After four flights on Saturday, Pilatus had a tailwheel issue that kept us grounded for the rest of that evening. However, on Sunday morning the airplane was again ready to go, and so were we. With (relatively) light winds around 15 mph, I was excited to continue to collect good measurements — the sun was even out! After a bit of troubleshooting for our instrument payload, we were able to launch the first flight Sunday around 10 am. Unfortunately, we again had a separation of the tailwheel support structure on landing — glues and epoxies just don’t work as well in the cold as they might in warmer climates. This resulted in another 2-hour delay before launching for our sixth flight of the campaign shortly after lunch. For this flight we changed the payload to carry two broadband longwave radiometers (along with our aerosol sampler, POPS, and our thermodynamic probe). For the first seven minutes of the flight, everything went very well, and the aircraft was flawlessly executing its box pattern. Then, things got a bit hairy — “We are losing altitude” and “I have no propulsion” are not words that you want to hear when operating an unmanned aircraft, particularly when the aircraft is on the furthest waypoint of the flight pattern… For some yet unexplained reason, the Pilatus propulsion system experienced a failure. Now any aircraft can experience propulsion system failures (even large passenger jets), and the flight crew (Doug and Tevis) handled it in a controlled, professional manner, gliding the aircraft down from its 100 meter cruise altitude for a safe landing. After getting a GPS fix on the approximate landing location, I went for a ride in the Kubota with Wessley King, one of the AMF-3 operators and our bear guard for the day, to recover the plane. After heading down to the beach, we were able to make visual contact with the aircraft, sitting beautifully approximately 100 yards offshore on the sea ice. It was really hard to tell that anything had gone wrong! The only damage was again to the same tailwheel block that had already failed twice and to one of the main gear struts which had popped out of its wheel hub due to a harder-than-usual landing. The airframe itself, along with all instruments were just fine, and our data system was still collecting measurements until I got to the plane and powered it down.
We took the plane back to the hangar for testing and could not find any obvious issues — the motor, power system, battery voltage and speed controller all seemed fine, and power and thrust testing also went well. There had also not been any indication of autopilot or radio communications failures, as both continued to work until the aircraft was powered down. This left us scratching our heads as to why the plane lost propulsion. We likely won’t have an answer to this question until we’ve had a chance to do more exhaustive system testing and analysis of the logged files back in Colorado — it is possible (but speculative at this point) that the extreme changes in temperature between the ambient “cold” conditions (around 5 F) and the warm temperatures experienced in flight impacted the performance of the electric motor. This event was certainly another reminder that what we are trying to accomplish is not at all easy, and that operating unmanned aircraft in the Arctic is something that pushes the limits of our current equipment. It also demonstrated that when the appropriate safety procedures are followed and risk mitigation is undertaken, system failure such as this does not have to have a bad outcome. Manned aircraft could have similar system failures and it is for this reason that pilots generally feel uncomfortable flying at 100 m altitude over the sea ice and why this is a great application for unmanned platforms. While the aircraft was easily repairable to a flyable state, given the uncertainty surrounding the cause of the propulsion failure, we made the decision to stop flying for the campaign and began packing up to head home.