Loss of control

The pilot had flown that morning from Bringelly to Kincumber to visit his brother and family. The flight was reported as normal and the helicopter as serviceable. Following lunch the pilot agreed to take his nephew for a ride "to wave a cheerio to friends at Wamberal". After takeoff he climbed to 1200 feet and set heading over Empire Bay. When he had passed Terrigal he descended quickly to 200 feet, and then flew north along the beach towards Spoon Bay. He observed that the sea was very rough, with large waves breaking on the headland and a strong southerly wind blowing. When he arrived over the headland, abeam Wamberal, the pilot initiated a turn to the right away from the land, and commenced to climb using cyclic control. His intention was to gain sufficient height during the turn so as to recross the coast and a built up area. The pilot increased the angle of bank to approximately 40 degrees, which he maintained during the climbing turn, allowing the airspeed to decrease to best climb speed. When the helicopter had climbed about 50 feet and turned through 120 degrees the pilot increased the collective pitch and power to continue the climb. He reported that the machine then appeared to buck and tend to corkscrew, without any accompanying unusual noises. It then lurched violently to the right and assumed a steep nose down attitude. There was no apparent response from application of left pedal which made the pilot believe that he had suffered a tail rotor failure. He lowered the collective control and pulled back on the cyclic in an attempt to raise the nose, which slowly came up and the lurching stopped. The helicopter was now facing towards the land, but too low and not close enough for the pilot to make an auto-rotational landing on the beach. The helicopter descended rapidly, and the pilot moved the cyclic control full back in an attempt to flare the machine for a ditching in the sea. At the same time he raised the collective and applied some power. The helicopter touched down relatively gently on the water, but sank almost immediately. Both occupants were able to evacuate through the cabin bubble and stayed together for about half an hour. When help did not appear to be forthcoming, the pilot, who was the stronger swimmer, decided to to swim against the current and surf to the shore to find help. After assuring himself that his nephew was alright, and floating satisfactorily, he set off for the beach, where he was assisted by police and other people who had witnessed the accident. An immediate air search by helicopters failed to locate the passenger in the sea. His body was eventually washed up on the shore about a week later. Attempts to locate and salvage the wreckage of the helicopter were unsuccessful, except for the recovery of the tail rotor and part of the tail boom. An inspection found no evidence that these components had suffered any pre-impact malfunction or failure. As the helicopter was in a steep turn to the right the pilot may have failed to recognise and then compensate, by applying left pedal, for the yaw, also to the right, which would have occurred as he increased the collective pitch. It is likely that he did not apply left pedal until the roll was well developed. Recovery should have been possible by rolling the helicopter level with cyclic and balancing with the pedals. When the pilot decreased the collective pitch, torque to the main rotor was reduced, slowing the yawing motion. As he was still applying full left pedal the tail rotor would have now taken effect, slowing or stopping the turn just prior to the helicopter entering the sea. The wind would have been creating a considerable amount of turbulence over the headland, but not enough to cause loss of control. Mechanical and/or tail rotor failure was considered unlikely. Without the tail rotor anti-torque effect the fuselage should have rotated at a fast rate to the right, and still been rotating at impact, even with the collective lever lowered. This accident was not the subject of a formal on-scene investigation.

The helicopter was carrying 360 litres of water based Herbicide and about 90 litres of fuel when it took off the second time that morning. Flying activity had commenced later than planned due to local fog and low cloud in the area to be sprayed. A level area on the bank of the creek was used for loading and the pilot availed himself of the open space over the creek to accelerate before commencing a climb out to the treatment area. This technique involved an immediate loss of ground effect as the helicopter moved off the elevated creek bank. The aircraft failed to achieve translational flight and the pilot "sensed" a power loss as it descended towards the fast flowing creek. The left skid fractured when it struck a log protruding from a low island. The impact rotated the helicopter through 180 degrees causing the pilot to lose sight of the only reasonable landing area on the opposite bank. The helicopter settled and balanced on sloping ground whilst the engine wound down due to fuel starvation caused by a fuel feed line separation in the initial impact. As the rotor system slowed down, the helicopter fell onto its damaged left side. The pilot was using a pad that was not suitable for the operation. The pad was limited by a fence line at one side which was high enough to prevent operations over it at high all up weights and a creek at the other side which caused an immediate loss of ground effect as soon as the helicopter moved off the pad. Trees and high terrain prevented other take-off options. The pilot was using the potential energy of pad height above the creek to gain airspeed quickly and early translational flight. On this occasion he did not obtain translational flight and the helicopter descended into the creek. The pilot overpitched the main rotor blades causing a loss of RPM and power. The engine and fuel components were examined in detail and no mechanical reason for a power loss was discovered. The helicopter was fitted with a spray system that was not authorised for the type in Australia, the performance combination of helicopter and spray system was unknown.

The pilot had arranged to take some friends for a flight and to have a flight check with a flying instructor beforehand. Initially the instructor was unhappy with the pilot's circuit planning and landings. The pilot was given some information on the aircraft and a circuit diagram and another check was scheduled for the following day. On the next day the instructor gave the pilot a substantial briefing prior to the flight. The pilot's flying had improved significantly and the instructor assessed him as competent to fly the aircraft. After a scenic flight in the local area the aircraft joined the circuit on right downwind for runway 12. Surface wind at the time was 120` at 15 knots. The aircraft was observed to touchdown aligned with the centreline, near the runway threshold, and a short time later was seen to bank left. It touched down on the grass within the left hand side flight strip heading about 40` to the left of the runway direction. At around this time the pilot elected to go-around and he applied full power. From this point the aircraft made a gentle left turn until it was tracking 330`. Initial impact with the ground occurred when the aircraft was descending at an angle of approximately 15` and banked at least 30` to the left. Indicated airspeed at this time was 35 knots. The aircraft landed heavily on the nosegear and rolled backwards before coming to rest 20 metres from the point of initial impact.

The pilot was herding a bull along a shallow valley and he decided to hover behind it. He entered the hover about 20 feet above the trees from a slight descent and with some right lateral movement. The heading of the helicopter was westerly in a wind from the south at 10 knots. The pilot then initiated a pedal turn into wind. The helicopter stopped turning when nearly into wind and the pilot added almost full left pedal to complete the turn. The helicopter then began to rotate smoothly and rapidly to the right, and the pilot applied full left pedal in an attempt to stop the yaw. As the application of full left pedal had no effect, the pilot closed the throttle but there was no perceptible reduction in the rate of turn. The helicopter rotated through at least 360 degrees before striking the trees and crashing on its side. The on site investigation did not reveal any mechanical defects or anomalies that would have contributed to the loss of tail rotor control.

At about 50 knots during the landing roll, the pilot discovered that the right wheel brake had failed. The aircraft had swerved to the left when both brakes had been applied. With the rudder control ineffective at low speed, the pilot intentionally ground looped the aircraft to avoid running into the perimeter drainage ditch. The right main gear collapsed and the right wing was damaged as a result of the ground loop. Examination of the right brake unit revealed that the master cylinder was corroded internally, allowing brake fluid to bypass the seal.

Circumstances:

The student was making his second solo flight. During the approach to land the student realised the helicopter was above the desired approach path. He elected to land on the grass surface a short distance beyond the helipad as he has been taught. The wind was blowing at 10 to 15 knots from the northeast, with the landing conducted in an east south east direction. The pilot said he began to enter the hover and turn the helicopter to the left, into wind. The aircraft continued the left turn, during which it began to move rearwards. The tail skid dug into the landing surface, resulting in the helicopter rolling to the left and striking the ground.

Significant Factors:

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

1. Improper compensation for wind conditions.

2. Pilot encountered unforeseen circumstances beyond his capability.

3. Improper operation of primary flight controls.

The pilot was engaged in a local flight and was practising hovering at various altitudes. With the helicopter heading 320 degrees, the pilot commenced a slow descent from about 15 feet. The wind was 310 degrees at 10 knots gusting to 20 knots. While descending to about three feet, the helicopter began a slow yaw to the right through 30 degrees. The pilot elected to let the yaw continue and carry out a 360 degree pivot turn. After turning through about 100 degrees, the pilot stopped the descent by increasing power and pulling on collective. Coincidentally, the pilot reported that the helicopter then rapidly increased the rate of turn despite the application of full left anti-torque pedal. After about one turn, the pilot said he lost all control and following about another 3-4 turns, the helicopter crashed onto its left side at about 45 degrees angle of bank. During the sequence several mainrotor blade ground strikes occurred. The pilot said that he was also aware of overcontrolling on the cyclic control and pulling on collective at one stage in his efforts to control the helicopter. The investigation could not find any mechanical cause for the loss of control. Both the pilot and passenger reported that at no time did the passenger touch the controls or have his feet near the anti-torque pedals. Studies on helicopter tail rotor vortex rings indicate that these rings can form with a relative wind from 220 to 320 degrees at 10 to 25 knots, with further indications that the worst areas are around 250 and 290 degrees relative. In addition, the conditions suitable for the formation of a vortex ring are most favourable when the tail rotor is moving to the left, as in a right pedal turn. At the point where power was increased, the tail rotor was in a critical position with a relative wind from 250 degrees and strength from the left of between 10 and 20 knots. If the tail rotor was close to a vortex ring state, increased power would, through the torque effect, produce a marked right yaw. Due to the vortex ring, left pedal input would have no effect.

Following the completion of a flight in the local training area the aircraft was landed on Runway 36 at Roma. At the time the wind was north-westerly at five to eight knots and there was a thunderstorm to the south-west of the field. During the landing roll the wind velocity suddenly increased to about 25 knots. The aircraft swung to the left and ran off the side of the runway. The pilot was unable to regain directional control but shut down the engine before the aircraft ran into a ditch. This accident was not the subject of an on-site investigation.

The pilot made an early morning DEPARTURE from the station strip. He reported that as the aircraft accelerated past 65 to 70 knots he selected gear up, however, the gear did not immediately retract and the gear safety override warning activated. The gear then retracted but shortly afterwards he heard a "clanging" noise which he assumed was coming from the engine. He immediately reduced power, selected gear down and turned back towards the strip for a forced landing. The turn was commenced from about 150 to 200 feet above the terrain, but after turning through about 160 degrees the aircraft stalled. The aircraft collided with the ground and after a ground run of approximately 122 metres the left wing struck two trees which yawed it through 180 degrees. The gear collapsed and the aircraft skidded backwards for 32 metres. An examination of the aircraft failed to detect any pre-impact malfunction with the engine or propeller. The source of the "clanging" noise reported by the pilot could not be established, although the post impact damage may have destroyed the evidence of the noise source. There was no other evidence of power loss. The pilot had been working very long hours over the preceding three days and was fatigued. It is possible that the level of fatigue had a deleterious affect on the pilot's ability to rapidly and correctly assess the action required. The pilot mistakenly identified a noise from an unknown source as an engine or propeller malfunction and reduced power by a substantial amount. The pilot then selected gear down and attempted a turn back manoeuvre from only 150 to 200 feet above ground level without increasing power. The aircraft subsequently stalled. The activation of the gear safety override system probably added to the confusion and reinforced in the pilot's mind that the aircraft had a serious malfunction.

Circumstances:

The pilot reported that, very shortly after liftoff from Runway 15, he heard a loud bang. He was advised by one of the passengers that the baggage door, situated on the upper right side of the fuselage above the wing trailing edge, had opened. Apart from a high noise level, aircraft behaviour seemed unchanged. The pilot said that he recalled a section in the Pilot's Operating Handbook which stated that the flight characteristics of the aircraft would not be affected by an unlatched door in flight. He reassured the passengers to this effect. As the aircraft climbed through about 150 feet, with landing gear still extended and flap still at 15 degrees, the pilot reduced engine power to what he thought was about 1800 RPM and turned left to land on the eastern section of Runway 06. He had assessed there was insufficient length of Runway 15 remaining to land straight ahead. As the aircraft turned with 15-20 degrees angle of bank, it was seen to adopt a nose-low/left-wing-low attitude and impact the runway. The pilot reported that the aircraft was in a nose low attitude and close to the ground when it rolled left. He applied right aileron and full power but was unable to prevent the left wing striking the runway surface. The initial impact was 83 metres left of the Runway 15 centreline on a heading of approximately 090 degrees magnetic. The left wing tip contacted the ground first followed by the nose. The aircraft skidded 51 metres before coming to rest. The landing gear collapsed during the impact sequence. Neither the pilot nor any of the passengers reported hearing the stall warning operate prior to impact. The surface wind at the time of the accident was estimated to have been from the south-east at 10 15 knots. There was no significant turbulence. On inspecting the aircraft, the baggage door was still attached to the airframe via its two hinges. The hold-open stay was broken. No fault was found with the door locking mechanism, either through the external locking handle, or the internal lever. The royalite plastic lining, including the protective cover for the internal lever, and insulation material had been torn from the inner face of the door and were found adjacent to the runway. The internal locking knob was securely stowed in the locked position. The pilot reported that it was his habit to check the door as he stepped on to the wing to enter the cockpit. As far as he could recall, the door was locked prior to the flight. Photographs taken of the aircraft shortly after the accident appear to show the external locking handle in the stowed position. However, it is possible for the handle to be stowed and the locking pins to be located outside the fuselage skin, thus leaving the door unlocked. Because of accident damage, the operation of the stall warning system could not be tested. The pilot could not recall the speed of the aircraft during the turn. It seems probable, in view of the aircraft configuration and the engine power setting, that the airspeed was closer to, rather than substantially higher than, the basic stalling speed of the aircraft. The luggage door is positioned on the upper right side of the fuselage above the wing trailing edge. It is hinged on the top of the fuselage with the hinge line parallel to the aircraft centreline. With the door open, the airflow pattern over the rudder/tailplane could be altered. There is no reference in the Aircraft Flight Manual or the Pilot's Operating Handbook to operations with the baggage door open. However, there is reference in the Operating Handbook to the actions to be taken in the event of the cabin door becoming unlatched in flight (flight characteristics unaffected). It was this that the pilot recalled at the time of the occurrence and which led him to reassure his passengers that there was no cause for concern. This knowledge could also have influenced his decision to turn left and land on the remaining section of the other runway.

Significant Factors:

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

1. The pilot probably did not adequately check the security of the baggage door before flight.

2. The baggage door opened as the aircraft took off.

3. In the subsequent landing, for reason(s) which could not be established positively, the pilot was unable to prevent the left wing from contacting the ground.

Recommendations:

There have been a number of accidents to M-20 aircraft involving in-flight opening of the baggage door. In at least two cases overseas, the accidents were fatal and involved loss of control at low speed. In a non-fatal accident overseas, airspeed and vertical speed indications became erratic after the baggage door came open and contributed to the pilot stalling the aircraft. In Queensland in 1984, in circumstances strikingly similar to the accident under discussion here, the pilot of an M-20 aircraft was attempting a landing after the baggage door opened shortly on take-off. He reported that the aircraft "fell away" as it crossed the end of the runway. The aircraft struck the runway surface right wing first, sustaining substantial damage. The pilot did not hear the stall warning sound. These examples indicate that the influence of an open baggage door on the flight characteristics of the Mooney 20 could be significant. Following the first fatal accident, a Service Bulletin dated 28 September 1988 was issued by Mooney Aircraft Corporation (SBM20-239) applying to various serial numbers of M20J Models to prevent in-flight opening of the baggage door. The Bulletin contained instructions to modify the inside latch on the baggage door and was incorporated in Australian Civil Aviation Authority Airworthiness Directive AD/M20/44 dated 23 February 1989. The baggage door inside latch was not a factor in this accident. However, that such modification action to the latch was considered necessary to prevent in-flight opening of the door raises a number of other aspects. These include:

1. Whether any warning device (such as a warning light in the cockpit) is necessary to alert the pilot that the door is not locked.

2. Whether any special technique is required during approach and landing to ensure safe control of the aircraft with the baggage door open.

3. Whether any warning should be included in the Aircraft Flight Manual about possible control problems in the event of the baggage door opening during flight.

4. Whether any flight test program is necessary to determine the handling characteristics of the aircraft with the baggage door open. It is recommended that the Civil Aviation Authority examine the above aspects with a view to ensuring the maximum safety level for Mooney 20 operations consistent with practicable economic considerations.