Validating Rotorcraft Simulations with Improved Wind Tunnel Measurements

Overview

Helicopters and tiltrotor aircraft provide many important civil and military services due to their ability to take off and land vertically without the need for an airport. To help develop more efficient rotorcraft designs, engineers need to better understand the complex airflow interactions caused by the spinning rotor. Each rotor blade produces tip vortices that interact with other blades and vortices, the fuselage, and other vehicle components. These interactions can affect the rotorcraft's performance and produce undesirable levels of noise and fuel burn.

Simulating moving rotor geometries and their complex flow fields is a very challenging computational problem. It is particularly difficult to accurately maintain the vortex wake without artificially diffusing the vortices and their influence on the flow field. We utilize a multidisciplinary approach that combines high-fidelity computational fluid dynamics (CFD) to model the flow field, computational structural dynamics (CSD) to model highly flexible rotor blades, and a trim algorithm to prescribe the blade motions for trimmed level flight. We are performing high-fidelity simulations of experimental rotor hover tests in order to improve and validate the accuracy of these computational models and design better experiments.

Project Details

Recent advancements in CFD simulation accuracy have achieved an order-of-magnitude improvement in hover performance prediction, and a 50% improvement in forward-flight aerodynamic loads prediction. This was accomplished through advancements in algorithm accuracy and turbulence modeling. One of the goals of NASA's Rotary Wing Project is to develop accurate multidisciplinary design analysis and optimization (MDAO) tools for new civil rotorcraft concepts using high-fidelity CFD codes. Validation of modern CFD codes with experimental data continues to be a pacing item in future rotorcraft MDAO.

The Rotary Wing Project is planning a new hover experiment to validate modern CFD codes. Higher levels of measurement accuracy are needed to validate the improved accuracy of CFD rotorcraft simulation. The primary blade performance parameter, figure-of-merit, can be difficult to measure due to its sensitivity to flow conditions and test facility interference. A 2% experimental error is not uncommon, yet recent CFD results have agreed with measured data to within 0.2%. High-resolution measurements of the vortex wake position, strength, and vortex core size are also needed to validate modern CFD codes.

Results and Impact

Particle image velocimetry (PIV) measurements from a planned hover experiment will help quantify the accuracy of CFD wake predictions. We have carried out a number of pre-test hover computations to better understand how different test facilities affect the rotor's performance and vortex wake geometries. These pre-test computations will help design a better experiment.

We have also performed detailed forward-flight simulations of a rotor with fuselage within the National Full-scale Aerodynamics Complex (NFAC) at NASA Ames Research Center. The goal is to better understand how the NFAC's 40-by-80-ft test section and test stand affect the measured rotor performance, vortex wake properties, and vortex interaction with the vehicle's fuselage. Overall, the fuselage had little affect on the rotor performance, i.e., the blade sectional normal forces and pitching moments. The fuselage was found to displace the rotor vortices somewhat, as compared to an isolated rotor without a fuselage. Finally, the test apparatus was shown to have little effect on the blade performance and vortex wake location.

Why HPC Matters

These flow simulations were made possible by NASA's Pleiades supercomputer. Solutions typically required 1,024 – 2,048 cores running for 1 – 2 weeks, with 5 – 24 hours of runtime per rotor revolution. One particularly large simulation case took 2 months to run using 2,048 Sandy Bridge cores on Pleiades. Computational model grid sizes varied from 150 – 500 million grid points, and generating movies of the unsteady flow simulations took up to 100 terabytes of disk space.

What's Next

The OVERFLOW CFD code will continue to be validated with existing and future high-resolution wind tunnel measurements. It will also be placed in an MDAO framework to examine and optimize future rotorcraft concepts.