Research within the TSTT center has focused on providing tools and technologies to increase the levels of interoperability of mesh-based methods for the analysis of PDEs and to fill specific technology gaps needed to increase the level of automation and reliability of these simulations. An examination of the technologies needed to generate and control the transformation of information within mesh-based simulation methods clearly indicates the types of technology developments needed (click here for more information). Current TSTT efforts are considering several of these including: interoperable mesh technologies, mesh quality improvement, enhancements to mesh generation software, high-order discretization methods, and the related terascale computing issues.
- Interoperable Meshing Technologies. Our emphasis is the creation of common interfaces for existing TSTT Center technologies that will allow them to interoperate with each other to provide fundamentally increased capabilities and to allow application scientists to easily switch among them. Our focus in the first year has been on low-level query operations for mesh access, the creation of CCA-compliant mesh components, and demonstrations of TSTT tool interoperability on a one-to-one basis.
- Mesh Quality Improvement. We are creating a new,stand alone mesh quality improvement toolkit that can be used with all TSTT technologies. The primary aim of this project is to provide a freely available, comprehensive software package that would accommodate a number of different mesh element types, quality metrics, and state-of-the-art topology modification and node point movement algorithms.
- Mesh Generation Enhancement. To achieve true interoperability among the different mesh types, it is important that the mesh generation tools all operate on similar infrastructure and provide similar capabilities, and we have identified three areas that require improvement: mesh generation for high-order discretization techniques, mesh quality improvement, and mesh evolution.
- High Order Discretization. Drawing on the extensive experience of researchers in the TSTT center, we will create a Discretization Library to ease the burden of developing applications using high-order approximations. This library will support commonly used operators and boundary conditions, will be extensible to provide application specific customization, and will be independent of the underlying mesh type and therefore interoperable with all TSTT meshing technology. In the first year, we have laid the groundwork for this library by extracting existing discretization technologies from center software frameworks.
- Terascale Computing. We will develop the algorithms necessary for efficient performance on terascale architectures. Our focus is on dynamic partitioning strategies for hybrid, adaptive computations, and the use of preprocessing tools to achieve optimized single processor performance.