RSS

Ditching After Bird Strike Probed by Safety Board

Thu, Feb 26, 2009 — David Evans

Articles, Featured

Despite the successful ditching of a US Airways A320 into the Hudson River 15 January 2009, and the saving of all lives aboard, the National Transportation Safety Board (NTSB) is conducting a full-blown investigation into the accident.

The airplane encountered a flock of birds and lost all power when the engines attempted to swallow feathered carcasses.

The accident airplane being craned out of the water. Note the missing left engine, torn off by the force of impact with the water.

The accident airplane being craned out of the water. Note the missing left engine, torn off by the force of impact with the water.

Pursuant to its investigation, the NTSB announced 19 February it will hold a two-day fact-finding hearing later this spring or summer (the precise dates to be announced later).

Mark Rosenker, the NTSB acting chairman, said, “Based on what we have learned so far about this accident, we know that many things went right.”

“But no matter how many things go right, we’ve found that each accident presents safety issues that we can learn from – both to further our investigation and, ultimately, to make the skies even safer. This hearing will move us closer toward those goals,” Rosenker said.

The hearing, when it’s finally held, will focus on these issues:

  • Training of crew members on emergency procedures.
  • Certification requirements for the Airbus A320 related to the structural integrity of the airframe during ditching.
  • Bird ingestion certification standards for transport-category turbofan engines.
  • New and developing technologies for detecting large groups of birds and procedures to avoid conflicts with birds in the vicinity of airports.

The airplane ingested at least two Canada geese – one into each engine – after taking off from New York’s LaGuardia Airport bound for Charlotte, NC. Canada geese can weigh up to 16 pounds, and engines are tested only to safely absorb the impact of 1.5 pound birds and keep functioning.

sullenberger

The airplane encountered the geese at about 3,200 feet. The airplane’s two engines, clogged with feathers, bone and muscle, and the fan blades damaged by impact, lost all thrust. Captain Chesley “Sully” Sullenberger, piloting the stricken aircraft, radioed the air traffic controller, “We may end up in the Hudson [River].”

It should be pointed out that Sullenburger was an experienced glider pilot. Without the thrust of his A320’s two jet engines, he was basically in a larger glider, now with 155 people aboard.

The transcript of the New York Terminal Radar Approach Control Facility (TRACON) provides a gripping account of the ditching (all times local; call sign of Sully’s US Airways plane is “Cactus 1549”; transcript edited):

Time Radio Transmission Message
3:25:51 pm Cactus 1549 Cactus 1549 seven hundred climbing five thousand.
3:27:32 TRACON Cactus 1549 turn left heading 270.
3:27:49 TRACON Tower, stop your departure, we got an emergency returning.
3:27:53 LaGuardia tower (LGA) Who is it?
3:27:54 TRACON It’s 1529 he ah bird strike he lost all engines he lost the thrust in the engines he is returning immediately.
3:27:59 LGA Cactus 1529 which engines?
3:28:01 TRACON He lost thrust in both engines he said.
3:28:03 LGA Got it.
3:28:05 TRACON Cactus 1549 if we can get it to you do you want to try and land [on] runway one three?
3:28:11 Cactus 1549 We’re unable we may end up in the Hudson.
3:28:31 TRACON Alright Cactus 1549 it’s going to be left traffic to runway three one.
3:28:34 Cactus 1549 Unable.
3:28:36 TRACON Okay, what do you need to land?
3:28:46 TRACON Cactus 1549 runway four is available if you want to make left traffic to runway four.
3:28:50 Cactus 1549 I am not sure if we can make any runway oh what’s over to our right? Anything in New Jersey maybe Teetorboro?
3:28:55 TRACON Okay, yeah, off to your right side is Teeterboro Airport.
3:29:02 TRACON Do you want to try and go to Teeterboro?
3:29:03 Cactus 1549 Yes.
3:29:05 TRACON Teterborough uh empire actually La Guardia departure got an emergency inbound.
3:29:10 Teterboro Okay, go ahead.
3:29:11 TRACON Cactus 1529 over the George Washington bridge wants to go to the airport right now.
3:29:14 Teterboro He wants to go to our airport, check. Does he need any assistance?
3:29:17 TRACON Ah, yes, he, ah, he was a bird strike. Can I get him in for runway one?
3:39:19 Teterboro Runway one that’s good.
3:29:21 TRACON Cactus 1529 turn right 280. You can land runway one at Teterboro.
3:29:25 Cactus 1549 We can’t do it.
3:29:26 TRACON Okay, which runway would you like at Teterboro?
3:29:28 Cactus 1549 We’re gonna be in the Hudson. (This is the last radio transmission from the airplane, which is being flown like a glider by Sullenberger.)
3:29:33 TRACON I’m sorry, say again, Cactus.
3:29:51 TRACON Cactus, ah, Cactus 1549 radar contact is lost, you also got Newark Airport off our two o’clock and about seven miles.
3:30:09 Eagle flight 4718 Two one zero un forty seven eighteen I don’t know, I think he said he was going in the Hudson.
3:30:14 TRACON Cactus 1529 uh you still on?
3:30:22 TRACON Cactus 1529 if you can, ah, you got, ah, runway two nine available at Newark off your two o’clock and seven miles.
3:31:30 UNKN Was that Cactus up by the Tappan Zee?
3:31:32 TRACON Uh, yeah, it was Cactus. He was just north of the ah George Washington Bridge when they had a bird strike.
3:33:47 TRACON Uh, I guess it was a double bird strike and he lost all thrust.

 

Bird strike was at about 3:27 pm at 3,200 feet. As shown, the aircraft was descending from that point until contact with the water.

Bird strike was at about 3:27 pm at 3,200 feet. As shown, the aircraft was descending from that point until contact with the water.

At this point, the US Airways plane had splashed into the Hudson River (right about the time Eagle flight 4718 said the stricken plane was “going in the Hudson”). Passenger Fred Berretta said later that he was expecting the plane to flip over and break apart. However, Sullenberger took care to strike the water tail first. By keeping the nose high, Sullenberger minimized the hydraulic forces on the engines and only one engine actually broke from the airplane on impact with the water.

Witness Ben Vonklemperer said he watched the landing from the 25th floor of his office building. “If someone’s going to land a plane in the water, this seemed the best possible way to do it,” Vonklemperer said. “The way they hit was very gradual. A very slow contact with the water.”

Total time of flight: about five minutes.

Note the fourth passenger from the left incorrectly wearing the life vest upside down. In moments of great stress, the simplest of instructions from the flight attendants could be confused.

Note the fourth passenger from the left incorrectly wearing the life vest upside down. In moments of great stress, the simplest of instructions from the flight attendants could be confused.

A ditching switch on the airplane increased the likelihood that it would float, at least long enough for all aboard to evacuate. The ditching switch effectively seals the airplane by closing valves and ventilation ports on the bottom of the fuselage. With the valves and ports shut, a “float line” is created; water will eventually seep in, and the airplane’s weight will drive it to the bottom. This feature is apparently unique to the Airbus family of airplanes. This important adjunct to occupant evacuation will doubtless be a mentioned in the final report of investigation, with a recommendation that it be adopted for all aircraft manufactured after a certain date.

The ditching switch, which sealed openings in the lower fuselage, contributed to the airplane’s staying afloat long enough to be safely evacuated. Note passengers on wing.

The ditching switch, which sealed openings in the lower fuselage, contributed to the airplane’s staying afloat long enough to be safely evacuated. Note passengers on wing.

There are a couple points worth considering even at this early juncture in the NTSB investigation. For one thing, with loss of all engine thrust, Sullenberger was basically flying a glider. Unfortunately, he probably had nil potential for any return to LaGuardia. According to contributing editor John Sampson (who, like Sullenberger, is an experienced glider and air transport pilot), Sullenberger “wisely opted NOT to try for the short runways and poor under/overruns of the short runways at Teterboro.”

“Cross-country glider pilots make regular and routine ‘outlandings,’ but rarely on water,” he added. “But a ditching along the Hudson was a good choice. Proof of the pudding!”

The Airbus A320 is a fly-by-wire airplane. That is, there are no cables connecting flight controls in the cockpit with aerodynamic surfaces such as spoilers, ailerons, rudder, etc. Rather, electrical signals from the cockpit are routed through the flight computers to the flight controls, which act in accordance with the electrical signals.

With the engines knocked out, the electric power from the engine-mounted generators was also lost. Either the ram air turbine (RAT) deployed to provide a modicum of electric and hydraulic power, or the auxiliary power unit (APU) was running.

Taking off with the APU running is not a standard operating procedure (SOP) requirement but is entirely at the discretion of the captain. Since LaGuardia’s runways are only 7,000 feet – adequate but not generous – a decision to power cabin pressurization through the APU instead of through the engines is valid, and would fit the kind of pilot this captain appears to be. In any event, some pilots start the APU against the possibility of birdstrike. Or the APU could have been started right after the birdstrike, when both engines showed reduced/no power.

With APU and/or RAT in use, the airplane would not have been in the Normal Law. Having lost both AC main busses, direct fly-by-wire laws disable all flight envelope protection. That is, if flown too slowly, the airplane WILL stall. Flown in Direct Law, the cockpit sidesticks are directly coupled to the controls via the computers, but without any of the stabilization feedbacks. In effect, the fly-by-wire aircraft is turned into a conventional airplane (albeit one compensated for configuration and center of gravity).

With this limited backup system, Sullenberger obviously kept the airplane above stall speed. It is estimated that his stall speed was around 110-115 knots. Based on radar returns, Sullenberger glided down the Hudson River at about 190 knots. The last radar return suggests the airplane was travelling at 153 knots just before touching the water. The pitch attitude of the aircraft at about 600 feet and at touchdown indicates a much higher speed than one close to stall.

The Primary Flight Display (PFD) for the captain (and likely the first officer, if the APU was running) would have had basic information for an aircraft in Alternate or Direct Law. That information would have been “Vls,” for Velocity Lowest Selectable, which provides sufficient protection above the stall speed (but not the automatic stall protection). Also displayed was “Vsw,” or Velocity Stall Warning, which is the actual stall speed. The airplane will stall if taken below that speed. Since the aircraft was in controlled flight right through the moment of touchdown, it was also above the stall speed.

It is suspected that Captain Sullenberger was “flying attitude” as opposed to strictly airspeed as he managed the energy of the airplane, bleeding speed for a controlled touchdown at a point he would have selected and kept in the same position in the windscreen while losing altitude and watching the speed very carefully.

In cases of forced landing, trading speed for altitude is key. But in this case, the “landing field” was almost unlimited, barring boat or ship traffic, so Sullenberger had the very best opportunity to control the reduction in airspeed while ensuring a pitch attitude for touchdown that was not too nose down and, more importantly, not too nose high. He might not have had that opportunity in any ground-based landing surface within his gliding range; both Teterboro and LaGuardia were poor choices for this reason. The necessary attitude was about 3-5 degrees nose up. Sullenberger seemed to keep that attitude right to touchdown as energy bled off. In fact, he would have managed the energy of the aircraft to achieve this state of affairs, meeting the touchdown point at the appropriate speed and pitch attitude – not near the stall speed but not too high a speed, either. Such energy management, while innately understandable to all airmen, nevertheless represents a superior example of flying under emergency conditions.

His technique made the touchdown a typical three-point “taildragger,” using the engines as the forward landing gear and the aft fuselage under-surface as the “tail wheel.” The structure absorbed a tremendous amount of forward energy in a very short period of time, but that is the way a forced landing is done. As the saying goes, “Fly the aircraft as far into the crash as possible.”

Once the bottom lips of the engine cowling immersed themselves in the water, very high loads on both engines and additional loads placed on the fuselage slowed the aircraft in a very short distance. Once the three points touched the water, everybody aboard was along for the ride until the aircraft stopped.

Exactly how Sullenberger accomplished the landing, the crew resource management (CRM) that was utilized, and the strengths and weaknesses of the Airbus fly-by-wire system in these circumstances, will be closely examined by the NTSB. There are doubtless lessons aplenty to be learned.

Of equal interest is the adequacy of LaGuadia’s bird control program and the resistance of current turbofan engines to bird ingestion.

In the United States, the population of Canada geese has increased about four-fold since 1970 and now tops 5.4 million birds in 2008 – up from just 1.2 million in 1970. As Richard Dolbeer and John Seubert, both recently retired from the Wildlife Services branch of the Department of Agriculture, wrote in a recent paper:

“The dramatic increase in the resident population of Canada geese should be of particular concern to the aviation industry because of the following four attributes of these geese: 1) large size (typically 8-10 lbs), which greatly exceeds the 4-lb bird certification standard for airframes and most engines, 2) flocking behavior which increases the likelihood of multiple bird strikes (44% of Canada goose strikes, 1990-2006), 3) attraction to airports for grazing and resting, and 4) year-round presence in urban environments near airports.”

Of the 1,349 Canada goose and likely Canada goose strikes (unidentified as to species but likely Canada geese) with civil aircraft reported during the seven year period (1990-2006), 716 of the impacts caused damage. Damage to one or more engines was reported in 174 cases.

This is the number of non-migratory Canada geese, or the potential population that lurks around all the time. Migratory geese brings the total North American population to 5.4 million birds.

This is the number of non-migratory Canada geese, or the potential population that lurks around all the time. Migratory geese brings the total North American population to 5.4 million birds.

Canada geese are the most massive bird commonly struck by aircraft in North America. Because of their size and flocking behavior, Canada geese have been responsible for a disproportionate amount of damage to aircraft. For example, from 1990-2006 Canada geese were involved in 4.4% of all reported bird strikes, but they caused 18% of the strikes causing damage and 26% of the reported costs.

In testimony 24 February before the House Transportation and Infrastructure, Peggy Gilligan, the FAA’s Associate Administrator for Safety, perhaps inadvertently revealed some of the problems with the FAA’s approach.

For example, she proudly proclaimed that the FAA has “collected over 100,000 voluntary wildlife strike reports” since 1990. The operative word is “voluntary.” The FAA does not require reporting and thus may have an undercounted assessment of the threat.

She indicated that the FAA has two pilot projects under way to assess whether Bird Radar can reliably detect flocks near airports. “We are planning additional testing at Chicago O’Hare, Dallas-Ft. Worth, and John F. Kennedy International airports starting later this year,” she announced. Here’s another example of the FAA engaging in tests years after a hazard was identified, and further years will pass before anything is operationally deployed.

She noted that the downed A320 was powered by CFM56-5B4/P engines, which were certified to meet these requirements:

Flocking birds – the engine must be able to ingest a flock of birds (seven 1.5 lt birds), not lose more than ¼ of its power and continue to run for 5 minutes at the takeoff power setting.

Single bird – the engine must be able to ingest a single large bird (4 lbs) and be able to shut down safety. Continued operation is not required. Rather, the engine is designed to shut down with no hazardous debris or fire.

Canada goose feather, a bit worse for wear and tear, retrieved from an engine of the US Air A320. The feather is from an adult goose, which can weigh considerably more than engines are tested and certified to ingest and keep functioning.

Canada goose feather, a bit worse for wear and tear, retrieved from an engine of the US Air A320. The feather is from an adult goose, which can weigh considerably more than engines are tested and certified to ingest and keep functioning.

Obviously, these standards are inadequate protection against Canada geese, which weigh more and flock together, posing the risk of multiple strikes on one airplane. Hardening the engines, though, implies enormous weight penalties.

The FAA may be dreaming if it thinks it can develop bird strike mitigation measures that will work beyond airport boundaries, let alone as high as 3,000 feet on departure. The US Airways jet was climbing through 3,200 feet when it encountered the flock of Canada geese.

What is not discussed by the FAA is a standalone system of detection and deflection, organic to the aircraft. For example, fuselage mounted lasers could perform double duty: as a futuristic ice detection, profiling and elimination technique; the system could also detect and then “burn” and deflect any avian intruders up to a mile or two ahead, in a fan-shaped array along the projected flight path.

Perhaps this approach could be assessed as the NTSB evaluates at the upcoming hearing “new and developing technologies” for dealing with the bird strike hazard.


Comments are closed.

Nolan Law Group