Abstract
In order to achieve a reasonable computing time while dealing with the fluid flow simulations through geometrically complex large physical domains such as porous media, the use of high-performance computing techniques on the large parallel computers is inevitable. Often billions of unknowns are calculated on a refined mesh that fits to the geometry to the certain measure, but not always all of them are necessary to be represented and analyzed. Depending on the engineering case, the results on a coarse scale might be significantly more representative than the fine one. To access to the differently scaled computed results, in this paper a specific approach of coupling two different scales has been adopted, macro- and micro- scale, in order to be able both to perform analysis on a rather large physical domain of porous media on the macro level without having an excessive data overhead or communication bottlenecks, so as to perform qualitative studies of a particular parameter within the micro model.
Original language | English (US) |
---|---|
Title of host publication | Proceedings of the Fourth International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering |
Publisher | Civil-Comp Press |
DOIs | |
State | Published - Mar 2 2015 |
Externally published | Yes |