Drone Manufacturer: 3DR
Drone Model: Solo
Country: United States of America
Type: Incident
Date: 2019-06-27
Applies: Daytime
Pilot Qualifications: Unknown Status
Pilot Flight Experience: Unknown Hours
Link to External Information About This submission: https://www.safecom.gov/safecom/19-0289
File Uploaded: None
Description
The morning of 6/27 {two days from possible auto-pilot overheat, one day from two normal flights without issue} we got to launch location at approximately 0630, set up the aircraft and associate equipment and prepared to launch. All systems checked out fine, GPS locked relatively quickly. We had been setting the camera trigger as the first waypoint, so in the checklist, when cycling through flight modes, Auto would also test camera triggers. The camera didn““t trigger when setting Auto.
We went though several steps to check the payload {manual trigger – no effect, set GPS distance to 0.5m and walk plane around – no effect, check wiring, reboot, etc.} We switched payloads to the Ricoh camera and still no communication to the airframe. We had been driving on severally wash-boarded roads all week with a very rigid pick-up truck and thought maybe a wire had wiggled loose somewhere, but didn““t feel comfortable working inside the avionics bay beyond visual inspection and making sure all clips were pushed in. I set the camera to interval shooting and we prepare to launch. Temperatures were in the upper 70s, surface winds – light and variable, 1-3 from the E aloft. We did a Loiter launch and climbed above the localized veg and terrain, approx 20m, and observed the aircraft.
It seems to be doing the weather-vaining, looking for the winds, kind of shallow roll-yaws back and forth. I put the aircraft in Althold and waited, it remained stationary for the most part. I gave up throttle and the aircraft began to spin, what appeared to be a torque yaw as it climbed. I countered with a yaw input and it stopped spinning. By this time the plane was about 70m AGL and was slowly drifting to the west as expected. I yawed and verified compass headings were correct in the HUD. I gave pitch and roll commands and the plane responded as expected. Systems seemingly functional I decided to send the plane on its mapping mission, one to buy time to trouble shoot with B. Forbes, and two, to see if any other systems may be experiencing issues, while hoping winds would increase out of the east, making Hover-Loiter more favorable.
The plane completed the mapping mission without issue and I hand flew a landing approach to Transition and input Loiter once in Hover. I flew in close to Home in Loiter and started descending, at about 20m AGL the plane did an aggressive wing dip to each side {Likely from variable winds, or surface rolling from veg}, in the wing dip I entered AltHold to stabilize, decided to climb, in climbing the plane torque yawed again. I countered with yaw input, I was able to move vertically while inputting yaw and balance the rotation.
Not wanting to land with controls countering each other, I decided to try a Loiter landing, but force a wind direction. I set up a landing approach, into the wind, about 90-degrees from my standing position approximately at 30M AGL adjusted the nose heading into the wind and to intercept the landing pad at forward flight. I entered Loiter, and gave slow forward flight while giving slow downward, flying an glide slope that would terminate at the pad with not changing rates. This seemed to keep the plane from seeking the wind and the decent was slow enough to not create yaw issues. About 1ft from the pad I stopped forward flight and continued descent and landed without issue.
At this point the flight team took conference and quickly agreed we were done, despite having one more section to fly. The payload was questionable and something unknown was affecting yaw. We informed the PI on the project and she agreed without issue.
The following day I contacted Adam Sloan and explained the behavior. He agreed that there was likely a loose connection affecting the payload response. He also theorize that two days prior when the aircraft auto-landed in the desert, perhaps there was some strain on the lower rear motor that is affecting its performance.
It likely couldn““t keep speed at full climb inputs, countering yaw. He speculated that the day prior it either didn““t require that level of performance, or that the motor was slowly failing and we didn““t see it until the last flight. He said if the motor can““t match the power output needed we““d see spinning, torque yaw in hover, but forward flight would not be affected.
This explanation matches well with what we saw from the ground, and the yaw issues only seemed problematic while climbing in AltHold {the most responsive hovering flight mode}. We are sending the airframe in to have the motor looked at, payload connectivity, and to evaluate if there are any residual issues associated with the overheat and auto-landing prior.
Reported Cause
Possible motor issue.