Latest papers in fluid mechanics

Author(s): Zhiqian Xin, Jiadong Wang, Xingyuan Mao, Bowen Jin, and Jian Deng

This study investigates the influence of varying degrees of freedom (DOFs) on the swimming performance of self-propelled undulatory swimmers navigating a straight path in three flow configurations: an unbounded fluid, near a solid wall, and in a side-by-side arrangement. Vertical and rotational DOFs…

[Phys. Rev. E 112, 035103] Published Fri Sep 19, 2025

Author(s): Martino Andrea Scarpolini, Giovanni Vagnoli, Fabio Guglietta, Roberto Verzicco, and Francesco Viola

Choosing the mounting position of an aortic prosthesis—intra- or supra-annular—remains debated, as clinical comparisons rarely isolate hemodynamic effects of valve replacement procedures. We perform fluid-structure interaction simulations of a patient-specific left heart, testing the same bioprosthetic valve in both configurations on the same patient, rationalizing mounting and valve size effects. Supra-annular implantation consistently lowers transvalvular pressure gradients, increases orifice area, and reduces shear and hemolysis risk (see figure). These results provide controlled evidence to guide implantation strategy and device selection in cases of patient-prosthesis mismatch risk.

[Phys. Rev. Fluids 10, 090501] Published Thu Sep 18, 2025

Author(s): R. Shapiro and A. Manela

The impact of gas rarefaction on the two-dimensional aerodynamic ground effect over a flat plate is analyzed. The free-molecular problem was studied analytically based on the collisionless Boltzmann equation and Maxwell boundary conditions, and compared with direct simulation Monte Carlo at finite Knudsen numbers (Kn). The results indicate that the ground invariably increases aerodynamic loading on the plate and shifts the maximum lift to lower angles of attack compared with the non-confined configuration (NC). While the ground may contribute negatively to the lift in the ideal-flow limit, its relative difference compared with NC is found to be significantly larger and positive at high Kn.

[Phys. Rev. Fluids 10, 093401] Published Thu Sep 18, 2025

Author(s): Jason K. Kabarowski, Aditya S. Khair, and Rahil N. Valani

A mathematical model is developed for the spontaneous Qunicke rotation of a dielectric sphere in an electric field, which accounts for fluid and particle inertia. The particle dynamics obey an integro-differential dynamical system that is a generalization of the celebrated Lorenz equations. Analysis and numerical solution of these modified Lorenz equations show that fluid inertia inhibits chaotic particle rotation, in qualitative agreement with prior experimental observations.

[Phys. Rev. Fluids 10, 093701] Published Thu Sep 18, 2025

Author(s): Haosen Liu and Benshuai Lyu

The stability characteristics of a compressible shear flow confined by two rigid plates are important in applications such as open-jet facilities and launching rockets. By performing linear spatiotemporal stability analysis, we identify the critical parametric regions within which absolute instability occurs. In particular, we show that at sufficiently high Mach numbers, a new type of absolutely unstable mode occurs, which arises from a feedback process due to the reflection of compressible waves by the rigid plates.

[Phys. Rev. Fluids 10, 093903] Published Thu Sep 18, 2025

Author(s): Yingtao Sun, Di Bian, Yuchen Wang, Kai Zhang, Jianfeng Zhou, and Zhigang Li

Superhydrophobic grooves offer substantial slip and drag reduction; however, real fluids are seldom completely clean. Using many-body dissipative particle dynamics simulations, we demonstrate that the presence of contaminant particles at the interface significantly decreases both local and effective slip. The primary factors influencing this effect are particle wettability and interfacial coverage, while particle size and mass have a minor role. The reduction in effective slip follows Philip’s model, providing a rule-of-thumb predictor and informing designs that can tolerate or manage contamination.

[Phys. Rev. Fluids 10, 094202] Published Thu Sep 18, 2025

Author(s): Daniel J. Booth and Thomas D. Montenegro-Johnson

Precise, localized flow control in microfluidic devices remains a difficult challenge. We demonstrate, theoretically, how active droplets might be harnessed to overcome this challenge. Active droplets are produced along the microchannel wall via stimulation of a responsive hydrogel, and the ensuing phoretic slip flows drive transport and mixing in the microfluidic device.

[Phys. Rev. Fluids 10, 094203] Published Thu Sep 18, 2025

Author(s): Jinyu Yao, Xinshu Zhang, Huawei Zhou, Xingyu Song, and Alessandro Toffoli

We develop a nonlinear wave reconstruction and prediction model with a dynamic averaging algorithm, in which shipborne radar images are used for data assimilation to improve the accuracy of wave reconstruction and prediction. Waves around the ship can be accurately predicted for the next few minutes under various sea states. Compared with the linear and second- order models, the new model includes the third-order nonlinear effects; thus, it significantly improves the prediction accuracy of extreme waves under rough sea states, providing effective safety guarantees for ship navigation and operations.

[Phys. Rev. Fluids 10, 094801] Published Thu Sep 18, 2025

Author(s): Samuel J. Petter, Benjamin C. Musci, Gokul Pathikonda, Prasoon Suchandra, and Devesh Ranjan

This study advances the understanding of blast-driven interface instabilities by transitioning from qualitative Mie scattering to quantitative planar particle image velocimetry (PIV). The velocity field and vorticity evolution reveal key insights into mixed-mode Richtmyer-Meshkov and Rayleigh-Taylor instabilities in a cylindrical geometry. High-Atwood number cases exhibit prolonged circulation growth, consistent with stronger turbulence and earlier mixing transition. The PIV data captures how pressure impulse and decay shape the instability beyond what Mie images alone can resolve.

[Phys. Rev. Fluids 10, 093902] Published Wed Sep 17, 2025

Author(s): Haoyu Wang, Isaac J. G. Lewis, Soohyeon Kang, Yuechao Wang, Leonardo P. Chamorro, and C. Ricardo Constante-Amores

We present data-driven models for predicting the motion of a freely settling sphere in a quiescent fluid using experimentally measured trajectories. Deterministic and stochastic neural differential equations reconstruct individual particle paths and capture the statistical features of settling dynamics without resolving the surrounding flow. Our results reveal the strengths of each modeling approach. Deterministic models excel at trajectory prediction, while stochastic models reproduce long-time statistical trends, thus providing a framework for reduced-order modeling of particulate flows.

[Phys. Rev. Fluids 10, 094402] Published Wed Sep 17, 2025

Author(s): Firoozeh Foroozan, Andrea Ianiro, Stefano Discetti, and Woutijn J. Baars

Experimental measurements were performed of convective heat transfer fluctuations beneath a grazing turbulent boundary layer flow. Spatiotemporal wall-temperature fields were acquired with an infrared camera and a heated-thin-foil sensor. Inferred Nusselt number fluctuations showed elongated features with scales similar to near-wall streaks. An analysis in the frequency–wavenumber domain revealed dispersive convection: larger streaks moved near freestream velocity, while smaller energetic features traveled at 10 times the friction velocity. These measurements provide a promising method for wall-based turbulence sensing and flow control.

[Phys. Rev. Fluids 10, 094904] Published Wed Sep 17, 2025

Author(s): Yue Hao, Charles Meneveau, and Tamer A. Zaki

Data assimilation provides a rigorous framework for integrating measurements with numerical simulations to estimate the flow. We developed a machine-learning-based assimilation strategy to infer unknown upstream flow conditions in a high-speed boundary layer from sparse wall-pressure measurements. Our method uses a Bayesian optimization to efficiently search for the optimal control parameters. Applied to a transitional boundary layer, the method accurately estimates the oncoming disturbances, and subsequent direct numerical simulation (DNS) predictions using the estimated conditions show excellent agreement with the true flow.

[Phys. Rev. Fluids 10, 094905] Published Wed Sep 17, 2025

Author(s): Kunihiko Taira, Georgios Rigas, and Kai Fukami

The fluid dynamics community has increasingly adopted machine learning to analyze, model, predict, and control a wide range of flows. This perspective article offers a critical assessment of the key challenges that must be addressed for deepening our understanding of flow physics and expanding the applicability of machine learning beyond fundamental research. We also highlight the importance of community-maintained datasets and open-source code repositories, as well as effective training of fluid mechanicians. We hope this paper sparks discussions and encourages collaborative efforts to advance the integration of machine learning in fluid dynamics.

[Phys. Rev. Fluids 10, 090701] Published Tue Sep 16, 2025

Author(s): Florian Voss and Uwe Thiele

Chemomechanical phenomena lie at the core of many biological and biomimetic systems. Particularly in the presence of free interfaces, such effects arise naturally due to chemically induced gradients of interfacial tension. We study a simple, thermodynamically consistent model for liquid drops on solid substrates that captures the coupling between an autocatalytic reaction of insoluble surfactants, the Marangoni effect and wetting dynamics. In the presence of chemical fuel, drops may exhibit complex self-organized motility modes like crawling and shuttling. The underlying chemomechanical feedback and the resulting bifurcation structure are studied in detail.

[Phys. Rev. Fluids 10, 094005] Published Tue Sep 16, 2025

Author(s): Ping Wang, Jinchi Li, Qingqing Wei, and Xiaojing Zheng

Channel and zero-pressure-gradient spatially developing turbulent boundary layer are the two canonical wall-bounded flows. Despite the long-standing controversies about their similarity, there is little attention paid to the similarity/dissimilarity between these two types of particle-laden turbulence, which is one of the most important topics in turbulence research. The particle distribution, turbulent statistics, and structures in the two kinds of particle-laden flow are thoroughly compared for the identical particle Stokes number and bulk volume fraction at turbulent Reynolds number of Reτ≈400. Qualitative and quantitative differences are observed throughout the turbulence region.

[Phys. Rev. Fluids 10, 094303] Published Tue Sep 16, 2025

Author(s): Guillermo L. Nozaleda, Javier Alaminos-Quesada, Cándido Gutiérrez-Montes, and Antonio L. Sánchez

Oscillatory flows in porous environments arise in both biological and technological systems, yet their time-averaged steady streaming has been largely overlooked. Here we analyze steady streaming in slender channels with porous interiors using a homogenized model with Darcy resistance. We find that porous media not only attenuate streaming compared to unobstructed channels but also alter its structure. These results provide new insight into fluid transport in oscillatory flows through porous environments.

[Phys. Rev. Fluids 10, 093103] Published Mon Sep 15, 2025

Author(s): Xiaoxiao Yang, Darius Marin, Charlotte Py, Olivier Cardoso, Anke Lindner, and Sandra Lerouge

Elastic instabilities and turbulence driven by elastic hoop stresses are likely to develop on top of shear-banding flows in giant micelles. We show the existence of a generic flow diagram in an operating space built on the curvature ratio Λ of the Taylor-Couette flow and the Weissenberg number Wi, which compares elastic and viscous stresses. Two different pathways to purely elastic turbulence are identified depending on Λ, with clear signatures in the stress response. The geometric scaling of the onset of elastic turbulence is found to be reminiscent of the Pakdel-McKinley criterion that recasts the different mechanisms and most unstable instability modes of flows with curved streamlines.

[Phys. Rev. Fluids 10, 093302] Published Mon Sep 15, 2025

Author(s): Romane Le Dizès Castell, Marc Prat, Noushine Shahidzadeh, and Sara Jabbari-Farouji

Drying of complex fluids in porous media is crucial for applications such as preserving cultural heritage materials, yet the role of sol–gel transitions in evaporation kinetics remains unclear. We develop a pore-network model to investigate the emergence of gel-like skin at the evaporation interface. By incorporating pore size gradients and a viscosity-dependent vapor pressure rule, the model captures skin formation. Its predictions quantitatively match experiments and explain the evaporation slowdown during sol–gel transition.

[Phys. Rev. Fluids 10, 094302] Published Mon Sep 15, 2025

Author(s): Bobae Johnson, Scott Weady, Zihan Zhang, Alison Kim, and Leif Ristroph

Ice melting is an important part of the climate system that involves complex fluid dynamics and interactive processes. Here we address the capsize problem in which melting-induced changes in size and shape of free floating ice can trigger it to rotate and turn over. Experiments show that “lab icebergs” lock to the waterline while gradually melting, then abruptly lose stability and roll over to assume a new posture, and this process repeats many times as the ice melts down. A particular angle of rotation is selected and, consequently, the ice tends towards a polygonal shape. These results are reproduced by a model that predicts the coupled shape-posture dynamics and uncovers the key mechanisms.

[Phys. Rev. Fluids 10, 093801] Published Fri Sep 12, 2025

Author(s): Jiajun Hu, Zhen Lu, and Yue Yang

Machine learning can accelerate the prediction of fluid flow around obstacles, but existing models often struggle to generalize to geometries not seen during training. We introduce a generative diffusion model that uses an obstacle’s geometry as a conditional prompt to predict the corresponding instantaneous flow field. Trained only on elementary shapes, the model demonstrates superior generalization by capturing key features like vortex shedding and pressure distributions for unseen and complex geometries. By generating more physically consistent results, it outperforms standard neural network and variational autoencoder models, showing promise for accelerating CFD workflows.

[Phys. Rev. Fluids 10, 094903] Published Fri Sep 12, 2025