Seminar: Fluctuating Multi-phase Hydrodynamics: from Boltzmann to Ising, and back

ABSTRACT

Fluctuating Hydrodynamics (FH) [1], naturally extends Navier-Stokes equations by including a stochastic stress tensor, capturing the mesoscopic effect of the agitation of the fluid molecules. FH has enabled the study of the surface tension curvature corrections in homogeneous and heterogeneous nucleation [2, 3],  of corrections to the dissipation spectra in turbulence [4], and of giant fluctuations in diffusion [5]. Mesoscopic kinetic models can simulate complex fluid flows at different scales, ranging from microfluidics to turbines, with a dramatic reduction of computational cost with respect to atomistic simulations. As a trade-off one needs to face the challenge of consistently mapping the micro-physics and the mesoscopic modelling with an appropriate coarse graining.

A consistent mesoscopic FH model, promises to successfully capture the intricate multi-scale physics of important phenomena posing long-standing challenges such as nucleation in fuel cells, cavitation in propellers and instabilities in injectors, where interface physics, thermal fluctuations and hydrodynamics, all come into play.

I will present recent results concerning mesoscopic and microscopic FH models aiming at defining a consistent coarse-graining transformation. I will show how hydrodynamic fluctuations in the multi-phase Shan-Chen lattice Boltzmann model (LBM) [6] display an unprecedented agreement between theory and numerics across all scales, yielding density structure factors that capture non-trivial experimental features. I will introduce a kinetic Ising model with distinguishable particles, featuring several parameters tuning time and length scales. Results show an excellent agreement with respect to a recently proposed coarse-graining procedure mapping Molecular Dynamics onto LBM [7], at a much lower computational cost. These results leverage advanced meta-programming techniques available in the "idea.deploy" [8] computational framework.

(1) L.D. Landau, E.M. Lifshitz, JETP, 1957, Vol. 5, No. 3, p. 512; (2) M. Gallo, F. Magaletti, D. Cocco, C.M. Casciola, JFM (2020), vol. 883, A14; (3) ML, L. Biferale, G. Falcucci, M. Sbragaglia and X. Shan, PRE 105, 015301 (2022); (4) R. M. McMullen, M. C. Krygier , J. R. Torczynski, M. A. Gallis, PRL 128, 114501 (2022); (5) A. Vailati, M. Giglio, Nature 390, 1997; (6) ML, L. Biferale, G. Falcucci, M. Sbragaglia, D. Yang and X. Shan, arXiv:2212.07848; (7) M. R. Parsa and A. J. Wagner, PRE 96, 013314 (2017); (8) https://github.com/lullimat/idea.deploy.

BIOGRAPHY

Dr. Lulli graduated in 2015 from the Physics dept. of the University of Rome Sapienza, on out-of-equilibrium Renormalization Group techniques and High-Performance Computing for the study of the 3D Ising spin-glass transition.  From 2015 to 2018, he worked on complex flows and the yield-stress transition in emulsions, in the Netherlands, at the Physics of Fluids group in Twente, in TU/e Eindhoven and at the Physics dept. of the University of Rome Tor Vergata. In 2018 he worked on structural glasses at the Applied Physics dept. of The Hong Kong Polytechnic University. From 2019 he has been working on the mesoscopic modelling of multi-phase/component interfaces at the Mechanics and Aerospace Engineering department of Southern University of Science and Technology (SUSTech). Recent works also pertain Stochastic Quantization of GR and Topological Quantum Neural Networks. He is now Research Associate at the Physics dept. of the Chinese University of Hong Kong and Visiting Scholar at SUSTech.