Hydraulic Problems Research Articles

An efficient parallelization technique for the coupled problems of fluid, gas and plasma mechanics in the grid environment

The development of efficient parallelization strategies for numerical simulation methods of fluid, gas and plasma mechanics remains one of the key technology challenges in modern scientific computing. The numerical models of gas and plasma dynamics based on the Navier-Stokes and electrodynamics equations require enormous computational efforts. For such cases, the use of parallel and distributed computing proved to be effective. The Grid computing environment could provide virtually unlimited computational resources and data storage, convenient task launch and monitoring tools, graphical user interfaces such as web portals and visualization systems. However, the deployment of traditional CFD solvers in the Grid environment remains very limited because basically it requires the cluster computing architecture. This study explores the applicability of distributed computing and Grid technologies for solving the weak-coupled problems of fluid, gas and plasma mechanics, including techniques of flow separation control like using plasma actuators to influence boundary layer structure. The adaptation techniques for the algorithms of coupled computational fluid dynamics and electrodynamics problems for distributed computations on grid and cloud infrastructure are presented. A parallel solver suitable for the Grid infrastructure has been developed and the test calculations in the distributed computing environment are performed. The simulation results for partially ionized separated flow behind the circular cylinder are analysed. Discussion includes some performance metrics and parallelization effectiveness estimation. The potential of the Grid infrastructure to provide a powerful and flexible computing environment for fast and efficient solution of weak-coupled problems of fluid, gas and plasma mechanics has been shown.

A general form of starting solutions of linear viscoelastic fluids for Couette flow

Couette flow, namely the flow of a fluid caused by moving boundary, is one of the fundamental problems in non-Newtonian fluid mechanics. Based on Laplace transform and the orthogonal basis function expansion, we provide a general method to solve the velocity field of the generalized linear viscoelastic fluid in Couette flow. This method can be applicable to the generalized cross sections and the generalized linear viscoelastic fluids. Through the method, the solution of Couette flow for the rectangular pipe and the generalized Burgers fluid can be obtained, which were rarely discussed before. The analytical solution consists of a quasi-steady solution and an approaching part, which is expressed by the sum of approaching terms. The analytical solution is composed of oscillatory solutions of multiple spatial modes. For example, the multiple resonance frequencies are found in Couette flow between parallel plates. The n-th resonance frequency is close to the oscillation frequency of n-order spatial mode. Low order spatial modes generally have stronger resonance.

Application of physics-informed neural networks (PINNs) solution to coupled thermal and hydraulic processes in silty sands

The accurate modeling of water and heat transport in soils is crucial for both geo-environmental and geothermal engineering. Traditional modeling methods are problematic because they require well-defined boundaries and initial conditions. Recently, physics-informed neural networks (PINNs), which incorporate partial differential equations (PDEs) to solve forward and inverse problems, have attracted increasing attention in machine learning research. In this study, we applied PINNs to tackle hydraulic and thermal transport coupling forward problems in silty sands. A fully connected deep neural network was utilized for training. This neural network model leverages automatic differentiation to apply the governing equations as constraints, based on the mathematical approximations established by the neural network itself. We conducted forward problems and compared the solutions derived from PINNs with those from Finite Element Method (FEM) simulations. The forward problem results demonstrate the PINNs model’s capability in predicting hydraulic transport, heat transport, and thermal–hydraulic coupling in silty sands under various boundary conditions. The PINNs exhibited great performance in simulating the thermal–hydraulic coupling problem. The accuracy of the PINNs solutions shows its potential for simulation in geotechnical engineering.

A priori physical information to aid generalization capabilities of neural networks for hydraulic modeling

The application of Neural Networks to river hydraulics and flood mapping is fledgling, despite the field suffering from data scarcity, a challenge for machine learning techniques. Consequently, many purely data-driven Neural Networks have shown limited capabilities when tasked with predicting new scenarios. In this work, we propose introducing physical information into the training phase in the form of a regularization term. Whereas this idea is formally borrowed from Physics-Informed Neural Networks, the proposed methodology does not necessarily resort to PDEs, making it suitable for scenarios with significant epistemic uncertainties, such as river hydraulics. The method enriches the information content of the dataset and appears highly versatile. It shows improved predictive capabilities for a highly controllable, synthetic hydraulic problem, even when extrapolating beyond the boundaries of the training dataset and in data-scarce scenarios. Therefore, our study lays the groundwork for future employment on real datasets from complex applications.

Evaluating machine learning techniques for fluid mechanics: Comparative analysis of accuracy and computational efficiency

This study focuses on applying machine learning (ML) techniques to fluid mechanics problems. Various ML techniques were used to create a series of case studies, where their accuracy and computational costs were compared, and behavior patterns in different problem types were analyzed. The goal is to evaluate the effectiveness and efficiency of ML techniques in fluid mechanics and to contribute to the field by comparing them with traditional methods. Case studies were also conducted using Computational Fluid Dynamics (CFD), and the results were compared with those from ML techniques in terms of accuracy and computational cost. For Case 1, after optimizing relevant parameters, the Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM) models all achieved an R² value above 0.9. However, in Case 2, only the ANN method surpassed this threshold, likely due to the limited data available. In Case 3, all models except for Linear Regression (LR) demonstrated predictive abilities above the 0.9 threshold after parameter optimization. The LR method was found to have low applicability to fluid mechanics problems, while SVM and ANN methods proved to be particularly effective tools after grid search optimization.

Computational and Experimental Investigation Of Water Surface Around Bridge Piers

Free surface flow around bridge piers is one of the most important problems of open channel hydraulics. The structural design of the bridge should be done correctly along with the hydraulic design by simulating the water surface around the bridge piers together with the scour that will occur around the bridge piers. If the hydraulic design of the bridge piers is not good enough, it can cause very serious damages in situations such as floods and overflows. The aim of this paper is to analyze the effect of bridge piers on the water surface experimentally and numerically. For the numerical analysis, FLOW-3D, a computational fluid dynamics package program that performs 3-dimensional solution, and HEC-RAS package program are used and the results are compared. In the study, flows with the same velocity for different number of bridge openings were studied and compared with the experimental results. In addition, the FLOW 3D package program, which performs solutions with the finite volume method, was compared with the experimental results by performing solutions with different mesh sizes. As a result, it was observed that the solutions performed with the FLOW-3D package program showed very close results with the experimental data.

Iterative methods for the infiltration advance problem

We examine the in ltration advance problem in irrigation hydraulics modeled through the Lewis-Milne equation. We propose two iterative methods that give reasonable results compared with existing formulas for various well-known in ltration functions. The method of successive approximations seen as a xed-point method and the variational iteration method (based on Lagrange multipliers method) produce formulas after few iterations and simple calculations compared to existing methods where the Laplace transform of the in ltration function has to be known.

A shear and elongational decomposition approach of the rate-of-deformation tensor for non-Newtonian flows with mixed kinematics

We propose a novel approach for non-Newtonian viscoelastic steady flows based on a decomposition of the rate-of-deformation tensor which here, in a simplified version, leads to an anisotropic generalised Newtonian fluid-like model with separated treatment of kinematics pertaining to shear and extensional flows. Care is taken to assure that the approach is objective and does not introduce spurious effects due to dependence on a specific reference frame. This is done by separating the rate-of-deformation tensor into shear and elongational components, with the local scalar shear and elongation rates being the second invariant of each tensor separately, and by modifying the method used to separate the rate-of-deformation tensor so that it becomes independent of superimposed rigid rotations, thus satisfying the principle of material indifference. We assess the model with two test cases: planar contraction flow, often employed to evaluate numerical methods or constitutive equations and for which experimentally observed corner vortex enhancement and pressure drop increase are seldom found in numerical simulations; and flow past confined or unconfined cylinders, for which experiments indicate a drag increase due to elasticity and most predictions give a drag decrease. With our anisotropic model, incorporating additional elongational-flow-related terms, large vortices and accentuated pressure drop coefficients can be predicted for the contraction problem and enhanced drag coefficient for the cylinder problems. This is the first work where problems of non-Newtonian fluid mechanics are solved numerically with separation of strain rate into shear and elongational components.

Locally-Verifiable Sufficient Conditions for Exactness of the Hierarchical B-spline Discrete de Rham Complex in $$\mathbb {R}^n$$

AbstractGiven a domain $$\Omega \subset \mathbb {R}^n$$ Ω ⊂ R n , the de Rham complex of differential forms arises naturally in the study of problems in electromagnetism and fluid mechanics defined on $$\Omega $$ Ω , and its discretization helps build stable numerical methods for such problems. For constructing such stable methods, one critical requirement is ensuring that the discrete subcomplex is cohomologically equivalent to the continuous complex. When $$\Omega $$ Ω is a hypercube, we thus require that the discrete subcomplex be exact. Focusing on such $$\Omega $$ Ω , we theoretically analyze the discrete de Rham complex built from hierarchical B-spline differential forms, i.e., the discrete differential forms are smooth splines and support adaptive refinements—these properties are key to enabling accurate and efficient numerical simulations. We provide locally-verifiable sufficient conditions that ensure that the discrete spline complex is exact. Numerical tests are presented to support the theoretical results, and the examples discussed include complexes that satisfy our prescribed conditions as well as those that violate them.

Effect of spatial variability of soil properties on slopes stability under rapid water drawdown conditions

Stability analysis of slopes is a fundamental problem of Soil Mechanics. Its main objective consists of evaluating the potential failure of the slope structure under a prescribed loading mode. A major component of the latter refers to the seepage forces induced by pore-pressure gradient, which is known to be responsible of destabilizing effects. This contribution employs a kinematic approach of limit analysis theory to obtain upper-bounds solutions to the stability problem of saturated slopes submitted to rapid water level drawdown. Random fields are used to model the uncertainty surrounding the spatial distribution of soil cohesion, friction angle, and permeability. In the context of effective stress governing the strength capacities of the soil material, it is shown that the seepage forces related to the water flow regime can be considered as external volumetric loads in the assessment of stability. The hydraulic problem governing the water filtration velocity is evaluated by resorting to an analytical variational approach, whose results are validated by comparisons with finite element solutions. The impact of hydraulic-related parameters on stability is first investigated for slopes within a deterministic framework. Subsequently, random fields are considered to take into account the variability related to the spatial distribution of the material properties, which are discretized using the Karhunen-Loève expansion with numerically computed eigenfunctions. The failure probability of slopes is assessed through Monte Carlo simulations. Several analyses have been performed to discuss the influence of the spatial variability of relevant problem parameters.

Model-Free Closed-Loop Control of Flow Past a Bluff Body: Methods, Applications, and Emerging Trends

Flow past one or multiple bluff bodies is almost ubiquitous in nature and industrial applications, and its rich underlying physics has made it one of the most typical problems in fluid mechanics and related disciplines. The search for ways to control such problems has attracted extensive attention from both the scientific and engineering fields, as this could potentially bring about benefits such as reduced drag, mitigated noise, suppressed vibration, and enhanced heat transfer. Flow control can be generally categorized into passive and active approaches, depending on whether there is an external energy input to the flow system. Active control is further divided into open-loop approaches and closed-loop approaches, depending on whether the controller depends on feedback signals extracted from the flow system. Unlike in many other applications of passive flow control and open-loop active flow control, theoretically advantageous closed-loop controls are quite rare in this area, due to the complicated features of flow systems. In this article, we review the recent progress in and future perspectives of flow past a single or multiple bluff bodies using model-free closed-loop control so as to outline the state-of-the-art research, determine the physical rationale, and point to some future research directions in this field.

Exploring different representations of hydraulic tomographic data for deep learning: Sequence or image

Exploring different representations of hydraulic tomographic data for deep learning: Sequence or image

Hydrodynamic simulation of coastal reservoir in response to flow control: a numerical study of the Saemangeum reservoir

Coastal reservoirs (CRs) are important water resources worldwide, but environmental problems can arise due to limited material mixing caused by the blocked propagation of tidal waves. The Saemangeum Reservoir is a large brackish reservoir in South Korea; owing to its complex fluctuation characteristics, accurate reproduction of its three-dimensional (3D) circulation and material distribution characteristics remains a challenge. Here, we aimed to improve the reproducibility of the Delft3D model through long-term precision monitoring and analyze problems occurring in CRs. The model reproduced the hydraulic flow and stratification phenomena in CRs. The Richardson number ( R i ) was higher than 100, and the Stratification Index ( SI ) was higher than 1 due to the density difference between the surface and bottom layers. The stratification strength of the Saemangeum Reservoir was, therefore, considered high. Moreover, a two-layer circulation occurred due to the strength of the eddy and the horizontal gradient of salinity. When the sluice gate was expanded, the area and number of days where hypoxia occurred decreased. It would also mitigate the retention of materials caused by the eddy and improve circulation. The findings of this study can be used as foundational data for solving and managing hydraulic problems within CRs.

Machine Learning-Based Predictions of Flow and Heat Transfer Characteristics in a Lid-Driven Cavity with a Rotating Cylinder

Machine learning-based predictions of heat transfer characteristics in lid-driven cavities are transforming the field of computational fluid dynamics (CFD). Lid-driven cavities are a fundamental problem in fluid mechanics, characterized by the motion of a fluid inside a square cavity driven by the motion of one of its walls. The goal of this study was to develop multiple machine-learning regression models and highlight the discrepancies between the predicted and actual average Nusselt numbers. Additionally, the study utilized physics-informed neural networks (PINNs) to model the flow and thermal behavior at both low and high Reynolds numbers. The results were compared among actual data from computational fluid dynamics (CFD) simulations, PINN models trained with CFD data, and purely PINN models created without any prior data input. The findings of this study showed that the random forest model exhibited an exceptional stability in its predictions, consistently maintaining low errors even as the Reynolds number increased compared with other machine-learning regression models. Further, the results of this study in terms of flow and thermal behavior within the cavity were found to depend significantly on the PINN method. The data-driven PINN exhibited a much lower mean average errors at both Reynolds numbers, while the physics-based PINN showed lower physics loss.

The dynamics of shock wave propagation far downstream of an abrupt area expansion

Predicting the transient flow fields that develop when a shock wave passes through an area expansion is a fundamental problem in compressible fluid mechanics and significant in many engineering applications. Experiments, large eddy simulations and geometrical shock dynamics are used to study the mechanism by which a normal shock wave that expands across an area expansion evolves into a uniform normal shock far downstream of it. This study analyses shock waves with moderate Mach numbers of 1.1–1.8 that expand at area ratios of up to 5. As the shock wave propagates into the expanded region, it experiences rapid deceleration, forming a non-uniform shock front. Impinging on the walls of the larger cross-section region, the shock wave reflects and generates a complex and highly transient shock pattern near the expansion region. We have found that as the shock front propagates further downstream, a laterally moving shock wave that intersects the shock front at a triple point reverberates laterally between the walls. This process effectively evens out the flow behind the incident shock front, thus reducing the variation of properties behind it. The extended duration of this process leads to significant pressure fluctuation behind the shock front. The results show that the evolution of the shock front can be scaled using the expanded region height and the velocity of the shock wave far downstream of the expansion. The results enabled the formulation of a simple empirical relation, allowing us to predict the shock velocity far downstream of gradual and abrupt area expansions.

A magnetohydrodynamic flow of a water-based hybrid nanofluid past a convectively heated rotating disk surface: A passive control of nanoparticles

Abstract One of the basic fluid mechanics problems of fluid flows over a revolving disk has both theoretical and real-world applications. The flow over a rotating disk has been the subject of numerous theoretical studies because it has many real-world applications in areas like rotating machinery, medical equipment, electronic devices, and computer storage. It is also crucial for engineering processes. Therefore, this article deals with a time-independent water-based hybrid nanofluid flow containing copper oxide and silver nanoparticles past a spinning disk. The Newtonian flow is taken into consideration in this analysis. The influence of magnetic field, thermophoresis, nonlinear thermal radiation, Brownian motion, and activation energy has been considered. The present analysis is modeled in a partial differential equation form and is then converted to ordinary differential equations using appropriate variables. A numerical solution using the bvp4c technique is accomplished using MATLAB software. The current results are matched with the previous literature and established a close relationship with previous studies. The purpose of this investigation is to numerically investigate the time-independent hybrid nanofluid flow comprising copper oxide and silver nanoparticles over a rotating disk surface. The results show that the increased magnetic parameters increase the friction force at the surface, which decreases the radial and azimuthal velocity distribution. At the sheet surface, the radial velocity of the hybrid nanofluid shows dominant performance compared to the nanofluid. On the other hand, the magnetic factor has dominant behavior on the azimuthal velocity component of the nanofluid flow compared to the hybrid nanofluid flow. The higher volume fraction and magnetic factor enhance the skin friction at the disk surface. Furthermore, greater surface drag is found for the hybrid nanofluid flow. The higher solid volume fraction, temperature ratio, and Biot number enhance the rate of heat transmission. Also, a higher rate of heat transmission is observed for the hybrid nanofluid flow.

Self-similar solution for laminar bubbly flow evolving from a vertical plate

The development of a bubble plume from a vertical gas-evolving electrode is driven by buoyancy and hydrodynamic bubble dispersion. This canonical fluid mechanics problem is relevant for both thermal and electrochemical processes. We adopt a mixture model formulation for the two-phase flow, considering variable density (beyond Boussinesq), viscosity and hydrodynamic bubble dispersion. Introducing a new change of coordinates, inspired by the Lees–Dorodnitsyn transformation, we obtain a new self-similar solution for the laminar boundary layer equations. The results predict a wall gas fraction and gas plume thickness that increase with height to the power of 1/5 before asymptotically reaching unity and scaling with height to the power 2/5, respectively. The vertical velocity scales with height to the power of 3/5. Our analysis shows that self-similarity is only possible if gas conservation is entirely formulated in terms of the gas specific volume instead of the gas fraction.

A Kronecker product accelerated efficient sparse Gaussian Process (E-SGP) for flow emulation

A Kronecker product accelerated efficient sparse Gaussian Process (E-SGP) for flow emulation

Dissipation Effects in the Tea Leaf Paradox

The Tea Leaf Paradox (TLP) describes unsteady fluid motions which help entrain and deposit suspended particles at the center of rotation. Various applications depend on the TLP for particle separations—spanning orders of magnitude in length scales—making it an important problem in fluid mechanics. Despite papers describing the phenomenon, the efficacy of particle separation using the TLP remains unclear as to the relative importance of, for example, hydrostatics, particle-fluid density ratio, wall friction, liquid bath aspect ratio and the rotation speed. The dynamics involved are notably complex and require a careful tuning of each variable. In this study, we have investigated the role of the limit of the aggregation dynamics in rotational flows within 3D-printed vessels of various sizes in tandem with particle imaging to probe the dissipation effects on the particle motions. We have found that the liquid bath aspect ratio limits how much aggregation may occur for a particle-fluid density ratio greater than unity (e.g., ρp/ρf>1), where ρp is the density of the particle and ρf is the ambient fluid density.

A comparative study of ensemble methods and multi-output classifiers for predictive maintenance of hydraulic systems

The maintenance and upkeep of hydraulic systems play a pivotal role in various industrial applications, from manufacturing processes to heavy machinery operations. Traditional strategies often face significant limitations, particularly concerning intervention time and the costs associated with delays between fault occurrence and maintenance. Over the past decades, proactive strategies have emerged with promising potential, primarily due to their predictive capabilities. These proactive approaches aim to anticipate faults and maintenance needs, thereby mitigating costs and operational disruptions associated with reactive approaches. This study explores the applicability of single output, ensemble methods, and the integration of multi-output classifiers in hydraulic predictive maintenance problems with relatively limited data. First, data is analyzed using Pearson correlation coefficients, and then feature extraction is conducted using recursive feature extraction, aiming to optimize the performance of predictive models, particularly Random Forest and CatBoost. Results show that the stacking ensemble method, incorporating LightGBM, XGBoost, CatBoost, and Random Forest, yields the most notable improvement, achieving a final accuracy of 98.63%. The results obtained in this study show satisfying performances for single-output models, ensemble methods, and multi-output models in predicting the health of hydraulic systems. Moreover, combining single output, ensemble methods, and the integration of multi-output classifiers has created a relatively reliable and robust predictive maintenance system.