What happened
On 11 December 2024, a Cessna 172R, registration SE-LGL, was conducting a private flight from Örebro Airport to Karlskoga Airport. Shortly after takeoff, upon reaching an altitude of approximately 2,126 feet, the pilot engaged the Garmin GFC 500 autopilot in heading and altitude hold mode. Because the aircraft had not been properly trimmed for the selected altitude, the autopilot initially used the pitch servo to adjust the elevator.
As the system attempted to maintain the reference altitude, the pitch trim servo activated to assist the pitch servo by commanding a nose-down trim. During this process, a warning light illuminated on the pilot's flight instrument. The aircraft's pitch angle began to increase, and the control yoke moved rapidly forward. Despite the pilot disengaging the autopilot, the aircraft entered an uncontrolled state, with the pitch angle reaching 56 degrees and airspeed dropping to 26 knots, triggering a stall. For approximately 48 seconds, the aircraft underwent extreme fluctuations in pitch and bank angles. The pilot eventually regained control only after pulling the circuit breaker for the autopilot system.
The investigation
SHK investigators examined the aircraft's flight data, extracted logs from the Garmin GI-275 and GTN-650 units, and reviewed footage from a cockpit GoPro camera. The investigation also included technical inspections of the autopilot servos and functional tests. Prior to the incident, the aircraft had recently undergone maintenance to replace three flight control servos and update the autopilot software following a service bulletin.
Findings
- The aircraft was not trimmed to the reference altitude at the moment the autopilot was engaged.
- The pitch trim servo continuously commanded nosedown trim, which contributed to the aerodynamic instability.
- The pilot believed the autopilot remained engaged during the uncontrolled phase, necessitating significant physical effort to counteract the system's inputs.
- The investigation remains ongoing, as the SHK has not yet been granted access to locked log files from the autopilot system, and the exact causal chain between the recent maintenance and the servo behavior is not yet fully established.