JetBlue Emergency Landing: LAX 050921

Today, shortly after takeoff from the Bob Hope Burbank airport, a JetBlue A-320 was unable to retract its nose gear.  The pilot informed ATC and the plane was vectored over to the Long Beach airport where it did a low flyby, so JetBlue maintenance personnel could look at the problem.  They were able to confirm that the nose gear had rotated to a position of about 90 degrees out of line to where it should have been.  That is why the gear could not be retracted into the wheel well.

The intended destination of the flight was New York Kennedy.  Obviously it could never fly that far with the landing gear extended, so the pilot had no choice but to fly in the Los Angeles Area for about 3 hours to burn off the weight of the major portion of fuel on board (there is no ability to dump fuel, on the A-320).  He had to do that to reduce the weight of the airplane so that it could be landed at the slowest possible speed.

To reduce the risk of the nose gear structure breaking loose, which would lead to more damage to the aircraft and increase the risk of fire, it would also be necessary to land without putting anymore weight on the nose wheel structure than absolutely necessary, and to delay its actually touching the runway for as long as possible.  I presume the pilot located passengers and their carryon bags as far aft as was feasible, so that when the plane touched down, the center of gravity being further aft helped to accomplished that objective.

To keep the nose gear from touching the pavement for as long as possible after the main gear contacts the runway, the pilot on some planes (like Boeing or Douglas airliners) might employ a non-standard technique of trimming the horizontal stabilizer towards the nose up position, during the actual landing flare, until it will go no further.  Then, as the nose began to drop, he would hold it off even longer with the elevators, until they too were pulled back to the full nose up position.  

That is a very difficult maneuver however, especially since pilots are not given simulator time to practice it.  The danger lies in the pilot running out of any pitch control at all, before he has gently lowered the nose gear to the runway.  If he does run out of pitch control prematurely, the nose gear may impact with too much downward force, and that would significantly increase the risk of its breaking off and causing additional damage to the plane.

Captain Scott Burke proved to have the required skills; no one could have improved on that performance.  He stopped right on the center line of that 12,000 foot runway (25L-----which was not an "auxiliary" runway, as one Sacramento TV News station claimed), with plenty of room to spare.  Best of all, the nose gear did not break loose.  

A split second before the nose gear touches the runway

The rear-most tire is dragging and about to catch fire

Rear tire is now on fire from friction generated heat

Investigation photos show the damage was limited to tires and rims only

I believe this is at least the seventh case of this type of accident (there may be even more than that), where the nose gear rotates to a position of 90 degrees, from where it should be, when the gear is down for takeoff and landing (this kind of problem could never happen on a Boeing airliner-----only the French come up with designs like that).  

On Feb. 16, 1999 an America West A-320-231, attempted to land at Columbus, Ohio (CMH), total of 31 on board.  Same kinds of ECAM warnings, when the landing gear was lowered.  That plane was finally forced to land with the nose gear rotated 90 degrees out of line.  Same result as this JetBlue plane, which landed safely at LAX today, except they evacuated down the emergency slides at CMH, because the pilot could see smoke curling up from underneath and he couldn't get the tower to answer his query----if any fire could be seen. 

Prior to that AWA emergency landing at CMH, the French had issued a service bulletin on the problem.  However, no ADs had yet been issued by either the French DGAC or the American FAA, so compliance was advisable, but not mandatory.  America West Airlines did not choose to heed that bulletin, and the emergency landing was the result.

Analysis of the previous cases revealed "...external hydraulic O-ring seals on the steering control module's selector valve were extruded (distorted out of the seal's groove). A small offset was found in the steering control valve."

Airbus further found that the problem would not be detectable by normal visual inspection procedures.  It appears that though the problem is rather rare, it does have to do with that particular design of the steering control module.  Quoting Airbus: 

...during landing gear extension, the brake and steering control unit (BSCU) would have been energized and hydraulic pressure would have been directed toward the steering servo valve. The BSCU would have then commanded a small rotation of the nose wheel to check for proper movement. Any disagreement between the commanded position and actual position of the nose wheel would have deactivated the nose wheel steering. However, if hydraulic pressure had bypassed the steering control valve, there would have been continued pressurization to the servo valve, and because of the servo valve's inherent offset, in-flight rotation of the nose wheels.

Procedures existed for removal of hydraulic pressure from the steering control module. However, once the nosewheel strut had deflected 90 degrees, the centering cam would have been rotated to a flat area, and would have been incapable of overriding the 3,000 PSI hydraulic system, and returning the nose wheels to a centered position.

Following the CMH emergency landing, both the French DGAC and the American FAA issued ADs on this problem, so that all steering control modules would be repaired and/or modified to prevent a recurrence.  Those ADs were issued in 1999.  All airlines had to comply with them within 12 months time. 

Then, it happened again on November 21, 2002.  This time the airline was United. When they attempted to retract the landing gear after takeoff at O'Hare, Chicago, they received a L/G SHOCK ABSORBER FAULT message on the upper ECAM (electronic centralized aircraft monitoring) display.  The captain moved the gear lever back to the down position, which then generated the AUTO FLT A/THR OFF message.  They could not use the autopilot after that.  

While the First Officer flew the plane in the Chicago area, the Captain went thru the book procedures for these types of messages, and also got on the "phone" to SAMC (system aircraft maintenance controller) at United's San Francisco maintenance base.  The captain explained to SAMC that the ECAM landing gear page, indicated all three landing gear were down and locked, and the gear doors were closed.  

SAMC then instructed the Captain to interrogate the Centralized Fault Display System (CFDS). That revealed the WHEEL N/W STEER FAULT message. The decision was then made to return to ORD.  The Captain was now concerned that he would not have any nose wheel steering during the landing roll. SAMC was of the opinion that the nosewheel steering might be recovered during the landing roll, when the nose strut was compressed.

The Captain then proceeded to land the  plane at ORD, on R04R (8,071 ft. long, by 150 ft. wide).  No one realized, up to that point, that the nose gear was turned sideways by 90 degrees.  The runway was wet and initially the landing was smooth and appeared to feel normal.  However, as they continued to slow down, the tower informed the crew that sparks could be seen coming from the nose gear.  It was about that time that they began to experience increased noise and vibration. The captain told the passengers to remain seated, while he assessed the situation outside the airplane. The ORD Fire Department arrived at the plane within two minutes. They determined there were no hazards (such as fire) outside the plane, and the passengers were spared the risks concomitant to an emergency evacuation down the escape slides.  

Both tires were blown, the rim of the left tire was ground all the way down to the axle.  The right tire rim was ground down too, but not as far.  The rest of the plane remained undamaged.

ANALYSIS AND FINDINGS-Excerpts from the Report:

The manufacturer of the NLG shock absorber is Messier-Dowty. The shock absorber is a double-acting, oleopneumatic type with no separator piston.

When the NLG was removed from the airplane, it was determined that the anti-rotation lugs at the top of the shock absorber were not properly seated in the backplate, which indexed and bolted to the inside of the shock strut assembly.

The shock absorber assembly was disassembled using the Job Instruction Cards (JIC) provided by UAL. These were the same JICs used when the strut was assembled during recent maintenance.

This teardown revealed the upper centering cam had been rotated 20 - 30 degrees when it was installed in the inner cylinder. With this condition, in order for the upper cam to mate properly with the lower cam, the sliding tube was rotated 20 to 30 degrees in relation to the aircraft centerline. In order for the axle to be perpendicular with the aircraft centerline when installed, the position of the anti-rotation lugs were rotated 20 - 30 degrees. This resulted in the anti-rotation lugs at the upper end of the strut not being properly engaged in the backplate slots.

Recent Maintenance

A heavy maintenance visit, C-check, on N804UA was completed on November 20, 2002. This maintenance was performed by a contract Federal Aviation Administration (FAA) Certified, Part 145 Repair Station

The NLG dynamic seal was replaced during the maintenance. In order for the seals to be replaced, the shock absorber assembly was removed was removed from the airplane and the sliding tube was removed from the inner cylinder. The seals were replaced, the shock absorber was reassembled, and reinstalled in the aircraft.

Although the maintenance was performed by a contract facility, UAL JICs were used. The assembly tasks on the job cards were derived from the FAA-Approved Airbus Aircraft Maintenance Manual (AMM).

The airplane flew one non-revenue ferry flight and two revenue flights prior to the incident takeoff. The airplane had accumulated a total of 9 hours of flight time between the maintenance and the incident.

UAL reported they had performed about 60 NLG dynamic changes in the past. The contract maitnenance facility had previously performed six NLG dynamic seal changes. The mechanic who performed the dynamic seal change on the incident NLG shock absorber had performed this job two times prior to the assembly for N804UA. The first time was with a trainer and the second time was on his own.

Additional Incidents

On November 1, 2002, an Airbus A320, being operated by JetBlue landed at the John F. Kennedy International (JFK) Airport, New York, New York, with its NLG turned 90 degrees. This airplane had come out of maintenance where the NLG dynamic seal was replaced approximately 3 days prior to the incident. The airplane had flown 15 cycles and 23 hours between the completion of the maintenance and the incident. The investigation into this incident revealed the same findings as were found on N804UA. The upper centering cam on the inner cylinder had been mis-installed. This resulted in the anti-rotation lugs not being perpendicular to the axle and consequently the lugs were not properly seated in the backplate slots.

Messier-Dowty reported that there have been two additional incidents involving the mis-assembly of the NLG shock absorber. One of these involved a Canada 3000 A320 and the other occurred in Ireland.

It sounds to me like this may be a case of designed-induced maintenance error.  

If you design a component, so that it can be installed improperly, sooner or later, human nature being what it is, that will eventually happen.  That is especially true, if the installation procedure is rather complex in nature. As the complexity of the operation increase, so does the risk of human error.  Conversely, if you design it so that it cannot be installed improperly (the old square pegs won't fit in round holes, concept), these kinds of incidents never happen in the first place.  

Frankly, it is beginning to look to me like the French designers haven't learned that basic design concept too well.  See the update below, which gives further evidence of how a bad design can induce repeated maintenance errors, over and over again.

In this present JetBlue accident, it is far too soon to make any judgment as to why the nose landing gear failed in the same manner as those previous occurrences.  Whether it is a design problem, or a maintenance problem, some combination of both, or even some totally new problem, which was not a factor in the previous accidents, remains to be determined by the FAA/NTSB investigation which is now beginning.  

For now, we can all be grateful that JetBlue management is so very, very picky about who it hires to fly its airplanes.

Update, September 26, 2005:

According to a study by the Transportation Safety Board of Canada, the nose-wheels of the A-319, 320 and the 321, also have a failure problem which is of a different nature, than the one which afflicted the JetBlue, United, AWA, Canadian and Ireland flights.  While the type of problem which forced JetBlue to make an emergency landing at LAX, has happened at least 7 times since 1989, the problem cited in the Canadian report has happened 67 times, since 1989. 

It is caused by nosewheel bearing failure and that in turn, appears to be caused by another design problem, which has not yet been resolved.  The investigation was triggered by a failure on an Air Canada Airbus.  

It is not certain as to why the bearings keep failing and causing wheels to fall off, but they suspect it may be a combination of the type and amounts of grease used, to lubricate the bearings, as well as the types of materials used in the bearings themselves. Proper tools and proper torques of the bearing nuts seems to be of vital importance also. 

Quoting excerpts from that report:

Investigations into these occurrences did not reveal a definitive determining factor. Various modifications were implemented in an attempt to resolve this type of occurrence: stiffening of the nosewheel axle with a steel sleeve, a spacer inserted between the nosewheel halves, and a larger nosewheel to accommodate larger wheel bearings. These modifications reduced the frequency of failure, but did not completely eliminate the failures.

Air Canada has taken steps to have the axles for its nose landing gears coated with SermeTel® to reduce the likelihood of axle failures from cadmium infusion as a result of high friction heat generated from bearing failures. SermeTel® is an anti-corrosive and chemical-resistant base applied as an initial coating prior to decorative coatings of epoxy resin and polyurethane paints.

It is an inorganic formula, consisting of an aqueous carrier containing a mixture of magnesium chromate, phosphates and silicates, and aluminum powder. Air Canada has also issued a maintenance alert and revised its wheel installation job card to stress the importance of a wheel inspection prior to installation and to ensure the recommended wheel installation tools and torques are used in accordance with the Aircraft Maintenance Manual.

The Air Canada wheel shop manuals were revised to raise the awareness of the importance of grease dams and seal installations.

In May 2004, Goodrich released Service Letter 1991 that recommends Mobil SHC–100 grease for the wheel bearings due to its superior adhesion properties, which also increase corrosion protection and bearing lubrication. Goodrich also issued Service Bulletin 3–1531–32–3 in July 2004, with new inspection procedures for bearing grease seals on Airbus A318, A319, A320 and A321 aircraft.

Airbus has designed an integrated retaining ring and seal that is being tested for qualification purposes. Testing includes roll and landing tests under different loads, as well as high-pressure water tests to demonstrate the enhanced performance of the new design in protecting the seal and bearing from external contaminants.