We are interested in the design and implementation of synthetic computational environments that facilitate simulation-based critical activities such as design, decision support, qualification, and certification related to the development, management, and maintenance of large physical systems. Such systems can extend from load-bearing complex structural systems to disaster-sensitive environmental or other multiphysics systems. Essential to this effort is the need for reducing computational time via the development of reduced order and surrogate models, along with domain and time parallelism implementations of the algorithms involved in solving systems of coupled partial differential equations describing the physics of the systems under investigation. A combination of heterogeneous hardware infrastructure extending from hand-held portable devices, to multi-core CPU and GPU architectures, to the computational GRID, along with that of the corresponding software that includes mobile agents and web/GRID services along with legacy applications, suggest a highly customizable meta-architecture and implementation of computational environments. The compositionality of aggregate computational units becomes a highly desirable property for visual-languages-oriented application synthesis environments.

 

key words

Compositional simulation; Data-driven simulation; Reduced order models; Surrogate models; Grid services; Heterogeneous computational infrastructure; Software agents;

Citizenship:  Open to U.S. citizens and permanent residents

Level:  Open to Postdoctoral applicants