You are here

Surface Mesh Refinement Guide Tool for Computational Fluid Dynamics Applications


OBJECTIVE: To develop a fluid dynamics-based diagnostic tool to guide the initial creation of computationally economical surface grids for computational fluid dynamics simulations of military aircraft, missiles and associated subcomponents. DESCRIPTION: For all of the continuous development in the field of Computational Fluid Dynamics (CFD), a certain degree of art is still required to obtain numerical solutions that can be critically judged to conform to physical reality. Much effort has been spent in creating technology that can automatically create computational meshes from geometry output by Computer-Aided Design/Computer-Aided Manufacturing (CAD/CAM) software [1-4]. These methods largely define quality in terms of the conformance of the mesh to the surface, and the geometric quality of the constituent cells of the mesh. However, these qualities are not the only requirements that a surface mesh must satisfy. The mesh density of the nodes representing the surface must also be sufficient to accurately capture gradients in the flow field near the surface [5-7]. Too coarse of a mesh causes sharp gradients in the flow field to suffer from numerical diffusion, where sudden changes in fluid properties or behavior are unrealistically"smeared"[8]. Too many points, and the time and computational expense required to achieve a converged CFD solution increase geometrically, if not exponentially, to the point where solution is impractical. This balance between physical fidelity and computational expense is a problem that CFD practitioners usually must address though intuition and experience. Current grid creation and refinement technologies require iterative CFD calculations to produce an acceptable surface mesh [9]. For complex geometries, such as those associated with military aircraft and missiles, this involves running test simulations that can require days to run on hundreds or thousands of computer processors in a modern supercomputer. For a sufficiently complex CFD model, additional time on the order of man-weeks may be required to manually refine surface meshes over geometrically complex features during successive iterations. Thus, a significant amount of time, effort and computational resources are required to complete aerodynamic analyses. For production engineering applications, time and available computational hardware may force the"acceptable"mesh to be of significantly lower quality than the optimum mesh, resulting in increased computational cost with lower solution fidelity [7]. The objective of this topic is to advance the state-of-the-art in initial surface mesh generation for CFD simulations of military aircraft and missiles of complex topography. The desired ultimate end-product would be a software tool that could use lower-order aerodynamic methods to identify regions on surfaces where surface mesh grid point density should be concentrated to capture significant features in the near flow field, while also identifying regions of benign flow where surface grid point density could be lessened to save computational expense. As most Army aircraft and missiles operate in the high Reynolds number regime, it is possible that the off-body flowfield could be assessed using inviscid methods, such as panel codes or vortex lattice methods, to estimate pressure gradients and local flow direction, while boundary layer methods could be used to estimate gradients within the boundary layer, separation bubbles and wakes. The results from this tool"s analysis could then be output in a format that could be visualized to guide CFD mesh creation and refinement. PHASE I: Identify innovative methods to guide the initial creation of computationally economical surface grids for computational fluid dynamics simulations of military aircraft, missiles, and associated subcomponents, based upon input of triangular or quadrilateral surface tessellations of the wetted outer mold line geometry. Develop a plan to implement these methodologies into a software tool. Provide preliminary verification and validation approaches to support the activity. Identify example surface geometries and computational and experimental data sets that will be used for verification and validation of tool operation. PHASE II: Develop a stand-alone software tool implementing the methodologies identified in Phase I. Demonstrate levels of computational efficiency and accuracy improvements in aerodynamic lift, drag and moment estimation gained by using CFD surface meshes developed from guidance produced by the tool versus results obtained by grids created using current, standard practice methods. International Traffic in Arms Regulation (ITAR) control is required. PHASE III: If successful, the product produced from Phase II will be a stand-alone tool that can guide the initial creation of computationally economical surface grids for computational fluid dynamics simulations of complete military aircraft, missiles, and wetted external components. The next, logical step for this capability would be to integrate the methodology into grid generation software to automate the creation of optimized meshes during the initial grid generation process. In addition to military uses, the resulting tools would be of value to civilian CFD practitioners. The time required to create grids for aerospace CFD simulations would be reduced, while the accuracy of the computed solutions would be increased by concentrating grid density in regions of most interesting flow behavior. This would allow the CFD solutions to be used as the basis for total aircraft aerodynamic performance estimation, estimation of engine power required, and to provide aerodynamic force and moment inputs to other engineering disciplinary tools such as computational structural mechanics and dynamics analysis software packages. REFERENCES: 1. Petersson , N.A.,"A Software Demonstration of'rap': Preparing CAD Geometries for Overlapping Grid Generation,"UCRL-JC-147260, Proceedings of the 8th International Conference on Numerical Grid Generation in Computational Field Simulations, International Society for Grid Generation , Birmingham, AL, Jun. 2002. 2. Haimes, R.,"Final Report for NAG1-02040, CAPRI: A Geometric Foundation for Computational Analysis and Design,"Document ID 20070022429, National Aeronautics and Space Administration, NASA Langley Research Center, VA, Oct. 2006. 3. Tomac, M., Rizzi, A., and Oppelstrup, J.,"From Geometry to CFD Grids An Automated Approach for Conceptual Design,"AIAA 2010-8240, American Institute of Aeronautics and Astronautics, Reston, VA, Aug. 2010. 4. Chan, W.M.,"Developments in Strategies and Software Tools for Overset Structured Grid Generation and Connectivity,"AIAA 2011-3051, American Institute of Aeronautics and Astronautics, Reston VA, Jun. 2011. 5. Knupp, P.,"Remarks on Mesh Quality,"AIAA 2008-933, American Institute of Aeronautics and Astronautics, Reston, VA, Jan. 2008. 6. Luke, S., Hebert, S., and Thompson, D.,"Theoretical and Practical Evaluation of Solver-Specific Mesh Quality (Invited),"AIAA 2008-934, American Institute of Aeronautics and Astronautics, Jan. 2008. 7. Thornburg, H.J.,"Overview of the PETTT Workshop on Mesh Quality/Resolution, Practice, Current Research, and Future Directions,"AIAA 2012-0606, American Institute of Aeronautics and Astronautics, Reston, VA, Jan. 2012. 8. Roy, C.J.,"Review of Discretization Error Estimators in Scientific Computing,"AIAA 2010-126, American Institute of Aeronautics and Astronautics, Reston, VA, Jan 2010. 9. Shih, A., Ito, Y., Koomullil, R., Kasmai, T., Jankun-Kelly, M., Thompson, D., and Brewer, D.,"Solution Adaptive Mesh Generation using Feature-Aligned Embedded Surface Meshes,"AIAA 2007-558, American Institute of Aeronautics and Astronautics, Reston, VA, Jan. 2007.
US Flag An Official Website of the United States Government