Nosewheel axle and two wheel hubs destroyed

Citation

(1998), "Nosewheel axle and two wheel hubs destroyed", Aircraft Engineering and Aerospace Technology, Vol. 70 No. 5. https://doi.org/10.1108/aeat.1998.12770eab.021

Nosewheel axle and two wheel hubs destroyed

Nosewheel axle and two wheel hubs destroyed

Keywords Aircraft, Bearing fatigue, Cadmium embrittlement, Landing gear, Safety

A Boeing 757-225 was engaged on two sectors, Birmingham to Malaga and return. At the end of the first sector, after an uneventful landing at Malaga, the aircraft left the runway via the rapid exit taxiway. At a speed of about 20kt a vibration was felt through the steering as the aircraft was turned left through approximately 120° onto the parallel taxiway. The commander initially thought that the vibration was owing to the taxiway surface, however, after 100-150 yards of straight taxi-ing the vibration returned again in a more marked fashion. The aircraft was brought to a stop immediately and ATC and the airport emergency services were informed. Passengers disembarked via steps and taken by bus from the taxiway.

On inspection the right-hand nosewheel was found canted over at an angle, the outer bearing having disintegrated. The operator's duty engineer, in Maintenance Control at Manchester was informed and despatched an authorised certifying engineer (a Licensed Aircraft Engineer or LAE), and a wheel change kit from Manchester by diverting another 757 bound for Tangier, to Malaga. The LAE visually examined the axle and found some axle damage that had been caused by the wheel bearing failure. He dressed out this damage, changed both nosewheels, and the aircraft took off on its return flight to Birmingham. The nose gear axle failed inboard of the right hand outer bearing land (Plate 12) as the aircraft turned off the runway at Birmingham.

Subsequent metallurgical examination showed that the fracture of the axle was the result of the degradation of the axle material properties, due to penetration of liquid cadmium from the surface plating into the steel wall of the axle (cadmium embrittlement), while it was subjected to heating and tensile stresses during the break-up of the bearing at Malaga. Additionally, the axle had been further weakened by mechanical damage to a depth of 33 per cent of the wall thickness, also caused during the break-up of the bearing at Malaga.

Plate 12 Nosewheel axle showing failure at right-hand end

The relevant chapter of the aircraft maintenance manual (AMM) called for an examination of the nose gear axle for overheating after a wheel bearing failure, and required the use of a borescope to enable the internal bore of the axle to be examined for evidence of overheating of the cadmium plating. The duty engineer was not aware of this check, neither was it known to the LAE, before the aircraft carrying the wheel change kit departed for Tangier via Malaga, nor was it found by the duty engineer during his subsequent document search. Consequently the examination was not carried out. The duty engineer continued to search for repair limits to the axle, but was unsuccessful (because no repairs were authorised).

Sequence of enginering events

Owing to various circumstances, the LAE was despatched on the aircraft for Tangier via Malaga, without an assistant, without the knowledge that an overheat check was required following a bearing failure, and without a borescope to enable him to carry out such a check. He did not have time to look at the AMM at Manchester and neither aircraft carried a copy, so he had no opportunity to see the sub-chapter containing the inspection requirement following a bearing failure. An AMM could have been obtained at Malaga or pages could have been faxed out by the duty engineer, albeit with some difficulty, but the LAE did not see the need to seek such additional information.

The maintenance organisation exposition issued by the operator and approved by the CAA defined the procedure, accepted by the CAA, to be used in the repair of aircraft. All certifying engineers employed by the operator would be expected to know these procedures and would be tested periodically on their content. In particular, the exposition required that repairs to aircraft structures which fall outside the aircraft structural repair manual are to be referred to the Technical Services Department, who would employ the aircraft manufacturer or an approved design organisation, to provide an approved repair scheme. The only method authorised by the aircraft manuals for the nosewheel axle was repair by replacement, any other method would therefore require an approved repair scheme. The exposition further required that all repairs were to be detailed and documented for record purposes. Later the duty engineer received a call from the aircraft on the ground at Malaga saying that it had been jacked, the wheel was off, and the axle was "not too bad", but the Spanish engineer was having trouble getting the bearing off. Later an avionics engineer took over as duty engineer in the UK maintenance base and passed this information on to the LAE who was by now airborne. Some time later the duty engineer was informed of the extent of the damage to the axle, and he was still concerned that no information about possible repair limits could be found in the aircraft manuals. He therefore contacted the Boeing 24-hour engineering support desk to see if the damage could be worked to limits. Boeing referred the duty information engineer to the AMM and asked for further information in the form of a sketch if the AMM did not find the information he needed. No further information was passed because the aircraft was declared serviceable by the LAE at Malaga and had taken off.

When the LAE arrived in Malaga he was asked about the the anticipated length of the delay. He missed the first cue of serious trouble because the damaged wheel was already loaded and was inaccessible because it had been placed behind luggage in the forward freight bay. He noticed that the bush fitted to the inside of the threaded end on the axle was damaged (Plate 13), as was the axle nut, but decided not to change the bush as sufficient locking holes remained available. He identified the mechanical damage between the bearing lands on the axle caused by the damaged wheel and bearing and assessed it as being 11Ž2in. long and 11Ž16in. deep. He could not see any signs of blueing or overheating on the outside of the axle and considered that it would be satisfactory for the aircraft to return to Birmingham, after the damage had been blended. These details were passed to the duty engineer in Manchester.

Plate 13 Bush removed from internal bore of axle

The LAE examined the inside bore of the axle after cleaning it as best he could, but as he only had a torch and not a borescope he could not see 7in. into the bore (as the AMM required). He therefore missed the signs of overheating inside the bore of the axle as the change of section behind the internal locking bush hid the damage from a torch inspection. Having decided that it would be satisfactory to blend the external damage to allow the return flight to proceed, the LAE did so with a half round file and emery cloth. He did not raise an acceptable deferred defect on what he considered to be a temporary repair as he had no drawings or blend limits to work to. In fact there were no limits published because no blending was allowed.

During the inspection of the axle he had to handle a distraction from the commander of the aircraft, who approached him about a refuelling problem. Later, during blending the same commander again approached him with his refuel problem; this time the LAE left the axle job to assist with the refuelling.

The LAE replaced the right-hand wheel, which was fitted without a problem, and changed the left-hand wheel using the photocopied maintenance manual extracts showing the appropriate torque figures. He contacted maintenance control one hour after arriving in Malaga, with brief details of the damage and stated that the aircraft was satisfactory for service but that an axle change should be planned when the schedule allowed. It was agreed that this would happen at Manchester after the Birmingham sector. The aircraft took off for the return flight to Birmingham at 2259 hours. (Flight crew duty time would have required the aircraft to take off from Malaga by 0200 hours.)

At 0121 hours the aircraft arrived at Birmingham. A replay of the flight data recorder gave the landing parameters for the last 12 flights. These indicated that the relevant parameters of the accident flight landing, where known, were no worse than average.

The aircraft turned off the runway via the rapid exit taxiway, at a speed of 12kt a "snap" was heard and vibration was felt throughout the steering. The aircraft was brought to a full stop over a distance of 20 yards and ATC and the airport emergency services were informed. Passengers were disembarked via steps and transferred by bus. Inspection revealed that the landing gear had failed inboard of the right-hand outer bearing land, in the region of the axle damage dressed out in Malaga.

Examination of components

The failed bearing at Malaga had achieved 440 landing cycles since installation. This life was slightly higher than the average time to rejection of a random selection of seven nosewheel bearings during tyre change. The bearing was visually examined by a representative of the manufacturer in the presence of AAIB. With the exception of water corrosion on the core bore, the inboard bearing was in reasonable condition. The outboard bearing was severely distressed with most of the debris exhibiting plastic deformation from softening at the high temperatures experienced during the bearing failure.

A total of 14 rollers out of a full complement of 27 were available for examination. The spherical grinding on the ends of some of the rollers was nearly undamaged, indicating that wear at this sliding contact area had not been a problem before the rapid collapse of the bearing. The bearing cone also exhibited sections of thrust rib in relatively undamaged condition, again endorsing a rapid collapse. Several rollers showed line etching caused by water corrosion, and, in some cases, fatigue damage at these sites of corrosion.

It was concluded that the outboard bearing had collapsed rapidly. This is characteristic of a cage failure and may be caused by corroded and fatigued spalled roller bodies wearing or "machining" the cage arms until they are sufficiently weakened to break. Corrosion of bearings on a variety of nosewheel applications is a well known phenomenon as the bearings are generally not backed-up with a brake pack which helps shield main wheel bearings from water egress.

Examination of the bearing indicated that there had not been a lubricant failure, although the grease may have carried contaminants. The grease recovered from the bearing was, therefore, examined in a laboratory of the Defence Evaluation and Research Agency at Farnborough, to check for the presence of water detergents or deicers. One grease sample was taken from the wheel carrying the failed bearing, and another sample was extracted from the bearing components in the laboratory. Unused Aeroshell 5 grease, the type used in the bearing, was also examined as a reference. The water content of the sample taken from the bearing was not significantly higher than that of the clean grease. Detergents and deicers were not found in the samples.

Visual and microscopic examination of the axle that failed at Birmingham gave the following information:

1. 1.

   A generally circumferential fracture had occurred between the wheel bearing lands at one end.

2. 2.

   There was considerable mechanical damage in the vicinity of the separation on the underside of the axle; some of this damage had been mechanically smoothed out. Measurements showed that up to 33 per cent of the wall thickness had been removed from this region.

3. 3.

   Some of the chromium plated lands on which the bearing sat varied in width around the circumference indicating a lack of concentricity between the axle centreline and the centreline used during the grinding of the lands (although the axle met all of the relevant drawing requirements).

Detailed examination of the fracture surface in the initiation region showed that it had been partially "wetted" with cadmium from the axle bore. It was also seen that cadmium in the bore at, and adjacent to, the initiation region had melted and re-solidified as globules. The melting point of cadmium is 321°C.

A section of the initiation region was extracted and polished. It showed that extensive cracking had occurred around the macrograins as a result of the penetration of liquid cadmium when the material had been subjected to tensile stress. This phenomenon is known as liquid cadmium embrittlement. Hardness tests indicated that a loss in tensile strength from approximately 118.4 tons/in.² to 113 tons/in.² had been caused by the localised heating when the mechanical damage occurred. However, the additional degradation by the liquid cadmium had further dramatically reduced the overall strength. Further hardness tests showed that a heat temperature of between 400°C and 500°C was reached when the axle sustained further mechanicaldamage at Birmingham(Plate 14).

The bearing failure was initiated by water corrosion of some of the rollers; this led to heavier roller damage from fatigue, and eventually to break-up of the bearing as the roller cage was broken. The lack of shielding from a brake pack made the Boeing 757 nose wheel bearings more susceptible to water ingress than the main wheel bearings. Nevertheless, the CAA MOR database showed that failures of this bearing on Boeing 757s have only occurred in the UK approximately once every two years, and that no axle failures have been reported.

The axle should have been rejected at Malaga on two counts: the area between the bearing lands was severely damaged and had suffered a 33 per cent reduction in wall thickness; and the application of the inspection contained in the AMM would have shown symptoms of cadmium embrittlement. The failure of the axle was therefore a result of deficiencies in:

• •

  the application of basic airframe trade knowledge during the inspection of the damaged axle;

  

  Plate 14 Axle showing damage caused by bearing failure ­ the abraded groove indicated by the arrow was caused after the axle failure at Birmingham

• •

  the knowledge of the check for cadmium embrittlement required by the AMM;

• •

  the provision of information for the use of the LAE; and

• •

  the quality oversight of the operation at Malaga.

Safety recommendations

It is recommended that the:

• •

  CAA requires organisations offering type rating courses to amend their syllabus to include the subject of cadmium embrittlement, where relevant;

• •

  CAA requires the operators to review their procedures for maintenance away from a main base, with the object of making them more robust, ensuring compliance with the AMM, and removing some of the pressure from the certifying engineers sent to rectify aircraft down route;

• •

  operator carefully defines his logistic and engineering priorities in a situation where rectification is required down-route.

Subsequent action has shown that the operator has reviewed procedures and implemented the necessary changes. The Quality Assurance Department now examines the content of all types of training courses and, from May 1998, they will operate their own line stations in the UK, with sufficient staff to facilitate down-route rescues.

Reference

AAIB Bulletin 4/98.