Humbling sim experience

[flyburg]

Had an interesting sim session today,

As a result of several high profile loss of control accidents recently my company introduced a special LOC sim session. Interesting training with a lot of background. My particular sim session was done with an ex air force guy. Lot of insight about recovering techniques! In my earlier years I did some acrobatics but no formal training and since then all my flying has been airline type. Aside from the occasional upset recovery during TQ's and occasional type recurrents not much experience in this regime.

One scenario in particular was eye watering/ego killing.

Airplane type: B747-400

Flying at 35.000 with an heavy airplane you climb to 37.000 which is just*
Possible with the weight. Reason for the climb, you are flying just below the clouds in an area of embedded TS and it is bumpy.

Starting the climb(with just a small margin between upper and lower red band but still within normal operating parameters) you notice the the wind shifting to more HW. You anticipate and notice a speed increase and start to reduce power. To no avail, the airspeed keeps inching towards the upper red band. You reduce power more( not excessively) but still, just after level of you get into the upper red band and get an overspeed warning! You reduce power even more, the airspeed drops but still in the upper red band and all of the sudden you get the stick shaker together with the overspend warning!!!

Respect the stick shaker, you add power but there is not much power left and you are on the wrong side of the power curve, increase in power does not increase airspeed so there is no other action than to descend.

As you descend trying to recover from the stick shaker you notice the wind changing again to a TW and increasing(it shifted to a HW during climb, now it shifts back to a TW) correspondingly, you notice the speed to drop even more, causing you to lower the nose even more to get out of the stick shaker.

The TW increases and the speed drops even more! During the entire event you also have the stall buffet. Somewhere, the overspeed warning stops..the stick shacker stops, you still have the stall buffet and your airspeed reads below 30(which is the minimum reading) with a large TW( in excess of 200IAS).

This all happens completely unexpected!!!! And in the timespan of about 30 seconds ( really, to the best of my recollection) you have gone from 37.000 to 22.000.

I'll stop here to see what you can make of this! I have no military training, consider myself an above average pilot), this completely caught me of guard and left me humbled enough to post it here.

Really interested to see what you can make of it and try to figure it out in a short timespan( remember, it happened to me in a short timespan)

I'll post the rest later but interested in opinions

[Hahn]

First thought: what did the temperature do?
Second: In this excessive wind changes you will (in the real atmosphere) have turbulence like hell.
Anyway: If you have to sacrifice altitude for speed, so be it. Bad over Tibet though.

[Check Airman]

Good that your company has learned from the AF accident. I was in the sim 2 months ago, and still doing stalls at 5000ft by chopping the power and watching the speed bleed off.

You've presented an interesting scenario. No doubt the ASI indication of 30kt was due to icing, and the resulting error in the wind readout when the TAS is compared to GS.

A 15000ft loss in altitude is certainly eye opening. I still wonder why some airlines (including my own) still want us to maintain altitude during a stall recovery:ugh:

[Microburst2002]

We had a session with a good unreliable speed scenario, the airplane ended up pitching up a lot, speed decaying. PF was unable to arrest the pitch up. This was a 320. When I prompted him to pitch down he said "I can't, I,m trying"
We had lost quite a few thousand feet already, the airplane slowing down and descending, the stall warning all the time. I remembered the Air France and started to use the THS handwheel to pitch the airplane down. Eventually it worked. The TRE told us later that due to some computer, I don't recall which one, the THS had frozen!

We learnt a lot.

By the way I became to two conclusions:
1- the most difficult part is realising that speed is unreliable
2- when speed is unreliable you don't know for how long it has been unreliable, you can be well in overspeed or about to stall, ergo safety is compromised and you should do the memory items, which at high altitude are illogical, but at least you have to disconnect everyithing, level off and fly a reasonable pitch with thrust as required, then troubleshoot

[737Jock]

Certainly gives a new perspective to large thrust variations at altitude.

On the A319/320 VMO/MMO is 350/.82 , however the VD/MD is 381/.89 and structural inspection is only required from VMO + 20, MMO + 0.04. (FCB15)

So I think in general there is still quite a bit of margin on the topside of the speedband, where this does not exist at the bottom. So maybe leave the thrust and accept an overspeed?

Difficult scenario though!

The aircraft should still take 2.5g up until VD/MD btw.

[flyburg]

Check airman,

Dude, thumbs up to you!!!

That was precisely what happened, during the climb, the pitot tubes iced over! Giving an erroneous speed increase (Alt increases, speed increases!!). For some reason accompanied by an increase in HW due to that same fault in the FMS, leading us to believe that the airspeed increase was accurate.

We reduced power for an erroneous airspeed increase leading us into an actual stall. As we recovered the alt went down and so the airspeed indication as well but this was not unnatural as the TW increased as well (due to the faulty IRS computation). Even though the low speed stall was recovered we still had the buffet ( which seamlessly went from low speed buffet to high speed buffet) together with an extremely low airspeed indication.

Short story short, when we finally interpreted the FPV and the airplane symbol, we realized there was no way we were in a stall and recovered using pitch and power settings with an unreliable airspeed situation

In the mean time, I'm ashamed to admit we had gone Mach 1.1, went from 35.000 to 15.000 in under a minute!

The entire sim session was about unusual attitude recovery, nose high speed low, nose low speed high etc, this scenario came at the end. Flying close to TS we fully expected another turbulence upset. This scenario came completely unexpected( guess some thought went into it!)

Maybe put AF in a different light.

Maybe people on here would have known better, so be it, but for me it was a humbling experience!!!!

[Check Airman]

One would have thought that after the Turkish incident in Amsterdam and AF447, AoA indications would be mandatory in transport jets. The data is already there, it's simply a matter of putting it on the PFD.

[captplaystation]

Amen to that, :ok: a real shame no-one wants to pay for it, & nobody has the cojones to mandate it :=

[Check Airman]

You raise an interesting point. No airline wants to pay for it, and no regulatory authority wants to mandate it. Yet, I've never met a pilot who would not like to see AoA in the cockpit. When it comes to safety, why are pilots not able to demand the installation of certain equipment?

I can envision a surgeon refusing to operate unless certain safety nets are in place. Why do we allow our bean counters to say that something is too expensive? They're not the ones trained to fly, we are. Shouldn't WE be the ones demanding the installation of AoA indicators?

[DozyWannabe]

Apropos of nothing, here's a link to a post I made regarding a sim session to perform experiments in a UAS situation.

Full disclosure : I'm not a pilot.

http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-85.html#post6793521

[stilton]

Better of bouncing along at FL350 until you have burned off more fuel and have a lot more margin for a climb :ok:

[Lyman]

AF Pilots threatened a walkout, unless AF comitted to an immediate change out of at least one Pitot (Thales).

"You raise an interesting point. No airline wants to pay for it, and no regulatory authority wants to mandate it. Yet, I've never met a pilot who would not like to see AoA in the cockpit. When it comes to safety, why are pilots not able to demand the installation of certain equipment?".....Check Airman

Airline/Regulator/Pilots. Without all three, flying comes to a halt, sooner or later.

A Stool cannot stand on two legs.

[Dan Winterland]

AoA is an essential tool for accident investigators and frequently thier reports refer to it - and more than one has commented that an AoA guage would be useful for the pilots. If you have the option of displaying the FPV (Flight Path Vector) on the PFD, then you can get a good idea - the difference between your attitude and the FPV is your angle of attack. I have flown military aircraft with AoA guages and it's one of the most useful intruments in the aircraft. I once encountered an ASI failure (common on this type due to pressure changes during manouevring) - the subsequent approach using AoA and power wasn't an issue.

But one thing I noticed in the recent increase in stall training post AF447 is the lack of awareness of pilots regarding the basics. When we did our instrument training and practiced partial panel, we should all have been briefed on the basics - i.e. that flight is an equasion where Attitude = Power + Performace. Take any one element away and we can still deduce it from the other two. For example, If I take the speed indication away, I can still aprroximate the speed to fly by setting attitude against power. At most cruise levels I know that for most weights on the A320, 2.5 degrees nose up and 85% N1 will give me a safe speed. Of course, this has been reinforced in the recent training, but I was suprised at how many of my colleagues had a problem remembering it.

[Microburst2002]

That is true.

I have seen many pilots have problems understanding how to fly the pitch/thrust tables in the 320. In some cases they want impossible things, like flying a given path (level, for instance) and pitch and thrust setting. You can't decide all of the terms of the equation!

I think it is due to the widespread reluctance to use thrust manually, so that when they see the table and they see a thrust setting on it, they want just to set that thrust and just forget thrust levers.

So they whole procedure becomes meaningless. The procedure states how it should be done, but I have never seen that issue properly discussed, and even some negative training regarding this procedure.

I was lucky because I was taught to fly according to that principle: Attitude = Power + Performace. So I have no problem with that. However I have the problem of coming to the conclusion that speed indications may be bull**** and then decide if the safe conduct of the flight is impacted or not (because it is nonsense, of course it is)

[BOAC]

flyburg - to answer your post#1, I think it a little unrealistic! Albeit a good training 'wake up'. Firstly, the decision to climb to a performance marginal level in area of TS is wrong, even if it is (presumably?) above cloud. The wind shifts were also quite exaggerated (for the exercise, naturally).

The great result is that power/attitude (?trim- if only!!!) was re-inforced. A very difficult scenario to handle and survive!

[Young Paul]

Yep. Our parallel sim experience was similar, and just as humbling. The old "unusual attitude" training was based on you (basically) flying into it, so firstly avoiding flying into it, and secondly flying out (basically) conventionally - "stall recovery". AF447 highlighted the fact that circumstances were possible where the aeroplane could very quickly be put in a situation where the aeroplane couldn't fly, where it had also lost a lot of its instruments and protections. Still recoverable - but not like anything I've seen in 25 years of flying.

Hat tip to the regulators and airlines for getting appropriate training in place. And scaring the life out of many pilots!

[HazelNuts39]

I'll stop here to see what you can make of this!The symptoms you describe indicate a frozen pitot with front end and drains blocked and the total pressure trapped inside. Did all indicators have the same symptoms?

... but this doesn't fit: You reduce power even more, the airspeed drops

[A37575]

Stall recovery practice at high altitude is an important part of jet transport training for type rating. In the 737 Classic, use 37,000 ft and maintain height as speed bleeds off after closing the thrust levers. Eventually severe buffet occurs quickly followed by stick shaker. The nose is then smoothly lowered to zero body attitude at the same time climb power applied. The aircraft needs to be trimmed to hold zero body angle in the descent recovery. When the IAS reaches Vref 40 plus 100 knots (typically 230 knots IAS) it is safe to ease out of the descent and the height loss is around 3000 ft. There is no way you should attempt recovery to level flight before reaching Vref 40 plus 100 knots since G buffet will occur again and further height loss will occur.

The Vref40 plus 100 knots comes from the FCTM advice on high altitude holding without an FMC.

Low level stalling with landing flap should be conducted at the outer marker height of typically 1200 ft agl. The worst case scenario of the Turkish Airlines 737 Amsterdam accident is replicated, where closed throttles and almost full back stabiliser trim is caused by the autopilot attempting to hold the ILS glide slope.

Recovery at that low altitude is generally successful providing autopilot and autothrottle are disengaged and full thrust applied. Apply immediate forward stabiliser trim and elevator while selecting between five and seven body angle up to counteract the strong pitch up that occurs at go-around thrust. Respect the stick shaker. Leave gear and flaps at landing flap setting until reaching Vref speed and climbing. Instinctive reaction to retract the flaps to 15 as in a normal go-around procedure, must be avoided since the speed will be around Vref 40 minus 24-30 knots when the stick shaker actuates and a full stall is then unavoidable - deadly at that low altitude. There is no buffet and stick shaker is the first aural indication of a stall.

All the above is an excellent confidence building exercise and a vital handling skill in IMC.

[misd-agin]

Pan Am 707 had an event over the Atlantic with a similar scenario. Rapid switch to tailwind that exceeded the airplane's ability to accelerate. Resulted in stall/descent to regain control.

[RAT 5]

"Recovery at that low altitude is generally successful providing autopilot and autothrottle are disengaged and full thrust applied. Apply immediate forward stabiliser trim and elevator while selecting between five and seven body angle up to counteract the strong pitch up that occurs at go-around thrust."

This is too similar to what was written originally in the FCTM. No feather in my arse, but when teaching TQ I always emphasised that a stall was an aerodynamic issue and had to be solved aerodynamically. Thus, first reduce AoA and then accelerate and avoid pitching up into a 2nd stall. FCTM has since been re-written to reflect this correct technique. It may only be a split second of a twinkling of an eye, but AoA first, then N1% umph and hang on.

[Microburst2002]

Gravity waves can scare you if you are near maximum FL (high weight) and the margin between VMO and VAPROT is small. In the updraft side, it makes you reduce target mach in order to reduce thrust and avoid overspeed. In this situations I hate soft altitude in the ALT CRZ mode. It makes the airplane gain some feet and increase the speed when you want quick A/THR reaction.

Then, just after you manage to recover the speed, with quite a low thrust, the downdraft side comes and everything changes to worse, the soft mode will again be late to react, some feet are lost, speed bleeds off, but in this circumstances, high altitude and low margin, gaining knots is much more difficult than losing them, so you can see the speed trend going well below green dot and the airplane seems unable to avoid the speed from decaying. And that damned soft altitude... Oh, wait, but we already have max CLB, and still unable to recover speed... Well, if then the pilot panics and decides to descend to a lower level (which is the best course of action), for Christ sake, don't let him pull for an OP DES!!!!!

[BOAC]

I thought 'gravity waves' were things that Star Trek got?:confused:

[Microburst2002]

hahaha

I meant Mountain Waves!

Anyway I think there is such things as gravity waves, which are basically mountain waves but with no mountains involved. But not sure.

Maybe I'll raise a thread about that, by the way...

[TTex600]

For Christ's sake, don't let him pull for Open Descent

Yep, I prefer to call it the "thrust idle" knob.

[The Range]

Microburst,
It is: power+ attitude= performance

[flyburg]

@ BOAC,

The decision to climb was not unrealistic in this particular scenario, still some margin left. The winds changing was more due to the failure of the pitot static system an not the actual winds changing I've been told that the indications we saw were natural for the particular failure although in the real world it should have been a clue!!

@ at hazelnuts39

In hindsight, you are right, come to think of it, the speed didn't drop hence we kept slowly reducing the power.

Many mistakes made for sure, and when you contemplate it behind a pc screen the answers will be much different! However this took seconds to develop in the sim as it could be in the real world!!( think between level off and the stick shaker was about 5 seconds).

Great training, I learned something, reinforces the feeling that current training for civilian pilots is lacking. For the last 11 years I have done al sim sessions with average to above average grades and never got into any trouble. However, this was an ego killer and I learned a lot more than from the standard V1 cut, come around do an NPA on AP, GA and finish of with a hand flown SE ILS!!


Greetings

[chrisN]

Microburst, there are indeed waves not associated with mountains. We sometime get them in East Anglia, UK, not noted for its sticking up bits. In my experience, and from what I have read, they do not generate such strong up and down velocity vectors as mountain waves, but for all I know there may be exceptions to that. Wave has also been seen on satellite photographs over the sea, far from any mountains – though with favourable conditions, mountains can generate waves a long way downwind as well as close to the triggering source.

One mechanism for wave without mountains AIUI is wind shear. Think lower high density fluid, with higher low density fluid having relative motion to it – such as sea/air interface. It seems that can happen higher in the atmosphere too, with adjacent layers.

(Glider pilot, not met man, and not ATPL)

[FlightPathOBN]

Good form,

My first sim check, 737-8, a 30 kt crosswind, temp -20C, snowing, with an iced runway...at night, in Deadhorse, AK....

BTW...full motion ...

Buzz Nelson was running the sim...

[bubbers44]

Hopefully we can stop the flow of low time cruise type pilots getting into the airlines, then monitoring autopilots and becoming captains with no actual flying experience hand flown. It probably isn't going to be allowed to happen because of the cost by the bean counters. It wasn't like that 30 years ago, you then had to know how to fly.

[Wizofoz]

Microburst,
It is: power+ attitude= performance

Only if the air is not moving vertically. Introduce a wave or up/downdraught in a critical performance situation and that goes out the window.

Also remember the same power+attitude gives vastly different performance depending on whether you are pre or post stall. AF447 was 16deg NU with full thrust- and descending very rapidly!!

[bubbers44]

Excuse me? No competent pilot in an airliner would expect performance at FL350 heavy at a 16 degree deck angle. No ONE. We can't let these inexperienced guys be together when the experienced guy has to take his break. Get one of the old 65+ year old guys that can handle it be the second pilot as CRUISE PILOT. They would know what to do. Trust me.

[Microburst2002]

Damn, I have to check more carefully what I quote!

Anyway I know you know what I meant, right?

Chrisn

Thanks a lot for that. I always wandered what the hell could make mountain waves without mountains. As for the latter, I find them very often in the same area and season, sometimes very intense.

[chrisN]

A bit more on mountain wave. In some circumstances, a localised area can have wave that has a vertical component the same as the wind speed. Unlikely though this might seem (need a diagram here – can’t find one to post), successively higher layers can pile up on each other, with a shorter local wavelength and more bending of the streamlines from horizontal to a steeper angle, until at a certain point they go vertical. In the lee of a Scottish Mountain, Morven (Aberdeenshire) on a day with about 10 knot wind, we found local lift of almost 10 knots for a while. Mountain waves are not always the classic sine wave shape.

This might be one explanation why even airliners experience up or down drafts which seem disproportionate to the general wind speed. As far as I know, however, this extreme is only found where the mountain shapes are right for it and at the same time the airmass characteristics (wind change with height, stable layer over unstable, etc.) lend themselves to this phenomenon.

I don’t know enough met to say if it could happen elsewhere than particular spots. I certainly don’t know what happens near jet streams.

It may be that the professionals don’t yet know that much, either. A lot of early work on wave came from gliding exploration of it, before the soundings and analysis/theories followed. For obvious reasons, there is not much gliding exploration of wave over the sea.

Chris N

[Wizofoz]

Excuse me? No competent pilot in an airliner would expect performance at FL350 heavy at a 16 degree deck angle. No ONE. We can't let these inexperienced guys be together when the experienced guy has to take his break. Get one of the old 65+ year old guys that can handle it be the second pilot as CRUISE PILOT. They would know what to do. Trust me.

Oh, completely true- but that was the attitude held THROUGHOUT- my point was simply that one attitude gives two different performances- pre stall and post stall.

[BOAC]

]The decision to climb was not unrealistic in this particular scenario, - not in your book, I take it, but not something I would have done with expected turbulence - whatever happened to commonsense?:ugh:CBs = turbulence= need extra margins, not less.
Flying at 35.000 with an heavy airplane you climb to 37.000 which is just*
Possible with the weight. {I read very little manoeuvre margin?) Reason for the climb, you are flying just below the clouds in an area of embedded TS and it is bumpy.

Starting the climb(with just a small margin between upper and lower red band (I read with a small manoeuvre margin which you then throw away) but still.....ChrisN and Micro - I suggest you do some met study (even Google) - and you will find that East Anglian 'waves' are indeed mountain waves, probably emanating form the Welsh uplands.

[Callsign Kilo]

The erroneous and exaggerated wind read out is to be expected with airspeed unreliable on the Boeing that I fly (I imagine the B744 is similar). A corrupt pitot static input causes the related ADIRU to produce all sorts of nonsense. IAS, TAS, Mach, W/V, stickshaker, overspeed, GPWS windsheer warnings, even the Altimeter can all be rendered unreliable. It is a very confusing situation to be confronted with and sobbers the mind completely. It comes as little wonder when you consider how accidents occur with this sort of failure(s).

Cudos to the OPs airline for understanding and exposing this type of occurrence. All crews should be confronted with this type of training. To combine it with upset recovery is also a fantastic idea as delayed recognition or as the OP suggests, operation withinin the limits of the aircraft flight envelope, may require the correct recovery technique to be applied.

It amazes me how many pilots appear bewildered when you mention the old adage, pitch x thrust = performance. It's disconcerting when you learn that so many are unaware of the importance of having a working knowledge of he fugures found within thr Performance Inflight section for flight with unreliable airspeed or the number one thing to do in a stall is to reduce the angle of attack. Seriously.

Credit to flyburg for posting his experiences in the sim, especially when he considers how he had felt humbled by the occurrence. However I'd be pretty sure that he now feels more confident about how to recognise and confront such a scenario, a scenario which in all likelihood has greater chances of occurring than an engine failure at V1.

[henra]

The winds changing was more due to the failure of the pitot static system an not the actual winds changing I've been told that the indications we saw were natural for the particular failure although in the real world it should have been a clue!!



Sounds absolutely possible. Wind speed is calculated based on TAS and GS (and hground track). TAS itself being calculated from IAS. And herein lies the problem. When IAS is unreliable, so is TAS and thus calculated wind speed.
Also the values might be possible. When cruising at 230kts IAS while having an IAS reading of 30kts due to icing the difference would be 200kts (IAS/CAS), so the FMS would consider this difference as wind speed. However during your M1.1 adventure the wind speed reading should have been even quite a bit higher, especially, when being at a lower altitude if the IAS reading was still below 30kts.
Do you know how they did simulate the windspeed in the Sim? Is there a program to emulate all effects of pitot icing ? Or was it manually controlled by your examiner.

Interesting and enlighteneing story, btw.

regards, Henra

[flyburg]

Nope, sorry, can't tell you how they simulated it. Will see if I can't find out.

Greetings.

[BOAC]

Normally simple button pushes on most sims. Again, modern sims can have 'lesson plans' built in which will do the job when asked.

[FERetd]

bubbers44 Quote " Hopefully we can stop the flow of low time cruise type pilots getting into the airlines, then monitoring autopilots and becoming captains with no actual flying experience hand flown. It probably isn't going to be allowed to happen because of the cost by the bean counters."

You are correct, of course. But although the bean counters love it, it is the regulator that allows this practice.

A well known airline based at the Fragrant Harbour uses low time Second Officers during criuse. These Second Officers do not even hold a full type rating, yet the aircraft is certificated to be operated by two pilots. This is criminal, in my opinion.

I wonder, then, what the definition of a pilot is.

I believe that the fare paying passenger has a right to expect two fully licenced and type rated pilots to be at the controls of the aircraft in which they are travelling.

Sadly, there are weak regulators who are nothing but whipping boys, merely endorsing the operator's demands.

The recently introduced MPL serves to illustrate the problem - it is not going to get any better, unfortunately.

[misd-agin]

You can't hope or will the airplane to fly. If it's not flying it needs less AOA. Period.

You can't say "I don't like the nose that low'. If the wing needs AOA reducing AOA is the driving force, not some desire for a 'normal' pitch attitude.

Did loss of airspeed in the sim. During recovery we had 10 degrees nose low at max power(actually restricted to 75% N1 due to EEC failure resulting from loss on data input!). CKA said "don't put the nose that low." Silly boy, AOA gauge demanded it. Pitch attitude was being driven by the AOA gauge.

If you've done acro or fighter flying you've probably seen amazingly slow airspeeds. Less than 1G, and low AOA, is what allows that. I've seen the airspeed indicator pegged against the stop (<60 kts), when the actual stall speed was 120+ kts, without being stalled. But not for long. ;)

[aguadalte]

Airbus Drivers,
Next time you go for a sim check, just ask your check air man to freeze all of your pitots during a climb (lets say passing 30.000'). Everything functioning normally, with AP ON...
Since pitots were frozen at the same time (no significant speed differences between ADR's), your ECAM will stay silent until high speed protection actuates and takes your aircraft into a deep stall...
You will have to force your aircraft status to change to Alternate Law, in order to be able to push your nose down...
That experience will tell you a lot about the aircraft you're flying...

[A-FLOOR]

Nope, sorry, can't tell you how they simulated it. Will see if I can't find out.

Greetings.The headwind/tailwind scenario as a result of the erroneous TAS computations with changing altitude with blocked pitots is absolutely realistic. (I work for the company that built your 747-400 simulators) The actual winds set the simulation were probably either calm or stable.

I urge any airline pilot who has never seen this effect firsthand to give it a try the next time you're in the sim. As flyburg has said, it is a real eye-opener with regards to AF447.

[Lyman]

AFLOOR, Is there any reason to include ICED probes? Not to? What do you think of the possibility 447 may have lost her reads via airmass volatility alone?

[AlphaZuluRomeo]

Since pitots were frozen at the same time (no significant speed differences between ADR's), your ECAM will stay silent until high speed protection actuates and takes your aircraft into a deep stall...

Uh?
Doesn't high AoA protection have a greater priority than overspeed protection :confused:

BTW, I also was under the impression that getting 3 pitots freezing at exactly the same time is... let's say extremely unlikely. Wrong?

[Lyman]

AZR

Hi. I recall from the "Another 447 avoided" that overspeed warning is primary over STALL warn...

Lyman

Fwiw: if the probes are blocked for the sim ride, simultaneously then it is not a true 447 profile....as above, freezing all three at exactly the same time is virtually impossible. I believe that is why BEA will not say ICE blockage caused UAS. If airmass/turbulence, or ICE, it happened concurrently, and frankly, IMHO, for all three to ice at all via micro granulae is far fetched. Any transition from 100 knots on the nose to 100 knots on the tail would cause extreme disruption at the nose, and the pitot probes, ( to a lesser extent, the statics).

[bubbers44]

We know unqualified pilots flying together is unsafe so if the regulators don't fix it maybe the press will have to. We will just have to be patient for another couple of AF type crashes. Sad, isn't it?

[Machinbird]

Bubbers
We know unqualified pilots flying together is unsafe so if the regulators don't fix it maybe the press will have to. We will just have to be patient for another couple of AF type crashes.I'm not trying to cause great mischief with this question, but how will we tell the qualified from the unqualified pilots? Both types will be wearing the uniform, sitting in the seats with the forward view, and carrying the ATP certificate in their wallet. Just what criterion will we be using?:E

I'll bet there guys&gals out there who fully believe they are qualified but would not fit either my or your definition of qualified.

So how do we get them properly qualified?

[bubbers44]

My airline, 30 years ago, had me take a check ride in a 4 engine turboprop I had never flown and do ILS's to a single engine approach to get hired. That is how you can find the pilots you want. We hand flew all approaches so the computer monitors would not be accepted for my job. That is how you hire a real pilot.

[bubbers44]

Yes, the airlines are hiring the monitor pilots because they are cheap. Maybe we can raise the standards to what we had 30 years ago. Maybe we need to see if the pilots now flying can handfly before we let them fly together. Automation is great but it should just help you fly, not be your only way to fly.

[elrey_b_jepp]

I doubt the Attitude indicator and FPV can show you the AOA.
The FPV has a trend vector with power inputs, therefore your nose may be level but FPV down like on an approach. That doesn't give you the AOA, am I right?

[HazelNuts39]

I also was under the impression that getting 3 pitots freezing at exactly the same time is... let's say extremely unlikely. Wrong? In this case they don't have to freeze at the same time, but they have to freeze in the same manner. If a pitot is blocked at the intake and at the drains simultaneously, you wouldn't notice anything until you change altitude.

[Young Paul]

1) I'm not convinced that this is simply a question of more/better flying experience. I agree about insufficiently qualified second officers, but the belief that there is something fundamentally missing in airline/type-rating flying training to me ranks with me alongside the idea that fly-by-wire is somehow "less safe". For the one-in-a-million event which led to this, there are the one-in-a-hundred events where overbearing, arrogant captains ended up effectively flying airliners "solo" because they had no confidence in their first officers (ie. were unable to operate an airliner as a crew, as the aircraft had been designed to).

2) Even if it is, I don't know how people expect to change this. There is much less military flying now - should tax rates be put up by 2% to pay for increased defence expenditure? Should pilot qualifications be increased in cost by £10000 to provide a greater diversity of experience? Who is prepared to pay for this? And how could anyone measure whether the extra training provided suitable economic benefit? And what extra experience should be provided? How would you know that you had provided all the training that would be needed to convert an "unsurvivable" event into not only "survivable" but "recoverable"?

3) The idea that thirty years ago we were in some kind of golden era where pilots were real pilots, men were real men and blahblahblah, is simply not borne out by the facts. The absolute rate of accidents has fallen and is falling - see the graphs on this page (http://en.wikipedia.org/wiki/Aviation_accidents_and_incidents) - whilst the number of commercial flights has increased. The argument that "we need pilots with somehow 'better' experience" is not statistically supportable.

To my mind, the approach of identifying a threat, then providing training to deal with it - steered by regulators and airlines - is entirely appropriate. The process that is happening now will ensure that professional pilots worldwide will have a greater appreciation of this combination of threats - it provides the extra experience that people say is missing.

[HazelNuts39]

I recall from the "Another 447 avoided" that overspeed warning is primary over STALL warn...Perhaps you misread. Overspeed warning took priority over A/P disconnect warning. There was no stall warning because the airplane remained in normal law.
Any transition from 100 knots on the nose to 100 knots on the tail would cause extreme disruption at the nose, and the pitot probes, ( to a lesser extent, the statics). It wouldn't cause disruption of pitot or static, but it would immediately deprive the airplane of almost all lift and drag. IOW in AF447 you wouldn't be able to pull 1.6 g at stall warning.

[Meikleour]

ray.boeing (http://www.pprune.org/members/233074-ray-boeing): No. The FPV will give you AOA indirectly. It is the difference between the aircraft attitude and the FPV angle. However I am under the impression that the FPV is not entirely inertially driven so may have errors added when the pitots/statics are obstructed.

[Microburst2002]

Which brings to mind is question:

Can static ports become frozen in flight? I know pitot tubes can, but static probes?

And are the chances that pitot tubes will freeze much eqrlier than static ports? i would say 99%. I deem that icing related unreliable speed situations in cruise flight will always be pitot related. So you will be able to rely on Altitude indications and FPV. Angle of attack can be derived from the difference between body angle and FPV.

Aguadalte

That scenario of the false oversped triggering undue high speed protection, when you are actually flying very slow, is dreadfull. But I believe you can still use manual pitch trim to achieve pitch down.

By way, what will happen in a Boeing in the same circumstances?

As I always say, the most difficult partof an unrliable speed scenario is becoming aware that speed is unreliable, in the first place. If you don't come to that conclusion timely, you are fu*ked.

Then, My opinion is that the procedure needs more refining. It is too "take off oriented". You can be at high level with "safe conduct of the fligh" absolutely "impacted". Then what do you do?

[AlphaZuluRomeo]

AZR
Hi. I recall from the "Another 447 avoided" that overspeed warning is primary over STALL warn...
Hi Lyman. :)
I'm sorry, you recall wrong here.
High AoA (Stall) protection has priority over other protections (IIRC, I stand to be corrected and hope I will not on that topic).
Stall warning has priority over other warnings.
Why? Because stall was deemed more dangerous.

In this case they don't have to freeze at the same time, but they have to freeze in the same manner. If a pitot is blocked at the intake and at the drains simultaneously, you wouldn't notice anything until you change altitude.
Hi HN39. :)
AFAIK, they have to freeze at the same time and in the same manner if to fool failure detection mechanisms (comparing in real time values from 3 probes).
My point was: such an event in real conditions is extremely unlikely, that's why detection are based on that assumption. Don't you agree?

[HazelNuts39]

AZR,

If the drain of a single pitot is blocked, and the intake of same pitot almost at same instant, then the total pressure inside the pitot is 'frozen', no change to detect. Same for a second and third pitot some time later. The IAS changes when the static pressure changes, i.e. the altitude. IAS increases with increasing altitude, and vice versa. The IAS from all frozen pitots change in the same manner, no differences to detect.

What IMO makes this scenario somewhat unlikely is the absence of turbulence in the simulation. Turbulence was present in all UAS incidents studied by BEA, and would cause altitude variations causing airspeed anomalies that would trigger detection unless several pitots froze simultaneously.

[Meikleour]

Microburst2002: Yes, static ports can get iced over in flight. In the `70s there was an incident invovling a Trident operated by BEA/BA which experienced freezing rain runback on the fuselage which blocked the static ports. It was in the BNN hold at the time. When the stick push operated it promptly dived through several levels of the hold missing other aircraft. My neighbour at the time was the F/O.

Regarding manual use of the THS to control pitch - yes it is always available and in some situations may be needed to overcome the ineffectiveness of the small elevators. This in my opinion is one of the great untaught "gotchas" of the FBW Airbus`s. The THS is NEVER NORMALLY moved by the crew in flight but as AF447 found out it may have helped. The same goes for the A320 loss at Perpignon.

[A-FLOOR]

AZR,

If the drain of a single pitot is blocked, and the intake of same pitot almost at same instant, then the total pressure inside the pitot is 'frozen', no change to detect. Same for a second and third pitot some time later. The IAS changes when the static pressure changes, i.e. the altitude. IAS increases with increasing altitude, and vice versa. The IAS from all frozen pitots change in the same manner, no differences to detect.

What IMO makes this scenario somewhat unlikely is the absence of turbulence in the simulation. Turbulence was present in all UAS incidents studied by BEA, and would cause altitude variations causing airspeed anomalies that would trigger detection unless several pitots froze simultaneously.The total pressure will freeze, but remember that the static ports were available and working 100% all the time in the AF447 scenario. IAS is still equal to Pdynamic, which is derived by substracting Pstatic from Ptotal.

The pressure variations in the turbulent air will still find their way to the IAS via the static ports, and this is of course also the reason why climbing will give you the impression aircraft is accelerating. What was also significant in the AF447 case is that the turbulence will cover up any other cues that the aircraft has stalled (stall buffet), and the FBW will mask any change in roll response of the flight controls due to the extreme AOA, as I recall that the roll rate response vs. stick deflection is linear in alternate law. FBW being FBW, it will always try to give you what you are asking of it. So the aircraft will appear to respond normally around the longitudinal axis, and the only thing telling you otherwise will be the compass heading turning in the opposite direction. But trying to figure that one out when you are descending at 10000fpm through a CB at night with a nose-up attitude with your aircraft telling you to pull up even more is an exercise of futility.

With regards to Airbus vs. Boeing: the Airbus overspeed mode was commanding an increase in pitch through the flight directors, which was followed religiously at least until the aircraft reached its apogee at 38000ft. This continuous backpressure on the stick caused the THS to trim all the way up, leading to the deep stall. I doubt they even realised that the THS was at the position it was in at any point as you are not supposed to fiddle with it yourself but let the automatics handle this, even though the captain correctly remarked on his return to the FD that the aircraft was stalled. It has been argued that if they just let go of the controls it would have allowed the aircraft to recover itself, and this is probably correct. This tells you how a Boeing would have fared in the same situation: the lack of autotrim on Boeings leaves the THS in the same position unless you conciously tell it to do something else using the trim switches, and this leaves open your avenue of sensing that the aircraft is stalled as you are unable to maintain a nose-up attitude as the aircraft loses airspeed. It's become pretty clear that in the same situation, a Boeing would not have deep-stalled. Not unless the crew trimmed up by themselves while not looking at the stab trim indicators.

[aguadalte]

Micro:
That scenario of the false oversped triggering undue high speed protection, when you are actually flying very slow, is dreadfull. But I believe you can still use manual pitch trim to achieve pitch down.

As I said, you will have to "change the Law" in order to pitch down, otherwise, in Normal Law, the Overspeed Protection will take control.

To those who say that, this is a very unlikely situation, I would like to ask if you think this would be absolutely impossible to happen?

[HazelNuts39]

The pressure variations in the turbulent air will still find their way to the IAS via the static ports, and this is of course also the reason why climbing will give you the impression aircraft is accelerating. What was also significant in the AF447 case is that the turbulence will cover up any other cues that the aircraft has stalled (stall buffet), and the FBW will mask any change in roll response of the flight controls due to the extreme AOA, as I recall that the roll rate response vs. stick deflection is linear in alternate law.
Altitude variations affect IAS via the static ports. "Turbulent air" as such does not.
Mach buffet is a violent high frequency shake of the aircraft that is very different from turbulence-induced g-variations. The airplane had left the turbulence before it stalled.
Once stalled, the airplane is effectively uncontrollable in roll, its reponse to a roll input is often opposite to that commanded, at any rate quite different from normal.

[AlphaZuluRomeo]

What was also significant in the AF447 case is that (...) the FBW will mask any change in roll response of the flight controls due to the extreme AOA, as I recall that the roll rate response vs. stick deflection is linear in alternate law. FBW being FBW, it will always try to give you what you are asking of it.
No. In Alternate 2 law, roll is direct. Meaning sidestick deflection gives proportionnal control surfaces deflection. This is different than the normal mode, and FBW doesn't try anything here, no "masking".

With regards to Airbus vs. Boeing: the Airbus overspeed mode was commanding an increase in pitch through the flight directors, which was followed religiously at least until the aircraft reached its apogee at 38000ft.
What flight are you talking about :confused:
If still AF447 : sorry, this one never get to overspeed.

(...) even though the captain correctly remarked on his return to the FD that the aircraft was stalled.
Same question. If that's AF447, I fail to see anything suggesting any of the crew members ever understood the stall situation.

This tells you how a Boeing would have fared in the same situation: the lack of autotrim on Boeings leaves the THS in the same position unless you conciously tell it to do something else using the trim switches, and this leaves open your avenue of sensing that the aircraft is stalled as you are unable to maintain a nose-up attitude as the aircraft loses airspeed.
Two remarks:
- It still remain to be proved that elevators only aren't enough to maintain stall (at a lower AoA than that obtained with the help of the THS, maybe). IIRC knowledgeable peope (of which I'm not) suggested otherwise.
- What do you mean by "the lack of autotrim on Boeings"?

[edit] May I suggest we move to the AF447 thread if I undestood correctly that was the topic at hand? We're hijacking flyburg's thread, here.

[A-FLOOR]

Altitude variations affect IAS via the static ports. "Turbulent air" as such does not.This may not be apparent as the output of the altitude calculation from the statis prerssure to the altimeter is filtered, but I assure you that it does.
Mach buffet is a violent high frequency shake of the aircraft that is very different from turbulence-induced g-variations. The airplane had left the turbulence before it stalled.I am talking about the stall buffet. The aircraft was at that point flying too slowly to induce a mach buffet
Once stalled, the airplane is effectively uncontrollable in roll, its reponse to a roll input is often opposite to that commanded, at any rate quite different from normal.There is most definitely a degree of residual positive roll control in an aircraft of conventional design when deep stalled. How much exactly for a particular aircraft in a particular circumstance is a question nobody can answer for sure as this has to my knowledge never been tried in a large aircraft (and lived through), but the windtunnel data, flight test data and computer models which are used in the aerodynamic models in simulators leave little room for doubt. I will concede however, that this is way beyond what any sim is certified for for flight training.

No. In Alternate 2 law, roll is direct. Meaning sidestick deflection gives proportionnal control surfaces deflection. This is different than the normal mode, and FBW doesn't try anything here, no "masking".Fair enough, it seems I had my modes mixed up.
What flight are you talking about :confused:
If still AF447 : sorry, this one never get to overspeed.


Same question. If that's AF447, I fail to see anything suggesting any of the crew members ever understood the stall situation.Have you read the CVR transcipt?
Two remarks:
- It still remain to be proved that elevators only aren't enough to maintain stall (at a lower AoA than that obtained with the help of the THS, maybe). IIRC knowledgeable peope (of which I'm not) suggested otherwise.My point was that with elevators alone the aircraft will drop the nose again when the elevator backpressure is released, but this may no longer be the case when the THS is trimmed fully aft.
- What do you mean by "the lack of autotrim on Boeings"?I know of no Boeing type which has an Airbus type autotrim system that automatically trims the THS to relieve pitch input. In manual flight, even the 777 and 787 have to be trimmed manually for airspeed using the trim switches. This does not take into account some functions where the trim is adjusted automatically with flap and speedbrake deployment, but the fact remains that the pilot is almost completely out of the loop with regards to the THS in the Airbus.

[HazelNuts39]

I am talking about the stall buffet. The aircraft was at that point flying too slowly to induce a mach buffetI am talking about the Mach buffet that AF447 was entering at 02:10:55 at M.67 that is shown on page 43 of Interim Report #3 and described in the associated text.

[AlphaZuluRomeo]

[edited] A-FLOOR, I moved my answer to the AF447 thread, here (http://www.pprune.org/tech-log/482356-af-447-thread-no-8-a-61.html#post7239297), in order to comply with my own remark re: "hijacking" of the present thread ;)

[henra]

With regards to Airbus vs. Boeing: the Airbus overspeed mode was commanding an increase in pitch through the flight directors, which was followed religiously at least until the aircraft reached its apogee at 38000ft. This continuous backpressure on the stick caused the THS to trim all the way up, leading to the deep stall.


Hmm, I'm afraid I'm not sure to which case you are referring?
In case of the A340 Zoom Climb it was overspeed protection that send the aircraft into a climb but it did not stall. AoA protection law kicked in and kept it at Alpha_Prot. No stall here, and no alternate law either.

If you are referring to AF447, there was no overspeed event, it was in alternate law 2 due to the ADIRU's signalling Unareliable airspeed (Icing of pitots) and it was sustained NU commands by the manually flying PF that caused the THS to finally support his efforts and go to a significant NU Trim setting. This combined with further mainly NU commands on the sidestick kept it in the stall.

Alpha_Max is a HARD protection which cannot be overriden as long as it's active (not the case in AF447 due to Alt2). High speed protection might win against Alpha_Prot but not against Alpha_Max
It is not THAT easy to fool the Airbus protections. That takes much more finesse and or bad luck (stuck AoA vanes in Perpignan).

@flyburg: Apologies for also taking part in the AF447 thread hijacking ;-)

[HazelNuts39]

In case of the A340 Zoom Climb it was overspeed protectionTwo A340 'level bust' incidents have been discussed recently on this forum: TC-JDN on 2/10/2000 and F-GLZU on 22/07/2011. In both cases there was an overspeed warning but the overspeed protection was not activated.

[henra]

Two A340 'level bust' incidents have been discussed recently on this forum: TC-JDN on 2/10/2000 and F-GLZU on 22/07/2011. In both cases there was an overspeed warning but the overspeed protection was not activated.


I was referring to the second one where at least the FDR trace shows 2 overspeed events right at the beginning of the incident.
See http://www.bea.aero/docspa/2011/f-zu110722/pdf/f-zu110722.pdf
page 18.
But indeed it seems the pitch up was due to Sidestick input although the AoA starts to increase prior to deflection of the S/S which I attributed to overspeed protection. Looking closer at the traces the pitch only starts to increase after the NU S/S commands.
In the text overspeed protection is also not explicitely mentioned, therefore I come to agree to your statement and stand corrected.

[AlphaZuluRomeo]

Ahem, sorry but...
"Au cours de l’événement, la protection haute vitesse ne s’est pas activée."
=> "Throughout the event, the high speed protection did not activate."
Footnote 4, page 3.

[john_tullamarine]

Yes, static ports can get iced over in flight.

That may be true.

However, the static is very sensitive to airflow variations. Put a minor obstruction on the skin somewhat upstream and the PEC can alter substantially from what the AFM suggests.

Keep in mind that a lot of effort in the certification test program goes into establishing the PEC information for the AFM.

[averow]

Good thoughts re use of THS as recourse of last resort. One perhaps could file this away under "what to do when all of the usual drills and memory items aren't working" file. I suspect however that the crew of AF447 never really knew that they were in a stall, nor what their real AoA was in the last minutes of flight......:sad:

[Microburst2002]

I know of no Boeing type which has an Airbus type autotrim system that automatically trims the THS to relieve pitch input. In manual flight, even the 777 and 787 have to be trimmed manually for airspeed using the trim switches.

B777 has no autoreim in that sense, but it is not a conventional THS either. It moves without pilot inputs, same as elevators, depending on actual speed and trim reference speed, which is what you select with the switches.

So, if speed is wrong, if speed is overspeed (for the computers, I mean) what would happen in a B777. Its protections are different tha those of airbus. No uncommanded pitch up or down, but there are forces induced in the control column to cue the pilot. In this case it would mis,ead the pilot, right? It would be too stiff for pushing, I deem...

Computers are computers. They have a big problem. What happens when they are doing the wrong thing without detecting it is tthe wrong thing.

[Spooky 2]

Just a small point of order here. Boeing offers the AOA in all of it's current airplanes. Delta and American are the only airlines that I know for sure have the AOA in the 737NG and 777 series aircraft.:ok:

[Young Paul]

"Computers are computers" - garbage in/garbage out.

However, the fact remains that accident rates continue to fall, not only per million flights, but in absolute terms. Yes, we need to continue to develop the software models, but the argument that flying is more dangerous because of increased computerisation is simply false. I am not convinced it's even possible to argue that aircraft with computerised protections are more dangerous than those without.

[Microburst2002]

Absolutely.

I'm talking about the different nature of computers as opposed to mechanical systems.

When computers are involved, wrong data that can't be dismissed by the computers as wrong, can make the computer make very nasty things. The more complex the system is, the more involved one computer in the system and the more systems interface that computer, the worse effects such a malfunction will have.

Wether it is a Boeing or an Airbus, computers are computers.

Computers are protected against that in various manners, like "watchdogs", "voting", control and monitoring dual channels and "dissimilar redundancy". When a computer is not working properly, it is very important that the computer is declared invalid, either by itself or by other computers, so that its outputs are neglected.

The simultaneous and identical freezing of all three pitot tubes will make all three computers be in error, which makes impossible to detect by voting. So it is necessary some added function to avoid that situation, such as probe ice detectors or a "reasonableness check of IAS" using data other than pitot, I think there is already something like that invented.

I'm still curious about what would happen in a B777

[henra]

Ahem, sorry but...
"Au cours de l’événement, la protection haute vitesse ne s’est pas activée."
=> "Throughout the event, the high speed protection did not activate."
Footnote 4, page 3.

Oooopsie, now that you point me to it....

Guess I need to give my glasses a good cleaning :}


On a more serious note:
Is there any documented case of an overspeed protection induced Pitch Up in an A330/340? Apart from this one where I wrongly assumed it was the overspped protection that initiated the climb I'm not aware of any other suspected case on these types. Did I oversee something?
Has it really never been used in anger?

[Linktrained]

MEIKLEOUR #59
AVEROW #72

If " The THS is NEVER NORMALLY moved by the Crew in flight... " it might tend to be forgotten.

If a car with hydraulic brakes has a hydraulic leak, the hand brake is a separate system which can be used... And it should be used, just occasionally, to remind the driver that it is still there.

( A family car had a hydraulic leak recently. Luckily the young driver DID remember. All was well...)