Challenger Hard Landing: Weather Or Pilot?
After a Challenger crashed on landing at a private airstrip in Texas in 2019, the NTSB conducted a limited investigation and decided the cause of the accident was wind shear. That’s an act of nature, and citing natural acts as causes tends to shift focus away from the actions of pilots. It is difficult then to take any useful lesson from the event, except to somehow avoid misfortune. However, I think there are some lessons to be learned from this accident.
The accident airplane was a Bombardier CL-600 2a12 (Model 601), N813WT. On board were two pilots, one flight attendant and six passengers. The FAR Part 91 flight left Fort Worth’s Meacham International Airport (KFTW) on Jan. 12, 2019, at about 1030 Central Daylight Time (CDT). The destination was Ox Ranch (01TX), a private airport about 25 nm northwest of Uvalde, Texas.
According to its website, Ox Ranch is a native and exotic animal hunting site. Giraffes, zebras, kangaroos, axis deer, bongo, kudu blackbucks, elk, hogs, red stags and whitetail deer range freely across its 18,000 acres, and there are lodgings for visitors. The owner built a 5,800-by-70-ft. runway in 2013 to accommodate guests, and the ranch requires pilots to obtain prior permission to land.
The flight arrived in the vicinity of the airport about 45 min. after takeoff from KFTW. The pilot-in-command (PIC) said in a statement that he flew over the runway to look for obstructions and animals and to see if there was a windsock, but he saw none. He flew a left visual traffic pattern to Runway 35 and set flaps to 45 deg. on final approach. He reported his airspeed as being Vref plus 10 kt. on 1 mi. final, and he said he retarded the throttles to idle when he was at 50 ft. above the runway.
The Challenger developed a high sink rate just before landing, hit hard, and the right main landing gear collapsed. The airplane slid off the right side of the runway, across a ditch and through an airport fence before coming to a stop. The crew and passengers escaped uninjured, but the airplane was damaged beyond economic repair and was sold to a salvage company.
The airplane wreckage was in a field on airport property, about 2,100 ft. east of Runway 35. There were two sets of parallel skid marks in the dirt that began about 1,600 ft. from the runway threshold and curved along the ground, ending at the airplane. There was a 3-in.-wide, 2-in.-deep and 1,500-ft.-long gouge in the runway asphalt. A portion of the airport perimeter fence was found embedded in the vertical fin leading to the vertical stabilizer. The right main landing gear was broken aft and collapsed under the right wing. The trailing edge of the right wing and spar structure at the right main landing gear wheel well were bent and broken upward. The inboard portion of the right inboard flap was broken upward. The nose landing gear was twisted sideways 90 deg. and collapsed into the wheel well. The pilot’s left windshield outer pane was shattered.
The PIC provided a lengthy narrative explanation of what happened on the NTSB Form 6120, “Pilot/Operator Aircraft Accident/Incident Report.” He said, “We encountered wind shear, airspeed dropped rapidly, and the wind shear forced us down to the runway.” He mentioned turbulence or wind gusts six times in his narrative.
The pilot’s statement is valuable to investigators. It is, of course, written from the pilot’s point of view and to his best recollection. It’s always necessary to validate what he says with other, perhaps more objective, facts.
The best way to get an objective measure of flight path and trajectory is from a flight data recorder, but the airplane wasn’t equipped with one. Another way that is becoming more common during investigations is using ADS-B data recorded by a nearby FAA antenna. Apparently, that wasn’t available either. The airplane had a Mark VII enhanced ground proximity warning system (EGPWS) installed, and it was recovered from the wreckage. The investigator shipped that unit to Honeywell’s lab for download.
There were enough EGPWS data points on this flight to establish that when the airplane was 50 ft. over the runway threshold, it showed a calibrated airspeed (CAS) of 123.6 kt., a groundspeed of 114 kt. and a sink rate of 1,257 fpm. One second later, the sink rate was 1,391 fpm, well in excess of the maximum allowed for aircraft certification for landing.
Determining the actual winds at the airfield at the time of the accident is difficult because the airport has no weather reporting facilities. The pilot reported that the winds were 290 variable to 350 deg. at 12 kt. with 10-kt. gusts. The source of his information was not reported, but it could have been from onboard instrument indications.
The NTSB meteorologist’s report showed somewhat more benign conditions. Citing meteorological observations (METARs), it said that nearest to the time of the accident, the wind at Uvalde’s Garner Field, 25 nm from Ox Ranch, was 320 deg. at 11 kt. with no gusts. The METAR at Laughlin AFB, 28 nm southwest of Ox Ranch, was wind 340 deg. at 16 kt. with no gusts about 30 min. before the accident. An unofficial reporting station 21 nm west-northwest of Ox Ranch reported wind 359 deg. at 14 kt., gusting to 24 kt. about 30 min. before the accident.
There is raised terrain just to the west of the airstrip and that could very well have induced low-level mechanical turbulence as a northwesterly wind blew across it. Although there was turbulence, it seems the wind was pretty steady, with occasional 5- to 10-kt. increases and closely aligned with the landing runway. These winds seem to be well within the limits of what a Challenger can handle.
The difference between the EGPWS calibrated airspeed and groundspeed could be accounted for by a 10-kt. direct headwind. It also could be accounted for by a wind of 290 deg. at 20 kt. (a 10-kt. headwind and a 17-kt. left crosswind component). The crosswind could have been fairly demanding to handle by the pilot or of minimal difficulty. It probably was somewhere in between.
There is no information to say if the pilot was cross controlling the airplane in preparation for a crosswind landing. If he was, that would have added drag, which could have slowed the airplane during the last part of the approach.
Swept-wing airplanes typically sink as they get slower before they stall. The airplane could have been getting slow. If in fact the approach speed was 123.6 and Vref plus 10 kt., the Vref speed was 113 kt. That seems low to me for a swept-wing jet with no slats.
The PIC, who was 42 years old, estimated his total flight time was 9,000 hr., with 3,500 hr. in the accident airplane make and model. He had 55 hr. in the last 90 days and 20 hr. in the last 30 days. He reported he had type ratings in the HS-125 Hawker, the G-1159 Gulfstream II and III, and the BBD-700 Global Express. Oddly, he did not report having a type rating in the CL 600. He may have just forgotten to add that, although the rest of the 6120 form he filled out was quite detailed. An insurance adjuster interviewed by the investigator stated that the accident flight was the first time the PIC had flown into Ox Ranch Airport.
The second-in-command (SIC), 31 years of age, reported 1,015 total flight hours, of which 25 were in the accident airplane make and model. All of his CL 600 flight time was in the last 90 days. He had an SIC type rating in an HS-125 and had completed a flight review in that aircraft just two months before the accident. He did not report any training on the CL 600. According to the insurance adjuster, the SIC “was not qualified in the airplane,” and they were conducting “a single-pilot operation on a windy day.”
The adjuster said the flight attendant was employed by American Airlines and “She got tossed around pretty good.”
The airplane was 36 years old. A Honeywell systems engineer said the EGPWS was capable of generating a wind-shear alert, but upon examination of the unit, she noted that feature had been disabled. The airplane was equipped with a cockpit voice recorder (CVR), as is required for this type of airplane. An NTSB specialist found that the CVR had a poor-quality audio recording with significant wow and flutter, indicating a problem with the tape or tape transport mechanism. The audio pertained to an earlier flight. The magnetic tape was broken, there was degraded magnetic material flaked from the tape, and no audio recording had been made of the accident flight.
Ox Ranch Airport is at an elevation of 1,293 ft. ASL. The single runway, 17/35, is asphalt paved. The fact that it is narrower than normal (70 ft.) could lead a pilot to experience the visual illusion of being high. If you think you’re high you may increase your descent rate and retard the power prematurely.
A representative of Ox Ranch said in an interview that the flight was a charter, but the PIC checked a box indicating the flight was a corporate flight. The picture that emerges is that of an elderly airplane, not maintained to Part 135 standards, operated by pilots, one or both of whom were not type qualified on the airplane, conducting a charter flight for which they had no commercial operating authority. It’s clear that regulations designed to minimize risk to charter customers were breached.
That still leaves open the question of whether better pilot skill or judgment could have prevented this wind-affected accident.
This case has distinct similarities to another that took place in Providence, Rhode Island, in 2007. A Bombardier CL-200-2B19 Regional Jet (CRJ-200LR) landed hard in a 9-deg. left bank at a descent rate of 18 fps and collapsed the left main landing gear. The first officer (FO), who was new to the airplane, allowed it to drift left of course during an ILS approach, and the captain took over at about 300 ft. Due to a miscommunication and unbeknownst to the captain, the FO retarded the throttles to idle. The surprised captain attempted to salvage the landing by applying power and pitch up in the flare, but he was too late. The airplane had descended at about 2,000 fpm late in the approach. Airspeed was approximately 132 kt. at touchdown. Due to the flare rotation and sink rate, the airplane exceeded the stall angle of attack, and the stall protection system (stick shaker and pusher) briefly activated.
The CRJ-200 is a direct descendent of the Challenger. It is stretched to carry 50 passengers, but the systems and flight characteristics are very similar. Both the Challenger and the CRJ are known to fly at relatively high approach speeds due to their clean swept wings and lack of leading-edge slats.
The CRJ accident FO, who until recently had been flying conventional training airplanes, did not appreciate the importance of precise airspeed control and maintaining adequate thrust until a safe touchdown is assured.
I think a highly plausible explanation of the Ox Ranch crash was that the PIC allowed the airplane to get slow late in the approach, probably after selecting landing flaps. He did not react immediately when an excessive sink rate began to develop by applying significant thrust to arrest the descent or go around. Instead, he retarded the throttles at 50 ft. in an attempt to salvage the landing. His ability to perceive and arrest the sink rate was made more difficult by strong winds and turbulence.
The lessons to be learned are similar to those learned by the CRJ FO. Maintain the appropriate wind-adjusted approach speed, be alert for developing sink rates, and avoid premature or abrupt thrust reduction, especially in windy conditions. If actual wind shear is forecast or reported, be prepared to go around or conduct a wind-shear escape maneuver.
The determination of probable cause for a weather accident is always debatable. Was it natural conditions or pilot actions? You must weigh how avoidable the weather conditions were and if the pilots took all available actions to prevent the accident.