I used to do functional check flights in our Boeing 747 squadron where one of the required items was to intentionally stall the aircraft to validate stall warning indications, either from the airplane itself or from its warning systems. One of the flight engineers in my squadron trusted with this duty retired from the Air Force and ended up with the same job flying for Airborne Express. He died during this mishap through no fault of his own.
The NTSB blames the pilot flying for inappropriate control inputs during the stall recovery and the company's lack of formal functional check evaluation flight procedures. All of that is certainly true but it misses a few points that can be instructive to any pilot flying a large aircraft:
- If you are tasked with doing a functional check evaluation flight, you need to think about the purpose of every maneuver. The purpose of stalling an aircraft intentionally during an FEF is to prove it provides adequate warning of the stall. You compute the numbers, if they don't pan out you abandon the maneuver.
- The old school thought process of a stall recovery losing minimum altitude misses the point of the exercise: to get the airplane out of danger. To do that you need to break the stall. To do that, you need to lower the angle of attack aggressively.
- If you are in charge of a cadre of "elite" pilots who take on these kinds of tasks, you must instill in them the idea that you must learn from previous mistakes and be willing to adopt techniques to mitigate risks.
- If you are an instructor pilot you must always remember that you are the pilot in command and there may come a time when you must drop the "mister nice guy" facade and take control.
What follows are quotes from the sources listed below, as well as my comments in blue.
- Date: 22 December 1996
- Type: Douglas DC-8
- Operator: Airborne Express
- Registration: N827AX
- Fatalities: 3 of 3 crew, 3 of 3 passengers
- Aircraft Fate: Destroyed
- Phase: En route
- Airports: (Departure) Greensboro/High Point-Piedmont Triad International Airport (GSO/KGSO), NC, USA; (Destination) Greensboro/High Point-Piedmont Triad International Airport (GSO/KGSO), NC, USA
[NTSB AAR-97/05, ¶1.1]
[NTSB AAR-97/05, ¶1.5.4] According to ABX records, the PNF on the accident flight had a total of 1.1 hours of flying experience as PIC (logged the previous day) of a DC-8 post-modification FEF. The PNF had 12.6 hours of experience as a nonflying second-in-command (SIC) on post-modification DC-8 FEFs and had conducted other, less-extensive FEFs in the DC-8 that did not involve a stall series. Between 1991 and 1993, the accident PNF had flown 15 FEFs as PIC in DC-9s, according to ABX records. Some of these DC-9 flights may have involved approach to stall. The accident flight PF had no experience as a pilot on a DC-8 post-modification FEF before the abbreviated December 21 FEF.
The purpose of the maneuver was to validate stall warning speeds, not to demonstrate stall recovery techniques. There is a difference. In a stall recovery demo, the throttles are generally brought way back, the shaker, pusher, nudger, or the wings themselves are used to provide the warning and the recovery is begun. Back in the 90's most operations graded these by measuring altitude loss. Of course this is wrong, you need to break the stall. But back then, the minimum altitude loss recovery was what a check airman wanted to see. But even that is beside the point. The point of this maneuver was to validate that the warning came up at the proper speed. How I've always done this is to pull only enough thrust so that the airspeed rolls back 2 knots per second. Once I got the warning, I pushed the power forward to gently recover.
[NTSB AAR-97/05, ¶1.11.1] FDR Data During Stall Entry and Subsequent Loss of Control
- FDR data indicated that the EPRs on all engines were at or near idle power as the airplane slowed to stall speed for the stall test maneuver and EPRs were increased by about 0.05 about 11 seconds before the pre-stall buffet was called. FDR data showed that at the time of impact, engine EPR levels were at or near idle. The full range of motion of the airplane's control column position (CCP) was about 37 degrees (5 degrees forward of neutral and 32 degrees aft of neutral). During the last 2 minutes of flight, the CCP peaked, at times, at 32 degrees aft, according to FDR data. A Safety Board performance study based on FDR data indicated that the pre-stall buffet began at 149 knots and that the stall occurred at 126 knots.
- FDR data indicated that immediately after the PF commanded and applied power to recover from the stall (at 1808:13), all four engines accelerated, although the No. 2 engine accelerated to a slightly lower EPR than the other engines during the power increase. The engines had stabilized at maximum EPR by 1808:18. At 1808:20, the airspeed was decaying from 130 knots and the CVR recorded sounds similar to engine compressor surges (popping sounds) that continued for 9 seconds. During the period of engine compressor surges (at 1808:25), the FDR recorded EPR reductions. The power was subsequently increased two additional times before the airplane impacted terrain, but the CVR recorded no further sounds linked to engine compressor surges.
- According to FDR data, at 1807:40, just before entering the stall sequence, the airplane's airspeed was diminishing from 180 knots. The airspeed had decreased to 126 knots at 1808:11. Between that time and 1809:20, the indicated airspeed fluctuated rapidly and significantly, consistent with erratic airspeed indications. The airplane impacted terrain at more than 240 knots.
- FDR data indicated that for about 14 seconds after the flight engineer noted the stall indication (at 1808:11) and through the time that the first roll excursion began at 1808:25, the airplane remained pitched nose up between 7 degrees and 14 degrees. CCP values ranged from about 5 degrees aft when the stall was noted to 20 degrees aft at 1808:25. During this period, the airplane's airspeed decayed from 145 knots to about 120 knots (the FDR was providing erratic airspeed indications below 126 knots, and was indicating about 90 knots). FDR data indicated that the airplane's pitch attitude then decreased from about 11 degrees above the horizon at 1808:25 to 12 degrees below the horizon at 1808:35. Thirteen seconds after the stall was called, vertical acceleration forces began to increase and remained (with minor fluctuations) consistently more than 1 G (with a maximum of 1.5 G) until impact.
The control column position and aircraft pitch indicate the pilot never attempted to break the stall, only to fight altitude loss. As a result he got into a deep stall.
- The airplane experienced four roll reversals between 1808:25 and impact at 1809:38. Ten seconds before impact, at 18:09:28, the airplane was 52 degrees nose down, in a 113-degree right roll (the largest roll angle recorded on the FDR). At 18:09:31, the airplane was 67 degrees nose down, the maximum pitch value recorded, and in a 79-degree right roll. At impact, the airplane was in a 26-degree-nose-down, 52-degree-left-wing-down-attitude.
[NTSB AAR-97/05, ¶1.16.1] According to Douglas DC-8-63 performance certification data, an aerodynamic stall is typically preceded by aerodynamic buffet about 15 knots above stall speed. Based on the flightcrew's calculation of 122 knots for the stall speed, the expected buffet airspeed would be 137 knots. Douglas data also indicated that the stall warning (stick shaker) activation point had a tolerance band of plus or minus 5 knots.
I did a fair number of these tests under Air Force rules where we briefed the stall number and the expected buffet speed so as to be able to know when the buffet should happen and when to abandon the maneuver.
[NTSB AAR-97/05, ¶1.17.3] According to the director of flight technical programs, he was the FEF PIC for most of the DC-8s that had earlier undergone major modification. From 1991 through 1994, the DC-8 flight standards manager was PIC for nine of these DC-8 post-modification FEFs, while the director of flight technical programs performed two of them. Beginning in October 1994, the director of flight technical programs instructed the accident PNF in FEF procedures on several flights during which the accident PNF served as SIC and nonflying pilot. The director of flight technical programs also told Safety Board investigators that he had provided the accident PNF with a simulator training session in which stall recoveries and unusual attitude recoveries were practiced. The manager said that the stall recovery procedure he taught the accident PNF was identical to the procedure taught to line pilots for their proficiency checks.
Herein lies a very big problem: the person leading the FEF program did not understand the reason they intentionally stalled the aircraft after modification and instilled in his pilots the stall recovery procedure was no different than what was used in a proficiency check.
[NTSB AAR-97/05, ¶1.18.1]
- FAA records indicate that a DC-8-63F operated by ABX experienced an in-flight loss of control incident during an FEF on May 16, 1991, while recovering from a stall in the landing configuration (landing gear and flaps down). The stall was being conducted as part of the FEF profile and involved engine compressor surges in the No. 1 and No. 2 engines during application of power in the stall recovery. The maneuver was begun at 13,000 feet msl, and the flight crew recovered control of the airplane at 7,000 feet msl. The PIC of the incident flight was the DC-8 flight standards manager at the time, and the PF in the right seat was the DC-8 equipment chief pilot.
- The flight engineer on the 1991 incident flight, now a DC-8 first officer, told Safety Board investigators that the flight crew calculated the stall speed and stick shaker speed, configured the airplane for the landing stall, and slowed to stall speed. He said that the flight crew was surprised when the stick shaker activation and buffet occurred simultaneously. The PF called for maximum power and the flight engineer advanced the throttles. The flight engineer said that he then heard popping noises on the left side of the airplane and noticed that the EPR gauges for the No. 1 and No. 2 engines showed no acceleration. The throttles were retarded to idle and advanced again by the flight engineer. The airplane then yawed to the left, rolled right and began a rapid, spiral descent. The PIC took the controls and reversed all four engines. The crew said they experienced a 1 1⁄2-turn spin.
- After the recovery from the stall/spin incident, the FEF was continued for nearly 3 hours until the FEF checklist was completed. FAA records indicate that the FAA Winston/Salem FSDO processed an enforcement action because of the flightcrew's failure to terminate the flight after the incident, and the pilots' check airmen status was temporarily suspended.
- As a follow-up to the incident, the POI assigned to ABX wrote in a June 7, 1991, letter to the Winston/Salem FSDO that he had met with ABX personnel and they agreed that the following changes would be made to the FEF profile: flightcrews would fly the FEF profile in the simulator before the actual evaluation flight; airborne maneuvers would be executed in an ATC-assigned altitude block depth ranging from 3,000 feet to 5,000 feet (see appendix D). Recovery was to be accomplished with pitch (lowering the nose), and power then slowly advanced to complete the recovery (and fly out of the buffet). "The stall maneuver will no longer use power [first] for recovery," the POI wrote in the June letter. He added in the letter, "In the simulator, lowering the nose three to four degrees results in an instant recovery. Power can then be slowly advanced to complete the recovery. All of this is being incorporated in their acceptance flight profile."
- Pilots reported using the new procedure on subsequent FEF flights. The PIC of the incident flight said that in performing stall series during subsequent evaluation flights he reduced pitch attitude more positively and "let the altitude go." He said that the airplane typically lost about 200 feet to 500 feet using this (modified) stall recovery procedure (compared to little or no altitude loss using the former power-recovery procedure). The flight engineer on the incident flight told Safety Board investigators that he was aware that the FEF stall recovery procedure had been changed following the incident. "The biggest change was to use a pitch recovery technique rather than a power recovery technique."
- ABX's director of flight technical programs, who had been on sick leave and then returned to flight status primarily flying DC-9 airplanes, told Safety Board investigators that he did not use the revised FEF stall recovery procedures when he subsequently flew several DC-8 post-modification FEFs. He added that he believed that if the engines were spooled up properly during the approach to stall, the resulting power recovery maneuver would be "okay."
- The POI said that he recalled having a disagreement with the director of flight technical programs about the stall procedures, arguing that ABX had been "performing the stall series in level flight like it was a precision [FAA-recommended stall recovery] maneuver." He said that he told ABX personnel, "Let's apply some common sense here." The POI said he told ABX managers and pilots that it was not necessary to hold altitude in such FEF scenarios.
[NTSB AAR-97/05, ¶2.2.1]
- For 8 seconds following the initiation of the stall recovery, from 1808:13 through 1808:21, the PF maintained the airplane's pitch attitude at between 10 degrees and 14 degrees ANU. During this 8-second period, the airspeed continued to decrease and the airplane entered a fully developed stall.
- The failure of the airplane to recover before entering the full stall resulted from the control column inputs the PF was making to maintain pitch attitude. The control column was moved aft by the PF, from 5 degrees aft (at 1808:11, just prior to initiating the recovery) to 20 degrees aft (14 seconds later). An increasingly downward flightpath angle coupled with a relatively constant pitch angle resulted in an increasing angle of attack. The increase in angle of attack, which placed the airplane farther into the stalled condition, may not have been perceived by the flightcrew unless they were closely monitoring the airspeed indicator. In addition, the vertical speed indicator and altimeter should have provided evidence of a developing sink rate and stall.
- Thereafter, the airplane began a series of roll reversals, and the airplane remained in an aerodynamic stall condition because the PF held significant back pressure on the control column all the way to impact. Each time the airplane developed a large nose-down pitch rate (combined with reductions in airspeed at 1808:25 and 1809:22), the PF responded with additional back pressure, according to FDR data on control column movement. In contrast, the appropriate pilot response to an uncommanded decrease in pitch attitude (which is, itself, an indication that the airplane is in a stall) would have been forward movement of the control column.
- During the attempted stall recovery, there were several indications of the PF's excessive aft control column inputs that should have suggested to the PNF that, as the PIC, he needed to correct the control inputs and recover from the stall.
- The PNF's role as an instructor pilot during the FEF maneuvers should have clarified the roles and responsibilities of the two pilots. As the instructor, his pilot-in-command authority should have been enhanced, and it should not have been difficult for him to exercise direct control to ensure the safety of the flight. However, the PNF's instructional role, like his command role, was informal on this flight. Therefore, his command and instructional authority may have remained unclear.
[NTSB AAR-97/05, ¶3.2] The National Transportation Safety Board determines that the probable causes of this accident were the inappropriate control inputs applied by the flying pilot during a stall recovery attempt, the failure of the nonflying pilot-in-command to recognize, address, and correct these inappropriate control inputs, and the failure of ABX to establish a formal functional evaluation flight program that included adequate program guidelines, requirements and pilot training for performance of these flights. Contributing to the causes of the accident were the inoperative stick shaker stall warning system and the ABX DC-8 flight training simulator's inadequate fidelity in reproducing the airplane's stall characteristics.
NTSB Aircraft Accident Report, AAR-97/05, Airborne Express N827AX, Douglas DC-8-63, Narrows, Virginia, December 22, 1996