Dresden 2026 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 41: Poster: Nonlinear Dynamics, Granular Matter, and Machine Learning

DY 41.16: Poster

Mittwoch, 11. März 2026, 15:00–18:00, P5

Computational modeling and design of self-stratifying colloidal materials — •Mayukh Kundu and Michael Howard — Auburn University, Auburn, United States

Mixtures of colloidal particles suspended in a solvent can spontaneously form layered structures during fast solvent drying. This process, called self-stratification, can be leveraged to fabricate multilayered colloidal materials in a single processing step. Existing models for simulating self-stratification are computationally expensive or inaccurate. I have developed a better model for simulating the phenomena using dynamic density functional theory (DDFT). DDFT is a continuum model that is systematically formulated from particle-level interactions and dynamics. As such, it incorporates physics that would be present in particle-based simulations but can access much larger length scales and longer time scales. DDFT has two key inputs: a thermodynamic model (free-energy functional) and a dynamics model (mobility tensor). DDFT model can be made faster using the simplest approximations of these inputs that give the desired accuracy. I systematically investigated approximations of both inputs to develop an accurate, efficient DDFT model for drying suspensions. I also coupled these drying simulations to an optimization strategy based on surrogate modeling to inverse design self-stratified coatings with targeted thickness and particle distribution. This work has the potential to reduce the time and resources required to create these novel materials in the laboratory.

Keywords: statistical physics; hydrodynamics; colloids; drying; machine learning