudinal flow component dominates over the cross-stream component, as also occurs in rivers. However, the data suggest that the reversal in direction of the cross-stream flow component compared with subaerial flows is important in determining the position and morphology of ‘point bars’ relative to bend apices. From analogy with fluvial compound channels, and fluvial theory, this reversal in secondary flow cell orientation is also expected to influence the spatial variations of grain size in submarine channel ‘point-bar’ deposits.

The influence of scale, slope and channel geometry on the flow dynamics of submarine channels

Submarine channels are major morphological features of the sea floor and are important in the transport of sediment to the deep ocean. Although much is known concerning the large-scale distribution of sediment within and surrounding submarine channels, there is little understanding of the fluid dynamic processes that control this sedimentation. Direct measurement of flow velocities and concentrations has proved to be extremely difficult within submarine channels, with the resultant paucity of direct observations making physical laboratory modelling a critical technique for examining the processes that operate in, and control, submarine channel development. Recent experimental and numerical studies have proposed a new model of secondary circulation within submarine channel bends, characterised by a reversal in the orientation of the secondary circulation cell relative to that found in meandering rivers. This new paradigm for submarine channels thus predicts basal flow from the inside to the outside of the bend at a bend apex, with an upper return flow directed towards the inner bend. The reversal in orientation of the secondary flow cell has been linked to the vertical distribution of downstream velocity and associated changes in centrifugal and pressure gradient forces. However, previous work has additionally proposed that shearing of the within-channel flow by overbank flow may also generate secondary flow reversal. This study assesses the applicability of the proposed submarine bend flow model against a range of key channel parameters. We demonstrate that the sense of secondary circulation is the same for all experimental conditions, strongly supporting the new model of secondary flow in submarine channels. Furthermore, investigation of overbank shear induced secondary circulation confirms for the first time that this mechanism can occur, and identifies the channel styles most likely to exhibit this effect. Such shear-induced circulation is, however, shown to be a secondary mechanism, with the vertical distribution of downstream velocity the principal mechanism. In certain channel configurations, the two mechanisms may act to augment one another.

Form roughness and the absence of secondary flow in a large confluence–diffluence, Rio Paraná, Argentina

AbstractConfluence–diffluence units are key elements within many river networks, having a major impact upon the routing of flow and sediment, and hence upon channel change. Although much progress has been made in understanding river confluences, and increasing attention is being paid to bifurcations and the important role of bifurcation asymmetry, most studies have been conducted in laboratory flumes or within small rivers with width:depth (aspect) ratios less than 50. This paper presents results of a field‐based study that details the bed morphology and 3D flow structure within a very large confluence–diffluence in the Río Paraná, Argentina, with a width:depth ratio of approximately 200. Flow within the confluence–diffluence is dominated largely by the bed roughness, in the form of sand dunes; coherent, channel‐scale, secondary flow cells, that have been identified as important aspects of the flow field within smaller channels, and assumed to be present within large rivers, are generally absent in this reach. This finding has profound implications for flow mixing rates, sediment transport rates and pathways, and thus the interpretation of confluence–diffluence morphology and sedimentology. Copyright © 2006 John Wiley & Sons, Ltd.

Flow structure in sinuous submarine channels: Velocity and turbulence structure of an experimental submarine channel

Submarine channels have long been considered analogous to meandering fluvial channels due to their similar planform geometry, and this has given rise to strong analogies in terms of the fluid dynamics of these channel types. However, there is a paucity of direct measurements detailing the structure of velocity or sediment concentration, and nothing is known about the structure of secondary flow, within submarine channels. Within fluvial channels, secondary circulation has been shown to be an important control of erosion, deposition, lateral sediment transport and overall bend stability. This paper presents a series of experiments in which saline gravity currents flowed through a submerged, fixed-form, sinuous channel model. Ultrasonic Doppler velocity profiling provided high-resolution, three-dimensional visualisation of flow fields, thereby allowing the first quantitative visualisation of velocity profiles and secondary flows of a saline gravity current within a sinuous submarine channel. The same model was also used to measure the structure of flow within a subaerial channel, thereby providing a detailed comparison between a submarine and fluvial channel. The data reveal that secondary flow cells in submarine channels are best developed at bend apexes and that the basal component of the flow cell moves from the inside to the outside of the bend, in a reversed direction to that expected from models of fluvial bend flow. These results are of prime importance when assessing the validity of comparisons between submarine and fluvial channels, and the role of secondary flow cells in the migration and evolution of meander bends. This work demonstrates that the fundamental flow processes controlling the evolution and migration of submarine and fluvial channels may be very different, and suggest that the form and evolution of submarine channels may also show different properties to their fluvial counterparts.

Local Heat/Mass Transfer Phenomena in Rotating Passage, Part 2: Angled Ribbed Passage

Mass transfer experiments and numerical computations are conducted to investigate the heat/mass transfer and flow characteristics in a rotating passage with a 180-deg turn. In the present study, rectangular 70-deg rib turbulators are attached in cross and parallel arrangements on the leading and trailing surfaces of the passage. The cross-sectional area of the passage is 20 × 40 mm 2 , and the pitch-to-rib height ratio (p/e) is 7.5. The rotation number ranges from 0.0 to 0.20, whereas the Reynolds number is constant at 10,000. The results reveal that the angled rib turbulators enhance significantly heat/mass transfer in the rotating two-pass duct by generating a secondary flow. For the cross-ribbed passage, a heat/mass transfer discrepancy is observed in the second pass even for the stationary case because one asymmetric cell of secondary flow is induced in the first pass and turning region. For the parallel-ribbed passage, the flow and heat/mass transfer characteristics are less subject to the rotation number than those of the former case due to the different flow structure. At the highest tested rotation number (Ro = 0.20), the turn-induced single vortex cell becomes identical regardless of the rib configuration so that the local heat/mass transfer distributions are similar in the turning region for all of the test sections.

Theoretical Analysis on the Secondary Flow in a Rotating Helical Pipe With an Elliptical Cross Section

In the present study, the flow in a rotating helical pipe with an elliptical cross section is considered. The axes of the elliptical cross section are in arbitrary directions. Using the perturbation method, the Navier-Stokes equations in a rotating helical coordinate system are solved. The combined effects of rotation, torsion, and geometry on the characteristics of secondary flow and fluid particle trajectory are discussed. Some new and interesting conclusions are obtained, such as how the number of secondary flow cells and the secondary flow intensity depends on the ratio of the Coroilis force to the centrifugal force. The results show that the increase of torsion has the tendency to transfer the structure of secondary flow into a saddle flow, and that the incline angle α increases or decreases the secondary flow intensity depending on the resultant force between the Corilois force and centrifugal force.

Experimental study on flow and local heat/mass transfer characteristics inside corrugated duct

The present study investigates the flow and heat/mass transfer characteristics of wavy duct for the primary surface heat exchanger application. Local heat/mass transfer coefficients on the corrugated duct side walls are measured using a naphthalene sublimation technique. The flow visualization technique and a numerical analysis using a commercial code, FLUENT, are used to understand the overall flow structures inside the duct. The corrugation angle of the wavy duct is 145° and the duct aspect ratio is 7.3. The Reynolds numbers, based on the duct hydraulic diameter, vary from 100 to 5000. The results show that complex secondary flows and transfer processes exist inside the wavy duct resulting in non-uniform distributions of the heat/mass transfer coefficients on the duct side walls. At low Reynolds numbers ( Re ⩽ 1000), relatively high heat/mass transfer regions like cell shapes appear on both pressure- and suction-side walls due to the secondary vortex flows called Taylor–Görtler vortices perpendicular to the main flow direction. However, at high Reynolds numbers ( Re > 1000), these secondary flow cells disappear and boundary layer type flow develop on pressure-side wall. On the suction-side wall, high heat/mass transfer region appears by the flow reattachment. The average heat/mass transfer coefficients are higher than those of the smooth circular duct and pressure drop increases due to the secondary flows inside wavy duct.

Large Eddy Simulation of Flow and Heat Transfer in a 90 deg Ribbed Duct With Rotation: Effect of Coriolis and Centrifugal Buoyancy Forces

Results from large eddy simulations (LES) of fully developed flow in a 90 deg ribbed duct are presented with rib pitch-to-height ratio P/e=10 and a rib height-to-hydraulic-diameter ratio e/Dh=0.1. Three rotation numbers Ro=0.18, 0.36, and 0.68 are studied at a nominal Reynolds number based on bulk velocity of 20 000. Centrifugal buoyancy effects are included at two Richardson numbers of Ri=12, 28 (Buoyancy parameter, Bo=0.12 and 0.30) for each rotation case. Heat transfer augmentation on the trailing side of the duct due to the action of Coriolis forces alone asymptotes to a value of 3.7±5% by Ro=0.2. On the other hand, augmentation ratios on the leading surface keep decreasing with an increase in rotation number with values ranging from 1.7 at Ro=0.18 to 1.2 at Ro=0.67. Secondary flow cells augment the heat transfer coefficient on the smooth walls by 20% to 30% over a stationary duct. Centrifugal buoyancy further strengthens the secondary flow cells in the duct cross-section which leads to an additional increase of 10% to 15%. Buoyancy also accentuates the augmentation of turbulence near the trailing wall of the duct and increases the heat transfer augmentation ratio 10% to 20% over the action of Coriolis forces alone. However, it does not have any significant effect at the leading side of the duct. The overall effect of buoyancy on heat transfer augmentation for the ribbed duct is found to be less than 10% over the effect of Coriolis forces alone. Friction on the other hand is augmented 15% to 20% at the highest buoyancy number studied. Comparison with available experiments in the literature show excellent agreement.

EFFECTS OF PERMEABLE SPUR DIKES ON FLOW AND RIVER BED EVOLUTION IN CURVED OPEN CHANNELS

The effects of several kinds of permeable spur dike on 3-D flow structures and river bed evolution were investigated in curved open channels experimentally and numerically. In this study, three-dimensional mean flow structures were measured in a curved open channel with spur dikes arranged along the outer bank. Experiments were conducted in movable beds and the effects of spur dikes on the local scour were examined. The secondary flow cell was generated in the same manner as the uniform rod roughness cases. Outer-bank arrangements of permeable spur dikes can reduce the outer-bank scour, but additional scour is generated in different regions. A traditional hydraulic structure, “Seigyu” has the similar effects on flow structures as the pile dikes. A 2-D numerical simulation model can be applied to these spur dike cases with the same assumption for the secondary flow dispersion terms.

Rapid vaporization of subcooled liquid in a capillary structure

A novel process is introduced for rapid vaporization of subcooled liquid in a capillary structure. The process consists of a low-thermal-conductivity porous wick, heated from a downward-facing grooved heating block that is in intimate contact with the upper surface of the wick structure. For such a specially configured heat transfer device, measurements show that vapor can be generated rather quickly once a sufficient amount of heat was applied. The mechanisms leading to the rapid vaporization of liquid are numerically investigated. It is found that the low thermal conductivity of the capillary structure and the presence of the extremely steep temperature gradients at the fin/porous structure interface due to the rather weak natural convection, reflected by small-scale secondary flow cells below the heated fins, are responsible for the rapid vaporization of subcooled liquid.

EXPERIMENTAL STUDY ON CONTROL OF BED EVOLUTION IN BENS BY USING CYLINDRICAL ROUGHNESS

Vegetation or pile dikes in a river may affect 3-D flow structures and can control the local scour and deposit in curved open channels. In this study, three-dimensional mean flow structures were measured in a curved open channel with various arrangements of cylindrical roughness elements. Furthermore, experiments were conducted in a movable bed and the effects of roughness on the local scour were examined. Some outer-bank arrangements of roughness can reduce the bed evolution. The secondary flow cell is generated over whole section through the roughness area and the spanwise deference of primary mean velocity becomes small in the case of submerged roughness. The streamwise variation of bed share stress indicatcs charactcristic behaviors affected by the arrangement of the roughness area, and it gives an explanation for the tendency of bed evolution.

Quasi-static and chaotic characteristics of the natural convection field in a vertical slot

In this paper, the time-dependent characteristics of natural convection in a vertical slot with hot and cold side walls were experimentally examined. In the experiment, the test liquid was a glycerine–water mixture (80 wt%) and the Prandtl number in the center of the slot was held constant at 295. On the other hand, the Grashof number was varied in the range between 370 and 3.06 × 103. As a result, it was revealed that low-pass filtering for the time evolution data of temperature made it possible to compute the largest Lyapunov exponents precisely. Secondary flow cells are observed in both chaotic and nonchaotic regions. Moreover, the secondary flow cells appeared in regions where the power spectrum of temperature fluctuation was regarded as significant and where the autocorrelation function decreased rapidly with time. © 2000 Scripta Technica, Heat Trans Asian Res, 30(1): 40–53, 2001

Impact of Secondary Flow on Bed Form and Sediment Transport in a Meandering Channel for Overbank Flow

An investigation of interaction between bed form and secondary flow in a compound meandering channel with straight floodplain banks for overbank was carried out. The mobile bed form in the main channel was measured by photogrammetry and the secondary flow was first visualized with a submergible video camera and then measured using three component laser Doppler anemometer. The results of the bed form and secondary flow during flood indicate that there exists a certain relationship between the bed form and secondary flow. The secondary flow induced by the floodplain flow in the main channel generates another secondary flow cell owing to the bed form change. Those secondary flow cells make sand bars crossing the main channel, particularly, in the cross-over region. Sediment transport rate reduces at an early stage of flooding owing to strong interaction between the floodplain flow and the main channel flow.