Development of an unstructured-grid based PNS/NS CFD solver provides advances in unstructured-grid technology to provide a more efficient analysis tool. Over a wide range of the missile flight trajectories, a PNS methodology is sufficient and a more cost effective approach to provide CFD analysis capability than a traditional NS approach. Overall memory and the CPU requirement can be reduced to provide a predictive CFD capability running on modern desktop platforms. With the latest announcement by Intel Corporation to provide 64-bit PC architecture, traditional expensive hardware requirements could be replaced with a more cost-effective class of PC desktops. The technical challenge is to merge the grid generation process with a hybrid PNS/NS CFD solver. The goal of this SBIR effort is to develop necessary techniques required in integrating grid generation and the data structure topology to the flow solver. Under a Phase I effort, a hybrid PNS/NS solver will be developed from an existing NS solver, and a 2D grid generator will be integrated into the flow solver. With a continuation of Phase II effort, the method developed will be extended to a 3D flow solver. With the software developed in object-oriented approach, other physical models such as chemically reacting flows, turbulent flows, multi-phase flows could all be added as a module to the developed baseline capability. Anticipated Benefits/Commercial Applications: The method and the resulting software as proposed by this effort will have cost saving benefits by reducing the hardware costs and by reducing the costs associated with the simulation effort. Also, the resulting software will be available as a module to existing commercial vendors to provide an integrated toolkit from grid generation to flow analysis to visualization within a single framework.
Keywords: CFD, hybrid PNS/NS, unstructured grid
