DPG Phi
Verhandlungen
Verhandlungen
DPG

Berlin 2008 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

SYMP: Symposium Computational Soft matter physics

SYMP 1: Computational soft matter physics

SYMP 1.4: Vortrag

Donnerstag, 28. Februar 2008, 10:45–11:00, H 0105

Scale bridging by hybrid simulations in microfluidics — •Martin Hecht and Jens Harting — Institute for Computational Physics, University of Stuttgart, 70569 Stuttgart, Germany

We have developed a hybrid molecular dynamics and stochastic rotation dynamics simulation method to simulate fluid flow close to walls for appplications in microfluidics. Moreover, this approach can be applied to various problems in soft condensed matter physics. Many important applications in this field deal with suspensions, membranes, macromolecules, or reacting fluids. However, the time and length scales in these mesoscopic systems are much above those which are accessible with conventional molecular dynamics simulations. Several mesoscopic simulation methods (lattice Boltzmann, stochastic rotation dynamics, dissipative particle dynamics) have been developed, which can deal with larger system sizes and larger simulation times. The prize one has to pay, to reduce the computational cost, is a simplification of the underlying processes on the molecular scale. For some applications, the details on the molecular length scale are essential in some regions of the system, whereas in other regions the resolution may be reduced. In this case, hybrid simulation techniques, which use different simulation methods for different regions of the system, are a possibility to minimize the computational costs of simulations under the constraint that in some regions the full physics has to be taken into account. In this talk we present our approach and first simulation results and discuss the key points of hybrid simulation approaches in general.

100% | Mobil-Ansicht | English Version | Kontakt/Impressum/Datenschutz
DPG-Physik > DPG-Verhandlungen > 2008 > Berlin