The ATSB has released its final report on the midair collision at a Bankstown Airport inbound reporting point that led to the introduction of Class D airspace at the former GAAP aerodromes.
On December 18 2008 a Liberty XL2, registered VH-XLY, collided in midair with a Cessna 152, registered VH-FMG, near Casula, NSW. The Liberty was engaged in a pilot licence flight test; the Cessna was engaged in pilot training. The pilots of both aircraft were operating in uncontrolled airspace under VFR in good weather.
The collision severely damaged the Cessna, which descended in an uncontrolled manner before impacting terrain, fatally injuring the two occupants. The Liberty was able to continue flying and landed at Bankstown Airport.
A number of the limitations of the see-and-avoid principle likely combined to limit the ability of the Liberty to re-sight the Cessna once they had lost sight of it. This accident is a salient reminder to all pilots that there are limitations with visual flight procedures and that, regardless of pilot’s experience, they need to remain vigilant at all times.
History of the flight
At about 1125 on the day of the accident the C152 and Liberty XL2 collided at an altitude of 1500ft AMSL over the Sydney suburb of Casula, which was located 10km south-west of Bankstown Airport. The Liberty had departed from Bankstown at about 0930 that morning
to conduct a PPL flight test. On board were a student pilot under test and a CASA approved
testing officer (ATO). At about 1030, the Cessna departed Bankstown to conduct a local training flight in the training area. On board were a student pilot and an instructor.
At 1103, the Liberty passed Prospect Reservoir and tracked south towards the Bringelly area before entering the Camden CTAF area. The recorded radar data showed that, as the Cessna was returning to Bankstown from the training area, the Liberty departed Camden and maintained the runway heading of 060° into the training area, before tracking towards the 2RN radio transmission tower, the southern inbound reporting point to Bankstown.
The radar data indicated that, about midway between the Camden CTAF area and 2RN, the Liberty conducted a left turn through 180°, followed by a turn to the right through 180°. The Liberty then flew in a northerly direction and flew above and behind the Cessna, which was about 600 ft below and tracking towards the 2RN reporting point. The Liberty continued to track north for a short time, then turned to the right, descended and tracked towards 2RN.
Following the merging of the radar returns in the vicinity of 2RN, one radar return faded and the other return continued to the vicinity of Bankstown Aerodrome. The fading radar return corresponded to the Cessna’s initial track towards Bankstown. Witnesses reported that the Cessna descended almost vertically before impacting the rear of a house in Casula.
Witnesses
At least 12 people witnessed either the collision between the two aircraft or the events immediately after the collision. One witness reported seeing a white aircraft with red stripes (C152) flying towards the city. That witness reported then seeing a second aircraft (Liberty) approaching the first aircraft from behind at an angle of about 45° and that it was travelling faster than the first aircraft. The witness reported observing both aircraft for about 30 seconds before the collision and reported that, “the nose of the second plane just collided in the left side of the tail of the first plane’. Another witness reported that they, “saw the tail of the red and white plane hit the front bottom part of the right hand side of the nose of the other plane”.
All of the witnesses that observed the collision reported that neither aircraft appeared to be manoeuvring or to take evasive action prior to the impact, and that both aircraft appeared to be flying in a straight line with their wings level. Following the collision, the witnesses described the Cessna pitching nose-down to a vertical attitude and that the tail of the aircraft broke off, but remained attached by what appeared to be cables. It was reported that the tail appeared to be rotating rapidly behind the Cessna as it descended vertically to the ground.
Recollection of events - Liberty pilots
Student under test
The student recalled entering the Camden CTAF and joining the circuit for runway 06 on the downwind leg. After a touch-and-go the student departed on the upwind leg of the circuit, initially maintaining 1300 ft AMSL until clear of the Camden area, and then climbing to 2500ft en route to the training area. The student stated that, after two steep turns at 45° angle of bank, first to the left and then to the right, he was asked to recover from an ATO-initiated aerodynamic stall. The ATO then asked him to return to Bankstown Aerodrome via 2RN.
The student reported that as he turned the aircraft to track towards 2RN, the ATO advised him of Cessna traffic on the right of the Liberty at about 45° to the nose. The student stated that he could not see the traffic initially but when he did sight it, it seemed to be higher than the Liberty, and appeared to be turning. He also reported that the ATO asked him to reduce speed to a slow cruise configuration of about 80 to 90 kts because of the traffic ahead of the Liberty, and that he complied with that request. In a later interview, the student reported that both he and the ATO lost sight of the Cessna after its initial sighting, and that they discussed losing sight of it.
The student indicated that he was busy preparing for the entry to the Bankstown control zone and could not specifically recall what the ATO was doing during the time the aircraft tracked towards 2RN. He did recall, however, that the ATO had some paperwork on his lap during the flight. The student also reported that he tuned the aircraft’s radio to the Bankstown ATIS frequency, noted the appropriate information and then made a radio broadcast to Bankstown at 2RN. He indicated that the air traffic control (ATC) response to the radio broadcast was not what he was expecting and he was about to ask the controller to repeat the response, when he saw the flash of an aircraft’s tail fin in the lower right corner of the front windscreen, followed almost immediately by a loud bang.
Approved testing officer
The ATO stated that during the flight back through the training area, he only sighted one aircraft, a Cessna. When queried about the Cessna, the ATO indicated that it was sighted after the student pilot had completed the steep turns and stall recovery, and that the Cessna was higher than their aircraft. He reported indicating the location of the Cessna to the student pilot. The ATO indicated that as the aircraft tracked towards 2RN, he briefly completed the sections of the test report form that dealt with steep turns and stall and recovery before closing his notebook on his lap. He then monitored the student’s preparation to report inbound. Sometime later, having himself lost sight of the Cessna, the ATO asked the student if he still had the Cessna sighted, to which the student had replied ‘no’. There was no evidence that either the ATO or the student pilot actively sought
to reacquire the Cessna at that time.
The ATO indicated that the Liberty normally cruises at about 110 to 115 kts, and that he had asked the student to slow the aircraft to about 90 kts after they had initially sighted the Cessna. The ATO reported that the student correctly obtained the ATIS for Bankstown before making a radio broadcast at 2RN.
The ATO indicated that almost immediately after the student’s inbound radio transmission, the Liberty collided with a Cessna aircraft. The ATO recalled the Cessna appearing in his 2 o’clock high position, at an angle of 45° to the right of the nose of the Liberty, descending at an angle of about 15° down to the horizon and in a left turn. He further reported that the impact, “was nearly a head-on collision”, and that he estimated the Cessna’s heading was about 290°.
Injuries to persons
The student and the instructor of the Cessna were fatally injured. The student of the Liberty did not report any injuries as a result of the accident. The ATO reported that he had a cut to the thumb on his right hand and some bruising to his left knee.
Personnel information
Cessna pilots
Student: SPL, 24.2 flying hours, Class 2 Medical.
Instructor: CPL with Grade 3 instructor rating, 423.9 flying hours (127.2 instructor hours), Class 1 Medical
Liberty pilots
Student: SPL, 88.2 flying hours, Class 1 Medical.
ATO: CPL with Grade 1 multi-engine instructor rating and a number of testing delegations from CASA, more than 27,000 flying hours.
Aircraft information
Both aircraft were maintained in accordance with CASA-approved systems of maintenance. The maintenance releases for both aircraft were reviewed and no outstanding defects were evident. A review of both aircraft’s logbooks did not reveal any outstanding defects or maintenance anomalies.
Meteorological information
The student pilot under test and the ATO both reported that the weather conditions were adequate for the VFR flight, with a visibility in excess of 10 km and a cloud base of about 3500ft.
Communications
Cessna
Examination of the recorded ATC transmissions during the time leading up to and after the collision revealed that the only transmission from the Cessna was associated with its departure to the training area about one hour before the collision. There were no inbound transmissions recorded from the Cessna prior to the collision. An examination of the damaged radio from the Cessna revealed that the active frequency was selected to 132.80 MHz (Bankstown ATC) and that the standby frequency was selected to 120.90 MHz (Bankstown ATIS). Impact damage precluded functional testing of the radio.
Liberty
The student under test reported making a radio broadcast approaching 2RN, but that the response was not as expected. A second broadcast seeking clarification was about to be made by the student when the collision occurred. The ATO reported that a correct inbound broadcast was made by the student under test, that it was acknowledged by Bankstown ATC and that the Liberty was cleared to enter the circuit on the crosswind leg.
Examination of the frequency revealed that there was no inbound radio transmission from the Liberty before the collision. An examination of the recorded information on the SMC frequency of 119.90 MHz revealed that the inbound broadcast from the Liberty was made on this frequency and indicated that the SMC controller advised the student that the broadcast was on the Bankstown ground frequency and to contact the tower on frequency 132.80 MHz. The location of the Liberty during those transmissions was to the south-west of 2RN and not in accordance with the published instructions for operations at Bankstown.
The first recorded inbound transmission on the tower frequency from the Liberty was the MAYDAY broadcast from the ATO following the collision. The ATO reported that he did not have to change the radio frequency or transmitter selector before making the MAYDAY broadcast. Examination of the upper communications radio revealed that the active frequency was 132.80 MHz (Bankstown ATC) and that the standby frequency was 124.55 MHz (Sydney Radar - South). The lower communications radio had 119.90 MHz (Bankstown ground) selected as the active frequency and 120.90 MHz (Bankstown ATIS) as the standby frequency.
Recorded radar data
The original radar recordings indicated that the Cessna and Liberty each had functioning transponders leading up to the collision and displayed secondary surveillance radar returns that indicated the selection in each case of transponder code 1200.
Wreckage and impact information
Overview of wreckage
Cessna
The Cessna was painted white with large red stripes on its fuselage that extended to the lower section of the tail area. The vertical fin was painted white with a red and black accent on each side, which extended around the leading edge of the fin. The Cessna impacted a house and the ground in a vertical, nose-down attitude.
There was significant compression damage to the leading edges of both wings, consistent with impact with the structure of the house and the ground A detached section of the Cessna’s vertical fin came to rest between two houses in an upright position after impacting the roof of a house to the west of the main wreckage. Examination of that piece of wreckage revealed blue paint transfer on the left side of the mass balance weight and upper section of the vertical fin.
The tail of the Cessna had separated about midway between the cabin and tail. The rudder, elevator and trim control cables remained intact, resulting in the separated section remaining connected to the main wreckage following the collision. The horizontal tail surfaces were relatively undamaged and remained attached to the tail of the aircraft. Control continuity to all flight controls was confirmed and there were no abnormalities with the engine of the Cessna. Examination of the portion of the vertical fin revealed that this damage was consistent with an impact with the Liberty’s exhaust pipe.
Liberty
The Liberty was painted white with blue, grey and aqua accents located along the fuselage. The propeller was painted white with red tips and black trailing surfaces. Damage to the Liberty consisted of impact damage to the right side of the aircraft’s nose cowling, with associated paint transfer and crush damage to the single exhaust pipe that was located on the lower-right side of the engine cowling. There was also damage to the right side of the nosewheel, paint transfer on the leading and trailing surfaces of the impact-damaged propeller and a break on the lower-right side of the cockpit windscreen.
Organisational and management information
Collision avoidance
Both aircraft were returning to Bankstown Aerodrome and were operating in Class G airspace at the time of the accident. The pilots were based at Bankstown Aerodrome and all four had experience, albeit to varying degrees, with the normal procedures for entry to the Bankstown control zone via 2RN.
Midair collision research
Midair collisions are relatively rare events in aviation. Research has indicated that when a midair collision occurs, the most likely place for it to occur will be in the circuit area of an aerodrome. Very few midair collisions occur outside of this area. In-flight visibility was found not to be a contributing factor in most, if not all of the collisions reviewed by the ATSB. The majority of the midair collisions involved one aircraft colliding with another from behind or converging from similar directions at an angle of less than 30°.
ANALYSIS
Introduction
There were some inconsistencies in witness recall, specifically that of the ATO, whose recollection of the immediate aspects of the collision, and of some of the events leading up to it, was at variance to that of the student, and inconsistent with the recorded radar data and the physical evidence. Although unable to be explained fully by the investigation, it was possible that the trauma associated with the event may have affected the ATO’s recollection of events.
The investigation was left to consider two major issues – were the pilots of the Cessna aware of the Liberty’s position and the potential collision risk, and were the pilots of the Liberty able to sight and maintain sight with the Cessna to avoid a collision.
Collision analysis
The collision occurred when the Liberty, which was travelling faster than the Cessna, impacted the left of the Cessna’s tailplane. The position of the damaged single exhaust pipe on the right of the Liberty’s engine cowling and corresponding impact mark on the left side of the tail of the Cessna served as a point of reference on which to orientate both aircraft at the time of the collision. When coupled with the tyre mark on the left side of the Cessna’s tail, the vertical and horizontal orientation of the Liberty at that time was able to be determined with a high degree of certainty. The location of the Cessna’s vertical fin at that time was confirmed by the paint transfer between both aircraft.
The impact from the Liberty deflected the Cessna’s vertical fin to the right and resulted in a twisting failure of the empennage of the Cessna towards the right, before its separation about midway between the cabin and tail. Once that structure failed and separated from the remainder of the aircraft, the flight control cables were the only connection between the tail and the remainder of the aircraft. The witness reports of the uncontrolled rotation of the separated tail section were explained by the vertical airflow over that aerofoil during the vertical descent.
The separation of the Cessna’s tail to the right resulted in the tail commencing to rotate in that direction and possibly completing one revolution before the Cessna’s tail-mounted white navigation light came into contact with the left underwing surface of the Liberty. That impact liberated the navigation light, which was the first item found along the wreckage trail. It is also possible that the student under test in the Liberty, as the handling pilot at the time of the collision, could have rolled the aircraft to the left as an instinctive reaction to the rapid bloom of the Cessna in the Liberty’s right windscreen, placing the Liberty’s left wing lower and closer to the Cessna’s tail.
The ATO’s account of the collision as being “...nearly a head-on collision” was unable to be reconciled with the physical evidence of the contact between the two aircraft.
Cessna pilots’ awareness of the Liberty
The position of the Liberty in the rear left quarter of the Cessna would have meant that to sight that aircraft the Cessna pilots would have to look behind and over their respective left shoulders. As the Liberty pilots gave their inbound call on the incorrect frequency, the pilots of the Cessna were not alerted to its close proximity and therefore had no reason to look towards the rear of their aircraft.
Liberty pilots’ sighting of the Cessna
The indication by the ATO to the student of a Cessna to the right of the Liberty after the steep turns and stalling exercises was consistent with the relative positions of both aircraft on the recorded radar data. The student’s recollection that the Cessna was higher than the Liberty and appeared to be turning contrasted with the recorded data, which showed that the Liberty was higher than the Cessna at that time. The identification by the student under test of red areas on the Cessna, and its identification by the ATO as a Cessna confirmed its conspicuity against the background over which it was flying, despite its small angular size at the time.
The investigation concluded that the Liberty pilots sighted the Cessna before the collision. Having sighted the Cessna, the pilots in the Liberty then had to consider if it presented a potential collision threat to their aircraft. The recollection by the ATO of asking the student under test to slow the aircraft, and student’s reported compliance with that request, was an indication that they possibly perceived the Cessna as a traffic risk as they approached 2RN. However, despite the recall of both pilots, the recorded radar data showed no reduction in the Liberty’s groundspeed. It is possible that the student pilot overlooked the request to slow down as he was preoccupied with preparing for the entry to the Bankstown control zone.
The ATO, as the pilot in command of the Liberty, retained ultimate responsibility for ensuring separation from the Cessna. If the ATO considered that a speed reduction was necessary, for whatever reason, he could have repeated the request or have intervened to take control of the aircraft to slow down or otherwise manoeuvre. Neither took place, and the investigation was unable to reconcile the difference between the pilots’ recall and the recorded radar data in respect of a reduction in the speed of the Liberty. Had the Liberty reduced speed as it tracked towards 2RN, it would have in all probability not collided with the Cessna.
As the Liberty tracked towards 2RN, the student under test would have become focussed inside the cockpit in preparation for the entry to the control zone. The ATO’s focus was also inside the cockpit, as he completed the test report form after the stalling and steep turns. Although the time taken would have been short, it was possible that, for some time, neither pilot was looking outside for the Cessna. Overall, it is reasonable to conclude that, once the ATO lost sight of the Cessna, a number of factors would have made its reacquisition difficult for the ATO.
Given that the pilots in the Liberty both sighted the Cessna as they turned towards 2RN, the discussion between the Liberty pilots about the location of the now unseen Cessna most likely took place prior to the student’s inbound radio broadcast. The combined duration of the inbound call by the student under test, and response by Bankstown Ground, meant that it was possible that there was sufficient time for the Liberty pilots to have resolved the collision threat before arriving at 2RN. However, there was no evidence of positive action by the crew of the Liberty to attempt to re-sight the Cessna and resolve the developing threat of collision.
Radio broadcast on the incorrect frequency
It may take up to 12.5 seconds to process the information presented and act upon it in order to avoid a collision. If there is no alert to the location of traffic, the probability of detecting it is low, until a short time before the collision when the target increases in visual size. When the Cessna was 12.5 seconds away from the collision, it still presented a very small angular target to the pilots of the Liberty. Once the time to impact reduced below 12.5 seconds, the likelihood of avoiding a collision was reduced for both aircraft.
The combination of the student pilot’s inbound radio broadcast and response from the SMC would have impacted on his attention to other tasks during the more than 12 seconds taken by those calls. The little more than two seconds between the end of the SMC’s transmission and the estimated time of the collision minimised any hope that the student might have avoided the collision. The student under test was most likely processing the unexpected response from the SMC and, having realised his error, was possibly in the process of switching the radio to the aerodrome control frequency at that time. That would explain why the ATO did not have to change radios to broadcast the MAYDAY call.
An inbound broadcast by the student under test on the correct frequency would have alerted the pilots of the Cessna of the presence of the Liberty virtually coincident with the commencement of the nominal 12.5-second collision avoidance response time, leaving the Cessna pilots the minimum time to acquire the Liberty and avoid the collision. Any delay in that acquisition decreased the opportunity to avoid the collision. As the Liberty was in an unexpected position behind the Cessna, and therefore more difficult to sight, it was unlikely that a broadcast by the student under test on the correct frequency would have allowed the Cessna pilots to react in time and avoid the collision.
The Cessna pilots may have been about to transmit their inbound report on the ATC frequency. However, even if that call had been made it would not have alerted the pilots in the Liberty of the Cessna’s position as the Liberty was on the incorrect frequency.
Liberty cockpit vision obstructions
As the headings between the two aircraft were about 20 degrees to 30 degrees different, with the Cessna passing from the right to left of the Liberty, this could have given the impression that the Cessna was turning in front of the Liberty. Although possibly explaining the Cessna’s apparent left turn as recalled by the ATO, it does not explain the ATO’s recollection of the Cessna being above the Liberty and almost tracking for a head-on collision.
It is therefore possible that the obstruction posed by the forward right cockpit pillar obstructed the ATO from sighting the Cessna, until its angular size exceeded the width of the pillar. At that time the collision was almost certain. The investigation concluded that the visual obstructions in the Liberty cockpit impacted on the likelihood that the Liberty crew might sight the Cessna.
Other issues
It could be expected that during a flight test a student would demonstrate flight manoeuvres consistent with their training, including the conduct of a comprehensive lookout. Although testing officers are allowed a certain degree of discretion in how they conduct a flight test, the conduct of two steep turns as a clearing exercise and immediate conduct of the stall sequence was difficult to reconcile with the immediate airspace being ‘clear around and below’ in accordance with CASA’s Flight Instructor Manual. Not least, the proximity of the Cessna 966m ahead and 500ft below the Liberty would suggest that in this case any lookout possible during the steep turns had been ineffective.
FINDINGS
Context
From the evidence available, the following findings are made, however they should not be read as apportioning blame or liability to any particular organisation or individual.
Contributing safety factors
- The ATO and the student under test in the Liberty initially sighted the Cessna, but subsequently lost sight of it.
- A number of the limitations of the see-and-avoid principle likely combined to limit the ability of the ATO and student under test in the Liberty to re-sight the Cessna once they had lost sight of it.
- The instructor and student pilot in the Cessna were probably not aware of the proximity of the Liberty.
Other safety factors
- The student under test in the Liberty made the inbound broadcast to Bankstown Aerodrome Control on the incorrect frequency.
As a result of a number of midair collisions in Australia, CASA conducted two reviews into operations at General Aviation Aerodrome Procedure (GAAP) aerodromes. Although this accident occurred in Class G airspace (uncontrolled airspace), due to its proximity to an inbound GAAP reporting point, CASA referred to it in those reviews. Those reviews made a number of recommendations to enhance the provision of GAAP procedures education and training and on the current implementation of GAAP. On June 3 2010, CASA introduced full Class D airspace at all aerodromes that previously operated under GAAP in Australia.
CAPTION: The C152 wreckage and crash site. (ATSB)