Flow field investigation in a rectangular shallow reservoir using UVP, LSPIV and numerical modelling

Low velocity and shallow-depth flow fields often are a challenge to most velocity measuring instruments. In the framework of a research project on reservoir sedimentation, the influence of the reservoir geometry on sediment transport and deposition was studied. An inexpensive and accurate technique for Large-Scale Particle Image Velocimetry (LSPIV) was developed to measure the surface velocity field in 2D. An Ultrasonic Doppler Velocity Profiler (UVP) and LSPIV techniques were used for verification and validation of the numerical simulations. The velocities measured by means of UVP allowed an instantaneous measurement of the 1D velocity profile over the whole flow depth. The turbulence large-scale structures and jet expansion in the basin have been determined based on UVP, LSPIV and numerical simulations. Vertical velocity distributions were defined to study the vertical velocity effect. UVP measurements confirm 2D flow map in shallow reservoir. LSPIV has potential to measure low velocities. The comparison between LSPIV, UVP and numerical simulation gives good agreements.     

Application of LSPIV to measure supercritical flow in steep channels with low relative submergence

Measuring flow discharge has always been one of the most important concerns of water experts. To measure discharge in streams using velocity-area method it is necessary to quantify average velocity of the flow. It is not feasible to measure velocity by contact approaches like current meters under certain conditions such as in flood periods or for very shallow flows. Flow surface image velocimetry methods as non-intrusive solutions have recently been widely utilized to measure discharge in open channels. One of these methods is a variety of PIV method named LSPIV which has been very popular due to the elimination of laser application. In this study, LSPIV was used to measure 2D velocity field over the surface of steep supercritical flow. The obtained surface velocity data were used to calculate Velocity Index (VI) which is multiplied by surface velocity to convert it to mean velocity and subsequently flow discharge. Also, a few relations were proposed to calculate the VI according to the slope and relative submergence. Since, Velocity Index has been so far mostly studied for subcritical conditions, results of this study may be applied for measuring supercritical flows. Eventually, the proposed method was verified to be used for discharge measurement and was proven quite precise in this regard.     

河流水面成像测速研究进展

大尺度粒子图像测速(LSPIV)是一种用于测量大面积(数百平方米)水体表面流速的非接触式全场流速测量技术。不仅可用于常规条件下明渠紊动特性和时均特性的研究,更具有极端条件下河道水流监测的应用潜力。然而受复杂现场环境的影响,该技术在天然河流的应用中面临着诸多挑战。从水流示踪方式、流场图像采集、水面目标增强、运动矢量估计、时均流场重建、水面流场定标及不确定度评估七个方面综述了国内外对河流水面成像测速(RSIV)方法的研究进展。通过对现有方法的总结和讨论,提出改进技术方法的研究需求,以使其成为水动力学和水文学中有力的测量工具。     

Image-based river discharge estimation by merging heterogeneous data with information entropy theory

An information entropy based approach for the discharge measurements is evaluated for the gaging of the Isère river at the Grenoble university campus. Over a four month period, six discharge measurements were made using a vessel-mounted aDcp. Simultaneously, particle tracking velocimetry (PTV) from video images was used to estimate surface velocities. The surface velocities are projected along the regularly surveyed river section of the Isère-Campus gaging station. The vertical velocity profile at each stream-wise location is approximated by a 1D entropy profile. Information entropy 1D velocity vertical profile depends on two parameters which are fitted using aDcp and surface velocity measurements. The inclusion of the surface velocities reduces the dispersion of the estimated entropy parameters. The measurements show that the two parameters are linearly related with a slope that is stage dependent and thus, surface velocity dependent. From there, the information entropy theory for 1D velocity distribution offers a protocol by which surface velocities only are used to compute the discharges. The protocol is calibrated with both aDcp and surface velocity measurements. It is finally validated with several events during which only surface velocities are measured. For the high water flood event the estimated discharge falls within 2% of the one estimated with the rating curve of the gaging station.     

Application of shallow-water acoustic tomography to measure flow direction and river discharge

High-frequency Fluvial Acoustic Tomography System (FATS) was initially used to measure flow velocity and river discharge in a mountainous river. The results showed the high-frequency FATS, not only improves the velocity resolution, but also reduces the minimum operational range from 76 m to 43 m in compare with the previous type of FATS. The analysis of sound wave propagation (Ray tracing) showed the bottom topography can be the reason of multi-ray paths of sound wave in the shallow freshwater rivers. A new formula based on the continuity equation introduced to estimate the variations of the angle between the flow direction and the FATS transmission line. The flow direction was measured using two crossed FATS transmission lines of 53 kHz and 30 kHz. The results compared to the up-looking ADCP (Acoustic Doppler Current Profiler) deployed near the intersection of the two lines which measured the changes in flow direction. The results affirmed the efficiency of the proposed method. Finally, the river discharge was estimated by both FAT systems and compared to the Rating Curve method and moving-boat ADCP estimates. The relative error of the FATS discharge measurements was less than 10%.     

River velocity measurements using optical flow algorithm and unoccupied aerial vehicles: A case study

Accurate and reliable measurements of river flow are critical for a multitude of hydrologic engineering applications. However, flow rate measurements using in-situ sensors are uncertain in many applications and physical measurements of velocity may not be practical due to inaccessible sites or flood conditions. Recent advances in remote sensing using unoccupied aerial vehicles have overcome these limitations through non-contact measurements of river velocities; however, existing approaches have several shortcomings, including the need for artificial tracers in the absence of debris and prior knowledge of tracer size, shape, and flow direction. This case study seeks to overcome these shortcomings through the development of a system that utilizes drones, video imaging, and state-of-the-art optical flow algorithms to measure velocity in rivers. This system was applied along Menomonee River in Wauwatosa, WI. To remotely sense river flow, a DJI Matrice 210 RTK drone equipped with a Zenmuse X5S camera was used to capture video. The video data from the drone was analyzed using optical flow algorithms to generate velocity estimations. River velocity was measured directly at point locations using a hand-held velocimeter. Results indicate that the optical flow algorithms estimate the magnitude of surface velocity to within 13–27% of hand-held measurements without the use of artificial seeding. These outcomes suggest that this system could be used as a possible method to measure velocities in rivers.     

Measuring river velocity and discharge with acoustic Doppler profilers

Acoustic Doppler profilers and associated software packages presently are being used to measure water velocity, channel bathymetry, and river discharge. The instruments have various configurations and frequencies; choice of the appropriate instrument depends on various factors including depth, width, and sediment load of the rivers being measured. The acoustic Doppler profilers are mounted on powerboats or small remote-controlled or tethered rafts or catamarans. Profilers enable users to make fast, accurate, and economical discharge measurements on large rivers and rivers with unsteady flow conditions because of flooding or irregular releases from reservoirs. This article describes the principles of operation, application of acoustic Doppler profilers to the measurement of velocity and discharge, and calibration and verification issues.     

Application of spherical-rod float image velocimetry for evaluating high flow rate in mountain rivers

Spherical-rod float image velocimetry (SFIV) is a convenient technique combining the positive functions of a rod float velocimetry (RFV) and large-scale particle image velocimetry (LSPIV) for measuring high flow rate in mountain rivers. The SFIV is the principle that the sphere allowing little image distortion according to the orientation is used as a floating tracer for LSPIV. The drifting distances of a spherical-rod float were calculated by geometrical interpretation of spherical images recorded in an experimental open channel and mountain rivers. The depth-reflecting velocities estimated by SFIV in the rivers as in the open channel coincided approximately with the velocities by visual observation from river bank despite of the long shooting distance, weather impact, and flow complicated by topography and bed materials. The velocity coefficients obtained from the experimental channel were used to evaluate depth-averaged velocity for river discharges. The high discharges estimated by SFIV in mountain rivers distributed mostly within the range of the rating curve established by RFV. The results show that the safe and efficient SFIV is a highly applicable technique in mountain rivers with the high flow rate and complex flow. In order to practically use SFIV in mountain rivers, additional studies are required for velocity coefficients depending on the water depth and draft.     

Application of video imagery techniques for low cost measurement of water surface velocity in open channels

Developments in digital video recording technology make the video imagery tools more popular for velocity measurement in water flows. This has especially been of large interest due to its inherent advantage of non-contact nature which is quite handy in extreme flow conditions. Particle Image Velocimetry (PIV), Particle Tracking Velocimetry (PTV) and Large Scale Particle Tracking Velocimetry (LSPTV) are applied to free surface channel flow for water surface velocity measurement. Experiments are conducted to measure either a single point velocity applying PTV or velocity profiles across the channel width applying PIV on the water surface in a rectang typical velocities of nearly 1 andular tilting flume for various flow conditions. Technical issues regarding tracer particle size and type, travel distance, lighting, recording speed, camera position, image distortion and state of flow are discussed. Measured data is compared to computational results obtained from a numerical model involving a non-linear turbulence model capable of predicting turbulence driven secondary flows. Confirmation of reasonable match between computational and experimental results whereby applying mutual collaboration of them for discharge measurement has been attested. In addition to discharge, boundary roughness has also been predicted as an outcome of the numerical solution.     

Flow resistance law under suspended sediment laden conditions

The uniform flow resistance equation, in the form due to Manning or Darcy-Weisbach, is widely applied to establish the stage-discharge relationship of a river cross-section. The application of this equation, namely the slope-area method, allows to indirectly measure the corresponding river discharge by measurements of bed slope, water level, cross-section area, wetted perimeter and an estimate of channel roughness. In this paper, a recently deduced flow resistance equation for open channel flow was tested during conditions of suspended sediment-laden flow. First, the flow resistance equation was determined by dimensional analysis and by applying the condition of incomplete self-similarity for the flow velocity profile. Then the analysis was developed by the following steps: (i) for sediment-laden flows characterized by known values of mean diameter and concentration of suspended sediments, a relationship (Eq. (28)) between the Γ function of the velocity profile, the channel slope and the Froude number was calibrated by the available measurements; and (ii) a relationship for estimating the Γ function (Eq. (29)) which also takes into account the mean concentration of suspended particles was also established. The theoretical flow resistance law (Eq. (26)) coupled with the relationship for estimating the Γ function (Eq. (28) or Eq. (29)), which is characterized by the applicability of a wide range of flow conditions, allowed to estimate the Darcy-Weisbach friction factor for flows with suspended-load. The analysis showed that for large-size mixtures the Darcy-Weisbach friction factor can be accurately estimated neglecting the effect of mean concentration of suspended sediments while for small-size mixtures the friction factor decreases when the mean sediment concentration increases.     

Prediction of stage–discharge curves in open-channels using a fixed-point velocity measurement

The single-point measurement method for discharge estimation, which was first introduced by Maghrebi, can be implemented to obtain the discharge at different stages of a river during a flood event. As an advantage of this method, discharge can be estimated automatically with a fixed measurement location in the river section or on the water surface, which is associated with minimum energy and cost consumptions. For the proposed model, we determine the isovel contours in a normalized form for the cross section of the flow. To do so, we need to apply the field or experimental data, concerning the cross sectional geometry at different stages and its roughness variation along the wetted perimeter to the model. Then we collect the data of the single fixed-point of velocity measurement at the flow section using a velocity current meter. To validate the method, it is applied to a flume with different cases of roughened walls. The obtained results of stage–discharge curves using the single point of measurement in comparison to the observed experimental ones show that this method can quickly and accurately estimate the flood discharges. The maximum deviation between the observed and calculated discharges for most of observations is less than 5%.     

基于径向基神经网络的明渠流量软测量方法

为克服现有明渠流量测量方法在监测自动化、测量准确度、测量成本和适用范围等方面存在的不足,在新兴的大尺度粒子图像测速(large-scale particle image velocimetry,LSPIV)技术的框架下,设计了一种基于径向基神经网络(radial basis function neural networ...     

A dip-corrected discharge estimation method for a river system

This work proposes a novel dip corrected velocity distribution model in combination with the entropy theory for discharge estimation in a braided river. A modified form of the dip correction factor is derived by considering the topographical complexities and applied for assessing velocity profiles in river cross-sections. The velocity profiles at different verticals are computed by employing Shanon's entropy theory. The depth-averaged velocities at different cross-sections are estimated from the computed vertical velocity profiles and substituted in the area-velocity method for the discharge calculation. The model is applied to two study areas, Majuli and Umananda of the Brahmaputra River, having both simple and braided sections. The validation of the model is performed using the observed discharge data available at the nearby gauge site for low flow condition. Results indicate that the integration of bed rugosity factor in dip corrected velocity distributions improves the accuracy of discharge estimations.     

Turbulent structures,integral length scale and turbulent kinetic energy (TKE) dissipation rate in compound channel flow

In the present study, high data rate measurements were obtained for the streamwise and vertical velocity components using 2D Laser Doppler Velocimeter. The turbulent field in a straight compound-channel flow was characterized for three different uniform flow water depths, corresponding to “deep flows”, “intermediate flows” and “shallow flows” conditions. Several methodologies were studied to process the data and to obtain autocorrelation functions, integral length scale and turbulence kinetic energy (TKE) dissipation rate. The Sample and Hold method was adopted to interpolate the unevenly spaced record and calculate the autocorrelation function; the integral-stop-value 1/e was used to estimate the integral length scale; and the TKE dissipation rate was estimated through the velocity energy spectrum. A double shear layer composed of two counter-rotating vertical oriented vortices, interacting with the secondary currents, is observed in the interface region for deep flow conditions. By decreasing the water depth, the interface region becomes dominated by a strong mixing layer of vertical oriented vortices with high TKE dissipation rate and large integral length scale, acting as a vertical wall to the weak secondary currents that develop at the main channel. The determination of the integral length scale permits to confirm the existence and the strength of these turbulence structures, unveiling the strong mixing layer as the origin of the largest integral length scales, even larger than the flow depth, and as the most efficient mechanism to redistribute turbulence generated at the bottom towards upper flow regions. Despite the high complexity of turbulence structures present in the flow, for all water depths, a linear dependence is depicted between integral length scale, TKE dissipation rate, and streamwise turbulence intensity.     

Determining the discharge rate from a submerged oil leak jet using ROV video

With expanded deep sea drilling in the Gulf of Mexico, and possibly the Arctic, it is imperative to have a technology available to quickly and accurately measure the discharge rate from a submerged oil leak jet. This paper describes an approach to measure the discharge rate using video from a Remotely Operated Vehicle (ROV). ROV video can be used to measure the velocity of visible features (turbulent eddies, vortices, entrained particles) on the boundary of an oil leak jet, from which the discharge rate can be estimated. This approach was first developed by the Flow Rate Technical Group (FRTG) Plume Team, of which the authors Savaş and Shaffer were members, during the response to the Deepwater Horizon (DWH) oil leak. Manual tracking of visible features produced the first accurate government estimates of the oil discharge rate from the DWH. However, for this approach to be practical as a routine response tool, software is required that automatically measures the velocity of visible features. To further develop this approach, experiments were conducted to simulate a submerged oil leak jet using a dye-colored water jet in the U.C. Berkeley Tow Tank facility. Jet exit diameters were 10.2 cm and 20.3 cm. With flow rates up to 11 gal/s, Reynolds numbers in the range of the DWH oil leak jets (up to 500,000) were achieved. The dye-colored water jets were recorded with high speed video and radial profiles of velocity were mapped with Laser Doppler Anemometry (LDA). Particle Image Velocimetry (PIV) software was applied to measure the velocity of visible features. The velocities measured with PIV software were in good agreement with the LDA measurements. Finally, the PIV software was applied to ROV video of the DWH oil leak jet. The measured velocities were 10–50% lower than manual measurements of velocity. More research is required to determine the reasons why PIV software produced much lower velocities than manual tracking for the DWH oil leak jet.     

On sampling bias in multiphase flows: Particle image velocimetry in bubbly flows

Measuring the liquid velocity and turbulence parameters in multiphase flows is a challenging task. In general, measurements based on optical methods are hindered by the presence of the gas phase. In the present work, it is shown that this leads to a sampling bias. Here, particle image velocimetry (PIV) is used to measure the liquid velocity and turbulence in a bubble column for different gas volume flow rates. As a result, passing bubbles lead to a significant sampling bias, which is evaluated by the mean liquid velocity and Reynolds stress tensor components. To overcome the sampling bias a window averaging procedure that waits a time depending on the locally distributed velocity information (hold processor) is derived. The procedure is demonstrated for an analytical test function. The PIV results obtained with the hold processor are reasonable for all values. By using the new procedure, reliable liquid velocity measurements in bubbly flows, which are vitally needed for CFD validation and modeling, are possible. In addition, the findings are general and can be applied to other flow situations and measuring techniques.     

A simple model for estimation of dimensionless isovel contours in open channels

A simple model is proposed for predicting the dimensionless isovel contours in straight ducts and open channels. It is assumed that each element of the boundary influences the velocity at an arbitrary point in the cross section. Then, the total effect of the boundary can be obtained using integration along the wetted perimeter. In this paper, power and logarithmic laws are used, while any velocity profile can be applied in the model. The model is applied to calculate the normalized isovel contours in rectangular channels. Then they are used, in combination with a single-point velocity measurement at a cross section of the uniform flow, to estimate the discharge. The kinetic energy and momentum correction factors, and the ratio of maximum to mean velocity, are also calculated from isovel patterns. Calibration and validation of the model are carried out by comparing the results obtained with measurements of the velocity in the main flow direction along the centerline of a rectangular flume as well as in the transverse direction. Each point of measurement can be used to estimate the discharge. Then, the estimated discharge is compared with the actual measured one. Depending on the position of the measurement, the deviation of the calculated and measured discharges will be altered. Model predictions are well correlated with experimental data for rectangular open channels.     

River surface target enhancement and background suppression for unseeded LSPIV

Large-Scale Particle Image Velocimetry (LSPIV) is an image-based technique for nonintrusive streamflow monitoring, where the visibility of flow tracers is one of the main limitations to its application in field conditions. Based on the target characteristics of flow tracers as well as the optical environment of river surface, the paper presents a target enhancement and background suppression method that innovatively combines near-infrared (NIR) imaging and spatial high-pass filtering (SHPF) to solve the above problem. An NIR smart camera was developed as the experimental instrument for image acquisition and preprocessing. Three sets of evaluations were performed at pixel-level, feature-level and vector-level. Results show that the NIR imaging not only enhances the contrast between targets and background, but also improves the peak signal-to-noise ratio (PSNR) of correlation plane in motion vector estimation. Moreover, the SHPF effectively suppresses the river background and strong noises, and consequently increases the percentage of correct vectors in the instantaneous flow field. Due to its strong operability, this method offers promising potential for the unseeded LSPIV.     

Portable unmanned surface vehicle that automatically measures flow velocity and direction in rivers

We developed an automatic measurement system for flow velocity and direction in natural rivers using an autonomously controlled unmanned surface vehicle (USV). Oncoming mainstream velocity was measured by the propulsion force required for the USV in order to preserve the position at a measurement point. To conduct such a field mission, the system runs by changing four characteristic control stages: 1) calculation of the tentative propulsion force, 2) navigation to the target point, 3) velocity measurement by staying at the target, and 4) detection of flow direction by flowing downstream. More than 20 indoor tests were conducted under several hydraulic conditions by varying streamwise velocity, and the calibration formula was obtained by interrelating the oncoming velocity magnitude with the propulsion force required to remain at the target. The attitude control was provided with side thrusters to improve the yaw stability of the USV in the oncoming current. The adjunctive work of the side thrusters was very effective. Field tests were conducted to examine the reliability and accuracy of the present automatic flow measurements in a river. Both local velocity and direction in the river flow were measured well by the USV. Error analysis was conducted by comparing with the existing velocimetry results, and the USV was found to possess a sufficient ability to meet practical performance for the flow measurement in a calm river with about 1 m/s velocity.     

Estimating the flow velocity and discharge of ADCP unmeasured area in tidal reach

Due to the ringing and side-lobe interference, acoustic Doppler current profiler (ADCP) is unable to accurately capture the complete velocity profile in open channels near the water surface and channel bottom, which are usually called unmeasured areas. At present, the flow velocities through the unmeasured areas are most commonly estimated using the power law with the power set to be the default value. However, since the flows are unsteady and nonuniform in tidal reaches, the velocity distribution model and corresponding parameters will vary with the bathymetric, tide period, etc. Therefore, the most common estimation with the power law may not be suitable in tidal reaches. In this paper, a simple determination method of the best model is proposed. Firstly, the parameters in three classical velocity distribution models, which are called power law, logarithmic law and parabolic law models, are solved by least squares based on the ADCP measured velocity cells. Then, the corresponding root-mean-square error (RMSE) of each model is used for the quantitative indicator that the model with the minimum RMSE is chosen as the best model. At last, the flow velocity and discharge of the unmeasured area are estimated by the best model. The experiments carried out in the tidal reach of Yangtze Estuary showed that vertical flow velocity distribution various with the bathymetry and tide period, and the best models averagely improved about 2.0% of the relative standard deviation (RSD) relative to the power law method in the discharge estimation, especially at some tide period the RSD of the best model was several times better than that of power law model. For Yangtze River with an annual average discharge of 3.0×10⁴ m³/s, the improvement should not be ignored. Therefore, it will be necessary to use the best model with minimum RMSE to estimate the flow velocity in tidal reach.