The NTSB last month opened the evidence docket into the nighttime crash of a Gulfstream IV at Hanscom Field (BED) in Bedford, Mass., on May 31 last year. A flight control issue, possibly the position of the gust lock, is one theory being considered to explain why the GIV went off the end of Runway 11, broke up and caught fire, killing all seven people aboard.
Although the “whys” behind this accident are not yet definitively revealed, the docket has delivered a number of facts that could ultimately explain what happened that night. For example, the experienced IS-BAO Stage 1-certified Gulfstream crew appeared to reject the takeoff too late in the ground roll to halt the aircraft. Main-gear tire skid marks appeared only on the last 1,440 feet of the 7,011-foot runway.
Some behavioral aspects of the crew related to preflight checks were also brought to light in the docket. The GIV’s quick-access data recorder (QAR) showed that the cockpit crew fully tested the aircraft’s flight controls before takeoff on only two of the previous 176 flights. A full control check was defined as stop-to-stop motion of the elevator, ailerons and rudder at some point between the beginning of the FDR power cycle and takeoff. If the gust lock had been engaged, the restrictions to conducting that control check might have raised awareness about a potential problem. Although investigators discovered the gust lock in the wreckage was not engaged, for some reason the flight power shutoff valve (FPSV) handle was pulled up. This would have removed hydraulic pressure from the flight controls. In that case, the Gulfstream system would revert to manual mode (akin to losing the power steering in an automobile).
The accident aircraft, N121JM, was registered to SK Travel and operated by Arizin Ventures under Part 91 regulations. The crew began the takeoff roll at 21:39:05 local time. Fifteen seconds later, they noticed a “rudder limit” warning indicating the rudder had reached one of its left-right travel limits. This prompted the pilot not flying to mention “rudder limit light is on.” The flying pilot asked if the copilot was on the rudders, to which he replied no. The takeoff continued until 21:39:59 (54 seconds elapsed time), when the copilot announced numerous times the “lock is on.” Fourteen seconds later (68 seconds after the takeoff began) the copilot said, “I can’t stop it.” The recording ended seven seconds later–one minute and 15 seconds elapsed. The GIV had reached a speed of 160 knots–well above rotation speed–before the crew made any attempts to slow the aircraft, and it went off the end of the runway at approximately 100 knots.
Systems Testing
Much of the cockpit remained intact, so the NTSB tested the throttle quadrant and the autothrottle systems recovered from the aircraft. Internally developed NTSB objectives were to look at simulated abnormal operations, as well as to validate engine power ratio (EPR) models, characterize aircraft power lever angle and engine EPR response during acceleration and validate the presence of an aerodynamic hold on the gust lock. The Board also reviewed engine data gathering, autothrottle disengage force tests, autothrottle/interlock disengage tests and high-speed taxi tests to characterize the effect of air load on the elevator gust lock release and autothrottle hold characterization.
NTSB system testing began with the facts uncovered during initial inspection of the cockpit. The thrust levers were found at between half and full throttle, and the thrust reverser handles were stowed. The control lock was discovered in the full-forward–gust locks not engaged–position. The speed brakes were also stowed.
Docket reports explained that Gulfstream’s autothrottle system would be expected to disengage if the crew attempted to take off with the gust lock in place. “When contacting the throttle interlock and when the autothrottle hold mode became active at 60 knots, there was no change in thrust or throttle position,” essentially meaning that pushing harder on the throttles would not translate into more thrust. In some tests, though, the NTSB learned that when attempting to remove the gust lock at the last minute if the aircraft was accelerating–a procedure not supported by Gulfstream–the lock could sometimes be disengaged and other times not. Investigators noted, “It [would have been] very difficult for the copilot to pull the gust lock handle back and unlatch it.”
Gulfstream reported that when hydraulic pressure is removed from the ailerons, both the left and right aileron surfaces can “float” in response to aerodynamic loads. The FDR showed that at approximately 21:40:05, one second after the call to “rotate” was made, the right and left ground spoilers and the right and left inboard flight spoilers began to “float” up to between 1 degree and 3 degrees, as both the left and right aileron surfaces moved in an upward direction. The FDR data further showed that at approximately 21:40:06 the yaw damper disengaged. The flight power shutoff valve can affect the operation of each of these systems.
Another report confirmed that the gust lock system interconnection with the throttle levers was intended to prevent advancing the throttle levers beyond a minimum power setting with the lock engaged. The pin that holds the gust lock in place was found sheared on N121JM, although it is unclear if this happened before takeoff or during the accident. Additional testing concluded that considerable force would be required to disengage the elevator hook at high speed on the ground.
Investigators found a pair of broken sunglasses inside the airplane’s throttle pedestal that impeded the gust lock sheave movement and halted the gust lock handle’s travel before reaching the full down position. Further testing revealed there were multiple ways the sunglasses could have partially restricted the movement of the gust lock sheave.
Finally, the docket notes an item in the GIV Line-Up checklist reminding crews, “At sixty (60) knots, the pilot shall confirm the elevators are free and the yoke has reached the neutral position. If the Flight Power Shutoff Handle (valve) is pulled at rotation due to a flight control problem, high pull forces will be required to achieve the takeoff attitude. There will be a delay in airplane rotation and, once airborne, a push force will be necessary to maintain the climb attitude. Application of forward trim will be required shortly after becoming airborne. To avoid running out of forward trim, reduce speed as necessary.”