Foggy Takeoff, Part 1, Cheyenne Loss Of Control
The hazard of taking off into an indefinite or low ceiling is underestimated by many pilots. It seems easy enough. Track the runway centerline, rotate and fly the assigned heading. Yet according to both NTSB and International Air Transport Association data, initial climb is the most common phase of flight where aircraft control is lost.
The reasons for this are many. During initial climb, attitude instruments may malfunction or lose electrical power or vacuum. Engines may fail and icy wings may lose lift. The CG may be too far aft or may shift. The pilot may be fatigued and inattentive. Gusty winds may erase the critical margin of lift needed to sustain flight. The pilot may allow the airspeed to decay and stall the airplane.
None of these things happened when a Piper PA-31T Cheyenne crashed on the morning of Dec. 28, 2019, at Lafayette, Louisiana.
The professionally flown business flight departed from Lafayette Regional Airport/Paul Fournet Field (KLFT) in foggy weather and remained airborne for only 1 min. It took the NTSB two years and five months to consider all the factors that could have caused the accident. Their conclusion cited one of the most common reasons for loss of control.
On the night before the accident, the pilot told his wife that the ceiling would be low when they took off. In the morning, a terminal area forecast (TAF) issued at 0700 indicated the ceiling would be 300 ft. and the visibility 4 mi. when they planned to take off, but at 0853 the actual observation was wind from 120 deg. at 5 kt.; visibility, 0.75 mi.; mist; vertical visibility, 200 ft.; temperature and dew point, 19 C; and altimeter setting, 29.97.
The flight’s destination was Atlanta’s DeKalb-Peachtree Airport (KPDK). On board were the aircraft owner’s daughter-in-law and grandson, two of his company’s vice presidents and a local TV sportscaster. The five passengers were planning to attend the LSU-Oklahoma football game at the Peach Bowl in Atlanta. The airplane was near its maximum allowable takeoff weight and near its aft CG limit.
At 0913, the pilot contacted clearance delivery. His radio calls were transcribed by the ATC facility: “Lafayette clearance Cheyenne four two charlie victors got india like to pick up the i-f-r to p-d-k.” The controller read the clearance and the pilot read back “We’re radar vectors McComb then direct we got two thousand we’ll get two three o ten minutes after twenty eight seven frequency and seventy two fifty four on the squawk for two charlie victor.”
At 0914, Lafayette ground control cleared the flight to taxi to Runway 22L via echo taxiway. At 0918, the pilot said he was ready for takeoff. The tower cleared him: “Cheyenne four two charlie victor Lafayette tower runway two two left turn right heading two four zero cleared for takeoff.” At 0918:35, the pilot said, “Two four o we're cleared to go here we go two charlie victor.”
ADS-B data revealed the flight passed through 150 ft. at 0920:05 and began a slight right turn toward its assigned heading of 240 deg.
At 0920:09, the tower said, “Cheyenne two charlie victor contact departure good day.” At 0920:11, the pilot said, “Two charlie victor goin’ departure,” and at 0920:20, the airplane began rolling past wings level and into a left bank. The Cheyenne was passing 474 ft. and its airspeed was 165 kt. The pilot checked in with departure control at 0920:26: “And departure Cheyenne four two charlie victor seven hundred for two turnin’ two forty.”
ATC cleared the pilot for further climb. “Cheyenne four two charlie victor Lafayette departure radar contact climb and maintain one zero thousand turn right heading three three zero.” The pilot replied, “Uh ten thousand three thirty on the heading two charlie victor.”
The Cheyenne continued to roll left at about 2 deg. per second. The airplane reached its highest point, 925 ft. MSL, at 0920:40. At that time it was in a 35-deg. left bank, and from there it began a left descending turn from which it would not recover. The last bank angle recorded was 75 deg.
The controller issued a low-altitude warning at 0920:58. “Low-altitude alert Cheyenne two charlie victor check your altitude immediately appears you're at three hundred feet.” The last ADS-B point received from the airplane was at 230 ft.
Twenty-three people either saw or heard the airplane as it descended out of the overcast. One witness was directly beneath the airplane as it broke out and described it as being in a hard left bank and “85% upside down.”
Another witness saw the wings level at the last second. The airplane hit trees and powerlines and landed hard in a post office parking lot across the street from his apartment. He said the landing gear was retracted and the airplane was moving at high speed.
The Cheyenne struck a car in the parking lot, causing it to roll several times and catch fire. The airplane continued out into an open field, leaving a trail of wreckage and coming to rest near the far end of the field.
Several people ran toward the burning airplane to attempt to rescue the people on board. One person heard cries for help from trapped passengers. Another, a former U.S. Marine Air Wing veteran, assisted the lone survivor to exit the burning wreckage. He was unable to rescue anyone else.
A Walmart Supercenter located only a few hundred feet northeast of the accident site was evacuated and closed as a precaution. The surrounding neighborhood was without electrical power for about 4 hr. as a result of the damage to the power lines.
The pilot and four passengers died. The woman who occupied the car that was struck was seriously injured, and two people in the nearby post office were injured by flying broken glass.
The NTSB conducted a major (Class 2) investigation. Class 2 investigations have a broad scope and may involve complex systems, multiple organizations or poor risk controls by the operator, manufacturer, maintainer or regulator. They are similar to Class 1 investigations, but the safety issues involved may not extend beyond one airplane type or operation. In the case of this accident, the safety board did not hold a public meeting or issue any recommendations.
Specialists in operations, air traffic control, systems, powerplants, weather, maintenance and electronic devices participated in the investigation, in addition to representatives from the FAA, Piper Aircraft, Hartzell Propeller, Pratt & Whitney Canada, the Transportation Safety Board of Canada and Transport Canada. An uncrewed aircraft system (UAS) aerial imagery specialist provided a detailed survey of the site.
The Piper Cheyenne involved in the accident was manufactured in 1980 and was powered by two P&WC PT6A-28 engines equipped with four-blade Hartzell propellers.
The debris field extended for 789 ft. along a heading of 315 deg. Only the main fuselage and inboard left wing made it to the final wreckage site. Most of both wings, both engines, the instrument panel, forward cabin, elevators and rudder were scattered along the debris path. Impact marks on the right tip tank showed contact with a light pole and on the left engine cowling with a tree. The left tip tank, which still had fuel in it, had struck a transmission line. The throttle quadrant was found near the car that was struck. Flight control continuity could not be established due to fragmentation of the wreckage.
The airplane was not equipped with a flight data or cockpit voice recorder. An NTSB performance specialist was able to obtain ADS-B data from the FAA and use it to calculate aircraft position and performance during the brief flight. Engine sounds recorded by residential security cameras were also used to produce a sound spectrum analysis. No data could be obtained from two iPhones retrieved from the wreckage.
To determine if any mechanical issues may have caused or contributed to the accident, two experienced NTSB engineers and a Piper accident investigator examined the wreckage carefully. Their examination included the flight controls, elevator trim, pitch trim servo, flaps, landing gear, pitot static system, stability augmentation system (SAS), stall margin indicator, angle of attack sensor and autopilot servos. None of these yielded any obvious faults.
The NTSB materials lab tested the light bulbs from the Bendix/King KAP315 auto flight annunciator. Stereo microscope examination showed that none of the autopilot flight progress lights were illuminated. If the autopilot was in use, one or more of these lights would have been illuminated. X-rays of the systems fault annunciator showed that none of them were illuminated either, indicating there were no systems malfunctions.
The Garmin GRS77 attitude heading reference system (AHRS) was recovered and bench tested. Internal circuitry tested normal, but no data from the accident flight could be recovered. A failure within the AHRS could have caused erroneous attitude and heading indications. However, if the AHRS did fail, red “Xs” would have been displayed in the instrument displays and presumably the pilot would have seen them. There was also an independent standby digital attitude indicator just below the primary flight display, so attitude information should have been available.
The airplane was equipped with a stall margin indicator. When it was recovered, the needle was pointing to the red and black stall warning area. However, analysis of the airplane’s performance showed that it remained at speeds above stall throughout the flight. Specifically, the lift coefficient calculated was always above the value at which the wing stalls.
Damage to the engines and propellers showed they were running at the time of impact. Analysis of the engine sounds showed the engines were running at about 2,000 to 2,100 rpm during the descent. Performance calculations and the indicator on one recovered torque gauge showed the power was reduced significantly during the last 5 sec. of flight.
In Part 2, we’ll discuss the pilot’s flying experience and the NTSB’s conclusions.