Geyser eruptions in storm sewer systems pose significant risks to urban infrastructure and public safety, driven by rapid air-water interactions and pressure surges. This study introduces a retrofitting system designed to mitigate geyser eruptions by enhancing air-water separation and stabilizing flow dynamics. The experimental setup, combined with numerical simulations, replicates field-scale conditions to evaluate the system's performance under varying air and water flow scenarios. Results demonstrate that the retrofitting system effectively reduces geyser heights, transitioning violent eruptions into moderate or non-eruption conditions, even under high-pressure gradients. Without retrofitting, violent eruptions occur when the air-to-water pressure head gradient ( S-AD) exceeds 5%. With retrofitting, eruptions remain moderate even at S-AD>10% . The system achieves this by minimizing the formation of large air bubbles, controlling pressure fluctuations, and directing air to escape through designated pathways. Notably, the retrofitting system maintains its efficacy across varying conduit configurations and lengths, ensuring adaptability to diverse urban stormwater systems.
Stepped dropshafts are key structures in urban deep tunnel drainage systems, yet their limits in conveyance capacity remain inadequately quantified. This study identifies an unfavourable hydraulic phenomenon, termed flow impingement, as the governing factor defining their conveyance capacity. Based on theoretical analysis and systematic experiments on seven physical models with varying step geometries, the onset of flow impingement is formulated as a hydraulic constraint governed by a dimensionless flow depth ratio. Results reveal that the flow depth ratio is primarily influenced by geometric parameters, including the relative vertical drop per turn, relative chute width, and relative step height, while showing negligible dependence on the dimensionless discharge. A multivariate regression model was developed to predict the flow depth ratio and was incorporated into the constraint equation. Consequently, a model of practical criterion for estimating the conveyance capacity is established. The proposed model demonstrates good agreement with independent literature data, suggesting its potential applicability in engineering design.
Rill flow is a major driver of soil erosion and a threat to watershed stability. As concentrated runoff evolves into gullies and steep step-pool channels, understanding its hydrodynamics becomes essential, especially once erosion rates diminish. In these conditions, strong vertical velocity gradients invalidate shallow-water assumptions. This work develops a section-averaged, non-hydrostatic rill-flow model that incorporates bed-form-related flow resistance while excluding sediment transport. A second-order finite-volume scheme with semi-implicit time stepping, solved through a Newton-Raphson method, is implemented. The model is evaluated using analytical solutions and laboratory data, showing satisfactory performance. Sensitivity analyses highlight the importance of accurate rill-width characterization. Experimental observations reveal rill widening at pools and the presence of bank cavities linked to secondary recirculating cells; incorporating these features significantly improves model accuracy. Studying steady flows in step-pool systems represents an initial step toward a comprehensive non-hydrostatic erosive flow model.
This study examines the aeration and turbulence characteristics of hydraulic jumps over vegetated rough beds. Three vegetation types including rigid acrylic rods, artificial turf and multi-leaf flexible plants were tested for Froude numbers between 2.8 and 5.5, with controlled equivalent roughness heights. Phase-detection probe measurements showed that all vegetated beds increased pre-aeration and impingement air entrainment at the jump toe. Vegetation morphology affects air-water flow properties more than vegetation density, especially at low Froude numbers. Flexible vegetation induced distinct turbulence modulation via shear layer interaction, despite lower flow resistance compared to rigid or turf-type roughness. The streamwise decay of characteristic void fraction, bubble count rate and velocity values are noted to be influenced by both approach flow conditions and vegetation dynamic response.
Monitoring dam-breaching experiments is challenging because imaging techniques like photogrammetry cannot penetrate turbid, aerated water to track bed elevation changes. This study presents a measurement campaign of small-scale dike breaching experiments using distributed optical fibre sensors (DOFS) to complement photogrammetric measurements. The fibre was placed transversally across the dike below the crest to monitor breach dynamics with minimal flow disturbance. Results show that optical fibre effectively tracks the breach bottom location and deepening process. However, careful post-treatment is necessary since optical fibres are sensitive to temperature gradients from infiltration. Strain and strain rate data distinguish between infiltration effects and mechanical processes such as erosion and block failure. Distributed optical fibre proves to be a promising tool that simultaneously generates valuable data on both infiltration and erosion, offering a powerful complement to traditional non-invasive imaging methods for experimental dam breach monitoring.
This study numerically investigates bilateral intakes around a broad-crested weir in a 90 degrees river bend. The influence of interior intake angle (30 degrees, 45 degrees, 60 degrees, 90 degrees) and the effects of the sill and weir on discharge distribution is analysed. Results indicate that a 30 degrees angle provides the most uniform flow and balanced discharge (each intake receives approximately 0.31 and 0.32 of the total flow at 35 l s-1). A 90 degrees angle produces severe flow separation, reducing interior intake efficiency to no more than 0.04. The sill substantially reduces interior intake discharge, especially at 60 degrees, where it falls to less than one-fifth of the no-sill value. Conversely, the broad-crested weir increases diverted discharge by factors of 2.9-14.4, depending on angle. This study demonstrates feasible water diversion from the inner bend and offers guidance for optimizing similar hydraulic systems.
This study experimentally investigates landslide-induced waves in reservoirs, addressing gaps in previous research that primarily focused on standard block shapes and surface waves. Three configurations of cylindrical concrete masses were released into reservoir models at depths of 35, 45 and 55 cm, with and without a 1.5 m downstream obstacle. Results indicate that obstacle placement can reduce wave height by up to 19.57%. Among the scenarios, parallel arrangements of sliding masses produce the highest wave energy, reaching 13.34 kJ, which scales to 4.2 & times; 109 kJ at the prototype. Increasing reservoir depth from 35 to 55 cm led to a 75.45% rise in wave energy and a 10.62% increase in maximum wave height. This study introduces a new experimental framework integrating mass geometry, obstacle placement, and reservoir depth to better quantify wave dynamics and energy transfer in reservoir settings.Key highlights
We present two models for predicting void-fraction distributions in hydraulic jumps, addressing the limitations of the conventional two-layer formulation in representing continuous profiles. The first model introduces a two-state convolution to describe the smooth transition between the turbulent shear and roller regions, while the second applies a superposition framework to capture the overlapping contributions from both layers. Validation against experimental data shows that both models improve the representation of the transition region and provide a more realistic connection between the upper and lower aerated layers. The proposed formulations not only reproduce measured profiles more accurately but also provide a continuous and physically consistent description across the entire flow depth. These models offer a practical tool for hydraulic design and advance the understanding of vertical mass transport in highly aerated flows.
This experimental study investigates air-water flow properties in a rectangular free-surface jet flow for 4.8 < Fr-0 < 11.4, 0.4 & times; 105 < Re < 2.1 & times; 105, and 35 < We < 120. Using a dual-tip conductivity probe, measurements of air concentration, bubble count rate, and droplet chord lengths were taken from the cavity to the far-field regions. Results show that air concentration is largely independent of Reynolds number, whereas the bubble count rate exhibits significant scale dependence. The jet flow is divided into four regions from the cavity to the far field, within which the time-averaged air concentration and the maximum bubble count rate exhibit distinct trends. Two formulae to predict the jet thickness and time averaged air concentration within the cavity region are given. Droplet size distributions shift markedly toward finer sprays after jet impingement, influenced by flow turbulence.
As critical components of fish migration channels, baffles are pivotal in adapting to water-level fluctuations and accommodating diverse fish species. This study presented an innovative modular and polymorphic baffle system (MPBS) designed to address the adaptability limitations of conventional fishways. The hydraulic characteristics of 12 different baffle configurations were investigated. The results show that the symmetric configuration with three pairs of vertical modules reduced the area with velocity > 1.0 m s(-1) by about 33%. A comparable reduction was achieved by half-submerging the vertical modules, which can provide the optimal hydraulic performance in terms of velocity uniformity and turbulence stability. The one-sided configuration with two and three pairs of modules substantially increased the area with velocity > 0.5 m s(-1) by nearly 96%. When the ratio of weir height to head difference decreased to 0.25, the hydraulic conditions became unfavourable for migration. These findings illuminate the potential of MPBS to enhance fishway adaptability, offering a promising solution for multi-species fish passage under varying hydraulic conditions.
The transport characteristics of semi-buoyant fish eggs significantly influence their drift, dispersal, and recruitment in rivers. Due to physical differences between eggs and flows, we hypothesize their motion is asynchronous with flow dynamics. Spherical surrogate particles mimicking real fish eggs in hydraulic properties were employed to investigate the three-dimensional movements in decelerating flows. Results indicated that surrogate eggs' longitudinal velocity lagged approximately 22% behind flow velocities, while their vertical velocity was strongly flow-dependent. Eggs dispersed evenly laterally along the main flow axis. Furthermore, increased travel distance led to higher egg loss rates, attributed to gravitational settling and trapping effect. These outcomes provide valuable insights into evaluating the distribution patterns of semi-buoyant fish eggs during long-distance river drifting.
Large eddy simulations have been conducted to analyse turbulent flow in partially-filled pipes, with a focus on understanding the impact of secondary currents on pressure fluctuations and turbulence structures near the pipe wall. Time-averaged secondary currents manifest as pairs of symmetric vortices around the pipe symmetric bisector in partially-filled conditions. These secondary currents exhibit a streamwise meandering pattern with a wavelength ranging from one to two times the pipe radius, which influences near-wall turbulence. Two-point correlation analysis indicates that the wall pressure fluctuates primarily due to sweeps and ejections exerting regular spikes in wall pressure. The secondary currents diminish the dominant area of fluid pressure and expand the regulatory influence of ejections and sweeps. Positive wall pressure peaks are associated with low-speed very-large-scale motions in fully-filled pipe flows, but secondary currents disrupt this connection by inhibiting very-large-scale motions in partially-filled pipes. Consequently, pressure fluctuations along the pipe perimeter are non-uniform, with locations where secondary currents point toward the pipe wall experiencing elevated pressure fluctuations.
Liquid sloshing in low-aspect-ratio tanks under complex marine motions threatens structural integrity through impact pressures, yet remains insufficiently characterized for compact marine reactors. This study combines a high-performance 6-degree-of-freedom (DOF) motion platform with a maximum rolling angle of 45 degrees and a 10 t payload, integrating dynamic pressure measurements to simulate severe marine excitations with periods as short as 4 s. Results indicate that pressure fluctuation amplitudes correlate positively with excitation amplitude and inversely with period. Multi-DOF excitations amplify pressure fluctuations, evidenced by spectral dominance of intensified main-frequency components. Five suppression configurations were evaluated. Cylindrical baffles achieved optimal suppression, reducing pressure amplitudes by up to 52% compared to uncontrolled cases. Frequency-domain analysis confirms that suppression devices not only reduce the pressure amplitude of the main frequency component but also effectively inhibit the generation of its multiple components. The work establishes a design framework for anti-sloshing structures, directly applicable to marine reactor systems.Highlight Multi-DOF sloshing tested on a heavy-duty 6-DOF platform for marine simulations.Study enhances sloshing insights in low-aspect-ratio tanks.Five types of sloshing suppression device studied, diverse applications proposed.52% max sloshing suppression achieved in this study.
Understanding how vegetation influences flood attenuation is essential for effective risk reduction and eco-engineering design. This study investigates the capability of deep neural networks (DNNs) to reproduce time-resolved hydrodynamics in dam-break flows through vegetation and to evaluate their potential as efficient surrogate models. Three-dimensional simulations were conducted using OpenFOAM and validated against laboratory experiments to generate training and testing data. The DNNs accurately reproduced both nondimensional water surface displacement and horizontal velocity within the training domain, achieving mean R-2 values of about 0.80 for the entire time series and 0.93 for the maximum value. However, prediction accuracy decreased for vegetation lengths and flow conditions outside the training range, indicating limited extrapolation capability. Overall, the results demonstrate that DNNs can efficiently complement costly hydraulic experiments and numerical models, enabling rapid assessment of vegetation effects on flooding for data-driven design of nature-based protection strategies.
The digital transition is often described as a recent disruption driven by artificial intelligence and data-intensive technologies. This paper argues instead that, in water engineering, it represents the maturation of a long-standing scientific trajectory rooted in physically based modelling, hydroinformatics, complex network theory, and evolutionary optimization. Digital tools are therefore interpreted as extensions of engineering reasoning rather than substitutes for it. The central contribution is the formal introduction of the phenomenological twin, defined as the physically grounded mathematical core of a digital twin architecture. While the digital twin integrates data, models and digital services, the phenomenological twin ensures causal consistency, interpretability and reliable generalization in physically governed systems. The paper clarifies its distinction from conventional physics-based digital twins, grey-box models, and purely data-driven approaches. By anchoring digital twin development in conservation laws and engineering modelling, the paper provides an epistemological framework for integrating machine learning within robust decision-support architectures.
