Dipartimento di Fisica
Università di Trieste
Strada Costiera 11
34151 Trieste (Italy)
Phone: +39 040 2240263
My research interests concern the physics of disordered states of matter, with particular focus on the microscopic mechanisms of glass formation and on glass structure. I am also interested in modeling the peculiar phase behavior, structure and dynamics of soft condensed matter.
My work is based on the methods of statistical physics and on computer simulations. Over the years I developed a computational approach based on reproducible research, high-performance computing and on high-level simulation frameworks like atooms, which I develop.
Recent papers and seminars
"Spatial structure of unstable normal modes in a glass-forming liquid"
M. Shimada, D. Coslovich, H. Mizuno, A. Ikeda, SciPost Physics 10, 001 (2021)
"Assessing the structural heterogeneity of supercooled liquids through community inference"
J. Paret, R. Jack, D. Coslovich, The Journal of Chemical Physics 152, 144502 (2020)
"A localization transition underlies the mode-coupling crossover of glasses"
D. Coslovich, A. Ninarello, L. Berthier, SciPost Physics 7, 077 (2019)
"Dynamic and thermodynamic crossover scenarios in the Kob-Andersen mixture: Insights from multi-CPU and multi-GPU simulations"
D. Coslovich, M. Ozawa, W. Kob, The European Physical Journal E 41, 62 (2018)
"Glass structure through the prism of clustering"
Glassy Systems and Inter-Disciplinary Applications, Institut d'Etudes Scientifiques de Cargèse (France), 2021
"Distributional clustering approach to the heterogeneity of supercooled liquids"
Digital meeting - Recent advances on the glass problem, CECAM (Switzerland), 2021
"Statistical inference of structural communities in supercooled liquids"
Laboratoire de Physiques de Solides, Université Paris-Sud, Paris (France), 2020
Meeting of the Simons collaboration "Cracking the glass problem", Royaumont (France), 2019
The codes below are available on my [git repository](https://framagit.org/coslo)
I developed a python framework called atooms to perform computer simulations and analyze their results. It provides an expressive, high-level interface to the main objects of particle simulations.
Atooms is a collection of python packages that provide a high-level, yet efficient framework to deal with particle-based simulations, such as molecular dynamics or Monte Carlo. It is composed by a base library and additional packages that implement complex simulation strategies or analysis tools.
Tutorial » Notebook » Public API »
Atooms-pt is the first simulation package I built on top of the atooms framework. It implements a multi-GPU parallel tempering simulation and relies on RUMD, an efficient molecular dynamics code developed by Glass and Time at the University of Roskilde.
The atooms-pp package provides python tools to compute static and dynamic correlation functions from particle-based simulation data.
Francesco Turci contributed a jupyter notebook showing how to compute static and dynamic correlations of a Lennard-Jones mixture using the postprocessing package.Tutorial » Notebook »
Transition path sampling
The atooms-tps package, developed in collaboration with Francesco Turci, provides a generic frontend to transition path sampling simulations, which allow to sample rare fluctuations in the trajectory space of a dynamical system.
Reproducible research and data
I use the zenodo data repository to store citeable datasets and workflow associated to my research papers, as well as code snapshots. Here are a few recent data sets of mine
- Dataset and workflow for "A localization transition underlies the mode-coupling crossover of glasses" arXiv:1811.03171 (2019)
- Dataset for "Dynamic and thermodynamic crossover scenarios in the Kob-Andersen mixture: Insights from multi-CPU and multi-GPU simulations" Eur. Phys. J. E 62, 41 (2018)
- Dataset for "Local order and crystallization of dense polydisperse hard spheres" J. Phys.: Condens. Matter 30, 144004 (2018)