This blog is the second attempt to share lessons and insights gained from studying a Cirrus fatal accident, Cirrus fatal #65 in SR20 N527MJ near Chapel Hill (was Sr 20 Down Chapel Hill).
(My first blog explored the loss of engine power at Buttonville Airport. See Reflections on Cirrus fatal #61 near Buttonville Airport in Canada)
My intent is not to second-guess the accident investigators, so I don't attempt to figure out what happened. I wasn't there. I don't have enough information.
I point you to the NTSB determination of probable cause:
Witnesses observed the airplane overfly the runway and enter a left traffic pattern. The airplane appeared to be faster than they were accustomed to seeing small airplanes operating. The airplane touched down hard on the runway surface and bounced several times before departing off the left side into the grass. The engine was described by the witnesses as operating at full power and the airplane appeared to by flying about 60 to 70 mph. The nose of the airplane was observed in a 45-degree nose up attitude and then leveled out back onto the ground. The airplane traveled 840 feet until the left wing collided with a tree and the airplane spun to the left and collided with the airport perimeter fence. The left and right flaps were in the retracted position. The Pilot's Operating Handbook for the airplane stated that the flaps are required to be extended 50 percent for a balked landing/go-around. No anomalies were noted during the examination of the airframe, flight controls, engine assembly, and accessories.
The National Transportation Safety Board determines the probable cause(s) of this ACC as follows:
The pilot's improper recovery from a bounced landing and subsequent improper go-around procedure, which resulted in a loss of directional control, runway excursion, and collision with a tree.
More information is available from the NTSB docket of public information, from which much of the detail here was gleaned. The docket contains several photographs, reports, witness statements, transcripts, selected for public disclosure.
What I do is ask what would I do if I faced similar situations that are described in the accident reports? How would I handle those challenges?
One aspect of this flight was that the pilot had never flown to Chapel Hill, NC, before. The Horace Williams Airport (KIGX) is a typical small general aviation airport with a single 4005' runway oriented east/west 09/27.
Looking at the satellite view, I notice that the runway is surrounded by trees closer than usual for half of the runway
From my experience, that likely affects the winds, causing crosswinds to shift direction aligned with the runway somewhere during my final approach. I also notice that the ramp is at the 27 end of the runway and there is no parallel taxiway, which probably induces arrivals to favor runway 09.
When approaching an unfamiliar airport, how do you choose which runway to land?
When the winds are closely aligned with the runway, the answer is easy. When the airport has a control tower, you will be told the active runway. When the airport has a UNICOM frequency, there is likely to be someone who can advise you with the active runway. When the airport has a CTAF frequency and there are planes in the pattern, then you can listen and figure out the active runway.
In the factual report of this Chapel Hill accident, the reported winds were 170 degrees at 4 knots, almost a direct crosswind. Apparently, there was little traffic.
Otherwise, no traffic and no advice, the choice is up to us. How would I choose? I would suspect the winds would shift, so if I choose, I could be wrong. But 4000' runways with a 4 knot tailwind is manageable, so even if I was wrong, I would prefer not to back taxi on runway 27, so I would probably choose runway 09, anticipate a tailwind, and manage my landings speeds with an expectation to go-around if things don't go well.
In this case, witnesses describe the plane flying downwind, making a “U” turn to final, landing on runway 09 with a tailwind according to the windsock next to the runway.
COPA Pilots have demonstrated a lot of problems with landing too fast. COPA has tracked some of this to misinformation from instructors fearful of tail strikes, some comes from the difference in sight picture, some comes from fear of slow flight. Whatever. Excessive speed during landings is killing us and damaging airplanes.
For a discussion of landing speeds, see this blog: A Photo Story of Landing a Cirrus
As a refresher, the speeds published in the Flight Operations Manual are 100 KIAS downwind, 90 KIAS base, 80 KIAS short final, 73-77 KIAS over the threshold (not the numbers), near stall at touchdown.
One tactic I use is to check my ground speed during final approach. Are the winds behaving as I expect them? Has the wind shifted? Normally, my ground speed would be lower than my airspeed as I land into the wind with a shorter rollout and less braking. If the ground speed is about the same, then the winds are calm and I should be able to land with book numbers. If the ground speed is higher, then I have a tailwind and can expect a longer float and rollout with perhaps greater braking with an increased possibility of needing to go around.
Witnesses to this accident report that the aircraft came in faster than usual and bounced.
Kyle Henn, the rear seat passenger, reported “The airplane touched down on the runway a made a " hop" back into the air, touched back down on the runway, made another "hop" went back into the air, touched back on the runway and made another "hop" before coming back down on the runway at about midfield.”
Chad Lewis, an Army Captain waiting for Kyle Henn and first on the scene, reported “On it's final approach, the plane was initially not in line with the runway, but after two minor corrections, one to the left, then one back to the right, it was lined up. It seemed to be coming in faster than usual for a small plane and according to the windsock beside the runway, the plane was landing with the wind, not against it.
Landing with a tailwind at an unfamiliar airport is challenging. Landing too fast adds energy to the situation. Don't know how this COPA Pilot practiced landings nor what his typical landing speeds were, but we know that the plane did not execute a smooth landing.
Bounce once, go around!
Bounce more than once, stay on the ground!
Landing hard happens occasionally. Sometimes from excess speed, sometimes from wind shear, sometimes from my ham-handed controls. If the plane bounces back up in the air, often I'm no longer flying straight but turned slightly. And if I am landing fast and touching down above stall speed, then the wings are still generating lift. Both of those situations prompt the advice to Go Around!
However, if you bounce more than once, you probably have had a prop strike and maybe you won't have the full power you need. The fatal accident in Caldwell, NJ, involved four bounces leaving a portion of a propeller blade on the runway resulting in an unsuccessful climb and ultimate crash. Hard decision, since prop strikes are apparently often not noticed in the cockpit.
In this accident, the plane bounced more than once including touching down on the nose wheel. Chad Lewis reported “The initial landing was pretty hard and there was a small bounce where all three wheels left the ground. When the plane came down the second time, the front wheel hit first and there was a bigger bounce. The plane bounced a third and fourth time, each time the front wheel hitting first and each bounce getting more pronounced.”
That describes pilot-induced oscillations. The pilot attempts to keep the plane on the ground for landing and causes bigger and bigger bounces. In a Cirrus, bounce once, go around!
Once you decide to go around, the POH procedure is clear:
1. Autopilot .....................................................................DISENGAGE2. Power Lever ........................................................FULL FORWARD3. Flaps ........................................................................................50%4. Airspeed....................................................................... 75-80 KIASAfter clear of obstacles:5. Flaps .........................................................................................UP
You need power … Now! But note the flaps change. Landing procedure is to use full flaps before short final approach. Now, in the go around, you select 50% flaps. Only after clear of obstacles should you raise the flaps.
By the way, given the COPA encouragement of CAPS during departure climb emergencies, I now consider 500' as my CAPS-viable altitude for clear of obstacles! That's when I raise flaps.
In an SR20, raising flaps too much or too early badly affects the climb performance. We have had two SR20 accidents implicating flaps. A fatal accident at Greenwood, SC, where the pilot took off over gross with no flaps, and failed to outclimb trees, stalled and crashed. Another non-fatal accident at Fish Haven, ID, where the pilot executed a go around but inadvertently selected 0% flaps causing the plane to sink and crash off the end of the runway.
In this crash at Chapel Hill, the investigators found the flaps in the wreckage fully retracted, in the UP position.
Tough combination: landing with excessive speed, bouncing several times, retracting flaps too early.
Furthermore, one witness located at a maintenance hangar reported “He heard a noise similar to a tail scrape, and heard an engine at full power. He looked towards runway 9 and observed Cirrus in a 45-degree nose up attitude in the vicinity of the last one third of the runway. He then observed the nose to level out back on the ground and the airplane was veering left off the left side of the runway in the grass.”
Such an extreme attitude is very unusual. Could be an exaggeration by the witness, could be a stress reaction by the pilot who wants the plane to climb by pulling back on the yoke, could be an aerodynamic reaction to a departure stall.
Ever notice that Cirrus runway excursions all seem to be on the left side of the runway? This one did. Here is the NTSB diagram of the ground scars found and the path of the aircraft.
Why left? I don't know for sure, but my suspicion is that COPA Pilots fail to apply sufficient right rudder to compensate for the left-turning tendency when applying full power, especially during a go around procedure. More right rudder is needed to keep the plane going straight!
One interesting aspect of this accident was the presence of an experienced pilot in the right seat accompanying the COPA Pilot. That pilot had zero time in a Cirrus but 12,500 hours in other planes.
When you fly with another pilot, how do you establish who does what and when? Recall deciding on the positive exchange of controls with an instructor? Do you do that with other pilots? Probably not. The instructor is ensuring that they can save themselves by taking control. In less formal settings, do we do the same thing?
For accident investigators, having two pilots in the front seats in an aircraft with dual controls, it can be impossible to determine who was flying. Have you been induced by other pilots to let them handle the controls? Even a landing?
Then when something bad happens during the flight, how does the pilot not flying deal with the temptation to intervene? How does the pilot in command deal with a surprise like that? Even with prior discussion and agreement, the adrenaline rush of an emergency combined with greater experience might override the commitment to let the pilot in command handle everything.
I encourage anyone in the plane to speak up if they sense a problem, but especially a pilot flying with me. However, I do establish my prominence as pilot in command, it's my airplane, so speak up but don't touch anything without my direction.
This accident at Chapel Hill is special because the widow of the pilot joined the COPA community and shared her life experiences after the accident. Deb Markwood posts regularly in the blog Looking Back, Looking Forward - A Widow's Perspective. She and I have communicated regularly about the accident investigation as she wants to understand more about what happened to her husband Tom. COPA has invited Deb to the Allentown CPPP in June.
My appreciation for and understanding of accident investigations comes from long conversations with two key people at Cirrus Aircraft, Brad Miller and Brannon Mayer. Their professional dedication to helping us understand what happened has helped me explain things clearly. Their discretion is appreciated and honored.
Finally, a lot of my understanding of better ways to fly my Cirrus come from interacting with the amazing CPPP instructors who routinely vie to teach at the COPA Pilot Proficiency Program. Their commitment to our safety is appreciated.
The Cirrus Owners & Pilots Association (aka, “COPA”) is a 501(c)7 non-profit corporation dedicated to serving its members.
COPA® is not in any way affiliated with Cirrus Aircraft, the manufacturer.Cirrus is a registered trademark of Cirrus Aircraft