Nano- and micro-confined fluid flows are often characterised by non-continuum effects that require special treatment beyond the scope of conventional continuum-fluid modelling. However, if the flow system has high-aspect-ratio components (e.g. long narrow channels) the computational cost of a fully molecular-based simulation can be prohibitive. In this talk we present some important elements of a heterogeneous molecular-continuum method that exploits the various degrees of scale separation in both time and space that are very often present in these types of flows. We demonstrate the ability of these techniques to predict the flow of water in aligned carbon nanotube (CNT) membranes: the tube diameters are 1-2 nm and the tube lengths (i.e. the membrane thicknesses) are 2-6 orders of magnitude larger. We compare our results with experimental data. We also find very good agreement with experimental results for a 1 mm thick membrane that has CNTs of diameter 1.59 nm. In this case, our hybrid multiscale simulation is orders of magnitude faster than a full molecular dynamics simulation.
|Publication status||Published - 22 Nov 2015|
|Event||68th Annual Meeting of the APS Division of Fluid Dynamics - Boston, United States|
Duration: 22 Nov 2015 → 24 Nov 2015
|Conference||68th Annual Meeting of the APS Division of Fluid Dynamics|
|Period||22/11/15 → 24/11/15|
- fluid flow
- carbon nanotube