Marian Muste

Large-scale particle image velocimetry for flow analysis in hydraulic engineering applications

Large-Scale Particle Image Velocimetry (LSPIV), as presented herein, is an extension of particle image velocimetry (PIV), which aims at providing velocity fields spanning large flow areas in laboratory or field conditions. Additional data, such as mappings of large-scale flow structures and discharges are readily obtainable using LSPIV. While the image- and data-processing algorithms are similar to conventional PIV, adjustments are required for illumination, seeding procedures, and pre-processing of the recorded images. This paper describes the implementation of video-based LSPIV in three hydraulic engineering applications covering surfaces from 4 m2 to 45,000 m2. These applications are: gas-transfer processes downstream a model spillway, ice conveyance through a model river confluence, and flood plain flow in a full-scale river. The special problems encountered in each of these experiments, as well as the selection and adjustments of the parameters to properly solve them, are examined. LSPIV has proven t...

River gauging using PIV techniques: a proof of concept experiment on the Iowa River

Abstract An image-based approach for discharge measurements is evaluated for river gauging in an experiment on the Iowa River at Iowa City, Iowa. Over a twenty-day period, ten discharge measurements were made using the image-based approach. A ten-minute video recording was made of the river flow for each measurement. Particle image velocimetry (PIV) was used to estimate surface velocities for the imaged area using naturally occurring foam as a flow tracer. The surface velocities were then estimated along a surveyed river section, and river discharge was computed using standard velocity–area methods over a selected cross-section. Several unique aspects of this experiment were the use of PIV for unseeded flow conditions, and the evaluation of discharge estimates over a range of flow conditions. A comparison of the PIV discharge measurements with traditional current meter measurements, which have been made at the site since 1984, showed that the PIV measurements were consistent with the observed stage–discharge relationship. Discharge for the experiment ranged from 50 to 300 m 3 s −1 , which covers a large portion of the existing rating curve. The experimental results suggest that image-based approach may be a reliable way of establishing a stage–discharge relationship at a site, perhaps even remotely, by making repeated measurement with a camera mounted at the site. Still, there are inherent limitations with the approach. These limitations include the need for recognizable tracer particles or flow patterns to detect motion, problems associated with shadows and reflections on the water surface, as well as the common the problem for all discharge measurement of the need for survey information on the channel cross-section.

Two-phase formulation of suspended sediment transport

Using a two-phase formulation, the vertical and horizontal momentum equations for sediment are used to obtain the concentration and velocity profiles of a dilute suspension of particles in a 2D uniform flow. Assuming the form of the vertical turbulent intensities and dilute concentrations of sediment, one can solve the equations analytically and compare them with experimental data. No empirical coefficients in the model are tuned to match individual experiments, for which the experimental data cover a large range of particle sizes and densities. The models are shown to accurately predict two experimentally observed but theoretically unexplained phenomena: the increased diffusive flux of large particles, and the measurable velocity lag of particles. The increased diffusion of large particles is shown to originate from the added diffusive nature of the sediments Reynolds stresses. The horizontal velocity lag of particles is due to an induced velocity, termed the drift velocity, resulting from the correlati...

Particle-image velocimetry for whole-field measurement of ice velocities

Particle-image velocimetry (PIV) is introduced herein as a useful experimental and field technique for determining and mapping whole fields of ice and water velocities. The technique, which is becoming extensively used in other applications of fluid mechanics, lends itself very well for this purpose. The primary requirements to use the technique are a suitable series of images of moving ice, or model ice in a hydraulic model, a computer equipped with a frame-grabber, and the pertinent software. The images may comprise direct video images, a series of photographs, or radar-generated records of moving ice. The present paper briefly describes the principles underlying PIV and shows how PIV can be adapted for use in hydraulic-model studies of ice-movement problems. The technique is demonstrated for a hydraulic model study of ice movement through a confluence of two rivers. The techniques utility suggests that it also could be a useful adjunct for remote sensing of ice movement on rivers, in coastal regions, or on oceans.

Sources of bias errors in flume experiments on suspended-sediment transport

Extensive research efforts in the last few decades have only partially elucidated the complexities of suspended-sediment transport. Lacking an adequate formulation and quantification of the interaction between suspended particles and the carrier liquid, it is common practice to combine sediment mechanics theory and empiricism to obtain predictive formulations. Flume data for suspended sediment transport, however, is incomplete and often inconsistent with respect to insights into sediment effects on water flow. Improvement of the data quality/reliability for future similar studies requires identification and evaluation of the sources of bias errors that might affect the experimental results. The present paper identifies and partially evaluates significant sources of bias errors in Hume investigations of suspended-sediment transport. Bias errors are discussed in the order in which they typically arise in the conduct of flume experiments. The paper considers first conceptual errors associated with the governing equations used for the design and interpretation of the experimental results. Considered next are bias errors generated during data acquisition and data reduction. As an outcome of bias error discussion, the paper recommends guidelines to avoid and reduce such errors, hence, to increase the reliability of the experimental results. Special emphasis is placed on illustrating the potential of the non-intrusive measurement instruments for removing bias errors.

End-to-End Cyberinfrastructure for Decision-Making Support in Watershed Management

AbstractOver the past decade, water-centric research has increasingly taken into consideration the interactions between the water cycle and the social, economic, and biogeophysical processes that drive watershed dynamics. In parallel, water management has made great strides in data sharing and collaborative modeling that support decision making through integrated planning and stakeholder involvement. Both research and management communities require data and simulation models that cover large spatial scales and workflows that enable investigations and decision making in real time with participation of multiple watershed actors. To efficiently accomplish their goals, these two communities are tapping into the capabilities of advanced cyberinfrastructure (CI) platforms that facilitate an understanding of watershed processes, knowledge management, visualization, interaction, and collaboration in multiple watershed science and engineering disciplines. This paper illustrates an implementation of an end-to-end C...

Standardized uncertainty analysis for hydrometry: a review of relevant approaches and implementation examples

Abstract The water-centric community has continuously made efforts to identify, assess and implement rigorous uncertainty analyses for routine hydrological measurements. This paper reviews some of the most relevant efforts and subsequently demonstrates that the Guide to the expression of uncertainty in measurement (GUM) is a good candidate for estimation of uncertainty intervals for hydrometry. The demonstration is made by implementing the GUM to typical hydrometric applications and comparing the analysis results with those obtained using the Monte Carlo method. The results show that hydrological measurements would benefit from the adoption of the GUM as the working standard, because of its soundness, the availability of software for practical implementation and potential for extending the GUM to hydrological/hydraulic numerical simulations. Editor D. Koutsoyiannis Citation Muste, M., Lee, K. and Bertrand-Krajewski, J.-L., 2012. Standardized uncertainty analysis for hydrometry: a review of relevant approaches and implementation examples. Hydrological Sciences Journal, 57 (4), 643–667.

Near-Transducer Errors in ADCP Measurements: Experimental Findings

Acoustic Doppler current profilers (ADCPs) are not able to accurately determine velocity near their transducers and near the bed. These limitations have restricted the use of ADCPs to flow depths that are large enough to allow acquisition of few directly measured velocity data that can be subsequently used to accurately estimate vertical velocity profiles and flow discharge in cross sections. While the causes that make ADCPs unable to collect data in the near-bed region are relatively well documented, the causes of near-transducer errors have not yet been fully understood and are only partly documented. We present results from an experimental study aimed at characterizing the systematic errors due to the combined effect of acoustic interference and instrument-induced flow disturbance near a Janus-configured ADCP. The study comprises: (1) concurrent measurements with an ADCP and an acoustic Doppler velocimeter (ADV) under the ADCP; (2) measurements of the flow disturbance produced by the ADCP in the vertical and horizontal planes; and (3) ADV measurements along the path of the acoustic beams ensonified by the ADCP during a measurement. Results suggest that ADCPs bias low the velocity profiles with respect to the undisturbed velocity profiles, mostly because of the flow disturbance induced by the ADCP, with acoustic effects playing a secondary role. For the range of flows we studied, both undisturbed and disturbed profiles exhibit similar shapes when plotted in dimensionless form, with the bulk flow velocity and the ADCP diameter (D) as characteristic scales. The differences between the undisturbed and the ADCP-disturbed profiles extend up to a distance of about 1.5D from the ADCP, except for the profiles measured at locations where the flow depth is close to D for which the boundary layer induced by the ADCP interacts with the one induced by the flume bed.

Flow and sediment transport measurements in a simulated ice-covered channel

Laboratory flume experiments conducted to illuminate flow field and sediment transport characteristics in ice-covered sand-bed channels are discussed. The overall purpose of the experiments was to examine how level ice covers affect flow distribution, flow turbulence characteristics, and sediment transport rates. The experiments were conducted with a nonrefrigerated flume fitted with a sand bed and plywood panels used to simulate cover-floating ice covers. A discriminator laser-doppler velocimeter was used to measure water and sediment particle velocities simultaneously. The addition of a cover and an increase in its roughness increases flow depth and decreases bulk flow velocity, thereby reducing sediment transport rates compared to those in open water flow. The Reynolds shear stress, turbulence intensities, and distribution of sediment size over the depth also change.

Acoustic mapping velocimetry

Knowledge of sediment dynamics in rivers is of great importance for various practical purposes. Despite its high relevance in riverine environment processes, the monitoring of sediment rates remains a major and challenging task for both suspended and bed load estimation. While the measurement of suspended load is currently an active area of testing with nonintrusive technologies (optical and acoustic), bed load measurement does not mark a similar progress. This paper describes an innovative combination of measurement techniques and analysis protocols that establishes the proof-of-concept for a promising technique, labeled herein Acoustic Mapping Velocimetry (AMV). The technique estimates bed load rates in rivers developing bed forms using a nonintrusive measurements approach. The raw information for AMV is collected with acoustic multibeam technology that in turn provides maps of the bathymetry over longitudinal swaths. As long as the acoustic maps can be acquired relatively quickly and the repetition rate for the mapping is commensurate with the movement of the bed forms, successive acoustic maps capture the progression of the bed form movement. Two-dimensional velocity maps associated with the bed form migration are obtained by implementing algorithms typically used in particle image velocimetry to acoustic maps converted in gray-level images. Furthermore, use of the obtained acoustic and velocity maps in conjunction with analytical formulations (e.g., Exner equation) enables estimation of multidirectional bed load rates over the whole imaged area. This paper presents a validation study of the AMV technique using a set of laboratory experiments.