In Italiano, per favore!

A Numerical Method for Plankton Simulation
It is possible to give a mathematical description of planktonic populations through advection-reaction-diffusion equations. We devised a numerical method that, in the limit of vanishing diffusivities, recovers the pure advection-reaction problem. Further details here. (Talk given at XIV WASCOM, Scicli, July 2007.)
On the stopping time of a bouncing ball
If we assume that a bouncing ball is a rigid object, and model the impacts by introducing a restitution coefficient, we find a convergent series of the times-of-flight. That's inelastic collapse: what happens after that? (you know it, I know it, but the restitution-coefficient model won't say!). On the other hand, if the series of the times-of-flight diverges, the ball never stops, ...or does it? (Talk given at the 2007 Symmetry and Perturbation Theory conference, Otranto, May 2007.)
Avoiding Inelastic Collapse in Granular Flows
The standard model standard model for interactions among grains in granular flows leads to the phenomenon of inelastic collapse: clusters of grains undergo infinite collisions in a finite time. This is not a physical singularity, but a shortcoming of the model. By tracking the dynamics of internal vibrational modes of the grains it is possible to avoid the collapse. (Talk given at the conference "Granular Matter: Mathematical Modeling and Physical Instances", Reggio Calabria, June 2005. )
Filling Gaps in Chaotic Time Series (In Italian)
The analysis of time series is particularly hard if the series has gaps (that is, missing data) having a length greater than the characteristic predictability time of the series. If the time series is the expression of a low-dimensional dynamics ruled by a chaotic attractor, I show how it is possible to fill the gaps with surrogate data which are consistent with the observed dynamics. (Internal talk given at my department, May 2005.)
Stommel, a hydrodynamic simulation code (C program)
Stommel solves the two-dimensional Boussinesq equations for temperature and streamfunction on a staggered grid with a finite-differences scheme. It has been used to study horizontal convection. The code is freely redistributable under the GPL license.
Simulations of Horizontal Convection
Even tough horizontal convection cannot become truly turbulent, it can show time-dependent non-periodic flows. Here I show the results of some numerical simulations carried out at low Prandtl number.
The Anti-Turbulence Theorem (November 2001)
Horizontal convection (that is convection forced by a horizontal temperature gradient, rather than by a vertical one, as in the Rayleigh-Bénard case) cannot become "truly turbulent" (Paparella & Young 2002). Of course, in order to give a proof of this statement, one needs first to agree on what is "true turbulence".

Ultimo aggiornamento: 03/12/2007