The pilot was performing a visual approach to a mountaintop airport in Burnsville, North Carolina, with an estimated density altitude of 9,860 feet, and a tailwind component.
He stated he “carried some extra speed” during the approach to compensate for wind and turbulence that he described as “burpy.”
As he started to flare the Cirrus SR22, he got pushed hard to his left. He lowered the nose and increased engine power to abort the landing, but the airplane struck trees and a small building before coming to rest inside the wood line on the left side of the runway.
The accident caused substantial damage to both wings and the fuselage, as well as minor injuries for the pilot.
Airport surveillance video depicted the airplane on approach, and as its attitude flattened in the landing flare, its nose and left wing deflected downward, and the airplane departed the left side of the runway where it struck trees and a building.
Probable Cause: The pilot’s failure to compensate for the high density altitude and the tailwind component while landing at a mountaintop airport, which resulted in an aerodynamic stall and a loss of aircraft control.
To download the final report. Click here. This will trigger a PDF download to your device.
This November 2021 accident report is provided by the National Transportation Safety Board. Published as an educational tool, it is intended to help pilots learn from the misfortunes of others.
I commented earlier but now that I’ve seen the video of the accident, there is no 50′ obstacle at the end and it appears the pilot flew an excellent approach. What I see is at the point he begins to round out, not flare, the left wing drops down and the plane rolls left. It all appears to happen smoothing not abruptly. The terrain certainly suggest this airport could be challenging on short final but what I don’t see is the pilot’s attempt to immediately correct. The plane rolls left. It might have been pushed somewhat to the left but the plane rolls and turns, and the pilot doesn’t roll the wing back instantly. With extra speed, the plane should have responded instantly. On the other hand, maybe he came in slower because it was a short runway and he began to stall the airplane, which would account for the roll left. I landed a float plane in gusty conditions in a small lake. The landing would be near a small island. I anticipated some wind issues near the island so I landed prior to reaching the sland. I touched down on both floats but was closer to the island than I wanted to be and sure enough I got caught in a sudden gust that lifted my right wing up and pulled both floats out of the water. I immediately corrected by holding the nose in the step attitude, adding power, and adding an aggressive crosswind correction. The plane responded instantly and we touched back down on the water under control. I am only guessing what might have happened to the Cirrus pilot but it looked recoverable from the back seat.
If you are a highly trained air carrier pilot, either 121 or 135, you will be adding 20 to 40% to the book numbers in order to operate in/out of an airport. Yet we teach student pilots that the minimum is okay for them. We (the industry including the FAA) teach the most vulnerable of pilots, the student, to rotate and climb at Vx and clear a 50’ obstacle in the minimum distance. In doing so we set up accidents such as this Cirrus disaster. This Cirrus pilot learned this bad behavior as a student pilot.
Vx (or less) climbs have a coffin corner between 20 and 200 feet. For example if the engine quits at 50’ the aircraft is dangerously close to the stall speed. Secondly the. Pushover back to the runway sets up an unrecoverable sink rate.
We should teach student pilots to only do Vy climbs and teach them to add a minimum of 40% to the book numbers for both takeoff and landing. Short Field Performance should be an endorsement, not a maneuver that we teach 20 hour students. Also this endorsement needs to be trained by someone qualified to teach it.
Did you know that the “book” numbers in most aircraft are extrapolated from two or three data points? In other words your book numbers haven’t really been test flown (other than the data points), they are a best guess.
I think manufacturers and instructors are obligated to teach the technique that will maximize the distance from obstacles on takeoffs so the pilots will be able to maintain the widest margin possible whenever conditions dictate. All airplanes other than Part 121 are vulnerable to some extent on takeoffs. Even in a Vy climb, the airplane is rotated at a lower speed and requires some time to accelerate to Vy in the initial climb, so you can still have an emergency right at Vx. Learning how to handle the pushover is something that should be done before solo – if done correctly it’s a non-event – incorrectly a potential disaster. Adding distance to the POH numbers is a great idea – for ex: AOPA recommends adding 50% to book takeoff and landing distances. But that wasn’t an issue in this accident – the SR22 landing distance (there was no obstacle) is 1300-1400ft (no wind) – runway 2900ft.
Well….actually it was THE factor in this accident. Had this Cirrus pilot been trained to add 40% to book numbers he would have needed about 5000 feet to land with a 10 knot tailwind.
As I mentioned to David Dow below, from what I see on the video (included in the accident report), it was windshear on short final just as he was going to flare. Up to that point, it looked like a perfect approach. See if you can look at the approach area on Google Maps 3D – the terrain is going to cause any winds to conflict in the threshold, plus there are trees fairly close to the runway which as you know can potential cause a great deal of turbulence. As David mentioned, the airport has had other accidents.
I am a flatland ( Ont. Canada )800 hr pilot with most time in a 200 Mooney. Learned to fly a C150 in the shortest narrow x wind strip in the world. My first trip in the Mooney was from Vancouver to Calgary with a RCAF instructor pilot. It was a learning experience. New plane, double the speed of the 150s,and 9000ft mountains. From Calgary to Ont no problem, flat land. Next time West and I had many I took a 3 hr mountain flying course. The instructor was From Montreal once again flat low land so he himself could certainly relate the difference. Was $ very very well spent.
This is a challenging airport. 28 crashes in the NTSB database with multiple fatalities. Pilot needs the appropriate training, needs to recognize and mitigate all potential risk factors, needs to bring his/her “A game” when landing at this airport. KAVL is about 1-hr away (by car). Easy drive. Much safer option, in many situations.
David Dow – thanks for that information. Got on Google Earth and checked the runway 32 approach topography in 3D – most unusual. Approximately first half of the threshold has steeply rising terrain from the left – second half of the threshold has fairly steeply descending terrain from the left. The report used the winds at KAVL which were a 90 degree left crosswind. Imagine the wild windshear’s a pilot could find from the close-in terrain at opposing angles. From the video it appeared the airplane was exactly where it should be and he was making a beautifully stabilized approach. But suddenly at about the time the flare was started, he rolls and turns left what looked like about 30 degrees. Looks to me like a windshear from the unusual terrain with no chance to correct (runway 50′ wide).
So why is he landing downwind?
I can’t find any comments by the pilot that mention a tailwind. The final report includes winds at the nearest airport (KAVL) at 230 degrees. That is a tailwind for KAVL, but only a 90 degree crosswind at the accident airport. So I don’t think there was a tailwind for the pilot.
did a 2000-hr pilot not realize he was landing downwind? carrying some extra speed & with a tailwind, & 2900-ft to work with, pilot would have gone into whatever was at the end of the runway… (i’m guessing the building’s condition was not a result of the crash…)
I am just starting lessons. I constantly see accident reports of higher hour pilots that a trainee would not do. Like landing downwind or not adjusting for the increased required landing speed in thin air at high altitude airports, combined with the added decrease in control. Did he check weather for turbulance or study the airport?
Per the video, when the airplane started the flare, it rolled left and appears to have changed ground track approximately 30 degrees to the left, and the report says it collided with that building. The video shows an instantaneous fire.
Let’s see, HD altitude, high OAT, downwind and hot on approach. What could possibly go wrong ? Maybe the Cirrus should stay where it was intended, a long paved runway.
Good thing he kept the speed up, huh?
It’s appropriate to add speed for turbulence. Check the video in the report – it looked like a perfect approach – then it appears he hit windshear as the flare was started. The airport has very, very unusual terrain (which can be seen on Google Maps 3D) which appears to bring in winds from the west at different vertical angles into the threshold (upsloping winds changing to downsloping).
Most of all. Bad judgement on behalf of the pilot.
A SR22 needs 2300’ to land over a 50’ obstacle. Add high density alt and a tailwind, a 2,900’ runway 50’ wide is an accident waiting to happen, and it did. Pilot had 2,000 hrs but obviously he didn’t think this through very well. Glad he wasn’t seriously hurt.
Note that the 2300 ft distance is based on coming over the threshold at 50 ft agl, on the 3 degree glide slope if you are using the VASI/PAPI, which puts the aiming point at 1,000 ft down the runway…probably not the best way to fly into this 2,900 ft runway….landing with little runway remaining.
Flying into this short runway, it would be better to come over the approach end at about 15-20 ft, so that the touch down is about 400 ft past the threshold, giving 2,000+ ft to stop.
Kelly Carnighan. From what I see on the video (included in the accident report), it was a loss of control at the flare after a perfect approach. There is no obstacle and doesn’t have anything to do with running long and out of runway. If you can check out the approach area on Google Maps 3D at a birdseye view, you can see a steep upslope on the west side which changes to a steep downslope, all within the threshold. I can’t imagine what sort of wild wind rotations that could cause in the threshold area. As David Dow mentioned there are 28 crashes in the NTSB database.
I think that the density altitude was closer to 6,800 ft, with an airport elevation of 4,400 ft and a max temp that day of 80 degF.
It was still a high DA, so the aircraft landing speed would have been higher than the pilot
saw on his asi.
This pilot had 2,000+ hrs in the aircraft, but no mention of any mountain flying.
Also the VFR chart warns of severe turbulence over the mountain range just south of the airport.
Oh, and an interesting comment in the NTSB final report;
” Pilots with limited or no training in mountain flying can be surprised about their aircraft’s different performance at high density altitude, often leading to serious or fatal accidents. Wind and other weather phenomena interacting with mountainous terrain often lead unsuspecting pilots into situations that are beyond their capabilities. “