Aircraft Accidents and Lessons Unlearned LXIX: Air France Flight AFR 4590

On Tuesday, July 25, 2000, at 14:44 (2:44 PM) Universal Time Co-ordinated (UTC) an Air France Concorde (AFC), registration F-BTSC, a charter flight, took off from Runway 26 Right at Roissy Charles de Gaulle Airport – CDG – bound for John F Kennedy International Airport – JFK. At some time before rotation, the Concorde’s number 2 tire – left main gear bogie, forward right tire – became damaged. Debris from the tire struck the wing structure; the number 5 tank was ruptured. A fire resulted; a loss of power followed, and the aircraft could not climb away nor retract its gear. The aircraft became uncontrollable and struck terrain in the form of a hotel. The report number f-sc000725a was adopted, which will be referred to as ‘the report’.

The Bureau Enquêtes-Accidents (BEA) led the official investigation into the accident. Per the provisions of Annex 13 to the Convention on International Civil Aviation (ICAO) British representatives from both the British Aircraft Corporation (BAC) – the airframe manufacturer and Rolls Royce – the engine manufacturer were added to the investigation team. Also, party to the investigation were the Federal Aviation Administration (FAA), the German aircraft accident investigation group: Bundesstelle für Flugunfalluntersuchung (BFU), and the United States (US) National Transportation Safety Board (NTSB). The NTSB – not the FAA – was chosen as the accredited representative for the US. This was a mistake; the NTSB had (has) no experience in aircraft maintenance; this would become evident later.

“In accordance with Annex 13 of the Convention on International Civil Aviation [ICAO], with EC directive 94/56 and with Law No 99-243 of 29 March 1999, the analysis of the accident and the conclusions and safety recommendations contained in this report are intended neither to apportion blame, nor to assess individual or collective responsibility. The sole objective is to draw lessons from this occurrence which may help to prevent future accidents or incidents.” The BEA ignored ICAO’s orders. The BEA apportioned blame incorrectly; they erroneously assessed individual responsibility.

The BEA Probable Causes: The accident was due to the following causes:

· High-speed passage of a tire over a part lost by an aircraft that had taken off five minutes earlier and the destruction of the tire.

· The ripping out of a large piece of tank in a complex process of transmission of the energy produced by the impact of a piece of tire at another point on the tank, this transmission associating deformation of the tank skin and the movement of the fuel, with perhaps contributory effect of other more minor shocks and/or a hydrodynamic pressure surge.

· Ignition of the leaking fuel by an electric arc in the landing gear bay or through contact with the hot parts of the engine with forward propagation of the flame causing a very large fire under the aircraft’s wing and severe loss of thrust on engine [number] 2 then engine [number] 1.

In addition, the impossibility of retracting the landing gear probably contributed to the retention and stabilization of the flame throughout the flight.

The BEA put great emphasis to an engine core cowl rub strip allegedly lost by a Continental Airlines (CAL) DC-10 as being the cause of the accident. Page 15 of the report shows ‘witness testimony was taken, but no testimonials were in the report. On news.bbc.co.uk/2/hi/europe/8492561.stm, CAL’s lawyers, in February 2010, said, “… they can prove the Concorde caught fire before it struck the titanium strip.” Whether or not the AFC was already on fire, the BEA’s decision to blame the accident on a rub strip was absurd. As an accredited representative, the NTSB failed to call the BEA out on this, primarily because of the NTSB’s complete lack of experience in maintenance specialties. By pursuing this course, the NTSB and the BEA also failed to find the true cause of the accident.

To avoid tangential issues, such as why the BEA didn’t investigate how Air France failed to install a main landing gear spacer during its last maintenance visit – the missing spacer may have contributed to the gear’s inability to retract – we will focus on the BEA’s failure to determine the engine fire’s origin.

On page 161 of the report, at 39 seconds into the takeoff roll, the following statements were made: “The aircraft was about 1700 meters from the threshold, in the area where the first parts of the water deflector were found. WAIT! The water deflector came off in pieces? It was probably at that moment that tire No 2 ran over the metallic strip.” Distances were given and then, “It is likely that this noise resulted from the damage to the tire. It was in fact in this area that the metallic strip and large piece of tire were found.”

It's unclear how the BEA surmised the rub strip – and not the water deflector – caused the accident or why the BEA used ambiguous words like ‘probably’ and ‘likely’ instead of facts or evidence. Was it possible the tire ran over the water deflector? The AFC’s weights provided on page 31 of the report raised questions, such as did the metal rub strip leave a mark in the runway when the 184,000-kilogram (405,650-pound) aircraft pushed down on it? Were there runway scrapes, indentations, dislodged tar? Figure 21 on page 61 shows the alleged rub strip after it was allegedly run over by the AFC.

My Maglite was run over by the main tire of a 200,000-pound B727-200 freighter in Eppley Airfield; the Maglite was crushed to zero, batteries and all. Why wasn’t the ‘rub strip’ flattened by the 405,000-pound Concorde? How did it maintain an edge up position? The rub strip appears unaffected by the AFC as does the ground below it. A core cowl rub strip doesn’t have that shape. Where is line 152 where it was found? Was it runway centerline or to the side? How was it found with reference to the AFC’s takeoff direction? Any runway damage from the rub strip? Why hadn’t the rub strip been ingested by a wing-mounted engine? Does it even resemble a DC-10 engine core cowl’s rub strip? If made from a tougher metal, how did it get so deformed before the AFC ran over it five minutes after the DC-10 departed?

Also, how many landings were on the number 2 tire? Should it have been replaced prior to flight? As found on page 26 of the report, Air France hadn’t used retread tires on its Concordes since January 15, 1996. Why? For what reason were retreads no longer used on Air France’s Concordes? Why did the tire rupture during takeoff? Did the BEA consider the fact the AFC’s number 2 main tire had been installed two months earlier, that a main tire’s sidewalls take a lot of punishment in that time. The numbers 2 and 3 tires were on the longest, 58 days and 69 days, respectively. The NTSB should have asked these questions. The NTSB, as accredited representatives, should’ve been looking into the tires’ conditions.

The report stated on page 105 in section 1.16.6.2.1 Disassembly and Repair of Wear Strips that replacing a rub strip is a ‘minor repair’ and that a minor repair does not imply replacement of structural elements. That may be an ICAO definition, but in the code of federal regulations (CFR), a minor repair is defined as ‘a repair other than a major repair’. A major repair, ‘means (1) if improperly done, might appreciably affect weight, balance, structural strength, performance, powerplant operation, flight characteristics or other qualities of airworthiness, or (2) that is not done according to accepted practices or cannot be done by elementary operations.’ The definition, which Continental (CAL) used per Title 14 CFR Part 1.1, did not prohibit the use of titanium as opposed to stainless steel. CAL could have used aluminum or even phenolic. The BEA’s ‘minor repair’ argument made no sense. Why did the tire cause the fire?

It was unlikely the rub strip caused the damage to the AFC’s number 2 main tire for several reasons. For one, the sample rub strip test results shown on page 100 of the report stated the Toulouse aeronautical test center – CEAT – created a sample ‘rub strip’ that was shown to have cut into the tire’s plies. However, the sample was spot welded to a plate to stabilize it during tests; this would have prevented natural movement of the rub strip as the test tire ran over it, so the results were skewed. Per Figure 51 on page 102, the rub strip’s arc shape was too long; the strip couldn’t have stood on its edge.

Second, as shown in Figure 48 on page 99, the test strip was bent at angles a titanium rub strip wouldn’t have been twisted to if it fell from a core cowl. The test strip’s shape was over exaggerated. The rub strip was too long and at too shallow an arc to maintain an ‘on-edge’ attitude. These two points alone disqualified the test results.

Third, the rub strip has a 90-degree angle on its leading edge that fits over the reverser cowl edge; this would have prevented it from falling off the CAL DC-10’s core cowl onto the ground. Instead, the rub strip, per figure 52, would’ve ended up inside the core cowl, not on the runway. In addition, the pressure from the core cowl latches would have prevented the strip from exiting the cowl unless someone standing on the runway opened the cowl. How did the BEA prove the rub strip was positioned just right to cut the tire; that it was on the centerline of the runway; that it was on the spot where the tire was damaged? Was that guesswork in place of facts? More importantly, to be clear, the BEA ignored the enormous damage done by a tire to focus instead on a rub strip the BEA could not even prove caused the tire failure. This was misdirection and/or ignorance.

This was why using the NTSB as the accredited representative was a mistake. The NTSB didn’t use aircraft mechanics to investigate maintenance issues. NTSB investigators didn’t understand the basics of how an engine cowl works or what the lifetime of a main tire is.

But if we assume the tire was cut … somehow … why would a tire failure cause so much damage and lead to an engine fire? This was a design issue; the root cause was that the Concorde was susceptible to this type of event from any main gear tire failure, for any reason. If the plies came off; a sidewall ruptured; or an overheated tire exploded, all could have led to the same events as AFR 4590. Take it one step further and compare it to another catastrophic event: the Space Shuttle Columbia. The shuttle was moving at between 686 and 922 kilometers per hour (km/h) or 426 and 573 miles per hour. The hard foam section that separated at this speed damaged the Columbia’s reinforced carbon fiber panels, where it compromised the shuttle’s heat shield. As far as could be determined, the hard foam didn’t penetrate the shuttle’s pressure vessel or wing section.

On AFR 4590, the number 2 tire tread penetrated the AFC’s outer wing skin and fuel tank as it was taking off at 200 km/h. Tire failures are more frequent than people realize. Therefore, to have a Concorde crash because of a tire failure was more a BAC design failure than a rub strip-induced problem. This BAC design failure – a major root cause – was missed (dismissed?) by both the BEA and the NTSB, as they searched for a completely irrelevant probable cause in the form of a rub strip.

However, the most telling piece of ignorance came in a British Broadcasting Company (BBC) article from December 2010, a decade after the accident. The journalist reported the Continental mechanic was cleared of charges of manslaughter – yes, manslaughter! – perhaps due to a lack of evidence. The judge, who found Continental financially responsible for the accident, confirmed the titanium debris dropped from a Continental DC-10 determined, “[The mechanic] should not have used titanium parts to make repairs on the DC-10 because the metal was known to be dangerous for aeroplane tires, and he should have used a softer metal, aluminum, instead, the court found.” This was the judge who found Continental Airlines financially responsible for BAC’s design error, saying that “… titanium was ‘known … to … be’ dangerous for aeroplane tires …” Perhaps the judge was a wannabe metallurgist … or something.

This is Aviation Lessons Unlearned (ALU) LXIX, the 69th accident article written to apply root cause analysis into failed accident reports to help aviators understand why aircraft crash, and not just from an aircraft maintenance perspective. After six years of writing these articles, the question that comes to mind is: Does anyone actually read these accident reports? These ALU articles are written using limited resources, and yet the investigation agencies’ accident reports are full of mistakes, distractions, inexperience, intentional disregards, and blunders that can only be classified as anti-safety.

This accident was misdirection on an international scale; BAC messed up their design and tried to blame a mechanic in the US for their failure. The BEA and NTSB helped them do it. It is truly discouraging that investigators fail to find facts; fail to apply root cause analysis; they rely heavily on emotions and hearsay; and don’t give accident investigations their best. And the aviation industry doesn’t become safer.