Software reuse methodologies for parallel and netcentric scientific computing
Abstract
The process of prototyping is part of every scientific inquiry, product design, and learning activity. New economic realities require the rapid prototyping of physical artifacts and rapid solutions to problems with numerous interrelated elements. This, in turn, requires the fast, accurate simulation of physical processes and design optimization using knowledge and computational models from multiple disciplines in science and engineering. Thus, the realization of rapid multidisciplinary prototyping is a new grand challenge. For the realization of this scenario the natural computational resource is a “computational grid” that connects the needed distributed high performance hardware and software resources used to simulate the elements of the artifact. Our research goal is to address this scenario in the context of parallel computing and Intranet/Internet based computational grids. To achieve this goal, in this dissertation we first address the “reuse” of legacy sequential scientific software to build parallel scientific applications. Second, we present a Web/Internet based server that “reuses” scientific problem solving environment (PSE) components to virtually place their capabilities at the fingertips of any scientist and engineer. Third, we describe the design and architecture of a generic multidisciplinary PSE (MPSE) framework that enables the “reuse” of legacy scientific code via encapsulation within a network of computational agents. We present the initial realization of this framework in the design of a prototype MPSE (GasTurbnLab), for supporting simulations needed for the design of efficient gas turbine engines.
Degree
Ph.D.
Advisors
Houstis, Purdue University.
Subject Area
Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.