January 2003 Volume 46 / Number 1
A boot in the posterior
Bruce Landsberg is the executive director of the AOPA Air Safety Foundation.
Animations and videos presented at the National Transportation Safety Board public hearing for American Airlines Flight 587 held October 29-November 1, 2001, are available at the NTSB Web site.
It's exceedingly rare to have an aircraft come apart, especially when it is flown within the design envelope. So there was great surprise and consternation when American Flight 587, an Airbus 300, shed its vertical stabilizer and rudder while flying well below maneuvering speed (VA). It would be understandable if the jet started shedding pieces at high speed after a robust control deflection, but we've all been told that if the aircraft is below VA, it will stall before bad things happen. What are the implications for pilots of light aircraft?
The preliminary facts in the case from the NTSB are as follows: On November 12, 2001, an Airbus A300-600R crashed shortly after takeoff from John F. Kennedy International Airport. The vertical stabilizer and rudder were found in Jamaica Bay, about one mile from the main wreckage, while the engines struck the ground several blocks north of the main wreckage.
Based on radar and flight data recorder (FDR) information, Flight 587 took off approximately 101 seconds behind a Boeing 747. The Airbus encountered two wake vortices generated by the 747. The second wake encounter occurred only 10 seconds before the FDR recording ceased, which coincided with the breakup of the airframe. During the second wake encounter the first officer, who was flying, appears to have made some strong control inputs — far stronger than any roll induced by the vortex. The aircraft experienced three strong lateral movements, two to the right, and then one to the left, which seemed to be caused by rudder movements. What is disturbing is that the indicated airspeed was about 255 knots, which is well below the design maneuvering speed of 273 KIAS for the Airbus. According to conventional, but perhaps flawed, wisdom the aircraft should have held together.
At this point in the investigation neither sabotage nor weather are considered factors. ATC clearances given to Flight 587, and spacing between the flight and the preceding 747, were in accordance with current guidelines. Nevertheless, the Airbus did have a wake encounter. There are no indications of a rudder system anomaly.
Likewise, the maintenance log shows no problems. Investigators looked for preexisting damage on the vertical stabilizer. This aircraft had been involved in a severe turbulence incident several years before but no damage was found. The NTSB noted that even if there were pre-existing damage to the stabilizer, the structure remained intact until massive overloads were sustained. Aerodynamic and internal stress calculations on the vertical stabilizer computed by Airbus, and independently by the NTSB, show that loading was significantly above the ultimate loads required by certification standards and actually near the structural test loads demonstrated during the certification process.
In a somewhat unusual move, the NTSB issued two critical recommendations to the FAA very early in the investigation. "For air carrier pilots, some training programs do not include information about the structural certification requirements for the rudder and vertical stabilizer on Transport category airplanes. Significantly, full opposite rudder inputs, even below the design maneuvering speed, may result in structural loads that exceed certification requirements."
There is a point of contention between American Airlines and Airbus regarding training of pilots in response to upsets, and this was a central point of recent NTSB hearings. In the mid-1990s American Airlines, using simulation, suggested that pilots use "coordinated rudder" input in response to a major diversion. Airbus claims that the training scenarios and simulators do not accurately replicate the aircraft response and have letters dating back to 1997 voicing a different opinion regarding aggressive use of rudder and the methodology of training.
As far as maneuvering speed is concerned, for Transport category aircraft the latest word is that the airframe will withstand one quick rudder deflection in one direction without damage to the airframe. There is no requirement for the airframe to withstand full deflection in one direction immediately followed by full deflection in the other direction.
According to the NTSB, "Pilots may have the impression that the rudder limiter systems, which limit rudder travel as airspeed increases to prevent a single full rudder input from overloading the structure, also prevent cyclic full rudder deflections from damaging the structure. This is not true. The structural certification requirements for Transport category airplanes do not take such maneuvers into account; therefore, such cyclic rudder inputs, even when a rudder limiter is in effect, can produce loads higher than those required for certification and that may exceed the structural capabilities of the aircraft."
Because of the confusion regarding the certification requirements for the vertical stabilizer, the NTSB asked the FAA to ensure that pilot training programs: "(1) explain the structural certification requirements for the rudder and vertical stabilizer; (2) explain that a full or nearly full rudder deflection in one direction followed by a full or nearly full rudder deflection in the opposite direction can result in potentially dangerous loads on the vertical stabilizer; and (3) explain that, on some aircraft, as speed increases, the maximum available rudder deflection can be obtained with comparatively light pedal forces and small pedal deflections." According to informed sources, this may be as little as 30 pounds of pressure and only a few inches of travel.
How does this relate to those of us flying light aircraft? A copy of an old FAA Flight Training Handbook noted, "Intentionally stalling an aircraft above its design maneuvering speed will impose a tremendous load factor."
The rudder is not specifically mentioned, and it seems more likely that the elevator controls would be the overcontrol device of choice. Out of sight, out of mind may apply here. We're typically focused on elevator or ailerons, which tend to be "more powerful" and more commonly used controls.
You may remember that VA is defined as the speed at which a full control deflection can be made abruptly and the aircraft will stall before any damage results to the airframe. The load factor squares as the stall speed doubles, so loading the wings, tail, or rudder enough can break the structure. A new aircraft, certificated in the Normal category, is limited to 3.8 Gs but has a safety margin of 50 percent so it shouldn't fail until 5.7 Gs are reached. Older aircraft may fail at lower G levels because of fatigue or if someone else explored the limits before. That's a good reason not to fly overloaded. It probably won't break the first time or even the fifth time that maximum gross weight is exceeded, but it would be really irritating to have the airframe fail in moderate turxulence, well below maneuvering speed, because some other pilots had used up the margin built into it. According to the FAA Flight Training Handbook, "The cumulative effect of such loads over a long period of time may tend to loosen and weaken vital parts so that actual failure may occur later when the aircraft is being operated in a normal manner."
In old aircraft, VA was estimated at 1.7 times the normal stall speed. If the airplane stalled at 42 knots in normal unaccelerated flight, 71 knots was as fast as you'd dare take it in turbulence. If you can't find VA in the pilot's operating handbook, use that formula. On newer aircraft the design maneuvering speed is published. For a 1979 Cessna 172N with a level-flight, flaps-up stall speed of 42 knots indicated airspeed (KIAS), the published VA is 97 KIAS at maximum gross weight.
Maneuvering speed works contrary to intuition — as the gross weight of the aircraft goes down, so does the maneuvering speed. In the aforementioned Skyhawk, at 1,950 pounds VA drops to 89 KIAS and at 1,600 pounds more than 80 KIAS will likely cause damage after a maximum control input.
The salient point that light-aircraft pilots should take away from this discussion is that large, rapid control movements, even below VA, should be done with consideration of extra load added to the structure. Rapid reversals may not break a light GA aircraft, but they sure aren't going to do the airframe any good. Unless the ground is imminently close, a smooth and relaxed hand (or foot) will beat a tight fist or a boot in the posterior every time.