AT76, en route, west-southwest of Sydney Australia, 2014

AT76, en route, west-southwest of Sydney Australia, 2014


On 20 February 2014, an ATR 72-600 crew mishandled their response to an intended airspeed adjustment whilst using VS mode during descent to Sydney and an upset involving opposite control inputs from the pilots caused an elevator disconnect. The senior cabin attendant sustained serious injury. After recovery of control, the flight was completed without further event. Post flight inspection did not discover damage to the aircraft which exceeded limit and ultimate loads on the stabilisers and the aircraft remained in service for a further five days until it was grounded for replacement of both horizontal and vertical stabilisers.

Event Details
Event Type
Flight Conditions
Flight Details
Type of Flight
Public Transport (Passenger)
Intended Destination
Take-off Commenced
Flight Airborne
Flight Completed
Phase of Flight
Deficient Crew Knowledge-handling, Inadequate Aircraft Operator Procedures
Inappropriate crew response - skills deficiency, Inappropriate crew response (technical fault), Manual Handling, AP/FD and/or ATHR status awareness
AP Status Awareness, Aircraft Loading, Temporary Control Loss
Dispatch of Unserviceable Aircraft
Damage or injury
Aircraft damage
Non-aircraft damage
Non-occupant Casualties
Number of Non-occupant Fatalities
Occupant Fatalities
Few occupants
Number of Occupant Fatalities
Off Airport Landing
Causal Factor Group(s)
Aircraft Operation
Aircraft Technical
Safety Recommendation(s)
None Made
Investigation Type


On 20 February 2014, the crew of an ATR 72-600 (VH-FVR) being operated by Virgin Australia Regional Airlines on a scheduled passenger flight from Canberra to Sydney experienced an in-flight upset during descent in day VMC involving a pitch control disconnect and serious injury to the senior cabin attendant. Following recovery, the remainder of the flight proceeded without further event. Following flight completion, extensive damage to the aircraft was not discovered and it continued in service for a further five days before being grounded.


An Investigation was carried out by the Australian Transport Safety Bureau (ATSB)FDR and CVR data was successfully downloaded and used in the Investigation and equivalent QAR data was made available to the aircraft manufacturer.

It was established that the First Officer had been PF. The flight was uneventful until the descent towards Sydney with the AP engaged in VS mode and with “a target airspeed of 235 knots”, when passing 8500 feet QNH a “rapid airspeed increase” was observed. The First Officer reduced the engine power setting and used control wheel steering (CWS) to temporarily disconnect the AP “before manually raising the aircraft’s nose to control the speed”, expecting that these actions would arrest the speed trend. The Captain stated that he had been unsure if the control inputs made would be sufficient to avoid exceeding Vmo (250 KIAS) so he had put one hand on the controls and disconnected the AP in order to raise the nose further.

With both pilots making simultaneous nose-up pitch inputs, the rapid pitch up produced a noticeable increase in ‘g’ loading which prompted the First Officer to immediately reverse his control input to nose down. Both pilots then reported noticing that “the controls suddenly felt different and ‘spongy’”. A series of system status warnings were annunciated as it became apparent that the aircraft was “under control at a stable attitude and speed” - it was reported as “being level or in a slight descent at an airspeed of about 230 knots”.

One of the warnings was ‘Pitch Disconnect’ indicating that the left and right elevator control systems had uncoupled from each other which allowed independent movement of the left and right elevators via the Captain’s and First Officer’s control columns respectively. After determining that neither column was jammed, it was decided that the Captain would take over as PF for the remainder of the flight which was completed without further event.

The upset associated with the occurrence of a flight control disconnect was found to have resulted in the senior Cabin Attendant sustaining serious injury.

Recorded flight data showed that there had been no atmospheric turbulence prior to crew intervention. It showed that “in the seconds leading up to the occurrence, there were a number of rapid increases in the recorded airspeed” with the First Officer making “three nose up control inputs” whilst the AP was temporarily disconnected. Just before the pitch disconnect occurred and “shortly before the First Officer made the third nose up input”, a small pitch up input was made on the Captain’s control column and then, when the First Officer reversed his pitch command, the Captain’s pitch up input “increased at about the same rate that the First Officer’s pitch up input decreased and the elevators uncoupled as per design for prevent jamming of the controls”. At disconnect, vertical acceleration was increasing through +2.8g to reach a maximum of 3.34g one second later and aircraft pitch attitude was increasing. The data showed “peak elevator deflections of +10.4° and -9.3° […] about 0.2 seconds after the pitch disconnect” and that about 0.25 seconds after these peak deflections, the Captain had moved his control column forward until both elevators were in similar positions. It was noted that “due to the data resolution and lack of a parameter that monitored the pilot’s disconnect button, it could not be determined if the autopilot disconnection was due to the load exceedance or the manual disconnection reported by the Captain”. It was noted from the CVR data that the first verbal indication of flight crew awareness of the pitch disconnect was about 6 seconds after the master warning caused by it had activated.

Significant structural damage caused to the aircraft by the “large asymmetric elevator deflections” when near to Vmo (which was not exceeded) was not identified until 5 days later during which time the aircraft continued to be operated on the basis that it was airworthy. Examination of the aircraft by investigators and ATR representatives then found extensive damage had been caused to the horizontal stabiliser (some is shown in the annotated photographs below) which included:

  • external damage to both the left and right horizontal stabilisers
  • fracture of the composite structure around the rear horizontal-to-vertical stabiliser attachment points
  • fracture of the front spar web
  • cracking of the horizontal-to-vertical stabiliser attachment support ribs
  • cracking of the attachment support structure
  • cracking and delamination of the skin panels at the rear spar.

Following these findings, ATR required that both the horizontal and vertical stabilisers should be replaced before release for any further flight.

An analysis based on data from the QAR showed that at the point of control disconnect, the Ultimate Load on the horizontal stabiliser had been exceeded by almost 50% for less than a second. It also showed that the Limit Loads for the vertical load on the horizontal stabiliser, the asymmetric moment on the horizontal stabiliser, the bending moment on the wing and the engine mounts had all been exceeded and that the vertical load on the wing had been reached.

In the earlier stages of the Investigation, the ATSB was not aware of any other in-flight pitch disconnect occurrences on ATR 42/72 aircraft so that it was considered existing procedural risk controls would be effective in maintaining a sufficiently low probability of a recurrence. However, when ATR responded to a request for details of any other pitch disconnect events of which they were aware, it became clear that there had been at least three other similar events where simultaneous opposing pitch control inputs had led to a ‘breakout’ of the pitch control system in flight. It was noted that two of these events remain under investigation by “other international agencies” so that verified details are not yet available and it appears that the circumstances surrounding the third event are not fully understood.

The horizontal-to-vertical stabiliser attachment with the fairings removed showing cracking of the composite structure around the rear attachment point [reproduced from the Official Report]

External damage to the tailplane and fairings indicated by marks/stickers [reproduced from the Official Report]


Once it became known that there had been other in-flight pitch disconnect occurrences, none of which had been the result of following the QRH response to a jammed control system, it became clear that the likelihood of an inadvertent in-flight pitch disconnect was higher than had first been considered. It was therefore determined that a Safety Issue existed as follows:

  • Inadvertent application of opposing pitch control inputs by flight crew can activate the pitch uncoupling mechanism which, in certain high-energy situations, can result in catastrophic damage to the aircraft structure before crews are able to react.

Safety Action taken has included the following:

  • ATR has:
    • Comprehensively revised FCOM content on “Procedures and Techniques - Flight Controls - Pitch”.
    • Revised FCOM and QRH “Procedures following failure - Flight Control - Pitch Disconnect”.
    • Created a new Job Card for a Detailed Visual Inspection (DVI) to be carried out after any reported pitch system disconnect in flight.
    • Revised the existing Job Card dedicated to the re-clutching of the pitch uncoupling mechanism to require action in accordance with the new DVI requirement after any report of a pitch system disconnection in flight.
    • Revised the AFM Supplement on ‘Ferry Flight with Pitch Elevators Disconnected’.
    • The Company has also contacted all Operators who have previously reported an in-flight pitch disconnect to advise them to carry out the new DVI procedure.
  • Virgin Australia Regional Airlines has:
    • Reviewed and revised (where necessary) policy and procedures associated with descent speeds, handover and takeover procedures, overspeed recovery and on-ground pitch disconnects.
    • Updated maintenance requirements following a pitch disconnect.

The Safety Issue identified has been deemed to be of sufficient magnitude to warrant the release of the Interim Report on which this summary is based and this was published on 15 June 2016.

Related Articles

SKYbrary Partners:

Safety knowledge contributed by: