The chief flying instructor was conducting initial multi engine endorsement training for a commercial pilot.  It was the student's fourth session.  After practising simulated engine failures at about 3000 ft in the training area, the aircraft was flown to Tyabb for circuit training.

On the third circuit, on base for landing to the south on the 1000 metre airstrip, the instructor failed the right engine by placing the mixture lever into the idle cut-off position. He told the student not to feather the right propeller but to continue with a touch-and-go landing, thereby experiencing an approach and landing with a failed engine and an unfeathered propeller.  The student selected two stages of flap and, on final, placed both propeller pitch levers into the full fine position.  She closed both throttles before touchdown.  During the landing roll the instructor raised the flaps and advanced the right mixture lever to the full rich position.  Then the student advanced both throttles fully, expecting take-off power on both engines.  At this point the indicated airspeed was about 70 knots. The minimum control speed for single engine operations is 69 knots.  The left engine produced take-off power but the right engine failed to deliver power.

Initially, when the aircraft began to yaw to the right, the instructor thought the student was having difficulty with directional control which she had experienced on previous landings. However, when the aircraft yawed further right, he identified a failed right engine and quickly took over the controls.  By then a collision with runway lights, gable markers and a shallow drainage ditch was imminent, despite the application of full left rudder by the instructor.  He chose to maintain full power on the left engine and become airborne, hoping to avoid the obstacles immediately ahead and find a clearer area to land. The aircraft flew for about 290 metres at a low height in a southerly direction within the airfield boundary but outside the runway gable markers with the instructor struggling to maintain control.  Then the right wing collided with a two metre high pile of old stumps.

The outboard section of the right wing, including the outboard fuel tank, was torn off.  The aircraft turned right through about 260 degrees while remaining upright. It then settled onto the ground and slid backwards, coming to rest about 30 metres beyond the stumps.  As the aircraft came to rest it caught fire.  The right-wing spar was completely broken, and the right engine was torn out of the airframe.  The pilot and the student escaped through flames.  Fire quickly gutted the cabin.

In the seconds between taking over the controls and impact with the stumps, the pilot did not attempt to feather the propeller of the failed engine because of the difficulty he encountered controlling the aircraft at low airspeed. If the RPM of the windmilling propeller had decreased below 800, which it probably had, then feathering would have been prevented by the design of the propeller mechanism.

The airframe was subsequently examined by engineers who found no fault which may have contributed to the accident.  No fault was found with the left engine which had produced take-off power on demand. No fault was found with the right engine.  Aviation gasoline was found in the fuel lines and in the fuel tanks.  There was adequate fuel on board for the flight. The cockpit switches were destroyed by fire.

From the time the instructor selected the mixture to idle cutoff until the student attempted to achieve take-off power during the attempted touch-and-go landing was probably less than two minutes.  The reason the fuel injected engine did not restart when the mixture lever was advanced was not determined.   Witnesses reported hearing a "backfire", probably from the right engine, at about the time full power was selected.

Initially when the right engine failed to deliver power, there was sufficient remaining runway ahead for the pilot(s) to have closed the throttles and stopped the aircraft safely before the end of the runway.  The student, who considered closing the left throttle, did not dare to do so once the instructor took over.  Having taken over the controls, the instructor was confronted with obstacles immediately ahead, which he cleared by becoming airborne. He was then confronted by another obstacle, the pile of stumps which he was unable to avoid.

Four or five times in this aircraft, recently, the instructor had successfully performed a touch-and-go landing after shutting down an engine with the mixture lever during multi-engine endorsement training; once per candidate towards the completion of their endorsement.  On all previous occasions the shutdown engine produced full power when the mixture lever was advanced, and the throttle opened fully.  The instructor's hope was that the training exercise would be a valuable experience for the students.  However, in hindsight he realised that the exercise enhanced the chances of an accident occurring if the shutdown engine did not restart quickly during the touch-and-go.

Since this accident it has become apparent that many multi engine flying instructors, approved testing officers and flying operations inspectors have varying opinions as to whether engines should be shut down completely in the circuit or at low level, as opposed to selecting zero thrust settings appropriate to the particular aircraft.

CONCLUSIONS

Findings

1. The instructor and student were properly licenced and qualified for the task.
2. The aircraft was serviceable and after the accident no faults were found that may have contributed to the accident.
3. The right engine did not deliver power when the mixture and throttle levers were advanced for the touch and go landing. The reason why the right engine did not restart was not determined.
4. The flying instructor underestimated the risk involved versus the training advantage of shutting down one engine at low level followed by a touch-and-go landing.

Significant factors

The following factors were considered relevant to the development of the accident:

1. The instructor shut the right engine down during approach.
2. The instructor briefed the student to fly the approach and land with the engine shut down and the propeller windmilling.
3. During the subsequent touch and go landing, the engine did not restart/deliver power when the mixture and throttle levers were advanced.
4. When the aircraft then began yawing to the right the instructor initially believed the student was having ongoing difficulty with directional control.
5. When the instructor identified that the yaw was due to asymmetric power and took control of the aircraft, the situation with aircraft performance, runway remaining and surrounding terrain was such that an accident was inevitable.

SAFETY ACTION

The following safety action is being taken:

1. The Civil Aviation Safety Authority (CASA) will release a Civil Aviation Advisory Publication containing an initial multi engine endorsement training syllabus.
2. The Bureau of Air Safety Investigation will conduct a research project on advanced training accidents.
3. Since this accident, CASA has held industry symposiums to discuss standardisation of multi engine training procedures.

Whilst turning onto a right base leg for runway 16R, with the autopilot disconnected, right rudder input was unsuccessful, and the rudder appeared 'frozen'. After two unsuccessful attempts to deflect the rudder to the right the pilot applied left rudder which responded, although requiring the application of higher-than-normal force on the rudder pedal. Subsequent application of right rudder was successful but felt extremely stiff. A light 'tramping' was also felt through the pedals. The number 1 yaw damper was disconnected temporarily with no apparent change in control feel or response. The aircraft subsequently landed without further incident.

An investigation found seized bearings at the pilot's aft rudder input quadrant, requiring excessive input forces for rudder deflection. The bearings, shaft and support frame assembly were changed, which resulted in the system returning to normal operation.

There had been a history of rudder input problems since the aircraft's last scheduled maintenance in the United Kingdom in February, when the quadrant shafts had been changed as an inspection sampling program requirement.

The aircraft was cleared for a visual approach to the duty runway (RWY) 11 and joined a 7km final from a right base position. The crew were advised by ATS of several military fighter aircraft taking off from RWY 29.

The captain advised ATS of his position and that a traffic confliction was about to occur as he saw two of the six fighter aircraft pass above and to the left of his aircraft. The last of the section of fighters passed within 1,000 metres of the aircraft at the same level.

Subsequent investigation revealed that a request for the use of a non-duty runway (RWY29) had been agreed to previously for the military aircraft, but there had been a delay with their departure. The radar controller failed to note that a non-duty runway was marked on the flight strips. Reflected radar returns were being experienced on the Surface Movement Radar and the controller misidentified these reflected returns as those of the military aircraft, and being relatively inexperienced cleared the military aircraft to take off in confliction with the approaching aircraft.

Whilst on climb to FL410 the pilot requested a descent to FL270 due to a minor engine problem. A short time later the pilot advised he was shutting down the left engine and requested a return to Sydney for a priority landing. He also requested the aerodrome fire service to be placed on standby. The aircraft subsequently landed without further incident.

It was later determined that the engine had suffered a 1st stage turbine blade failure. A metallurgical report subsequently determined that the failure was caused by fatigue crack growth which initiated from a stress concentrator in one of the blade cooling passages. It appeared that the stress concentrator was created during blade manufacture.

At the end of the landing roll the pilot noticed that the aircraft was leaning to the left and thought that the tyre had deflated. The lean increased and the aircraft then rolled forward onto its nose with the propeller striking the ground.

It was found that the nose gear leg had broken from its support assembly and collapsed.

Airworthiness Directive AD/PL-12/1 amdt 2 requires a visual inspection of the support assembly every 100 hours time in service. The aircraft had flown about 60 hours since the last inspection.

OE-LAV, a Boeing B767 aircraft (flight number LDA2) enroute from Melbourne to Singapore at Flight Level 350 (FL350) suffered a reduction of oil pressure in one engine. The problem could not be rectified in flight and the engine was shut down.

The crew notified Perth ATC of the problem, advising that they were commencing an emergency descent to FL240 and required a diversion to Darwin. They were advised of opposite direction traffic, HS-TMB, a DC10 aircraft (flight number THA991), at FL330 which should be passing their present position at that time.

They were initially cleared to FL340, but the crew advised that their TCAS equipment had shown they had passed THA991 and were now diverting right of track and requiring an immediate descent to FL240.

THA 991 was instructed to descend to FL290, but as ATC could not provide a positive separation standard between the two aircraft instructed LDA2 to maintain its own separation while descending to FL240 and track direct to Darwin.

During a stock mustering operation, the head stockman, who was a passenger in the helicopter, asked the pilot to land so that he could give instructions to stockmen on the ground. The pilot landed the helicopter and commenced to secure the controls.

The stockman left his seat, stood up on the right step attached to the skid gear, and began to signal the stockmen by waving his right arm. The stockman's arm contacted the main rotor disc causing two fingers to be severed. After first aid, the stockman was transported in the helicopter to the Cloncurry Hospital where it was discovered that he had also broken his right wrist and arm.

The helicopter was slightly damaged with a broken right centre Perspex panel which was struck by the stockman's arm as it was flung from the rotor disc.

The stockman was a very experienced helicopter passenger. He had not been specifically briefed on helicopter safety aspects for this particular flight. The pilot was concerned with securing the helicopter after landing and did not see the event. He first noticed something was wrong when he saw the stockman writhing on the ground.

The pilot had been goat spotting and decided to land and wait for the accompanying helicopter to refuel.

The passenger was not feeling well, and the pilot had landed on the strip two years previously. At that time a road, that crossed the strip mid-way along it, was level with the strip surface and did not create any problems. Consequently, the pilot only conducted a cursory inspection, of the strip, prior to landing, and did not notice that the surface conditions had changed.

As the aircraft decelerated through about 15 knots, during the landing roll, it crossed the road. The depth of the road surface, below the strip surface, caused the aircraft to bounce and as it touched down again the nosewheel collided with the ridge on the far side of the road and collapsed.

Recent heavy rain and grading by the local council had changed the profile of the road, dropping its surface well below that of the strip.

The left side cabin door opened in flight, shortly after take-off. The pilot reduced power and attempted to close the door. When he found he was unable to close the door against the aerodynamic forces, he transmitted a "MAYDAY" call, advising of his intention to force land the aircraft. The aircraft was landed in a paddock with a soft surface, resulting in the nose landing gear being torn off.

A post flight inspection found no defects with the door latches. It is likely that the pilot failed to latch the cabin door properly before flight.

The aircraft was making an approach to a gravel surfaced strip in CAVOK, light wind conditions with no turbulence. The pilot reported that full flaps were extended, and a final approach speed of 75 knots was maintained until touchdown, which was slightly short of the strip threshold.

The right main wheel struck soft ground, dislodging the right main landing gear. As the aircraft continued onto the strip the remaining landing gears were also dislodged, resulting in major damage to the aircraft. All four occupants were able to exit the aircraft safely.