Exploiting parallelism in the solution of scientific and computational engineering problems requires significant expertise and effort on the part of application developers. We describe a framework targeted at the class of discrete-time grid simulations that hides much of the complexity inherent in t
Elements of a Computational Framework for Coupled Simulations
✍ Scribed by Hermann G. Matthies; Rainer Niekamp; Martin Hautefeuille; Dominik Jürgens; Christophe Kassiotis; Tarin Srisupattarawanit; Jean-Baptiste Colliat; Adnan Ibrahimbegović
- Publisher
- John Wiley and Sons
- Year
- 2010
- Tongue
- English
- Weight
- 71 KB
- Volume
- 33
- Category
- Article
- ISSN
- 0936-7195
No coin nor oath required. For personal study only.
✦ Synopsis
Abstract
Coupled problems arise because certain aspects of some systems have previously been modelled separately, assuming (implicitly) an uncoupled behaviour. More often then not, computational approaches and mature software have been developed for each of these aspects in isolation. If one has to consider now the coupled problem, it would be advantageous if this previous work could be put to good use. This, in particular, concerns the developed software. In this presentation we will analyse the requirements for this to be achieved, and outline some of the possible solutions. Several tasks have to be accomplished: Non‐matching spatial grids and geometric representations have to be joined, fulfilling certain consistency requirements. Similarly non‐matching time‐stepping schemes have tobe combined. After outlining a possible mathematical formulation for the coupling, the joining of the software will be considered. We view the original software which is modelling some aspect as a software‐engineering component, and we describe a component‐based approach to achieve a consistently coupled and (if so desired) distributed computational simulation (© 2010 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)
📜 SIMILAR VOLUMES
## Abstract Peer‐to‐peer paradigm is more and more studied by the distributed computing community. Indeed, this type of architecture has interesting properties like the absence of centralized topology, fault tolerance or dynamic reorganization of the network. However, managing these networks is com