995 papers
Fluid dynamics explores how liquids and gases move, shaping everything from weather patterns to the flow of blood through our veins. This field bridges the gap between abstract mathematical equations and the tangible forces that drive our physical world, offering insights into turbulence, aerodynamics, and fluid behavior in complex environments.
On Gist.Science, we process every new preprint in this category directly from arXiv to make cutting-edge research accessible to everyone. Each paper is transformed into a clear, plain-language overview alongside a detailed technical summary, ensuring both students and experts can grasp the latest findings without getting lost in dense jargon.
Below, you will find the most recent studies in fluid dynamics, curated and explained for a broader audience.
This paper introduces a fully data-free Physics-Informed Neural Network framework that solves compressible inviscid flows up to Mach 15 around a cylinder by integrating a hybrid convolutional architecture, Mach-guided dynamic residual scaling, and specialized loss constraints to overcome spectral bias, gradient pathologies, and shock-capturing instabilities.
This study demonstrates that graph neural networks can effectively reconstruct fine-scale structures in turbulent reacting flows on complex, unstructured meshes, overcoming the limitations of traditional deep learning models restricted to uniform grids.
This study experimentally demonstrates that rectangular orifice synthetic jet actuators can effectively disrupt the rotational coherence of wall-bounded vortices by up to 70%, thereby recovering pressure in the vortex wake and offering a viable strategy for vortex mitigation when the vortex's position and size are known.
This paper presents a unified cross-dimensional discrete Boltzmann framework for simulating compressible flows with tunable specific heat ratios, utilizing a high-symmetry discrete velocity set for Galilean invariance and an operator-splitting scheme to extend a one-dimensional kinetic formulation to multi-dimensional systems, all of which are validated against standard benchmark problems.
This paper presents a numerical study using the vortex filament model to demonstrate that superfluid He flowing past a rough wall at sustains a polarized ultraquantum turbulence state above a critical velocity, characterized by a parabolic velocity profile with wall slip and a friction force proportional to the flow speed.
This study demonstrates that applying the Hilbert transform directly to standard Proper Orthogonal Decomposition (POD) modes offers a computationally efficient and artifact-free method for identifying propagating turbulent structures in a sphere's near wake, effectively replicating the results of the more complex Hilbert POD technique.
This paper validates the hypothesis that structure tensors provide a sufficient statistical description for turbulence by demonstrating that equivariant neural networks utilizing these tensors yield significantly more accurate Reynolds-averaged Navier-Stokes closures for the rapid pressure-strain term compared to existing models.
This paper investigates the stability of vortex-induced vibrations in multiple freely oscillating bodies by deriving a Linear Arbitrary Lagrangian Eulerian method and proposing a low-cost impedance-based criterion to accurately predict instability thresholds, which are validated through global stability analysis and extensive parametric studies on tandem and three-body cylinder systems.
By studying viscoelastic flows in curved pipes, this paper reveals that turbulence at vanishing inertia arises from the competition between two hydrodynamic instabilities leading to co-existing turbulent states, thereby challenging the traditional distinction between elastic and elasto-inertial turbulence.
This paper presents a Bayesian framework for identifying flame impulse responses that utilizes a physically motivated distributed time delay model and Bayesian model comparison to automatically select the optimal model complexity, thereby producing robust, physically consistent results with fewer spurious features and reduced data requirements compared to traditional system identification methods.