Extending Airframe Noise Simulations from Scale Models to Full-Scale Aircraft

Overview

Reducing and mitigating aircraft noise in communities near airports is an important goal for NASA. To efficiently design and develop viable noise reduction technologies, engineers depend on high-fidelity simulation tools with accurate predictive capabilities. Researchers supporting the Environmentally Responsible Aviation (ERA) project—part of NASA’s Aeronautics Research Mission Directorate—are working to accelerate the development of these tools.

Noise generated by the airframe is a primary source of aircraft noise during approach and landing. Our current study is part of a comprehensive computational effort designed to advance the state of the art in airframe noise prediction from scale models to system-level, full-aircraft configurations.

Project Details

As a continuation of our previous work, which was based on an 18% scale half-span model of a Gulfstream aircraft, we have simulated airflow over a full-span, full-scale baseline aircraft in landing configuration (with wing flaps deflected 39 degrees, and the main landing gear deployed). To capture the primary airframe noise sources for this aircraft, we focused our simulations on accurately resolving the local flow fields at three locations: the flap tips, the regions near the main landing gear, and the interaction zones between these two areas.

To date, we have completed coarse- and medium-resolution computations; fine-resolution results are in progress. We used Exa Corporation’s PowerFLOW Lattice-Boltzmann flow solver to perform these time-accurate simulations. To obtain our medium-resolution results, we used approximately two billion volumetric cells to discretize the space surrounding the aircraft surfaces.

Results and Impact

Simulations of this kind present numerous challenges due to extreme geometric complexities and the wide range of spatio-temporal resolution that is required. To the best of our knowledge, the simulations performed for this study constitute the first attempt to predict airframe noise radiating from a full-scale aircraft in landing configuration with most of the geometrical details included. Through a direct comparison with acoustic measurements obtained during a 2006 flight test, we have demonstrated that high-fidelity computer simulations can accurately predict the very complex fluid dynamic processes that generate airframe noise. Our study will help NASA mature existing technologies and eventually produce new ones that will confine aircraft noise footprints within airport boundaries.

Why HPC Matters

Our full-scale aircraft simulations constitute the first step toward solving an important research challenge to reduce airframe noise. As such, these complex simulations are very resource intensive. All of the computations for this study were performed on NASA’s Pleiades supercomputer. Our medium-resolution simulations required approximately 4,000 processors and 1.5 million processor hours, and generated 40-50 terabytes of data. The rendering and visualization of these large datasets were made possible by the NASA Advanced Supercomputing (NAS) Division’s visualization team.

What's Next

In the near term, we will repeat these simulations using a finer-resolution grid in order to confirm the accuracy of the computed far-field noise at higher frequencies. We plan to subsequently extend the fine-resolution simulations to evaluate the aeroacoustic performance of several noise reduction technologies for full-scale aircraft flaps and landing gear.