TPS students 'stay classy' during test management project Published March 11, 2015 By Rebecca Amber Staff writer EDWARDS AIR FORCE BASE, Calif. -- It's Test Management Project season at the U.S. Air Force Test Pilot School. In order to earn a master's degree in flight test, each student must successfully complete a TMP, the school's equivalent of a thesis, and present the results in June. HAVE CLASS, is a controlled L1 adaptive stability study. In the group there are two B-52 pilots, one F-16 pilot and two engineers who are responsible for planning and conducting the flights. Chris Cotting, master instructor of Flying Qualities is the staff advisor for the project. "The way this particular type of control system works is - I don't necessarily have to tell it everything about the airplane. All I have to do is give it some general information and tell it how I want it to fly then the control system will sort out the details," said Cotting. To test the technology, the students flew the L1 adaptive controller on Calspan's variable stability Learjet 10 times within the first two weeks of March. On each flight, the jet is injected with a different failure to see if the L1 adaptive controller can override the problem and maintain nominal aircraft handling qualities. Or in extreme cases, the software would take the aircraft from an un-flyable state to one that is relatively easy to manage. For example, the Learjet's flight computer might be programmed to act like the horizontal tail is half its actual size or the center of gravity has shifted unexpectedly. "The L1 controller basically rejects that and says, 'I don't care, I'm going to fly it the same no matter what' and it does it without knowing which configuration we're giving to the Learjet," said Cotting. In a perfect test, the airplane would fly the same the entire time, without any noticeable changes. Capt. Maggie Blackstun, TPS engineering student, went into the project thinking L1 would "fix everything." "It doesn't fix everything, but it does make everything a lot better and a lot safer," said Blackstun. "The underlying dynamics of the failure state are still there, but L1 is improving them so much." And when the aircraft is flying as expected, L1 does not interfere. B-52 pilot, Capt. Bandy Jeffrey, is the team's project manager. "We understood how the system worked, but we were here to evaluate the extent to which it worked. Everything hasn't been perfect, but it's definitely very interesting," said Jeffrey. "It's definitely made bad planes fly better." One of the things that makes this particular controller special is that it's predictable and the results are repeatable. The flight crew is able to determine how much error the controller can handle before take-off. "We know how much uncertainty we can deal with before we leave the ground," said Cotting. "And that's actually, for this type of work, a big deal." According to Cotting, if the controller is found to be successful, it will have major safety implications. For instance, if a C-17 were loaded with the wrong center of gravity, the control system would account for that. Normally, if an F-15 or F-16 was damaged in battle, the pilot would struggle to bring the aircraft home safely. With the L1, the control system would do all of the hard work, allowing the pilot to fly normally. The system also has the potential to reduce developmental costs in new aircraft. According to Cotting, it would eliminate the need for the extensive wind tunnel testing every new airplane undergoes. "This particular type of controller doesn't need all that. If you can give it a couple of points along the way, it can sort out all of that." said Cotting. The technology for the L1 adaptive controller was first funded by the Air Force Office of Scientific Research and developed in 2004 by University of Illinois professor Naira Hovakimyan and colleague Chengyu Cao. The controller's metrics were validated in 2006 when the technology was flown for the first time in the Rascal platform unmanned aerial vehicle at the Naval Postgraduate School using piccolo autopilot. Then in 2009 NASA began testing the controller on the AirSTAR Dynamically Scaled Generic Transport Model research aircraft. In 2011 the technology was tested on the SIMONA motion-based research simulator at the Delft University of Technology. The testing at TPS is the first time the controller has been tested in a manned aircraft. Hovakimyan and two University of Illinois students came to Edwards for that portion of the project. Hovakimyan hopes the data collected during the project can be used to pursue certification from the Federal Aviation Administration for commercialization. "We are very happy with the fact that all of our theoretical claims have been exactly, precisely verified in every flight," said Hovakimyan. She hopes the technology they have been developing over the past 10 years will help reduce aviation accidents someday. "Human nature is to control and correct," explained Hovakimyan. "You don't want the safety of the aircraft to be dependent on pilot training. As long as it's flyable and controllable you want your autopilot to fly it so well that the pilot does not have to take corrective actions." Since many of the normal ways flight control testing is done don't apply here, the students are forced to expand the flight envelope of the controller with new, unconventional methods of testing. "We've learned a ton about controls, we've learned a ton about different failure states, what a good flying airplane is, what a poorly flying airplane is," said Blackstun. Test pilots and engineers usually only see part of the test projects they are assigned to. Test Management Projects are designed to teach TPS students how to take an initial idea all the way through a flight test program in a confined, controlled way. According to Cotting, "We're giving them a taste of the whole process." Findings of the HAVE CLASS project will be presented as part of the graduation festivities in June.