Programs in Physics & Physical Chemistry
|[Licence| Download | New Version Template] aeel_v1_0.tar.gz(1914 Kbytes)|
|Manuscript Title: FILMPAR: A parallel algorithm designed for the efficient and accurate computation of thin film flow on functional surfaces containing micro-structure|
|Authors: Y C Lee, H M Thompson, P H Gaskell|
|Program title: FILMPAR|
|Catalogue identifier: AEEL_v1_0|
Distribution format: tar.gz
|Journal reference: Comput. Phys. Commun. 180(2009)2634|
|Programming language: C++ and MPI.|
|Computer: Desktop, server.|
|Operating system: Unix/Linux Mac OS X.|
|Has the code been vectorised or parallelized?: Yes. Tested with up to 128 processors|
|RAM: 512 MBytes|
|Keywords: Multigrid, parallel computing, thin film flow, lubrication equations.|
|PACS: 47.11.Bc, 47.15.gm.|
External routines: GNU C/C++, MPI
Nature of problem:
Thin film flows over functional substrates containing well-defined single and complex topographical features are of enormous significance, having a wide variety of engineering, industrial and physical applications. However, despite recent modelling advances, the accurate numerical solution of the equations governing such problems is still at a relatively early stage. Indeed, recent studies employing a simplifying long-wave approximation have shown that highly efficient numerical methods are necessary to solve the resulting lubrication equations in order to achieve the level of grid resolution required to accurately capture the effects of micro- and nano-scale topographical features.
A portable parallel multigrid algorithm has been developed for the above purpose, for the particular case of flow over submerged topographical features. Within the multigrid framework adopted, a W-cycle is used to accelerate convergence in respect of the time dependent nature of the problem, with relaxation sweeps performed using a fixed number of pre- and post- Red-Black Gauss-Seidel Newton iterations. In addition, the algorithm incorporates automatic adaptive time-stepping to avoid the computational expense associated with repeated time-step failure.
1.31 minutes using 128 processors on BlueGene/P with a problem size of over 16.7 million nodes.
|Disclaimer | ScienceDirect | CPC Journal | CPC | QUB|