A Spinning Yarn
Buzzing and hot-dogging are the leading spin scenarios, often by highly qualified pilots who ought to know better. As Pat Veillette recently reported in Aviation Safety, the solution may not be more spin training, but more training in good judgement.
This article appears in the May 2002 edition of Aviation Safety and is reprinted here by permission.
Anyone picking a doctor doesn't want medieval snake-oil remedies, conjecture, or anecdotes, nor remedies guided by obvious conflict of interest or conjecture. They want approved remedies that have withstood the methodology of sound science, been submitted for the scrutiny of peer review, and been accepted by the guiding scientific bodies.
Unfortunately, there isn't comparable scrutiny in aviation. And when it comes to spins, most people rely on misinformation, outdated statistics, anecdotal hangar stories, or self-serving opinions.
Measuring progress in combating the problem of spin accidents is a difficult process, and it's been unclear for years whether the industry has made any progress in that regard. Are we stuck on a never-ending cycle?
To answer this question, I launched an investigation of spin accidents, beginning with 1994. I chose 1994 as a starting point because the FAA modified stall- and spin-training requirements in 1993, and I wanted to see if the changes made any difference in the accident record. In addition, accident reports from 1994 on are fairly easy to access, while those in preceding years become more difficult.
There were 11,302 general aviation airplane accidents in the period I studied, 1994 through 2000. When you consider that it requires injuries or "substantial damage" for the NTSB to classify an unfortunate event as an official accident, well, that's a lot of bent airplanes.
During this seven-year period, there were 394 spin accidents on record, or roughly 3.5 percent of the total accidents. Earlier research papers on the topic show that the percentage of spin accidents has certainly subsided from several decades ago.
There were 2,288 fatal accidents in the general aviation airplane population in the timeframe of this study. Of these, 324 (roughly 14 percent) were caused by spins. Again, this is down from earlier years. This is relatively good news.
With 324 of the 394 spin accidents resulting in fatalities, it's clear that this is a dangerous flight regime. Spin training has become popular in some circles, but there remain substantial questions about how effective it is. There are more than 100 pilots in this database who had extensive spin training -- many with outstanding backgrounds -- who still fell into a spin accident.
The ability to perform a spin recovery is often an academic argument, simply because 90 percent of the spins occurred at altitudes that were too low for recovery.
Phase of Flight
(Click chart for hi-res version)
Perhaps predictably, nearly 36 percent of all spin accidents occurred while the aircraft was in maneuvering flight. In fact, 84 of the spin accidents (21 percent) occurred when the pilot was performing what accident investigators sometimes call "ostentatious displays at low altitude" -- otherwise known as buzzing. For a myriad of reasons, you have no safety margin when flying low, and it doesn't take a Ph.D. in rocket science to see that. The accident reports have some head-shaking examples of really bad judgment.
Takeoff was the next most common phase of flight, with 32.7 percent of accidents. This isn't surprising in and of itself, but what was unexpected is that nearly half of the takeoff spin accidents were due to the pilot showing off on takeoff and using pitch and bank attitudes far in excess of any safety margin.
Since deliberately showing off on takeoff is so closely related to buzzing, it becomes obvious that more than a third of all spin accidents were the result of really bad pilot judgment very close to the ground. The lesson is clear: If you avoid the temptation to show off or buzz friends, your chances of an inadvertent spin accident are dramatically lower.
Nearly all of these accidents occurred at less than about 300 feet AGL, so any spin recovery technique would have been useless. In fact, we found that a disturbing number had been spin trained, even to the point of flying competitive aerobatics. It's pilot judgment that needs to be improved.
Roughly one-third of the takeoff stall/spins resulted from engine failures. Most of them also occurred within a few hundred feet of the ground. In fact, in many cases, the aircraft had barely dropped off on a wing as it struck the ground.
Only a few of the pilots attempted the "impossible turn" [Ed: returning to the runway], a sign that the training has finally convinced enough pilots of the futility of that attempt from low altitude.
I've experienced two actual engine failures on takeoff, both on the same day. Most surprising to me was how rapidly the event happens. Pilots need to be spring-loaded to push the nose down at the first sign of power loss on initial climb, but it's still a tough thing to do.
Twenty-four spin accidents during takeoff were partly caused by a combination of high density altitude, heavy weights, and adverse winds, all of which led to a failure of the pilot to maintain sufficient flying airspeed. Again, this is a deadly combination, and the conditions are plainly obvious, but several pilots keep stumbling on this error year after year.
Approximately 18 percent of spin accidents occur during landing. Of these, slightly more than half occurred during an emergency landing. Surprisingly, only one or two of the classic "turning base to final" scenarios were in the database. Most of the landing spin accidents were officially attributed to the pilot being distracted or preoccupied with a mechanical failure, followed by the pilot's failure to maintain sufficient flying speed.
During any flight regime, normal or emergency, the single most important task you must accomplish is flying the aircraft. "Aviate, navigate, and communicate," is an old saying that remains true today. Even if your hands are on fire, you still have to fly the aircraft.
Go-arounds were the next most common phase of flight involving spin accidents, representing about 7.1 percent of accidents. Almost all of these involved a delayed decision to go around, followed by poor control of the pitch and power, sometimes forgetting to put the flaps in the correct position for climb, and finally not recognizing that the airspeed had decayed below an acceptable level.
Cruise was the last phase of flight involving spin accidents (6.3 percent). Most of these occurred in IMC, and nearly half occurred in icing conditions that exceeded the aircraft's anti-icing or de-icing capabilities. In other words, the aircraft was coming out of the sky regardless of any attempt by the pilot to recover from the spin.
In a previous study I did on the altitude lost during an incipient spin and recovery, I found it took hundreds of feet to recover from an incipient spin using the optimal recovery technique. Less-than-optimal recovery technique significantly increased the altitude lost.
Out of curiosity, I plotted the altitudes cited in the spin accidents. More than 90 percent of them started at such a low altitude that the spin was unrecoverable. The best spin pilot in the world couldn't have recovered these aircraft prior to ground impact, simply because there was not enough altitude available. This has very important implications for the stall/spin problem -- from both operational and training standpoints.
Obviously, pilots need to avoid the stall/spin envelope so close to the ground. In addition, the emphasis in training needs to be on preventing the high angles of attack that initiate the spin sequence, as well as proper use of rudder as the airplane approaches the stall. Knowing how to execute a full spin recovery is rather irrelevant at such low altitudes.
[Ed: A selection of spin accident narratives involving experienced pilots.]
A Beech T-34B was observed by the airport manager to climb to about 10 feet above the runway, retract its landing gear, and then pulled up into a 75- to 80-degree climb angle. At the top of the climb, the aircraft turned left past a 90-degree bank, then descended at approximately a 70-degree pitch down angle and turned about 135 degrees to the left when it struck the ground. The ATP/CFI in the front seat had over 4,000 hours of flight time, including aerobatic training. The CFI in the rear seat had over 2,100 hours.
The pilot of the unlimited class YAK-54 stated he was going to demonstrate some aerobatic maneuvers to the pilot-rated passenger. Witnesses observed the aircraft enter into an inverted right spin at a lower-than-recommended altitude. The aircraft did not recover from the spin and impacted the ground. The PIC was a general officer in the Air Force and commander of Alaskan Air Command. The general had flown more than a dozen different fighters, bombers, tankers, and experimental aircraft, including the F-117 Nighthawk, the B-1B Lancer, and the X-29. He was a command military pilot with more than 4,100 military flight hours, in addition to several thousand civilian flight hours. He held waivers from the FAA allowing him to perform low-altitude aerobatics and had competed in many civilian aerobatic competitions.
The unicom operator saw the Cessna 172RG rotate about midfield at the high-density-altitude airport (7,542 feet) and assume a nose-high climb attitude. The airplane reached about 100 feet above the runway, then appeared to stall and roll off on the left wing. Four occupants were killed in the impact. The PIC was a U.S. Marine Corps pilot, as was the right seat passenger.
Witnesses observed the PA-18 approximately 20 feet off the ground when it stalled and spun "straight down nose first." The two commercial pilots on board were killed. The rear seat occupant was a commercial pilot who was also acting as an instructor pilot. His last medical indicated that he had logged 19,800 hours.
The student pilot, formerly commercial-rated, took off, maintained a level flight attitude just above the runway for 5 seconds, then put the airplane into a steep climb. Witnesses then observed the aircraft descend uncontrolled, killing the pilot. The pilot's commercial license had been permanently revoked two years earlier for flying during a suspension for low altitude flying.
The North American T-6G-NA began a high-G pull-up. At the top of the climb, the aircraft nosed over and began a slight turn left that became a hard left that turned into a partial snap roll with a resulting spin to the left. After three to four turns the aircraft struck the ground. The aircraft's POH states that the aircraft will lose 500 feet of altitude for each full spin rotation. The PIC was an International Council of Air Shows "Ace" and zero-altitude aerobatics examiner.
Witnesses observed the Beech S35 attempting a full-throttle climb at a low altitude before rolling inverted and descending in a near vertical spinning attitude. The two occupants were both retired airline pilots. The PIC had over 18,000 flight hours. The POH states, "In a cross control stall, the airplane often stalls with little warning. The nose may pitch down, the inside wing may suddenly drop, and the airplane may continue to roll to an inverted position. This is usually the beginning of a spin ... The pilot must be able to recognize when this stall is imminent and must take immediate action to prevent a completely stalled condition ... Recovery may be impossible prior to ground contact due to the low altitude."
If you think spin accidents are the province of inexperienced pilots, you'll have to alter your expectations. Private pilots were involved in 46.4 percent of the spin accidents, and student pilots accounted for 3.8 percent. That means the other half were commercial pilots, ATPs, instructors -- not the neophytes you might expect.
But before you read too much into this, there is more of the data to consider. For instance, a large number of spin accidents occurred in experimental aircraft, which opens up a whole new set of questions. In addition, because so many of these accidents involved low-altitude buzzing, it's obvious that pilot judgment, rather than experience level, is a primary factor in these accidents.
For many years I have been staunchly "pro-spin training." I had been spin-trained, both in the civilian and military worlds, and I believe it made me a better aviator. I have given spin training to many pilots over the years, in both powered airplanes and gliders. Fellow spin-training advocates have stated that spin training makes a pilot more aware of the factors leading up to a spin, more capable of proficiently handling the aircraft at high angles of attack, less likely to enter into an inadvertent spin, and more likely to recover.
That said, I was shocked at the outstanding credentials of many of spin accident pilots. The seven-year trend of data certainly casts doubt on my enthusiasm for spin training. Commercial pilots were involved in 35.5 percent of the spin accidents. Airline transport pilots were involved in 11.7 percent of the spin accidents.
The credentials of many of these pilots were quite impressive. Some of the accident pilots were highly accomplished military pilots. Some were "fighter weapons school" instructors. Some were certified by aerobatic associations, were accomplished airshow pilots, held "low-altitude aerobatic waivers," or were aerobatic examiners. Many had graduated from flight training programs specializing in unusual-attitude recoveries, spins, and aerobatics. In fact, many were even accomplished aerobatic or spin instructors. Several were test pilots. Even a highly experienced Reno air racer with amazing experience and credentials was involved in an unfortunate spin accident.
More than a quarter (29 percent) of the accidents involved pilots who had documented spin training. Some of these involved pilots deliberately doing spins, but 83 involved spin-trained pilots who failed to recognize and react to an inadvertent spin in a timely manner. Unfortunately, being spin-trained doesn't necessarily mean the pilot will be immune from having a spin accident.
Forty-nine of the accidents (12.4 percent) involved pilots who had aerobatic training and certification. Some of these spins occurred while practicing aerobatics, though 27 involved acro-trained pilots who failed to recognize and recover from an inadvertent spin in a timely manner. Many of these victims had attended brand-name aerobatic schools or emergency maneuver training.
In theory, they should have been highly capable of preventing and recovering from an incipient spin. Unfortunately the data shows that being a spin instructor doesn't guarantee the prevention of a fatal spin accident, either deliberate or inadvertent.
Twenty-three of the accidents (6.1 percent) involved military or former military fixed-wing pilots. This is another revealing statistic because the military services have such in-depth spin training programs. I remember being required as a lieutenant to demonstrate deliberate spins as part of the training curriculum. That was supplemented by hours of classroom instruction on the topic, and numerous hours spent studying the spin section in the aircraft manual.
We were quizzed on the topic almost daily. The program was designed well for the curriculum, since the most obvious spin risk in the training curriculum occurred during aerobatic practice, which occurred at 8,000 feet AGL and higher. Unfortunately, this in-depth and rigorous training may not transfer directly to the aircraft and spin scenarios typical of general aviation flying.
Almost one-fifth of the spin accidents (17.5 percent) involved a flight instructor. More than a decade ago, I did a study of stall/spin accidents that was published by the National Research Council. Among its findings were the lack of standardization and quality control of flight instructor candidate preparation, particularly in regards to stall/spin knowledge and application. Apparently not much has changed there.
It seems apparent that any pilot, regardless of experience or spin training, can fall victim to a fatal spin accident.
What You're Spinning
Trends among the aircraft involved in spin accidents are as enlightening as the profiles of the kinds of pilots who crash. One of the few rigorous scientific studies published by an independent reputable scientific society found that aircraft design was actually responsible for the decline in stall/spin accident rates, and changes in training had little overall effect.
Since I teach graduate engineers and medical students in an interdisciplinary doctoral program on injury and accident prevention, I study the medical and occupational safety journals extensively. The "systems safety" method has consistently been determined in many industrial settings to be the most effective method for preventing accidents and injury. The first step is to reduce the risk (hint: low-altitude buzzing). The next step is to change the design of the equipment to incorporate safety features. Third, incorporate warning devices. Training and procedures have been proven time and again to be the least effective methods for preventing accidents.
Statistically speaking, one of the problems with assessing the risk of a particular airplane is that flight times are so uncertain. Without having a good handle on flight time, it's difficult to say with conviction that one category of airplane is more spin accident-prone than another. However, I still think there are some things you can take at face value.
Experimental and amateur-built aircraft were involved in 95 of the spin accidents (24 percent). The best guess among the feds is that amateur-built aircraft account for only 2 percent of all flying hours. Even if that number is wildly wrong, it's still clear that there is a big problem here.
Spin Training: A Waste of Time?
Spin training obviously teaches you how to recover from spins, but more importantly it helps pilots avoid inadvertent spins in the first place.
The spin is the result of yaw and roll at a high angle of attack. Avoiding the yawing moment -- proper rudder/aileron coordination -- is an integral part of any attempt to solve the problem with unintentional spins.
Unfortunately, pilots tend to spend little time and effort in the realm of high angles of attack. Because they've been cautioned about stall/spin scenarios, they keep the speed up on landing approach, which leads to problems like runway overruns and runway loss of control.
Spin training programs are often combined with intro aerobatics for two simple reasons. First, aerobatic airplanes are spin friendly and responsive -- both in getting into and getting out of a spin.
Second, and perhaps more important, spins that happen during aerobatic training are more like the inadvertent spins that interrupt "regular" flying. A conventional training scenario in which the spin is a planned and isolated occurrence doesn't have the training value of a real unexpected spin.
For spin training to work, it has to go far beyond a simple one-turn spin to the left or right. It has to look at the scenarios that lead up to the spin and give pilots the confidence and tools to properly fly in those regimes.
Is spin training a waste of time? Not if it's done right and you have reasonable expectations. But it still won't protect you from extreme lapses of judgment.
-- Ken Ibold
Spins involve an aerodynamic region that features angles of attack exceeding 45 degrees. Even the best computational fluid dynamic analyses have a hard time accurately modeling such extraordinary angles of attack.
NASA did some flight testing in the 1980s on an airplane on which they changed the trailing edge of the wing from a 90-degree angle to a fillet for drag reduction. That small change completely changed the aircraft's spin characteristics. Seemingly small and insignificant changes to an airframe can result in catastrophic changes to an aircraft's spin characteristics.
Another example of this is the changes in the spin characteristics in the Piper Tomahawk due to the changes in the wing construction from the prototype to the later manufactured models.
Therein lies part of the problem in the experimental aircraft world. There are many variables during the construction phase, and it's easy to build a wing that has a slight warp to it, or perhaps some looseness in the ribs, or maybe some minor changes to the airfoil, or maybe some minor surface roughness that isn't detectable to the eye or touch but that changes the nature of the boundary layer.
A second factor relating to experimental aircraft is that there is no assurance that a homebuilt can recover from an incipient spin. The number of variables involved in spin recovery is huge, including tail size, blanking out of the empennage surfaces at high angles of attack, prop wash effects, aircraft weight distribution, and hundreds of other parameters.
Third, there is no assurance that the so-called "standard" spin recovery technique will work on any given experimental. Finally, many builders ignore the need for transition training and are the test pilot for their creations. Without someone familiar with the type showing them the nuances, such pilots can get into serious trouble.
Incidentally, from a historical and statistical perspective, the earlier studies that examined spin accident rates were done at a time when amateur-built aircraft were fairly rare.
Older generation taildraggers -- J-3 Cubs, Taylorcraft, PA-18's, etc. -- were involved in a disproportionate share of spin accidents, about 21.6 percent of the total. Many of these aircraft readily spin, and most do not have stall warning devices.
High-winged aircraft were involved in 63 of the spin accidents (16 percent). It appears that the earlier models of these high-winged training aircraft were more commonly involved in spin accidents. Design changes on the later models included drooped leading edges along the outer span, less control deflection and less flap deflection, all of which will help the stall resistance of the aircraft.
Large flap deflections were a definite trend in the database. The drawback of big flaps is that the drag can quickly decelerate the aircraft and lead to high angles of attack in a hurry. Large flaps were a factor in nine of the go-around accidents, four takeoff accidents, and 13 landing accidents.
By the way, most of these aircraft do not recover from a spin with the flaps deployed. The pilots could have possessed the ultimate in spin recovery skills, but the airflow disruption over the tail created by the flaps inhibits spin recovery and makes any attempt at spin recovery futile. This highlights the importance of stall recognition and prevention. Proper flap management is vital.
Aerobatic aircraft were involved in 46 spin accidents (11.7 percent). This is another area where the spin accident rate is probably disproportionate to the aircraft's use. Most of these aircraft were certified to recover from a fully developed six-turn spin. Ease of spin entry is one characteristic rather common to most aerobatic aircraft, making spin recovery training usually one of the first steps in an aerobatic training course. Nearly all of the aerobatic aircraft spin accidents involved pilots who had some level of spin training.
Tricks and Traps
- Choose an aircraft with spin-resistant handling qualities.
- Be prepared for an engine failure on takeoff and have a plan for handling it.
- Practice good engine maintenance, proper preflight inspections, and engine warm-up procedures.
- Be proactive in detecting and making corrections for the wind when flying a traffic pattern.
- Manage the pitch, power, and flaps in a smooth, methodical manner.
- After an engine failure, remember to keep flying the aircraft until it comes to rest.
- Remember that maintaining aircraft control is always the most important task.
- Buzz or make other low-altitude, high-bank, and high-pitch-angle maneuvers.
- Exceed acceptable parameters in glide path, runway alignment, sink rate, or airspeed during a landing approach. If you do, make a decision sooner rather than later to go around.
- Ignore the takeoff risk factors, especially high density altitude, heavy weight, and gusty wind conditions.
Twin-engine aircraft were also involved in 46 spin accidents. Eight of these involved an actual engine failure. The multi-engine spins involve a different set of scenarios and problems than single-engine spins.
High-performance aircraft were involved in 31 spin accidents, most of which occurred in IMC or icing conditions. Pilot experience was cited as a contributing factor in roughly half of these accidents, in that the pilots may have been flying too much aircraft for their training and experience level.
Once again, it is pilot judgment rather than stick and rudder skills that appear to be a primary factor in the error chain. An additional aspect with this category of aircraft is that most are not certified to recover from a fully developed spin. They are tested to the requirements of the normal category, which is basically an incipient spin, and there is no assurance that the aircraft will recover from a fully developed spin.
In addition, an ice-covered aircraft has absolutely no guarantee that it will recover from a spin, period. That's a whole new ballgame and, given all of the variability involved, you would then become a real test pilot.
Low-winged training aircraft were involved in 28 of the spin accidents. Some of the more common low-winged training aircraft now incorporate stall warning horns, drooped leading edges and restricted control movement, all of which make an aircraft less prone to spinning.
The spin debate rages on because most people rely on anecdotal evidence to support their opinions. But some things are clear. Avoiding high angles of attack in maneuvers close to the ground is vitally important.
Since 90 percent of spin accidents are caused by spins that occur at an altitude insufficient for a spin recovery, and since many of the spins occur in an aircraft configuration in which a spin recovery is highly unlikely, it really becomes a moot point whether knowing a spin recovery would make much difference. Preventing the stall is far more important in the typical spin accident error chain.
There's been a lot of hype about spin training, and pilots have to decide for themselves whether it gives much bang for the training buck. It's hard to get too enthusiastic for spin training when you see the accident narratives involving accomplished aerobatic pilots who fall into the spin trap.
There are many proactive decisions pilots can make that will substantially lower their chances of a spin accident -- without spending a bundle on specialized training. Improving pilot judgment and basic airmanship would go a long way in making inadvertent spin accidents far fewer.
Pat Veillette is an aviation safety researcher working in the training department of a large carrier.