Circumstances:

The aircraft was scheduled to conduct three round trips between Townsville and Cape Richards, with a stop at Orpheus Island on some legs, during the day. The pilot involved in the accident flew the first and third trips and another pilot flew the second. After the first trip the pilot reported that he pumped the floats out and considered that the quantity of water removed was normal. On the third trip he picked up a "standby" passenger at Orpheus Island. On arrival at Cape Richards the scheduled six passengers were loaded. A witness employed by the resort to handle the aircraft and passengers on the island reported that the floats appeared to be sitting in the water such that the water was above the normal water line on the floats. Examination of the aircraft loading indicated that the aircraft was overweight, and the centre of gravity was just inside the rear limit. The pilot reported taxiing at 1613 hours with eight persons on board. At 1624 hours he reported that he was returning to unload one passenger. In that time two take-off attempts into the north-east were made. The wind in the bay where the attempts were made was a light northerly. The pilot again reported taxiing at 1634 hours with seven persons on board. A further two take-off attempts were made. On the final attempt the pilot did not taxi as far into the bay as on previous occasions. The take-off was continued well out beyond the shelter of the island into an area where the wind was easterly at about 10 knots, and the swell was 1 to 1.5 metres. The pilot reported that the aircraft had attained an indicated airspeed of 55 knots, and he intended to fly it off the water at 57 knots. The right float had lifted from the water, and it hit a wave which pushed the right wing up. The pilot was unable to lift the left wing which hit the water, causing the aircraft to cartwheel.

Significant Factors:

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

1. The pilot selected the incorrect take-off direction for the wind conditions prevailing.

2. The pilot continued the take-off into an area of unsuitable swell. This accident was not the subject of an on-scene investigation.

After taking off in calm, fine weather conditions, the pilot climbed the aircraft to a height of about 500 above the ground and commenced a series of manoeuvres incorporating steep angles of bank. During the nose-down segment of a wing-over, the aircraft was seen to be near vertical and to accelerate rapidly before pulling up into a steep turn at about 200 feet above the ground. The turn was tightened to at least 60 degrees of bank, and shortly afterwards the right wing failed. The aircraft fell to the ground at high speed and collided with a fence post after sliding for 18 metres. The aircraft had been substantially modified by the pilot, and was being operated at well in excess of the design weight. The wing spar had failed as a result of the overload forces applied to the structure during the sequence of manoeuvres.

After a normal preflight inspection and takeoff, the pilot selected the landing gear up. The indicators were later checked and the pilot found that all the up lights were indicating normally, but one down light was also on. During the short flight the pilot operated the landing gear a number of times. He was satisfied throughout the flight that the landing gear was retracting; his only concern being whether the gear would extend again on his return flight. During the gear checking the aircraft arrived in the circuit area of the destination. The pilot made routine radio calls to his company and Townsville Flight Service, checked for boats in the landing area and noted that there were only four landing gear lights illuminated. On touchdown the pilot was unable to prevent the aircraft from pitching excessively nose down. The aircraft came to rest with the engine and some of the wings in the water. Subsequent actions by the pilot and a nearby boat resulted in the aircraft being righted and then towed to a mooring site. The landing gear had been in the down position for the landing. Since the landing gear indicating system uses four blue lights to indicate gear up and four green lights to indicate gear down, the pilot concedes that he had not noted the colour of the lights adequately in order to determine whether the gear was up or down.

The pilot flew to a private airstrip for a visit with fellow aviators. After about one and a half hours he decided to fly home. The aircraft took off to the north and climbed straight ahead to about 100 feet above the ground before making a steep, 180 degree, left turn and descending towards witnesses standing in front of a hangar. About 100 metres south of the witnesses the pilot commenced a left barrel roll from an estimated altitude 60 feet and an airspeed of about 85 knots. The aircraft managed to get over the top of the barrel roll but then descended rapidly. It impacted the ground in a south westerly direction with a nose-low, wings level attitude. It bounced off the ground and came to rest in about 10 metres with the engine dislodged from the airframe and the landing gear collapsed. It has been determined that the aircraft was serviceable prior to the accident. Neither the weather conditions nor the pilot's medical status contributed to the accident. Subsequent inflight checks in a Chipmunk aircraft have confirmed that a complete barrel roll cannot be achieved from an altitude of 60 feet with an entry airspeed of 85 knots. The pilot had been instructed and found competent to carry out barrel rolls at higher altitudes in his Chipmunk in 1966. He gave no prior warning that he was going to attempt to carry out a low level barrel roll.

The pilot was flying a night approach to a beach landing area marked by a circle of lights consisting of three motor vehicles and six flashing strobe lights. The approach was flown over water in conditions described as very dark, with no visible horizon below 500 feet altitude. The pilot reported that the approach appeared to be proceeding normally when, at what he thought was about 200 feet altitude, the helicopter suddenly yawed right and struck the water. No fault was found with the helicopter which might have contributed to the accident. Wreckage examination indicated that the helicopter contacted the water while in a right skid and that the main and tail rotors were under power at impact.

The pilots were competing in a triangular navigation exercise and were in the vicinity of the final turning point. The aircraft were orbiting in a thermal when they collided at a height of about 3500 feet above the ground. Both aircraft became uncontrollable and the pilots parachuted to safety. The investigation established that both aircraft were orbitting to the right when they collided. Examination of the wreckage showed that the right wing of VH-GXY hit the left wing of VH-GGV. The impact was on the leading edge of each wing. About two metres of the wing of VH-GXY was severed, and about one metre of the wing of VH-GGV. Neither pilot had been aware of the presence of the other aircraft in the thermal.

The pilots were competing in a race as part of the World Gliding Championships. The gliders were in a group climbing in a thermal, when they collided at about 3900 feet above mean sea level. The Italian pilot in LB did not see the other aircraft, while the Polish pilot of VH-GSO only became aware of the proximity of LB at the last moment. The collision severed the right half of the horizontal stabiliser of LB, however the pilot was able to retain control of the aircraft and landed without further damage.

The aircraft were part of a four aircraft formation taking part in a photography session. On return to Darwin it was intended that the aircraft land one after the other on the same runway. The first aircraft landed and the pilot requested a clearance to turn off the runway onto a cross strip. A clearance was not available due to other traffic on that strip and the pilot was instructed to continue taxying along the runway. As the second aircraft commenced the flare for landing it collided with the aircraft already on the runway. The flight had been poorly planned and the briefing conducted before DEPARTURE was inadequate. The pilot flying the second aircraft to land was inexperienced on the aircraft type and had no formation flying experience.

The pilots were competing in the Austraglide '86 international gliding championships. During a cross-country exercise a number of gliders were thermalling in the same general area. The pilot of D-2870 entered a thermal at about 4000 feet above ground level. His entry was made via a 45 degree bank right turn, but after turning through about 450 degrees the left wingtip of his aircraft contacted the forward under-fuselage area of VH-HNZ. This aircraft had been in a left turn with about 12 degrees angle of bank. Following the collision, both aircraft remained under control and were flown to the planned destination without further incident. Neither pilot saw the other aircraft prior to the collision. The pilot of VH-HNZ had been turning to adjust his track and had been watching other gliders well below his level. Visibility forward and downward is somewhat restricted in this particular model of glider. The pilot of D-2870 advised that he was turning towards the sun and climbing when the collision occurred.

Arrangements had been made for the aerobatic team to perform an air display over the water at Portsea, prior to the start of a large Regatta. These arrangements included advice to the organisers of the area in which the display would be made. The team carried out a successful practice session the night before. On the morning of the Regatta they proceeded to Portsea where change over was made to a discrete VHF radio frequency so that necessary radio communications could be made between team members and the display was commenced. The display sequences were completed as planned until after about 10 minutes the final sequence was commenced from a vee formation with the leader in front and numbers two and three positioned on each side of him. For this manoeuvre the aircraft were to pull up in turn, on a count of three, from level flight at about 500 feet to complete five eighths of a loop, followed by a 180 degree roll to upright and then dive to level off at 500 feet in line astern and depart the area. The first to pull up was to be number two, followed in turn by numbers one and three. For a significant proportion of the manoeuvre, pilots of the following aircraft are unable to see the aircraft ahead.

Successful conduct of the manoeuvre required that all team members perform compact loops of similar diameter. As planned and on the call of the leader the number two pilot commenced his pull up, followed by number one on a count of three and then by number three, also on a count of three. About three quarters of the way up the loop the leader realised something was wrong as he could see number two too early and called for number two to speed up. There was no response, but this was not considered unusual. As he pulled over the top of his loop the leader again lost sight of number two. The leader descended initially as planned but then levelled off, concerned about the location of number two and realising he would have to move sideways, to the left away from the shore crowds. He looked right and left and was suddenly hit from behind and above by number two aircraft. The aircraft collided at a height of about 750 feet. The number two aircraft was very badly damaged in the collision and descended steeply into the water, contacting the sail of a yacht as it fell. The leader's aircraft was significantly damaged by the collision but he was able to maintain partial control and was forced to limit engine power due the control difficulties. As a result of the damage, the aircraft was unable to maintain height.

A ditching was made in a clear area adjacent to some yachts. The pilot escaped from his sinking aircraft and was assisted by a person from a nearby yacht. He was subsequently taken on board a rescue vessel. The weather was fine with good visibility, the temperature was 18 degrees and there was a light westerly wind. The leader's aircraft was not recovered, but he reported there were no operating problems with it prior to the collision. About 70 percent of the number two aircraft was recovered. Examination of this did not reveal evidence of any pre-existing defects which might have contributed to the accident. The aircraft used for the display were all of the the same general type but had different performance capabilities. This was primarily because of different engines and aircraft weights. As was normal procedure, the leader flew the aircraft with the least performance, the aim of which was to allow a performance margin for the other two aircraft which in most team manoeuvres, had to be positioned with respect to the leader. A disadvantage in this type of manoeuvre is that it is necessary for the other two pilots to limit the engine power and ensure that they fly the same path. In his pre-display briefing the leader made a point of drawing attention to this aspect.

An amateur video was taken of the accident flight and this indicated that the total time from the start of the pull-up by the number two pilot, until the collision, was about 15 seconds. The video showed that on the pull-up, number two aircraft climbed much higher than normal and that during descent it completed some two and three quarter rolls to the left prior to the collision with the number one aircraft. Because of the relative positions of the two aircraft from the top of their respective climbs it would have been very difficult for either pilot to have seen the other in time to avoid the collision. Opinions were obtained from experienced aerobatic pilots concerning the rolling manouvers of the number two aircraft. In the main these opinions indicated that the manoeuvres were controlled and that the aircraft maintained the required line during descent. Earlier in the display sequence, a team manoeuvre with some similarity to that on which the accident happened took place. On this occasion, however, the number three pulled up first, followed by the number one, both of which then completed a manoeuvre similar to that during which the accident happened. Number two pulled up last and because he has to delay before his next sequence, he sometimes executed a large diameter pull-up followed by a diving manoeuvre which included one and a half rolls. A possible explanation for what happened on the accident sequence was that the pilot inadvertently employed a procedure generally similar to that he sometimes used earlier in the display.

The accident manoeuvre had only recently been added to the sequence, being completed three times previously, including once at an air display two weeks earlier and twice during practice sessions. In aerobatic terms the accident manoeuvre was not difficult. All pilots in the team were experienced aerobatic pilots with approvals to conduct aerobatics down to a height of 500 feet above ground level. The pilots in numbers one and three aircraft had been in the team from its formation some ten years earlier. The pilot in the number two aircraft had joined as a part-time member in 1982 and had been a full-time member since 1985. Air displays of this type require that the performing aircraft not track or manoeuvre towards spectators within a horizontal distance of 500 metres or pass within 200 metres horizontally from spectators. There were many official, spectator and competition boats in the area. The leader indicated that there were also large clear areas of water. He reported he conducted the display with the aim of remaining over the clear areas. At times there were water craft moving into the display area and he altered the path of the runs to avoid these.