Probabilistic physics-constrained reconstruction of three-dimensional flow fields
- 1 April 2026
- journal article
- Published by AIP Publishing in Physics of Fluids
- Vol. 38 (4)
Abstract
Accurate reconstruction of three-dimensional flow fields is essential for resolving fine-scale flow structures and understanding unsteady flow physics in ship and ocean engineering applications. However, while direct numerical simulation provides high-fidelity solutions of the Navier–Stokes equations, its prohibitive computational cost limits its applicability to large-scale and high-Reynolds-number flows. Reduced-order and data-driven reconstruction approaches offer improved efficiency, yet often fail to preserve inter-slice coherence, fine-scale vortical structures, and physical consistency when extrapolating beyond the training regime. To address these challenges from a physical-consistency perspective, this study investigates a physics-constrained framework for three-dimensional flow-field reconstruction. The governing physical constraints are embedded into the reconstruction process through a probabilistic formulation derived from the vorticity–velocity representation of the Navier–Stokes equations, enabling improved enforcement of incompressibility and dynamical consistency. A spatiotemporal reconstruction architecture is constructed to capture multi-scale spatial features and temporal evolution characteristics of unsteady flows, while a physics-informed attention mechanism is employed to enhance the recovery of fine-scale flow structures in regions with strong spatial gradients. The proposed framework is assessed using turbulent flow data from the Johns Hopkins Turbulence Database, transient boundary layer flows, and experimental tomographic particle image velocimetry measurements. Quantitative comparisons with high-fidelity reference data demonstrate that the physics-constrained reconstruction significantly improves the preservation of fine-scale flow structures and extrapolation performance compared with conventional deterministic approaches, indicating its effectiveness for physically consistent three-dimensional flow-field reconstruction at reduced computational cost.Keywords
Funding Information
- National Natural Science Foundation of China (52401362)
- Natural Science Foundation of Liaoning Province (2025-BS-0231)
This publication has 30 references indexed in Scilit:
- A multi-scale hybrid attention Swin-transformer-based model for the super-resolution reconstruction of turbulenceNonlinear Dynamics, 2025
- PhySR: Physics-informed deep super-resolution for spatiotemporal dataJournal of Computational Physics, 2023
- Deep learning for fluid velocity field estimation: A reviewOcean Engineering, 2023
- Physics-informed deep learning for incompressible laminar flowsTheoretical and Applied Mechanics Letters, 2020
- A high-order multiscale approach to turbulence for compact nodal schemesComputer Methods in Applied Mechanics and Engineering, 2020
- Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equationsJournal of Computational Physics, 2019
- Adaptive detrending to accelerate convolutional gated recurrent unit training for contextual video recognitionNeural Networks, 2018
- Bayesian deep convolutional encoder–decoder networks for surrogate modeling and uncertainty quantificationJournal of Computational Physics, 2018
- Resolution effects and scaling in numerical simulations of passive scalar mixing in turbulencePhysica D: Nonlinear Phenomena, 2010
- A Compact-Difference Scheme for the Navier–Stokes Equations in Vorticity–Velocity FormulationJournal of Computational Physics, 2000