Mdv: an Intelligent System for Multidisciplinary Visualization

One of the grand challenges in the Computational Aerosciences Project (CAS) of the High Performance Computing and Communications Program (HPCCP) involves performing multidisciplinary simulations in which two or more phenomena , such as uid dynamics and structural dynamics , are strongly coupled. This challenge has, in turn, given visualization researchers the challenge of developing tools to investigate the interaction eeectively. Visualization researchers must ensure that these tools are accessible to the scientist, producing visualizations easily and rapidly. They must also ensure that these representations are useful and do not contain information that might lead to misinterpretation. This paper presents the ongoing development of an intelligent, assistant-based visualization system to aid scientists in visual analysis of multidisciplinary simulations for computational aerosciences. This system is based on a data model and integrated knowledge base that supports the rapid production of visualizations. The data model addresses the issue of managing complex, heterogeneous datasets. The knowledge base contains expertise from the sci-entiic domains, visual perception and graphic design. The information from both sources is utilized by the user interface in the production of visualiza-tions based on the analysis goals of the scientist.