Reducing Fatal Accidents: No Low-Hanging Fruit

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

At last month's Society of Aviation and Flight Educators symposium in Atlanta, we spent a lot of time talking about the overall GA fatal rate of accidents, but not much time was spent on the specifics of the accidents themselves. So in this blog, I'm examining one year's worth of fatal accidents. The included chart shows the results.

You can do the same exercise from the NTSB's data. Pick a year—any year—and it will take about a half day to skim through and sort the results. This turns out not to be so easy. To make much sense of it, you have to sort into broad categories that don't necessarily correspond to the same categories others might use. For example, traditionally, CFITs refer to aircraft that are flown into terrain or water while under control and usually related to some sort of IFR operation. I used a broader definition. If the airplane was flown into the ground under control in any kind of operation, regardless of weather, I called that a CFIT. But I also have a category called "low flying," where the buzz jobs and stupid pilot tricks go. Some of these could just as easily be called CFITs, or loss of control and some might even be stalls. A further caution: The NTSB's reports sometimes lack detail, are vague and , I suspect, are flat out wrong at times.

For instance, our sister magazine, KITPLANES, found that the NTSB's analysis of amateur-built safety trends was significantly in error because it miscoded many airplanes as experimental but which were actually ultralights or certified aircraft. These errors accounted for a considerable swing in the actual accident rate. So, caveat emptor. For my sweep of the 2008 data, I included only accidents involving certified aircraft in the U.S. I did not include ultralights, helicopters or amateur-built airplanes. The total was 218 accidents which yielded at least one fatality.

click for a larger version

Even with this lack of granularity in the data, it is possible to gain a broad glimpse of accident causes and results. I don't have the space here to cover everything, so I'll hit only the highlights. There's nothing much new about any of this, by the way. Stall-related crashes are obviously a big player and regardless of methodology, every analysis I've seen seems to confirm this. Every year, about 20 percent of all fatal accidents are stall related.

I see two ways to look at this. One is that about 50 pilots every year kill themselves because of surprise stalls. The other is that many thousands more don't. On a per flight basis, then, stall incidence is low, suggesting that there's probably not much fundamentally wrong with stall training doctrine. Could a change in that training or awareness push the number from 50 to 40? Or 30? Maybe, but I think it will be difficult to yield measurable results. I thought the best recommendation SAFE came up with was also the most specific: Add angle-of-attack and load factor awareness to stall training doctrine. Still, a long shot.

Loss of control was also a leading cause. There's probably some overlap here with stalls and other causes, again because of lack of detail. For our routine sweeps of accidents, we see a lot of runway loss of control or R-LOCs. These are rarely fatal. The fatal loss-of-control accidents seem to involve just plummeting into the ground or some object for no apparent reason. In many of these, there's no clear pattern of bad judgment or any pattern at all.

Speaking of bad judgment, there's a lot of it and some of these accidents just show stunningly poor risk awareness. I would say about a third of the fatals involve such fundamentally poor decision making that you just don't know where to begin. SAFE's Bob Wright thinks these people are write offs that no amount of outreach out will save. I agree. They are nothing but accidents looking for a grid reference. We can't help them.

You can decide for yourself if this accident was one of those. But in reading the details, it made me wonder if someone --or several someones—in the local community should have stepped up and questioned the advisability of this particular operation. That smacks of big brotherism, but is a little of that worth it to avoid killing people who don't sense the risk they're taking? You can't ask the people in the airplane, because they're dead.

So, bottom line, my guess is that in a few more than 100 fatal accidents, the pilots just drew a bad card which they might—emphasize might—have avoided with better skills, proficiency and risk awareness. This is going to be a very small nail to hit, regardless of the size of the hammer. I think it's worth the effort, even if it doesn't actually reduce the fatal accident rate. Better risk awareness just makes better pilots. And that could help some avoid accidents entirely, much less a fatal one.

Comments (100)

I've arrived at very much the same conclusion. If one gets an instrument rating, avoids vfr into imc, avoids buzzing, is proficient enough in the airplane to avoid stalls, and avoids tstorms and ice, you have avoided 95% of the accidents. The "bold pilots" we all know have likely sealed their fates, I think they know deep down they are really pushing their limits, and it is probably very limited what we can do about it.

Posted by: Josh Johnson | June 2, 2011 5:38 AM    Report this comment

I'm a CFI. I told a friend of mine he shouldn't fly any more. On a subsequent flight he ran his Bonanza through a fence. The person in the accident likely wouldn't have listened either. Non-aviation people have no basis on which to make a judgment whether to fly with someone. There's no solution to this problem. It's the same with drivers.

Posted by: DENNIS WOLF | June 2, 2011 6:44 AM    Report this comment

Very interesting article, Paul. Thank you for writing it.

My own interest these days is in the home built community (that accounts for around 20 percent of the GA fleet) and also in the related part of GA where owner/operators fly planes vs. pilots who fly planes owned by others such as FBOs or commercial operators.

When a pilot flies a plane owned by someone else he faces a determination by the owner that he is properly prepared and fit to fly. When the plane is owned by the pilot there is nobody to "Second guess" the pilot's judgment. In the case of home builders the problem is even greater. Home builders often build planes their pilot skills just don't measure up to. Since they own the plane there is nobody who can tell them they can't fly it.

Even in the case where the FAA or other government entity takes administrative action (like pulling the pilot's medical or pilot certificate) there just isn't any way to prevent the owner from taking his plane out for a flight.

I would really like to see an analysis of accidents involving owners flying their own planes. I suspect the accident rate in this class of operations is considerably higher than the kind where there is a separate owner.

Sadly, even if this scenario is at work (like it was in your example accident CHI08FA156) I don't see there is any acceptable way to avoid this problem.

Posted by: PAUL MULWITZ | June 2, 2011 7:23 AM    Report this comment

"No Low-Hanging Fruit "

I agree completely. Aerodynamics are well known, we have better instruments, planes, training, and weather forecasts than ever before. The only thing that has not advanced in the last 100 years are people.

GA means non-professional and that means ordinary "people" of all types. You have to accept that people crash in planes just like they crash in cars. NHTSA gave up on driver training and accepted crashes to the point of generating a "crash safety rating".

The car industry is way ahead in its thinking. It accepts that crashes will happen and happen frequently. In fact, car makers real concern is being sued when their cars don't protect idiots from poor driving!

It's time that GA give up on the idea that it's immune from people with bad judgment. You can only do so much and then it's up to the NTSB to pick up the pieces.

Posted by: Mark Fraser | June 2, 2011 8:24 AM    Report this comment

IIRC, the referenced accident pilot was the owner of the FBO at that airport (and may have owned the airport). I doubt anyone at the airport would say anything too strongly to him.

GA will never be perfect. Neither will airline flight (ref: AF447). The best we can do (for all concerned) is to be aware of the issues.

As the local FAASafety coordinator says at the meetings: "the ones here aren't the pilots I worry about". The pilots we need to worry about aren't reading articles like this. Unfortunately.

Posted by: leave blank | June 2, 2011 8:31 AM    Report this comment

After 100 years of increasing regulations, additional solutions required to "fix" the problem are incompatible with the freedom we get by flying. If we really want to reduce the accident rate to zero, then maybe GA should be done away with. We're damnned either way it seems.

Posted by: A Richie | June 2, 2011 8:37 AM    Report this comment


What a great article! I think you are on the mark with this one...if the 86 yr old pilot in the link you provided was MY pilot, you can bet the farm that I'd be in the right seat and my hands very close to those controls, especially if I knew about the car accident.

My grandmother is having serious issues slowing down and accepting that she can't do everything that she could when she was 20 (she's 90 now). I tried to help her in the kitchen last week and that was a huge mistake...

The "low hanging fruit" according to six-sigma methodology should be the easiest to correct. So as a CFI, show a student how to buzz something...a little instruction goes a long way toward being safe.

And if you're too old to fly...hang it up. You have nothing left to prove to anyone!

Posted by: R. Doe | June 2, 2011 8:39 AM    Report this comment

Paul made a very important point: just because someone does not have a Medical or Ticket does not prevent them from flying. Just look at the "Bare-foot bandit". The same applies to driving -- there are many who lose their license (for various reasons -- DUI, medical, child support arrearages) but continue to drive.

Short of pre-emptively throwing those people in jail, there is no way to prevent them from flying/driving. Note that I only suggest jail after a conviction, not pre-crime.

Posted by: leave blank | June 2, 2011 8:43 AM    Report this comment

3) THIS IS A FIXABLE AREA. The other scenarios as described by Paul B. are writeoffs attributable to Darwin’s theories and natural selection. We can’t help that the guys that are predestined to kill themselves. Theyre all over the world, theyre just using aluminum construction in these cases.
4) FAA Stall training always using pitch and power to recover. In LOTOT there is no power available.
5) Spin training always has altitude available. LOTOT does not.
6) If you are a pilot and would like to see the FB post or add a comment on the subject I will accept your friend request. I'm probably not the most experienced or decorated CFI out there, but just thinking maybe together we can put our resources together and make some changes in this obvious (IMHO)gap. I welcome your views and experience.
Thank you.
Dan Gryder

Posted by: Dan Gryder | June 2, 2011 9:23 AM    Report this comment

Dan, with respect, students know that "emergencies" are staged and that they are not in real danger. They go through the motions and they pass their reviews.

When an emergency is REAL then it's a completely different mind set. Some people panic, some focus on the problem and not on flying, some handle it just fine. You can't predict how people will react when it's not a drill and they are the PIC and actually "going in". It's darn hard to actually PUSH and aim at the ground when it's no longer just a "procedure" but real...

Posted by: Mark Fraser | June 2, 2011 10:28 AM    Report this comment

Like so many topics I wonder at the basic assumptions. In this case I wonder at the accidents per hours flown: No-one has ever asked me how many hours I flew in a give year, so it makes me suspicious. If it's based on fuel consumption how do they account for the autogas burners and gliders? And there's the hand wringing going on at Bob Miller Inc. over aviation kills per year that amount to a fraction of one day of auto kills. I know Bob Miller has a following, and he maintains the public hates GA because we occasionally make a mess of our planes, which scares off potential pilots. I just don't see the hate or that crashes have anything to do with private pilot starts unless they were personally involved in one. From the interested people I talk to it's the cost and FAA hoops that have little to do with flying a plane or safety that turns them off.

I've lost a number of friends flying single air tanker fire support in the Dromeder. I've never flown it, but I'm told it has poor control feel and easy to inadvertently stall. I wonder how many other GA aircraft are that way, and if an angle of attack indicator with aural as well as visual annunciators have any merit? If carrier pilots find them useful some of us flivver flyers might as well.

Posted by: THOMAS M CONNOR | June 2, 2011 10:38 AM    Report this comment

I'm assuming the hours flown comes from medical or insurance records, probably the former. They do ask you how many hours you have flown since your last flight physical. There's a similar question on my insurance renewal form, but I don't know if non-owners always have insurance. Perhaps you need to supply proof of insurance to rent; I haven't rented for many years and can't recall that. Regardless, the hours-flown by certificated pilots information is definitely available.

Posted by: DAVID CHULJIAN | June 2, 2011 10:41 AM    Report this comment

Paul, thanks for mentioning my presentation at the SAFE Pilot Training Reform Symposium. As an attendee, you can expect us to issue our preliminary report VERY soon.

I will slightly modify your description of what I said, since it is only the deliberate risk takers that I would say we can't do much about. Regarding those with poor judgement, it is not judgement we are trying to teach but rather the science and practical application of risk management, including risk identification, assessment and mitigation. We need to do this from day one. My own analysis of accidents for clients shows that the root cause of 60-75 per cent of fatals is the result of poor risk management, not lack of physical skills. As you suggested, the NTSB data is deeply flawed and the true story can only be found by performing root cause analysis. The hot news to follow is what Cessna, King Schools, and Redbird are doing to revamp the Cessna Pilot Centers training program, to introduce risk management on day one of training.

Bob Wright, Wright Aviation Solutions (and SAFE Symposium organizer)

Posted by: Robert Wright | June 2, 2011 10:54 AM    Report this comment

Given the continued persistent prominence of stall related accidents, why is there no push on aircraft manufacturers to install angle of attack indicators in the planes they sell? Why isn't there a greater effort to encourage aircraft owners and flight schools to install these in the aircraft they operate? I seem to remember an Aviation Consumer article from years ago that favorably reviewed the inexpensive versions of AOA indicators. Maybe this is about to change, as I notice both Sport Aviation and Aviation Safety have articles this month on these simple devices. But why is it taking so long? Why does Cirrus go to all the trouble and certification costs that it obviously has taken on to reduce risk and yet not install these on their aircraft?

Posted by: Robert Davison | June 2, 2011 11:14 AM    Report this comment

LOTOT: Blessed with an old SAC runway we can do multiple options on each pass and gently introduce turning a Vx climb into a power loss situation. As Dan says, the angle change is rather dramatic. The first few times we make it a partial power loss so the PF can catch up, and proceed from there based on comfort levels. Do it with a crosswind to keep it interesting, and be sure tower knows your intentions so when they get calls from 'concerned bystanders' they have an answer.

Rudder control is another. It's interesting how many fly with feet flat on the floor. If I see it I add min-controllable airspeed to a flight review, including dutch rolls to bring the point home. Interestingly, some just don't get it.

While not a high cause of fatals, many have no clue how to make a crosswind takeoff or landing. With our huge runway I make a game of it like the King school video on the topic, working on maintaining centerline, then move left and right with the controls crossed to build skill.

Posted by: THOMAS M CONNOR | June 2, 2011 11:20 AM    Report this comment

Finally, I see little interest in airspeed control on final approach to land or picking an aim point. Anywhere will do. I really like to make a bunch of power off approaches, landing on a different runway if it presents the best option, using slips to control speed and aim point. It helps break the mindset of landing on what you planned on before the engine quit. Again, let tower know you have such a plan, or ask for 'any runway, multiple options'. We've also landed on taxiways or at least made low approaches to them, because it was the best option for the situation.

Posted by: THOMAS M CONNOR | June 2, 2011 11:20 AM    Report this comment

"You can't fix stupid." - Ron White, Comedian

You can teach someone all the right things, and once they get their pilot certificate, they're free to make any decision they want while operating an aircraft. I agree that adding risk assessment/mitigation/elimination in scenario based training is definitely a must, I also know that no matter how much you pound on that nail, sometimes, they go crooked.

I've seen pilots jump into airplanes without a pre-flight. They ignore 14 CFR 91.103. I'm sure someone trained them properly while they were coming up the line. It's the complacency that develops from doing the same thing over and over that leads people into a dangerous comfort zone of thinking, "this is the same flight I do everyday". To that, I say, wrong. Today is different than yesterday and you better treat this "routine flight" with the utmost professionalism and respect. But people don't, and Darwin comes into play.

"To err is human..." - Alexander Pope

Posted by: William Wang | June 2, 2011 11:27 AM    Report this comment

Paul, your 'low hanging fruit' analogy is excellent and your factual analysis of 2008's accidents supports it. When I first read about SAFE's goals I was flabbergasted. To suggest that people are not flying because they think it's dangerous is not only preposterous, but self-serving for "Wright Inc" and some are falling for it.A little training on risk management, CRM,FAA allowing instalation of angle-of-attack indicators as OPTIONAL, and application of common sense will help but the people Wright is trying to save are likely beyond saving. I have been flying for 40 years and have seen lots of folks that shouldn't be flying but I'm not part of the aviation police. We have all we can tolerate of THAT with the FAA. In the end, usafe pilots exist, aviation is unforgiving of mistakes and at some point, the two will meet. End of story. There is a point of diminishing returns on investment.Mr. Wright has parochial interests which ALSO need to be factored into this discussion and any subsequent outcome/change in training standards. People reading this blog are likely the "old" pilots not the "bold" pilots.

Posted by: Larry Stencel | June 2, 2011 11:39 AM    Report this comment

I remember some high performance sailplanes had especially dramatic stall/spin behaviors, e.g. roll upside down. Much later I remember the first time I stalled a 172. It was very stable and I could only tell we were stalled only be looking at the rapid sink rate on the instruments.

I was wondering how 3 very experienced pilots could keep an Airbus in deep stall from 38000 feet. They had lost much of their instrumentation. If an airbus has a ‘well behaved’ stall perhaps it was indistinguishable from being in a severe, sustained ‘downdraft’ that might occur.

Although unlikely to be a LOTOT situation, it might be an unconscious “raise the nose to stretch the glide” situation. Is it hard to recognize a stall in a ‘modern’ plane when the behavior no longer appears abnormal?

Posted by: Ray St-Laurent | June 2, 2011 11:42 AM    Report this comment

"Regardless, the hours-flown by certificated pilots information is definitely available."

The hours reported on the medical are 'as logged or estimated.' I've been told to put in a number, but don't sweat it because it has nothing to do with the medical exam and neither the AME or FAA flight surgeon care. If you look at the accident report Paul provided in his narrative, the accident pilot did exactly that and the report listed numbers that jumped from 30k to 55k to 50k and such. If the feds cannot classify a plane as experimental or spam-can I have serious doubts that they take the hours reported on the medical seriously.

The airman app is the same way; At CFI renewal clinics they say to leave it blank because there is no hourly requirement for CFI renewal.

I've had people say they got questionaires from the feds asking about hours flown, but in 25 years I've never gotten one, and they know where I live, hence my skepticism of anything even remotely reliable.

Institutions such as CAP and flight schools track aircraft use to the tenth, so they might be a source of some data, but I suspect that projecting it on the worldwide fleet of N-numbered airframes and licenses is at best a wag.

Posted by: THOMAS M CONNOR | June 2, 2011 11:49 AM    Report this comment

>>>> why is there no push on aircraft manufacturers to install angle of attack indicators

Posted by: PAUL MULWITZ | June 2, 2011 11:56 AM    Report this comment

Two comments: One, I agree that you can't do anything with pilots who are determined to kill themselves, other than hope they don't take anybody else with them. I know one young pilot (20, maybe?) who may have perfectly good skills, but I will probably never know that since I would be extremely reluctant to fly with him unless he developes some risk assessment ability. He doesn't seem to have any now.
Two, I have trouble taking NTSB analyses of accident causes at face value since an incident involving a friend of mine. He over-ran the runway attempting to abort a takeoff. The problem was a mechanical one with the plane (one that particular model was known for). The NTSB tried to say it was pilot error; that he hadn't used the whole runway. I was one of two pilots who witnessed the incident, and I knew that wasn't true, but they never talked to me. As far as I know, they got all their information second hand from the emergency responders, who weren't there at the time. As far as I know, they also didn't speak to the A&P who checked the plane out and confirmed the mechanical problem. They only backed off from the "pilot error" theory after my friend bombarded them with the correct information.

Posted by: John Worsley | June 2, 2011 12:18 PM    Report this comment

In defense of the Colgan pilot that crashed near Buffalo, the pilot was reported to have received training in tailplane stalls because the plane involved is known to have such problems. It was not mentioned in the NTSB summary tho it may have been in the detailed narrative. We'll never know what the pilot was thinking, but if he thought the tail had iced up he performed the school solution. Failing to power up or recognize that it wasn't working is another matter however.

Posted by: THOMAS M CONNOR | June 2, 2011 12:41 PM    Report this comment

As I mentioned previously, there is a general lack of basic skills in the pilots I encounter. The school of acronyms offer little help in that regard. I've taught CRM, risk assessment and blah blah blah and it boils down to a lot of common sense. Some of the risk assessment strategies have become a list of checklists that employ acronyms of acronyms that in the end add little to the flight. The plus of such systems is that it replaces the stultifying thought process of 'don't go' if there is a chance of encountering a cloud or weather or the plane or pilot are not up to snuff to 'analyze it to death and still don't go.' I much prefer the naval way of risk management: In a single pilot cockpit, if you commit and recognize two errors be careful. Recognize three errors and call it a day. In a TRAINED multi-crew cockpit more are allowed, but discuss it. From what little we know of the air France A330 that stalled from 38kft I think that was part of the problem: they weren't talking. As an aside and in support of my lack of basics theory is this: The Air France A330 had a backup iron gyro AI in the center stack yet three pilots apparently ignored it. Assuming that the plane was in the 500klb category, once it passed thru 20kft with a vertical speed in the neighborhood of 120 kts they all knew they were dead and did the best they could to keep the debris field as small as possible. We'll learn more if they release the CVR transcripts.

Posted by: THOMAS M CONNOR | June 2, 2011 12:41 PM    Report this comment

Stall/spin is at the top of the list, and AOA indicators are still not mandatory?

So some pilots will kill themselves regardless. But why not find out if mandatory AOA on the panel will decrease the fatality rate?

Posted by: Patrick Underwood | June 2, 2011 1:01 PM    Report this comment


Why do you think an AOA will change things when the planes already have a stall warning device that provides the same information in simpler form?

Posted by: PAUL MULWITZ | June 2, 2011 1:07 PM    Report this comment

Given the continued persistent prominence of stall related accidents, why is there no push on aircraft manufacturers to install angle of attack indicators in the planes they sell? Why isn't there a greater effort to encourage aircraft owners and flight schools to install these in the aircraft they operate? I seem to remember an Aviation Consumer article from years ago that favorably reviewed the inexpensive versions of AOA indicators. Maybe this is about to change, as I notice both Sport Aviation and Aviation Safety have articles this month on these simple devices. But why is it taking so long? Why does Cirrus go to all the trouble and certification costs that it obviously has taken on to reduce risk and yet not install these on their aircraft?

Posted by: Robert Davison | June 2, 2011 1:16 PM    Report this comment

One reason, from Aviation Safety, is that when we want to establish an angle of attack providing the greatest lift for a given airspeed or configuration, the stall warning is next to useless. It only alerts us when it might be too late.

Posted by: Robert Davison | June 2, 2011 1:18 PM    Report this comment

Besides Aviation Safety, the other recent article is in this month's AOPA Pilot (not Sport Aviation as I wrote in my first post, which for some reason was re-posted). The AOPA Pilot article offers an other reason for AOA indicators --"But the stall warning system is a simple on/off switch that doesn’t show trends. In cases where the angle of attack is rapidly increasing and/or airspeed is quickly diminishing, there can be little time from the blare to the aerodynamic stall."

Posted by: Robert Davison | June 2, 2011 1:37 PM    Report this comment

Never forget that GA is full of amateur pilots. By definition that means "pilot error" accidents are expected. The reality is that 20 hours of dual time (spent any way you hack it) is still just a license to learn. At that point it's up to the individual and individuals act like individuals.

That's WHY all the GA safety symposium in the world never make a dent in accident rates (though the food at such events is pretty good).

Posted by: Mark Fraser | June 2, 2011 1:46 PM    Report this comment

No LHF?!?!? What?!?! LOTOT is THE LHF of all the LHF's.
The FAA put the scenario in for Glider, and AMEL, but it is not in ASEL PTS! It is not taught, practiced, Oral'd or flight checked or BFR'd by any regulation. And yet it is the number one GA fatality for ASEL. If you got an instructor that teaches LOTOT and has you ready, keep him. It's not if. Its when. Are you ready? There it is. Push.

Posted by: Dan Gryder | June 2, 2011 2:50 PM    Report this comment

"SAFE's Bob Wright thinks these people are write offs that no amount of outreach out will save. I agree. They are nothing but accidents looking for a grid reference. We can't help them."

Never forget -- half the pilots in the country are below average.

Posted by: Gary Dikkers | June 2, 2011 4:14 PM    Report this comment

"Given the continued persistent prominence of stall related accidents, why is there no push on aircraft manufacturers to install angle of attack indicators in the planes they sell? Why isn't there a greater effort to encourage aircraft owners and flight schools to install these in the aircraft they operate?"


Yes, yes, yes. Why isn't there a push to put AOA indicators in GA airplanes? That single step could do more to prevent stall/spin accidents in the traffic pattern than a million flying safety symposiums.

Posted by: Gary Dikkers | June 2, 2011 4:18 PM    Report this comment

1. Does a stall warning horn really provide all the information an AOA indicator does?
2. I don't know if it will help. It would be nice to find out.

As airspeeds and stall concepts are being beaten into my soul during my training, I also don't quite understand how so many people let stalls develop to the point of not being able to recover. But if a little instrument cuts down the number of fatalities (and again, I don't know if it will) I'd be all for it.

Posted by: Patrick Underwood | June 2, 2011 4:30 PM    Report this comment

Again, IF stall/spin accidents are the biggest cause of fatalities, AND there is a little instrument that might help pilots avoid those types of accidents, WHY is there not a serious push to find out once and for all?

And why does the US Navy believe AOA indicators work? Is the Navy just a bunch of nervous Nellies or something?

Posted by: Patrick Underwood | June 2, 2011 4:42 PM    Report this comment

AOA vs stall warning? AOA wins hands down! It gives a constant, easily readable graphic indication of how the aircraft is flying. Properly installed and understood, it gives continuous info regarding approach speed and departure speed.
In a LOTOT incident, the needle will spring for the roof, and the pilot just pushes util it is back in the right place. By the time a stall warning goes off, the aircraft is decelerating so rapidly a stall is almost assured.
BTW, my Airbus may have alpha vanes all over the nose, but we don't get a direct read-out. It gets analysed by the computer and displayed as the spread between Vmin and Vmax. I wish we had the direct gauge as well.

Posted by: Peter Buckley | June 2, 2011 5:34 PM    Report this comment


My understanding is that the stall warning provides the same information as an AOA when used in a light plane except it only gives one measurement point while the AOA gives a continuous measure of the nearness to a stall.

An AOA indicator would give more information that allows the pilot to fly closer to the edge in a number of different flight conditions. However, when it comes to stall warning they both give roughly the same warning. That is: "You are nearing a stall."

Getting more information isn't necessarily better. It might be that the real problem that leads to pilots stalling their planes on landing approaches is they are overloaded and can't keep up with all the information they need. Adding more information is a step in the wrong direction in this case.

Many planes (LSA and experimental) don't have any stall warning or AOA, and yet they don't have a lot more stall accidents than certified planes. It turns out that for light planes the airspeed indicator is plenty of information to keep the plane from stalling while still being able to land on available runways. This could be a lot different for a plane weighing 500,000 pounds trying to land on a runway that for it is marginal in length. In that case it is important to make the final approach at the lowest safe speed possible.

Posted by: PAUL MULWITZ | June 2, 2011 5:58 PM    Report this comment


Interesting analysis. I believe that technology could help with stalls. But, after that, it does look like there may not be much to be done!

Lost Control is unpreventable because we don't know the cause. The only solution we know of for CFIT, Spatial D, Mechanical, Low Fly, and Breakup is the same: preaching. That has probably achieved what it's going to achieve.

With Fuel, few aircraft with a fuel totalizer seem to run dry, but other than that we're back to... preaching.

There are cheap technological solutions to Midair - look up FLARM - but ADS-B has been blessed. Until then we're back to... preaching.

On the "technology could help" point about stalls, an AOA gauge won't help if the pilot isn't looking. What we need is an audio AOA warning. Like sailplane varios, it would have a variable audio output that starts to chirp/whine/hum/talk at an AOA corresponding to Vy at 1g, becoming more urgent at the AOA for Vx, then up to a squeal at the stall. Why Vy? It's the highest AOA used in normal operations (a short-field takeoff or landing is a special operation) and it leaves some time for response.

A simple alternative is to design the airplane so that it doesn't have enough elevator authority to stall - like the Ercoupe. By all accounts this worked fine! A 4-seat version would be difficult, because of the required CG range.

Posted by: Thomas Boyle | June 2, 2011 6:02 PM    Report this comment

"And why does the US Navy believe AOA indicators work? Is the Navy just a bunch of nervous Nellies or something?"

Because the Navy has to do something no other segment of aviation does: stuff airplanes onto a short deck and arrest them with cables. To keep all that stuff from breaking--airplane, arrestment gear--the touchdown energy has to be in a pretty narrow band. Best way to do that is an AOA indicator and airspeed. Plus, Navy doctrine has always been big on angle of attack and especially pitch controlling airspeed.

I think AOAs are a terrific idea. We've reviewed them. But I'm not sure they would make the slightest dent in the stall/spin record. I think it has more to do with what's between the ears than what's on the panel.

Posted by: Paul Bertorelli | June 2, 2011 6:50 PM    Report this comment

>>> A simple alternative is to design the airplane so that it doesn't have enough elevator authority to stall - like the Ercoupe.

Posted by: PAUL MULWITZ | June 2, 2011 7:12 PM    Report this comment

"I find it odd that in the list of accident causes, weather doesn't rate a mention."

This is a flaw in my taxonomy and maybe thinking. I didn't include six "weather" accidents in the chart of lack of space. But the VFR-into-IMC imply weather, although a few were blackhole accidents. Some of the CFITs also involved weather.

But I tend to believe there are few accidents really caused by weather. These days, pilots don't have weather sneak up on them without forewarning, if they ever did. Those are bad judgment accidents in which weather played a role.

Others view it differently.

Posted by: Paul Bertorelli | June 2, 2011 7:38 PM    Report this comment

I respectfully disagree with any contributors suggesting we can't improve the statistics. We can, and we will. There would be no airplanes if it weren't for people who believed in beating the odds.
Some simple ideas:
More and better simulators. Full motion sims are cheaper and safer to do emergency training in.
More emphasis on good aviating skills. The rudder is your friend!
LOTOT stalls - 150% of Vs have a simple stick/yoke pusher system. Or a really loud horn.
All flights of any kind should require simple registration of movements using your pilot cert#. FAA database knows exactly who is in the air at all times. IF VFR pilots are within 5 miles of IFR conditions, they are warned. Again at 3. Again at 2 and again at 1. The data and the systems exist to automate tracking and communications to help pilots fly safe.
weather is huge and needs to be improved. Why can't i rely on more systems to keep me abreast of changing conditions AUTOMATICALLY?
Same with terrain. Same with other obstructions. Same with airspace. We have to use the technology and the data already in place to give us better information, in a more timely fashion, and know that my privilege to fly has obligations with serious ramifications if i don't comply.
We can keep improving. Each of us is implicated in the statistics - we have contributed in our own way,we've all made mistakes and we must take ownership of improving them further for the good of GA.

Posted by: Joe Goebel | June 2, 2011 7:48 PM    Report this comment

I think "judgement" is a tricky word. A 5000 hour pilot with a few close calls a little wiser for it sees judgement very differently than a 250 hour pilot who is trying to be prudent but may not have the experiences under his/her belt to always make the best call.

Posted by: Joe Goebel | June 2, 2011 7:52 PM    Report this comment

"Those are bad judgment accidents in which weather played a role."

I agree with that - my point really was that stall, CFIT etc. are often just the end-game is a sequence that started well before the impact - often the crucial errors are made before the plane even takes off. As such, concentrating on stall avoidance/recognition/recovery etc. does little to fix the real problems. I like the saying "good pilots use their superior judgement to avoid having to use their superior skills". About the only time that aircraft handling skills are crucial is on landing - and landing accidents break aircraft more often than people. To reduce the fatal accident rate we have to improve judgement - better flying skills or equipment will have little effect.

Posted by: Clyde Stubbs | June 2, 2011 7:54 PM    Report this comment

"Because the Navy has to do something no other segment of aviation does: stuff airplanes onto a short deck and arrest them with cables."


Yes, the Navy has AOA to do that, but I flew fighters in the Air Force, and we also fly with AOA indicators. AOA is of tremendous help landing on a carrier, but it is also helpful in other ways. Use it right and you automatically know how hard to pull for best turn rate and/or smallest turn radius -- something that can be vital in air combat maneuvers, doing a circling approach under a low ceiling, or during the final turn in the traffic pattern. (Had Corey Lidle had an AOA indicator in his Cirrus SR20, laid the airplane on its side to 75 deg of bank, and pulled to best turn AOA, he probably would have made that turn over the East River. Had he pulled the nose up and done a tight wingover or whifferdill he probably would also have made it, but that's another topic for another time.)

Lose an engine, just hold the AOA that gives the best glide angle, or slowest rate of descent -- your choice. Same on a climb. Hold the AOA for best climb rate or steepest climb angle. Works no matter your fuel weight or power setting.

I've heard the main reason AOA has never caught on for GA is that they have been expensive, but prices have come down lately. Now there is no reason for not using AOA except ignorance, or being stubborn and too set in your ways.

Posted by: Gary Dikkers | June 2, 2011 9:59 PM    Report this comment

After being a CFI since 1965 and either a check pilot, FAA ops Inspector or DPE since 1967 it is obvious that we have not made any real improvment in teaching and or testing. Good design prevents more pilots & CFI's from breaking airplanes on take-off or landing than skill. Most pilots believe rudders are are just a nice foot rest. A high percentage of pilots Think that stalls have more to do with pitch than with elevator positions.

No need for new regs or PTS's, CFI's and Examiners simply must be sure our students can actually fly to the existing standards. Most Cannnot.

The numbers like + or - 10 degrees or whatever in the PTS's are not near as important as task descriptions like these from the Pvt. ASEL:

. selects a suitable touchdown point.

Maintains a stabilized approach and recommended airspeed . . .

Makes smooth, timely, and correct control application during the roundout & touchdown.

Touches down smoothly at approximate stalling speed.

Touches down at or within 400 ft. beyond a specified point, with no drift, and with the airplane's longitudinal axis aligned with and over the runway center.

Maintains crosswind correction and directional control throughout the approach and landing sequence.

If we could just get pilots & CFI's to do this one thing we would have a drastic effect on the overall accident rate. There are many other tasks where the task description requires the applicant to perform to a fairly high level of proficiency.

Posted by: THOMAS INGLIMA | June 2, 2011 11:55 PM    Report this comment

Paul, I'm curious why you excluded ultralights and EAB. There is an active discussion among ultralight pilots and former ultralight instructors about accidents caused by people transitioning from heavier GA aircraft to very light aircraft. With the introduction of light sport, the FAA has created a situation were it is very difficult for people to receive proper transition training.

Posted by: DANA NICKERSON | June 3, 2011 8:04 AM    Report this comment

Because our audience isn't ultralights and because the investigations appear even less complete than the certified airplanes. Also, I know squat about that segment, especially flight characteristics.

I don't think I came across any LSA fatals, but I'm not sure.

Posted by: Paul Bertorelli | June 3, 2011 8:12 AM    Report this comment

Joe said: "I respectfully disagree with any contributors suggesting we can't improve the statistics. We can, and we will. There would be no airplanes if it weren't for people who believed in beating the odds"

Airplanes would never have been created without people WILLING to risk fatal accidents on every flight. People today seem to forget that fatal accidents were "the norm" in aviation. Today pilots have great training, great airplanes, superior weather information, outstanding instrumentation.

All those improvements have made private aviation SO SAFE in fact that fatal accidents are now seen as a problem instead of the norm. When you reach that point of safety then there is no "Low-Hanging Fruit" because you've already reached the level where safe flight operation is now the norm.

Posted by: Mark Fraser | June 3, 2011 8:27 AM    Report this comment


Part of the problem with your comment "I don't think I came across any LSA fatals" is that LSA doesn't have just one definition. I know there were LSA fatal accidents in 2008 because there were some in-flight structure failures of Zodiac XLs that year.

So, what exactly is an LSA?

One definition is in FAR 1.1 which defines any aircraft that meets the specifications as Light-Sport. These are the planes it is legal for Sport Pilots to fly (as well as higher rated pilots reducing their privileges to Sport Pilots to get out from under the FAA medical certificate oppression). It is this definition that is most useful when talking about handling characteristics of LSA as well as pilot transition issues. Airplanes that meet this definition might have one of several different kinds of airworthiness certificates. A partial list includes E-AB, standard type certificates, S-LSA, and E-LSA.

Another definition of LSA is planes that have airworthiness certificates that include "LSA" in their name. These might be S-LSA which are factory built airplanes that (are claimed to) meet the ASTM standards for LSA. They could also be E-LSA which might be one of a couple of completely different aircraft types. One type of E-LSA is "Fat Ultralights". Another is "Kit" planes that exactly meet the design of an S-LSA from the same supplier. Still another is an aircraft that was S-LSA and converted to E-LSA by its owner.

Posted by: PAUL MULWITZ | June 3, 2011 8:32 AM    Report this comment

I'm aware of the LSA definitions and know enough of the models to tell the difference. I tossed the experimentals and intended to keep the LSAs in, I just can't remember specifically what they were.

Could be a couple of ultralights were really LSAs, but I couldn't tell for sure.

Posted by: Paul Bertorelli | June 3, 2011 8:59 AM    Report this comment

Paul Mulwitz,

The Ercoupe, as I'm sure you're aware, was not a canard. It actually could be stalled (you had to zoom it a bit) but would recover immediately. The key insight was that the deadly stalls are typically entered inadvertently, and a zoom stall tends to be a deliberate maneuver. Whether the aircraft will recover from a stall depends on the wing and tail and the pitching moments involved; whether it can be stalled inadvertently from level flight depends on elevator authority. There is a problem with the highly-stall-resistant design, however, and it is that a low-authority elevator will only work within a narrow CG range. The Ercoupe's side-by-side seating gave it the required narrow loading range.

Generally speaking a canard actually will recover from a full stall, in which both the canard and main wing are fully stalled - just as a conventional layout will recover even if both the main wing and tail are fully stalled. In both cases, the forward surface is more heavily loaded and will sink faster, leading to a recovery.

Posted by: Thomas Boyle | June 3, 2011 9:49 AM    Report this comment


Studies like yours have a lot of value no matter what choices you make in selecting accidents to consider. Whether you include LSAs or Exerimentals or not you will get much the same results. Most of the accidents are pilot errors that wouldn't happen if the pilots were doing their basic job - flying the plane. There are a couple of special cases that arise from special aircraft that could make a small difference in your results but that is all they would do.

Experimentals have a somewhat higher occurrence of fuel starvation than other planes. This is not a cause of accidents since pilots should be able to convert an engine failure into a reasonably safe emergency landing. If they fail to make a good landing then the accident is blamed on that rather than the fuel problem. Also, like most landing accidents these tend to be non-fatal.

LSA include a number of aircraft types that are not particularly interesting to main-stream airplane fliers. While fatalities in powered parachutes and gyrocopters are tragic the lessons for airplane pilots from these accidents are few and far between.

In the end the complications of different aircraft types don't really change the basic story that blatant pilot error is the primary cause for most accidents. Your study brings that out just as well as studies with other selection criteria.

Posted by: PAUL MULWITZ | June 3, 2011 10:55 AM    Report this comment

Thomas, you might recall that the 2-seat conventional American(Grumman)also eliminated stalls/spins. The FAA installed a placard. It worked.

Point being is that the word "don't" should be enough to completely eliminate such problems. Prevention is infinitely better than recovery training. Once you're in a stall/spin then you've already broken multiple warnings and exceeded you piloting skills.

If you can't follow the word "don't" then Aviation may not be for you...

Posted by: Mark Fraser | June 3, 2011 11:10 AM    Report this comment

Mark, I'm not familiar with the placard in the Grumman. But experience shows that many highly experienced aviators have stalled aircraft through inattention. No matter how much we like to tell ourselves otherwise, we are human: our attention wanders at times. "Don't" isn't enough.

Posted by: Thomas Boyle | June 3, 2011 11:13 AM    Report this comment

Thomas, The placard is "Spins Prohibited". It worked.

I'm human. In 40 years of flying I've never been so inattentive in the left seat that a stall ever "happened". Seriously, if your mind so far off from flying the aircraft them land immediately. Excuses bend aircraft.

Posted by: Mark Fraser | June 3, 2011 12:11 PM    Report this comment

I've been noticing alot of comments recently regarding AOA indicators, particularly as they relate to the AF 447 accident. The thing that puzzles me, however, is why does anyone think that a little AOA indicator is going to make a difference when an entire flight crew ignores a large glass display that shows the nose 15 degrees above the horizon???? AOA indicators may be useful items, but it is a whole lot cheaper to adjust training standards to emphasis the need to PUSH to maintain the lifeblood of airspeed than it is to require that every GA airplane be outfitted with AOA indicators. Just my .02.

Posted by: ANDREW HOCHHAUS | June 3, 2011 12:48 PM    Report this comment

Nobody understands why the AF447 crew did what they did. (See the ATP guys arguing at It immediately reminded me of a DC-8 crash during a post-maint. test flight (Airborne Express?) where, in a 3-person cockpit in a planned stall series, the crew held the stick back all the way from 15k. Utterly incomprehensible to someone who is relentlessly drilled on "get the nose down and power in."

Gross stupidity, or just being overwhelmed by events in a one-in-a-million situation... neither of these obviates the idea that AOA indicators could help (and I note that so far an Airbus pilot and an AF fighter pilot agree here).

Posted by: Patrick Underwood | June 3, 2011 1:16 PM    Report this comment


40 years of flying have taught you to be overconfident, I'm afraid.

Yes, most of us will never be that inattentive - but largely because we will never be sufficiently distracted. But it only takes a very low probability of doing it, multiplied by a very large number of flights, to wind up with a lot of deaths. Highly experienced, skilled pilots have done it. There's video online of Mark Newman (UK aerobatic glider champion) doing it - yes, he was low and it was turbulent, and he was distracted, and that's the point. It's easy to avoid stalling, if you're thinking about not doing it. It's harder, though, when you're thinking about something else. And, yes, Newman was surely thinking about flying the aircraft (he was hardly thinking about his tax return!). (Newman survived. You'll find the crash on youtube.) I've seen video of airshow pilots doing it in everything from jet fighters to Tiger Moths, in situations from high-g turns to a simple, no-fancy-stuff takeoff (okay, the engine quit, but we're talking professional airshow pilot, and he got distracted). If they can do it, most of us can.

Posted by: Thomas Boyle | June 3, 2011 2:20 PM    Report this comment

"we're talking professional airshow pilot, and he got distracted). If they can do it, most of us can."

Not me.
I'm not a show-off like an airshow pilot.
I take flying dead seriously.
That's what keeps people from being overconfident and/or dead. I've had no issue at all with 2 dead engine events at low altitude. Pay attention and it's easy (duh!).

Posted by: Mark Fraser | June 3, 2011 9:15 PM    Report this comment

I found the Airborne Express story in "Aircraft Accident Analysis: Final Reports" by Walters & Sumwah. An interesting bit: "Other significant findings by the Safety Board were that the accident may have been prevented if an angle-of-attack indicator had been available for use by the flight crew..."

This accident, in which "the PF held the column aft all the way to ground impact", took place at night, in IMC. That makes me wonder how many of these GA stall/spin accidents occurred in the same conditions--without a reference to the natural horizon.

By the way, the stall warning had started at 149 knots, and the airplane stalled at 126.

Posted by: Patrick Underwood | June 3, 2011 9:36 PM    Report this comment

Regarding stalls and the Ercoupe - you have to approach at a significantly higher airspeed to avoid three pointing the landing than most other GA aircraft (I fly short final @ 80mph). I feel this decreases safety. The Ercoupe has a fatal accident rate 4x that of a 172 so I wouldn't use it as the poster child for safety. As for AOA, nice idea, however I don't think today's pilots need another gizmo to be playing with - they need to fly the airplane. Those at significant risk of stall-spin accidents would do well to get a couple hours dual instruction every few months especially if they're flying infrequently.

Posted by: Josh Johnson | June 4, 2011 6:32 AM    Report this comment

I'm always fascinated by the same recurring themes that come up in these accident rate reduction blogs. They seem to be

a) There's nothing that can be done

b) "If people flew as well as I do, they wouldn't have accidents" - along with its corollary, "Anyone who doesn't fly as well as I do has no business flying." Invariably from someone with 10,000 hours of airtime.

c) "If people got more flight instruction, they wouldn't have accidents." Almost invariably from an instructor.

Posted by: Finbar Sheehy | June 4, 2011 5:58 PM    Report this comment


You left out d) Wouldn't things be better if GA aircraft had AOA indicators?

Posted by: Gary Dikkers | June 4, 2011 8:34 PM    Report this comment

Living pilots ARE doing it right. That's the only metric that's provable(we're alive). What's not provable or recordable is training stopping a particular future GA accident(it's speculation).

So yea, those of us doing it right are not dying. Those doing the training are still "hoping" they can change human nature...

Posted by: Mark Fraser | June 4, 2011 8:49 PM    Report this comment

"You left out d) Wouldn't things be better if GA aircraft had AOA indicators?"

No, not really. If there is no solid training foundation, the AOA indicator is just one more expensive thing to install for pilots to ignore.

Posted by: ANDREW HOCHHAUS | June 4, 2011 11:37 PM    Report this comment

Besides, an AOA indicator isn't going to help in a single engine loss of power on T/O. Going back to what Dan Gryder has been saying, your reaction to push forward and get the nose down must be sring-loaded and immediate. If you are waiting to process what an instrument like an AOA indicator is telling you then it is too late.

Posted by: ANDREW HOCHHAUS | June 4, 2011 11:45 PM    Report this comment

>> If you are waiting to process what an instrument like an AOA indicator is telling you then it is too late. > Besides, an AOA indicator isn't going to help in a single engine loss of power on T/O...

Posted by: Gary Dikkers | June 5, 2011 12:03 AM    Report this comment

" If you are waiting to process what an instrument like an AOA indicator is telling you then it is too late. "


I take it you've never flow with an AOA indicator. Learning to use it is a skill quickly acquired, and processing what it tells you is almost instantaneous.

One of the problems with the stall warning horn is that it tells you too long before the actual stall. There are times when it's just fine to fly around with the stall warning horn on constantly but not be in any danger of actually stalling -- for example practicing slow flight. With an AOA indicator you can tell exactly when the wing will stall -- something that is invaluable when flying a slow, power-on, stabilized approach, or needing to make a max performance turn. (Of course experienced pilots should be able to tell the same by the feedback through the flight controls, and fly right at the edge of the burble -- but it requires continual practice to keep that skill finely honed.)

" Besides, an AOA indicator isn't going to help in a single engine loss of power on T/O..."

Yes, it will. The pilot should immediately maintain the AOA for best glide. And if you find yourself needing to try and make the "impossible turn" you'll have a better chance if you can fly right at the edge of the stall. With an AOA, the impossible turn would be slightly less impossible.

Posted by: Gary Dikkers | June 5, 2011 12:04 AM    Report this comment

Training is not really the issue, it is the attitude of some pilots that, "hey, I've got this flying thing all figured out and dont need no stinkin trainin!" that is the problem. We do have evidence that recurrent training works - the military, airlines, and corporate flight departments are doing a much safer job flying large complicated aircraft than we do in little ones. If we want to live long lives as aviators and die at an old age in bed we would be wise to continually improve our educations as pilots. This not only includes cfi time, but also reading and staying current on what's new, attending safety meetings and seminars, learning all you can about your machine, and learning all about your equipment ( even if all you've got is a handheld GPS )

Posted by: Josh Johnson | June 5, 2011 6:37 AM    Report this comment

Gary, You're right. That one seems to be more recent but is now very popular.

Mark, I wasn't arguing, just making an observation. However, I have to disagree with your follow-up: absence of evidence is not evidence of absence. While we do know that (most of) the dead were doing it wrong, we don't know that those who are living are doing it right, only that we haven't been killed yet by any bad habits we may have.

Josh, I have 2 thoughts about the comparison to airline pilots. One is that the cost/benefit tradeoff for additional training for an airline pilot is very different, and may justify a level of training that makes absolutely no sense for private pilots. The other is that I don't know how much of the safety benefit comes from training, per se, and how much from operating flights that are both highly-constrained in terms of their operating parameters, and are also the same mission, day in and day out.

As an aside, I've never flown with an AOA indicator, but I always liked the stall warning on the C152. It is an audio sound made by air sucked past a reed, so the sound varies with AOA rather than being a binary warning (on/off). I felt I could maintain a high AOA fairly precisely by the sound the buzzer made.

Posted by: Finbar Sheehy | June 5, 2011 9:35 AM    Report this comment

Speculating only, but with Airbus experience...
AF447 occurred at night, in moderate to severe turbulence to pilots who are trained to trust the protections the aircraft provides. In the sim, I've done a take-off by accelerating to V1 and pulling the stick all the way back and right. The aircraft climbed out in a nice right turn, protections doing their thing. Stall recovery is the same... TOGA thrust and pull back. The airplane maintains the AOA required for the best climb (it can see it, we can't).

Once in alternate law due to perceived ASI failures, the protections are gone. The airplane tells you this on the central screen, below the engine gauges, in 18pt type. So if you read this and if you can process this while in the middle of everything else and if you can revert to flying in a non-protected aircraft style, you may recover. I know we do little if any training in these situations, but I will be requesting it in the future.

Posted by: Peter Buckley | June 5, 2011 8:45 PM    Report this comment

Dissenting vote –

If you believe, as I do, that a pilot in a moment of stress will act in accordance with his/her strongest conditioning, then consider:

Stall spin: relatively sudden onset.

Level of cognitive processing: low

Most likely response to high-stress sudden-onset event: behavior in accordance with strongest conditioning.

Early Private pilot conditioning in response to initial stall indications: upon hearing the warning horn, continue to apply back pressure to the stick or yoke until full stall is achieved. Multiple repetitions.

Result: pilots trained to respond to warning horn by: a) ignoring the horn while problem gets worse and b) applying improper control response.

Oddly, commercial pilots are trained to recover at immanent stall by applying the correct response.

However, we continue for PPL to instill with primacy and frequency the conditioned response of ignoring a warning signal while applying a control input opposite of the desired input in an actual situation while under duress.

Where Private pilots are concerned, we are, under the watchful eye of FAA and in accordance with the PTS making the stall warning horn a conditioned signal to pull back on the yoke.

Not a question of smart or stupid where pilot response is concerned. Happens too fast for higher-level cognition.

Nothing can be done? If we keep doing what we’re doing, we’ll keep getting what we are getting…

Posted by: Nick Frisch | June 7, 2011 9:12 PM    Report this comment

A stall results in only one way, the pilot pulling the control wheel...
We teach how to stall, not how not to stall.
The early Cessna 150 and 172 had an emergency procedure for exiting inadvertent weather...turn loose of the control wheel, ease the rudder into a standard rate turn for one minute and ease out with the rudder.
I have asked almost 200 flight instructors and examiners about this and none ever had read it..."and it's an emergency procedure!!"
The airplane will fly fine all by itself, It was designed to fly. Someone has to teach the technique.

Posted by: Robert Reser | June 8, 2011 9:00 AM    Report this comment

I have use an Alpha System AOA in a 182 for over 4 years. Stall warning devices have one point typically set at 5-8 knots above stall. AOA devices become active well before that with a caution cue followed by an approach speed cue and finally red cues of an impending stall. AOA are great for steep turn to final situations and stabilized approaches. You fly slower with better control with an AOA.

How about a structure approach to risk management in training for the student to create written guidelines beyond which the will not go. That is when we feel lance a solution in the air rather than a well thought out approach while safely on the ground can go a long way toward breaking the chain of event. After writing down these thgouths then print them put them in a 3 hole punch and file them in a 3 ring binder. Now you have your own personal Operations Manual just like the professionals.

Posted by: Charles Lloyd | June 8, 2011 4:14 PM    Report this comment

The "teaching how to stall, not how not to stall" comment reminds me of a pet peeve I have about stall training - especially in gliders. As we all know, if a wing drops at the stall, you want to try to pick it up with the rudder, not the aileron. This was duly drummed into my head by both my original power instructor and almost every gliding instructor I've ever done stalls with.

I believe this is a dangerously wrong instruction.

The correct instruction is, at the stall, push.

I realized this after accidentally spinning a C-150 not once, but twice in a row while practicing stalls (fortunately, with an instructor aboard). This particular C-150 tended to drop the left wing smartly in an approach configuration stall. Each time, I did what I was told and tried to stop the wing drop with the rudder. That didn't work - and I was so focused on the rudder that... I didn't push!

Low-time pilots are going to be able to think somewhere between zero and one thought when the airplane stalls unexpectedly. If they can manage one thought, that thought needs to be "push!" and not "pick up the wing with the rudder."

As an added bonus, if you push first, the wing will unstall quickly, and then the ailerons will work as normal, so no exotic rudder-based wing-leveling technique is required.

By all means teach wing-pick-up tricks as part of an aerobatics or advanced airplane handling course. But I believe it's a dangerous thing to introduce to student or low-time pilots.

Posted by: Finbar Sheehy | June 8, 2011 10:12 PM    Report this comment

"GA means non-professional and that means ordinary "people" of all types. You have to accept that people crash in planes just like they crash in cars."

This a choice made by the pilot. He trades off safety for convenience, cost, etc. That this is the case is shown by the vastly superior accident rates achieved in some sectors of busibess aviation. Those who choose to fly like the pros and invest in professional grade training will achieve much lower accident rates than the general population of GA pilots. Its just that simple. So make your choice.

Posted by: R Boswell | June 9, 2011 8:09 PM    Report this comment

>>> Those who choose to fly like the pros and invest in professional grade training will achieve much lower accident rates

Posted by: PAUL MULWITZ | June 9, 2011 9:47 PM    Report this comment

"To say that GA pilots are non-professional is to use the simple definition that professional means you do it for money. However, I believe anyone can fly with professionalism no matter how much they are paid."

"Professional" to me means "flies with skill, awareness & judgment." Pay has nothing to do with it.

Posted by: R Boswell | June 10, 2011 6:33 PM    Report this comment

"so no exotic rudder-based wing-leveling technique is required."

Excuse me? Have we come to the point that using the rudder is now an "exotic" technique? And here all along I've been teaching that it's just basic airmanship to use your feet while flying an airplane.

Slowly, slowly we sink to the lowest common denominator.

Posted by: Paul Bertorelli | June 12, 2011 6:02 AM    Report this comment


You're really going to try to claim a glider pilot doesn't "get it" about using the rudder?

And you're going to attack the idea of focusing on what's most important (unloading a stalled wing) to make the best use of the limited panic-affected cognitive abilities of low-time pilots as "sinking to the lowest common denominatir?"


I was referring to the use of rudder to control roll as an "exotic technique." it's a non-standard - and therefore counterintuitive - use of the controls and is relevant only on some aircraft, only when the wing is stalled. I'm comfortable with calling that an exotic technique. And I'm very comfortable with getting students and low-time pilots to focus on unloading the stalled wing, rather than having them focus on a - yes, exotic - technique for momentarily leveling the wings while still stalled.

Should I take it that your defensive response was because you've been teaching students to focus on the rudder when the airplane stalls?

The technique itself is fine, and many instructors teach it - although I wish most of them wouldn't. It has its place in aerobatics training. It's just the wrong priority for a low-time pilot trying to stay alive while freaked out. It requires a non-standard, non-intuitive use of controls to address a secondary problem. Yes, it's exotic.

Posted by: Finbar Sheehy | June 12, 2011 9:38 AM    Report this comment

To be even clearer on what I believe student pilots should have in mind, it is this: "whenever you are flying at a low speed, i.e., in the traffic pattern or loitering for sightseeing, and you are surprised by a sudden rolling motion, push, then level the wings."

A fairly standard stall demonstration would illustrate the point, i.e., bring the airplane to the threshold of a stall and try to level the wing with the ailerons. (On some airplanes this will work, so for the illustration it's best to choose one where it doesn't.) Now lower the nose and try again - success!

There's no harm in going on to point out two other things: first, that flying the airplane with the ball centered (standard use of the rudder) will help to minimize the tendency for a wing to drop and, conversely, having it far off center will produce a sharp wing drop; second, that you can even move the wing drop from one side to the other, at the threshold of a stall, with the rudder.

The second point about the rudder does show that you can pick up a dropped wing with the rudder, but the focus of the exercise is to illustrate importance of making a habit of flying with the ball centered (the true meaning of "seat of the pants," since with practice you don't need to watch the ball that much - and, again, emphasizing an important reason for the standard use of the rudder).

The overall focus, however, should be on lowering the nose first, in response to an unexpected wing drop in slow flight.

Posted by: Finbar Sheehy | June 12, 2011 9:55 AM    Report this comment

I was referring to the use of rudder to control roll as an "exotic technique." it's a non-standard - and therefore counterintuitive - use of the controls and is relevant only on some aircraft, only when the wing is stalled."

I'd tell you that it's decidedly standard, Finbar, in every aircraft. Years ago, when I was in primary for power airplanes, the acronym AER was used, for aileron, elevator rudder, which was--and is--the order of loss of effectiveness of control in slow flight and stalls.

Typically, an examiner will still ask a candidate: If a wing drops during a stall, how do you raise it? The answer many want is with opposite rudder, not aileron, which is the intuitive response.

I don't know about others, but I still it teach it this way. Especially when exploring extended or aggravated stalls. If the airplane yaws left or drops the left wing, respond with right rudder. In some airplanes, just breaking the stall with pitch won't prevent the spin entry, but rudder will. I consider this very basic stuff, not exotic at all.

Further, in cruise flight, controlling roll with your feet is a useful skill. And I remember the date I was taught it: May 1970. The examiner on my PP checkride, upon seeing my struggle to keep the airplane level while refolding a sectional said..."Son, just use your feet."

I'll admit I am a hopeless dinosaur making this old school argument. But I yam what I yam.

Posted by: Paul Bertorelli | June 12, 2011 10:04 AM    Report this comment

Hi Paul, I understand that these things work; snap rolls are the ultimate demonstration of rudder-for-roll. I also understand the "gotcha" games examiners play and that students have to be told what the examiner will want to hear. I, too, hold the wings level with the rudder in cruise flight when using both hands for something - but while that works, it's inefficient and I'd argue is not the standard use of the rudder. However, that gets us into semantics.

My primary point is to ask any instructors who are reading this, to please focus on teaching primary reactions that may save their students' lives (as well as the lives of their passengers). Unload the wing in a stall. If you have an unexpected wing drop at low speed, unload the wing. As long as the wing isn't stalled, their natural reaction (pick up the wing with ailerons, the instinct first learned) will work, instead of killing everyone aboard. Teach that, instead of trying to teach them to override their instinct after being surprised.

If I have persuaded any instructors to change their behavior, perhaps I may someday indirectly save a life or two. There's not much I can do about anyone who has decided to be set in their ways. But maybe even some of them will give it some thought now and again. You never know!


Posted by: Finbar Sheehy | June 12, 2011 1:49 PM    Report this comment


I have no idea what you mean when you say "Unload the wing". Do you roll into a 90 degree bank so there is no load (no lift) from the wings?

Perhaps your method of teaching is appropriate for gliders. In powered planes (at least all the ones I have flown) it is the rudder that will save your life in an uncoordinated stall. Without rapid rudder correction in this case many planes will enter a spin. Since most non-commercial pilots have never experienced a spin this is likely to be fatal.

Of all the "Old wives tales" I have heard, I do believe there is no "Natural" way to recover from a deep spin. Only the prescribed (and hopefully trained) recovery with rudder first and push on the stick will get you out of it before crashing into the ground.

Posted by: PAUL MULWITZ | June 12, 2011 2:48 PM    Report this comment

Hi Paul M.,

I mean "push the stick forward to reduce the angle of attack - and therefore the load on the wing." When the wing drops at low speed, push the stick/yoke before you apply aileron. Just that.

This is not about gliders: It was an airplane that first taught me the lesson about focusing on AOA, not yaw, in a stall; and I learned it the hard way, by focusing on rudder rather than elevator first (I spun anyway, twice). When I started doing things the other way around, I got much better results - not surprising, since stalls are primarily an AOA problem, not a rudder problem. Yes, you may need rudder to recover from a developed spin. But - feel free to try the experiment and let us know if I'm wrong. In the kind of situation actually likely to happen inadvertently, i.e., a gradually approached stall, with no more than "natural" uncoordination (e.g., feet on the floor, but not deliberately looking for trouble with a large rudder deflection), when the wing drops apply forward stick/yoke, then normal aileron against the dropped wing. I've flown quite a few airplanes, and only one did not respond by stopping the roll when the stick went forward, then rolling in the conventional way when the aileron was applied. (The one exception snap-rolled inverted at the stall, with the ball centered right up to the stall break; it had a rigging problem.)


Posted by: Finbar Sheehy | June 12, 2011 3:13 PM    Report this comment

Hi Finbar,

OK, I believe relaxing back pressure on the stick is a good first action for dealing with any stall. This should reduce the angle of attack and probably eliminate the stall. Of course this refers to "Normal" stalls of the approach or departure variety. This is unlikely to help much in an accelerated stall. (I hate accelerated stalls - the most violent maneuver I have ever experienced in an airplane.)

My comment about unloading the wings comes from an anecdotal comment from the only person I know who survived a "Flutter" incident in a Zodiac XL. He experienced "Vibrations" and decided to roll to a 90 degree bank to unload the wings and lived to tell about it. Most XL accident reports suggest the pilot pulled up the nose to slow the plane down (the normal suggested action in a flutter incident). Those pilots all died as a result of massive structural failure including wing separation. I now believe these were all "Wing Flutter" incidents rather than the more common aileron (or other control) flutter.

To my understanding a properly designed and rigged airplane will only drop a wing in a stall if uncoordinated. This is cruising for a spin. If properly coordinated, a docile plane will either drop the nose or just mush in a stall and the rudder issue will not come up.

My only real problem with teaching low skilled pilots to kick the rudder in a stall is some might kick the wrong rudder direction - almost guaranteeing a spin.

Posted by: PAUL MULWITZ | June 12, 2011 5:25 PM    Report this comment

Hi Paul,

That's really interesting, about the Zodiac. Was the airplane checked over afterward, for indications of what happened?

What do you mean by "wing flutter" incidents? How would roll-to-vertical have helped? (Clearly it didn't hurt!)

The situation I'm thinking of on the stall front, is that if you approach the stall gradually, your first indication that there's a problem may be when the airplane rolls opposite an aileron input. Given the docile stall of many planes, the nose may not pitch over and the sink rate may not be the first thing noticed. Even the aileron input itself if likely to be subconscious, so the first conscious realization that something is wrong will be when the wing drops "unexpectedly."

Posted by: Finbar Sheehy | June 12, 2011 6:24 PM    Report this comment

Hi Finbar,

I don't really want to discuss the Zodiac issue at length in this blog. It is clearly off the topic here, and it can get pretty messy.

The short version is there has been a major redesign of the Zodiac XL structure to eliminate the rash of in-flight failures and no accidents have been reported since the upgrade was released. If you would like to get more information from me on that whole can of worms please contact me offline at psm(a)

Wing flutter is something new to me. Apparently it is a diverging oscillation in the entire wing rather than a small area like a single control surface oscillating on its hinge. It seems to come from a structure that is not stiff enough, but it is not even a little bit clear to me why many planes with the same design have no problem but occasionally some unlucky flight experiences this problem. I am guessing the vibrations require the wing to be loaded (i.e. generating lift) to be sustained. When the wing was put out of service by eliminating the air flow the vibration stopped.

Posted by: PAUL MULWITZ | June 12, 2011 8:33 PM    Report this comment

Having attended the SAFE conference I was intrigued by the difference in instructor certification in that the Canadians have a tiered system where, it appears instructors are "mentored" and earn the ability to sign off applicants for advanced ratings or situation as they progress in the system. I then went looking for accident data to determine if a difference exists between the two systems/countries. So far it seems that eh Canadian report accident data in "per registered aircraft" and we report in hours flown. So, I'll keep looking but am interested in anecdotal data. Thanks.

Posted by: William Ross | June 30, 2011 5:02 PM    Report this comment

Great thread!

On the subject of AOA's--Gary Dikkers has it exactly right. Not only are they effective in PREVENTING stalls by showing the lift reserve, but he nailed it when he said that you can learn a lot from the AOA.

The first King Air 90s had a simplified AOA built into the glareshield--right in front of the pilot. It was just a stagnation tab, and identical to the stall warning tab (both were made by Safe Flight Instrument). It sensed the stagnation point on the wing. It was effective--indicated airspeed was irrelevant--and it took into account power application, bank angle, aircraft weight, density altitude, and flap position--all things that affect the stall. Just fly the center diamond for a normal approach--the high diamond in gusty conditions (you could watch the AOA react as gusts hit the aircraft) and the low diamond for short-field landings.

The FAA said that since it only sensed stagnation, that it wasn't a "real" AOA--and Beech stopped installing it. Too bad--as Dikkers mentioned, you can learn a lot from it. I see that Safe Flight is now offering it again.

Posted by: jim hanson | December 29, 2011 12:48 PM    Report this comment

In regard to the "pick up the wing with rudder" vs. "unload the wing"--BOTH are correct.

If the stall OCCURS and a wing drops--the rudder is the most effective in picking it up.

As a glider instructor, though, we teach pilots to live on the edge of stalls--tight spirals when centering thermals, with attention diverted outside the cockpit. Gliders don't have stall warners--and don't need them.

Pilots are taught to automatically relax back pressure (and G's) at the first indication of a stall ("unload the wing"), then get right back to thermaling.

It's stall PREVENTION vs. RECOVERY--both techniques are correct when used in either prevention or recovery.

A similar recovery/prevention technique is used for upset training. Pilots are told "step on the sky" (use rudder on the "sky" side of the attitude indicator) and ALSO to "unload the wing" (a stall can't occur with an unloaded wing). Again, both are correct in their proper context--there is no "one size fits all" prevention/recovery.

Posted by: jim hanson | December 29, 2011 12:57 PM    Report this comment

Paul--thank you so much for your insights into accidents--obtained by trying to make sense of raw data.

I've been reading aviation magazines for 50 years now--and a common lament from anyone that tried to make sense of the information is "FAA has no idea how many hours are flown--or by what aircraft." You would think that the government would have rectified this lack of information, given its importance.

Getting the info from medical records doesn't help--it's only an estimate--a pilot can put down anything he believes to be true--and it doesn't capture gliders, balloons, ultralights, or LSAs. INSTEAD--how about having the AI that does the signoff inspections on the aircraft report the info? EVERY aircraft has to have some kind of inspection. Report the hours flown in the last year, and the aircraft type. To preserve privacy--N-numbers need not be supplied.

Within ONE YEAR, you would have started to build a meaningful database--without a big intrusion into privacy. It would also give meaningful information as to the number of hours flown by aircraft type (supporting or refuting the need for an AD, for example--on that type. It could also be mined to determine the relative safety of particular engines, for example. Right now, all we have is an "educated guess."

Other than "not invented here", why DOESN'T the government get this information, instead of forcing those interested in safety to guess at the number of hours flown.

Posted by: jim hanson | December 29, 2011 1:07 PM    Report this comment

[part 1] Paul and some of the studio audience are giving up too easily. We have no choice: we HAVE TO change human behavior. Name the activity: flying, driving, mowing your lawn, working in a machine shop, chopping onions for dinner; it's lethal unless you have a proactive attitude towards safety.
Remember, we're only about four generations into an era in which the common man operates complex machinery. Human attitudes improve with painful slowness. We are still deeply invested in our last-century attitudes.
It's a mistake to focus exclusively on the minutae, whether it be stall awareness, texting while driving, or standing in an unsafe position wielding a chainsaw. No, step back and look at the big picture. A person who doesn't have a strong culturally-imbedded notion of safety through competence and accident avoidance through taking command, will not respond to the 10th or 12th don't-text-and-drive message.
-- John Schubert, Limeport-dot-org

Posted by: John Schubert | December 29, 2011 1:30 PM    Report this comment

[part 2] But when we get people to absorb these notions, and make these notions a part of their personality, they change. They become proactive, seeking safety info in every activity. They become early adopters, performing an act of safety even if they're the only ones doing so. Translated into specific aviation crash causes, this would include taking pride in the good judgment of giving up one's flying privileges when one gets too old. It would include finding the crabbiest and most demanding flight instructor around for one's BFR, to deliberately practice flying under stress, and not quitting the BFR until you'd puked doing hoodwork. Excellent change will occur if it becomes more of a human norm to take pride in being careful, and if we extend this norm to more socio-economic groups (I think we've already penetrated the suburban minivan moms pretty well, but haven't done so well with refugees from impoverished third world countries).
This is a long process, and it will never have perfect results.
But saying, "No, the buzz job morons will always be with us" is giving up too easily.
John Schubert, Limeport-dot-org

Posted by: John Schubert | December 29, 2011 1:31 PM    Report this comment

We have no choice: we HAVE TO change human behavior."

Oh Puleeze! Who appointed you hallway monitor? The anointed have invaded all parts of our lives, dictating what we eat, drink, wear, drive and fly. 100% of all births result in death so lighten up, no matter how many safety Nazis there are nobody gets out of this show alive.

The sad part is that most do-gooders appear to be strong on opinion but weak on cause and effect analysis. The unintended consequences are ignored, and often the real accident causes or how to prevent them are ignored as well. For example: Airbags. They were designed to keep the moron alive who sits on seat belts. It was a failure and it took years to realize they are useless in a longitudinal crash without also using a belt and harness. In the interim the safety nazis killed more who sat on their belts believing the bag alone would save them. The current popularity of aircraft airbags (and parachutes) is an interesting study in human behavior that isn't well supported by the data.

Posted by: THOMAS M CONNOR | December 29, 2011 11:57 PM    Report this comment

just two things:
1) no airplane in the world can stall by itself, someone else must pull the elevator for that (full up trim position is unable even to activate stall warning). so an airplane doesn't stall but it's stalled by the pilot in any case.
2) every airplane is provided with an angle of attack indicator: the elevator position. you cannot possibly be unaware of its position, you have it in your hands (your life as well).

Posted by: Gabriel Sedioli | December 30, 2011 8:21 AM    Report this comment

Changing behavior part du:I suspect the same is true for the current crusade against drivers using cell phones. Many claim they observed drivers with a phone glued to the ear doing something stupid. I suspect they just forgot about the drivers who cut them off while slurping coffee or talking to passengers or oblivious to other vehicles. How the phone Nazis plan to exempt the average cop is unclear: With a microphone in one hand, the other on a keyboard and radar gun at the ready while speeding only because he can get away with it.

Then there are the environmental zealots who have successfully banned many useful materials with little to no evidence of danger or toxicity. I have a long list, but you get the point: the anointed are not working in your best interest.

Google "Top 10 Causes of Accidental Death." Aside from car crashes the list is surprisingly mundane. Why is it that we have so many experts spending billions on risks that kill a fraction of a percent a year but the big killers have no zealots. Is it because trial lawyers can't find deep pockets to sue for falls, drowning and electrocutions?

Posted by: THOMAS M CONNOR | December 30, 2011 12:14 PM    Report this comment

phpBB :

Could not connect to the database

Posted by: THOMAS M CONNOR | December 30, 2011 12:15 PM    Report this comment

With regard to the stall warner as an AOA indicatory: I can slip a 100 and 200 series Cessna to the right and stall it into a spin without a hint of stall warning horn. That's because those Cessnas have the warner on the left wing so it's fooled by the rising wing that's rising because it's still flying. Cessna apparently did it that way because P-factor and other forces turn the plane to the left, hence spins to the left are more common than to the right. If there were dual stall warners then they might be a form of AOA indicator. But remember:

The FAA claim that a wing will stall at any airspeed and any attitude. Google 'VG chart' for a refresher. This is where AOA vs pitch attitude can get confusing because they aren't the same. It's also where Airbus rudders broke off because the FAA had us memorizing too little about Va. Oops!

Posted by: THOMAS M CONNOR | December 31, 2011 11:09 PM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration