In its final report on the crash of an Israel Aerospace Industries (IAI) 1124A Westwind near Huntsville, Ala. on June 18, 2014, the NTSB was unable to identify a probable cause.
With three airline-transport-rated pilots on board (one a pilot proficiency examiner, PPE), the jet departed from Huntsville International Airport-Carl T. Jones Field (HSV), in Huntsville, Ala., on a check ride flight and was seen to dip to the right, wobble then dive sharply to the right and crash into a field immediately after takeoff. The airplane was destroyed by the impact and subsequent post-crash fire, and all on board were killed. The NTSB found partial deployment of the right thrust reverser, along with throttle positioning consistent with a V1 engine cut exercise the PPE was known to perform during pilot proficiency checks. The airplane belonged to SynFuels Holdings Finance, and was being flown under Part 91.
According to the company’s chief pilot, the purpose of the flight was to conduct pilot-in-command (PIC) proficiency checks for two experienced company contract pilots, as required by FAR 61.58. The chief pilot told the NTSB in interviews that the in-airplane training was more efficient than flight simulator training in terms of time, indicating that the training program he had used in the past took five days, whereas training with the airplane took only two days. The pilots involved in the accident were all highly experienced ATPs, type-rated in the 1124A.
The chief pilot knew the PPE’s check ride routine, reporting that he had been given PIC proficiency checks by the accident PPE in the accident airplane. During those flights the PPE would be seated in the right seat and would give an engine cut after V1 speed was attained, when the flight was in a positive rate of climb at either 10 to 15 feet agl, or 20 to 30 feet agl. The PPE would reach down to raise the landing gear and then would retard the right thrust lever. He also indicated that since owning the airplane, there had been no issues with the thrust reverser system of either engine.
The airplane was equipped with a Fairchild GA100 CVR designed to record 30 minutes of analog audio on a continuous loop tape in a four-channel format. A sound spectrum study of the CVR recording was performed to determine engine performance and thrust reverser deployment during the previous full-stop landing and during the accident takeoff. The study also included plots with sound analysis data combined with the aircraft’s AlliedSignal EGPWS data, which all indicated that the airplane accelerated to about 135 KCAS when a change in radio altimeter depicted a climb and the acoustic analysis indicated the N2 speed of one engine continued to decrease. The air traffic controller who cleared the flight for takeoff later reported seeing the airplane rotate about halfway down the runway. While climbing 18 feet above the runway and at 148 knots calibrated airspeed (KCAS), the pilot commanded gear up. Less than a second later while at 149 KCAS and 33 feet radio altitude in a wings-level attitude, a rattling sound began that continued until the end of the recording. Acoustic analysis of the sound indicated it was consistent with thrust reverser deployment. While in a left bank of 0.7 degrees and 147 KCAS at 53 feet, the PPE asked what happened, followed by a peak in background noise about 1.5 seconds later. At that same time the acoustic analysis indicated one engine was operating at 96 percent N2, and the other engine had decreased to 92 percent N2.
The controller who had cleared the flight to take off stated that when the aircraft was at about 50 feet, he noted, “…the airplane’s right wing [dipped] a little but the pilot seemed to correct [for it].” About 11 seconds after the first EGPWS recorded data point, there was a right roll of 0.7 degrees; the airplane at that time was 83 feet above the runway centerline about 4,650 feet down the runway at 142 KCAS. Shortly thereafter, while at 88 feet flying at 139 KCAS, in a seven-degree right bank with the landing gear extended, with one engine at about 96 percent N2 and the other engine at about 91 percent N2, the pilot asked, “What’d you do? What happened?” The PPE replied, “I don’t know.” The airplane continued to climb while drifting to the right of the runway centerline.
Seconds later one engine was operating at about 96 percent N2 and the other engine was operating at about 84 percent N2 (sound analysis cannot determine which engine), while at 104 feet and a right bank of 8.4 degrees at 137 KCAS. In a strained voice the PPE told the PIC to get the gear up; the EGPWS data indicated the landing gear was retracted one second later. The airplane climbed to a maximum altitude of 108 feet while drifting to the right with one engine indicating 96 percent N2. The right bank angle continued to increase from this point until the end of the EGPWS recording.
Both engines were removed from the crash site, disassembled, bench-tested, CT-scanned and analyzed. The results of the CT scanning of the left thrust reverser actuator revealed the actuator appeared to be in the extended position and the pawls appeared to be extended out over the piston sleeve. The actuator switch moving contacts appeared to be in contact with the lower switch arm. The results of the CT scanning of the right thrust reverser actuator revealed the actuator was in a non-extended position and the pawls were retracted within the piston sleeve. The actuator switch moving contacts appeared to be in contact with the upper switch arm. Also, a crack was noted in the right actuator switch housing. Investigators found no particles or obstructions in either thrust reverser control valve passages, and the moving seals of both thrust reverser control valves were in contact with their respective mating surfaces. The examination of the left thrust reverser actuator revealed indicator lights showing the locking pawls in the locked position.
Operational testing revealed no evidence of pre-impact failure or malfunction. Following bench testing it was disassembled and one end of the retaining ring key was missing. The locking pawls were all in place and had normal wear marks. Examination of the right thrust reverser actuator indicator lights showed the locking pawls were in the unlocked position. Operational testing revealed no evidence of pre-impact failure or malfunction. Following bench testing it was disassembled and the locking pawls were all in place and had normal wear marks.
A typical Westwind has several interlocks and a throttle detent, plus piggyback levers designed to prevent deployment of a thrust reverser in flight. A representative of the airframe manufacturer reported no known cases of inadvertent in-flight thrust reverser deployment on the 1124A. A review of the NTSB database for accidents and incidents involving the 1124 series from 1982 to the present also showed no accidents where thrust reverser deployment was identified as the probable cause, a factor or finding.
IAI provided the NTSB with its Engineering Report No. 4820/10483, titled, “Westwind Model 1124–Dynamic Aircraft Response to Inadvertent Inflight Thrust Reverser Deployment,” dated March 25, 1976, showing calculations and analysis performed for takeoff and cruise configurations to predict the airplane’s flight behavior and controllability in response to thrust reverser deployment in flight. The analysis concluded that the aircraft’s longitudinal and lateral motions were not coupled, and that the lateral motions were affected by thrust reverser deployment only slightly and were readily controllable by the pilot and/or autopilot. Testing was performed at 15,000 feet msl on an aircraft that was modified with a switch to permit deployment of the thrust reverser in flight.
Key to the investigation was the position of the throttle levers in the quadrant found in the wreckage. The left and right thrust levers were three-quarters forward. The left thrust lever was bent to the right, and the piggyback lever was at half travel toward the deploy detent, while the right piggyback lever was in close proximity to the deploy position. A gouge on the right side of the slot of the throttle quadrant associated with the adjacent securing nut correlated with the thrust lever being at two-thirds of its travel. Looking more closely at the right thrust reverser piggyback lever and adjacent portion of the throttle quadrant, the NTSB found a linear gouge on the right side of the quadrant slightly aft of the deploy detent. The gouge was parallel to the resting position of the piggyback lever. The NTSB was unable to come up with a probable cause for the accident.