Next Generation Data Management of Large-Scale CFD Simulations

Large-scale, 3D simulations of complex configurations using Computational Fluid Dynamics (CFD) have become increasingly critical in the design of aircraft, aerospace vehicles and propulsion systems. The CFD analysis process consists of three phases: pre-processing, solver and post-processing. Multi-disciplinary techniques, which analyze phenomena such as fluid-structure interaction, introduce additional computational complexity and require non-trivial coupling between CFD and structures codes. In designing software for each of the three analytical phases, software developers tend to look at what is optimal and necessary for their particular operation. Considerations such as data sharing, archiving and interoperability are given less attention than efficiency and accuracy of physical models. As the pace of simulation increases, the sheer volume of numerical data requires that a stable and easy to operate methodology be made available to developers and users of pre-processing, post-processing and solver tools. The offeror proposes to create a software toolkit that provides an intelligent storage and retrieval mechanism for large-scale CFD simulation data, combining software libraries, database/compression methodologies and web browser-style control and query tools. This toolkit will assist those developing and maintaining solver codes as well as the analyst/designer community and greatly accelerate aerospace vehicle synthesis by enabling data sharing, storage efficiency and interoperability.