Crash pilot’s name may be cleared—23 years later

Press, 12 April 1984, Page 21

Crash pilot’s name may be cleared —23 years later

By

ERIC BEARDSLEY,

information officer, University of

Canterbury

A long-standing difference between the Air Accidents Investigation Office of the Ministry of Transport and a University of Canterbury metallurgical expert, Professor L. A. Erasmus, may be solved once and for all by further laboratory investigations at Ham made possible by a $2OOO Civil Aviation research grant. The difference of opinion, which has parallels with the disputes over the Air New Zealand DC-10 crash in Antarctica, is about the cause of the crash of an Aero Commander plane, ZK-BWA, owned by Bay of Plenty Airways, with the loss of the pilot and five passengers, on Mount Ruapehu on November 21, 1961.

landing very heavily at Tuelakes airport, Oklahoma city, in 1958. The damage was not detected during the subsequent repair, and the weakening of the starboard wing structure would have resulted in additional load being transmitted to the main (front) spar structure, leading to the initiation of a fatigue crack in the front spar lower cap. The report concluded that, everything else being equal, the starboard wing was probably capable of remaining intact and there was therefore only one conclusion: that the starboard propeller and possibly an adjacent portion of the bottom of the fuselage struck a rock outcrop on top of the mountain. Vibration caused by the damaged propeller, coupled with violent turbulence, caused the fatigue crack in the front spar lower cap to propagate rapidly to failure, with consequent separation of the starboard wing. The verdict caused dismay among some pilots and an informal group was established to try to clear Mr Bartlett of blame. Eventually, nine years after the crash, members of the group approached Professor Erasmus, now head of the Mechanical Engineering Department at the University of Canterbury, to seek his assistance. He had arrived in New Zealand from England in 1965 and had no knowledge of the crash. But after reading the accident report he agreed to examine a section of the front spar lower cap because an insinuation in the report troubled him. This was that the fatigue crack in the spar cap could have become dormant — contrary to the normal behaviour of fatigue cracks. If the fatigue crack had nucleated and begun propagation under normal operating loads, propagation would normally have continued at an increased rate unless operating loads were appreciably reduced. Unfortunately, the sections of

Just before noon that day the twin-engine aircraft was over the north-east slope of Ruapehu when the starboard wing, complete with engine, was seen to fall from the fuselage. As the various structural sections plunged to the ground, the fuselage burst into flames and the fuselage and port wing were almost completely destroyed. The starboard wing, engine, and propeller were found as a complete unit on a snow-covered slope at 7400 ft. Subsequent investigations have been based largely on what this unit yields to the expert eye, aided by the electron microscope, and on a series of four photographs recovered from a damaged camera belonging to one of the passengers. The University of Canterbury investigation, which caused the initial crash report to be withdrawn, also resulted in the award to Professor Erasmus of the Wigram memorial medal of the New Zealand division of the Royal Aeronautical Society. The first air accident report attributed the crash to a decision by the pilot, Alfred Bartlett, to fly close to the summit of Ruapehu, where turbulence or some pilot manoeuvre caused the starboard propeller to hit a rock outcrop. That investigation showed up long-standing fractures in the starboard, rear spar upper and lower caps (the wing load-carrying structure.) It was surmised that they were caused initially by the plane

fractured spar cap had been mislaid but the accident investigators sent Professor Erasmus tip sections of the starboard propeller blade instead. On examining them he soon found a major flaw in the initial report. Had the blades hit an outcrop of rock on top of the mountain, fine mineral particles would almost certainly have become embedded in the metal, or there would have been abrasive score marks on the blades. But there was no sign of either.

“All the major damage to the blade tips is consistent with contact with another metal surface while under power,” Professor Erasmus concluded. “A small particle of aluminium alloy fused on to the rear surface of one of the blades appears to have become welded on to this blade by a high pressure sliding action at the time of general damage to this blade tip.” Shortly after Professor Erasmus published that report the remnants of the plane’s fractured spar caps were sent to him by the Accidents Investigation office. And when they were subjected to examination by electron microscope his doubts about the official cause of the crash grew. The tests are difficult for a lay person to comprehend but as Professor Erasmus amassed more data from them he became convinced that the crash was the final and inevitable stage of failure, possibly starting with the heavy landing at Oklahoma. He suggested that the failure of the rear caps imposed additional stress on the main spar lower cap, resulting eventually in the initiation and propagation of a fatigue crack. This crack propagated pro-

gressively over a long period, probably amounting to some thousands of flying hours. By November 21, 1961, the fatigue crack would have propagated through 40 per cent of the main spar lower cap area. In other words, the fatigue crack seen after the accident existed almost in its entirety before the plane left Wellington on the fatal flight. “With each increment of fatigtie crack propagation the starboard wing structure was correspondingly weakened and its load-carry-ing capacity impaired,” he said. “This continued to a point where final fracture of the front spar lower cap was imminent when next severely stressed. Starboard wing failure was imminent and inevitable when the aircraft next encountered turbulent flying conditions.”

Professor Erasmus thought the final stage would probably not have been instantaneous because other structural elements in the wing root would also have to go. In the end, the wing pivoted forward about the front spar and the propeller blades, still under power, slashed into the fuselage. Nearly a year after that finding, in 1972, the Accidents Investigation Office withdrew its original report and issued another. It conceded that the starboard wing front spar was defective and that some of the damage to the propeller tips was caused by metal to metal contact; but it also introduced new argument to suggest that the pilot did not behave responsibly. Among this argument was a suggestion that while some of the propeller damage was caused by contact with metal, the bending of the three blade tips was identical and could have been caused by striking snow. However, the disappearance of 75mm of one of the blade tips made it impossible to establish any bend radius on that blade.

The original accident report noted that the complete tip of this blade was missing and searches were made for it during the initial visit to the crash site and later when much of the snow on the mountain had melted. But when Professor Erasmus began his inquiry he noted that an Air Department photograph of the wreckage showed the complete tip attached to this propeller blade. As a result, the new accident report said that at some time between the investigators’ first and second visits to the crash, the tip was removed by an unauthorised person and never recovered.

On this point, Professor Erasmus

comments that there is no basis for the claim that the blade tip bend radii were identical. To back the claim that that starboard propeller struck snow, the 1972 report drew on four photographs recovered from a damaged camera found in the wreckage. They were taken by a passenger while the plane was over the mountain and each one included part of the instrument panel coaming. It was assumed that the camera was aimed “more or less directly ahead.” The plane’s height when the last photograph was taken was calculated to have been 9305 ft, plus or minus 71ft. The height of Paretetaitonga peak (1650 ft away) is 9025 ft. Civil aviation regulations required the plane to be at least 500 ft above the highest point of the terrain within a 2000 ft radius of the aircraft position. From this photographic evidence the 1972 report concluded that at some point near Paretetaitonga peak the pilot altered course appreciably to starboard and began a dive towards the crater lake on the summit. The already weakened starboard wing began to yield under positive loading and while the plane was above the lee slopes it failed and separated from the fuselage. The report suggested that some of the leading edge and rear surface damage to starboard propeller blades was produced when the blades struck some part of the plane, possibly the tail plane, after wing separation. “Forward tip bending incurred by the blades of the starboard propeller is characteristic of that observed on other blades known to have made contact with water or other terrain while in flight under power, and its severity and uniformity suggest that bending may not have been produced at the same time as leading edge and rear surface damage was incurred,” the 1972 report said.

“The available evidence does not prove conclusively that the starboard propeller blades came into contact with mountain terrain before wing separation occurred; neither does it prove conclusively that they did not.” The report attributed the cause of the crash to in-flight structural failure and separation from the fuselage of the starboard wing, but added: “It appears to the Chief Inspector of Air Accidents that a degree of responsibility for the accident may be attributed to the deceased pilot in command in that the evidence shows that he unnecessarily flew the aircraft below the minimum safe height prescribed — and that had he not done so the accident might not have occurred.”

However, subsequent tests by

Professor Erasmus with cameras in similar aircraft disproved that the aircraft was diving when the fourth photograph was taken by the passenger. Moreover, his plotting of the aircraft’s heading when each photograph was taken indicated that the highest ridge overflown just after the fourth photograph was taken had an altitude of 8500 ft. The aircraft would then have been about 9300 ft.

This contravened the regulation, but there were indications from the three earlier photographs that the pilot intended climbing to above the regulation height. The loss of a mere 200 ft in altitude between the third and fourth photographs could be attributed to strong down draughts or to wing failure which would have been progressing fairly rapidly at that stage. Professor Erasmus found a similar case in Canada in 1964. There an Aero Commander crashed just after take-off because of — according to the official report — “failure of the wing main spar ... under flight loads as a result of weakness caused by fatigue cracks.” The main crack was found in an identical position and to be of identical size to that on ZK-BWA.

The fatigue crack on the New Zealand plane was being extended with each flight and it was only coincidental that final failure be-

gan over Ruapehu, says Professor Erasmus. “It is therefore quite unreasonable to suggest that if the height regulation had not been contravened the accident might not have occurred.

“The second report admits that the detected presence of the defective spar caps would have invalidated the certificate of airworthiness and maintenance release document and that the aircraft was physically not airworthy immediately before the final flight. It argues, however, that both documents should be regarded as valid. Unfortunately, valid documents do not produce an airworthy aircraft. Nor did they prevent wing failure and the resultant accident.”

Professor Erasmus began the protracted investigations with some reluctance initially, but he became closely interested, though he says he has tried to keep that interest to technical argument only. But he did feel at one stage that the only step left was to call for a public inquiry. That will not be necessary. Last month, Professor Erasmus received an official copy of the report on the accident, printed by the Government Printer. It is essentially the 1972 report and contains the following statement by the Minister of Transport (Mr Gair): “This report approved for

release as a public document and Aircraft Accident Report No. 1192 dated February 14, 1961 declared a nullity.”

There are two curious features about the new report: The crash was on November 21, 1961, and not even the strongest advocates of an inquiry suggested that the report was written before the crash occurred. The slip is emphasised by being repeated on the cover of the report. The initial report was, of course, published in February, 1962.

• More significant is the omission of a clause contained in the 1972 report attributing “a degree of responsibility for the accident" to the pilot. But other conclusions suggesting there was pilot error remain.

Professor Erasmus hopes to dispel those final doubts by proving, with the aid of the $2OOO research grant, that propeller tips would bend forward after impinging against the fuselage. His findings will be awaited with considerable interest by relatives and friends of the pilot. They are hoping that nearly a quarter of a century after the crash Alfred Bartlett will be completely exonerated.