Particle-based modeling on meso-scales
The aim of this school is to provide a solid training on methods which are using particles in order to model matter at the mesoscopic scale. There have been a number of significant achievements in the last years on the topic and this school will have as objective to present these techniques in a pedagogical and unified manner in the form of a week of training by lectures and hands-on exercises. The target consists of PhD students or researchers needing to get aware of those techniques in their research.
Necessity for coarse-graining (CG) molecular modeling has been often argued in materials science contexts like biomaterials or polymer dynamics : indeed very often the time and space scales of interest are out of range for a full-atom method. Besides, the details of the complete dynamics are often not relevant in the study of the process under examination. This CG necessity is even greater since experimental work is nowadays exploring more and more the lower scales, close to the micro and even nanoscale. The necessity to address the question of the minimal complexity needed to account for observed behavior at those scales is at the heart of the coarse-graining rationale.
There have been also developments of multi-scale methods in modeling techniques starting from an atomistic level : often one needs a truly microscopic modeling for a well-defined part of the system, while all the rest is acting like a fluctuating continuous background. In those techniques, the possibility to have particle-based methods for the large-scale level of description is interesting as those methods are easier to interface with truly microscopic ones, avoiding the delicate interaction between grids and particles.
The topic of this school appears very timely, given the large number of workshops dedicated to mesoscopic scales in the last years.
The content of the tutorial will associate the stochastic techniques used in statistical physics to describe the mesoscale with coarse-grained dynamics as developed in the numerical modeling community. More explicitely :
20 hours of classes
1. Theoretical foundations of mesoscale modeling (Pep Español, 2 hours)
2. Particle-based mesoscale methods (Patrick Warren, 6 hours)
2 main parts :
(1) from Langevin dynamics to Brownian Dynamics (BD), with various thermostats
(2) from BD to DPD (Dissipative Particle Dynamics)
3. Kinetic modeling (MJ Ruiz, Marisol Ripoll, Ignacio Pagonabarraga, 9 hours)
(a) theoretical foundation to kinetic theory (3 hours by MJR)
(b) application to Multiple-Particle collision dynamic (also known as SRD, stochastic rotation dynamics) 3 hours by Marisol Ripoll
(c) the Lattice-Boltzmann technique, 3 hours by Ignacio Pagonabarraga
4. Hybrid methods (R Delgado, 3 hours) : comining MD and hydrodynamics
20 hours of practicals equally divided into the following modules, self-explanatory :
1. Langevin thermostats
2. DPD code
3. MPC code
4. Lattice Boltzmann code
5. An hydrid code
Codes will be homemade, provided by lecturers who will be asked to make them as elementary as possible.
Michel Mareschal ( ULB ) - Organiser
Ignacio Pagonabarraga ( CECAM ) - Organiser