During initial climb, the right propeller of the DHC-8-315 (Dash 8) aircraft auto-feathered. The flight crew retarded the right engine power lever, declared a PAN (radio code indicating uncertainty or alert) condition, and completed an uneventful single engine return to Sydney airport.
The aircraft was fitted with two Pratt and Whitney Canada PW123E engines. The flight data recorder (FDR) indicated that the right engine over-torqued to 120 percent for 7 seconds after the propeller feathered. The FDR also indicated that the left engine over-torqued to 117 percent for 20 seconds. The engine manufacturer's transient over-torque limits were not exceeded.
Maintenance personnel found that a loose connection of the right engine torque signal conditioning unit (TSCU) connector pins resulted in an intermittent electrical connection. The TSCU was replaced as a precaution, and the connector was cleaned and reseated. Following a flight test, the aircraft was returned to service.
The propeller auto-feather system, when selected, was designed to automatically feather the propeller during take-off if the engine torque decreased below about 22 percent rated torque. Interlock features in the auto-feather logic and control circuits provided arming control and prevented auto-feather of the operating propeller, once the auto-feather sequence for one of the propellers was initiated. The system provided for relaying a 'power uptrim' (engine power increase) signal to the operating engine.
The ATSB investigation into two previous occurrences (199905044 and 200002853) determined that in the earlier occurrence the electronic engine controller electrical connector was contaminated with water, while a faulty TSCU was found in the other.
The two previous Australian occurrences were also documented on the Australian Civil Aviation Safety Authority's database. A search of the engine manufacturer's database and the Service Difficulty Databases for Transport Canada and the United States of America Federal Aviation Administration, revealed twenty-three similar occurrences in the period from 19 June 1993 to 27 October 2001.
Of the twenty-six worldwide events reported, four were confirmed in-flight engine shutdowns (IFSD). Nineteen were attributed to electrical problems (harness and/or connector or torque signal conditioning unit). Fourteen events occurred during initial climb out and ten during the take-off roll.
Aircraft and engine manufacturer background information
The aircraft manufacturer advised that their data indicated that propeller auto-feathering as described in this incident was a result of loss of torque signal to the TSCU, most likely due to "connector intermittencies". Improvements to the system included design changes to strengthen the torque signal, and flight crew procedural changes. The aircraft manufacturer considered that the present decrease in reported occurrences reflected the success of these changes.
The engine manufacturer reported that the occurrences were associated with an intermittent loss of torque signal. They recommended, when an operator experienced one or more occurrences, that the operator conduct a fleet-wide electrical harness inspection, clean the connectors and enhance connector tightening procedures.
Service bulletins and operator letters
On 25 May 1993, the engine manufacturer issued Service Bulletin (SB) 21269 addressing the application of shrink tubing to the TSCU wiring harness to provide protection from moisture ingress and loosening of the connectors.
On 19 December 1995, the engine manufacturer issued SB 21456 addressing spurious `uptrims' and activation of the auto-feather control system when the system was in the armed condition. Those problems were attributed to the torque sensor air gap not being optimised. The procedures were described for decreasing the torque sensor air gap by replacing a spacer in the unit. The modification improved signal strength and reduced sensitivity to electrical 'noise'.
On 31 January 1996, the engine manufacturer issued SB 21463 addressing fretting of the TSCU electrical connector socket pins. The modification involved replacing the existing wiring harness with one that included a more secure connector assembly with sockets less susceptible to fretting.
On 11 December 1997, the engine manufacturer issued Operator Message Number (OMN) 464 informing operators of two recent IFSD auto-feather events and advised them that those events may have been the result of incorrect tightening torque on the TSCU connectors. They recommended that the connectors be inspected for security, and if found loose, should be disconnected and inspected for contamination and moisture.
On 28 February 2000, the engine manufacturer issued OMN 602 informing operators of recent IFSD auto-feather events and reiterated procedures currently published in the Aircraft Flight Manual. The letter strongly recommended that the operator's review, with their flight crews, the correct procedures to follow with respect to any engine or propeller malfunction on take-off. The letter also noted that previous events had indicated that flight crews often retard the power levers of both engines, thereby cancelling the `power uptrim' signal to the operating engine.
Compliance with these bulletins and messages was not mandatory, however the maintenance organisation implemented the requirements of SBs 21269, 21456 and 21463.
|Date:||27 October 2001||Investigation status:||Completed|
|Time:||1345 hours ESuT|
|State:||New South Wales||Occurrence type:||Propeller/rotor malfunction|
|Release date:||19 December 2002||Occurrence category:||Incident|
|Report status:||Final||Highest injury level:||None|
|Aircraft manufacturer||de Havilland Canada|
|Type of operation||Air Transport High Capacity|
|Damage to aircraft||Nil|
|Departure point||Sydney, NSW|
|Departure time||1345 hours ESuT|