FIELDTURB

Complex Fluids Dynamics

The research developed within the framework of the specific initiative FieldTurb focuses on the problem of the dynamics of complex and turbulent fluids, and on the transport and interaction of “Particles and Fields” within them. The aim of the project is to achieve a deeper understanding of fundamental issues related to the general problems of classical field theories for systems out of equilibrium on macro-, micro-, and nanometric scales, as well as numerous applied problems concerning non-ideal turbulence, including fluids with complex rheology, relativistic flows, and control issues.

Example of generation using diffusion models of Lagrangian trajectories in developed turbulence.
This powerful technique allows one to solve both data augmentation and data assimilation or super-resolution problems. Ref. [1]

These problems are addressed by combining the traditional tools of theoretical physics with interdisciplinary methodologies. In particular, the research group has developed an approach that integrates theory with numerical simulations and data analysis, based on many years of experience in using massively parallel distributed computing systems and, in recent years, through the integration of innovative Machine Learning (ML) and Data Assimilation techniques.
This integrated approach has also become necessary due to the recent shift in focus from fundamental questions to non-ideal situations, which are characterized by various mechanisms of symmetry breaking.

Multiphase fluids and emulsions exhibit non-Newtonian behavior that significantly affects their heat transport properties under thermal convection Ref. [2], e.g., when placed between two walls at different temperatures and gravity effects are non-negligible.
The Tor Vergata group has released TLBfind, an open-access code for the numerical simulation of such systems Ref [3] .

Paradigmatic examples include the effects of anisotropy or inhomogeneity due to spatial confinement and the presence of other phases. Similarly, for complex fluids on a microscopic scale, such as foams and emulsions, a complete characterisation is still lacking, with many open challenges for theory, numerical simulations, and experiments.
Furthermore, the fluid mechanics community is currently heavily engaged in researching solutions to many problems related to the use of machine learning techniques: from data assimilation to super-resolution, from small-scale modelling to the navigation of Lagrangian agents in complex flows, such as in the problem of olfactory search or achieving a given target.
These highly challenging issues find applications in fields ranging from environmental monitoring to micromedicine and behaviour studies.

Useful reference
National Url
Massimo Cencini
Luca Biferale
Mauro Sbragaglia
Michele Buzzicotti