Abstract:Experimental measurements and numerical simulations of turbulent flows are characterized by a tradeoff between accuracy and resolution. In this study, we combine accurate sparse pointwise mean velocity measurements with the Reynolds-averaged Navier-Stokes (RANS) equations using data assimilation methods. Importantly, we bridge the gap between data assimilation (DA) using physics-informed neural networks (PINNs) and variational methods based on classical spatial discretization of the flow equations, by comparing both approaches on the same turbulent flow case. First, by constraining the PINN with sparse data and the underdetermined RANS equations without closure, we show that the mean flow is reconstructed to a higher accuracy than a RANS solver using the Spalart-Allmaras (SA) turbulence model. Second, we propose the SA turbulence model augmented PINN (PINN-DA-SA), which outperforms the former approach by up to 73% reduction in mean velocity reconstruction error with coarse measurements. The additional SA physics constraints improve flow reconstructions in regions with high velocity and pressure gradients and separation. Third, we compare the PINN-DA-SA approach to a variational data assimilation using the same sparse velocity measurements and physics constraints. The PINN-DA-SA achieves lower reconstruction error across a range of data resolutions. This is attributed to discretization errors in the variational methodology that are avoided by PINNs. We demonstrate the method using high-fidelity measurements from direct numerical simulation of the turbulent periodic hill at Re=5600 . https://doi.org/10.1103/PhysRevFluids.9.034605 Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI. Published by the American Physical Society

State estimation in minimal turbulent channel flow: A comparative study of 4DVar and PINN

Flow Reconstruction with Uncertainty Quantification from Noisy Measurements Based on Bayesian Physics-Informed Neural Networks

Turbulence model augmented physics informed neural networks for mean flow reconstruction

Turbulence model augmented physics-informed neural networks for mean-flow reconstruction

Dense velocity reconstruction from particle image velocimetry/particle tracking velocimetry using a physics-informed neural network

Mean flow reconstruction of unsteady flows using physics-informed neural networks

Reconstruction of the turbulent flow field with sparse measurements using physics-informed neural network

RANS-PINN based Simulation Surrogates for Predicting Turbulent Flows

Finite Volume Physical Informed Neural Network (FV-PINN) with Reduced Derivative Order for Incompressible Flows

Turbulence Modeling for Physics-Informed Neural Networks: Comparison of Different RANS Models for the Backward-Facing Step Flow

Assimilating experimental data of a mean three-dimensional separated flow using physics-informed neural networks

Applying physics informed neural network for flow data assimilation

Enhancement of PIV measurements via physics-informed neural networks

Leveraging turbulence data with physics informed neural networks

Comparative Assessment for Pressure Field Reconstruction Based on Physics-Informed Neural Network

New insights into experimental stratified flows obtained through physics-informed neural networks

Learning time-averaged turbulent flow field of jet in crossflow from limited observations using physics-informed neural networks

Physics-informed neural network to augment experimental data: an application to stratified flows

HDNet: Physics-Inspired Neural Network for Flow Estimation based on Helmholtz Decomposition

A practical approach to flow field reconstruction with sparse or incomplete data through physics informed neural network

Using Parametric PINNs for Predicting Internal and External Turbulent Flows