Updated: NTSB Probe Into MD-87 Houston Overrun Accident Finds Hard Evidence
The NTSB released an update to its ongoing investigation into the runway overrun of a McDonnell Douglas MD-87 in Houston last month (October 19, 2021). The converted airliner was heading to Boston to watch the hometown Astros play the Red Sox at Fenway Park in Boston, Massachusetts. The lightly loaded jet never got off the ground.
After the crew had initiated a takeoff abort far down the runway, the twin jet went through two fences, across a county road and continued into a field, where it caught fire and burned to the ground, except for the tail section. Miraculously, all 21 passengers and crew onboard escaped the inferno. No one was even seriously injured. (The NTSB at one point suggested that the evacuation itself, which usually involves injuries, should be studied so it could learn something from this accident to save lives in future mishaps.)
The MD-87 was introduced in the mid-80s as the second to last version of a line of narrow body airliners derived from the Douglas DC-9 and eventually built by McDonnell Douglas and, later, by Boeing. Although it was only carrying 21 passengers and crew at the time of the crash, the MD-87 can be outfitted to accommodate as many as 130 passengers and four to five crew members.
The airplane’s weight apparently had nothing to do with the mishap. After the crash, when investigators used an elevated lift to examine the tail section of the jet, they found what amounts to a smoking gun. They wrote, “the airplane’s empennage/tail section was not consumed by the postcrash fire and both the left and right elevators were found in a trailing edge down position.” Investigators then used an elevated lift to examine the tail, and there they found the probable cause of the crash, though they did not use that term. “Investigators found,” the NTSB wrote, “that the airplane’s left and right elevators were jammed in a TED [trailing edge down] position and could not be moved when manipulated by hand. Both inboard actuating cranks for both elevator’s geared tabs were bent outboard, and their respective links were bent (see figures 5) Further, both actuating cranks and links were found locked in an overcenter position beyond their normal range of travel. The elevators, horizontal stabilizer, and control tabs have been retained by the NTSB for further examination.” In other words, the elevator mechanisms were damaged and immovable. The takeaway is that the crew had no choice but to abort. The airplane was not going flying no matter how hard they pulled to rotate.
But how did this happen? Shouldn’t the jammed linkage have revealed itself on the preflight controls check? The surprising answers are, no and no. They did link to a similar accident in which an MD-83 overran the runway at Ypsilanti, Michigan, in 2017 in which 116 people onboard all got off with just a single minor injury among them. In that crash, there was no post-crash fire. In the investigation into that mishap, the NTSB found the same elevator anomaly in that aircraft, which has the same design elevator. It concluded that the elevator, which has no built-in gust lock, was damaged by high winds prior to the flight. “When the airplane is parked, each elevator is free to move independently within the confines of its mechanical stops if acted upon by an external force,” the report’s author wrote, “such as wind or manipulation by maintenance personnel. The elevator system (by design) has no gust lock, and the elevators are not interconnected. Initial examination of the elevator control system by investigators found that most of the system components forward of the engines had been consumed by the postcrash fire; therefore, full flight control continuity could not be established.” MD-80-series airliners don’t have a built in gust lock, and the crew have no way to check whether or not the elevator is working until the plane is traveling at high speed and ready to leave the ground. So, there was no gust lock to remove before flight and the crew apparently didn’t miss any pre-flight checks of the elevators’ functionality.
Investigators will be also be focusing on crew performance. “According to flight crew interviews, during the takeoff roll, the first officer made the 80-knot, V1, and Vr callouts. At the rotate callout, the captain tried to pull back on the control column but indicated that it felt like it was ‘in concrete.’ About the time the first officer made the V2 callout, the captain informed the first officer of the problem. The first officer also attempted to pull back on the control column but was unable to pull it aft. The first officer called abort and retarded the thrust levers, and the captain deployed the thrust reversers. The captain indicated that the autobrake system was in rejected takeoff (RTO) mode, and it applied maximum wheel braking.”
A few things stand out immediately. The first is that the twin jet had takeoff speed. The callout of Vr (that the plane had reached rotation speed) clearly indicates that. The other two speeds, V1 (that the plane had passed the abort speed) and V2 (that it could successfully fly on one engine) show that the plane was accelerating as is normally. Yet, when the captain pulled back on the control column to rotate, nothing happened. Less than nothing. It was, again, he said, as though it were encased “in concrete.”
The V1 decision speed is instructive here. The speed is that at which if there are any anomalies, the crew will still takeoff and deal with the problem when in the air. How is this point determined? It is the point at which the airplane can no longer be stopped on the remaining runway. By the time the crew started to abort, the plane was traveling on the ground at well beyond the V1 decision point, so by definition it couldn’t stop on the remaining runway, and it didn’t. In fact, at that point it was traveling down the runway at 158 mph, according to the NTSB, a speed at which it probably stood no chance of stopping before departing the airport property (for which there is no published V-speed).
This crash, at least the second of an 80-series MD aircraft, is interesting in another way. There is clearly a problem, and it is only by extraordinary luck that no one perished in either the Houston or the Ypsilanti mishaps. The NTSB will likely look to prevent further mishaps in the type by working with manufacturers and maintenance experts to find and recommend some corrective mechanical action or pilot pre-flight inspection technique to spot damaged and potentially jammed elevators before flight and before a similar runway overrun, potentially with great loss of life, happens again.