Who Would Want to Manufacture an Aircraft?
How can a bolt cause a fatal accident? Our resident solicitor and former airline pilot PAUL CLOUGH investigates.
On the 19 December 2014, the Queensland Court of Appeal delivered a judgment in an aircraft personal injury claim. There are usually two ways to process personal injury claims arising from incidents or accidents of aircraft. The common way is to sue pursuant to the Civil Aviation (Carriers Liability) Act and the other is pursuant to the local state equivalent of the Personal Injury Procedure Act (PIPA). The conduct of each method is quite different. Under the Carriers Liability procedure, the plaintiff does not have to prove negligence but simply that an accident or incident occurred between the time the plaintiff commenced boarding the aircraft until after the plaintiff has left the aircraft, that the plaintiff has suffered damages and that an accident or incident has occurred. Additionally, the damages are limited to $500,000 maximum per person and there is a strict two-year time limit on making the claim.
Under PIPA, the plaintiff has to prove the negligence of the defendant and the damages suffered. The time limit is three years extendable in some circumstances and the quantum of damages is at large. Certain restrictions are placed on the State Law to prevent minor claims being blown out of proportion both in damages and time.
In the judgment being considered, McDermott v Robinson Helicopter Company Inc  QCA 357, it would appear that the state system of personal injury law was utilised. Perhaps the two-year time limit had expired and the plaintiff had only the three year state law limitation available. The accident occurred on 30 May 2004 and the claim was filed in the Supreme Court in 2007.
Two of the three appeal judges were in a majority and one judge dissented. The judgment was primarily focused on the reasoning for each decision rather than a searching critique of the evidence of fact.
In summary, the fatal accident was caused by a bolt, called in the judgment “Bolt 4”, that “began to rotate loose shortly after it was assembled, and, in any event, more than 100 hours before the accident”. It was assumed that the bolt having become loose then created two cracks in the forward flexplate. This resulted in a loss of power to the rotor and hence of control of the helicopter, which resulted in the fatal accident. The above words are taken verbatim from the judgment of Alan Wilson J.
To complicate matters, the helicopter was obviously being worked hard. There were two 100 hourly inspections carried out on 27 February 2004 by LAME Fisher and 12 May 2004 by LAME Bray. Additionally, 28/29 pilot inspections of the forward flexplate were performed. None of these persons detected any rotation of Bolt 4.
Bolt 4 together with the other three bolts holding the flexplate is a critical fastener. Security of these bolts is of paramount importance to the ongoing use of the helicopter. The bolts themselves are tightened on installation by a torque wrench, a palnut is installed as a secondary locking device and thirdly, torque stripes are painted upon each of the four bolts to provide a easily viewed security marker.
Remember the plaintiff has to prove negligence in the defendant to establish liability in the defendant. Robinson Helicopter Company Inc had built the chopper and sold it into Australia. When it was sold and delivered it can be accepted that bolt 4 was installed and tightened correctly. No evidence was adduced that Bolt 4 had been intentionally loosened at any time from when it left the Robinson plant until the accident.
The 3 judges of the Court of Appeal all became obsessed with the third type of security device associated with the bolt, namely the torque stripe painted onto the bolt; the flexplate and the screwed thread leading to a housing. Various witnesses of varying degrees of qualification gave evidence about the torque stripes, how they were painted, whether they deteriorate, whether they will rotate with the bolt, whether they should be renewed. One of the judges, Wilson J, opined that each of the LAMEs who completed the two 100 hourly inspections should have applied a simple spanner to the bolt to confirm that it was tight. So far as it concerned Robinson, the approved maintenance data contained in the maintenance manual was the subject of close scrutiny. The Robinson Maintenance Manual (MM) simply required inspecting LAMEs conducting the 100 hourly to “verify security” and “correct installation”.
Ultimately, the two judges in the majority were of the opinion that the requirements of the MM for inspecting LAMEs was inadequate. It did not help that Robinson changed this aspect of the inspection procedure in the MM after the accident to require the torque stripes to be conspicuously painted on in lacquer rather the paint; that they “must” rather than “should” be painted across all parts of the bolt, the nuts, and the part being bolted.
From a lawyer’s point of view, this judgment seems to be a valuable guide for Aircraft manufacturers to spell out in their MMs very specifically the inspections to be carried out by LAMEs in the course of periodic maintenance. In particular, for critical fasteners in and about aircraft and helicopters, the duty of care necessary in the detail of the actual inspection can only be provided by much more than “verify security” and “correct installation”. Some hands-on verification is necessary to satisfy the high level of duty of care for critical rotating part security.
From a practical LAME engineering point of view, this judgment is a mish mash of assumed facts upon which is strapped a finding of a lack of duty of care in the writing of a maintenance manual. It is difficult to conclude that a LAME reading this judgment will understand what he is expected to inspect and what he can rely upon that has been completed satisfactorily in the past.
I mention only two of many fundamental assumed facts upon which the judgment is based and which do not ring true as a “matter of fact”. Firstly, the trial judge accepted that the case focused not upon how the bolt became loose but rather why its looseness was not detected (Para 46). The trial judge said, “It was very difficult to identify when the bolt was left loose but, in any event, the evidence fell short of establishing that it must have been left loose before the service carried out on 17 February 2004 (by LAME Fisher) (Para 62)”. That must mean that he thought the looseness occurred after 17 February 2014. At Para 67 the trial judge found that, “Rotation commenced at Bolt 4 shortly after the joint was incorrectly assembled, and more than 100 hours before the accident. If the torque strip had been correctly applied at the time when the bolted joint was incorrectly assembled, it would have broken relatively shortly thereafter, and would have been visible on a subsequent 100 hourly inspection”. (Reasons, Para 157). What is the evidence to support all of these findings? “Incorrect assembly”? Was the logbook in evidence indicating that the bolt was disassembled for some reason to need reassembly? Who was the LAME that did the disassembly and reassembly and why? Are there two signatures of two LAMEs in the logbook to certify the control run? We do not know. The finding of incorrect assembly was made without foundation in evidence. How do we know that the bolt rotation commenced “more than 100 hours before the accident”? No expert testified to this hypothesis. There was the evidence of the 28/29 pilots and two LAMEs who inspected the bolt that nothing was untoward in the torque stripe. Again the commencement of the rotation of the bolt was pure conjecture on the part of both the trial judge and the three judges of the Court of Appeal. As a personal aside, your correspondent flew Bell 47G2s for about 1500 hours. The helicopter is a mass of vibrating, rotating and moving parts. If a nut becomes loose and does not have an effective secondary or tertiary locking device it will vibrate from the bolt within minutes rather than over more than 100 hours. If the flexplate rotated at some speed, a loose bolt would manifest itself within minutes of looseness being achieved. It was my experience to see doorframe nuts and handles vibrate loose in a very short space of time in helicopters, hence we had lock wires on many hand-operated bolts nuts and hand-pieces. This phenomenon did not see the light of day as evidence. Again, the timing rate of rotation of the bolt was a figment of the judge’s imagination rather than evidence.
The other matter that seems to have been given short shrift was the other two inspection criteria and how the LAMEs and pilots perceived their relevance. Whilst the pal nut was mentioned as a secondary locking device, no LAME, pilot or expert commented upon its veracity or relevance to ensuring that Bolt 4 did not rotate loose to the catastrophic culmination in the accident. No mention of evidence as to where and how the pal nut secured the bolt from rotating off was mentioned in 22 pages of judgment. If the pal nut was doing its work and there was no evidence to suggest otherwise the bolt was secure and the evidence of the LAMEs to the effect that they verified the security of Bolt 4 should not have been discarded out of hand.
The wording of the maintenance manual was sufficient for a LAME who not only verified the security of the bolt but also the security of the secondary locking device, the pal nut, to fulfill any perceived heightened duty of care in the manufacturer.
The focus of the four judges concerned with the torque stripe blindsided the benches to the other complementary security features namely the torque limits of the bolt, the lack of evidence that it had been disassembled and reassembled and the function of the pal nut.
In all the judgment is much like a curate’s egg – only good in parts.
Watch this space…