New (2010) Stall Recovery's @ high altitudes [Archive]

PBL

13th Jun 2010, 11:29

Yes, let's indeed not confuse stall with approach to stall. Here is an example of such confusion:
and the approach to stall wasnt what i meant though,..
the part i meant was where you were past stick shaker and didnt correct accordingly... point where stick pusher kicked in.

If you have a stick pusher on your airplane, then you are doing approach-to-stall recovery, because the pusher is there to stop you stalling the airplane, either because somebody didn't like what they saw in the wind tunnel, or somebody else decided there wasn't enough data about it.

Do not mistake the procedure with approach to stall, this "new" procedure is a remainder of the stall recovery, .....
There is an FOT from Airbus out after extensive Flight and Sim Tests

You think people have been going out doing stalls in Airbuses to see what happens?

The aerodynamic characteristics of a modern commercial jet are determined in wind tunnels, mostly for certification purposes. The wind tunnel work determines how the wing behaves, not the whole airplane. It determines at what point lift drops off sharply, and so forth. There is a lot of data, but there is by no means everything. The point of "stall" is not necessarily defined as being where the lift drops off sharply; it is more often defined by a degree of buffet (I refer to the regs).

"Sim tests" cannot help anyone with stalls. Simulators are only veridical at most up to the point of defined-stall (which may be at lower AoA than lift-break). So anything you can do in a sim is an approach-to-stall procedure.

As far as I know, this is a result of the recommendations following the Colgan Air crash, during which the PF reacted inappropriately at approach-to-stall, at stick pusher.

PBL

Tankengine

13th Jun 2010, 12:58

I hate to state facts here!:E
but
There is a new procedure from Airbus for the 330 [newly arrived at my airline anyway]

It basically states to lower the nose [with a note that this may need a reduction in thrust] then start to smoothly increase thrust while accelerating and to go flaps1 [LE devices] if below 20000' [sorry BOAC:ouch:]

Underslung engines may create more pitch up at low speeds/high AOA than elevator can pitch down!:eek:

Centaurus

13th Jun 2010, 13:05

and the reaction of the FAA trained FO was to try and use full rudder to lift the wing.

FAA trained or not, it is a fact that using rudder to lift a wing is commonly taught in flying schools in Australia. Judging also by the questions that appear in the Pprune Instructor Forum, it happens elsewhere as well. It seems to be handed down from generation to generation of young flying instructors who in turn once they achieve Grade One instructor status teach new instructors.

It has proved impossible to completely eradicate this technique throughout many general aviation flying schools. But you would think when type rating courses are conducted this erroneous teaching would soon be discovered and rectified.

FullWings

13th Jun 2010, 13:17

I think guiones has it right. Reality is just being restated. As you fly higher, the reduction in max. thrust available means that the speed at which you "fall behind the drag curve" increases, possibly to well above the stalling speed. At that point you have no choice but to descend (the aeroplane will do it for you), either by maintaining the same IAS and waiting for the thrust to come back :ooh: or by reducing pitch (and drag) and trading some potential energy for kinetic (height vs. speed) to get back earlier to a thrust>drag regime.

At progressively higher altitudes, stall recovery moves steadily towards the technique you would use in an unpowered aircraft, compared with lower down where minimum height loss is more important. The training doing the rounds at the moment may be because previous exercises concentrated more on low altitude mishaps?

guiones

13th Jun 2010, 14:40

PBL, that is exactly what I am saying.

Airbus EV (Flight Test) did extensive stall behavior test on actual aircraft and then it compared it to the Sim behavior before they published the FOT.

G

guiones

13th Jun 2010, 14:42

Tankengine, that is the FOT I mentioned in my posts.

The reduction in power is "you may" if you need it to regain or maintain pitch control.

G

Mansfield

13th Jun 2010, 14:54

AC 25-7, Flight Test Guide For Certification of Transport Category Airplanes, Chapter 8, Para 228(b) states the following:

Some airplanes require artificial stall warning systems, such as stick shakers, to compensate for a lack of clearly identifiable natural stall warning to show compliance with the stall warning requirements of § 25.207. Similarly, some airplanes require a stall identification device or system (e.g., stick pusher, automatic inboard slat segment retraction, auto-trim, etc.) to compensate for an inability to meet the stalling definitions of § 25.201 or the stall characteristics requirements of § 25.203.

The stick pusher does not function as a stall warning device; that is the purpose of the stick shaker. The pusher is a stall identification device. It represents exactly the same thing as CLmax, and is used when CLmax either cannot be clearly identified or when the stall characteristics around CLmax are particularly unsatisfactory.

With respect to stall testing, the manufacturer does indeed conduct stall testing in flight, including at altitude. I have appended additional excerpts from AC 25-7A below. This language may give a better understanding of the issue from the authorities point of view.

I suspect that any new material on this topic from the manufacturers is a result of a rather extensive effort being put forth in the industry to understand loss of control accidents. I believe it would be a mistake to assume that it originates solely with the Buffalo accident.

AC 25-7A, Chapter 2, Section 6 – Stalls.

b. Explanation.

(1) The purpose of stall testing is threefold:

(i) To define the minimum inflight airspeeds and how they vary with weight, altitude, and airplane configuration (stall speeds).

(ii) To demonstrate that handling qualities are adequate to allow a safe recovery from the highest angle of attack attainable in normal flight (stall characteristics).

(iii) To determine that there is adequate pre-stall warning (either aerodynamic or artificial) to allow the pilot time to recover from any probable high angle of attack condition without inadvertently stalling the airplane.

(2) During this testing, the angle of attack should be increased at least to the point where the following two conditions are satisfied:

(i) Attainment of an angle of attack measurably greater than that for maximum lift, except when the stall is defined by a stall identification device (e.g., stick pusher).

(ii) Clear indication to the pilot through the inherent flight characteristics or stall identification device (e.g., stick pusher) that the airplane is stalled.

(3) The airplane is considered to be fully stalled when any one or a combination of the characteristics listed below occurs to give the pilot a clear and distinctive indication that he should stop any further increase in angle of attack, at which time recovery should be initiated using normal techniques. The stall speed is defined as the minimum speed reached during the maneuver, except for those airplanes that require stall identification devices (see paragraph 29b(3)(iv), below).

(i) The pitch control reaches the aft stop and is held full aft for two seconds, or until the pitch attitude stops increasing, whichever occurs later. In the case of turning flight stalls, recovery may be initiated once the pitch control reaches the aft stop when accompanied by a rolling motion that is not immediately controllable (provided the rolling motion complies with § 25.203(c)).

(ii) An uncommanded, distinctive and easily recognizable nose down pitch that cannot be readily arrested. This nose down pitch may be accompanied by a rolling motion that is not immediately controllable, provided that the rolling motion complies with § 25.203(b) or (c), as appropriate.

(iii) The airplane demonstrates an unmistakable, inherent aerodynamic warning of a magnitude and severity that is a strong and effective deterrent to further speed reduction. This deterrent level of aerodynamic warning (i.e., buffet) must be of a much greater magnitude than the initial buffet ordinarily associated with stall warning. An example is a large transport airplane that exhibits “deterrent buffet” with flaps up and is characterized
by an intensity that inhibits reading cockpit instruments and would require a strong determined effort by the pilot to increase the angle of attack any further.

(iv) The activation point of a stall identification device that is a strong and effective deterrent to further speed reduction. Paragraph 228 of this AC presents guidance material for demonstrating compliance of stall identification systems with the regulatory requirements of Part 25 of the FAR.

BOAC

13th Jun 2010, 15:02

and to go flaps1 [LE devices] if below 20000' [sorry BOAChttp://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/shiner.gif]
- no apologies needed my reply was to a post about high-level stalls where deploying LE devices will probably put you on your back.

Centaurus - let's not confuse folk here please! The use of rudder to lift a stalled wing is ESSENTIAL and correct and MUST be taught. Use of aileron on a fully stalled wing will merely increase the wing drop (basic aerodynamics?) and will probably cause a spin.

Above the stall (ie stick shake) it is fine to use aileron.

s_bakmeijer

13th Jun 2010, 17:19

great opinions everywhere and also experience of other aircraft,.. what i am currently flying is an md80. as where we have st.shaker and st.pusher.
where the shaker kicks is, you're approaching the stall, and the pusher kicks in where the wings stall.
also when i mentioned lift improvement devices,, that also 'slats' not flaps. these have certains speed limitations aswell, which i guess at the point of stall the current ias/mach will be below the flaps limitations, didnt double check that very accuratly in the sim, but at altitude, the simulator recreated a full stall between 160/170 KIAS (mach: didnt notice)

but i mean, it's awful,... you will get so many warnings b4 a stall, and still it happens.
my main reason for this thread.
what i could see in the simulator... the md88 (md80 series) will give you these warnings:
-airspeed drop (obvious)
-210KIAS -> Landing gear aural warning
-210KIAS -> Langing gear 3 red lights
-FMA-> Low SPD
-<200KIAS Buffeting starts very subtile, but noticeable and ever increasing
-stick shaker
-stick pusher

our demonstration we being conducted at FL370 (max md80) with AT off, clean config.

1.recovery at st.shaker keeping AP ON and firewall throttles took about 40seconds, 4000' alt loss

2.recovery at st.shaker w AP OFF trying to maintain ALT with throttles to firewall took also 40seconds, 2000' alt loss

3. recovery at st.shaker w AP OFF deliberately lowering nose to horizon immediately and THROTTLE REMAIN IDLE took 4seconds, alt loss was +/- 2000'

4. recovery at st.shaker w AP OFF, lowering nose to horizon Throttle to firewall took also 4 seconds, and alt loss was >1000'

----
all in all a great eye opener. we knew that the sim couldn't simulate a complete stall accurately but, for training purposes this was a very very great test ! ( i'm <1yr on jet)

OPINIONS pls.

120class

13th Jun 2010, 17:27

One should read John Farley's book especially the chapters on stalling and stall recovery technique.

PBL

13th Jun 2010, 18:11

I find it hard to believe that anyone with an understanding of aerodynamics would disagree with anything I said, but it seems to have happened. Sorry.

If there is a pusher, the pusher will kick in at a determined point. You can take that, if you like, to be a definition of "stall", namely the point at which the engineers decided the pusher should activate. It is not necessarily the same thing as CLmax, as Mansfield suggested it was.

Mansfield also quotes the certification regulations, to show that flight tests are performed at various combinations of parameters to a defined point of stall. Well, yes indeed. Those have all been done during certification of the airplanes we are talking about.

So what data is it that a given manufacturer didn't have, that they would have gone back up to get because the regulators were worried again about LOC? Can anybody here say?

On to semantics. If an airplane is equipped with a stick pusher, then, granted if you will, a stick pusher activates at the point of defined "stall". That is, you get slower and slower, up to a specific airspeed at which stick pusher activates, and over you go. At what point in that entire process can you be said to be "stalled"? A suggestion: at no point, or maybe at just one. So, if I may be permitted to insist on accuracy, you are performing a manoeuvre which can be best described as recovery from approach to stall, not as recovery from a stalled state. The pusher does not let you get into a stalled state. Isn't this obvious?

An analogy. You walk up to the edge of a lake with shallow banks. You let your shoes go right up to the edge of the water. The water touches the very tips of your shoes, but no more. Are you in the lake? Are you wet? Have you been wetted? If you step back, are you recovering yourself to dry land from being in the lake? Answer as you will. I would say, you went up to the edge of the lake but you didn't recover from being in it. Similarly with a stick pusher and stall. You went up to the point of stall but you're not recovering a stall.

I am happy to agree that the current action is a general response to recent LOC accidents if that wording seems more appropriate to some.

PBL

guiones

13th Jun 2010, 20:53

PBL, this is from your post:

"You think people have been going out doing stalls in Airbuses to see what happens?

The aerodynamic characteristics of a modern commercial jet are determined in wind tunnels, mostly for certification purposes. The wind tunnel work determines how the wing behaves, not the whole airplane. It determines at what point lift drops off sharply, and so forth. There is a lot of data, but there is by no means everything. The point of "stall" is not necessarily defined as being where the lift drops off sharply; it is more often defined by a degree of buffet (I refer to the regs).

"Sim tests" cannot help anyone with stalls. Simulators are only veridical at most up to the point of defined-stall (which may be at lower AoA than lift-break). So anything you can do in a sim is an approach-to-stall procedure."

Yes indeed, Airbus in this case thought it was worthwhile for writing the FOT to conduct additional many stall tests on the aircraft and compare them to simulator behavior, not only rely on certification data. At the same time it has all the data required for its operators to simulate stall recovery in the simulator, this will not be sim tests as you state it but training on stall recovery.

PBL

14th Jun 2010, 03:24

Giuones and others,

I seem to have trouble explaining myself and getting my points across. Since my career is partly based on explaining myself and getting my points across, I find this situation somewhat puzzling. I have met it before, often, on PPRuNe, which is partly why I no longer contribute on a regular basis.

I have tried before the strategy of writing long, pedantic pieces to try to get my points across. It doesn't always work, but let me try it just once here. If this doesn't work, I'll just stop, for I don't know any other ways of trying.

s bakmeijer was talking about recovery from stall in an MD-80 series jet. Now, I don't know anyone who worked on the MD-80 wing, but I do believe I can guess how they thought and worked in general terms. I do work on a regular basis with an aerodynamicist who was responsible for the wings on some other well-known commercial jets. And I am not yet sure that everyone here is thinking the issue through. So, at the risk of being pedantic, let me tell a story which I think is very likely to be true of, say, the MD-80-series wing design process.

There you are, designing a wing, looking at the wind tunnel experiments, and you notice that, in particular circumstances, when the airstream separates from the upper surface and the lift drops off (it may do so suddenly or gradually; I don't know the story for the MD-80-series), the turbulent wash blanks out the tail (or flows over the place where the tail should be; it may be that your wind tunnel model doesn't have a tail attached).

Now, you and everyone else knows that you cannot certify the aircraft if this is allowed to happen in real life, for it is a certification requirement that it not do so.

So you put in a stick pusher. When you do so, you have to choose at what point you design it to activate. Do you design it to activate at the point at which the airstream separates? Well, that would be unwise, indeed useless, because at that point, as you have seen in the wind tunnel, the tail is blanked, and that is exactly what you are trying to avoid. You design it to activate *just before* that point, so that the separation does not occur, the tail is not in danger of being blanked, and you and your DI can go home and sleep comfortably.

Mostly, lift increases up to the range at which that potentially-blanking airflow separation occurs. You are triggering the stick pusher just before it gets to that point, so you are not triggering it "at" CL_max in physical terms, but just before that point.

And maybe you tell your certification agency: we aerodynamicists define the "point of stall" to be the point at which we trigger the pusher. Because there is some leeway allowed to you to do so if you wish. The stall phenomenon is actually a range of aerodynamic behavior over a range of airspeed and you can set the point for definitional purposes more or less wherever you choose to do so in this range.

Years later, some pilots, say, MD-80 pilots, say that stick pusher activates "at the stall" and call what happens "stall recovery", because that is what everyone else in their community says. So be it. But if you want to talk about aerodynamic phenomena, as I often do, this phraseology does not accurately represent what is going on.

Guiones, I know that flight tests were done and sims data was validated for this whole endeavor. But I believe that it misdescribes what was done to say that the airplane was stalled and recovered and the sims were modified to reflect that behavior. I think you will find that the airplane was taken up to the point at which unpleasant handling characteristics started to occur (maybe up to the engineering-department-defined "point of stall" - which may or may not be at the point of maximum lift, depending on your airplane - and maybe to just before, but not over that point, since going over it is possibly dangerous and people who do these tests generally do not want to kill themselves, and their bosses don't want them to do it either). And that simulator behavior was validated up to that point. But sim behavior may not be trusted after this break, for the simple reason that nobody has flown the airplane after this break to know what it really does then.

And you can ask an aerodynamicist responsible for a specific wing, about behavior post-stall: "what happens here?" And, depending on what you are asking about, the answer may very well be "I don't think anybody knows". I know this because it happens to me a lot.

Can we agree now? I am not going to belabor it further, because if this way of putting things does not suffice to explain where I am coming from, I don't know what I can say that will.

PBL

Pugilistic Animus

14th Jun 2010, 04:40

FAA trained or not, it is a fact that using rudder to lift a wing is commonly taught in flying schools

the use of the rudder to lift the wing is required if you are flying an airplane without boundary layer control non-frise-type ailerons or if you are an aerobatic student

YouTube - Falling Leaf - Instructor Demo (http://www.youtube.com/watch?v=ZOPsQn2Mksg)

none of that applies to transport aircraft

mad_jock

14th Jun 2010, 12:43

let's not confuse folk here please! The use of rudder to lift a stalled wing is ESSENTIAL and correct and MUST be taught. Use of aileron on a fully stalled wing will merely increase the wing drop (basic aerodynamics?) and will probably cause a spin.

Current spam can training and also what I have been taught in turboprops is that rudder should be used to hold the current roll attitude. Then the wings rolled level again after the aircraft is unstalled, as per a stall in a turn. The rudder should be used to hold the wing from dropping further but its not to be used to lift it back to level.

I will standby to be corrected though.

Mansfield

14th Jun 2010, 12:59

Perhaps a better way to come at this is to refer to Mad (Flt) Scientist's excellent explanation of the pusher concept back in 2007:

http://www.pprune.org/tech-log/270567-b747-400-no-stick-pusher.html#post3213867

PBL, I believe the disagreement lies not with the point at which a stick pusher is triggered, but with the premise that it be interpreted as an approach to stall. The manufacturers and certification authorities intend for the pilot to interpret the stick pusher as an indication that a stall has occurred. Now, aerodynamically, the pusher may operate before the actual stall is reached, however it is defined, but this seems to me to be a technical point and not a practical one from the pilot's viewpoint.

s_bakmeijer and all: for a more representative look at these types of events in service, take a look at the following reports from the NASA Aviation Safety Reporting System:

ASRS - Aviation Safety Reporting System (http://asrs.arc.nasa.gov/search/database.html)

376201
418260
441448
663390
665350
672119
749437

I believe you'll find three MD-80 series aircraft, and one additional that is unidentified but most likely also a DC-9 or MD-80. These provide some insight into how these types of events take place.

The major difficulty with in service stalls is that at the point of occurrence, the dynamics of the actual event and the flight crew's mental model of the situation are not aligned. Further, the behavior of the airplane in the fully stalled condition may well not be what the simulator has modeled, as PBL has said; this is particularly true of premature stalls resulting from ice accretion. The event is usually a terrifying surprise, developing very rapidly and possibly becoming somewhat violent. Pilot reactions can be surprising.

Although the response of the captain in the Buffalo accident was tragic, it was not as atypical as the Board seemed to think. In the icing accident database that I maintain, I currently have sixteen events in which I am able to document control column pull forces following a stall warning. These include AA 903 at West Palm Beach, ABX N827AX at Narrows, Virginia and a number of EMB-120 and ATR events. I am certain that none of these pilots would have done this in the simulator; but then, what they encountered in the airplane was like nothing they had ever seen in the sim.

john_tullamarine

14th Jun 2010, 22:34

All simulator data packages contain NO valid data beyond the "g" break.

Depends on what the airframe and sim OEMs got up to in the commercial arrangements for the airframe data pack.

More importantly, the pilot does not know to what extent the sim responses are validated or extrapolated. Indeed, two apparently identical sims can be tweaked a little differently by two sim operators to give not quite the same experience to the pilot.

Underslung engines may create more pitch up at low speeds/high AOA than elevator can pitch down!

Caused by the low thrust line and nacelle lip normal force - both destabilising. If the stability characteristics are unacceptable, then one would expect to see an SAS of some sort incorporated to give the pilot the impression of stability. Excluding research aircraft and odd ball system failures, one should not see a pitch unstable aircraft out on the line.

using rudder to lift a wing is commonly taught in flying schools

Probably not good technique at/in the stall due to the risk of spin outcomes. A search of John Farley's comments on stalling are a useful read.

The use of rudder to lift a stalled wing is ESSENTIAL and correct and MUST be taught. Use of aileron on a fully stalled wing will merely increase the wing drop (basic aerodynamics?) and will probably cause a spin.

Very much a case of following the OEM's guidance for a specific Type. However, rudder mixed with stall is a recipe for potentially interesting results. For other than ancient Types, the aileron should be functional throughout .. this, naturally, will depend on wing section variations and whether all or part of the wing is stalled.

BOAC

15th Jun 2010, 07:47

this, naturally, will depend on wing section variations and whether all or part of the wing is stalled. - exactly the problem, John. Here we are all talking about 'stall' in different ways, from light buffet/stick shake through to fully stalled. I refer, as I said, to fully stalled.

Fully stalled and aileron will be a disaster, as will allowing the ensuing yaw to the dropped wing. Using aileron to pick up a stalled 737 wing at 80kts will simply make the wing drop faster.

At stick shake it should be fine

Some of our 'recent' low speed/extreme pitch accidents and near accidents will have been fully stalled, I suggest?

As expressed in posts #12 and #13, "Approach to stall recovery techniques are irrelevant to actual stall recovery". Full stall recovery has long been ignored in airline flying - largely because the training (rightly) is to avoid such. Sadly that seems to have failed.

guiones

15th Jun 2010, 21:21

PBL, this is why you have trouble explaining yourself, two contradictory statements in two posts, is not us that fail to understand, but you that contradict yourself:

"You think people have been going out doing stalls in Airbuses to see what happens?"

"Guiones, I know that flight tests were done and sims data was validated for this whole endeavor."

On the aerodynamic descriptions I do agree with you; but I am form the Pilot side and we must call it approach to stall and stall to get the difference across, it may not be perfect from your point of view, but we must make a difference between the two recovery procedures. And by the way, some people have flown past the break, most are just not around to tell us about it.

G

PBL

16th Jun 2010, 09:32

Oh, dear, here we go again.

Guiones, the advice I give to most people who are tempted to think I contradict myself is that it is most likely to be a clear indication to themselves that they have misunderstood.

In your case, you have misunderstood the concept "contradiction" as well as what I said. Let me explain.

First, in a very simple case, two assertions A and B are contradictory if their conjunction is logically guaranteed to be false.

In the two sentences you quote, there is only one assertion, not two. The first sentence contains a question, not an assertion. It was a question to you, indeed one which I believe you have not answered. When you give an answer, if it contradicts the second sentence (my assertion), then I will be contradicting you, not contradicting myself.

Enough of trivial explanations of logic, back to aero.

So, answer the question: say what you think those flight tests did. Then I'll see if I can find out what they did, and get back to you. It might take me a while.

PBL

guiones

16th Jun 2010, 14:29

PBL, with your question you challenged my assertion that Airbus was conducting flight tests; with your second statement you confirmed it, here in the US we call that a Clinton answer.

I already said what the tests did, compare aircraft behavior to sim behavior in order to write the FOT.

I do not need you to find out anything for me as far as the tests, I have spoken to one of the pilots involved.

G

PBL

16th Jun 2010, 15:29

Guiones,

I don't participate in these forums to engage in silly word games. I participate to engage in discussion about matters of mutual interest in aeronautics. So one last word from me on this to the (hopefully) wise:

PBL, with your question you challenged my assertion that Airbus was conducting flight tests

You are mistaken. I asked

You think people have been going out doing stalls in Airbuses to see what happens?

And you still haven't answered.

PBL

guiones

16th Jun 2010, 16:57

PBL:

To answer your question, again.

Yes, Airbus did conduct tests and this were done to validate aircraft behavior vs sim behavior for the FOT.

PM me if you ever come to Miami and we will discuss in person, I WILL BUY THE BEERS.

My last post on this matter, got to go out to fly.

G

DFC

19th Jun 2010, 09:11

To pick up on two very important points mentioned previously.

1. Stall has absolutely nothing to do with airspeed. It has everything to do with angle of attack. It is entirely possible to fly any aircraft at an airspeed of less than 10 Knots but not stalled - because the angle of attack is less than that at which the stall occurs.

2. Unless one is intending to do some aerobatics or aligning the aircraft with the centerline on landing, with all engines operating the rudder is only used to prevent yaw. Yes that is to prevent yaw. Or as some people like to say - keep the ball in the middle.

In order to "pick up a wing" using rudder - stalled, near the stall or at Vmo - one must yaw the aircraft. The only time I would expect anyone to intentionally yaw the aircraft at or close to the stall is when they want to experience the combination of stall and yaw = spin.

The stall recovery (or prevention) actions are basic and simple - reduce angle of attack and prevent yaw.

Height loss above 20,000 ft in a stall recovery (or avoidance) is not an issue that is worried about. If you are 20,000ft above the MSA then where is the problem in using some gravity to assist a prompt and effective recovery.

Finally, in a multi, it is not simply Vmca and engine failure that is an issue in stall recovery which involves accelerating to TOGA thrust / firewalling the levers but simple things like unequal acceleration times that can introduce unwanted yaw at the critical time when it is least desired.

Reduce AOA and prevent yaw.

BOAC

19th Jun 2010, 11:28

It is entirely possible to fly any aircraft at an airspeed of less than 10 Knots but not stalled - because the angle of attack is less than that at which the stall occurs. - a little 'pie in the sky', DFC - you need to add 'but not in any sort of controlled flight nor for very long', just as it is not possible to stall an aircraft at zero g, but again, not for long..

Desk Jockey

19th Jun 2010, 12:54

I hope this adds something here. From my types you will see I haven't been hands on for a while.
What I was taught in engineering training is that the pusher in the 737 was called a stick nudger and is purely there to properly identify the stall by pushing the column forward gently. The stick shaker similarly is there to replicate the pre stall buffet for those aircraft whose characteristics need additional identification.
(Now showing my age)
The Trident :D and the BAC 1-11 had true stick pushers and were a recovery system. Being a T tail configuration they couldn't be allowed to stall fully as they wouldn't recover after the tail enters turbulent air from the wing as some poor souls found out in testing.
In those aircraft the stick pusher operates rapidly pushing the control columns forward while the tail is still in clean air. Pitch control is this case was taken away from your good selves as it wasnt possible to restrain the columns. In the case of the 1-11 accompanied by very loud klaxons.

Tee Emm

19th Jun 2010, 14:13

Height loss above 20,000 ft in a stall recovery (or avoidance) is not an issue that is worried about. If you are 20,000ft above the MSA then where is the problem in using some gravity to assist a prompt and effective recovery.If the recovery is made correctly, then as you say, no problem. But if in IMC and particularly at night, the sticker shaker operating suddenly will almost certainly be an unexpected event catching a crew by surprise. It would then take only a few harsh control movements to cause serious overcontrolling and unless the pilot was on the ball with his basic instrument flying skill, an unusual atttitude can result. Now you have a deadly serious problem while in IMC. Realistic unusual attitude recovery training is rarely conducted in airline recurrency training. It is seen as unnecessary and a waste of simulator time. Even if this training is conducted it might only be once every couple of years. Not exactly effective recurrency.

PBL

19th Jun 2010, 14:48

Why, I wonder, didn't someone like BOAC pick up on this?

Stall has absolutely nothing to do with airspeed.

That is so wrong. It must be corrected.

The angle of attack at which a modern commercial jet stalls at 0.8M is something around half of the angle of attack at which it stalls at Vs1g.

PBL

A321COBI

19th Jun 2010, 15:21

stall has everything to do with airspeed
a stall is a reduction in the lift coefficient generated by an airfol as angle of attack increases
so it obviously had something to do with airspeed, somebody dropped out of pilot taining school:=

BOAC

19th Jun 2010, 15:34

Basically because it was correct! Your commentThe angle of attack at which a modern commercial jet stalls at 0.8M is something around half of the angle of attack at which it stalls at Vs1g. is totally confusing.

Firstly 0.8M has nothing really to do with 'airspeed'
Secondly you are mixing a Mach effect at an unstated 'g' with a stall at '1g' at an unstated airspeed! Absolute confusion. It will still stall at 'Vs1G' at high level, will it not? Just a different Vs due to airflow Mach effects.

It has everything to do with angle of attack. - is correct. Whatever the angle of attack at which an aerofoil will stall (yes, at whatever Mach number), 'airspeed' is irrelevant. I can stall the same high performance jet at sea level at 140kts straight and level 1g or 400kts in a manouevre.

However, all of this is pretty much irrelevant to the thread.

PBL

20th Jun 2010, 06:51

Basically because it was correct!

No, it's not correct. The angle of attack at which a high-performance wing will stall is a function of the speed of the air flowing over it, exactly contrary to what DFC said in hisher first sentence.

These AoAs, for different airspeeds, are established during certification of the airplane. You put a wing in a wind tunnel, and look at its behavior at different airflow speeds. The AoA at which, say, C_L drops off sharply is a function of the airspeed.

I must say I am somewhat surprised at your response. I had awaited something different.

Your comment
[PBL: The angle of attack at which a modern commercial jet stalls at 0.8M is something around half of the angle of attack at which it stalls at Vs1g.]
is totally confusing.

I am sorry you are confused. I think it is a precise statement, and I don't know how to say it more precisely. Let me try.
I can put numbers on these, for a specific wing X:
"The AoA at which wing X stalls at 0.8M" = S;
"The AoA of wing X at Vs1g" = T
S is round about half of T for wing X.
This won't necessarily be true for all high-performance wings X, but it is certainly true for some of them.
Less confusing?

Secondly you are mixing a Mach effect at an unstated 'g' with a stall at '1g' at an unstated airspeed! Absolute confusion. It will still stall at 'Vs1G' at high level, will it not? Just a different Vs due to airflow Mach effects.

Vs1g is a specific airspeed. What do you mean, "unstated airspeed"?

The AoA at which a wing will stall, and how it will stall, is dependent on Mach effects. These must be determined when certifying aircraft.

I don't agree that this is "off topic". Some posters are talking about stall recovery at high altitude, and many of them seem to be confused about aerodynamics, even what "stall" may mean. The vocabulary I am using is one which I use on an almost daily basis in my work with an aerodynamicist. I am gradually learning through these exchanges on PPRuNe that many pilots seem to be unaware of the aerodynamic phenomena which govern the behavior of their airplane. I don't think this disconnect can be healthy.

As I explained earlier, the "stall" is generally a range of phenomena playing out over a range of airspeeds. The certification-defined "stall speed" is the one at which (crudely) handling becomes compromised due to buffet. This may be, often is, lower than the speed at which C_L falls off. I don't agree with Mansfield that this is not a "practical" distinction. It becomes very practical when considering gusts at altitude. For certification, one has to take the wing to defined-stall ("uncomfortable buffet") and then subject it to gusts of a defined magnitude, and make sure nothing untoward happens, such as suddenly losing all your lift.

PBL

BOAC

20th Jun 2010, 07:58

Vs1g is a specific airspeed. - but varies with altitude, no? Sounds like a variable to me.

It really is of no significance to a pilot how a wing stalls - it is, as you demonstrate, academic and for wind tunnels and test pilots - not airline pilots. What is significant is what to do
a) to avoid it
b) to recover if it does

None of us have time to think (or worry) about little lambda feet creeping along our wing, of boundary layer shocks, of vortex formation etc in an aeroplane full of passengers. Yes, the confusion remains over what is a 'stall' as I have said, but in real terms it matters not. Avoid or recover. People like you can worry about S/T.
I am gradually learning through these exchanges on PPRuNe that many pilots seem to be unaware of the aerodynamic phenomena which govern the behavior of their airplane. I don't think this disconnect can be healthy. - that is the essence of this. The disconnect is between those who hold a mouse and those who hold a control column, and in my opinion reasonable.

PBL

20th Jun 2010, 09:28

[PBL: Vs1g is a specific airspeed.]
- but varies with altitude, no?
Curiouser and curiouser, said Alice.

It varies with dynamic pressure. Maybe I can mention the term EAS and leave the rest to you?

It really is of no significance to a pilot how a wing stalls

What an extraordinary view. It appears to me to conflict with your earlier apparent interest in how a wing stalls:
my reply was to a post about high-level stalls where deploying LE devices will probably put you on your back........ The use of rudder to lift a stalled wing is ESSENTIAL and correct and MUST be taught. Use of aileron on a fully stalled wing will merely increase the wing drop (basic aerodynamics?) and will probably cause a spin.

Say you fly in or near a thunderstorm, and get hit with a strong gust. Is it all the same to you if the airplane suddenly pitches up more strongly than you can pitch down, or if your elevators don't appear to function any more to control pitch? Wouldn't it be preferable to have at least a modicum of understanding as to what might be going on, so that you can effect your desire
b) to recover if it does?

[PBL: many pilots seem to be unaware of the aerodynamic phenomena which govern the behavior of their airplane. I don't think this disconnect can be healthy.]
...... The disconnect is between those who hold a mouse and those who hold a control column, and in my opinion reasonable.

You are welcome to your opinion. Sorry I misjudged you. I had understood from your previous posts that you were interested in what happens with a "fully stalled" wing.

PBL

HazelNuts39

20th Jun 2010, 14:28

So what data is it that a given manufacturer didn't have, that they would have gone back up to get because the regulators were worried again about LOC? Can anybody here say?
Perhaps you might find this article interesting: Airbus testpilot views on stall recovery training (http://www.smartcockpit.com/pdf/flightops/flyingtechnique/3)
If an airplane is equipped with a stick pusher, then, granted if you will, a stick pusher activates at the point of defined "stall". That is, you get slower and slower, up to a specific airspeed at which stick pusher activates, and over you go. At what point in that entire process can you be said to be "stalled"? A suggestion: at no point, or maybe at just one. So, if I may be permitted to insist on accuracy, you are performing a manoeuvre which can be best described as recovery from approach to stall, not as recovery from a stalled state. The pusher does not let you get into a stalled state. Isn't this obvious? That depends. A stickpusher is usually fitted if an airplane, in some or all configurations and cg positions, does not exhibit a natural pitch down tendency at the stall. The 'firing point' of the stickpusher is usually just after the point of maximum lift, i.e. just after the airplane is stalled. The purpose of the stickpusher is not to prevent the stall, but to prevent the airplane from getting deep into the stall. Some manufacturers call their stickpusher euphemistically a Post-Stall Recovery System. For the recovery it is immaterial whether the post stall pitch-down is natural or generated by a stick-pusher.

HN39

PBL

20th Jun 2010, 16:27

Perhaps you might find this article interesting:

I don't think so. I am much more inclined to cut out of the discussion. I am getting tired of expounding basic aerodynamics and being told it's all wrong.

PBL

A321COBI

20th Jun 2010, 16:32

I don't think so. I am much more inclined to cut out of the discussion. I am getting tired of expounding basic aerodynamics and being told it's all wrong

I agree, far too basic and its annoying when people tell you whats involved in your job :mad:

BOAC

20th Jun 2010, 17:38

PBL - before you go 'outside' - personally I am interested, yes. For the great mass of all of the airline pilots in the world, it is not particularly relevant.

I actually spent 4+ years of my military life flying an aircraft that spent a large proportion of its time partly stalled, in pre-stall buffet, at speeds from 170kts in the circuit, through 300+kts in combat to M1.3+.

I don't do it now.

Pugilistic Animus

21st Jun 2010, 10:47

don't think so. I am much more inclined to cut out of the discussion. I am getting tired of expounding basic aerodynamics and being told it's all wrong. stalls, as mentioned before, can occur in any attitude, at any thrust setting and at any airspeed:=--- also maximum loading is placed upon the wing at the stall,... that maximum loading varies with airspeed,

A stall at Va results in a limit load

and a stall above Va will usually result in structural failure due to ultimate loads!!!

Edit: so, therefore the AoA of the stall remains constant...;)

misd-agin

21st Jun 2010, 17:15

Comment made in training course dealing with high altitude jet upsets - the first upset typical results in a loss of X (1-3K?) feet. The second upset results in the average loss of Y (7K?) feet.

Scenario - upset one starts. Crew is startled. Rocking and rolling, buffet, etc. Nose drops and altitude loss starts. Plane starts flying again. So far, so good.

Then the crew tries to minimize the altitude loss and the second upset starts. A perfectly flying wing, well below typical flying speed and typical AOA, is exposed to an increase in AOA/G(BOAC's point) in an attempt to reduce the descent rate. Critical AOA is reached very quickly with different results from a low altitude/high AOA event. Bam, upset 2 is entered.

The course mentioned the major injuries, and a/c damage, result from getting into the second upset. One is a basic a 1G event, while the other is an 'accelerated' stall event. As BOAC alluded to, they can be interesting as heck in fighters but they're a bad place to explore in a jet liner.

John Farley

22nd Jun 2010, 08:23

Sounds like a good course.

Dan Winterland

23rd Jul 2010, 07:54

I have just picked up the latest amendment for my type's QRH which now includes Stall Recovery as a memory item. It looks as if they are trying to re-enforce the point that power is not to be used, particularly pertinent of the FBW Airbus where there is no pitch up with power application in normal or alternate law. If you are stalling, you could possibly be in direct law and you will get a pitch up.

IIRC, there was a B737 in the UK (Bournemouth?) which got to a very low speed because the pilot didn't handle the thrust induced pitch up. this is probably what has lead to the manufacturers re-examining thier advice and procedures.

The emphasis seems to be on the fact that the Standard Stall Recovery taught in the early lessons in light pistons is not applicable to all types, and that any recovery actions require consideration - or a type specific automatic response promulgated by the publication of memory actions.

But it's been an interesting thread, in respect of some of the advice being given such as using the rudder to keep the wings level at the stall, and that the stall is speed dependant. I thought these notions had been eradicated years ago. I'm suprised they still seem to exist!

BOAC

23rd Jul 2010, 08:49

Dan - out of interest, which flight controls do you suggest you should use at the point of stall to control wing drop?

There are many, many more examples of underslung engine pitch/power near LOC than the BOH example.

You say that the latest amendment says 'no power'? For the avoidance of possible confusion, is that for all stalls - including low-level/approach? I appreciate we are on a 'high-level stall' thread but you quote a low-level example (BOH), and do I assume this is an AB amendment?

PBL

23rd Jul 2010, 10:12

in respect of some of the advice being given such as ...... that the stall is speed dependant. I thought these notions had been eradicated years ago. I'm suprised they still seem to exist!

To surprise you even more, here is a recent diagram from Boeing (http://www.boeing.com/commercial/aeromagazine/aero_12/attack_fig4.html) showing clearly that C_L_Max is dependent on Mach number.

If you think that the point of C_L_Max is the point of stall (and let me caution about that, because that is not what the certification regulations say, but it is what many pilots seem to believe), then maybe you could ask yourself why Mr. Boeing is showing point of stall dependent on Mach number. Maybe because it is?

PBL

Mad (Flt) Scientist

23rd Jul 2010, 16:51

And, to reinforce PBL's point, while that Boeing diagram may be principally thinking about Mach effect from, say, ).5 through to 0.8, there are aerofoils where the Mach effect is perceptible all the way down to very low numbers - 0.2 or lower - a speed regime where Mach effects are not normally expected.

BOAC

23rd Jul 2010, 19:10

While you very learned and clever chaps crunch your LaPlaces and Fouriers, to bring this back to the OP's question - from a piloting point of view- compressibility etc has no relevance on stalling. We are no longer in the days of Canberras and B-47s and in 'uncharted territory' but the driver nowadays has guidelines of 'normal' pitch attitudes at different altitudes and does not need to know or care about shock separation, and modern EFIS has buffet margins clearly displayed anyway. As long as he/she can differentiate between high-speed and low-speed buffet and carry out the appropriate recovery action, that is enough.

HazelNuts39

23rd Jul 2010, 20:38

As long as he/she can differentiate between high-speed and low-speed buffet and carry out the appropriate recovery action, that is enough.
BOAC;
Sorry if this getting a bit off track, but I've been wondering for some time, in the context of loss of airspeed indication, would a pilot always be able to distinguish between high-speed and low speed buffet? Does it feel differently, and how would he 'learn' to differentiate?

EDIT:: I noticed in the BOEING paper that their stall warning schedule (fig.10) is designed to allow that distinction.

regards,
HN39

CONF iture

23rd Jul 2010, 22:37

You say that the latest amendment says 'no power'? For the avoidance of possible confusion, is that for all stalls - including low-level/approach? I appreciate we are on a 'high-level stall' thread but you quote a low-level example (BOH), and do I assume this is an AB amendment?
This is for all stalls, low level included.
The procedure (as described by Airbus) is now to SMOOTHLY INCREASE THRUST AS NEEDED but only when out of stall.

Clearly, the following Note is in direct relation with Perpignan, and probably with Amsterdam (Boeing) as well :
In case of lack of pitch authority, reducing thrust may be necessary.

Dan Winterland

24th Jul 2010, 04:49

Thanks CONFiture. What I should have said is 'no increase in thrust'. And a note at the bottom of the new Airbus procedure states that "If a risk of ground contact exists, once clearly out of the stall (no longer stall indications), establish smoothly a positive climb gradient" It is clearly meant for all levels. In the case of a stall warning at lift-off, there is another new memory items which is to select TOGA thrust and 15 degrees nose up.

BOAC asked ''Dan - out of interest, which flight controls do you suggest you should use at the point of stall to control wing drop?''

It depends on the type. In a straight wing type, you don't. The stalling exercise I learned many years ago and then later taught for seven years in the RAF training world had the pilot use rudder only to prevent further yaw. The wings were levelled with ailerons once the buffet had gone. So you don't try and control wing drop at the stall. You let it happen and sort it out after you have regained full control.

With swept wing aircraft, the recovery can include the use of ailerons. the only swept wing aircraft I regularly (or intentionally!) stalled was the Hawk, and IIRC, we used the ailerons to keep the winds level at the point of stall - although it's twenty five years ago since I last flew it. And the new Airbus memory item states:

- NOSE DOWN PITCH CONTROL..................APPLY
- BANK..................................................WINGS LEVEL

PBL. Thanks for the diagram. I had to stare at it for a long time before I could work out what it was trying to tell me. I thinks it's what I always understood - that lift decreases with compressability. I conceed that it may have a small effect, but in practical terms - it isn't that important. The main factor which affects stall speed as far as pilot handing is concerned is the load factor.

HazelNuts39 asked: "Sorry if this getting a bit off track, but I've been wondering for some time, in the context of loss of airspeed indication, would a pilot always be able to distinguish between high-speed and low speed buffet? Does it feel differently, and how would he 'learn' to differentiate?"

High speed buffet tends to be at a higher frequency. But the big clue should be the part of the flight envelope at which it occurs - for which the power setting, attitude and slipstream would all give major clues to even the least aware pilot as to which buffet he is in. However, there are some aircraft powerful enough to push themselves into "Coffin Corner" - the point at which the low speed buffet and high speed buffet meet. The U2 is in this category. And I used to fly the Victor, which at lighter weights could fly into this area.

BOAC

24th Jul 2010, 06:56

HN - Dan W has summed up the 'buffets' nicely. There is no way that the 'routine pilot' should get anywhere near the often misnamed 'coffin corner', and as Dan says providing the pilot is aware of what he/she is doing vis a vis pitch attitude there should be no great difficulty in distinguishing between the two. As Dan says they are different, but I would not expect a 'diagnosis' based on 'what it feels like' to take place. We are well-provided now with target pitch attitudes for different stages of flight and should know which we are feeling from that.

Dan - I am completely with you on the correct sequence of stall recovery ie pitch then level wings, but you do need to be very careful with aileron even on a swept wing. These do not exhibit the classic straight wing stall and aileron is useable while the a/c is in buffet BUT if you look at the FDR from PGF you will see the classic 'Stalling II' demonstration of use of aileron increasing wing drop on a fully stalled wing. That is how they got to those bank angles. I suspect that your successful use of aileron at 'the point of the stall' in a Hawk was not actually at the 'stall' - the Lightning, which as I have said, spent much of its time in pre-stall buffet would respond readily to aileron until it didn't and then you were in a spin - and it was that quick, and the key was to prevent yaw into the dropping wing which always leads to autorotation. I have several (unexpected) 'flick rolls' in my log-book to prove that. The lesson has to be driven home - forget aileron UNTIL you have unstalled the wing. It can and does kill. That is all I am trying to say. It is of concern that people seem to either forget or not be aware.

HazelNuts39

24th Jul 2010, 08:04

Dan Winterland;

Thanks for your explanation on buffet.

regards,
HN39

TheChitterneFlyer

24th Jul 2010, 09:25

What a convoluted thread this has turned-out to be! I think everyone will agree that the most dangerous phase of flight to experience the stall, or the approach to a stall, is the intermmediate/final approach phase where the loss in altitude (during recovery) will 'hurt' if we don't get it right.

Whilst that we're all professionals and that we all (I hope) understand the theory behind these discussions, it's true to say that the regulators do, sometimes, have to 'spell things out' to those who don't. The Training Manual; part of the Ops Manual, must provide a standardised procedure for stall recovery technique. Hence, the basic concept for stall recovery (when within the area as stated above) is to reduce alpha. The application of thrust alone will innevitably increase alpha. I believe that this is more of an 'awareness' issue; in practice, you'd no-doubt lower the nose and increase thrust whilst also 'controlling' the tendancy for thrust induced pitch-up.

We're all running around in circles and getting 'bogged-down' with semantics. If we were all to be in the same room and having the very same discussion we'd no doubt all agree with each other... this ain't rocket science!

TCF

BOAC

24th Jul 2010, 10:44

whilst also 'controlling' the tendancy for thrust induced pitch-up. - yes - in my opinion I would have preferred the manufacturers to have focussed on 'unusual attitudes'+pitch/power couples rather than fiddling with basic stall recovery techniques. We are seeing many examples of failure to control this, and 'tweaking' the stall recovery will not stop these a/c losing control in a g/a. A few years back Boeing pushed out a series of 'U/Ps' plus notes and videos for a recurrent sim exercise.

Personally if I am stalling near the ground I want and need power.The procedure (as described by Airbus) is now to SMOOTHLY INCREASE THRUST AS NEEDED but only when out of stall.- hmm!

As with others here, I am still unclear what this amendment means by the STALL in terms of recovery - stall warning or full stall?

Dan Winterland

24th Jul 2010, 17:32

Hi BOAC. It's the stall warning. The Airbus Flight Crew Operating Manual 3 defines the point at which recovery should be made, the new procedures are an amendment to the current procedures and deal purely with the recovery actions. However, you will only get a stall warning in a reversion flight control modes.

The FBW Airbus will normally be flying in Normal Law where stalling is theoretically impossible. The protections will not let you do it. If there are flight computer degradations or an upset which has not been prevented by the normal protections, then you may be in Alternate Law. If you lose more computers or ADCs, then you may be in Direct Law which means that the control surface deflection is proportional to stick movement - i.e. like a conventional aircraft. The aircraft can be stalled in Alternate or Direct Law, but there is a very definate stall warning associated with his. In Normal and Alternate law, the aircraft will not pitch up with power application. Only in Direct law.

So the instances where these new memory items may be required will be very rare. I have never been out of normal law in 4000hrs of Airbus flying. My company has a fleet of 30+ and we have had one case of Direct law in the last 5 years - casued by a double RAD ALT failure where the aircraft reverts to Direct Law when the gear is lowered.

And as for the new procedure, I suggest any pilot would apply a healthy dose of airmanship if required. I think if I was to stall at 200' on approach, I would use some power.

As for the use of ailerons, I don't remember exactly at what stage of the stall we used them down to while flying the Hawk- I will ask someone who has flown it more recently. As for the Airbus procedure, it's in the memory item procedure therefore, it is approved by the manufacturer. Perhaps it's because in Alternate and Direct law that VSW (Velocity Stall Warning) is somewhat higher than the actual stall speed and the ailerons are still safe to use and limiting the load factor by reducing the angle of bank is the main priority.

I'm in agreement about the different characteristics of swept wing stalling. Earlier types tended to pitch up as the tips stalled first and the centre of pressure moved forwards. Two of my previous types exhibited this trend (Victor and VC10) and as both had T tails, this was definately not a good thing! Some other types exhibit a classic nose down pitch. The B747 was originally designed with no stall warning or protection at all. It was only the British ARB which demanded a warning system which led to a stick shaker being fitted. Which is why the B747's device looks like an afterthought - an electric motor with an eccentric weight attached, and clamped to the control column.

The VC10 has a duplicated shaker and warning system. The pusher actuates before the stall and the system has a 'rate computer' which takes a feed from the AoA sensor and triggers the system earlier if there is a nose up pitch detected. The Victor had no stall recovery aids - except for deploying the brake chute in the vague hope it would pitch the nose down before the weak link broke. More than one Victor met it's demise in stalling.

CONF iture

24th Jul 2010, 18:29

However, you will only get a stall warning in a reversion flight control modes.
Except in Perpignan where one AoA sensor was considered as damaged by the system, even if actually it was the only one to tell the truth but the system didn't deem as necessary to advise the crew of the actual AoA DISAGREE.

When you know the Major role that Airbus gave to those AoA sensors in their Protection features, I think it was absolutely necessary to advise the crew that something was wrong with the readings.

To me, clearly, the Airbus philosophy is ALSO to be blamed in the Perpignan accident.

411A

25th Jul 2010, 04:18

I wonder...all this discussion about stalls and using rudder (or not) or ailerons (or not) during recovery.
Has anyone here actually stalled a large swept-wing airliner?
Is so, what were the results?

Having discussed this subject with Lockheed production test pilots, whom have actually intensionally stalled an L1011 during acceptance tests, rudder is never ever used to 'lift a wing' when the airplane is fully stalled, either in the clean, approach or landing configuration.
To use rudder to 'lift a wing' in the fully stalled condition, is to invite a rather abrupt up-side down condition, IE: inverted flight.
Not good for keeping ones breakfast down, me thinks.:eek:

Personally, I have completed acceptance tests with L1011 aircraft where we approached the fully stalled condition.
Rudder was never used during recovery...stab (the L1011 has an all-flying stab, unique to civil jet transport aircraft), ailerons and power, only.
No rudder.

Tee Emm

25th Jul 2010, 05:16

forget aileron UNTIL you have unstalled the wing.

And providing you have instantly pushed the stick forward when the stall is detected, then that takes less than two seconds to unstall the wing. So would it be correct to say that for all intents and purposes the stick forward and aileron as needed to roll wings level is simultaneous.

On the other hand, to take this rudder thing to a ridiculous degree I was astonished to have a senior flying school instructor tell me that on final approach in a Cessna 152 which has a Vref of 54 knots,his students are taught never to use the ailerons below 60 knots IAS to correct bank angle. Instead they must ensure the wheel is central and control bank angle by carefully "skidding" the wings level using rudder in case they are close to the stall.

When a CFI of a flying school teaches that to his student instructors on an instructors course, then no wonder the standard of the flying industry where I come from is appalling. The regulator,in the form of local flight ops inspectors, have no idea this goes on under the very noses, because they are too busy travelling around the countryside trawling through flying school paperwork audits and writing audit reports criticising the operator for using non standard sub-paragraph format in their Operations Manuals. OK - that is a generalisation; but believe me its not far from the truth all the same.

PantLoad

25th Jul 2010, 05:31

Let's see, if I were to try to get an aircraft to spin, how would I go about it????

Fly safe,

PantLoad

BOAC

25th Jul 2010, 07:23

So would it be correct to say that for all intents and purposes the stick forward and aileron as needed to roll wings level is simultaneous. - absolutely - unstall the wing and ailerons are back on the menu - but the problem is that there were pilots (now dead) who tried full aileron to pick up a wing on an Airbus in 'flight' at less than 46kts and thereby managed an extreme bank angle.

Yaw towards a dropped wing is, of course, an excellent way to get the nose back down in an otherwise uncontrolled power/pitch situation (and as taught for upset recovery) providing height is sufficient, but alpha must be kept under control.

Airclues

25th Jul 2010, 08:10

Has anyone here actually stalled a large swept-wing airliner?
Is so, what were the results?

In the early 80's I was co-pilot on several C of A air tests on the Boeing 747 when a full stall was completed (I believe that the UKCAA was the only authority that required this).
Although there was a significant nose drop, I can't remember there being any wing drop. There was a severe buffet which made it difficult to read the instruments (I don't know how the test engineers came up with their figures). The height loss was about 5000ft. In the mid 80's the UKCAA fell into line with the rest of the world and cancelled the full stall requirement.

Dave

Thread drift;

On 3/8/83 we carried out an air test on G-AWNB. We were given permission to do the stalls in the Portland Danger Area so that we could be sure that there was no traffic below us. When we had completed the stalls we were invited to carry out a GCA approach at Portland N.A.S. Was anyone there at the time? If so, do you have any photos?

Goprdon

25th Jul 2010, 10:38

To the following CAA Flight Crew Training Notice I would add:
Do not try to pick a wing up with rudder, it may work on some aircraft but on others you will spin.Do use the rudder to prevent yaw.
Do move the control column forward to lower the nose and reduce the AoA.
Do follow the manufacturer's instructions regarding stall/spin recovery.
Do re-read PantLoad's post above at Post69.
Now for the notice:

SAFETY REGULATION GROUP FLIGHT CREW TRAINING NOTICE
01/2010
Applicability: RETRE, TRIE, TRE, SFE, TRI, SFI Effective: Immediate
STALL RECOVERY TECHNIQUE
1 Recent observations by CAA Training Inspectors have raised concerns that some instructors (both SFIs and TRIs) have been teaching inappropriate stall recovery techniques. It would appear that these instructors have been encouraging their trainees to maintain altitude during recovery from an approach to a stall. The technique that has been advised is to apply maximum power and allow the aircraft to accelerate out of this high alpha stall-warning regime. There is no mention of any requirement to reduce the angle of attack – indeed one trainee was briefed that “he may need to increase back pressure in order to maintain altitude”.
2 It could be argued that with all stall warning devices working correctly on an uncontaminated wing, such a recovery technique may well allow the aircraft to accelerate out of danger with no height loss at the lower to medium altitudes. The concern is that should a crew be faced with anything other than this idealised set of circumstances, they may apply this technique indiscriminately with potentially disastrous consequences.
3 The standard stall recovery technique should therefore always emphasise the requirement to reduce the angle of attack so as to ensure the prompt return of the wing to full controllability. The reduction in angle of attack (and consequential height loss) will be minimal when the approach to the stall is recognised early, and the correct recovery action is initiated without delay.
NOTE: Any manufacturer’s recommended stall recovery techniques must always be followed, and will take precedence over the technique described above should there be any conflicting advice.
4 Any queries as a result of this FCTN should be addressed to Head of Flight Crew Standards
Captain David McCorquodale Head of Flight Crew Standards
21 April 2010

PBL

26th Jul 2010, 12:49

PBL. Thanks for the diagram. I had to stare at it for a long time before I could work out what it was trying to tell me. I thinks it's what I always understood - that lift decreases with compressability.

What I was hoping it would tell you is that C_L_Max occurs at different angles of attack depending on Mach number. Which shows (if you take C_L_Max to represent the point of stall) that the point of stall is dependent on Mach number, contrary to your suggestion that the "notion[.] ... that stall is speed dependent .... ha[s] been eradicated years ago"

BOAC seems to want to maintain the view that this stuff is "theory" (whatever that may mean; we are talking about the physical characteristics of the airfoils you chaps fly, as demonstrated by wind tunnel data). I find that a bit odd, since Boeing Aero is a magazine which is designed to address practical issues for pilots, so one may take it that Boeing thinks this "theory" is relevant to flying their airplanes.

BOAC says from a piloting point of view- compressibility etc has no relevance on stalling

I can't see that. The diagrams tell me a different story. Not to you?

the driver nowadays has guidelines of 'normal' pitch attitudes at different altitudes and does not need to know or care about shock separation,

Shock separation occurs with overspeed. We were talking about stall.

.....As long as he/she can differentiate between high-speed and low-speed buffet and carry out the appropriate recovery action, that is enough.

With respect, no it's not enough. There are two different kinds of high-speed buffet: that caused by approach to stall and that caused by overspeed. The "appropriate recovery actions" are different in those two cases and the pilot needs to distinguish them. Anecdotal simulator evidence, from a reliable source, suggests to me that not all line pilots can.

BTW, I am curious as to what the new Airbus procedures are for stall warning at cruise altitudes. It used to be to reduce back-pressure on the stick.

Also BTW, I am surprised that some people are suggesting you can't stall an Airbus (say, an A330) in Normal Law. You can, obviously, if you encounter a strong-enough gust. You can also stall it in Alternate Law, for the same reason. And in Alternate Law, in case of an ADR DISAGREE, high-AoA protection is lost in any case, so you can stall it without a gust in that situation.

PBL

Capt Pit Bull

26th Jul 2010, 12:56

Been chewing over this one a bit.

One thing that occured to me is that if I tallied up all the time in the sim spent with stall warning / stick shake / stick push going off, on 95% of the time the exercise was windshear recovery at fairly low level, with large back forces on the control column and overpowering the stick push.

(several differnt types involved).

Now, regardless of what I know intellectually, you have to wonder what the muscle memory will make of it all.

pb

john_tullamarine

27th Jul 2010, 07:04

If anyone has made the observation earlier in the thread, my apologies.

As far as I can see, PBL has been pretty much a lone voice in the wilderness extolling the influence of M. Certainly chaps such as JF and MFS, naturally enough, will be well aware of the following ..

For routine activities we tend to think of lift and drag only being functions of alpha and that works fine and beaut for most low speed activities in typical aircraft.

However, for many circumstances one needs to delve just a little deeper to get the story. If one digs into any of the standard engineering texts, one finds that the variables include -

(a) alpha

(b) a length measure of some sort, usually taken to be mean chord

(c) density

(d) dynamic viscosity

(e) local velocity

(f) sonic velocity

If one plays with some dimensional analysis sums, and that's always good fun, this list can be simplified to CL and CD being dependent on

(a) alpha

(b) Re (Reynolds Number)

(c) M (Mach Number)

For most cases, we can ignore the effects of Re and M and simplify to the usual CL = etc...

However, if there is any significant variation in Re or M throughout the flight regime, then one ought not to be surprised if the simplified sum gives a dodgy answer ...

It's not helped by pilot training only talking about the simplified sum without qualifying that story with the usual engineering caveats.

As we so often find, a little knowledge can generate the most stimulating of spirited discussions.

mad_jock

27th Jul 2010, 07:53

For those made of very stern stuff...

Navier-Stokes Equations (http://www.grc.nasa.gov/WWW/K-12/airplane/nseqs.html)

The thought of that exam still sends shivers down my back.

error_401

27th Jul 2010, 08:34

5 cents worth into the thread.

I've been an observer on an acceptance flight for an ATR72 in Toulouse with test pilot from ATR on right hand seat and flight test engineer on jump.

Aircraft is kept in level flight with a bit over flight idle power until:

1st stick shaker activates
2nd stick pusher activates

I wanted to know about the margins and this is what I've been told: There should be about 2 degrees AoA more before the wing actually stalls. The stick pusher will activate while the wing is still stable to fly.

IMHO: I would always go for pitch and then power. At height I would accept an altitude loss. I hope NEVER TO FIND myself in a FAR/JAR 25 aircraft near stall and then experience a wing drop because it has fully stalled!:\
I would always refrain from using rudder else as to minimize sideslip and then only after the wing is flyable again.

What do we learn:
- Modern aircraft even ATR with conventional controls have no pronounced low speed buffet therefore stick shaker
- Stick pusher activates WELL BEFORE entering a wing stall else it would be pointless to have it but is based on AoA.
- Nevertheless I agree with BOAC/PBL when you leave normal flight envelopes you're out on your own and it may stall before stick pusher.

rudderrudderrat

27th Jul 2010, 08:45

Hi JT,

Forgive my ignorance - but shouldn't there be an "area term" amongst the variables?
If you include chord, then don't you need span?

BOAC

27th Jul 2010, 09:34

and it may stall before stick pusher. - leaving aside system failures and icing which introduce a whole new 'ball game', I maintain that a properly designed a/c should not exhibit that characteristic - after all, stall warning computations should be corrected for Mach in modern a/c.

"I hope NEVER TO FIND myself in a FAR/JAR 25 aircraft near stall and then experience a wing drop because it has fully stalled!" amen to that, but it is apparent that people are.

rrat - I venture to suggest that JT is talking about 2-D flow?

m_j - many thanks for revisiting the scene of the odd recurring nightmare:mad:

jt - "It's not helped by pilot training only talking about the simplified sum without qualifying that story with the usual engineering caveats." agreed, but it should be remembered that not all pilots are BSc level aeronautical engineers - some are just capable of the 3 times tables. Surely the simple lessons are:

Stick to the limits from the manufacturer - IAS/Mach/Altitude
If something untoward happens when you do 'X', undo 'X'
Learn the correct recovery actions if the problem persists.

One would hope this would work across the airline pilot spectrum.

mad_jock

27th Jul 2010, 09:42

There isn't an area term because the foil is presumed to be infinately long so you don't have issues with air movement in the z axis.

So one of the boundary conditions is that the length considered is of unit length. So as the area equals chord x length it simplifies it down to just chord.

2D analysis of this stuff (which is what we are doing by stipulating unit length) is number crunching intensive. 3D anaylis will blow you and you computers socks off. If you then include time as well as turbulent flow you will decide sod this for a game of solders those fluids boys are off there heads I am sticking to structures and thermo things made out of steel. That fibre stuff is a pile of cack as well.

and BAOC I resent that comment, BEng Mech can do at least up to ther times 12 table.

BOAC

27th Jul 2010, 10:07

BEng Mech can do at least up to ther times 12 table - not wot we 'aeros' used to say about them.....................I guess they use their fingers to do that?:)

mad_jock

27th Jul 2010, 16:01

Well we did have a special course in it.

Had to learn how to deal with inches and associated constants for oil,pressure vessel and pipe work.

Learned how ****e yank steel was (shocking bad for both fatigue and corrosion). Always buy from Japan cause the Brit stuff was ****e as well after they shut down Ravenscraig.

All was going well until we had to convert critical crack intensity's which I think from memory are Nmm^0.75 to the yank equiv. At which point I was forced to go to the pub.

john_tullamarine

27th Jul 2010, 22:54

shouldn't there be an "area term" amongst the variables

comes in with the dimensionless coefficients

not all pilots are BSc level aeronautical engineers - some are just capable of the 3 times tables

or course - this is the reason that explanations are couched in different terms for different folks, according to their knowledge, needs and state of inebriation.

I guess they use their fingers to do that?

on occasion, depending on the quality of the grape, I have to resort to counting on my fingers and toes .. 11 is about as far as I can go.

Stick to the limits from the manufacturer - IAS/Mach/Altitude
If something untoward happens when you do 'X', undo 'X'
Learn the correct recovery actions if the problem persists.

now that's a terse set of good advice and well worth committing to memory.

inches

thanks be to tradition .. shall we confuse them all with slugs, poundals, ergs, and such like ?

mad_jock

28th Jul 2010, 07:19

Currently ploughing my way through this.

God Created the Integers: The Mathematical Breakthroughs That Changed History edited by Steven Hawkins.

Keeps the grey matter stimulated.

john_tullamarine

28th Jul 2010, 10:36

Steven Hawkins

now, you're a brave man ..

When one contemplates Hawkins, what a most incredible fellow ...

mad_jock

28th Jul 2010, 11:16

Its not really written by him he supply's the editing and the commentry on the particular topic under discussion and how it relates to things we use every day.

The full title is

God Created The Intergers, The mathematical breakthroughs that changed History

It starts off with Euclid 325-265BC

And finishes with Alan Turing 1912-1954

Currently working my way through Euclid Elements just now. I will admit its not everyones idea of fun. The most sobering though is that all Royal Navy midshipmen had to learn it by the time they were 14.

There is quite a bit of Mech eng interest stuff in it as well apart from goemetry with:

Newtons, Motion of Bodies
Fourier, Propagation of Heat
Cauchy, Calculus

john_tullamarine

28th Jul 2010, 11:20

That's unfair .. now I'm going to want to go off and get a copy ...

PBL

28th Jul 2010, 12:13

While the English and their former colonists have persisted with the adaptation of a mercantile system of units to science, I am not at all in accord with the suggestion to confuse them all with slugs, poundals, ergs, and such like because I hold it unseemly to berate the slug, poor creature. It is a necessary part of a system of consistent units.

I myself prefer the units which Mother Nature intended, in which the speed of light is 3 x 10^9, gravity is 10, the speed of sound in dry air at 15° is a third of a thousand, and, most quaintly, the Newton is about half an apple. But I do grant privileged status to the nautical mile as a minute of great-circle arc.

The point made by JT, that the essential coefficients of lift, drag, and moment are strict functions of AoA, Mach number, and Reynolds number, along with the derivation, is elegantly expressed in almost exactly those terms in Section 5.3, pp 262-8, of John D Anderson Jr., Introduction to Flight, Sixth Edition, McGraw-Hill 2008.

I tried summarising what I think are the important, sometimes revealing, points in this discussion, but it's up to about 4 pp so far, so rather than post it here I'll clean it up a little and likely post it on our blog, and publish a link here when I do.

PBL

Centaurus

28th Jul 2010, 13:02

As a simulator instructor on the 737-300 and after reading these posts I am beginning to wonder if perhaps I am teaching an incorrect technique when discussing stick shaker recovery at high altitudes. After all, we can only use the advice given in the relevant manufacturer's FCTM and sometimes that advice is quite lacking in more detail leaving it to the reader to fill in the gaps.

In the case of the 737-300 the FCTM states " Ground contact not a factor..at first indication of stall (buffet or stick shaker...apply maximum thrust, smoothly decrease the pitch attitude to approximately 5 degrees above the horizon and level the wings...as the engines accelerate, counteract the nose up pitch tendency with positive forward control column pressure and nose down trim..at altitudes above 20,000 ft pitch attitudes of less than 5 degrees may be necessary to achieve acceptable acceleration...accelerate to maneuvering speed and stop the rate of descent...correct back to target altitude.

From my observation in the simulator of high altitude stall recovery (above 31,000 ft for example) it is the application of maximum thrust that often causes a delay in early recovery because of the marked pitch up that occurs as high power cuts in. Unless the pilot accurately pins zero body attitude as the aircraft accelerates downhill, the aircraft tries to lift the nose and a "G" buffet returns. Perhaps the accent should be on instant lowering of the attitude to commence speed increase and only when at a appropriate speed above the initial buffet (say 30 knots?) thrust should increased? In the case above for training purposes, we close both thrust levers and maintain level flight while slowly decelerating towards the stick shaker speed.

rudderrudderrat

28th Jul 2010, 13:31

Hi Centaurus,

Post 72 from Goprdon quotes the CAA guidelines:
" The standard stall recovery technique should therefore always emphasise the requirement to reduce the angle of attack so as to ensure the prompt return of the wing to full controllability. The reduction in angle of attack (and consequential height loss) will be minimal when the approach to the stall is recognised early, and the correct recovery action is initiated without delay."

So I think it's emphasising that a reduction in angle of attack is the most important thing, secondly the application of a controllable amount of power. The problem with under slung wing engines is the pitch up couple. (TOGA power might not be controllable by the elevator at very low speeds - 737 at AMS etc.) As the speed increases - then you will be in a position to control the application of even more power.

Edit. The problem with the simulator is the lack of feeling of the "delta g" during the recovery. Most crews will trigger a secondary stall stick shake because of their over enthusiastic pitch up. In real life - they will be more aware of the "g".

PBL

28th Jul 2010, 14:40

The blog post is now there: Understanding Aerodynamics of Stalls (http://www.abnormaldistribution.org/2010/07/28/understanding-aerodynamics-of-stalls/)

PBL

BOAC

28th Jul 2010, 16:49

Forgetting 'aerodynamics' and concentrating instead on 'stability and control', there appear to be two major contributors to recent 'underslung' accidents/incidents involving near or complete LOC following an unintentional low-speed 'event'.

1) The way that a/c flight systems are allowed to automatically continue trimming tailplanes below normal flying speeds without some form of warning to crews that this is happening

2) A failure to recognise and handle the resulting excessive pitch and low speeds following the application of recovery power.

As I and others have said over the last few years, historically initial and recurrent training has 'compartmentalised' stalling and unusual attitude recovery ie "now we will look at stalls and recoveries" and "next session we will look at nose-high low-speed recoveries". Those of us who have air-tested down to the stall know that you always STOP trimming at xkts above the stall speed so you do not get the excessive nose-up couple. We are now arriving increasingly at events where the a/c has done exactly the opposite and crews are not ready for it. With the increasing ratios of those pilots in the industry who have not got much real experience of unusual attitudes, it is apparent, to me anyway, that 'stall' recoveries MUST now lead straight into the nose-high low-speed recoveries as an exercise, meaning that I would suggest allowing trimming right down to stick shake. Recovery from these conditions is indeed simple if you are prepared for it and take the correct actions.

411A

28th Jul 2010, 20:52

1) The way that a/c flight systems are allowed to automatically continue trimming tailplanes below normal flying speeds without some form of warning to crews that this is happening

Except for the L1011...simply a better design, out the door with ship one.;)
Now, some might say...'EAL401 proved the L1011 was not up to standard'...and I would say...'IF you follow the manufactureres instructions, it was fully compliant, and still is.
IE: RTFB.
Rocket science it ain't.
Gosh, what an absolute surprise.:rolleyes:

john_tullamarine

28th Jul 2010, 23:12

In the case of the 737-300 ..

With the caveat that the OEM's advice for a particular aircraft is likely to be based on sensible FT work and, therefore, one should observe it .. there are a couple of considerations with engines sticking out in the front like most of today's twins -

(a) if the engine is underslung (as nearly all are) the low thrust couple doesn't help

(b) if the upset has been allowed to progress to a high alpha, there is a problem observed once power is applied. As power comes up, with the change in flow around the front of the nacelle (and through the prop disc for turboprops) there arises a significant up-force at the nacelle lip (or prop disc).

Both of these are destabilising and reduce the longitudinal static stability. The end result is that the pilot finds him/herself in hot water.

The main emphasis for any recovery should be skewed toward reducing alpha and then, at an appropriate time for the Type, increase power.

you always STOP trimming at xkts above the stall speed

and the certification stall data (and the AFM techniques) are based on a similar approach to the matter ie, if the trim has run too far nose up then the pilot's existing problem are complicated further.

Gosh, what an absolute surprise

.. well, after all, it IS a Lockheed product .. not like those other agricultural animals.

HazelNuts39

29th Jul 2010, 10:59

You think people have been going out doing stalls in Airbuses to see what happens?
The aerodynamic characteristics of a modern commercial jet are determined in wind tunnels, mostly for certification purposes. For type certification of Transport Category airplanes the 'applicant' (manufacturer) must conduct an extensive stall test programme. Details and objectives can be found in reference 1. Wind tunnel testing does not produce the evidence required for certification, but is primarily a design tool and assists in preparing the flight test crew for the more critical conditions to be tested, as explained in reference 1.
So what data is it that a given manufacturer didn't have, that they would have gone back up to get because the regulators were worried again about LOC? Can anybody here say?The major manufacturers contributing to an industry-wide effort to improve pilot training aids for upset recovery, including recovery from stalls, drew on their extensive experience. They did not have to 'go back' to get it. A detailed reply to your question 'straight from the horse's mouth' can be found in ref.2.
Another, Airclues, replied In the early 80’s I was co-pilot on several C[ertificate] of A[irworthiness] air tests on the Boeing 747 when a full stall was completed (I believe that the UKCAA was the only authority that required this) and described his experiences. In other words, actually high-alpha-stalling large commercial aircraft, even for certification, is ancient history.
Perhaps you are confusing (type) certication with the UK CAA practice for periodic renewal (from memory: every 5 years) of the C of A of individual 'Performance Group A' aircraft on the UK Register. The UK ARB (later CAA) requires a performance flight test as a condition for renewal of the CofA. The flight test includes stalls and verification of climb performance with simulated engine failure (2nd and 4th TO climb segments). To my knowledge, the CAA is the only authority to require these tests, but possibly some former 'Commonwealth' countries follow CAA practice. The CAA requirement for CofA renewal performance testing has nothing to do with type certification testing.

Reference 1: Advisory Circular no.AC-25-7A (http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgAdvisoryCircular.nsf/0/C2614E27B49BF38686256BA300696689?OpenDocument&Highlight=ac%2025-7a) Flight Test Guide for Certification of Transport Category Airplanes.

Reference 2: Airbus Flight Test views on Upset Recovery Training (http://www.smartcockpit.com/pdf/flightops/flyingtechnique/2)

PBL

29th Jul 2010, 12:55

HN39,

Your reference 1 will be quite useful for those here who don't know what is taken to determine a "full stall". I point out -again - that the flight tests are only performed to determine Vs1g. That is one point - a very, very important point, but still only one - on the C_L / AoA / speed-MachNo graph collection.

Condition (i) for "fully stalled" says "pitch control reaches the aft stop and is held full aft for two seconds, or until the pitch attitude stop increasing, whichever occurs later". You can obviously do that in both Normal and Alternate Law (that is, Alternate 1) on an Airbus without actually reaching, let alone going beyond, AoA for C_L_Max (and I mean the real C_L_Max). Regs define that as "fully stalled", but it is not aerodynamically at the point of serious loss of lift, or even severe buffet.

The major manufacturers contributing to an industry-wide effort to improve pilot training aids for upset recovery, including recovery from stalls, drew on their extensive experience. They did not have to 'go back' to get it.

That is what I understood also, and therefore why I was querying guiones on what he thought was done, since he was saying a lot of further testing and calibration was performed (he apparently knows some of the participants, but hasn't yet said).

Perhaps you are confusing (type) certication with the UK CAA practice for periodic renewal (from memory: every 5 years) of the C of A of individual 'Performance Group A' aircraft on the UK Register

Yes, perhaps I am. Perhaps Airclues can clarify which he meant?

PBL

HazelNuts39

29th Jul 2010, 14:09

You can obviously do that in both Normal and Alternate Law (that is, Alternate 1) on an Airbus without actually reaching, let alone going beyond, AoA for C_L_Max (and I mean the real C_L_Max).
I didn't think the discussion was limited to Airbus A320/330/340/350/380. Can you really authoritatively state that Airbus never stalled any of those airplanes beyond the g-break? Bear in mind that certification is not limited to normal configurations and conditions but also considers icing, system failures and their consequences and probabilities (ref. FAR 25.1309 etc), turning flight stalls, accelerated stalls, low and high altitudes ... What about other manufacturers?

Anyway, I have considerable difficulty in accepting your statement that "In other words, actually high-alpha-stalling large commercial aircraft, even for certification, is ancient history." From my perspective, it doesn't do justice to the rigidity, thoroughness and expense in money and manpower of the type certification process of large airplanes.

regards,
HN39

Mad (Flt) Scientist

29th Jul 2010, 16:08

I didn't think the discussion was limited to Airbus A320/330/340/350/380. Can you really authoritatively state that Airbus never stalled any of those airplanes beyond the g-break? Bear in mind that certification is not limited to normal configurations and conditions but also considers icing, system failures and their consequences and probabilities (ref. FAR 25.1309 etc), turning flight stalls, accelerated stalls, low and high altitudes ... What about other manufacturers?

In most cases, the full suite of certification test for stalling are not required for failure conditions and the like, since it is assumed that either (a) a procedure will be applied which will reduce the probability of a full stall sufficiently or (b) the probability of the failure itself, in combination with the stalling probability, is so low as to be acceptable.

The most obvious example is failure of a stall protection system - obviously, if the aircraft needs such a system in order to be compliant with the basic requirements, it will never be able to meet those same requirements with that system failed. And it is in general not required to do so; the criteria for "Continued Safe Flight and Landing" - the condition to be demonstrated subsequent to the failure - are less stringent that the full certification requirements.

PBL

29th Jul 2010, 17:44

I didn't think the discussion was limited to Airbus A320/330/340/350/380. Can you really authoritatively state that Airbus never stalled any of those airplanes beyond the g-break?

I don't work for Airbus, and their flight test program is likely proprietary information, so I am not in a position to state authoritatively anything about it at all.

Anyway, I have considerable difficulty in accepting your statement that "In other words, actually high-alpha-stalling large commercial aircraft, even for certification, is ancient history." From my perspective, it doesn't do justice to the rigidity, thoroughness and expense in money and manpower of the type certification process of large airplanes.

Opinions may differ and yours is a reasonable point of view (with which I do not necessarily agree). But I am wondering whether I overstated the case and am inquiring. The wording of statements causes some problems here, as I pointed out in my blog note. Here is a further example: BOAC made a valid point that if you are "fully stalled" you don't want to try lifting wings with aileron. But he meant something different from CS 25, for which an Airbus holding Alpha_Max (less than C_L_Max) for two seconds counts as "fully stalled". I prefer (what I take to be) BOAC's meaning, but you can't argue with regs; if they say "chicken and onion pie" means the left windshield pane, then the left windshield pane it is, and we have to find another word for our lunch.

PBL

john_tullamarine

29th Jul 2010, 20:02

Your reference 1 will be quite useful for those here who don't know what is taken to determine a "full stall".

One of the problems with these sorts of discussions for the newchums is that there is a very important caveat which MUST be considered - the Design Standards (which include stall protocols) are living animals. Over the years the Standards evolve and what was the rule yesterday can change quite significantly today.

Hence, for all Type specific discussions going beyond a bit of a yarn on PPRuNe, one MUST

(a) check the TCDS for the relevant frozen Design Standard revision(s) applicable to the particular certification

(b) refer to that/those (generally now superseded) documents rather than the current words

For US flight test interpretation, we need to look at the relevant revision either of 25-7 or 23-8.

Most importantly - if one tries to run a stall iaw the "wrong" Standard revision, one can be in for a short sharp wakeup call when the bird does something totally unexpected.

For instance, I can recall a very experienced TP relating a tale about a particular popular light cabin class twin. The original stall protocols went to the initial stall with a prompt recovery. During a training period with a TP student, the student thought it might be a good idea to hold the aircraft into the stall .. whereupon they found themselves promptly in an inverted spin .. there are many traps for young players if folk fancy themselves as putative TPs without having put in the hard yards with the TP training and post grad exposure/experience.

PBL

29th Jul 2010, 20:48

Hence, for all Type specific discussions going beyond a bit of a yarn on PPRuNe, one MUST

(a) check the TCDS for the relevant frozen Design Standard revision(s) applicable to the particular certification

(b) refer to that/those (generally now superseded) documents rather than the current words

Yes, indeed!

In fact, I'm dealing with such issues this very week.

PBL

safetypee

29th Jul 2010, 21:59

One of the problems in this matter has been the poor wording in Manuals and similarly, yet different, the view taken by the FAA – three views at least, certification, regulatory, and training/checking.
If the objective is to train to the approach to the stall, as may have been the case, then this requires specific aspects of awareness and flight techniques.
With more recent focus on Loss of Control, then training has been diverted to ‘upset recovery’ vice a stalled aircraft. This training is very context dependent.
Only now is stall training being considered again in detail, and this too depends on context and may require specific awareness and ‘aircraft type dependent’ techniques.

Attempting to unravel poor descriptions, undo weak training practices, and simultaneously introduce new programs will be fraught with problems – thus we have arrived.

As much as BOAC might not like certification regulations ;), a good understanding of these will aid training requirements and an understanding of manufacturers’ recommendations.
Although the requirements in CS25 / FAR25 etc, might not be in accord with academic theory, they are the basis on which particular aircraft have been judged – what aircraft do in practice and not what they should do in the classroom.

Of the commercial types which I was associated with, the stall evaluation was extensive. It progressed well beyond any training stall or concept of an upset manoeuvre, in search of the academic T tail deep stall (wind tunnel predicted), but which as far as we know, is not encountered – (from Vss, hard stick back, 5 deg/sec pitch rate, resulting in at least 5 deg stall alpha over-swing).
Turning stalls did induce a wing drop, for which lateral control can be used to aid recovery (as per CS25), but this may not be generic advice as control response in such conditions will be type dependent.
It’s nice to ensure that the inner wing stalls before the tip and thus enable conventional ailerons to be used.

Of course things don’t always go as planned. With development and increasing weight the stalls became more variable with a notable roll off. Extensive aerodynamic instrumentation failed to disclose a reason and the wt/cg certification was limited by the maximum roll angle with use of full control (aileron and rudder) in the stall.
Subsequently, slight wing twist at high weight has been identified as the likely problem.
None of the above would be of direct relevance to a training program, except perhaps a better understanding of why this manufacturer in this particular aircraft type includes the use the aileron or rudder control as required to control roll angle during a stall recovery.

IMHO the issue which the regulators (training) should be looking at is why aircraft are deviating from safe flight into the speed/manoeuvre safety buffer prior to stall warning and an actual stall; what’s generating these problems.
Also, as noted by BOAC, understand trim; there are some significant differences both in theory and application amongst modern aircraft types.

john_tullamarine

29th Jul 2010, 22:34

three views at least, certification, regulatory, and training/checking

And that's been a sticking point for years probably in just about every Contracting State.

The training paradigm traditionally has been about minimum loss of height without regard to how the design, certification and FT processes intermesh. The need for a reduction in alpha to get away from the main stalling problem as a first action appears never to have got into the training regime's thought processes and that has been the main safety disconnect in the whole thing ....

(stands by for JF to wade in with technical TP stuff ...)

BOAC

30th Jul 2010, 07:38

SP - ;) noted. I have absolutely nothing against the certification requirements - there has to be a base-line, be it right or in need of 'tweaking', against which a/c can be judged. My comments on this thread against the 'theoreticians'/certificators' has been that we are seeing more and more near or complete LOC at excessive pitch angles which arrive at FULLY stalled a/c.

I know from airtests that you cannot properly 'stall' a 737 in line with the airtest schedule- there is no more 'up elevator' available to stall the wing and we settle for declaring a benign gentle nose drop as the 'point of stall' when in fact it represents the limit of pitch control, so the 1g clean stall really has merit only in checking that the stall warning functions correctly and that there are no adverse excursions in roll, yaw or pitch NEAR the stall. The actual airtest stall has no relevance to the situations in which crews seem to be finding (?placing?) themselves.

Unfortunately the 'training system' has followed this philosophy and in my view, as I have often said, we need to thoroughly overhaul the training system for the glass cockpit and I completely endorse SP's last paragraph. Wittingly or unwittingly we have allowed ourselves to be 'automated' into a world where 'it can't happen'. Wandering only slightly off topic, PJ2's post on the TIP crash says "It is just an airplane, and it requires flying skill; it is not a platform which requires mere "managing" and this is where I see the whole system failing - it encourages crew to 'manage' a complex and clever system and seems to forget the fact that when you press a few big red buttons the a/c and crew SHOULD be capable of just flying the a/c. The very point of my thread on the Safety Forum on the 'Computers in the cockpit etc'.

PBL

30th Jul 2010, 16:17

My comments on this thread against the 'theoreticians'/certificators' has been that we are seeing more and more near or complete LOC at excessive pitch angles which arrive at FULLY stalled a/c.

One of the difficulties with making comments "against" specific people is that what starts out as a legitimate difference in emphasis can turn into an artificial polarity, and I would rather not see this happen in this case because this is a complex set of issues with plenty of potential for confusion.

That said, I would like to address, first, the issue of whether "we are seeing more and more near or complete LOC...", and, second, whether it is possible to train for it, or whether there is no practical alternative to training for avoidance.

This morning, for lack of anything better to do after having read DO-178A a few times and commented on it to others, I went into Flight International's annual safety reviews for 2007, 2008, 2009, to identify LOC with stalls. I just looked at fatal accidents with scheduled passenger flights, non-scheduled passenger flights, freight flights, and regional/commuter flights (FI's categories).

Let's look at the commercial large aircraft stuff first. If it's OK, I won't count Antonov 12's because there are a lot, but I will count the other Antonov's.

2007: Total:11

Passenger:
01.01 Adam Air, B737, Indonesia, Departure from cruise flight
09.01 Aeriantur, An-26B, Iraq, Crashed on Approach
07.03 Garuda, B737, Yogyakarta, Runway overrun
17.03, UTAir, Tu134A, Samara, Landed short
05.05 Kenya, B737, Cameroon, Descent into terrain after TO
28.06 Angola, B737, Angola, Landing accident
17.07 TAM, A320, Sao Paolo, Runway overrun

Non-passenger:
23.07 Djibouti, An-26, Ethiopia, Engine failure on climb out
26.08 Great Lake Business, An-32, Congo, Engine failure on climb out
20.09 Arctic Circle Air, Shorts Skyvan, Mystic Lake, Alaska, failed to climb on ferry flight after damage
04.10 El Sam, An-26, Congo, Engine failure after TO

2008: Total:10

Scheduled Passenger
15.04, DC-9, Congo, Runway overrun
30.05 TCAC A320, Tegucigalpa, Runway overrun
10.06 A310, Sudan, fire after landing
20.08 Spanair, Madrid, TO accident
24.08 Itek B737, Bishkek, Approach to Landing
14.09 Aeroflot Nord, B737, Perm, Approach to Landing

Non-passenger
30.06, Ababeel, Il-76, Sudan, Crashed after TO, perhaps engine fire
06.07 USA Jet Airlines, DC-9, Saltillo, Mexico, "crashed on a road 800m from airport". Not said whether TO or landing
30.08 Conviasa, B737, Ecuador, CFIT
27.11 XL Airways, A320, Perpignan, LOC during acceptance testing

2009: Total:14

Scheduled Passenger
25.02 Turkish, B737, Amsterdam, Approach to Landing
12.02 Colgan, Q400, Approach to Landing
09.04, Aviastar Mandiri, BA146, Indonesia, CFIT on approach to landing
01.06 Air France, A330, Departure from Cruise Flight
30.06 Yemenia, A310, near Moroni, Approach to Landing
15.07, Caspian, Tu 154M, Departure from Cruise Flight
24.07, Aria, Il-62M, Runway overrun
04.08 Bangkok, ATR-72, Thailand, Lost control after landing

A Q400 counts as a large commercial for my purposes here.

Non-passenger
09.03 Aerolift, Il-76, Entebbe, Crashed on TO
23.03 FedEx, MD-11 Tokyo, LOC on landing
26.05 Service Air, An-26, Congo, Approach to Landing (short final)
21.09, Sudan Airways, Boeing 707, Sharjah, crash after TO
01.11 Russian Interior Ministry, Il-76, apparent LOC on TO
28.11 Avient, MD-11, Shanghai, LOC after tailstrike on TO

2010:

It's a week too early for David's list, so let me guess

25.01 Ethiopian, B737, nr Beirut, Departure from Climbout after TO
12.05 Afriqiyah, A330, Tripoli, Approach to Landing
22.05 Air India Express, B737, Mangalore, Runway overrun
28.07 AirBlue, A321, Islamabad, Approach to Landing

Now, which of these involved full stalls at high AoA whose recovery would have avoided the accident?

2007: Adam Air? Questionable: they had lost control before this because they weren't paying attention to flying the airplane
Ariantur? Who knows?
Kenya? There were other things wrong

2008: Spanair? Hardly; better to have your high-lift devices out
Bishkek? Who knows?
Perm? The MAK said otherwise
XL Airways

2009: Turkish. The report said that there was no altitude for stall recovery. There was lots wrong with their handling before the AC stalled
Yemenia? Who knows?
Colgan. The aircraft was stalled from controlled flight by the PF's actions.

So I will go for XL Air and Colgan, and maybe Turkish. That's 3 out of 39. That doesn't seem like a rash of high-AoA-LOC accidents to me.

Now, there is nothing wrong with anticipating a trend and thinking of what to do about it. But to me the trend is not yet clear.

Now, what to do about it? Train for high-AoA-LOC recoveries? Well, I don't think that's feasible, but what do I know, I don't fly these big high-performance airplanes. But I do talk to and work with people who have designed and analysed them.

I don't think it is feasible because what an airplane does in these regimes is highly specific and not really known. Real upsets are not flight-tested; this comes from wind tunnels. Simulators are no help, because outside the envelope there is no guarantee of veridicality. So, if you can't train for it, you can only train to avoid it. Here is the view of William Wainwright, Airbus Chief Test Pilot, in Issue 24 of FAST:
We manufacturers were very concerned over the types of manoeuvres being flown in simulators and the conclusions that were being drawn from them. Simulators, like any computer system, are only as good as the data that goes into them. That means the data package that is given to the simulator manufacturer. And we test pilots do not deliberately lose control of our aircraft just to get data for the simulator. And even when that happens, one isolated incident does not provide much information because of the very complicated equations that govern dynamic manoeuvres involving non-linear aerodynamics and inertia effects. The complete data package includes a part that is drawn from actual flight tests, a part that uses wind tunnel data, and the rest which is pure extrapolation. It should be obvious that firm conclusions about aircraft behaviour can only be drawn from the parts of the flight envelope that are based on hard data. This in fact means being not far from the centre of the flight envelope; the part that is used in normal service. It does not cover the edges of the envelope. I should also add that most of the data actually collected in flight is from quasi-static manoeuvres. Thus, dynamic manoeuvring is not very well represented. In fact, a typical data package has flight test data for the areas described in Table 1. In other words, you have reasonable cover up to quite high sideslips and quite high angles of attack (AOA), but not at the same time. Furthermore, the matching between aircraft stalling tests and the simulator concentrates mainly on the longitudinal axis. This means that the simulator model is able to correctly reproduce the stalling speeds and the pitching behaviour, but fidelity is not ensured for rolling efficiency (based on a simplified model of wind tunnel data) or for possible asymmetric stalling of the wings. Also, the range for one engine inoperative is much less than the range for all engines operating and linear interpolation is assumed between low and high Mach numbers. Wind tunnel data goes further. For example, a typical data package would cover the areas described in table 2. In fact, this is a perfectly adequate coverage to conduct all normal training needs. But it is insufficient to evaluate recovery techniques from loss of control incidents. Whereas, the training managers were all in the habit of demonstrating the handling characteristics beyond the stall; often telling their trainees that the rudder is far more effective than aileron and induces less drag and has no vices! In short, they were developing handling techniques from simulators that were outside their guaranteed domain. Simulators can be used for upset training, but the training should be confined to the normal flight envelope. For example, training should stop at the stall warning. They are “ virtual” aircraft and they should not be used to develop techniques at the edges of the flight envelope. This is work for test pilots and flight test engineers using their knowledge gained from flight testing the “ real” aircraft.

PBL

BOAC

30th Jul 2010, 20:53

You ignored the word 'near' in my post? An impressive list of irrelevant accidents. There have been several 'near' LOC in recent years which did not result in accidents - shall we ignore them?

Moving on:
Turkish:
"The report said that there was no altitude for stall recovery." Quite irrelevant to my point - should not have been there.
"There was lots wrong with their handling before the AC stalled" Totally relevant to my point. Likewise PGF, BOH, LHR, CPH, etc etc

"Train for high-AoA-LOC recoveries? Well, I don't think that's feasible, but what do I know, I don't fly these big high-performance airplanes. But I do talk to and work with people who have designed and analysed them." yes it is feasible (and done sometimes by those who understand) and you really should talk to the people who do fly them if you are to engage in this discussion on piloting.

"Real upsets are not flight-tested; this comes from wind tunnels." Unfortunately becoming untrue - quite a few pilots are inadvertently 'flight testing'

"So, if you can't train for it, you can only train to avoid it." Well, (1) you can - and (2) yes you must

WW:
"Whereas, the training managers were all in the habit of demonstrating the handling characteristics beyond the stall;" They obviously believed what the sim sales and design people told them?

"often telling their trainees that the rudder is far more effective than aileron and induces less drag" Correct on the real thing. I know not in the sim. The only roll control you have in the real aircraft and if correctly applied it works and is NOT dangerous as some seem to think. Wrongly used = big vices.

Lastly I am watching a trend, not anticipating one. Do your analytic programmes allow you to factor in reduced training, experience, increasing subservience to automation, less 'piloting' and rest to accident prevention?

The big 'poke you in your eyes'/'kick you in the balls' message right now is TRIM

PBL

30th Jul 2010, 21:34

BOAC,

I took your proposal seriously enough to spend a number of hours today trying to figure out if it was true or not. I thought the results I obtained were worth broadcasting, as a contribution to discussion. I'm sure they are not the end of the story.

I can't at present join the dots in your argument that people have to be trained to recover upsets in large commercial airplanes. I just can't see how that can realistically be done. Neither, apparently, can Airbus's main man. If you have a proposal, please do bring it up for discussion.

PBL

punkalouver

30th Jul 2010, 22:26

BOAC,

I can't at present join the dots in your argument that people have to be trained to recover upsets in large commercial airplanes. I just can't see how that can realistically be done. Neither, apparently, can Airbus's main man. If you have a proposal, please do bring it up for discussion.

PBL

Quite simple actually. The training would be done in groundschool.

BOAC

31st Jul 2010, 07:46

it is apparent, to me anyway, that 'stall' recoveries MUST now lead straight into the nose-high low-speed recoveries as an exercise, meaning that I would suggest allowing trimming right down to stick shake. Recovery from these conditions is indeed simple if you are prepared for it and take the correct actions.

Yes, groundschool to introduce the topic - excellent idea - but in the sim it is so easy. None of the 'unusual positions' I flew a few years back on a Boeing triggered sim exercise were done 'out-of-trim' but could have been. Better to see the a/c stand on its tail in a box 8 ft off the ground that can be 'frozen' than in an a/c 800ft up that cannot.

PBL

31st Jul 2010, 08:47

OK, BOAC, you are concerned about the specific condition in which a pilot executing a vertical escape manoeuvre of some sort is unaware of the position of the longitudinal trim. There are two ways in which heshe could be unaware, namely (a) it is displayed and available, but not noticed, or (b) the position is not displayed in a readily-assimilable manner during the manoeuvre.

Situation (a) is what I call an attention failure (the "A" in the PARDIA flow-of-information classification). Situation (b) is a violation of what I call the Mutual Cognition of Relevant Parameters (MCRP) Criterion in my IET System Safety Keynote last year (Securing the Interface; I just noticed that we failed to put a copy on the WWW. I'll remedy that later). I think you are concerned about both situations, but I feel it appropriate to distinguish them.

You suggest it's a trend. It may be, or it may not be, I can't tell. What would be very helpful is to have a definitive list of situations in which (b) has occurred. I listed fatal accidents in the last four years in my post yesterday; maybe, if you have the time and motivation, you could write a definitive list of those incidents in which situation (b) has occurred and has arguably adversely affected the safety of flight. I think from my list only Turkish and XL Airways fit (there may be some speculation concerning others).

This is not an old concern. It arose during the Nagoya accident to a China Air Lines A300 in 1994. The report also identified other upsets (over Paris; over Moscow) in which automated trimming of the THS had led to pitch-ups and high-AoA situations in this line of aircraft (A300/310).

I advised counsel for the plaintiffs in the civil proceedings concerning the Nagoya accident in 1998 (the trial was in 1999, I believe). They didn't use my advice in court AFAIK, but what I advised them was identical with what the court finally found.

You point out, though, that it is not only Airbus aircraft which engender situation (b). Which other aircraft are subject to it?

PBL

BOAC

31st Jul 2010, 10:14

Well, I have the time but after your second paragraph I'm fresh out of motivation:)

Centaurus

31st Jul 2010, 10:29

2009: Turkish. The report said that there was no altitude for stall recovery. There was lots wrong with their handling before the AC stalled

Not quite sure about your assessment on that accident. From what I read of the accident the thrust levers closed because of a fault in the radio altimeter and the crew simply watched it happen. They had plenty of altitude to recover when the airspeed was already approaching the stall but were obviously nonplussed and took no action until too late. In the 737 simulator we have reproduced the sequence of events and at 300 ft above runway level at stick shaker with an awful lot of back trim already applied by the autopilot, a competent pilot could recover providing he got on to the problem promptly and was not afraid to use significant stabiliser trim to aid elevator control.

Mansfield

31st Jul 2010, 10:52

While you guys are duking it out over whether a trend exists with regard to stall accidents, I thought I'd just point out that at the rather large joint conference in Toronto next week, AIAA will host three sessions which may be enlightening. I've taken the liberty of posting the agendas for these sessions below. I would suggest that the existence of these sessions, and the amount of effort that has gone into the development of the papers to be presented, rather strongly supports BOAC's notion that a) this is a problem, and b) a lot can be done about it.

I'll be there, but I'm funded. Anyone else?

Aircraft Loss of Control I: The Problem and Technical Challenges:

2:00 PM-2:30 PM

AIAA-2010-8002. Loss of Control – Avoidance, Recognition and Recovery: Reducing the Leading Cause of Accidents J. M. Cox
2:30 PM-3:00 PM

AIAA-2010-8003. Accident Lessons for Stall Upset Recovery Training D. A. Crider
3:00 PM-3:30 PM

AIAA-2010-8004. Aircraft Loss-of-Control Accident and Incident Analysis C. M. Belcastro
3:30 PM-4:00 PM

AIAA-2010-8005. Loss-of-Control: Perspectives on Flight Dynamics and Control of Impaired Aircraft H. G. Kwatny; J. Dongmo; R. Allen; B. Chang; G. Bajpai No itinerary selected
4:00 PM-4:30 PM

AIAA-2010-8006. Human Factors of Aircraft Loss of Control Accidents S. Casner No itinerary selected
4:30 PM-5:00 PM

AIAA-2010-8007. Aircraft Loss of Control Causal Factors and Technical Challenges S. R. Jacobson
5:00 PM-5:30 PM

AIAA-2010-8008. Technical Challenges of Upset Recovery Training: Simulating the Element of Surprise J. S. Burki-Cohen
5:30 PM-6:00 PM

AIAA-2010-8009. Results of a Comprehensive In-Flight Simulation-Based Advanced Maneuver & Upset Recovery Training Study J. Priest

Aircraft Loss of Control II: Potential System Solutions

9:00 AM-9:30 AM

AIAA-2010-8139. Some Thoughts on Reducing the Risk of Aircraft Loss of Control D. Bateman
9:30 AM-10:00 AM

AIAA-2010-8140. Examination of Icing Induced Loss of Control and its Mitigations A. L. Reehorst; H. E. Addy; R. O. Colantonio
10:00 AM-10:30 AM

AIAA-2010-8141. Development and Implementation of a Model-Driven Envelope Protection System for In-Flight Ice Contamination D. R. Gingras; B. Barnhart; R. Ranaudo
10:30 AM-11:00 AM

AIAA-2010-8142. Future Concepts for Preventing Aircraft Loss-of-Control Accidents C. M. Belcastro; S. R. Jacobson
11:00 AM-11:30 AM

AIAA-2010-8143. Validation and Verification of Future Integrated Safety-Critical Systems Operating Under Off-Nominal Conditions C. M. Belcastro

Aircraft Loss of Control III: Upset Simulation and Training

2:00 PM-2:30 PM

AIAA-2010-7791. What Really Can Be Done in Simulation to Improve Upset Training? S. K. Advani; J. A. Schroeder; B. Burks
2:30 PM-3:00 PM

AIAA-2010-7792. Simulation Modeling for Off-Nominal Conditions - Where Are We Today? G. H. Shah; J. V. Foster; K. Cunningham
3:00 PM-3:30 PM

AIAA-2010-7793. Improvement of Stall-Regime Aerodynamics Modeling for Aircraft Training Simulations D. R. Gingras; J. N. Ralston
3:30 PM-4:00 PM

AIAA-2010-7794. Developing Scenarios for Research into Upset Recovery Simulation L. ****e; M. Grigorev; V. Biryukov; E. Groen
4:00 PM-4:30 PM

AIAA-2010-7795. Approach to Stall Training in Simulators D. Carbaugh
4:30 PM-5:00 PM

AIAA-2010-7796. Training to Prevent Upset J. Drappier
5:00 PM-5:30 PM

AIAA-2010-7797. Ground Based Simulation of Airplane Upset Recovery Using an Enhanced Aircraft Model F. Liu ; P. R. Grant
5:30 PM-6:00 PM

AIAA-2010-7798. Effectiveness of Sustained G Simulation in Loss of Control and Upset Recovery Training S. Glaser; P. Comtois

AIAA-2010-8141. Development and Implementation of a Model-Driven Envelope Protection System for In-Flight Ice Contamination D. R. Gingras; B. Barnhart; R. Ranaudo
10:30 AM-11:00 AM

AIAA-2010-8142. Future Concepts for Preventing Aircraft Loss-of-Control Accidents C. M. Belcastro; S. R. Jacobson
11:00 AM-11:30 AM

AIAA-2010-8143. Validation and Verification of Future Integrated Safety-Critical Systems Operating Under Off-Nominal Conditions C. M. Belcastro

BOAC

31st Jul 2010, 11:05

Mansfield - excellent news. Thank you. I did not know about this.

PBL

31st Jul 2010, 18:33

Centaurus,

Not quite sure about your assessment on that accident.

I said: 2009: Turkish. The report said that there was no altitude for stall recovery. There was lots wrong with their handling before the AC stalled

Let me rephrase. The report said on pp 6-7 that when the aircraft stalled "the height remaining ... was insufficient for a recovery". There was lots wrong with their handling before the AC stalled. There were also other things wrong which did not concern their handling. Better?

Mansfield,

I didn't think we were talking about LOC accidents in general. As far as I can tell, BOAC is mainly concerned, not about LOC accidents in general, but about a specific kind of LOC accident. But if we are, then of the list I gave, many are LOC accidents. Let me specifically indicate those I think are LOC, although there is room for quibbling:

2007:
01.01 Adam Air, B737, Indonesia, Departure from cruise flight
09.01 Aeriantur, An-26B, Iraq, Crashed on Approach
05.05 Kenya, B737, Cameroon, Descent into terrain after TO
23.07 Djibouti, An-26, Ethiopia, Engine failure on climb out

2008:

20.08 Spanair, Madrid, TO accident
24.08 Itek B737, Bishkek, Approach to Landing
14.09 Aeroflot Nord, B737, Perm, Approach to Landing
30.06, Ababeel, Il-76, Sudan, Crashed after TO, perhaps engine fire
06.07 USA Jet Airlines, DC-9, Saltillo, Mexico, "crashed on a road 800m from airport". Not said whether TO or landing
27.11 XL Airways, A320, Perpignan, LOC during acceptance testing

2009:

Scheduled Passenger
25.02 Turkish, B737, Amsterdam, Approach to Landing
12.02 Colgan, Q400, Approach to Landing
01.06 Air France, A330, Departure from Cruise Flight
30.06 Yemenia, A310, near Moroni, Approach to Landing
15.07, Caspian, Tu 154M, Departure from Cruise Flight
04.08 Bangkok, ATR-72, Thailand, Lost control after landing
09.03 Aerolift, Il-76, Entebbe, Crashed on TO
23.03 FedEx, MD-11 Tokyo, LOC on landing
26.05 Service Air, An-26, Congo, Approach to Landing (short final)
21.09, Sudan Airways, Boeing 707, Sharjah, crash after TO
01.11 Russian Interior Ministry, Il-76, apparent LOC on TO
28.11 Avient, MD-11, Shanghai, LOC after tailstrike on TO

2010:

25.01 Ethiopian, B737, nr Beirut, Departure from Climbout after TO
12.05 Afriqiyah, A330, Tripoli, Approach to Landing

That is 24 from 39. Over half and just under 2/3 of all fatal accidents to large commercial aircraft in the period. Even if I take out those about which I am unsure, there are still 18 sure LOCs, just under 1/2. It is widely agreed that LOC has taken over from CFIT as being the largest fatal-accident category. It is also an issue for the military. NASA has been working on LOC for a while, I understand, but I didn't know Don Bateman was working on it also.

BOAC has stressed, if I understand him right, that he is most interested in high-AoA departures which involve automatic trim movement not necessarily clearly annunciated to the pilot. As I said earlier, that appears to be 2 or 3 of those 39.

BOAC,

I have the time but after your second paragraph I'm fresh out of motivation

Pity.

PBL

Pugilistic Animus

31st Jul 2010, 19:11

I said this before: the stall is the most dangerous thing in aviation,...and at the same time it really is not:)

my final answer

BOAC

31st Jul 2010, 19:25

As far as I can tell, BOAC is mainly concerned, not about LOC accidents in general, but about a specific kind of LOC accident. - yes! Posted in #63 and for those who slept through it again in #93. This thread is where theory and practice part company and why my 'motivation' is 'diverging on the negative axis' (=getting less).

'Control' to those who operate these machines simply means flying from A to B and touching down at B in such a way as to not puncture any lips or break teeth. Failing to achieve that at B is generally a 'loss of control' somewhere along the line, be it through sabotage, incompetence, structural failure, CFIT or whatever. All your list qualify. However, my particular concern now has been focussed here on the pitch problem.

'Mansfield's conference' would appear to be in accord. Perhaps try to get a transcript of the second event (3pm) on day 1 and that at 5:30pm and the next. Then have a day off and pick it up at 2:30 on day 3 and stay with it.

Mansfield - If there is any link to a summary afterwards I would be grateful.

BOAC

2nd Aug 2010, 15:42

IGH - my initial 'simple' solution (which will not suit all, I know.........) is based on the fact that we need to retain the ability to trim ALL stages of flight, including very low speed, but that we should have:-
a) A warning that we have gone past 'sensible' with a possible 'stop'
b) A requirement to actively over-ride the stop when needed.

It appears to be the 'un-noticed' extreme trim position that is causing the problem, so let's 'notice' it, loud and clear and obviously?

This is not a big technological challenge.

Pugilistic Animus

2nd Aug 2010, 18:24

Mad (Flt) Scientist

4th Aug 2010, 16:18

(p-pinfinity)/(1/2rhoV^2)/Cpi=1/[1-M^2]^.5+ [M^2/1+{1-M^2}^.5]*(cpi/2)

or maybe just 1/[1-M^2]^.5

no further comments:\:oh:

:8

Actually, no, mot just the "standard" Mach effect. Stall AOA ends up being a powerful function of freestream Mach, due to transonic effects in the (very accelerated) boundary layer. As a result the freestream Mach is not the Mach directly having the effect.

What you're alluding to is fine at more 'normal' AOAs.

Pugilistic Animus

5th Aug 2010, 07:03

I guess experimental lift slope curve data is the only way to go; although I did ASSume at least part of the leftward alpha shift of those curves with increasing M would have been due to that relation, and of course in this case Mach and Reynold's numbers intertwine inseparably :\...it must be a true supercritical design...I don't suppose you could provide the NACA designation for the airfoil:oh::O..would a Cm change also affect the Alpha effective?

:)

error_401

5th Aug 2010, 15:30

For the out of trim situation.

The ATR I flew had a very valuable feature. Every time the trim was actuated more than a couple seconds the trim whooler would give an aural warning. Very useful with A/P on when slowing down. Always triggered looking at the trim and energy state of the aircraft. You knew immediately that something was changing more rapidly than it should.

error_401

5th Aug 2010, 17:02

And about how pilots (for instance me) sense stall:

At the bottom I have listed two references. The FAR reference shows something about the discussion which might be of interest. In fact the statement is that there has to be a warning whatsoever to indicate a possible high AoA situation before C/L MAX. It also states that a stall warning shall be given 7 percent over the determined stall speed.

The whole certification never goes over the absolute limit. No one wants to crash a multimillion testbed. Interesting to learn from the certification documentation that the aircraft hast to be controllable by use of the normal devices up to certificated stall.

IMHO this means we're safe to recover as long as we react with the first warning and do not take it over to fully stalled wing.

It also states IMO the two possibilities to fulfill the requirements for stall testing whereas when fitted with a stall identification system there is no need to take the plane over C/L MAX.

The aircraft I flew and actually fly did and do not show aerodynamic buffet at high AoA. Both were fitted with STICK SHAKER and STICK PUSHER for that particular reason. Previous type gave no aerodynamic indication. Actual type does have an aerodynamic buffet inducer on leading edge - still for some reason it has been fitted with both warning devices as well and FCOM does not rely on buffet.

To the discussion about the speeds: When flying IMHO I give a sh... if the warning is IAS/CAS/EAS/M dependent, corrected or not. If one of the clues triggers be it aerodynamic or artificial there is normally no time to wonder if it may have another margin of 4 kts indicated to reach C/L MAX or not because we are at 30'000 feet. The manufacturer had to show that these devices would alert before it got out of hands. Anyways I may be in a speed range where my hairs at the back of my neck stand already up or should.

INDICATION
The only way as indication of impending approach to stall on current type I have is the red and black band coming up from the bottom of my PFD to "indicate" AoA C/L MAX. FCOM states that this is slightly before reaching aerodynamic C/L MAX.

PERCEPTION
To perception comes attitude vs. power vs. change in altitude = experience. If on A/P no direct clues are given. Need to assemble information from 3 different locations PFD/ENG DISPLAY/TRIM BAND. Additional clues could be airspeed noise as it get's silent or missing engine noise.
Handflown it may become mushy and soft.

AS FOR THE SPEED DISCUSSION
The only thing I have on the front end is airspeed - indicated airspeed IMO. I say indicated airspeed because this is what I have. An indication on my PFD of actual measured and corrected for whatever the ADC does speed of the air mass I'm in. And I have a booklet based on mass which tells me what speed is still save (1.3 Vs) and I don't know if it is Stall1g or the older certification and to be honest - I don't care because I have no influence on that whatsoever. Still AoA and the margin above C/L MAX when STICK SHAKER or STICK PUSHER triggers may account for any deviation with altitude M, R number etc. (at least I hope the manufacturer did so)

The aerodynamic buffeting from the leading edge vortex generator may as well be already covered by the STICK SHAKER or STICK PUSHER.

RECOVERY
It looks as I've been one of the lucky guys to train o.k. from the beginning. Lower the nose apply power as far as practicable control roll with ailerons was what we were taught from the beginning. From own experience in stalling a Piper Warrior to recovery from spins I have learned: Piper stalled nicely power off. Was fluttering down like a leaf in one after another secondary stalls and only application of rudder put it wings over. Ailerons were still working fine.
I wanted to go a bit further after CPL/IR training and used an EXTRA and VOTEC aerobatic plane on several occasions to train stall and spin recovery from fully stalled conditions. Obviously a treat in these A/C but the lesson learned was again - NO RUDDER on "impending stall". Any excessive rudder application had me upside down before I could think and made things much worse. Still the ailerons neutral nose down worked to NOT FULLY stall the A/C.

MY POINT
It makes quite a difference if recovering from an approach to stall or a fully developed stall with it's unknowns.

Funny that all manufacturers tell about the same story about stall recovery.
Funny that all aircraft I know disconnect A/P when stall warning is given.
Finally RTFM!

MY QUESTION ABOUT TRAINING
Maybe the trouble with use of rudder is that basic training started to mix up stall recovery with spin recovery at some point?

MY ADVICE
Know your pitch and power for the most important configurations. Know your trim.

MY CONCLUSION
The whole discussion is about stall recovery. But there is no such thing in training. It ends at impending stall except if you take aerobatics or manufacturer defined stall speed!

Simulators are NOT ABLE to represent a fully stalled condition.

Altitude helps - a lot.

It depends if you are approaching a stall and start recovery in time and you are still within the envelope that was tested and agreed upon as: "Shows a behavior which can be handled with normal piloting skills...". If your aircraft whatever type goes wing over and dives at 45 deg nose down you're on your own as a test pilot. I guess that in large transport aircraft the only thing that counts then is good luck and enough energy say altitude below. The day I may have some extra money to spend I'd like to go to the 3D SIM and train that.

To clarify I would put the recommendations into three different sections.

Approach to stall where there IS margin and the first warnings kick in timely. Still the aircraft is flyable and ailerons efficient. You don't know what the margin is so better be quick. Loss of altitude included. You were on the lucky side.
Stalled to the point where we are over the apex of C/L MAX. The relatively sudden drop in lift may cause a lot of trouble such as nose down or it wings over for whatever disturbance which made stalling asymmetrical. Here you are outside the envelope and it will get very interesting. Good luck!
After the primary stall where you find yourself in whatever position possible depending on attitude, energy state, aircraft type and you're out of any tested envelope on the type such as a spin. Good luck!

Some information worth considering:
AERO : Upset Recovery (http://www.boeing.com/commercial/aeromagazine/aero_03/fo/fo01/index.html)
FAR AC 25-7A page 189 ff (http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgAdvisoryCircular.nsf/0/c2614e27b49bf38686256ba300696689/$FILE/AC%2025-7A%20Change%201.pdf)

And some information from NTSB reports on what happens in a full stall. Better be high up.

DCA08FA075 (http://www.ntsb.gov/ntsb/brief2.asp?ev_id=20080703X00974&ntsbno=DCA08FA075&akey=1)

CHI06IA127 (http://www.ntsb.gov/ntsb/brief.asp?ev_id=20060512X00563&key=1)

DCA97MA049 (http://www.ntsb.gov/ntsb/brief.asp?ev_id=20001208X07893&key=1)

DCA97MA016 (http://www.ntsb.gov/ntsb/brief.asp?ev_id=20001208X07150&key=1)

DCA03IA005 (http://www.ntsb.gov/ntsb/brief2.asp?ev_id=20021127X05535&ntsbno=DCA03IA005&akey=1)

EXECUTIVE SUMMARY (http://civilaviation.nic.in/coi-fin/report.htm#FACTUAL%20INFORMATION)

Still I think the training given in the original #1 post is sensible.

Keep the blue up and safe landings.

deefer dog

6th Aug 2010, 00:06

(p-pinfinity)/(1/2rhoV^2)/Cpi=1/[1-M^2]^.5+ [M^2/1+{1-M^2}^.5]*(cpi/2)

or maybe just 1/[1-M^2]^.5

With the greatest of respect we are talking about stall recovery procedures for jet engined aircraft flying at high levels. Those of us who dare to ventue up there use a formulae that better suits our purpose, and one that is understood by those with less than level 2 English.

We know we have to sacrifice something for the screw up that just happened, irrespective of whether the fault was ours or not. If the screw up was a stall at high level....the penalty is altitude.

If we stall at high level, we are going down! Any recovery action should recognise this fact.

PBL

6th Aug 2010, 07:57

Nice post from error_401.

It occurs to me that no one (including me) has yet cited the certification requirements, although I have referred to them. The definition of "stall" is in CS 25 201 (d). See also the accompanying comments in AMJ 25 201. The current document is downloadable from The EASA Certification Specifications WWW page (http://www.easa.europa.eu/agency-measures/certification-specifications.php). I should point out that most transports in use today were certified to forerunners of this document, not this precise document, and that engineers are very particular as to precisely which version of the document a given aircraft was certified to.

Notice that none of the characteristics in CS 25 201 (d) refer to C_L_Max.

Notice that stall characteristics are only required to be demonstrated from (a) level flight, at low speed (so, Vs1g stall), with engine power appropriate for a slow deceleration rate; and (b) a moderate level banked turn under similar conditions.

PBL

BOAC

6th Aug 2010, 08:08

MY QUESTION ABOUT TRAINING
Maybe the trouble with use of rudder is that basic training started to mix up stall recovery with spin recovery at some point? - possible, but what is not properly understood is that there is no other roll control available on a fully stalled wing. If 'use of rudder' put you 'on your back' you used it incorrectly.:) Not quite sure why you were using it "on impending stall"?

As you say, we should never get to the point of 'fully stalled', so aileron usage remains quite practical, but sadly some of our colleagues are managing too. Hopefully the initiatives coming out of the Toronto conference will go a long way to address this.

PBL

6th Aug 2010, 09:39

BOAC,

the Toronto session at the 2010 AIAA Guidance, Navigation and Control annual meeting was just one waypoint in a long history of targeted research on LOC.

Any "initiative" started long before, with targeted research programs. For example, here an article on Loss of Control from the AAIA organ in March 2003 (http://www.aiaa.org/aerospace/Article.cfm?issuetocid=322&ArchiveIssueID=36). NASA Langley's work was well represented at the 2010 AAIA GNC meeting, it seems, as one would expect from such a research program.

I can't find the papers yet on the Langley server, but there appear to have been two contributions from NASA Langley on safety research to the AAIA GNC annual meeting last year (Chicago, 2009). Here is one: Aircraft Accident Prevention: Loss of Control Analysis (http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20090030517_2009030886.pdf), and here is the other: Methodologies for Adaptive Flight Envelope Estimation and Protection (http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20090029978_2009030104.pdf). Neither appears to be directly related to your concerns about unremarked longitudinal trim changes.

PBL

BOAC

6th Aug 2010, 11:13

A helpful contact has provided some useful content from Toronto.

Mansfield

6th Aug 2010, 13:06

The presentations in Toronto were certainly not the first on loss of control. But the landscape has changed in the last year; not only is loss of control clearly identified as the leading cause of accidents in transport category aircraft worldwide, but the lobby efforts brought to bear by the Colgan families have generated a great deal of pressure for change in training.

The former has brought us the original point of this thread, namely, changes to stall recovery procedures from the major manufacturers. The latter has brought us the 1500 hour requirement with the associated training requirements.

As to error 401's post...the whooler on the ATR was definitely a wise idea. Trim motion is a cue that has become seriously deficient in modern designs...I still miss the trim wheel in the 727!

That said, we have to be careful about generalizations:

"Lower the nose apply power as far as practicable control roll with ailerons was what we were taught from the beginning. From own experience in stalling a Piper Warrior to recovery from spins I have learned: Piper stalled nicely power off. Was fluttering down like a leaf in one after another secondary stalls and only application of rudder put it wings over. Ailerons were still working fine." But the ailerons on a swept wing jet may not be working fine. That was a major point made by the manufacturers last week at Toronto. A fully stalled jet transport is laterally unstable and roll control will not correct the problem.

Rather than stomp the rudder, the manufacturers emphasize unloading the wing and "unstalling" it, getting flow reattachment and hence regaining aileron control.

"Go to the gym and get arms strong enough to pitch down even when trimmed pitch up. You should not be flying a turboprop if you are not able to hold ailerons with engine failure at V1. One should really be able to hold against the pitch up moment when applying power on a jet or at least be aware of it."

Again, Airbus pointed out that in the event that the pitch trim has been run up to the stall, so that the airplane is fully trimmed to the speed at which the stall takes place, the elevators may not have adequate authority to overcome the pitch up generated by the application of full thrust on underslung engines. Going the the gym has nothing to do with it...

You really can't mix apples and oranges in many areas, even if you can in others. Single engine trainers are not the best at teaching large jet characteristics...and high wing turboprops do not necessarily behave the same way as jets with underslung engines and swept wings do. This is what gets us in so much trouble...

PEI_3721

6th Aug 2010, 18:03

The trim-in-motion cues are useful; the requirement originated from certification in the days when trims could runaway – they shouldn’t in modern aircraft.
However, the more valuable cue is the trimmed condition of the aircraft. In conventional aircraft this is indicated by the residual control force at constant attitude. If there is an unwanted force then trim can be changed – another cue for trim-in-motion. Most modern autopilot / auto trim systems have low / high airspeed cut-outs; beware those that don’t.

I have seen some very bad trim practices, e.g. trimming during the flare, which if used by habit in other situations can result in unexpected loss of speed – pilots are reluctant to hold a force. IMHO, there are pilots who appear not to understand the principles of ‘trim’, nor have they been taught how to trim.
Trim is often used as a datum about which the aircraft can be manoeuvred. In roll there is no requirement to trim with manoeuvre, but in pitch, if the intent is to change speed, this requires that the trim datum must also change.

Try the following as a LOC sim exercise; with hands-off (autopilot engaged) mistrim in roll up to the max fuel asymmetry limit, as the crew take over disengage the autopilot. A real event – 737, mis handled fuel cross transfer – 120 deg roll!
Repeat with a pitch offset – more difficult to engineer, and the crew’s response should be considerably different – why; compare levels of surprise, expectation, normality.

I put aside modern (unconventional) aircraft which may have different trim philosophies and thus potentially unique problems and solutions.
Anyone recall the Hunter trim follow-up system (aah Hawker), how similar is this concept to the Airbus?
What happened if the follow-up was switched off during a tight turn, use actions for a conventional aircraft, but in an Airbus – look at the ECAM, manual trim?

Mansfield check PMs

PEI_3721

6th Aug 2010, 19:27

Error 401, #125, re “ The whole certification never goes over the limit.”
See #103. Stall certification is one of the most extensive series of tests during aircraft development. Aircraft have to be recoverable in situations up-to, during, and post stall; you should be able to affect a safe recovery at any time (altitude permitting) providing the situation is not mis handled; and therein is a problem.

Re “ wing leading edge device” This may have been a ‘stall breaker’ to deliberately induce a loss of lift on the wing inboard section. For a swept wing, this provides the necessary nose down pitch at the stall – for recovery; it’s not intended as a warning or indication, e.g. Avro RJ.

The EFIS low speed symbols may only be an indication of the airspeed at which stall warning / stall may occur; the level-flight accuracy is generally good, but in manoeuvres errors can increase.
The display is for airspeed, and unless your aircraft has novel weight-CL computation, the stall airspeed may only be an approximation (albeit quite accurate) based on trim curves and alpha, e.g. Avro RJ.

‘If on autopilot’; a modern autoflight system should disconnect the autopilot and autotrim system at stick shake, providing additional warning (AP disc) of the low speed (but not stalled) situation.

‘Recovery’; fly the aircraft (use the controls) as recommended by the manufacturer. If the nose is lowered – AOA reduced, and airspeed is increasing, the wing is producing lift (not much) and the controls should work (not well), the aircraft will not spin. The wing may have stalled, but the fin / rudder will not have, nor the tailplane
Somehow basic training has generated a great fear or phobia about spinning. First the aircraft has to be stalled, and the condition maintained; second there has to be yaw / roll, etc, etc.
There is an incipient phase where recovery may be self evident; in large aircraft, mass / inertia effects may provide additional time, or more likely spin resistance.
Recovery – yes, basically all aircraft are the same.

Have knowledge, use this knowledge, negate bias or false knowledge, control surprise, and follow the manufacturer’s recommendations.
And as you state: -
Funny that all manufacturers tell about the same story about stall recovery.
Funny that all aircraft I know disconnect A/P when stall warning is given.
Finally RTFM! http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/thumbs.gif

“Attitude helps”; yes, in defining situations which have progressed beyond normal operations – the period of identification and avoidance. As above, pilots require more 'thinking attitude'.

“… whatever type goes wing over and dives at 45 deg nose down you're on your own…”, no not so; follow the standard recovery procedures.
Don’t focus on CL MAX, that’s for the classroom, unless the aircraft has a CL gauge.

The aspects of stalling which might not have been tested are generally combinations of control mishandling; also, system failure or extreme weather encounter (including wake turb).
Training stalls can be approached with confidence; an actual stall can be avoided or recovered with similar confidence. Just don’t stall near the ground.
At altitude, don’t rush; the aircraft is still ‘flying’ albeit below the stall speed, so fly the aircraft.

Pugilistic Animus

6th Aug 2010, 20:21

of course those are the nice perfect stalls...don't forget the wing-snapper types...FAR 25.343 has a huge loophole:}:ouch:

Pugilistic Animus

6th Aug 2010, 21:28

or as mentioned before...the secondary stall, caused by incomplete recovery I think adrenaline, if you know to pitch down' will take care of any positive pitchning moments induced by thrust application... I mean If a 100 lb woman can lift a car,...and a stewardess can survive a 35000' fall....you'll overcome it because a seconday stall is guaranteed to be more much more fun than the initial upset:}

or as Langwiesche puts it -to paraphrase- AoA is the difference between where an airplane looks like it's going and where it actually is going...there's plenty of horrible math to back that:\

error_401

7th Aug 2010, 07:35

Thanks for all comments and valuable insights

@BOAC - I fully agree. Fully stalled you are left with rudder only. Now in that situation I would give it a try. We tried both ways into bank = spin out of bank depended on situation and pitch up = nose went up nearly vertical or rolled over. But then again this was in an aerobatic single.
BTW. To have it flip was part of the fun ;) with an EXTRA 300 at 5'000 AGL. We also managed some recoveries.

@ Mansfield - Agree the notion was merely about the forces involved if pitch trim was at the very back. This will be type dependent. Sadly true that maybe elevator has not enough authority. I'll modify it later as it has more to do with asymmetrical flight and applies only to aircraft with conventional controls.

@ PEI 3721 - I should have added: "Never ever give up"

Centaurus

7th Aug 2010, 08:39

A real event – 737, mis handled fuel cross transfer – 120 deg roll!
Repeat with a pitch offset –

Now that surprises me. During type rating training using the 737-300 Cat D simulator, the instructor briefs there will be a 1500 kg difference between wing tanks and asks the pilot to conduct a final approach and land. This is to give the crew the experience at fuel imbalance if it should occur due to lack of attention when cross-feeding.

In our simulator there is a slight roll but easily controlled. Aileron trim is not needed. This backs up the FCTM statement that lateral control is not significantly affected when operating with fuel beyond normal balance limit and that the primary purpose of fuel balance limitations on Boeing aircraft is for structural life of the airframe and landing gear and not for controllability.

BOAC

7th Aug 2010, 09:03

Now that surprises me.- real event, very close to hull loss.

PBL

8th Aug 2010, 16:20

While looking through the WWW for papers related to the 2010 AIAA GNC conference papers on LOC, I came across a RAeS FOG report from 2008 or later (there is no date on it) on upset recovery that I thought might interest some: Airplane Upset Recovery Training; History, Core Concepts, and Mitigation (http://www.safeopsys.com/docs/RAES_URT_MASTER.pdf)

In Section 2.2.8 is a statement highlighted in boldface, about swept-wing aircraft, expressing what many have said here: in upset recoveries, USE OF RUDDER IS generally NOT RECOMMENDED and in some cases expressly forbidden

The working paper was put on the WWW by the company for which J. M. Cox works.

PBL

BOAC

8th Aug 2010, 17:00

I quite agree with that. Unfortunately you have mis or selectively quoted, because 2.2.8 concerns upset recoveries and not specifically stall recoveries, which were covered in previous sections and 2.2.8. commences with 'High speed high altitude flight'

If you read on a little, paras 3.3 1a and 3.3 3 b specifically discuss the way rudder should be used in LOC recoveries. Rudder is also taught as a means of inducing roll at extreme pitch angles durng upset training and from the same document from which you quote

If control provided by the ailerons and spoilers is ineffective, rudder input may be required to induce a rolling manoeuvre for recovery. A small amount of rudder input is sufficient. Excessive rudder applied too quickly, or held too long, can result in loss of lateral and directional control.

The simple message, however, is if you do not know what to do with the rudder AND how to do it, it is best to leave it alone. Applying incorrect rudder is a sure path to a further disaster than the one you are in.

Pugilistic Animus

8th Aug 2010, 20:48

The Rudder is probably only useful in a spin; a spin entered by having applied the rudder and side slipped in the first place:}

The only other thing I can think of is some 'uncertified' mechanical failure that would require as much rolling moment as possible---unfortunately the one example we've seen was that MD-80 jack-screw malfunction on the Stab and was a completely unsurvivable event:(

PBL

9th Aug 2010, 07:41

BOAC,

I am not sure what you are quibbling about. Of course I have selectively quoted - all quoting is selective.

The FOG in 2.2.8 is addressing loss of lateral stability (wing drop) and says rudder use to arrest it is not recommended and may be expressly forbidden. As far as I know, wing drop is most likely to occur during progress into a stall, so, apparently unlike you, I read this comment as being highly relevant to stall.

In Section 3 they talk generally about unusual attitude recovery, and the advice is intended to be clear: destall the wing before anything else.

When talking in Section 3 of unusual attitudes with decreasing AS at high angles of bank in the case in which ailerons and spoilers are ineffective, they do suggest to try a little rudder to induce a roll, and guard this advice quite specifically. Well, of course they say that - what other option do you have if all other roll control is demonstrably ineffective and you need to roll?

I read them as saying: priority number one, destall the wing; if a wing drops, don't use rudder to arrest it, unless you get into high angles of bank while still stalled, and need to roll to drop the nose, in which case a little rudder, briefly, may be required to induce a roll.

Given the propensity of some on this forum to read a quote or an interpretation as the opinion of the writer, rather than of those quoted, let me emphasise that in the case of discrepant readings it is up to the FOG to clarify what they mean, not me. I am just reading what they wrote. But if one wants to be sure what they meant, I do know a couple of people in the group so can ask directly if it's something major.

I do note that in the three places in which they mention rudder use, in only one of them do they mention the possibility of structural damage in injudicious use of rudder. I guess this is because they are primarily talking about very low airspeed situations.

I also note that this is not the only document concerning upset recovery. A U.S. working group also prepared the upset recovery aid, a fairly large document, between 2004 and 2008. It may be found on the FAA WWW site: the Upset Recovery Training Aid (http://www.faa.gov/other_visit/aviation_industry/airline_operators/training/media/AP_UpsetRecovery_Book.pdf), as well as by the FSF.

PBL

jcbmack

9th Aug 2010, 08:17

PA and MFS know what they are doing!

Pugilistic Animus

9th Aug 2010, 17:06

PBL

no plane is certified for rudder reversals or any other rapid control reversals, even at or below Va

http://www.pprune.org/tech-log/372145-forward-slips-jet-safe.html

Pugilistic Animus

9th Aug 2010, 18:00

New (2010) Stall Recovery's @ high altitudes

New Ideas in Aviation:}

HazelNuts39

12th Aug 2010, 11:32

Quote:
Originally Posted by Pugilistic Animus
(p-pinfinity)/(1/2rhoV^2)/Cpi=1/[1-M^2]^.5+ [M^2/1+{1-M^2}^.5]*(cpi/2)

or maybe just 1/[1-M^2]^.5 (end of PA quote)

Actually, no, mot just the "standard" Mach effect. Stall AOA ends up being a powerful function of freestream Mach, due to transonic effects in the (very accelerated) boundary layer. As a result the freestream Mach is not the Mach directly having the effect.

What you're alluding to is fine at more 'normal' AOAs.

To illustrate this further, I would redraw the diagram as shown here:
cL_AoA_M003.jpg (http://docs.google.com/leaf?id=0B0CqniOzW0rjOTAyNjU1ZTYtMGQ2MS00M2Q5LTgwY2EtNDg0Yzl kOThkNTU3&hl=en_GB&authkey=CLjqzvkP)

The (sub-critical) Mach effect cited by PA changes the lift-curve slope below cL-max. For swept wings, one could perhaps use the component of Mach normal to the wing 1/4-chord line in these equations?

With regard to stall AoA, for Mach between M1 (M1 in my diagram is low, say M=0,15) and M2 the flow is sub-critical (or low super-critical without causing shockwaves), and cL-max changes little, if at all. Above M2 the flow at high AoA is sufficiently super-critical to cause (local) shock-induced separation, and then cL-max reduces rapidly with increasing Mach.

Vs1g speeds published for one modern wide-body would seem to indicate M2=0,275 for that airplane in clean configuration.

regards,
HN39

EDIT:: For an experimental illustration of these aspects, see:
NACA Technical Note No. 1390 (http://naca.central.cranfield.ac.uk/reports/1947/naca-tn-1390.pdf)

Pugilistic Animus

13th Aug 2010, 21:23

HazelNuts39---nice handwriting ;)

since that relation is used by the United States, I knew it'd be right:}

there also be an assumption of infinite aspect ratio so vorticity circulation, would be a factor...that alone would account for some divergence from expected, also, there will be some finite measure of slip 'eta' and other Rn effects that would need to be accounted for, let's not forget 'tunnel wall error' of course many partial solutions to this aspect of the NS equation have been published which are helpful in CFD renderings ..of the problem the total problem is so difficult that with out the airfoil ordinates dc/alpha and dalpha/dM relations it would be untenable to precisely predict any composite airfoil characteristics with high accuracy...of course aerodynamics is an experimental science and not theoretical:)... I mean, we can now make a pile of dog cr#p fly if we wanna:}

HazelNuts39

13th Aug 2010, 22:24

Pugilistic Animus;

thanks for your reply. Prandtl-Glauert for normal Mach-component resulted in satisfactory correlation of CL-AoA-Mach with Flight Simulator data.

Gulfcapt

14th Aug 2010, 10:19

Great thread guys! :ok: Speaking from the point of view of an aerodynamics user rather than expert, I sure appreciate the time many have taken to explain things.

I'm with PA on the "new training." This was an item covered during my last initial type in 2005. I think it was a special emphasis item due to that Pinnacle RJ over the US midwest.

FWIW (and it may not be worth a thing...), the Gulfstream 5 has a normalized AoA indicator. Pitch Limit Indicator (PLI) shows up at .70, shaker at .85 and pusher at 1.0. Gulfstream used to publish great LRC tables that utilized AoA, but they ceased that with the G-4.

Best,
GC

Pugilistic Animus

14th Aug 2010, 18:42

Hazelnut39:
Von Karman was Prandtl's student---both, researchers--- especially Von Karman ---with his student Hsue Tsien contributed greatly to our understanding of high speed/supersonic aerodynamics, compressibility and viscosity:)

Gulfcapt:
You are correct!
No new ideas in aviation:}

:)

Ask21

2nd Sep 2010, 08:23

FAA trained or not, it is a fact that using rudder to lift a wing is commonly taught in flying schools in Australia. Judging also by the questions that appear in the PPRuNe Instructor Forum, it happens elsewhere as well. It seems to be handed down from generation to generation of young flying instructors who in turn once they achieve Grade One instructor status teach new instructors.

It has proved impossible to completely eradicate this technique throughout many general aviation flying schools. But you would think when type rating courses are conducted this erroneous teaching would soon be discovered and rectified.

From my training in a sailplane it was common training procedure to induce both stall and spin and recover from it. Stall was induced at 1000 meters /3000 feet by pulling the stick - speed slow down - increased angle of attack. To recover simply put stick forward. To induce spin - give a strong rudder at the close to stall state. That would cause one wing to drop as a stone and give a spin (rotation) To RECOVER from spin wing stick/ailron was in neutral but strong rudder to opposite side of spin to reduce rotation - it did work. Quite a lot loss of height. Does training on jets include recovery from states of spin? Or is there a mix of procedures. Seems to me that rudder should not be used unless you have a completely stalled single wing / spin situation - in that case I don't think ailron will work anyway. If ailron works I guess it would mean that the wing was not completely stalled in the first place. On the other hand it should be known that heavy rudder at close to stall angle /incompete stall could induce a complete stall of one wing and a following spin situation.

It was also common knowledge that a good designed wing would stall from the medial side first, close to the hull. That is to preserve ailron-funcion as long as possible. The lateral part of the wing will be designed with a lower angle of attac than the medial. When stall is approached there will be a shaking due to in the plane as inner part of the wings stall and experience turbulence in stead of laminar flow.

I would recommend pilots to take a training tour in a sailplane to experience a fully stalled situation and spin and practice of its recovery.