Aeronautics

Overview

Mitigation of aircraft noise is one of the key goals of NASA’s Aeronautics Research Mission Directorate. Simulation-based prediction of airframe noise—a major factor in the total aircraft noise generated during approach and landing—is essential for designing practical noise reduction strategies that can benefit communities near major airports.

High-fidelity, simulation-based airframe noise prediction for full-scale, complete large civil aircraft in landing configuration is a grand challenge for the aerospace community. Only a few years ago, such simulations were deemed unattainable in the foreseeable future. The computational approach we used in this project was built upon the knowledge and experience gained from our previous high-fidelity aeroacoustic simulations of model- and full-scale complete Gulfstream aircraft. The present extension of this computational methodology to a full-scale landing gear on a large civil transport constitutes a promising initial attempt to meet this grand challenge.

Project Details

Under a NASA partnership with the Boeing Company on airframe noise research, we are using Exa Corporation’s Lattice Boltzmann code PowerFLOW to investigate the complex, unsteady flow field around a Boeing 777 nose landing gear, in order to obtain greater insight into the noise-producing mechanisms and identify the location of the major sources.

For large, twin-aisle civil aircraft, the undercarriage is the most prominent source of airframe noise. Due to its many subcomponents with intricate geometries, landing gear systems produce very complex flow fields that are nonlinear and highly interactive. Physics-based prediction of landing-gear noise requires a fundamental understanding of the time-dependent flow field produced by each component and the resulting interactions among these flow fields. To achieve this, we are running simulations on NASA’s Pleiades supercomputer with two goals: to accurately compute the far-field noise signature of the full-scale nose landing gear of a Boeing 777, and to validate the predicted results against measured acoustic data obtained from flight tests of the same aircraft.

Results and Impact

Our initial simulations were run on coarse- and medium-level resolution meshes to establish best practices for achieving the numerical, spatial, and temporal resolution required to properly capture the key noise-producing flow features.

Comparison of the predicted levels and frequency content of the noise spectrum on the ground with flight test data revealed good agreement. These encouraging results bode well for extending our current simulation approach from the nose gear to the more complex main landing gear of large civil transports, and then applying the methodology to the complete aircraft.

Why HPC Matters

High-fidelity simulations of the unsteady flow around complex landing gear require very large calculations that can only be run on supercomputers. Our medium-resolution simulation required 6.5 billion cells to discretize the volume surrounding the aircraft nose landing gear, fuselage, and wing. The simulation was run on 5,000 Pleiades cores and required over 1 million processor hours. To compute the far-field noise spectrum, the simulation generated over 50 terabytes of data. The combined capabilities of Pleiades and the post-processing expertise available at the NASA Advanced Supercomputing facility have enabled high resolution simulation and visualization of the prominent flow features over a broad range of spatial-temporal scales.

What’s Next

To capture the high-frequency segment of the far-field noise spectrum, we will perform the simulations on a finer spatial grid. We anticipate that these fine-resolution simulations will comprise 20 billion cells and require 4.5 million processor hours, distributed over 8,000 to 10,000 Pleiades cores.

More Information

Mehdi R. Khorrami, NASA Langley Research Center
Mehdi.R.Khorrami@nasa.gov