How to best describe a classical fluid? And moreover why does it
present an interesting subject of study?
The term fluid incorporates both gases and liquids and implies a
disordered state of matter. Real fluids are nontrivial many-body
systems due to their complex interparticle interactions and the
resulting collective motion. Detailed theoretical description on the
microscopic level is thus difficult. However it is possible to capture
and predict the central aspects of their behaviour by using more
simple model fluids which focus on the essential physical features
of the system. A couple of such examples are a system of purely
hard-spheres (that already enables study of packing effects) or a
fluid of Lennard-Jones particles.
This latter case is interesting in equilibrium study since particles
exhibiting a combination of attraction and repulsion in their
interaction potential can undergo a phase transition. This then
leads to the coexistence of liquid and gas phases. Starting from a
fundamental idea of van der Waals, namely that the packing
pattern of particles is mostly dominated by the repulsive part of
their pair potential while the attractive part can be regarded as a
perturbation, we will show how to treat accurately such systems
using modern methods of statistical mechanics. One of the main
results presented will be a study of the liquid-gas interface which,
despite being a commonplace phenomenon, remains a
controversial `hot topic' in the field.
For the second part of the presentation we will leave the world of
equilibrium physics and consider the nonequilibrium dynamics of
classical fluids for which the particles undergo Brownian motion.
We will focus on hard-spheres in three dimensions and investigate
the collective dynamics of the system in response to various timedependent external potentials. The main point of interest is that
our new theoretical approach, the so-called superadiabatic-DDFT,
captures memory effects and enables first-principles predictions to
be made, in excellent agreement with direct many-particle
simulations.