Gear collapse

Citation

(1998), "Gear collapse", Aircraft Engineering and Aerospace Technology, Vol. 70 No. 1. https://doi.org/10.1108/aeat.1998.12770aab.018

Gear collapse

Gear collapse

A Fokker F27 MK 050 was on a scheduled service from Rotterdam to London Heathrow with four crew and 41 passengers on board. As the aircraft approached the London Terminal Manoeuvring Area (TMA) the crew were instructed to hold at Lambourne (LAM) VOR. After three holds the aircraft was radar vectored for an ILS approach to Runway 27 left. The landing gear was selected down normally but the crew noticed that the left main landing gear showed an unsafe indication. The commander, who was the handling pilot, therefore decided to carry out a go-around in order to investigate the problem. Before initiating the go-around manoeuvre however, he asked the first officer to consult the "Main gear unsafe after down selection" check list to see whether he should raise the gear. The first item on the check list was to recycle the gear. The gear was therefore raised during the go-around as the first part of that procedure.

After the go-around which was carried out at 11.53 hrs, the crew were instructed by ATC to take up a heading of 120° M and climb to 3,000ft. ATC then asked for the reason for the go-around and whether the crew wished to make a further approach. The crew needed time for their investigation, so ATC suggested that the aircraft should hold at Epsom NDB. The aircraft entered the hold at 12.01 hrs and the crew returned to the check list. They re-selected the gear down, completing the cycle. This however was also unsuccessful. The first officer then left the flight deck to check visually if the left main gear was down and locked by looking for alignment of painted red lines on the landing gear strut. His inspection showed that the gear was not in a safe condition. On his return to the flight deck the crew initiated the "Alternate Down Procedure" As this was also unsuccessful the commander, in accordance with the check list, requester clearance from AC to carry out a level "2g" (60° bank) turn at 172 kt. This manoeuvre however, did not affect the left main gear unsafe indication.

At this stage the crew were now committed to an emergency landing. At 12.12 hrs they transmitted to ATC that "We have the intention to make an emergency landing at Heathrow and we need approximately 15 minutes for preparation. The landing gear system is down but not locked so there is a high risk of the gear collapsing after touchdown". The airport emergency services were alerted and brought to an "aircraft accident imminent" status. At 12.22 hrs the Heathrow Director informed the crew that the surface wind was now favouring a landing on the easterly runway and that Runway 09R would be used. The weather was passed to crew as visibility 8 km, scattered cloud at 400 ft, overcast cloud at 600 ft, and a surface wind of 140°/4 to 7 kts.

In preparation for the landing the crew actioned the "One main gear up or unsafe" checklist. At 12.23 hrs ATC asked if the crew were ready for the approach. The crew replied that "The cabin is still being prepared so we need at least one or two more holds". At 12.38 hrs, having updated ATC in the number of persons on board, the crew declared a "mayday" and stated that: "We have 600 kg of fuel left which gives us a low fuel emergency upon landing and we expect a collapse of the left-hand gear upon touchdown". At 12.48 hrs, after a very comprehensive preparation of the aircraft and cabin, the aircraft, having been radar vectored, intercepted the ILS to runway 09R. The aircraft's track had taken it clear, to the south-west of the built-up areas of the outskirts of London. The aircraft however had not presented any danger to areas beneath its flight path as the emergency would only develop on landing.

The aircraft's touchdown was normal at 12.53 hrs right main wheel first. About five seconds after all the landing gears were in contact the left main landing gear collapsed and the aircraft left wing tip, left propeller and the rear left portion of the fuselage contacted the runway. The aircraft veered to the left coming to rest on the hard surface clear of the runway in Block 81. A full aircraft evacuation was then carried out with approximately half the passengers leaving the aircraft via the front left door whilst the remainder exited via the half open rear door. The commander was the last to leave the aircraft via the rear left door. The front left door, which is hinged at its lower edge and incorporates several steps, opened normally but the door adopted a horizontal attitude because of the fuselage's close proximity to the ground. The rear left door opened normally but its lower corner contacted the ground with the door only 90° open. Although the width available for evacuation was reduced the passengers evacuated the aircraft without difficulty.

Flight recorders

The Flight Data Recorder (FDR) and Cockpit Voice Recorder (CVR) were removed from the aircraft and replayed at the AAIB. The recorded data on the FDR showed that the aircraft had made the initial go-around at an altitude of 1,040 ft agl. The aircraft then levelled at 3,000 ft agl and joined the holding pattern over Epsom. During the second hold, whilst carrying out the unsuccessful attempt to lock down the left main gear, the aircraft had made a tight 360° turn which resulted in a recorded maximum vertical acceleration of 1.95g. The Epsom hold was then re-established and flown a further seven times. At 12.53 hrs the aircraft touched down at 99kt with flap 35 selected rolled 0.5° to the right. The right main gear touched first. Five seconds later the FDR recorded that all three oleos were on the ground. Ten seconds after the initial touchdown, at a speed of 60 kts, a sharp roll of 13° to the left and a large vertical acceleration were recorded as the left main gear collapsed. The aircraft began to slew to the left, finally coming to rest, after a further 12 seconds, on a heading of 017°M. During this time the left hand propeller came into contact with the runway and slowed from 80 per cent RPM to 63 per cent RPM. The sound of the propeller blade tip striking the ground was audible on the CVR recording. Once the aircraft came to rest the non-handling pilot pulled the engine 1 fire handle whilst the Captain set the parking brake. The Captain then called for the on-ground emergency checklist and the First Officer pulled the No 2 engine fire handle and set the fuel levers to off before the CVR and FDR recordings terminated.

Engineering aspects

The aircraft came to a halt on a taxiway having slewed though about 90° to the left. It rested on its nose and right main landing gears, the fuselage underside and the left wing tip. The left main landing gear had partially retracted into its well. For recovery the aircraft's left wing was lifted using air bags and then supported on a jack. The left gear began to extend as the wing was lifted and as it emerged from the wheel well it could be seen that the platform which forms one half of the over-centring stop in the lock-link was not fully secured and was slightly displaced. The platform was aligned and taped in place. As the wing was lifted further and the gear reached full extension the lock-link went over-centre and the gear locked down.

One of the two bolts used to secure the platform was found to be missing and the other one was loose. At the position of the missing bolt contact marks showed that a bolt had been present at one time. The head of the remaining bolt was drilled for wire locking but no wire was present. The wheel well was searched for the missing bolt. This was not found but a washer was found which matched the washer on the remaining bolt for type, size and colour. The platform was free to rotate about the remaining bolt and some damage marks showed that it had fouled an adjacent stepped surface of the lock-link and this had prevented the lock-link from achieving the overcentred condition. When this condition was replicated the lock-link was seen to be in the position described by the crew in which the ground lock-pin holes in the two sections of the lock-pin were misaligned by one diameter. No other defects were apparent in the left main landing gear, the extension jack and the lock-link over-centring spring, which were tested separately, and the investigation concentrated on the retention and locking of the platform.

On manufacture, when the initial assembly of the lock-link has been completed and after locking wire has been fitted, the link is coated with primer paint and a cosmetic silver paint. Evidence for the presence of locking wire, should, therefore be visible in the paint coating on the bolt head as well as in contact marks on its metallic surface (it is cadmium plated). The head of the remaining bolt was examined closely and at one of the three locking wire holes there was damage on the edge of the hole at each end which appeared consistent with the use of locking wire. The other two holes were clear of any such contact damage. Microscopic examination of the head of the recovered bolt revealed no evidence that it had been untightened at any time, in fact the condition of the painted surface indicated that it had not. (Under the paint there were some contact marks on the hexagon faces consistent with spanner or socket application in the tightening direction.) The paint on the bolt head had suffered some in-service damage or erosion and there were two small impact marks on the top edges of the head almost diametrically opposite one another. It could not be determined what the cause, or effect, of these impacts might have been.

Maintenance and overhaul documentation

The landing gear manufacturer's assembly drawing for the lock-link did not show locking wire drawn at the location of the two bolts which secure the platform and a drawing note (specifying wire locking) which had been applied to other locations which required wire locking was not applied to the bolts. The bolts which were shown on the drawing (and supplied in the kit of parts for assembly) were specified as NAS 1303-15h. On this bolt, holes are provided in the hexagon head for wire locking. The assembly drawing referred to a manufacturing standard document which itself referred to a company process specification which defined standard practices for component assembly. This contained the statement "When a drawing does not call out this specification together with an item reference, the appropriate instructions in paragraph 4 shall apply, in the absence of specific instructions to the contrary". Paragraph 4 contains the process specifications including wire locking.

The lock-link was assembled in accordance with a document containing a list of numbered operations and inspections for the numbered operations were recorded on separate sheets. The subject lock-link was assembled as one of a batch of four, all being recorded on one set of paperwork. Three other batches of four were assembled at around the same time by the same team, and for the investigation, these were considered to form one batch and an attempt was made to trace these. In the instruction which included the final torque tightening of the platform bolts there was a note giving the process specification for the torquing of the bolts but no instruction or note on the locking of the bolts. Nevertheless, staff at the landing gear manufacturer stated that it was standard practice to apply wire locking to all nuts or bolts to which it could be applied and the physical evidence from the remaining bolt was that wire had been present in one of the holes in the bolt at some time. After the incident, the assembly drawing, as the master reference document for the build of the lock-link (though it was only used within the company), was re-issued with wire locking shown and appropriate notes applied.

The landing gear manufacturer also found that the Component Maintenance Manual (CMM) for the lock-link did not contain an instruction to wire lock the platform bolts even though there were specific instructions to this effect for other fasteners. The CMM is the document that would be used by any agency other than the manufacturer that might overhaul a Fokker 50 landing gear or one of its component parts. This information did not affect the investigation of the collapse of this particular gear as this gear had not been overhauled but the CMM, which was in the process of being re-issued while the investigation was in progress, was amended to include an instruction for the wire locking of the platform bolts.

The landing gear manufacturer reported that the company had recorded some cases of problems with the quality of wire locking between 1991 and 1995. The outcome of an investigation which was carried out was to improve the training of fitters in the required procedures and standards.

Maintenance history

The aircraft's records did not show that the landing gear on this aircraft had been changed or overhauled since the aircraft had been built, so the gear had been last assembled during its original build by the manufacturer. There were two anomalies between the left landing gear component serial numbers recorded in the records and those found on the aircraft, but these were single digit differences and were probably transposition errors. There was no discrepancy between the lock-link serial number as found and as recorded.

Detailed inspection of the landing gear is required in the scheduled "B" Checks at 650 hour intervals: "Examine the MLG and make sure that; all the bolts, the nuts and the attachment points are correctly installed. All the split pins, the lockwires and the lock plates are correctly installed and intact".

The last "B" Check had been conducted in October 1996 at 9,552 airframe hours and 9,443 cycles.

A search of the records of unscheduled maintenance or rectification revealed no work directly affecting the platform and the stores record showed that this component had never been replaced. The downlock microswitch is contained within an open housing or bracket the front face of which is immediately behind the heads of the platform bolts. It was considered that if work was carried out on the microswitch, it was possible that the platform bolts could be mistaken for the front retention bolts of the bracket. In fact, the front of the bracket is secured by two countersunk screws transversely installed, but its rear flange is held by two bolts of the same size as the platform bolts which are not wire locked. It is not necessary to remove the bracket from the lock-link in order to remove or change the microswitch as it can be extracted with the bracket remaining in place. The last action recorded affecting the microswitch arose from the rectification of an indication fault in the right main gear in February 1994. During rectification the left microswitch system was cleaned and adjusted as well as the right. However, adjustment of the microswitch operation is made to the striker bolt, not to the microswitch itself, and should not have involved any disturbance of the bracket, its attachment bolts and screws or the platform bolts and there would seem to be little opportunity for their mistaken disturbance in that operation.

Safety actions

On the day following the incident in December 1996, Fokker Services issued an All Operators Message reporting the occurrence and recommending that an inspection for the presence of correct wire locking should be considered. As a result of the AOM, a report was returned of a broken locking wire on another aircraft, with one section of the wire completely missing, and five days after the incident, the AAIB and the Accident and Incident Investigation Bureau (AIIB) of The Netherlands Aviation Safety Board issued recommendations to The Netherlands Airworthiness Authority (RLD) that Fokker's recommendations be strengthened, and that they require operators to carry out inspections.

The AAIB recommended that The Netherlands regulatory authority should require operators of the Fokker 50 to inspect the retention bolts of the maingear down lock platforms to confirm the presence and condition of the locking wire between the two bolt heads, report back and repair any deficiency found, after advice from the manufacturer.

On 18 December 1996, Fokker announced the issue of a Service Bulletin which was mandated by RLD and the RLD later raised an Action Record Sheet which required Fokker Services to revise the downlock platform retention design and provide changes to comply with JAR's.

In the feedback from operators, out of the existing 210 aircraft, information was obtained on 138 plus information on 11 spare lock-link units. Two additional anomalies were reported; in one case the lock wire was wrongly applied and in the other (overhauled) case the wire was missing. Included in the report were ten other units from the manufacturing batch containing the subject lock-link. (Amongst the lock-links not traced some had been overhauled and therefore their original condition was lost.) No anomalies were reported in these units.

The actions of the crew in following the emergency procedures, left the normal landing gear selector and the "Alternate" gear selector in the "Down" position. The "Alternate" selector operates a cable system which, in the "Down" position, opens a dump valve which vents hydraulic pump pressure and mechanically releases the uplocks of the main gear and the nose gear front doors. In this condition, with the normal gear selector also selected "Down", hydraulic pressure on both sides of the actuator is vented so that there is no hydraulic resistance to the leg descending under its own weight and aerodynamic loads. None of these functions were relevant to the circumstances in this accident and once the aircraft had touched down the hydraulic system offered no resistance to the leg retracting, allowing the wing tip, fuselage and left propeller to contact the ground.

Figure 1 Landing gear downlock mechanism

In view of this a further recommendation was made that Fokker Services consider, in the light of this accident and considering the variety of emergency situations which may arise, including leakage in the hydraulic system, whether changes to the landing gear emergency extension procedures are possible in order either to enable flight crew to decide when an "Alternate Down" selection is appropriate, or, to return the landing gear hydraulic system finally to "Normal" operation with "Down" selected to pressurise the extension side of the landing gear actuator and reduce the likelihood of a landing gear retraction after landing.

Fokker Services should ensure that any changes are consistent with the design philosophy of the system and the check list.

ReferenceAAIB Bulletin 8/97.