December 13, 2012
Fly-by-wire control system are standard issue in military aircraft -- including drones -- and in major new transport aircraft. Within three years, Diamond says it hopes to offer such systems for light aircraft and it's further along the developmental road than we imagined.
When we visited Diamond Aircraft's skunkworks in Austria earlier this year, the company showed us a crude fly-by-wire developmental project that, eight months later, has flown some 50 hours and will soon be ready to enter its second developmental phase, according to Diamond CEO Christian Dries. The system is installed in a DA42 testbed and Dries says if it can overcome certification hurdlesno small feat, to be sureit could be certified and available as an option as early as 2015.
But why would you want such a thing? Dries says as Diamond envisions the system, it's a more sophisticated and capable version of envelope protection that's now a common feature in a higher end autopilots. That means upsets would be either prevented or automatically recovered and the airplane would be capable of autoland with or without initiated pilot commands.
In the DA42 test article, the FBW system is set up rather like a human controlled autopilot: "We have the mechanical system on the left side and we have the control stick for the fly by wire on the right side. From the right side, we are activating servos and these servos are, for the moment, linked to the mechanical steering," Dries said. This was done at the insistence of regulators, who want some proving hours before turning Diamond loose to pursue the next step: direct logic control through servos linked to the principle control surfaces. That stage will be followed by envelope protection and autoland, the software for which Diamond developed experience in building an unmanned helicopter drone.
"Initially, we are testing the fly by wire system. It has no additional features in the computer. That is the first step. This basic system works fine," Dries told AVweb in an interview this week.
That sounds impressive as far as it goes, but what about feedback to the pilot? "At the moment, there's just the control stick and it doesn't have too much feedback. In the next step, we will try to simulate the feedback. But as you know, this concept is used in airliners, which don't have that much feedback either," Dries said, adding that adapting the different control feel is a question of training. In the beginning, it was a little different than the real airplane, but we had to adjust some gain and modes, now it flies like the regular airplane," Dries said.
For now, the control laws are processed through a purpose-made computer measuring about 5 by 10 inches, but as development work continues, that will be miniaturized. Diamond will also need a capable autopilot to make envelope recovery and autoland work.
"I have to tell you," Dries says, "the GFC700 autopilot from Garmin is the best autopilot I can imagine. But, for whatever reasons, Garmin may not allow us to interfere in the system. So this may be a completely new development for us and our partners."
Diamond's planned end-state capability for the FBW system are ambitious. "The main reason, of course, that we do fly-by-wire systems is envelope protection. The next step for us is the autoland system, which we call internally the electronic parachute. And this should help the pilot, either a private pilot or a single-pilot operation, if the pilot is not capable of flying the airplane. It will land automatically.
Diamond envisions a system in which the electronics continually monitor the pilots alertness and if it detects lack of response -- say due to hypoxia or illness -- it would automatically transmit an emergency transponder code and initiate autoland at the nearest suitable airport. Dries says the software to achieve this -- and certifying it -- is the most difficult challenge, the hardware less so.
Dries says regulators are requiring the same statistical reliability in the FBW systems that airliners are certified to under FAR Part 25. That means a failure probability of 10 to 9th, which Dries says will obviate the need for parallel backup controls. Nonetheless, the control servos may be doubled up and controlled by two or multi-channel protocols. The DA42 and the new DA52 already have three-deep electrical redundancy and there's headroom to improve that, Dries told us.
And the market? "Don't forget, an airplane with this system can be used as an unmanned airplane. It can do everything by itself," Dries says, meaning the FBW system will fit right into Diamond's booming airborne sensor system. But it's also intended for manned aircraft, first the DA42 and DA52, but eventually everything.
"I hope that once it's on the market and if it's not too expensive, it can be installed in almost every aircraft," Dries says. Diamond will like to bring FBW in as an option under $100,000, but it might cost as much as $150,000. If it's accepted by the market in volume, the price would come down. Dries has no illusions about the certification hurdles but Diamond has generally met its cert projections for piston aircraft. Dries believes FBW could be available as early as 2015. And once it is, such technology could open the door to a host of new designs, perhaps some designed to be intentionally unstable to take advantage of lower drag planforms and lifting surfaces.