Institut Charles Sadron News
Titre : Viscoleastic memory functions in glassy and polymer liquids
Responsable : Baschnagel Jörg (TSP)
Description du projet
Viscoelastic fluids are ubiquitous. Examples include glassy liquids or polymer melts. A key property of these fluids are long-lasting dynamic memory effects correlating the flow at a given time to the history of prior motion. Theoretically, these memory effects are described by viscoelastic memory functions (VMFs) which determine the response of the fluid to deformation on various length and time scales. It is a fundamental unsolved problem of condensed matter physics to develop accurate approximation for the VMFs. Progress in developing such approximations has stagnated in the past, since theory shows that the time evolution of the VMFs is not generated by the classical Newtonian dynamics, but by an intricate ``projected'' dynamics. The projected dynamics cannot be reproduced by standard molecular dynamics (MD) simulations, thereby disabling direct calculation of the exact VMFs to guide the development of approximations. Here we suggest to close this gap by a new, constrained MD (CMD) method that mimics the projected dynamics directly in the simulation. Through two parallel doctoral theses, we seek to create the CMD method, to make it publicly available by implementation in the opensource LAMMPS code for high-performance computing, and to apply CMD to explore the spatio-temporal dependence of the VMFs for models of glassy [1] and polymer liquids [2]. [1] L. Klochko, J. Baschnagel, J. P. Wittmer, H. Meyer, O. Benzerara, A. N. Semenov, "Theory of length-scale dependent relaxation moduli and stress fluctuations in glass-forming and viscoelastic liquids", J. Chem. Phys. 156, 164505 (2022). [2] A. N. Semenov, J. Farago, H. Meyer, "Length-scale dependent relaxation shear modulus and viscoelastic hydrodynamic interactions in polymer liquids", J. Chem. Phys. 136, 244905 (2012).
Informations complémentaires
We are looking for physicists or theoretical chemists having a good knowledge of statistical physics, experience in programming (e.g., python, C++, etc.), and familiarity with LINUX. Successful candidates must have a strong interest in both theory and numerical work. Numerical work involves code development (e.g., implementing CMD in the C++ LAMMPS code, programming of data analysis tools, etc.), carrying out large-scale MD simulations, and data curation. Development of CMD and analysis of the simulation data critically hinge on the synergy provided by applying the underlying theory that also requires further extension. The positions can be filled at any time, but no later than Oct 1, 2025.
Fin de l'offre : 13/06/2025
Lien externe : https://emploi.cnrs.fr/Offres/Doctorant/UPR22-JORBAS-001/Default.aspx?lang=EN