The pilot was conducting the last of his spraying tasks for the day. The paddock had an uphill slope and there were two groups of tall trees at the uphill end. The first run was conducted up the slope but during the subsequent pull up and procedure turn, strong turbulence was encountered. The left wing struck branches in one group of trees, control was lost and the aircraft struck the ground heavily. The fuel tank ruptured, a fire broke out and the wreckage was completely gutted. The spraying was commenced in strong wind conditions and the procedure turn was executed in the lee of the trees. When the turbulence was encountered the pilot banked the aircraft and attempted to fly between the two stands of trees but was unable to prevent the left wing striking a number of branches.

The pilot had been advised that the aircraft propeller was defective and he indicated to witnesses that he intended only to conduct taxying. However, the aircraft was seen to become airborne and start to fly a circuit. On base leg the engine noise grew louder and the aircraft fell to the ground with the rotor stationary. Wreckage examination revealed that part of the propeller had separated in flight and the remaining propeller section had impacted the rotor following failure of the engine mounts due to excessive vibration.

Circumstances:

The aircraft was landing at night on runway 11C, at the end of a multi-sector cargo flight. Weather conditions were fine and clear with a light wind from the north-east. The pilot said the touchdown was normal. However, the aircraft started to pull to the right, the right tyre deflated, and the right main landing gear collapsed. Skid marks on the runway were consistent with the right wheel being locked from about 400 metres beyond the landing threshold. The skid continued for a further 124 metres, parallel to the centreline, before the right tyre deflated. The skid continued for a further 151 metres before the right main landing gear collapsed, resulting in the aircraft departing the runway to the right. A subsequent examination revealed no mechanical fault with the wheel brakes or wheel bearings. Evidence was found which was consistent with an excessive application of the right brake.

After about 80 metres of ground roll following a normal touchdown the nose began to drop, followed by the right and left wings, and the aircraft slid to a halt with the gear retracted. No mechanical fault or defect was subsequently found with the aircraft. The weight of available evidence indicated that the pilot had probably inadvertently selected the gear up shortly after touchdown.

The pilot was conducting his first solo circuit and landing. Following the initial touchdown the aircraft ballooned twice, and the pilot was unable to maintain directional control after the final touchdown. The aircraft collided with another glider which was being prepared for launching. Immediately prior to this particular flight, the pilot had successfully completed a check flight with an instructor. It was considered likely that the aircraft had been affected by a crosswind component during the latter stages of the landing, and the pilot had been unable to react in sufficient time to avoid a collision with the parked glider.

During the first take-off for top dressing operations, the aircraft commenced to swing to the left. The pilot applied opposite rudder and braking but was unable to control the swing. The side loads imposed on the right hand gear caused it to fail at its attachment to the stub wing, and fold up. The left hand oleo strut piston had separated from its cylinder, which then failed at its attachment, and fell from the aircraft. The gear then started to fold back, bringing the piston into contact with the tyre, which gave a braking affect and caused the aircraft to swing. A Transavia Service Bulletin and an Airworthiness Directive require a mandatory inspection to be carried out on all oleo struts of this type after each 100 hours time in service, in order to prevent a failure of this nature from occurring. The aircraft log books could not be located, and it could not be established whether the required inspections had been carried out.

The aircraft Captain had received a Garuda Route Clearance Unit presentation on the Perth route prior to departure from Bali and he had made one previous flight into Perth as the First Officer on a DC-10 in May 1988. After an evidently uneventful flight the aircraft was making an approach to Runway 03 at Perth. The cloud base at the time was approximately 800 feet above sea level. Air Traffic Control had used radar vectoring to position the aircraft for the approach and the crew was then instructed to make a pilot intercept of the appropriate radial from the Perth VOR. The aircraft did not accurately follow the required radial and the aircraft diverged up to about one kilometre either side of the published final approach path. The First Officer who was visually monitoring the approach was unable to sight the runway lighting and an overshoot was commenced when the aircraft was about three kilometres from the runway threshold. At about the same time as the Captain commenced his overshoot procedure the Perth Tower controller who had monitored the aircraft's approach from the time it had descended below the cloud base decided that the aircraft was not in a position to make a normal landing. He therefore instructed the aircraft to overshoot. The subsequent approach and landing were without incident. The investigation revealed that the details of the approach published by the Jeppesen Company in the USA contained an error of 5 degrees in the direction of the VOR radial required for the final approach path for the particular runway. These approach details were contained in company documentation available to the crew. Whether this error had any relevance to the incident circumstances was not established. It was noted that there are areas of significant lighting in the vicinity of Perth Airport which could hamper visual acquisition of the runway lighting under conditions of poor visibility. A precision approach aid such as ILS is not available for Runway 03 and runway approach lighting is not installed.

Circumstances:

The aircraft was climbing 600 ft after take-off when the pilot noticed a buzzing noise. While attempting to isolate the source of the buzzing two loud bangs were heard the right engine fan RPM reduced to zero and the generator dropped offline. Sparks and flames were seen in the right engine inlet area. The pilot shut down the right engine and executed a single engine landing. Initial investigation showed that the low compressor stator had ejected through the compressor casing with parts of the engine and cowlings falling into a residential area. This examination revealed that the low compressor stator also known as the fan stator had been rotated at high speed. The resultant high heat and milling action cut and severed the fan casing which dropped into the fan with destructive results. The stator had split and ejected through the severed case. The outer ring of the stator is positioned axially by mating snaps and is anchored against rotation by a series of 1/16 inch rivets. These rivets had sheared allowing the fan air flow to rotate the stator. The manufacturer advised that they had received reports of occasional loose or missing rivets and had also experienced two previous incidents of total rivet shear associated with stator spin but without the major break-up that occurred in this instance. Rivet failure has been attributed to a resonant condition in the fan area occurring at transient rotor speeds. It is also suspected that foreign object damage (FOD) and birdstrikes could by introducing a single event overload be a contributory factor. One of the two previous stator spin incidents occurred concurrent with a multiple bird strike. The failed engine suffered a bird-strike incident in June 1989. A visual and borescope inspection in accordance with the manufacturer's requirements found no evidence of damage. The engine was internally washed and then satisfactorily test run. It operated for a further 580 cycles and 620 hours before the rivet failure. The manufacturer has introduced two Service Bulletin (SB) modifications in an endeavour to rectify the rivet failures. SB7264 replaced the 1/16-inch rivets with 1/8-inch rivets and SB7268 fitted a silicon vibration dampening ring to the outer periphery of the fan case. Both recommended accomplishment when disassembly afforded access to the area; however, neither was mandatory.

The failed engine was fitted with the silicon vibration dampening ring whilst undergoing the bird-strike inspection some 16 months prior to the incident. However, the larger rivets had not been incorporated as the area had not been opened up since the SB was issued in January 1988. The rivets had not been closely inspected while the engine was in service because a rivet inspection was not specifically called for in the maintenance schedules nor was it listed as a requirement of the FOD/bird-strike inspection. The SBs themselves did not alert operators to the potential for rivet failures, nor did they advise that rivet failures had occurred in service. The rivet replacement SB did not list the engines to which it was applicable. Accordingly, the maintenance organisation responsible for this engine would not necessarily have been alerted that this engine was among those that needed to have the smaller rivets replaced. The reasons given for introducing the vibration dampening ring and for replacing the rivets were unlikely to alert maintenance organisations to the importance of the rivet replacement nor would they be sufficient to initiate specific inspection of these rivets during routine maintenance or bird-strike inspections. As a result of this incident the manufacturer recognised the possible effects of minor bird strikes on the stator securing rivets and called for a field inspection of the subject rivets. Out of 452 responses 43 (9.5) indicated some form of rivet discrepancy. There were no reports of discrepancies with the modified rivets. The manufacturer has revised Service Bulletin SB7264 to upgrade the compliance recommendation such that the rivet replacement be accomplished at either the first workshop visit or prior to the next flight following any foreign body ingestion. The manufacturer also issued specific instructions for inspection of the rivets which will be incorporated into the maintenance manuals and revised the format for Service Bulletins so as to provide operators with more details of the required maintenance action.

Significant Factors:

The following factors were considered relevant to the development of the incident.

1. Engine design efficiency.

2. The rivets securing the low-pressure compressor stator had a history of failure.

3. The engine had previously suffered a bird-strike incident.

4. The rivets were not inspected as there was no requirement to do so.

5. The rivets were not replaced as the engine had not been subjected to the compliance requirement.

6. The rivets securing the low-pressure compressor stator failed.

7. The engine was severely damaged when the low-pressure compressor stator rotated at high speed and severed the compressor casing.

The pilot completed his pre-takeoff checks before launching for a local flight. He subsequently advised that while he was manoeuvring behind a glider ahead of him in the circuit the canopy rattled and flew open. He was unable to close the canopy and shortly afterwards it tore free from the glider structure. A successful and uneventful landing was then carried out. Investigation revealed that it was possible to close the canopy locking handle fully without the locking pins engaging in the lock studs. In this type of glider the unsecured canopy was then not easy to detect from inside the cockpit. The pilot had limited experience on type and was not aware of the difficulty in detecting an unsecured canopy. Physical checks for canopy security by pushing on the canopy were discouraged by the gliding club for fear of damage. The launching crew had not noticed that the canopy was unlocked during the hookup procedures.

Surface conditions were; strong south-easterly winds of 30 knots with gusts to 40 knots, water surface choppy. The pilot made three unsuccessful attempts to align the aircraft for a takeoff. On his fourth attempt he was successful and the takeoff run was commenced. Shortly afterwards, a gust of wind lifted the left wing, the right float dug into the water, and the aircraft nosed over. Both occupants swam clear of the aircraft which had come to rest upside down on a submerged mudbank. The takeoff attempt had been made on the town reach of the river in a crosswind. An into-wind takeoff would have been possible if the pilot had taxied the aircraft to the next reach of the river.