8 min read

During my 45 years of observing and writing about aviation, pilot upset training is a topic that has waxed and waned. For the past few years the idea of learning how to recover from an extreme attitude is in ascendance. But the reason upset training emphasis falls in and out of favor is because it just doesn’t work.

The typical pilot upset training course takes pilots who fly normal airplanes for transportation reasons and puts them in an aerobatic-capable airplane. Most often the training airplane is something like an unlimited aerobatic piston single such as an Extra. The more costly programs use ex-military jet trainers such as the Czechoslovakian L-39 that was developed for the Eastern Bloc and is now widely available in the U.S.

The trainee pilot gets some basic ground school instructing how to recover from a spin, or how to roll to the closest horizon, even when inverted. Then, in an aerobatic airplane with an aerobatic instructor, the trainee rolls and tumbles through extreme attitudes. And, of course, the flights are always done in visual conditions.

Upset training certainly won’t make you a worse pilot. And the courses can be eye-opening for a pilot who has never flown beyond a 60 degree bank, or 20 or 30 degrees nose up or down. If you’re lucky enough to draw the right instructor, the course can even be fun.


Does training in an L-39 really transfer to a Citation or a Baron?

But what you learn tumbling through the sky in an upset recovery course won’t do any good in real life flying in a business jet, or even a personal piston single.

The reason upset training doesn’t help in the real world is that it’s built on a false premise. That premise is that upsets happen to well trained and careful pilots who are following all procedures. That’s just not true. Upsets are not some aerodynamic phenomenon lurking in the atmosphere to grab pilots following well structured procedures.

To experience an upset a pilot must first lose control of the airplane. Simply replace the term upset with loss of control and the issue becomes more clear.

Advocates of upset training say the careful pilot may lose control because of unforecast or unreported extreme turbulence, or after being surprised by an instrument failure, or maybe because he just wasn’t paying attention in the dark or in the clouds and the airplane went out of control. Or maybe he tried to climb too high for the weight of the airplane and the air temperature and the airplane stalled. Or maybe he poked his nose into a thunderstorm and found extreme turbulence. Or, even more likely, maybe he mishandled an autopilot or flight management system failure and lost control.

However it happens, our upset-trained pilot who loses control of the airplane must, in a second or two after departure from controlled flight, remember how he moved the stick, pushed the rudder, and managed the throttle in that L-39 and recovered. Huh? Our guy is so careless or inattentive, or poorly trained in normal flying, that he lost control, but in the panicked few seconds that follow, he remembers the upset training and responds perfectly. I don’t think so.

Even if unforeseen events were to force your normal category airplane into a spin, what does the upset trained pilot do? During the upset training, recovery from a spin was quick and positive. Rudder to stop rotation, pop in forward stick, and the aerobatic airplane is flying again.

Just about any aerobatic airplane will fly itself out of a spin if the pilot takes hands and feet off the controls. In fact, that was one of the many flying qualities requirements for the U.S. military’s T-6A Texan trainer. It had to demonstrate hands off recovery from even an inverted spin. But is your King Air, or Citation, or Embraer going to recover from a spin even with perfect rudder and stick input? Or your Baron or PC-12? Maybe, but it hasn’t been demonstrated.

Wouldn’t it be much more valuable than upset training to train and practice our basic instrument flying skills? Deal with instrument and equipment failures in high fidelity simulators? And stay out of thunderstorms, and within the safe operating envelope to avoid an upset in the first place? I think so.

But, I can hear you yelling at the computer—what about a wake vortex? You’re right. Flying into a powerful wake vortex has caused a number of pilots to lose control and crash. But essentially all high risk vortex encounters happen close to the ground on takeoff or approach. With so little altitude available, is our upset-trained ace going to roll a business jet through a full 180 degrees or more and not lose more than the altitude available?

It’s one thing to flip an Extra or L-39 through a roll in a couple seconds, but try it in a business jet, or turboprop, with flaps and gear down. In non-aerobatic airplanes it’s going to take an excruciatingly long time to complete a roll, even with the aid of the vortex, and the nose will drop so much no matter how you pull, recovery within hundreds of feet is impossible.

The behavior of vortex wakes and how to avoid them is well understood. If we follow wake avoidance guidance the odds of a vortex encounter are very low.


In a Citabria, recovery from an upset is pretty simple, but what about on short final?

As in all things aviation, wake avoidance is not absolutely perfect. The most recent fatal crash blamed on a wake I recall happened to a Baron pilot taking off at Memphis several years ago. The Baron had just lifted off when it flipped inverted in clear air and light winds. Investigators determined the cause was a wake from an airplane that had departed well beyond the time and distance parameters that would have normally made its wake a threat. The wake cause seems improbable, but there is no other reason for the loss of control. And there is no reason to believe that any level of pilot skill by the Baron pilot could have avoided the crash at such a low altitude after liftoff.

OK, but what about that Air France Airbus crew over the Atlantic, at night, in an area of very active thunderstorms where the crew lost all three airspeed inputs because, investigators determine, the pitot tubes all froze over?

That overstressed crew still had attitude information, altitude data, and ground speed information. But it was a black night in who-knows-what kind of turbulence with the crew advisory system (CAS) messages scrolling by faster than the tape out of an IRS examiner’s adding machine. The CAS messages were accompanied by aural beeps and buzzes, and no doubt spoken warning words from the system.

Could they have maintained control without any airspeed indication if they had been upset trained? Probably, if the confusion and stress hadn’t been so high. The cockpit workload would have been so over-the-top in such a situation that managing it would take lots of luck, as well as lots of skill. If you think you can’t be overloaded in an airplane, just get in a Level D sim with an instructor determined to give you more emergencies, failures, and complex approach procedures than you can handle. Overload can and will happen to any of us.

Instead of flying mild aerobatics in an L-39, there is a technical solution that could have saved the day for that Airbus crew over the Atlantic. It’s synthetic airspeed. The onboard systems knew the altitude, air temperature, groundspeed, ground track, magnetic heading, and—until just seconds before the pitots froze—wind direction and velocity. Given those data, a sharp private pilot student could work backwards with his E6B to come up with a very close estimate of indicated airspeed. The computer power found even in most piston singles could do the same thing if properly programmed.

With a close approximation of indicated airspeed/Mach, the Air France crew almost certainly would have avoided the deep stall that doomed them. They had all been trained to lower the nose to recover from a stall since their earliest days as student pilots, so it wasn’t more training that was required, upset or otherwise. They simply needed life-saving airspeed information the onboard computers could have produced.

An upset training course can certainly make any of us more precise pilots, and probably more cautious pilots. Enhancing stick and rudder skills is always a good thing. But as a solution to loss of control in flight the upset training just isn’t it. Avoiding the loss of control is what matters in every case, just as avoiding flying into the ground is of paramount importance. And the more we train and practice maintaining safe margins, the safer all of our flying will be, whether we have flown inverted or performed a roll, or not.

Mac McClellan
66 replies
  1. John Royce
    John Royce says:

    It seems the author has either jumped to some pretty big and inaccurate conclusions, or witnessed some pretty shocking UPRT training..
    Given the blatantly wrong idea that “most aerobatic aeroplanes will recover from a spin themselves” (can’t think of a single aeros monoplane that would) the former would seem to be the case…..
    Here in EASA land we are a month into mandatory Upset Prevention and Recovery Training prior to initial type ratings, and it is becoming clear that after 20 or so students completing the short course, the watering down of PPL/CPL syllabi to remove spinning and minimise time in more usual attitudes has come home to roost.
    Whilst the bureaucrats are partly to blame, many grass roots instructors are equally poorly equipped to be exposed to, never mind deal with, abnormal G loads and extreme uncommanded pitch and bank angles. A case of the partially sighted leading the blind… Perhaps UPRT should be extended to pre FlC requirements?
    FCL 745 dictates that prevention is the emphasis in theoretical training, with recovery a big but not exhaustive part of the practical flying side. Simple exposure to the sort of upsets which most students have never experienced, may well minimise pilot freeze and cognitive impairment in a future real life scenario – a reoccurring theme in many accidents.
    Practical experimenting with the aerodynamics at work – aircraft stability, changes to lift vector etc is something often never undertaken before, and the theory long since forgotten after ATPL exams. Such exercises really do tie everything together in the student mind for possibly the first time in their journey to commercial flying.
    And such understanding equips the pilot to deal with upsets much quicker and with a much more balanced and procedural approach. Perhaps it makes a difference of seconds from upset to stabilised flight – but in many accident reports it is clear that seconds were the difference between life and death.
    Of course such flights are carried out on aircraft certified to deal with student errors and more dramatic upsets, but a good training provider will keep recovery SOPs inside heavy jet load tolerances, utilise MCC procedures, and with general procedures in line with transport aircraft wherever possible (to use the author’s example of AA447 pilots “they would just follow training from initial flight training days automatically and push the nose down” then the logic continues that such a reflex would have resulted in a large immediate thrust increase. It may be standard on SEP trainers, but we all know such a reflex is asking for a secondary stall in most airliners. So early training isn’t a lot of use.
    The author is of course absolutely correct on two points. Any stick and rudder time is useful. And it is tremendous fun!

    • M Winger
      M Winger says:

      Huh (as you said in your column)?

      First, I agree with John in this matter, but maybe for a few other reasons.

      The introductory flight to what you call “upset training” (geeze, new tech talk, so is there a difference?) would mostly be performed in a Citabria due to its availability. Next, as a peruse the article you seem to be all over the aviation/weather spectrum with so many tails of woe scaring the heck out of all of us.

      I get it. You want us aware of the potential of being prepared for the whatever, but so much for the apparent little time you have for this report.

      Clearly the story, training, and consideration is warranted for the experience of possible learning of inverted flight conditions, so John R. thanks!

    • Shane
      Shane says:

      Interesting! Im a new pilot (PPL) I have less than 100hrs on my books. So I’m not going to pretend to speak with some false bravado of aviation experience or wisdom. That said! I did do (my choice as my CFI had offered it) spin training and upset training in a Cessna 150. NO I didn’t do loopy loops or barrel roles or massive G load turns like one would in an L39. But we did pull some G’s and more than doubled our stall speeds in those high bank and extraordinary steep turns!

      WHY? it was simple for me. I demanded the training though uncomfortable as it initially was, I fly my wife and kids around! Sure I or even maybe a super skilled pilot couldn’t recover close the ground on some of the mentioned incidents above. BUT I sure knew what it felt like to stall at high speeds I better understood how and when its likely to happen! I also got a better understanding about flying by the seat of your pants (translating what those forces pushing your butt around in the seat mean).

      Maybe my experience wasn’t “true upset” training. but I feel it was life saving training!

  2. Brent Dolphin
    Brent Dolphin says:

    With AF447 I don’t think there was any amount of other training that would have saved the aircraft. The most junior pilot was PIC and the other pilot present did not know the junior pilot was holding the stick back the entire time, even at “go around” thrust. When the captain came in he knew within moments what was happening as he could see both sticks. (Airbus dos not mirror stick movements from one side to the other). The stall warning also kept blaring which should be a good indicator you are in a deep stall with 20* nose up at full thrust and losing altitude.

    To best understand, read the CVR transcript. It is very chilling.

    I agree that upset recovery is probably not useful in an emergency when people are already stressed (unless you practice it in commercial aircraft all the time) but it is a good eye opener on how to analyze a situation, and it is good fun. On a side note, we did it under the hood also.

    • Jim
      Jim says:

      Upset training would have or should be teaching AOA and AOA would have saved Air France.

      Upset training also would have helped reduce the excessive rudder dithering in wake in Nov 2001, a crash that had sufficient altitude to handle the wake but instead tore its own tail off.

  3. Bob H
    Bob H says:

    BD: “With AF447 I don’t think there was any amount of other training that would have saved the aircraft…..The stall warning also kept blaring which should be a good indicator you are in a deep stall with 20* nose up at full thrust and losing altitude.”

    What you are suggesting is that professional pilots can’t be trained to recognize a stall condition and take appropriate action. The captain was indeed taking appropriate action. It was the 1st officer who was in a white knuckle panic unable to provide correct inputs. Simulator training certainly could have improved the 1st officer’s performance. In addition, to Mac’s point, better “other training” could have prevented the conditions that led to the upset in the first place.

    • Brent Dolphin
      Brent Dolphin says:

      Bob, what I am suggesting is that the junior of the three pilots here would have had a considerable amount of training already and stall recovery is one of the basics covered after only a few hours of training. Another few runs through a HF simulator would not have made him a better pilot. The stall warning had been active for nearly a minute and the first officer had told him to release the back pressure. Some people just don’t have the ability to take appropriate action when things aren’t in the checklist.

      I would also say that if there were an indicator on each side of what each pilot was doing with the stick, the FO would have been able to double check whether the JO did release back pressure as he had been instructed.

      Also, what Dave said on over reliance on technology and automation.

      George, the stall warning went on for 54 seconds* and stopped as the captain came back to the flight deck and even without the warning he recognized right away what was happening.

      The FO did push the side stick forward a couple times, but the FBW on an airbus averages the inputs so when the JO was still puling back the system averaged the inputs to neutral.


      “While the stall alarm sounded continuously for 54s the captain, urgently called back from a rest break, re-entered the cockpit just as it ceased. The warnings then became intermittent, owing to A330 logic that cuts out the alarm if airspeeds become invalid.”

      The big question in the flight is why did Bonin pull back on the side stick to climb when the autopilot disconnected.

  4. Dave M
    Dave M says:

    If you listen to the the AF tapes and see the FDR simulation you can actually see the junior FO had the side stick pitching the nose up and initiated a climb. The plane did climb to prob its max alt for the current weight with max thrust and then went into a deep stall. It had a nose up indication to about 10000 feet. The CA did not notice this until that point where he finally said from the Jumpseat to pitch down. The first problem is the CA (PIC) left for his required break right before the aircraft entered the severe weather conditions – Very Poor PIC decision making. The senior FO that took his place was an office working pilot who was on the flight to get current. His failure was to not take control until 10000 feet and try to recover which left limited alt and he had to pull the nose back up when the ground prox warning system started alerting. This goes back to basic aerodynamics, basic piloting skills and lack of understanding. Over reliance on automation, poor primary training and very poor CA (PIC) decision making is what lead to this disaster.

  5. George Haeh
    George Haeh says:

    In AF447, the stall warning halted when the airspeed went below the threshold set by the software (you don’t need the stall warning blaring during taxi). I recall from the BEA report that two nose down inputs stopped when the stall warning came on as the lower limit of the stall warning band was reached.

    As for vortices aka rotors, I encountered one downwind of a ridge at 300′ AGL as revealed by one of the flight recorders in my glider. The shear in a vortex can be as much as twice the wind speed.

    I have a running argument with instructors (except in a full aerobatic course) that advocate holding pro spin controls until one full rotation has been completed as it builds the wrong muscle memory.

    That muscle memory might have you smacking the ground halfway through your first (actually last) rotation spinning out of a turn to final.

    Staying alive in a mishandled or disrupted turn to final requires catching the wing drop before the spin is allowed to develop further.

  6. YARS (Tom Yarsley)
    YARS (Tom Yarsley) says:

    Mac is absolutely correct.
    This is why mandatory spin training is of NO use in a situation in which a pilot has his/her head so far up his/her butt, that s/he UNINTENTIONALLY stalls/spins an aircraft in a base-to-final turn. NONE.

    THAT neat trick would require the flipping of a skull switch that has two labeled positions: Moron and Genius. Somebody please tell me, as soon as you locate that switch.

    • ChrisBow Piloto
      ChrisBow Piloto says:

      That switch is there and I have watched it get flipped. The moron part usually happens because of distraction and the genius part wakes up when they realize what is going on. With proper training, a person can completely make that switch in less than a second.

    • Eric
      Eric says:

      I can agree that full spin training in a highly aerobatic airplane might have limited benefit for many pilots, but, spin entry (and avoidance/prevention) is extremely important, and upset training can be beneficial for all.
      The stall/spin on base-to-final example is not a spin (there’s too little altitude available to spin before hitting the ground), it is a spin entry. Good training, even beginning at pre-solo stage, should impress upon the pilot the rapid roll to inverted from a skidding stall and how to avoid it. As the DPE before my ATP checkride said, “it is IMPOSSIBLE to spin in a slip” (before I get a thousand WTFs, when you stall in a slip the up-wing, the same side with rudder displaced, stalls/drops allowing time to recover. If you continue to hold the rudder after the wing drops, then you transition from slip into a skid). Always remember excess rudder into the down wing = bad! Act like a wing walker and step on the “Up” wing.

      What’s great about upset training for all pilots? It can reduce the startle factor. It can reinforce proper rudder technique. And it can reinforce the need to unload first and then roll (break the stall first. After all, stall speed at zero G is zero knots in all airplanes).

      • Buzz
        Buzz says:

        Having just completed Upset Recovery Training, I agree with Eric and I believe that the events of AF447 demonstrate the thesis behind such training quite well. FO was panicked and clutch reflex kicked in resulting in his holding the stick back.

        Mother Nature and our fight or flight reflexes are not always compatible with aviation. I agree with Eric that Upset Recover Training can help with recognizing the psycho-dynamics of the moment and taking actions appropriate to recovery which include the counterintuitive unloading of the elevator.

        Upset recovery training can be fun (was in my case) and I think has made me a better stick and rudder pilot in GA aircraft. I’m a lot more comfortable on the vertical plane looking at terra firma spinning towards me, but, that’s with the knowledge that I’m in an airplane that is equipped to do that. I’m not so certain I’d be comfortable doing it in a Phenom.

        Like any training, there are limits to utility some of which diminish over time with latency. Am I glad I took it? Yes. Is it a magic bullet? No? Does it all add up to making me a better pilot and more aware of my situation? I think so but at the end of the day, the objective is to avoid those circumstances which create the risk of an upset that has to be recovered from . If you fail in this regard, you increase your risks dramatically.

  7. Seth
    Seth says:

    Where to start…

    Since the Air France accident has been the focus of most of the comments, I guess I’ll start there. One of the many causal factors that came out of the accident investigation was a near complete lack of knowledge and understanding of the loss of reliable airspeed emergency checklist. Apparently this training (at the time) was only seen once during initial qualification training. This checklist was never called for, nor any of the steps taken. The loss of reliable airspeed is what led to the myriad of failure annunciations the pilots had to deal with when the computers lost a basic, but critically important input to determine and control the aircraft’s flight path.

    Another causal factor is the Airbus responds to the last stick input given, regardless of which stick gave the input. While the relief captain had been pushing forward to recover from the stall, the first officer in his panicked state had repeatedly input aft stick, keeping the aircraft in a stalled condition all the way down to the ocean. There’s no direct connection or feedback between the captain’s and first officer’s stick, so the relief captain had no idea his nose down input had been overridden by the first officer.

    The current statistics bear out that the avoidance style of training will not prevent loss of control-inflight (LOC-I) accidents from occurring. The vast majority of transport category accidents in recent history have been caused by stalling the aircraft (nearly 40%), followed by flight control and disorientation respectively (combined altogether, roughly 60% of all transport category accidents). As such, government agencies and industry got together and studied the problem at length.

    The new mantra is avoid, recognize, and recover. The recover piece had long been ignored as being unnecessary for highly trained professional pilots, with all the focus put on avoidance and some on recognition. But the statistics show in recent accidents that pilots were not recognizing upsets properly NOR taking the appropriate corrective action. This demonstrated in blood the avoidance and “properly trained” pilot model was not working and needed to be addressed.

    The Extended Envelope Training (EET) at my airline now consists of academics where you refresh on swept wing aircraft aerodynamics at low and high altitude, stall characteristics (low & high), upset recovery (low & high), and a review of several recent LOC-I accidents to discuss whether or not the crew took the proper action following the avoid, recognize, and recover model as a guide to the discussion.

    This is followed up with a lesson in a Level D simulator where you prebrief a shortened version of the academics to refresh and reinforce, focusing on how to implement of the recovery techniques discussed in class in our particular aircraft. You go into the sim with a partner, both getting hands-on training as the pilot flying utilizing the recovery techniques to proficiency in several configurations in the low and high altitude regimes to see the differences in aircraft response, then swapping midpoint so you also get exposure to the roll the pilot monitoring is supposed to play during an upset, supporting the pilot flying during the recovery. The AoA and G-loading is displayed for the pilot during the maneuvers, and the instructor shows the student the V/G plot to show them where their recovery fell in regard to the flight envelope, repeating recoveries as necessary to get them safely within the aircraft structural limits. This is fantastic and fun training.

    As much as I like Mr. Mac and some of the things he’s written over the years, this piece ignores the current accident statistics and opines the training is basically useless. There are examples where this training has saved lives, and data that shows when pilots are only trained to avoid upsets, they’re left untrained and unprepared when an upset occurs, resulting in accidents and deaths. Training to “avoid only“ doesn’t account for the fact that sometimes the environment deals you a bad situation no matter how professional you are as a pilot, and ignores the fact that professional pilots are humans that make mistakes that put them into an upset situations, without any training to recover. And that’s where Mr. Mac is patently wrong in his opinion.

    • John Zimmerman
      John Zimmerman says:

      I’m headed to FlightSafety this fall to complete their UPRT course in the PC-12 simulator (the airplane I fly most often). I’ll be interested to see how this goes. Here’s my bias: I generally agree with Mac that flying an L-39 or an Extra doesn’t do much for recovering from an upset in a big airplane. Fun, yes. Good for basic stick and rudder skills, yes. But not a lot of transfer to a Pilatus (or whatever).

      However, I’ve heard from multiple Pilatus pilots that doing a UPRT course in the simulator–a full motion sim that’s the same model as the airplane you fly, with actual test data from the manufacturer–is a different animal. Maybe UPRT isn’t good or bad by itself, it just depends on how it’s done.

      So I’m a skeptic but I’m cautiously optimistic about doing it in the sim. I’ll report back on this.

    Mac MCCLELLAN says:

    Every year during the recurrent training required by FAR 61.58 for pilots of airplanes that require a crew of more than one pilot, or are turbojet powered, we have done a form of upset training in the simulator.
    The way it usually works is the sim instructor tells you to cover your eyes and then he–or your sim partner–maneuvers the sim into an unusual attitude. The most effective instructors not only toss the sim around, they run the pitch trim. Many jets have a “clacker” that sounds when the trim is in motion so you can hear that, but you don’t know if the trim is running nose up, or nose down. And all of this is done in simulated IMC conditions, not visual.
    Then, after the airplane is in some extreme attitude, your told to look up and recover. Obviously, the airplane is going to be either gaining speed like crazy nose down in some extreme bank angle, or pointed up and very near a stall. And it is probably wildly out of trim so the stick forces can be confusing. So it’s pull or push, and chop power or add all of it immediately. And then try to get the wing level. Best to do all of that at once.
    The best instructors I’ve had in more than 40 years of doing this also dick around with the instruments in front of you. Used to be either the attitude indicator, or airspeed, but now it’s the PFD that they would fail. So when you look up you may see everything level, but with radical airspeed and vertical speed. Or, maybe you see nothing but red flags. Easier. Either way you need to look across the cockpit and compare the displays, and then look at the standby instruments to break a possible tie.
    I know that some airlines or training programs have taken the unusual attitude sim training to extremes, such as recovering from inverted or at least beyond 90 degree banks. Or maybe even a fully developed stall. I haven’t done those, only the unusual and rapidly changing attitude that required immediate action to recover.
    Whatever you think of sim training, I have found this part of recurrent training more realistic to transportation type flying than the aerobatic training I have done in aerobatic airplanes.
    Mac Mc

    • RichR
      RichR says:

      Done this both ways…military TACAIR in full motion simulator where large flight test $$ were spent to accurately emulate post stall/spin flight modes with training supplemented by routine TACAIR maneuvering flights (intentional stall/spins prohibited). Now operating on the shoestring end flying ex-mil primary trainer thru routine acro. Full motion sims are great for this, but I doubt that TACAIR type flight test programs have been applied to validating transport aircraft post stall or extreme maneuvering behaviors…so I’d assume commercial sims are an approximation in this regime, good for developing procedural emphasis, but may also have, like upset trainer airframes, significant behavioral inaccuracies. Whether you spend to get “Upset Training” or upset training with your local CFI or buddy with a C150, there’s a lot to be said for getting used to using entire control throw and experiencing post stall deck angles, breaks, and windscreens full of sky or ground to remove panic from the initial response. I routinely bring PICs flying who are uncomfortable in these situations, we brief, fly and learn/confirm the world does not end beyond 45 AOB or 20 pitch (these are on my dime, I’m not pushing product). With all the mandated hours now required for ATP, it certainly seems like an opportunity to take the time to add in more maneuvering training.

      On avoiding upset situations…with reliance on autopilots, when “George” runs out of airspeed/programming and PIC was distracted with laptop or pax, what kind of scenario comes next?…even the “responsible pilot” may end up here.

      In AF mishap, besides pilot performance, it highlights many man/machine interface issues that should be addressed or avoided in design stage.

  9. Enderson Rafael
    Enderson Rafael says:

    The AF447 is a very complex scenario. From lack of system understanding, including lack of procedures application (eg: unreliable airspeed memory items) to design flaws, it has much to do with the MAXs crashes actually (with a big load on training instead of design, opposite to what the media in general says). The industry incorporated the high altitude stalls training since then, focusing on regaining control instead of altitude. And as suggested, new airliners such as the B787 and A350, have the synthetic speed feature (AoA generated speed, GPS generated altitude), although primarily the memory items (pitch and power), remain the same. The speed info is gonna be used much later, after you’ve got it flying safely.

  10. Guy
    Guy says:

    The other thing that upset recovery training cannot replicate (whether conducted in an airplane or simulator) is the “startle factor”. Initial reactions to any situation are critical, but if the brain cannot rapidly analyse the problem, the aircraft may run out of altitude and/or airspeed before a solution is implemented.

  11. Pete
    Pete says:

    This is an interesting discussion.

    Having embarked on my journey as a UPRT Instructor at my airline I’d like to pitch in with some of reasons I believe what we are doing is valuable.

    From my research and learnings in our course the majority of Air transport category aircraft that have had a stalling event did not recognise they were actually stalled.

    That fact alone leads us down the road of appropriate startle effect training with the first point being Recognition.

    I mean seriously ……..what is the rush. A stick shaker is not actually a stall …….aircraft dependant it’s a percentage above the wing actually stalling.

    If you do nothing else other recognise the problem and take a breath you can at least over come the Startle and recover appropriately with enough height loss to keep the G load to a minimum.

    Another example of an upset is a Vmo over speed. I see pilots recover an over speed by chopping the power , pulling back on the stick to remove the clacker and get out of the red.

    Again ……what’s the rush. It’s Vmo not Vne. The wings won’t rip off and if you take a breath, figure out what you need to do and then again Gently do it.

    I see great value on changing the way Industry teaches upsets. This needs to happen at the highest level and more importantly at the beginning of any career in aviation.

    My thoughts for what it’s worth.

    • Mac McClellan
      Mac McClellan says:

      Hi Pete,
      You reinforce my point. Look for Vmo in an Extra? It’s not there. Upset recovery practice in a Level D sim is very useful. Flopping around in VMC in an aerobatic airplane fun, but not useful in transport flying.
      Mac Mc

  12. Bert
    Bert says:

    Mac’s last point is the most important: maintain safe margins. For example, flying the traffic pattern at 1.3 Vso, coordinated turns and a maximum bank angle of 30 degrees will make a stall/spin accident impossible in most of the airplanes we fly.

    • Warren Webb Jr
      Warren Webb Jr says:

      I know from flying with countless pilots of all ratings, including even other cfi’s and part 135 pilots, that there has been a massive failure to teach a simple pitch reference which is illustrated in Figure 3-8 on page 3-9 of the FAA Airplane Flying Handbook. The proper visual pitch reference is a point on the glareshield usually directly above the attitude indicator (in all cases it’s the point directly in front of the pilot’s or co-pilot’s nose). That point works exactly the same as the pitch dot on the attitude indicator staying vertically stationary when the aircraft is rolled – in a descent on base it’s about 8 inches below the horizon. Roll into the turn on final, keep that point the same 8 inches below the horizon, and let the banked lifting force of the wing turn the airplane in a normal manner. Instead what I think happens is the pilot (usually making a late aggressive turn) not only rolls aggressively but uses back pressure to increase the rate of turn while fixating toward the runway resulting in the stall. That point on the glareshield, if it had been observed, would have probably gone up to or above the horizon indicating the pilot lost awareness of the pitch position and simply stalled the airplane.

      I agree very much with maintaining safety margins, but limiting the bank angle, which I’ve heard other cfi’s recommend, may be problematic. It may not produce the turning performance desired resulting in the pilot using back pressure to tighten the turn which can lead directly to the stall. An appropriate pilot can do this test at a safe altitude. Establish a descent simulating configuration and speed normally used on base (approximately 70 kts in a C172). Note the distance of the glareshield pitch reference point below the horizon. Chop power and roll into a 60 degree bank in either direction making absolutely sure the pitch reference point stays at least the same distance below the horizon and that no airspeed is lost. Guaranteed there will be no stall. The C172R indicated stall speed with a 60 degree bank would be about 50 kts in level flight, even less in the descent, so there’s the good safety margin.

      I’m not saying any pilot can or should roll into just any bank angle base to final. My point is that generally the REAL PROBLEM is losing awareness of the pitch position (and airspeed I might add) and applying excessive or any back pressure which raises the nose from the proper descent position. If there is any undesired sink at any point of the descent do the normal – an increase in power as needed to hold airspeed with a simultaneous increase in pitch (ex: from -6 to -4 on the attitude indicator) as needed to hold glideslope.

  13. Erik
    Erik says:

    I can not speak to how airliners react in stall/spin/ upset situations having never flown one, or even a simulator for one.

    I will also state that an Extras or similar high performance aerobatic have absurdly quick control response.

    What I can speak to is that the biz jets and turbo prop aircraft I have flown all respond at least as quickly to control inputs as aerobatic trainers on the lower end of the spectrum such as Citabrias, Stearmans, etc. It seems to me that training using both low performance aircraft and advanced sims should provide the best all around upset training.

    When you push the nose up in an inverted recovery to upright and are hanging from your belts it is an very uncomfortable feeling. Practicing it will ingrain it.

    As for spin training, if you have not experienced what an airplane feels like right as it departs you will never be able to recognize it and prevent it from happening. I certainly agree that upset/spin training should not take place solely vfr, however partial panel work is a viable option in many trainers.

  14. ChrisBow Piloto
    ChrisBow Piloto says:

    I respect your experience and most of this article, but your assumption that a plot trained in upset recovery will not know what to do quickly after an upset due to negligence is completely wrong. I have experienced this in firefighter training and driving and I have friends who talk about this type of thing in the military. It becomes muscle memory and will happen like a reflex (assuming they have done it a few times, which may also be a false assumption). It is amazing how quickly a well trained human can go from being oblivious to a situation to being in complete control of it.

  15. Jack Stovall
    Jack Stovall says:

    Upset recovery practiced in a C150 Aerobat, Aerobatic Bonanza, and the like are excellent airplanes to get upset training because they are exactly like airplanes that people fly every day. Upset recovery is more about spin prevention than spin recovery and this can be practiced in these type airplanes more safely than in standard category airplanes. You can also legally put the student in many unusual attitudes including inverted and teach them how to recover. It’s not about going out and flying common aerobatic maneuvers but about showing how to get out of situations you might inadvertently find yourself in during your day to day flying. It makes you more confident and knowledgeable as a pilot just as IMC practice and landing practice do only more so. The key is to FLY AND STAY PROFICIENT!

    • Paul Bergman
      Paul Bergman says:

      I agree Jack! I did the upset training at BruceAir in Seattle in an Extra300L. Did I know in advance what to do? Yes! My surprise was just how easy it is to start a spin and how quickly it can flip over the plane.
      I now understand some of the areas that pilots are not trained in sufficiently to recognize/not get into the situation that has the potential to spin, i.e., just how close many pilots come to a departure stall or a landing stall by simply not understanding the ‘don’t s’ and recognize them early on before soloing.
      I am working on a presentation for pilots that is an eye opener and not really discussed that much, or from my particular viewpoint.
      I just got Rich Stowell’s book, Stall/Spin Awareness, but have not read past the first few pages just yet. I’ll see if he touches on the exact thinking I will use for prevention training & awareness.
      I am an Advanced Ground & Instrument Ground Instructor.

  16. Rodrigo Guerra
    Rodrigo Guerra says:

    Recovery from unusual attitudes may not be as efficient as we’d like, but it is no waste of time at all, since it develops confidence regarding where the limits of the airplane are and develops the sensitivity to perceive their closeness, while bringing prevention into focus, in real flying. It does not replace appropriate flight planning and does not compensate for recklessness in any level, so the main point for safety in aviation is still staying ahead of unforeseen events.

    I’m just ashamed of the seemingly superiority mr McClellan apparently thinks he has in relation to pilots who got involved, or even killed in accidents that may have been prevented somehow (maybe with an upset recovery training, but unquestionably Some further training). Disdain and contempt for those from whose accidents we try to learn from are at least objectionable. I hope you really never find yourself in any kind of upset, mister.

  17. Jeremy
    Jeremy says:

    The idea that upset training doesn’t work is absurd. I have two very close friends that are alive today because the training they received in extra 300 enabled them to recover a king air 300. It is true that many pilots will go an entire career without loosing control of an airplane. But if more pilots had upset recovery training then we would see less fatal accidents due to loss of control. I have personally seen pilots in a simulator loose control of a 737 in a rudder hard over on final as has killed many people. I have also seen the same situation with a pilot that has had upset training with no knowledge of the event about to happen safely recover the 737 sim after a rudder hard over. How many lives would upset recovery training have to save before you would feel that it worked?

      Mac MCCLELLAN says:

      Hi Jeremy,
      There is only one specific data point of how upset training played a role in an accident that I am aware of. And that’s not a positive one.
      It was the American Airbus A300 out of JFK in November of 2001. The airplane crashed because the vertical fin broke off. The NTSB found the probable cause of the fin failure to be the pilot flying’s aggressive and overuse of the rudder in response to the turbulence and minor upset caused by hitting the wake of a 747.
      The pilot had been through his airline’s upset training that emphasized use of rudder in recovery from unusual attitudes, something one would do in a light airplane or aerobatic airplane.
      What many of us learned from that investigation is that–despite what we had been taught all of our flying lives–full and abrupt use of flight controls can break the airframe even when flying below maneuvering speed, as this pilot was. The reason is that the certification rules deal with the loads of only a full application of the flight control, and then returning it to neutral. The pilot of the Airbus moved the rudder aggressively in both directions which took the structure outside of the certification load envelope and the fin broke off.
      Mac Mc

      • Dan Johnson
        Dan Johnson says:

        It also should be pointed out at the Airbus flight control system as designed provided essentially no stick force gradient to the rudder pedals. The mishap pilot was able to use full travel without using hardly any force.

        UPRT training would have had him unloading the airplane to one-half G, thus stopping any stall and providing powerful ailerons to roll the aircraft to a stable position. Any graduate of UPT or a basic acro course knows the relationship between control effectiveness and loading in both pitch and roll.

        I’ve got to say Mac, as an airline and GA guy I’m extremely disappointed you’ve chosen to preach that UPRT training is useless. It’s a spurious argument based on the flawed conclusion that because current UPRT training isn’t as good as it could be or wholly applicable to every UAS and LOCi, that it shouldn’t be undertaken.

        EVERY pilot should have some basic aerobatic and spin training. Even one hop is orders of magnitude better than no training.

  18. JWS
    JWS says:

    Note the progression in the article and the author’s response to it. First ‘upset training is trash’, then ‘upset training in aerobatic airplanes is trash’ to ‘upset training for multi-million dollar airplanes is trash’ to ‘upset training in million-dollar simulators for multi-million dollar airplanes might be okay’. No data were used to support his initial blanket declaration, in part perhaps, because saves as a result of such training don’t make news. Like the one which saved me.

    Unfortunately, this article is being widely distributed by Sporty’s eNews to be read mostly by normal people, not ATP rated turbine drivers. IMO, the potential for lost lives as a result cannot be dismissed.

  19. Andrew
    Andrew says:

    One of the items I have remembered from my upset training (and reinforced in instrument and commercial training) is to level the wings before pulling or pushing. Less chance of over g’ing and exceeding design limit margins.

  20. Bruce
    Bruce says:

    Heres a comment from a basic CPSEL pilot and upset training. Instead of my instructor putting the plane into a unusual attitude, he would have me under the hood, head down, eyes closed but still on the controls. In other words, I put the plane into an unusual attitude rather than him doing so. After a time, I’d be instructed to look up, recover and maintain normal flight. Sometimes, over that timeperiod, he would fail an instrument (the guy loved post it notes) and I’d have to recover partial panel without knowing it until I got my eyes back on the panel. Three lessons learned. Vertigo strikes quickly and brutally and most of the time YOU put the plane into an unusual attitude by YOUR inputs, not a second party aboard. And its a machine and machines have failures.

  21. Joe
    Joe says:

    Most interesting conversations. My two cents: AF447 lost airspeed and air data information. Was not attitude information fully available? The cockpit confusion overwhelmed them. The pilot flying failed to release the control. I always thought attitude plus power equals performance. It would seem all they had to do was put the airplane in level flight attitude and trim the engines for cruise power. UPSET training would have certainly ingrained better basic flying skills. I guess Monday morning stuff is always easy.. but the one thing missing was the 3 cell mag light to convince the guy flying to release the control.

  22. Sparky
    Sparky says:

    One of the first comments posted was that they didn’t think any airplane would recover from spins hands off. The T-34 and the T-2 both Navy trainers would both recover from a spin hands off.
    I did with both of those planes.
    The T-2 would also do upright and inverted spins and a Lomcovok. Fleet pilots in all the tactical airplanes, F-14 A-6 EA-6B A-7 F-18 would do a flight or two in the T-2 and practice spin recovery using the prescribed spin recover for the fleet airplane they flew and it always worked in the T-2. Experience from the fleet pilots who had this training did reduce the number of loss of control mishaps, but it didn’t eliminate them. startle factor and the loss of control too low (typically below 10,000′ AGL) to allow for a safe recovery.
    Perhaps the acro airplanes are too close to neutral stability to effect a hands of spin recovery. I suspect the spin certified Cessnas will recover hands off from a spin too.

    I don’t believe Transport Category airplanes are required to do a full stall even in certification so any simulator training is based on computer modeling of what is likely to happen and not real world flight test data. And if they did stall the plane in flight test any modeling of Incipient spin and post stall gyrations are merely best guess.
    A swept wing airplane is a whole different animal at the stall AOA.
    One of the Navy Planes I flew (EA-6B Prowler) was never successfully spun and recovered during flight test. The procedures had you with two hands neutral controls feet off the rudder pedals with stick slightly forward wait until you identify that you are in a fully developed spin and then the anti spin controls are opposite rudder, stick full aft and into the direction of the turn hold for 3 seconds and neutralize the controls with stick forward of neutral.
    It was all done on instruments and you had the benefit of an AOA display to see if you were below stall AOA.

    • RichR
      RichR says:

      …or maybe the secret was flying with (good) ECMOs knowing they’d whack you upside the helmet with their knee board! :)

  23. Pablo C. Branco
    Pablo C. Branco says:

    WOW! I am SHOCKED. This is probably the most ignorant article about flying I have ever read. Sadly and surprisingly, published on an outlet I (used to) respect. I am in awe that McClellan thinks this way being such an experienced aviator (or is he?) and that AirFacts would even publish this.

    • Pablo C. Branco
      Pablo C. Branco says:

      It is just not about “teaching people to recover from a spin” – this article is looking at this very narrowly. Upset Recovery training is about teaching people how to be comfortable in an airplane (ANY airplane!) in any attitude and know what to do from there. Arguments like “only pilots who made mistakes put themselves in upsets” are not only incorrect, but shortsighted. We all, at some point, made mistakes as pilots. Some of us knew how to correct from them. This article advocates that if you simply NEVER make a mistake, you will never need to recover from them – well, go ahead and throw the first stone if you’re a pilot and never made a mistake.

      The best remedy to create enough awareness to make better pilots at AVOIDING situations is to actually cross those lines in a safe and controlled environment that only UPRT can provide and that would be a major benefit for, for example, base to final stalls. It happens because pilots can’t identify it is coming, but what if they have seen it before? Then, they will recognize it. And avoid it. Some things just can’t be learnt from reading books and watching documentaries!

      I’m sad to say but this article advocates AGAINST safety in flying.

      Its not just about “teaching how to recover from a spin” – it is about teaching people how to fly better and be safer. If you haven’t been to an UPRT school, do it, then come read this article again and see if you still agree with it. I am most certainly happy I have; more than once!

  24. Kenny P
    Kenny P says:

    Yay, it’s my turn, and my opinion of why the training is of little value is somewhat different. I have a friend who spent thousands of dollars on tactical training with firearms. He wants to protect his family, and he learned all kinds of stuff—four years ago. He’s forgotten most of it, I’m sure, because he has no way to continually practice. And one cannot be vigilant every second of the day.
    My take on the whole deal (even though I carry a pistol) is that my family and I are better served by avoiding such situations where I may need my training. Same with my flying: avoid the situations that cause upset. As Mac says, it ain’t rocket surgery.
    There may come, in everyone’s life, a situation that cannot be foreseen. At that moment, your fate may be sealed, or you may have a slight opportunity to get the cards to fall in your favor. Me? I don’t worry about that situation. I’ll deal with it if and when it comes.
    For the record, I’ve spun a 150 and my Skyhawk numerous times. There is plenty of warning that needs to be ignored before either rolls over on its back.

  25. Jim
    Jim says:

    I took Rich Stowell’s “EMT” course – twice, actually – many years ago. In my opinion, it was VERY helpful. Two things stand out: firstly, if the airplane is grossly out of shape, I don’t freak out and freeze; I’ve seen that attitude before. It is hard to appreciate how traumatic it is for a pilot who’s never, say, spun to experience it the first time, and especially if unexpected. The second thing is somewhat specific: we (at altitude, of course) performed several turn-to-final-skidding upsets. The thing I remember most was not how violent it was, but how GENTLE. I was expecting the upset, but didn’t realize we were in “trouble” until we were spinning. The wing just gently dropped, and when I tried to pick it up with aileron, it dropped more, until we were in a spin. There is no recovery if you are on base-to-final; you are dead. Since that day, I’ve been a Nazi about having the ball centered in the pattern. I often fly with low-time pilots – and I always tell them, “If that ball is not centered in the pattern, I will yell at you. Count on it.”

    • J Dubb
      J Dubb says:

      I’ve often wondered why it would not be better to simply not use the rudder in the pattern. Then every turn will be a slip, which is inherently stall resistant

  26. George Catalano
    George Catalano says:

    McClellan – I am shocked you actually wrote this article. Maybe you were out of subjects matters to write about so figured why not kick the hornets nest…..???
    In the present aviation world we pump out hundreds of button pushing new pilots that can barely negotiate a cross wind, use their rudder pedals for a foot rests and get the sweats when they are passed a 30 degree bank. Basic aerobatic training and upset training can do nothing but good for these young and old pilots alike.
    Sure there is a disconnect between flying a T-34 upside down and a 737 – but if you have ever been to driving a school where they teach you how to get control of an ‘out of control’ car, you can connect the dots here. Upset training teaches you to understand aerodynamics better by feeling the edges of the flight envelopes. Seeing unusual attitudes in a controlled and safe environment so that the brain can process the visual and physical parameters of what is happening.
    I’ve been flying a long time and learned back in the mid 1970’s, so unfortunately I’m in the camp that would require tail wheel training, spin training and some basic aerobatics. I have taught many students aerobatics and everyone of them have had the same takeaway – it helps them be a better and safer pilot….end of story.

  27. Gennaro
    Gennaro says:

    If we don’t have a sound reason that explains why an airplane fly’s we certainly cannot expect pilots to agree on upset training. I lean toward agreement with Mac. For me the principal benifit of upset training would be that it is fun.

  28. atpfirefly
    atpfirefly says:

    I would suggest that this article is over reaching on the subject of UPRT. It certainly opens the door to many and varied opinions. The main thing I have found that relates more to upset training than any other is the fact that the pilot who goes through this type of training has a sort of epiphany in terms of encountering something other than a steep turn or spin. Spin training starts the process of instilling more comfort for a pilot in their confidence to control the aircraft and UPRT takes the process to a completely different level in terms of confidence in their own abilities to have “been there, done that”. Pilots are all different, but the one thing that seems to help immensely is simple exposure to the unusual attitudes associated with UPRT…it instills a level of confidence in the pilot that would otherwise not be there. I have done many sessions in Level D sims related to unusual attitude recovery, every session was an education relative to that specific type and how it can bite you and what you need to do to recover. After “Sully” landed on the Potomac, that gave the Sim instructors the idea to put everyone through a dual power failure in the corporate jets. It was an interesting exercise and made the pilots aware that a dual failure didn’t necessarily lead to a complete disaster. The comments related to “car control” training, more crosswind training, and tailwind training ALL have their places in the training environment. I participated in some ab initio training programs for some Airlines, some of those programs grabbed applicants off of college campuses and were “guaranteed jobs”….no aptitude required. These types of individuals in many cases didn’t even have driving licenses. For them, every bit of training they can be exposed to is badly needed. These particular pilots, in most cases, have never seen anything beyond a steep turn and they truly have “never been there, or done that” in terms of a real upset. Anything more than a steep turn and they are lost in terms of what to do or how to even approach a recovery. AF447 is in fact the perfect example, the co-pilot was holding full aft stick with no idea what he was doing. I personally was extremely fortunate to have towed and taught in gliders as I started my career. This meant that early on, I dealt with tailwheel aircraft (Super Cubs) in formation flying, cross winds, and many engineless landings. What I didn’t realize until later was that I distinctly had a leg up on aircraft control over other pilots who I now had the chance to compare myself against in two pilot cockpits. It was nothing I did specifically, but incidentally, to acquiring better aircraft control over many pilots I flew with. The main thrust of these comments is that since I was fortunate to have had that background, I was able to appreciate the fact that this type of flying and training definitely has it’s advantages. The article written here didn’t include comments from pilots who expressed their thoughts of their abilities before and after UPRT training. Those comparison comments would be where the true value of UPRT training lies I would think.

  29. Jim C.
    Jim C. says:

    I’m a single engine piston instrument rated and current pilot with 2100 hours. I’ve never flown a turbine and am not ever planning to, so I have to ask a pretty basic question. I understand that in some transport category aircraft one pilot can pull backward on his/her stick and the other can push full forward. The aircraft then “splits” the elevators, one up, one down. In my uninformed opinion, this seems like a knuckled headed way to design an airplane so there must be some reason why you’d want to twist the tail in some obscure condition. In addition to AAF447 I believe this was also a factor in the crash of an Egyptian airliner years ago where it was speculated one of the pilots was suicidal, fighting the inputs of the other pilot. Doesn’t it make more sense to couple the control yokes together so that you instantly know what the other pilot is trying to do and at least have a chance of figuring out if you have a panicked or suicidal individual working against your control inputs?
    And I’m willing to give Mac a pass here. Perhaps what he’s saying here is politically incorrect to aerobatic instructors and flight schools, but he’s not saying upset training is counterproductive. He’s just calling a spade a spade based on his flying experience and doesn’t feel the extra time and expense translates to a useful advantage in the real life aircraft you’ll be flying. And nope, in 2100 hours of IFR flying in real conditions with lotsa bumps I haven’t lost control. Yet. But I think I’ll be ready when it happens and am always thinking, “What if…?”

  30. Tim
    Tim says:

    This article shows to me the total lack of credibility of many aviation sites and the way many ‘pilots’ jump into the fray spewing ‘knowledge’ that is grossly incorrect.

    I am a proponent of UPRT and for many reasons. Pilot training today, on the whole, is very inadequate. They rush the students through and graduate them on the other side with barely minimal skills. UPRT should be taught if only to expose the ‘pilot’ to the entire envelope of a planes capability – and beyond – so that they are able to gain the self-confidence that they CAN recover the airplane should the unexpected happen. That can be done in a variety of aircraft and the more advanced aircraft a pilot seeks to fly, the more advanced the trainer should be.

    There are some misinformations stated about the Air France accident – yes the FO was flying and he was relatively inexperienced (as MANY are these days) yet the FO was (at least some of the time) following published and trained Airbus procedures – hold the side stick full aft and let the computers save you (One of the many things I do not like about the Airbus philosophy). At that point in time Airbus taught that the illustrious Airbus could not stall because of their innovative ‘laws’ of protection (what a joke). But they did not take their training so far as to teach that when the ‘laws’ all go away the pilot is left with ‘an airplane’ – FLY IT. But, alas, the pilots were dumbed down to believe the computer would save them. It didn’t. The captain, upon returning to the cockpit amidst the horror of the event, immediately recognized the situation and the errors of the FO flying… but too late.
    Yes, pilots need to learn the basic skills and stay out of the upsets in the first place. However, pilots should also be trained in UPRT to explore the entire envelope of an aircrafts capabilities to be proficient in controlling the aircraft whenever and for whatever is thrown at them – by themselves or because of someone else. It just makes sense.

  31. Jacques Perrault
    Jacques Perrault says:

    I will limit my comments to spin training. I was trained in Canada alomst 50 years ago, when incipient spin training was part of the curriculum. Not only did one understand how the airplane got into it (and what to do to get out of it) but the actual situation was extremely powerful to burn this to memory and act as a deterrent to avoid it.

  32. Stephen Phoenix
    Stephen Phoenix says:

    Actually, I think one of the bigger deficiencies in training, to prevent or recover from LOC, is in stall training. Early on we show and teach stall recoveries in something like a 172. The students learn and get their license and they are “safe from stalls”. They have seen stalls only at the forward edge of the CG envelope at light weight. Think the airplane handles a bit differently at max weight and aft CG? The time to find out is not after an engine failure turning crosswind in a loaded airplane. I think simulators are perhaps ideal for this type of training.

  33. John Raby
    John Raby says:

    I still remember Thrust and Attitude equals performance. In the AF accident both attitude displays and thrust read outs were available. Pretty simple really. Not enough focus on basic skills and maintaining a truly professional attitude to our work.

  34. Jack Morris
    Jack Morris says:

    This is part of a transcript from a PBS NOVA investigation of AF447. What this program suggests is that seasoned airline pilots should know about certain power and pitch combinations that will permit an airline without airspeed indicators to fly in a safe and stable attitude.

    The entire transcript is available at this link:

    What follows is an excerpt where they used a SIM to create the abnormal AF447 flight situation and then the simple solution to rectify it. Seems to me that every transport pilot should know this principle and when to use it.


    NARRATOR: As they edge around the storm, Alder triggers the critical moment of Flight 447: he fails all three airspeed indicators.

    SIMULATOR CO-PILOT: Okay, we have NAV ADR 1 fault. We have unreliable airspeed.

    NARRATOR: The automatic flight control systems shut down.

    SIMULATOR CO-PILOT: We’re flying with no auto-pilot or auto-thrust.

    SIMULATOR CAPTAIN: Okay. Autopilot’s off. I have control.

    SIMULATOR CO-PILOT: You have control.

    NARRATOR: If their actual airspeed rises or falls by as little as 10 knots, they could suffer a catastrophic loss of control. But the pilot uses standard procedures, learned in training. He moves the throttle levers to set thrust at exactly 85 percent.

    SIMULATOR CAPTAIN: And I’m selecting…I’ve got 85 percent set.

    NARRATOR: Then, he raises the elevators to pitch the nose up at precisely five degrees. With engines at 85 percent power, and five degrees upward pitch, the aircraft should always settle at the same safe speed.

    MARTIN ALDER: It’s quite possible to fly the aircraft to actually quite high standards of accuracy, something in the region of about five knots or so of the, of the desired target speed would be quite achievable for most crews.

    NARRATOR: Without airspeed indicators, pitch-and-power is a pilot’s lifeline. These pilots ignore any fault messages until they are safely in control.

    SIMULATOR CAPTAIN: I’ve got five degrees on the standby, four degrees on the other.

    MARTIN ALDER: Flying the airplane is the prime objective. No matter how attractive the messages might be to anybody on the flight deck, you both concentrate in ensuring that the person who should be flying the airplane is flying it, and flying it in the manner which is safe.

    SIMULATOR CAPTAIN: Okay, so we’ve got what I think is basic control of the attitude, we’re bumping a little bit with the weather, but, generally, that’s safe.

    NARRATOR: With no reliable airspeed indicators, maintaining pitch and power can get t
    them to the nearest airport. The emergency is over.




    UPRT was not now to me her in EUROPE although I obtained my FAA licences in 1969 in FRANKFURT USAF AIRBASE

    In my career always kept updatet and current via
    My annual requalification flight
    Annual recurrent AEROBATIC training with a CFI
    Annual UNUSUAL POSITIONS training
    Annual EMERGENCY revision training
    In a SEL activity en operation in CIVILIAN and MILITARY
    By the way a made a mistake on the mentioned airplane it is a
    SV4B or STAMPE

    congratulation on YOUR VERY INTRESTING READING

  36. John Raby
    John Raby says:

    Thanks Jack. Good description of the basics. Ironically light aircraft can be better situated as their basic autopilots can maintain pitch and or altitude by simple analog inputs not reliant on ADIRU type devices.

  37. Scott Glaser
    Scott Glaser says:

    This is article is factually incorrect. There is a large body of research that demonstrate that in aircraft URRT is the only effective solution to the short comings in modern pilot training that has led to the loss of control epidemic.

    I am Senior VP of Operations for Flight Research, Inc. (FRI) the leading provider of Jet URRT. I’ll be clear that my position on URRT is not a result of my being a part of FRI. I joined FRI because I believe in the company’s program and approach.

    Loss of control in flight is the #1 cause of aviation fatalities, worldwide, across all sectors of aviation. Again, URRT is the solution – as proven by data. There are a few references talking about the subject below. The key for a URRT course to be effective is that it must meet certain parameters. This article is in an opinion and not fact based. An effective URRT program addresses the questions mentioned in the above article.

    Aircraft Response
    The aircraft used for URRT must emulate the handling of the airplane the pilot flies operationally. If the pilot flies a business jet, the training must be done in an airplane that flies like a business jet. In our course we use a Sabreliner for the non-inverted work and Aermacchi Impala for the inverted work. The Saberliner is a swept wing business jet and emulates that aircraft that our customers fly. The Impala, while it looks like a fighter, does not fly like one. It has handling qualities and roll rates very similar to business jets. So, the article is correct that a business jet pilot should not do URRT in an Extra or Citabria. But URRT is effective in the correct airplane.

    Spins are an optional maneuver, however, there is high training value in spinning an airplane. It provides confidence in all the stages of handling the airplane up to the spin. It helps the pilots understand what the progression of events leading to a departure look like so if they should start to develop in real life, they will be recognized.

    Wake Encounters Only Happen Too Close to the Ground For Training to Matter
    First – there was an accident in the last year where a Challenger flew through the wake of A380 at cruise. The airplane rolled 3+ times, was totaled and there were several injuries. So no – wakes don’t only happen close to the ground. They do also happen on departure and approach, however, the altitudes are typically in the low thousands where there is in fact plenty of altitude to recover. Proper recoveries can be completed with little or no altitude loss – even in a business jet.

    The crew had bad data and the aircraft was not responding as expected. This is covered in URRT. When things go bad, unload the airplane. Had they do so, we wouldn’t be having this discussion.

    Simply training avoidance of a risk in any safety item is not an acceptable solution. URRT covers prevention and prevention should be practiced at all costs. However, many accidents, AF447 is one, were not avoidable by the crew. Upsets will happen. Modern civil pilot training does not provide pilots the skills to recovery. Per a significant body of scientific research, effective URRT is the answer.

    Richard A. Leland, Rodney O. Rogers, Ph.D., Albert Boquet, Ph.D., Scott T. Glaser “An Experiment to
    Evaluate Transfer of Upset Recovery Training Conducted Using Two Different Flight Simulation Devices”,
    Environmental Tectonics Corporation, 2009.

    Richard A. Leland. “Spatial Orientation Training for Aircraft Operators: A Pilot’s Workbook for Spatial
    Disorientation Training.”. Environmental Tectonics Corporation.

    Gawron, J. Valerie “Airplane Upset Training Evaluation Report”, NASA/CR-2002-211405, National Aeronautics and Space Administration, May 2002

    Go, Tiauw H. Go, Burki-Cohen, Judith, Soja, Nancy N. “The Effect of Simulator Motion on Pilot Training and Evaluation”, AIAA 2000-4296


    Loss of Control In-flight (LOC-I) is the leading cause of fatalities in every sector of aviation according to the NTSB and a diversity of other informed entities to include ICAO, the FAA, and the Commercial Aviation Safety Team (CAST). For commercial airlines worldwide, for example, LOC-I now accounts for over 49% of all fatalities. This relative contribution has increased in relative severity to an all-time high over the last seven years of this report.

    In recent NTSB findings over the last 10 years between 2009 and 2018 narrowed down to the professional pilot sectors of aviation, including Part 135 and executive/corporate flight operations, LOC-I tops both the number of fatalities and the number fatal accidents. These are hardly ‘careless’ and ‘poorly trained’ pilots.

    Upset Prevention and Recovery Training (UPRT) is consistently supported by aviation industry organizations. In-flight UPRT is now recommended by ICAO prior to Commercial licensing, is currently required by EASA, and will be soon by ten countries in Asia. It is recommended by the International Air Transport Association (IATA) for airlines, is required for all ATP pilots as part of the ATP Certification Training Program, and for all airline pilots in the US beginning in 2020.

    Over three decades of firsthand experience delivering UPRT solutions to airlines, government agencies, corporate flight departments, insurance agencies, and flight schools provides Aviation Performance Solutions ( with a basis for clarification of some basic facts about UPRT and LOC-I prevention that can establish a foundation from which a more robust and insightful discussion can be presented. Because taking solution-oriented steps to reduce the risk of LOC-I is essential, we need to address the numerous misconceptions presented in this article individually.

    This subject is far too important and extensive to be addressed comprehensively in this remarks section. We have prepared an account of 14 high level topics that encompass misconceptions or incorrect statements made in this article. This commentary can be found in the article “14 Issues with ‘Why Upset Training Just Doesn’t Work’ Published by Air Facts Journal” found on the APS website listed above.

    Here is a summary of the topics addressed:
    1. Recovery Empowers Prevention
    2. Regulatory and Industry Acceptance of UPRT
    3. Aerobatics are not UPRT
    4. Spin Prevention is Key
    5. Instrument Recoveries in UPRT
    6. It’s Not About Fun
    7. Transfer of Skill
    8. Simulator Limitations in UPRT
    9. Human Factors and Air France 447
    10. Technological “Silver Bullets”
    11. Airplanes vs. Simulators for UPRT
    12. Wake Vortex Encounters
    13. UPRT Instructor Qualification
    14. The False Premise of Existing Flight Training

    The root cause of the persistence of LOC-I as the #1 fatal threat in aviation is that existing pilot licensing training does not prepare pilots adequately for the differences they can encounter aerodynamically, psychologically, or physiologically beyond the normal flight envelope. Regardless of the technology employed, the flight crew always remains the last line of defense for the control of an aircraft’s flight path. Simply put, the ‘been there, seen that’ value of UPRT is just the beginning of the development of the resilience, discipline, and cognitive agility in crisis required to both arrest an escalating airplane upset, and recover if necessary.

    Randall Brooks, VP Training – Aviation Performance Solutions

  39. Steve Stowe
    Steve Stowe says:

    I’ll argue that UPRT does work. Even though I started as a fighter pilot and am a USAF Test Pilot School graduate with many real life stall and spin tests under my belt, I’ve found civilian UPRT to provide transferable skills to the airplanes I’ve flown and tested commercially. The key is to use a reliable UPRT provider. As an airline pilot and later as a OEM test pilot of airliners and business jets, I’ve evaluated a number of UPRT providers for various flight departments, including the National Test Pilot School’s UPRT program (Impala), the joint FAA & NASA UPRT research program in Calspan’s (then Veridian) vari-stab Learjet and Bonanza F-33, the APS program (Extra 300 and S211), and also with other OEM test pilots developed some Level D sim UPRT exercises for ICAO’s AUPRTA Rev3.

    One of the biggest misconceptions in this article is UPRT “wont do any good in real life business jet flying.” For me, it’s been quite the opposite. While some procedure steps may be type specific, most of the skills UPRT provides will transfer to “real life flying.” For example, unloading (reducing AoA to regain control) and using bank angle (nose high) to tilt the lift vector to minimize airspeed loss works in any airplane.

    With over 50 years of flying (received the FAA Master Pilot Award last year) and in my day to day world of stall testing and other tests that put the jet in extremes of attitudes and corners of the envelope, I often use those skill sets taught in UPRT. On stalls, we had years of negative training in commercial sim rating courses and recurrents when the FAA insisted that stall recoveries (actually stick shaker recovery) should be done in level flight using thrust alone. Try that in your business jet at cruise altitude and see what happens. Finally, they are now teaching to unload first, as taught in UPRT.

    UPRT provides pilots with instinctive skills that take over during the “startle factor.” The article says a panicked pilot won’t remember training, but that is exactly when training will instinctively kick in, almost automatically even though startled out of your mind. And if you are a commercial pilot, your employer’s insurance premiums may even be reduced if the pilot cadre has completed UPRT.

    My recommendation to our industry on UPRT – take it, and take it seriously.

    Steve Stowe
    President, Aircraft Certification Experts, Inc.

  40. Geoffrey W McCarthy MD DipAvMed
    Geoffrey W McCarthy MD DipAvMed says:

    Mac, I always read your columns, but this is an egregious and dangerous message: “Just don’t make a mistake, and you will never need to recover from an unusual attitude”. For the record, please respond with your aerobatic experience. More importantly, please tell readers your experience level in **actual** swept wing stalls. I will bet you have never intentionally, or accidentally, stalled a swept wing jet, mostly because the FAA and manufacturers will not let you. Full disclosure from me: I was taught to stall an USAF F-100 Super Sabre, mostly to experience the utter lack of stall break, and to experience the effectiveness of the rudder: you can maintain a perfectly upright attitude while descending 4-5000ft/min. Later, flying for the RAF, I stalled…and spun…the excellent BAE Hawk. Same characteristics. Plus, as co-pilot, with a test pilot in Command, i experienced some buffet in a DeHavilland Comet MkIV, a better warning than the Hun or Hawk. The pilots in AF447 had never experienced a swept wing stall, had minimal if any UPRT, and everybody died.
    Worse, you do not appreciate the *primary, not secondary,* result – my view here -of UPRT and sim training: Reduction of “arousal” when the real situation is encountered. Yes pilots panic when confronted with a life-threatening situation, but if experienced in the sim multiple times, are much more likely to maintain some ability to recall and think and perform.
    So, to sum up: you advocate cancelling UPRT, as mandated by EASA, and advocated by FAA, because pilots just need to follow the rules, and be better trained. You, like the NTSB willfully ignore the reality of human performance under stress. I am a HF expert and also Patient Safety expert. I shall take your message to your local hospital, and advocate that they stop all emergency simulation: after all, your docs should just follow the rules, no?

  41. Brian Lloyd
    Brian Lloyd says:

    Two years ago, while flying around the world in my Mooney, following Amelia Earhart’s route, I was approaching Yangon in Myanmar. Yangon approach gave me a vector in a direction I wasn’t really wild about going but since neither their radar nor my stormscope saw anything, I accepted the vector in the interests of expediency. 30 seconds after entering the cumulus I was upset. An immature CB had turned into a mature CB. The aircraft rolled approximately 130 degrees left and ended up about 30 degrees nose-down as well. Since I practice AND teach UPRT my response was immediate and instinctive. Unload, roll, recover back to approximately level, and get the h–l out of there.

    The bottom line here is: Upset prevention and recover training works, even in small GA airplanes. I have experienced an extreme turbulence upset, responded as per training, and survived to tell you about it. In fact, as it was happening, my response was so instinctive that I felt almost like an observer as my hands and feet did the right thing. I even said to myself, “That really works!”

    There is only one point I agree with you on — that on the basis of one UPRT training session you are going to do the right thing 5 years hence? Probably not. Just like we practice landings, approaches, and other fun flying stuff, we have to practice UPRT too. I practice every week. My UPRT responses are always fresh and new. The success of my upset event shows that I was proficient. But I work to maintain that proficiency, just like I do shooting approaches, doing holds, stalls, steep turns, slow flight, and basic aerobatics.

    But bottom line Mac? UPRT works. I am living (important word there) proof that it does.

  42. Peter E Rossi
    Peter E Rossi says:

    I am only a rank amateur with about 750 hours in SEL. After getting my private pilot cert, I enrolled in the EMT class at CP Aviation in Santa Paula, CA. I quickly learned that my stick and rudder skills were very limited. I learned many things other than spin recovery. I learned not to be afraid of full deflection of controls. I learned how to fly with simulated control failures. All landings were power off 180s with no flaps. I learned so much about rudders and coordinated flight. We also flew my DA40 very much on the feathered edge. I learned to slip so much better. I gained huge confidence. Just last week there was a 30 knot shear from headwind to tailwind on final at KVNY. I would have been in a lot of trouble without that training. I very much believe that this kind of training could very well be the difference between life and death. I’m plan on more training and aerobatics. Why was Bob Hoover considered the best and safest pilot who ever lived? Because his whole career was upset training on steroids. It’s not how many hours you have, but how you’ve spent those hours. The hours I spent in EMT are the most valuable, by far, of my training. When I sit on a commercial flight, I hope that the flying pilot has had training like this.

  43. Gary Graebe
    Gary Graebe says:

    In most cases the use of references (VFR) outside the aircraft will prevent or avoid LOCI..Not enrolling in UPRT probably won’t kill you ,but why take the chance. On the other hand,if you frequently fly in IMC darkness and weather UPRT may be beneficial. Given the seriousness of the threat,UPRT will provide simple priorities and exit strategies for survival. Like the old lady said about chicken soup “It could’t Hurt”.

  44. Ron Rapp
    Ron Rapp says:

    Though I’ve been a fan of Mac’s columns for many years, it does seem that he’s shanked the ball pretty badly on this one.

    I wrote a post for AOPA’s Opinion Leaders blog about six years ago which pointed out a number of reasons why upset recovery training works and is imperative for aviators. So much the better if it can be accomplished “in type”, which incidentally addresses one of Mac’s major criticisms of such training.

    That used to be rare, but it’s becoming far more common. On my two most recent type ratings, both programs had optional type-specific upset recovery courses for their Level D simulators which were informative and eye opening. The G650, for example, gives you a fairly violent half snap roll when stalled. We were given opportunities to see upsets at both low (1500 AGL) and high altitudes (FL550). I thought the training was excellent.

    I personally have experienced a flight control failure in the middle of an IAC aerobatic sequence, automation failures, weather related upsets (Banning Pass, anyone?), and more than one wake turbulence encounter — the most recent of which was on an arrival procedure into LAX at about 7000 feet. I lived through them all. Would I have faired so well without the training? It’s impossible to say with absolute certainty, but I tend to think not.

    For me, at least, upset recovery is the most important training I’ve had.


Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *