Factual context, fresh insights and logical implications from last year’s crash of Asiana Airlines Flight 214 stand to influence settled beliefs about today’s state of crash-related performance of people, equipment and systems throughout commercial air transport. However, substantive changes to aircraft systems, equipment or practices affecting airplane cabin crashworthiness, occupant protection and emergency response likely would be premature pending the U.S. National Transportation Safety Board’s (NTSB’s) completion of the investigation, several subject matter experts told the agency’s Dec. 11, 2013, investigative hearing in Washington.
The NTSB’s preliminary report, last updated in August 2013, summarizes the accident this way: “On July 6, 2013, about 1128 Pacific Daylight Time, Asiana Airlines Flight 214, a Boeing 777-200ER, registration HL7742, impacted the sea wall and subsequently the runway during landing on Runway 28L at San Francisco International Airport (SFO), San Francisco, California. Of the four flight crewmembers, 12 flight attendants and 291 passengers, about 182 were transported to the hospital with injuries and three passengers were fatally injured. The airplane was destroyed by impact forces and postcrash fire. The regularly scheduled passenger flight was operating under the provisions of [U.S. Federal Aviation Regulations] Part 129 between Incheon International Airport, Seoul, South Korea, and SFO. Visual meteorological conditions prevailed at the time of the accident.”
The U.S. Federal Aviation Administration (FAA) recently has published two reports on trends in accident survivability, said Jeff Gardlin, aerospace engineer, Transport Airplane Directorate, FAA. “What they basically show is that the accident rate and the fatality rate in accidents are both dropping, and that the percentage of accidents that are survivable is increasing,” he said. “It’s [also] pretty clear that there is an increase in traffic, and that it is likely that we will see accidents even at a very low rate.”
Major components of FAA’s occupant-survivability approach since the 1980s have included measures clearly relevant to this accident, such as reducing flammability/heat-release of large interior surfaces of the cabin and seat cushions; manufacturing new airplanes and retrofitting older fleets with advanced insulation that resists penetration of the fuselage by external fire; improving exit path markings and lighting; requiring seats dynamically tested to withstand loads up to 16 times the normal acceleration of gravity (16 g) forward along the longitudinal axis; and improving escape slides with changes such as radiant heat–reflective fabrics and low smoke toxicity, he said.
Bruce Wallace, associate technical fellow for evacuation systems engineering, The Boeing Co., and member of the NTSB survival factors team for the accident, described what he called “key events in the impact sequence,” the 777 design features that contributed to the protection of occupants during the crash and how these features performed in the accident. During the initial impact, the widebody jet’s main landing gear and aft fuselage struck the seawall at an airspeed of more than 100 kt (185 kph), he said.
“The severe impact loads caused the main landing gear to separate from the wings,” Wallace said. “The impact also resulted in extensive damage to the aft fuselage structure and separation of the tail. The aft galley, one large cargo container and a great deal of baggage were released onto the runway. The airplane continued to travel down the runway resting on the nose gear and the engines. As the airplane continued down the runway, the left engine separated and the airplane became partially airborne. With the nose gear in contact with the ground, the airplane rotated counterclockwise approximately 330 degrees and impacted the ground in a downward and sideways direction.
“As the airplane slid to a stop, the lower fuselage was peeled open toward the right side of the airplane. The right engine separated and came to rest next to the fuselage. A fire ignited in the right engine oil tank outside the fuselage. From video review of the accident, black smoke can first be seen coming out of door 1L [first main door from front, left side] approximately 15 minutes after the airplane came to a stop. Two escape slides inflated inside the cabin, but all passengers and crew were evacuated from the airplane using alternate exits (see “ARFF Timing Insights”). This [brief summary] does not diminish the fact that there were fatalities and injuries.”
The occupant-safety considerations employed in the 777 design specifically for crash survival reflect the entire airline industry’s decades of evolutionary improvements, he said. “Airplane safety and survivability of airplane interiors emphasize three areas: surviving an impact, surviving a fire and airplane evacuation,” Wallace said. “Keeping the severity of the multiple impacts in this accident in mind, the airplane performed extremely well with respect to each of these goals during … the impact sequence, despite being subject to several severe impacts that likely exceeded the design goals. The passenger seating area remained intact, and the overhead stow bins did not fall on passengers or block their evacuation.”
The separation of the engines and main landing gear, by design, from high-impact forces falls into the category of enabling occupants to survive a post-crash fire. That feature is intended to prevent fuel-tank rupture, and in this accident, the fuel tanks did not rupture. Wallace said, “If a fire occurs, insulation blankets [ASW, 4/08] and cabin materials are designed to resist the spread of fire. … Although fire did occur in the right engine oil tank, its propagation was slowed significantly to allow for evacuation of all passengers and crew.”
A fundamental factor in enabling occupants to evacuate the airplane in this situation was the escape slide systems. For the 777-200 series, the associated timed test for airworthiness certification by the FAA demonstrated that up to 440 occupants, motivated to act urgently in a high-density seat configuration, can evacuate within 90 seconds in night emergency-lighting conditions with a standards-presumed scattering of luggage in aisles and half the slides inoperative (ASW, 1/07).
Regarding the slides that inflated inside the accident aircraft, he said the NTSB’s final determination of the failure mode — tentatively believed to be catastrophic failure of the release mechanism during the impacts — is awaiting the conclusion of the investigation, as are any possible safety improvements. Wallace said that in his Boeing career, he had never known of aircraft damage that resulted in interior slide deployments, and John O’Donnell, president, Air Cruisers (the slide manufacturer), agreed that this failure was unprecedented.
“Some of the features that expedite evacuation are simple-operation doors; seats and other interior components designed to stay secure and not block the aisles; and automatic, self-inflating escape slides,” Wallace said. “Despite the extensive structural damage to the airplane in this accident, the doors opened, the seats and interior components stayed clear of the aisles, and the occupants were evacuated from the airplane.
“It is difficult to foresee, and design to, all possible events that may occur during an airplane accident. This accident included multiple, extremely severe impacts that exceeded the design and certification requirements. Despite this, the 777 occupant-safety features performed extremely well and contributed to the high survival rate. This performance of the 777 airplane highlights the benefits of the work that the regulators, operators and suppliers have done — along with Boeing — to increase airplane safety and accident survivability.”
Survival also was strongly influenced by the design, testing and standards-compliant manufacturing of all types of passenger and crewmember seats in the accident airplane, he said. Airplane cabin components and all items of mass must meet standards requiring that they remain restrained and have enough strength not to fail under static (gradually applied and sustained) loads or dynamic loads (those applied as a sudden impulse). “One [requirement example] is that static loads on all components in the airplane [withstand] 3-g side load; the seats themselves are 3-g side load,” Wallace said. “In addition, we do dynamic testing of 16-g and 14-g [loads] to demonstrate that we can protect the occupants in those kinds of dynamic loads.”
He also cited the overhead stow bins remaining intact as an example of standards improvement based on company analysis and NTSB recommendations. “Over time, as we evaluate airplane accidents, we do discover areas for improvement — and stow bins are one of those,” Wallace said. “The stow-bin systems themselves are designed to the 9 g forward, 6 g down, 3 g side, 3 g up and 1.5 g aft [load criteria]. In addition to that, we do have some flight and ground loads that increase the up-and-down loads, so we design [and test] our stow bins to meet them.”
The survival factors group studied the post-crash condition of the accident airplane’s 16 flight attendant jump seats, located at the eight pairs of exit doors. “Jump seats at [door pairs] 1 through 3 were all intact,” he said. “Door 1R had a twisted seat pan, and door 3L had some deformity on that seat pan, as well, but they were all intact. At door 4R, we lost the aft end of the airplane, and a significant amount of the floor structure underneath it. Two jump seats did end up out on the runway [during] the initial damage to the tail end of the airplane. There was a third jump seat that ended up right-aft of the airplane on the final impact in the final resting zone. So those were all in the area of the airplane [in which impact loads] exceeded the normal 9-g forces that we design our components to.” He also noted that the sequence of initial vertical impact loads, followed by loads during the fuselage rotation and then a significant side load were “not typical of what we have seen with accidents.”
Asked whether, overall, seats in the accident airplane — all manufactured and installed to the 16-g standard to protect occupants in a survivable crash — performed as expected, Wallace cited the fire damage as one factor that precluded a complete assessment. All fatalities involved passengers who had been subjected to the most severe forces while seated in the far-aft rows 41 and 42, he added.
“All the seat legs themselves that we could inspect [in the fire area] were still mounted to the floor tracks,” he said. “Where we had the significant structural damage, the seats were, in general, leaning back. Some of the legs had come loose or had been fractured. There were a lot of legs bent to the left. … Since we had lost the lower portion of the fuselage back there, the floor that supports [seats] was pushed up on the right side; it was very badly damaged. … Without that structure, it’s difficult for us to really consider [the seat performance]. … It’s hard to tell what the forces were on the seats. … They were still able to protect many of the occupants [and] performed pretty well, considering the damage to the airplane.”
One reasonable way of assessing fire-resistance performance of cabin materials in the accident airplane was simply to look at the outcome, Wallace agreed with the NTSB questioners. “The first sign of dark smoke coming out of door 1L was about 15 minutes in [i.e., after initial impact], and that would indicate … with the attitude of the airplane, it was the highest point, the smoke would move that way,” he said. “But … in this accident, the fire was slowed greatly enough that we could get everybody evacuated — including the rescue team that was able to get [out] the people that could not get out themselves.”
Wallace was among witnesses at the investigative hearing who agreed with the view that, given that the investigation is still under way with many details not finalized, the likely best course for the industry is to reserve judgment about the exact lessons learned. Nevertheless, some suggested that one avenue for improvement may involve several characteristics of the impact sequence of Asiana Airlines Flight 214 that have not been seen in any other accidents.
Ensuring Ability to Evacuate
“[The FAA now has] initiatives under way to address airframe-level crashworthiness,” said the FAA’s Gardlin. “We’re looking at means of providing occupant protection or occupant-injury criteria that more closely correlate with what we’re interested in, which is the ability of someone to evacuate an airplane after an accident. The standard we have now has been, to some extent, [carried] across from automotive standards. They are valid standards, but they are not specifically geared to [aviation] problems. We’re looking at novel ways of giving passengers safety information that they might actually assimilate and use, and we have an extensive research and development–prioritization process under way that we are hoping to implement that covers all the areas that I mentioned.”
FAA dynamic-testing regulations require that head-impact protection be provided for transport category airplane occupants, said Richard DeWeese, coordinator, Biodynamic Research Team, FAA Civil Aerospace Medical Institute. “Depending on the seat design — particularly how far apart the passenger rows are — this can be done in different ways. If they are close enough together, then the seat back can be used as an effective energy-absorber. The person flails over onto the seat back, and it absorbs that head-impact energy, reducing the risk of injury. If the rows are far enough apart, then the person flailing forward with just a lap belt restraining them would miss the seat in front entirely, again protecting the head from impact. It’s the seats that fall in between [that are problematic], where the seat back is too far away to really be much of an energy-absorber but it’s close enough to present an injury risk. That’s where something like a shoulder belt or an inflatable restraint system can be used to provide head-impact protection.” In the accident airplane cabin, only the flight attendants and the passengers in first/business-class seats had lap belts and shoulder harnesses so that these occupants would have protection at least equivalent to occupants of economy-class seats, he said. He added that such restraints also are superior in the level of protection afforded.
ARFF Timing Insights
The December 2013 testimony before the U.S. National Transportation Safety Board by subject matter experts in aircraft rescue and fire fighting (ARFF) in part provided insights into the time factor involved in responding to Asiana Airlines Flight 214 — a critical context for the performance of the accident aircraft and its occupant survival–related equipage.
“The initial impact … occurred at 1127,” said Dale Carnes, assistant deputy chief, San Francisco Fire Department. “Approximately 12 seconds later, the U.S. Federal Aviation Administration tower dispatched an ‘Alert 3 in progress’ to all three fire stations. At 1131, two minutes after being dispatched and three minutes after the initial impact, the first ARFF unit, Rescue 88, arrived on scene, followed 37 seconds later by Rescue 9. Upon seeing that the initial passenger-egress paths on the left side of the aircraft were not threatened, both units immediately attacked the fire in engine no. 2, which was resting against the right side of the aircraft.
“By 1133, a little less than six minutes post-impact, all seven airport firefighting companies [teams of firefighters led by fire officers] and two paramedic units were on scene. Paramedics were beginning to collect casualties and to initiate triage. Approximately one minute later, the first of 56 ground ambulances arrived on scene. At 1138, the ARFF personnel entered the aircraft for interior search and rescue. By 1146, all trapped passengers and those who had remained behind to assist had been removed from the aircraft. By 1218, approximately 19 minutes after the impact, all fire in the fuselage had been extinguished with the support of companies from San Francisco and mutual aid [vehicles and personnel] from San Mateo County.
“At 1301, the last patient from the airfield was transported by ambulance. All ambulatory passengers had been relocated to the terminals, and a significant number of those passengers later self-reported injuries and underwent secondary triage. By 1758, the last of those passengers were transported from the terminals to area hospitals. All told, 56 ground ambulances, two medical helicopters and two buses transported 179 patients to 12 area hospitals.”
Interior hose lines failed to suppress the large fuselage fire that was spreading rapidly during ARFF operations, Carnes said. Upon the completion of evacuation and rescue of at least four trapped people, the operator of a vehicle equipped with a high-reach extendable turret and a piercing nozzle penetrated the crown of the fuselage three times to inject firefighting agents that were successful in extinguishing the fire. “We had a total of nine of our 22 personnel [and some non-ARFF airport staff] that were inside the aircraft at any one time during the rescue,” Carnes added. The NTSB’s investigation continues into the cause and manner of death of one of the passengers found fatally injured outside the aircraft.