Field Investigation on Hydroabrasion in High-Speed Sediment-Laden Flows at Sediment Bypass Tunnels
(2020)
Wear due to sediment particles in fluid flows, also termed hydroabrasion’ or simply ‘abrasion’, is an omnipresent issue at hydraulic structures as well as in bedrock rivers. However, interactions between flow field, particle motion, channel topography, material properties and abrasion have rarely been investigated on a prototype scale, leaving many open questions as to their quantitative interrelations. Therefore, we investigated hydroabrasion in a multi‐year field study at two Swiss Sediment Bypass Tunnels (SBTs). Abrasion depths of various invert materials, hydraulics and sediment transport conditions were determined and used to compute the abrasion coefficients kv of different abrasion models for high‐strength concrete and granite. The results reveal that these models are useful to estimate spatially averaged abrasion rates. The kv‐value is about one order of magnitude higher for granite than for high‐strength concrete, hence, using material‐specific abrasion coefficients enhances the prediction accuracy. Three‐dimensional flow structures, i.e., secondary currents occurring both, in the straight and curved sections of the tunnels cause incision channels, while also longitudinally undulating abrasion patterns were observed. Furthermore, hydroabrasion concentrated along joints and protruding edges. The maximum abrasion depths were roughly twice the mean abrasion depths, irrespective of hydraulics, sediment transport
conditions and invert material.
Sediment transport in high-speed flows over a fixed bed. 2: Particle impacts and abrasion prediction
(2017)
Single bed load particle impacts were experimentally investigated in supercritical open channel flow over a fixed planar bed of low relative roughness height simulating high-gradient non-alluvial mountain streams as well as hydraulic structures. Particle impact characteristics (impact velocity, impact angle, Stokes number, restitution and dynamic friction coefficients) were determined for a wide range of hydraulic parameters and particle properties. Particle impact velocity scaled with the particle velocity, and the vertical particle impact velocity increased with excess transport stage. Particle impact and rebound angles were low and decreased with transport stage. Analysis of the particle impacts with the bed revealed almost no viscous damping effects with high normal restitution coefficients exceeding unity. The normal and resultant Stokes numbers were high and above critical thresholds for viscous damping. These results are attributed to the coherent turbulent structures near the wall region, i.e. bursting motion with ejection and sweep events responsible for turbulence generation and particle transport. The tangential restitution coefficients were slightly below unity and the dynamic friction coefficients were lower than for alluvial bed data, revealing that only a small amount of horizontal energy was transferred to the bed. The abrasion prediction model formed by Sklar and Dietrich in 2004 was revised based on the new equations on vertical impact velocity and hop length covering various bed configurations. The abrasion coefficient kv was found to be vary around kv ~ 105 for hard materials (tensile strength ft > 1 MPa), one order of magnitude lower than the value assumed so far for Sklar and Dietrich's model.
Particle dynamics are investigated experimentally in supercritical high-speed open channel flow over a fixed planar bed of low relative roughness height simulating flows in high-gradient non-alluvial mountain streams and hydraulic structures. Non-dimensional equations were developed for transport mode, particle velocity, hop length and hop height accounting for a wide range of literature data encompassing sub- and supercritical flow conditions as well as planar and alluvial bed configurations. Particles were dominantly transported in saltation and particle trajectories on planar beds were rather flat and long compared with alluvial bed data due to (1) increased lift forces by spinning motion, (2) strongly downward directed secondary currents, and (3) a planar flume bed where variation in particle reflection and damping effects were minor. The analysis of particle saltation trajectories revealed that the rising and falling limbs were almost symmetrical contradicting alluvial bed data. Furthermore, no or negligible effect of particle size and shape on particle dynamics were found. Implications of experimental findings for mechanistic saltation-abrasion models are briefly discussed.
An experimental investigation of supercritical uniform and gradually varied open channel flows is presented for a wide range of Froude numbers and flume width-to-flow depth aspect ratios. The instantaneous streamwise and vertical flow velocities were measured in a laboratory flume over the entire width using a two dimensional–laser Doppler anemometry (2D-LDA) system to determine turbulence intensities, and bed and Reynolds shear stresses. The mean velocity patterns show undulation across the flume, indicating the presence of counterrotating secondary current cells. These currents redistribute turbulence intensities and bed and Reynolds shear stresses across the flume. For aspect ratios ≤ 4−5, i.e., narrow open channel flow, the velocity-dip phenomenon is identified both in the streamwise velocity and the Reynolds shear stress distributions. For high aspect ratios, i.e., wide open channel flow, the strength of secondary currents diminish toward the flume center, resulting in a 2D flow farther away from the walls and no velocity-dip phenomenon. Froude number effects on the flow characteristics are less pronounced compared to the aspect ratio effects. At high Froude numbers, the results for narrow and wide open channel flows agree well with literature data. The log-law holds in the inner region across the entire flume width for all investigated Froude numbers and aspect ratios. The Reynolds shear stress distribution agrees well with the computed spanwise bed shear stress distribution. At the flume side walls, the bed shear stresses are 20–50 % higher than the mean values. These results are verified with an engineering example in which high sediment transport and corresponding deep abrasion patterns at the side walls were observed.
In order to prevent reservoir sedimentation, sediment bypass tunnels can be an efficient countermeasure restoring sediment continuity of impounded rivers. Although supercritical open channel flow conditions in these tunnelsprevent tun-nel blockage, in combination with the high bypassed sediment volume it may lead tosevere abrasion damages on inverts. Consequently,wear termed hydroabra-sionoccurs. Based on laboratoryexperiments, a theoretical model was devel-oped to predict abrasion rates and service life timeof sediment bypass tunnels. Insituexperiments are further conducted for model calibration to provide an abrasion prediction approach for field applications.Finally,recommendations concerning the hydraulic design of the tunnel as well as the structural design ofthe invertareprovided.
The derivation of an abrasion prediction model for concrete hydraulic structures valid in supercritical flows is presented herein. The state of the art saltation-abrasion model from Sklar and Dietrich (2004) is modified using the findings of a recent research pro-ject on the design and layout of sediment bypass tunnels. The model correlates the im-pacting parameters with the invert material properties by an abrasion coefficient kv. The value of this coefficient is verified by a similarity analysis to bedrock abrasion in river systems applying a correlation between the abrasion rate and the bed material strength. A sensitivity analysis reveals that the saltation-abrasion model is highly dependent on an adequate estimation of kv. However, as a first order estimate the proposed model en-ables the practical engineer to estimate abrasion at hydraulic structures prone to super-critical flows.
Supercritical sediment-laden open channel flows occur in many hydraulic structures including dam outlets, weirs, and bypass tunnels. Due to high flow velocities and sediment flux severe problems such as erosion and abrasion damages are expected in these structures (Jacobs et al., 2001). Sediment bypass tunnels (SBT), as an effective measure to decrease reservoir sedimentation by bypassing sediments during floods, are exceptionally prone to high abrasion causing significant annual maintenance cost (Sumi et al., 2004; Auel and Boes, 2011). The Laboratory of Hydraulics, Hydrology and Glaciology (VAW) of ETH Zurich conducted a laboratory study to counteract these negative effects (Auel, 2014). The main goals of the project were to analyze the fundamental physical processes in supercritical flows as present in SBTs by investigating the mean and turbulence flow characteristics (Auel et al., 2014a), particle motion (Auel et al., 2014b; 2015b), and abrasion development caused by transported sediment. Besides new insights into the three listed topics, paramount interest is given to their inter-relations and the development of an easily applicable abrasion prediction model (Auel et al., 2015a). This paper presents selected results on the second topic, i.e. the analysis of saltation trajectories of single coarse particles in supercritical flow.
Worldwide, a large number of reservoirs impounded by dams are rapidly filling up with sediments. As on a global level the loss of reservoir volume due to sedimentation increases faster than the creation of new storage volume, the sustainability of reservoirs may be questioned if no countermeasures are taken. This paper gives an overview of the amount and the processes of reservoir sedimentation and its impact on dams and reservoirs. Furthermore, sediment bypass tunnels as a countermeasure for small to medium sized reservoirs are discussed with their pros and cons. The issue of hydroabrasion is highlighted, and the main design features to be applied for sediment bypass tunnels are given.
Single glass sphere motion recordings were conducted in a transitional-rough bed open channel at steady and highly supercritical flow similar to hydraulic conditions in sediment bypass tunnels. A high speed camera with a maximum resolution of 2,560 × 2,160 pixels was used to record the movement of bedload particles with diameters of D = 5.3, 10.3 and 17.5 mm. An in-house developed Particle Tracking Velocimetry (PTV) program was used to determine the transport mode and velocities of each particle for a wide range of Froude numbers up to Fo = 6. The relative roughness defined as the ratio of the bed roughness height ks to the water depth h varied from ks/h = 0.02–0.03. Particles were observed to move in rolling and saltation modes depending on the Shields number. The particle velocity shows a linearly increasing relationship with both friction velocity and Froude number nearly independent on the particle diameter. A linear relationship was also found between rolling and saltating particle velocities indicating that particle velocity does not depend on the transport mode in the range of the investigated hydraulic conditions. Scaling of particle velocity with the wave celerity plotted as a function of the Froude number adequately merged external data sets with the present data. As a consequence, a linear fit for a large Froude number range was obtained.