Research progress in hydrofoil cavitation prediction and suppression methods

Physics of Fluids, Journal Year: 2025, Volume and Issue: 37(1)

Published: Jan. 1, 2025

To reduce the adverse damage caused by cavitation phenomena to hydraulic machinery, such as surface erosion of equipment, increased mechanical vibration, and decreased service life, this review summarizes from aspects instability mechanisms, prediction methods, suppression methods. In terms flow two main mechanisms that affect shedding cloud cavitation, reentrant jet, bubbly shock wave, were thoroughly summarized. It is pointed out behavior cavity greatly influenced thickness jet relative cavity, wave also one important factors in vortex dynamics. a detailed comparison analysis made between traditional methods based on numerical simulation currently popular neural networks. The former mainly includes models turbulence models, while latter application chain physics-informed network, pressure–velocity long short-term memory, other networks prediction. artificial intelligence predictive have advantages model order reduction accurate field feature parameters. active passive Finally, current research status hydrofoil article discusses looks forward direction development.

Language: Английский

---

Airfoil Flowfield Prediction Based on Proper Orthogonal Decomposition with Deep Learning

Journal of Aircraft, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 15

Published: Feb. 12, 2025

In aerodynamics, the flowfield around an airfoil is typically solved using experiments or computational fluid dynamics (CFD), while they are often computationally expensive. Reduced-order models (ROMs) can effectively balance accuracy and efficiency of CFD, deep learning also be integrated with ROMs to approximate flowfields, thereby avoiding need solve complex Navier–Stokes (N-S) equations. Using NACA0012 as a case study, proper orthogonal decomposition–gated recurrent unit (POD-GRU) model proposed for approximation purpose. For approximation, CFD simulation data used training, POD employed dimensionality reduction, GRU network predict coefficients reconstruction. This first time that has been airfoil. The achieves prediction error within 8% high-fidelity models, its cost single operating condition merely 0.4% required by traditional simulations. remarkable make method particularly suitable applications requiring rapid response, such control mitigate stall, flow interference, enhancing lift during flight. Furthermore, demonstrates robust predictive capabilities across varying inflow velocities angles attack, showing significant potential engineering where both speed precision critical.

Language: Английский

---

A graph neural network surrogate model for multi-objective fluid-acoustic shape optimization

Computer Methods in Applied Mechanics and Engineering, Journal Year: 2025, Volume and Issue: 441, P. 117921 - 117921

Published: April 4, 2025

Language: Английский

---

RegStack machine learning model for accurate prediction of tidal stream turbine performance and biofouling

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127766 - 127766

Published: April 1, 2025

Language: Английский

---

Biomimetic Hydrofoil Propulsion: Harnessing the Propulsive Capabilities of Sea Turtles and Penguins for Robotics

Biomimetics, Journal Year: 2025, Volume and Issue: 10(5), P. 272 - 272

Published: April 28, 2025

This review synthesizes current research on hydrofoil-propelled robots inspired by the swimming mechanisms of sea turtles and penguins. It begins summarizing kinematics these organisms, highlighting their superior aquatic performance as primary motivation for biomimetic design. Next, established analytical methods characterizing hydrofoil locomotion patterns are presented, along with a clear delineation decoupled motion components exhibited turtle flippers penguin wings. Such decoupling provides systematic framework guiding design driving mechanisms. Building this biomechanical foundation, critically examines recent advances in flexible hydrofoils that enhance propulsion efficiency through three synergistic to thrust generation, while identifying key challenges material durability non-linear fluid–structure interactions. The then surveys existing actuation systems, which commonly reproduce coupled motions multiple degrees freedom (DOFs). Finally, representative examined: turtle-inspired forelimbs typically incorporate DOFs, whereas penguin-inspired wings usually offer two DOFs. By aligning robotic designs source offers critical insights advance development systems enhanced performance.

Language: Английский

---

Fast prediction of propeller dynamic wake based on deep learning

Physics of Fluids, Journal Year: 2024, Volume and Issue: 36(8)

Published: Aug. 1, 2024

Efficiently predicting the wake of propellers is great importance for achieving propeller design optimization. In this work, deep learning (DL) method called convolutional neural networks (PWCNN) proposed, which combines transformer encoder and dilated block to capture multi-scale characteristics wakes. Computational fluid dynamics (CFD) simulations are conducted using delayed detached eddy simulation model generate extensive high-fidelity data operating under different conditions required DL. PWCNN takes predicted at previous time step update input iteratively predicts future steps achieve dynamic prediction. The good agreement between DL prediction CFD results, with mean relative error velocity components less than 2.36% 15 steps, proves that can efficiently spatiotemporal evolution characteristic Furthermore, predict changes reasonable accuracy unseen conditions, further confirming generality proposed in forecasting wake.

Language: Английский

---

Study of the hydrodynamic characteristics of the blade based on a bionic hydrofoil at low flow velocity

Ocean Engineering, Journal Year: 2024, Volume and Issue: 318, P. 120102 - 120102

Published: Dec. 19, 2024

Language: Английский

---