Parallel Geometry-Based Mesh Adaptation for Supersonic Jet Acoustic Analysis
As interest in supersonic transport grows, mitigating noise during critical flight phases such as landing and takeoff remains a significant challenge. NASA’s Commercial Supersonic Technology (CST) project aims to address this issue by developing advanced propulsion systems that exceed current noise regulations. A key aspect of this research focuses on dual-stream nozzles, which enhance propulsion efficiency while reducing noise. However, the complex geometries of these nozzles create meshing challenges that affect simulation accuracy and computational time. To overcome these obstacles, the Launch, Ascent, and Vehicle Aerodynamics (LAVA) team at NASA’s Ames Research Center introduced a novel MPI/OpenMP parallel geometry-based mesh adaptation algorithm that provides resolution of complex geometric boundaries without adding degrees of freedom. This technology has proven to simplify the grid generation process required for complex nozzle configurations studied as a part of the CST project. These simplifications maintain accuracy in acoustic predictions compared to traditional meshing strategies. This new technology greatly reduces the time-to-solution, enabling faster high-fidelity acoustic analysis for a variety of complex nozzle configurations.
Developing a novel adaptation strategy focused on reducing memory usage and improving computational efficiency was made possible with NASA's supercomputing resources. These innovations in mesh generation significantly reduce manual labor while maintaining the acoustic prediction accuracy crucial for the success of NASA’s CST project.
Chase Ashby,
NASA Ames Research Center
Quick Facts
- The LAVA team utilized NASA's Aitken supercomputer cluster to develop and test their MPI/OpenMP parallel mesh adaptation algorithm, optimizing performance with Intel’s VTUNE profiler and conducting jet acoustic simulations for NASA's CST project.
- The new meshing strategy eliminated the 40% reduction in time-step size and a 15% increase in mesh count resulting from a traditional body-fitted meshing for lobed mixers in jet nozzle geometries.
- These results will improve the efficiency of noise reduction simulations and streamline the design process for quieter supersonic aircraft propulsion systems.
The next steps involve integrating geometry-based and adjoint-based adaptation techniques to optimize the mesh for key quantities of interest.
Researchers
- Chase Ashby, NASA Ames Research Center Please enable Javascript to see email address.
- Jeffrey Housman, NASA Ames Research Center Please enable Javascript to see email address.
