Performance predictor

By Joe Stumpe|February 2018

Researchers work on a computational tool to predict the performance of long, narrow wings

As the University of Michigan’s unmanned X-HALE plane takes off, its composite wings flex upward by about .85 meters, or about 30 percent of the aircraft’s half wingspan, before returning to their more horizontal position as the plane gains altitude.

Are longer, more flexible wings like these the way of the future for commercial transport aircraft? Many observers think so because of the fuel efficiency these wings promise, and in fact the trend has begun to a degree. The wings of the Airbus A350 and Boeing 787 are notable for their flexing, but even bolder and more futuristic designs will require exploring “uncharted territory for the industry’s methodologies,” says University of Michigan professor Carlos Cesnik.

He is among the researchers working to chart this new territory through the fields of flight dynamics and aeroelasticity — the science of how aerodynamic forces distort nonrigid structures and are in turn affected by them. Cesnik isn’t a fan of wing flexibility for its own sake.

Rather, it’s the fact that longer wings could cut fuel consumption in future airplanes because of their superior lift-to-drag ratios. Those wings would need to be flexible to be viable, because stiffening them would add unacceptable weight. Such wings have a high aspect ratio, a reference to their length compared to their width.

Cesnik notes that X-HALE and commercial airliners “are very different configurations. However, the flexibility problem and other issues are very similar.”

The trend in airliners is toward higher aspect ratios, but so far those ratios are not as high as that of the X-HALE. NASA has identified high-aspect ratios as a key technology for commercial air transports of the future.

Achieving higher aspect ratios would be a lot more complicated than simply attaching longer wings to a fuselage. Longer wings don’t just bend more than traditional ones. They twist and respond to flight conditions and also interact with rigid fuselages in ways that are still are not completely understood. For instance, the up-and-down motion of a plane body during turbulence might be faster than the slow bending of a long, flexible wing — or vice versa. Flexibility “can create an unstable mode that doesn’t exist with more rigid wings,” says Jessica Jones, one of Cesnik’s former grad students who now designs aircraft for Aurora Flight Sciences, a subsidiary of Boeing.

Twenty years ago, it was impossible to predict or model such motions because the available computational models relied on linear equations. So Cesnik began developing software called UM-NAST (for the University of Michigan’s Nonlinear Simulation Toolbox). Although not finished, UM-NAST is already in the hands of designers in some aerospace companies, Cesnik says.

Traditionally, the movement, or “deflection” in engineering terms, of wings has been so small that it can be analyzed with traditional calculations. “Mathematically, the problem is linear,” Cesnik says. “The relationship between the input and output of things, if you double the input, you double the output.”

But as the length and flexibility of wings increase, that linear behavior no longer applies.

“This is where the challenges are,” Cesnik says. “Computationally, the problem becomes more difficult.”

Cesnik works with grad students to explore different vehicle configurations and flight scenarios.

“It is like flying these very flexible aircraft in a virtual world,” he says.

Jones says the data needed “is usually fairly detailed — length of wings, weight of wings; the entire aircraft is defined in that way. Then we specify what speed; if there are any external disturbances like turbulence. Then we basically solve a series of equations that tells us how the aircraft system responds to these inputs.”

Designers can explore the aircraft’s envelope in the safety of this virtual world. “We fly the airplane at its limits — at very low speeds and very high speeds, subjecting the aircraft to increasingly large disturbances to study how it reacts. This can give us valuable information about vehicle stability and modes of failure before the aircraft ever takes off.”

Cesnik says UM-NAST “allows us to be aggressive in the conditions we want to simulate with relatively minor consequences. Also, “since it is a computer, we can just pick it up from where it ends and re-start it and/or figure out what was missing in the [simulated plane] and improve the design.”

As powerful as UM-NAST is, it is not yet in final form. That’s where X-HALE comes in. Cesnik’s team built and designed X-HALE (for “Experimental High-Altitude, Long-Endurance”) a decade ago. Its flights from a small runway outside Ann Arbor help UM researchers validate the predictions of their UM-MAST computer simulations, providing missing fundamental data on the stability of long-winged planes; the loads on them as they maneuver and their responses to wind gusts. X-HALE’s sensors record the shape of the wing as the aircraft flies at different speeds and trajectories. “In order to be accurate, we need experimental data to compare it with,” Cesnik says. “We start by taking computational tools and use them to design the experiment. We fly it and we collect the data and we see if the data we collect actually matches with the predictions we have done. When it doesn’t, we figure out why so that it will predict accurately next time.”

The university flies two X-HALE planes each weighing 12 kilograms with a 6-meter wingspan. A wing-boom-tail type aircraft, X-HALE is powered by five motors and carries four data collection devices.

Despite its name, X-HALE flies at a maximum altitude of about 122 meters due to legal restrictions, for up to 30 minutes at a time. The data it collects is “scaled up” to that of larger, faster, longer-flying aircraft.

X-HALE has the dents to show that it’s made hard landings “multiple times, but we never take it lightly,” Cesnik says.

“It is costly to fix physical things and takes time. However, our approach is to take ‘manageable risk’ and be aggressive in our experiments as well. It is a fine balance to keep engineering progressing.”

Research on high-aspect-ratio wings and other potential innovations is driven primarily by two factors: the desire for unmanned planes that can fly at a high altitude for a long time, and the realization that growth in commercial airline traffic requires aircraft that burn less fuel and create fewer emissions.

Currently, most commercial aircraft wings have an aspect ratio of about 9, which means the square of the wingspan is nine times larger than its area. Calculations have shown that aspect ratios of 13 to 15 can significantly improve fuel efficiency.

Increased wing span is already part of new designs, although in much smaller increments than a plane like X-HALE with its aspect ratio of 30.

Unmanned, solar-powered planes with wing lengths on the order of X-HALE have been attempted in the past. In 2015, Google’s Solara 50 crashed shortly after takeoff in New Mexico. In 2016, Facebook’s Aquila crashed 90 minutes after takeoff in Arizona. Studying the failures of those aircraft is another aspect of the research. In the crashes of both the Google and Facebook airplanes, Cesnik says, it’s believed that excessive speed led to structural deformation of the wings and ultimately failure.

Asked when he expects high-aspect-ratio wings to become mainstream, Cesnik notes that it’s starting to happen. “We keep moving forward in terms of developing new tools. The level of understanding has grown significantly, but there’s still a lot to know.”

From his perspective, he says with a laugh, “I think it’s a lifelong pursuit.”