KITWARE INC — Department of Energy SBIR Phase II: The manufacturing and engineering industry needs more cost effective and easy-to-use high-

Description

The manufacturing and engineering industry needs more cost effective and easy-to-use high- performance computing (HPC) modeling and simulation tools to design, test, and analyze innovative components and products. For 92% of the small- to medium-sized manufacturers with less than 100 employees, an HPC system is most likely not part of their development environment. This class of manufacturers typically relies on two-dimensional computer-aided design software and a limited amount of desktop modeling and simulation software. This under-utilization of HPC tools by this class of manufactures is of chief concern to the Department of Energy. The proposed simulation framework and suite of tools will be developed with one eye on eliminating the barriers to adoption of HPC modeling and simulation tools for manufacturing and engineering, and the other eye on increasing the efficiency and effectiveness of manufacturing and engineering practices in solving real-world problems. By using these tools, manufacturers will be able to improve their design cycles and reduce their need of developing expensive physical prototypes. In Phase I, a framework referred to as the Simulation Model Based Architecture was developed, capable of supporting the entire lifecycle of a simulation. The approach used was to leverage existing open-source toolkits wherever possible so that the resulting system could also be distributed as open-source. The system provided several mechanisms by which HPC systems can be leveraged. In Phase II, the major focus will be improving the ease of use of the system and extending its core functionality. To address ease of use, the customization capabilities will be enhanced and workflow wizards will be developed to guide the user through their specific simulation workflow. In terms of core functionality, the modeling and meshing capabilities will be enhanced and the mechanisms for data exchange with various simulation environments will be improved. In addition, more complex workflows such as design optimization will be supported. Modular frameworks such as the one proposed here lend themselves to a service business model; vendors with the skills to provide technology integration services can partner with their customers and collaborators to build valuable, competitive products and services. Such an open, modular framework provides significant business opportunities. Beyond the manufacturing community, the technology developed here will also benefit the larger scientific computing community since the code will be released under open source licenses (non-reciprocal Apache or BSD) within open source communities. Thus researchers, educators, and commercial enterprises will be able tune the suite to their particular workflow, easily research specialized areas of interest, and leverage international communities to develop leading edge technology.