Reconstructing velocity and pressure from sparse noisy particle tracks using Physics-Informed Neural Networks

arXiv (Cornell University) · 2022 · 6 citations

• Computer Science
• Artificial Intelligence
• Physics

Volume-resolving imaging techniques are rapidly advancing progress in experimental fluid mechanics. However, reconstructing the full and structured Eulerian velocity and pressure fields from sparse and noisy particle tracks obtained experimentally remains a significant challenge. We introduce a new method for this reconstruction, based on Physics-Informed Neural Networks (PINNs). The method uses a Neural Network regularized by the Navier-Stokes equations to interpolate the velocity data and simultaneously determine the pressure field. We compare this approach to the state-of-the-art Constrained Cost Minimization method [1]. Using data from direct numerical simulations and various types of synthetically generated particle tracks, we show that PINNs are able to accurately reconstruct both velocity and pressure even in regions with low particle density and small accelerations. PINNs are also robust against increasing the distance between particles and the noise in the measurements, when studied under synthetic and experimental conditions.

Spectral Universality of Elastoinertial Turbulence

Physical Review Letters · 2021 · 40 citations

• Physics
• Mechanics
• Classical mechanics

Dissolving small amounts of polymer into a Newtonian fluid can dramatically change the dynamics of transitional and turbulent flows. We investigate the spatiotemporal dynamics of a submerged jet of dilute polymer solution entering a quiescent bath of Newtonian fluid. High-speed digital Schlieren imaging is used to quantify the evolution of Lagrangian features in the jet revealing a rich sequence of transitional and turbulent states. At high levels of viscoelasticity, we identify a new distinct transitional pathway to elastoinertial turbulence (EIT) that does not feature the conventional turbulent bursts and instead proceeds via a shear-layer instability that produces elongated filaments of polymer due to the nonlinear effects of viscoelasticity. Even though the pathways to the EIT state can be different, and within EIT the spatial details of the turbulent structures vary systematically with polymer microstructure and concentration, there is a universality in the power-law spectral decay of EIT with frequency, f^{-3}, independent of fluid rheology and flow parameters.

Two-point stress–strain-rate correlation structure and non-local eddy viscosity in turbulent flows

Journal of Fluid Mechanics · 2021 · 42 citations

• Mechanics
• Materials science
• Physics

Abstract