Share |

Revolutionizing rotorcraft design with CFD

Italian polymath Leonardo da Vinci provided an early blueprint for what eventually became the modern helicopter. His sketch, dated 1493, details a platform and a helical screw surrounded by what he called stiff linen. Da Vinci predicted that turning the contraption sharply would cause it to rise into the air in a spiral.

Leonardo da Vinci sketch. Helical screw.

Although the French and Germans built and piloted helicopters as early as 1907, none of their designs were as commercially viable as that of Russian born designer Igor Sikorsky. Sikorsky’s 1940 VS-300 could safely fly in six directions, which made it the primary model for all modern-day rotorcraft. To maneuver, rotary wing aircraft (or rotorcraft) use lift created by rotor blades revolving around a central mast. Several rotor blades surrounding a single mast are called a rotor.

Rotorcraft include helicopters, cyclopters, autogyros, and gyrodynes. Like their fixed-wing counterparts, these aircraft experience exposure to unpredictable weather conditions, shear winds, and turbulence. When high velocity air meets slower moving air, clear air turbulence (CAT) often poses a significant safety threat – especially because it is undetectable by conventional radar. This meeting of air occurs naturally at very high altitudes and in mountainous regions. CAT can also form when slow moving air comes in contact with high-speed air around a helicopter’s rotor.

Helicopters also face exposure to other turbulence, such as self-generated vortices. (Think of vortices as a spinning motion around an imagined axis, similar to spinning whirlpools of water formed by boat paddles. In this case, helicopter rotor blades are the paddles, and water is replaced by air.) An axis occurs at each blade tip of the rotor, and the spinning of each blade creates vortices. This phenomenon, called blade vortex interaction (BVI), happens repeatedly as rotating blades create their own tip vortices and intersect with vortices created by other blades.

BVI is difficult to predict using computational fluid dynamics (CFD). However, researcher's from NASA's Rotary Wing project are developing physics-based CFD tools to enhance the predictive accuracy of flow fields.  Comprehensive codes previously under-predicted the figure of merit (FM), a key hover performance parameter, by anywhere from two to six per cent. “These percentages seem small but their impact is huge,” says Neal Chaderjian, a research scientist at NASA’s Ames Research Center. Chaderjian was on hand at the SC12 conference in Salt Lake City, Utah, to discuss and explain the breakthrough research.

Overhead view of V22 rotor blade tip vortices and turbulent worms. Image courtesy Neal Chaderjian.

“My job is to advance the numerical prediction capability and accuracy of rotorcraft aeromechanics using CFD methods. Aeromechanics combines aerodynamics – flow fields generated by rotorcraft – and structures, as most rotor blades bend and twist, which affects aerodynamics.  The current engineering approach uses comprehensive codes (simplified, low-order flow models) and CFD (high-order, first-principles flow simulations).  Rotorcraft design relies more on the comprehensive codes because they can get answers in one minute or less, while CFD can take hours to days to obtain high-fidelity solutions.” 

Chaderjian has improved current CFD code to solve the Navier-Stokes equation of motion for fluids, which is both time-dependent and intensive. “Accurately resolving the viscous flow regions on the rotor blades requires high spatial resolution and time accuracy,” he said. Other improvements include adaptive mesh refinement (AMR) to resolve rotor wakes. “The AMR improvement allows the vortices and rotor wake to be resolved by grids four times as fine,” Chaderjian explains. During simulation, grid sizes varied from 60 to 750 million grid points. Visit NASA's photostream on Flickr to see a full-motion simulation of a UH-60 helicopter rotor, and the same helicopter's rotor vortex wake.

This additional efficiency and accuracy enable Chaderjian to demonstrate fine-grain details of blade vortices and turbulent flows never seen before, and for the first time, the FM is predicted to within the experimental error range for V22 Osprey tiltrotor and UH-60 Black Hawk helicopter rotors in hover (recall: prior to these advancements the FM under-predictions were two to six percent). To put these improvements into perspective, every half-percent deviation the FM is from error, one fewer person is able to safely occupy the rotorcraft – we are witnessing tremendous achievement. 

V22 Osprey. Photo courtesy Michael Pereckas.

Accuracy, however, comes with a price: “This requires a supercomputer with large computer memory, disk space, and many processors to obtain solutions within a few days to a week or more.  Anything less limits the state-of-the-art accuracy of the flow simulations,” noted Chaderjian. The flow simulations were made possible by NASA’s Pleiades supercomputer. Solutions required around 1,500-4,600 cores running for two to three weeks.

High-resolution CFD flow simulations more accurately reflect rotorcraft aeromechanics, and improve design prediction capabilities. This ultimately reduces the number of expensive wind tunnel and flight test experiments. Because the research is cutting edge, Chaderjian predicts computational methods developed by NASA’s rotary wing researchers will take five to ten years to make their way into the industrial design process. “The end result will be a rotorcraft design that has improved performance, along with reduced production, flight, and maintenance costs, and a smaller impact on the environment.”

Your rating: None Average: 4.8 (8 votes)