Computational Aeroacoustic Open-Rotor Simulations for Green Aviation

Overview

Over the last few years, there has been renewed interest in pursuing contra-rotating, open-rotor propulsion technology due to its potential to significantly reduce aircraft fuel consumption. A main concern for the design of such systems is that they must meet community and cabin noise standards. Hence, reliable noise prediction capabilities are essential to enable novel and fuel-efficient open-rotor systems. Immersed boundary and structured curvilinear grid approaches have reached a level of maturity to where they are now frequently employed for specific real-world applications within NASA, which makes these noise prediction models more reliable tools for aircraft design.

Project Details

In recent years, NASA initiated several research efforts that have assessed the agency's capabilities for predicting noise source and for analyzing contra-rotating open rotors. This project demonstrates that our higher-order methods provide the ability to perform aeroacoustic analysis of wake-dominated flow fields. One of the key challenges of these models is to conduct efficient high-fidelity simulations of the relatively complex geometry. Our first-of-a-kind aeroacoustic simulation employs an immersed boundary method for an open-rotor application. In addition to applying the immersed boundary method to this moving boundary problem, we conducted a detailed aeroacoustic analysis of the noise-generation mechanisms encountered in the open-rotor flow.

To gain confidence in our approach, we compared the simulation data—both immersed boundary and curvilinear-based— to available experimental wind tunnel data. The noise-generation mechanisms were analyzed by employing spectral analysis, proper orthogonal decomposition, and the causality method. Several low- to high-fidelity simulation tools were used to provide noise predictions, and their results were compared to the experimental data.

Results and Impact

We employed higher-order accurate immersed boundary and structured curvilinear methods to simulate the highly complex flow field around an open rotor. Three key tasks were performed:

• Simulating the open rotor with our novel higher-order immersed boundary method.
• Simulating the open rotor with our boundary layer resolving, body-fitted structured curvilinear approach.
• Analyzing in detail the most dominant noise-generation mechanisms of the open-rotor propulsion system.

Several postprocessing tools (such as Fourier transforms, proper orthogonal decomposition, and the causality method) were used to obtain a detailed understanding of the noise-generation process. We found that a wind tunnel sensor plate contributed significantly to the broadband noise due to interactions with the rotors and hub. Ongoing work includes mesh refinement studies, and simulations of a low-speed case. Through these higher-order accurate simulations, NASA continues to advance the state-of-the-art in computational fluid dynamics (CFD) simulations for green aviation.

Why HPC Matters

Using NASA's Pleiades supercomputer, the simulations consumed approximately three weeks of wall-clock time on 500 – 2,000 Intel Haswell cores. These simulations would not be possible without the use of a large cluster like Pleiades. The visualization team at the NASA Advanced Supercomputing (NAS) facility helped generate ultra-high-quality renderings of the open-rotor flow physics.

More Media

See additional images/videos related to this demo in our media gallery.