Final Report Summary - FFLOWCCS (Fluid Flow in Complex and Curved Spaces)
The subject of the ERC project was the application of fluid dynamics, in particular in the numerical realization of the Lattice Boltzmann method, to problems with deforming boundaries, with particular emphasis on porous media, curved spaces and quantum fluids. The porous media analysis was sided by statistical modelling and filtration experiments.
The single most important discovery which is an immediate consequence of the ERC grant is that ejections from filters follow a power law statistics. More generally porous media, that undergo deposition and erosion of material, release matter in form of bursts, which similar to earthquakes can be very large. This discovery was established experimentally, verified numerically with a sophisticated Lattice Boltzmann method and understood as well from the local point of view on the pore scale using DEM simulations and tomographic measurements as well as a collective phenomenon using an electrical analog. This discovery has fundamental consequences not only for the life-time of filters but also for the release of contaminants from the soil or sand production from oil wells. Also concerning statistical models for the clogging phenomena for porous media a substantial advance was achieved finding under which conditions abrupt phase transitions could be expected.
Huge progress has been achieved in the extension of the Lattice Boltzmann method to relativistic and quantum systems: New families of polynomials for Jüttner, Fermi-Dirac and Bose-Einstein distributions were introduced and ultra-relativistic simulations became possible. Additionally the formalism was generalized to arbitrarily curved spaces by including the spatial metric into a covariant formulation. Subsequently for the first time a Lattice Wigner model was developed. As unexpected spin-off, a novel DFT method to calculate molecular structures by solving the Kohn-Sham equation with Lattice Boltzmann was invented and tested, comparing in its first versions already quite favorably in performance with today’s standard methods. Lattice Boltzmann simulations were used to gain insight about the temporal sequence of solar flares, study instabilities of relativistic gases, modelling debris flow and simulate the flow through “campylotic”, i.e. randomly curved spaces. Particularly enlightening was the application to graphene, since its electronic flow can be described as a relativistic fluid.
Another novelty was the implementation of strongly deformable boundaries using subdivision shell elements, leading to the discovery of a yet unknown morphology of membranes deformed by wires, which we called the “warped” morphology. A new type of fluid structure FEM-LBM coupling was implemented by considering wall deformations through changes in the spatial metric of the fluid solver which can for instance simulate waving flags.
The single most important discovery which is an immediate consequence of the ERC grant is that ejections from filters follow a power law statistics. More generally porous media, that undergo deposition and erosion of material, release matter in form of bursts, which similar to earthquakes can be very large. This discovery was established experimentally, verified numerically with a sophisticated Lattice Boltzmann method and understood as well from the local point of view on the pore scale using DEM simulations and tomographic measurements as well as a collective phenomenon using an electrical analog. This discovery has fundamental consequences not only for the life-time of filters but also for the release of contaminants from the soil or sand production from oil wells. Also concerning statistical models for the clogging phenomena for porous media a substantial advance was achieved finding under which conditions abrupt phase transitions could be expected.
Huge progress has been achieved in the extension of the Lattice Boltzmann method to relativistic and quantum systems: New families of polynomials for Jüttner, Fermi-Dirac and Bose-Einstein distributions were introduced and ultra-relativistic simulations became possible. Additionally the formalism was generalized to arbitrarily curved spaces by including the spatial metric into a covariant formulation. Subsequently for the first time a Lattice Wigner model was developed. As unexpected spin-off, a novel DFT method to calculate molecular structures by solving the Kohn-Sham equation with Lattice Boltzmann was invented and tested, comparing in its first versions already quite favorably in performance with today’s standard methods. Lattice Boltzmann simulations were used to gain insight about the temporal sequence of solar flares, study instabilities of relativistic gases, modelling debris flow and simulate the flow through “campylotic”, i.e. randomly curved spaces. Particularly enlightening was the application to graphene, since its electronic flow can be described as a relativistic fluid.
Another novelty was the implementation of strongly deformable boundaries using subdivision shell elements, leading to the discovery of a yet unknown morphology of membranes deformed by wires, which we called the “warped” morphology. A new type of fluid structure FEM-LBM coupling was implemented by considering wall deformations through changes in the spatial metric of the fluid solver which can for instance simulate waving flags.