Tech Log The very best in practical technical discussion on the web


Old 28th Jun 2009, 21:42
  #2461 (permalink)  
Join Date: Aug 2005
Location: fairly close to the colonial capitol
Age: 52
Posts: 1,686

The rudder travel limiter function was covered earlier, but it was last mentioned scores of posts back over a week ago. Let me share my recollection on the travel limiter.

The A330 will keep limits set at last known good airspeed should there be a loss of air data. If this condition were to be sustained, the system becomes 'unlocked' only upon slat/flap (?) extension IIRC.

Rudder/VS part deux...
Even if the rudder was fully 'unlocked', I see no reason for the PF to touch the rudder pedals at cruise. They don't even make for a good foot rest.
vapilot2004 is offline  
Old 28th Jun 2009, 22:27
  #2462 (permalink)  
Join Date: Mar 2009
Location: S23W046
Age: 70
Posts: 57
Pitot Antiicing System

Reading the CRASH-AERIEN forum of french pilots I found the following post adressing the Antiicing System of the pitots. Reading 400 pages the author shows profound system knowledge and post also relevant schematics. I found it too interesting to not post it here. In the following he talks about the pitot heating logic:

au sol a une temperature relativement faible, je parle de rechauffe en temperature mais j aurais pus parler d augmentation de tension dans la resistance mais bon on fait simple, de l ordre de 25degrés. en vol ce sont les sonde de detection de givrage( ice detect) qui vont en fct de la detection envoyer un signal de rechauffage ou non au calculateur phc (probe heat computer). deux type severe ou non qui modifie ainsi le rechauffage (en gros 200 degres). ces ice detect sont surtout la pour couper le rechaufage quand il n y en a pas besoin. ensuite on a un mode de surpassement en mettant le push button "probe window heat" sur on, la c est pleine power tout le temps (encore faut il le mettre sur on). mais le hic du systeme c est que les sondes ne sont pas situé au meme endroit, deux a gauche captain et stand by (9d1 et 9d3) et une a droite celle du first officer (9d2) avec la prise du vent relatif differente, tout un merdier!!! alors cela ne givre pas au meme moment forcement, c est pourquoi la logique des adrs et qu a partir d une certaine derive d un des circuit les autres le shoot car on ne peut avoir une vitesse differente sur le pfd captain et celui du copi.
A relevant foto:


At the ground it is kept to a temperature relatively low, about 25°C, I am talking of temperature in respect of heating, could have talked about increase of voltage in the resistor, but want to keep it simple. Inflight the Ice Detectors will send according to what they sense a signal for heating ON or OFF to the Probe Heater Computer (PHC). There are two modes 'severe' or 'not severe' that modify the heating (roundabout 200°C).
These Ice Detectors are primarily there to switch off heating when it is not required.
There is an override mode, if one pushes the 'Probe Window Heat' button to ON, which will provide constantly full heating (but you have to switch it to ON).
But the problem of this system cosists of the probes not beeing installed at the same location, two at the left side, captain and standby (9d1 and 9d3), and the First Officer's (9d2) one at the right side, with different sensing of the relative airstream, all that beeing a s....t!!!
Now this will not necessarily ice up at the very same moment, that's why the logic of the ADRs, as soon as a certain drift of the values of one circuit is sensed, are excluding it, because one cannot have different speeds on captains and copilots PFD.

For me that could be an answer to why the problem with pitot-icing seems to be more prominent on A32/3/40 than other types.

I remember very well that AB pilots at Lufthansa learned a big deal about Ground Sensing Logic after the Warsaw accident, they did not have all details in their Docs before ('...they don't need to know?').
I write this because sometimes there is more behind a system, than would be visibly to us pilots. Having said this, I do not want to imply any bad intention, only thinking that sometimes the 'designer/engineer-pilot interface' has to be looked at more critically.
Flyinheavy is offline  
Old 28th Jun 2009, 22:27
  #2463 (permalink)  
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by protectthehornet
I recall when the 737 flipped on its back due to rudder quickly we got hardover training.
Don't forget that Boeing fought it tooth and nail at the time, and spent a considerable amount of money coming up with a report that tried to shift responsibility to the pilots. It took two accidents with the loss of all on board before Boeing acknowledged there might be a problem.

Originally Posted by Will Fraser
Or, could the PF, acting on cue, push the nose down, even add power, though the a/c is anything but stalled?"
The opposite happened in two B757 incidents where a pitot blockage caused an overspeed warning, which caused the PF to retard the throttles and stall it, maintaining that action right through the stick shaker. It seems to be a human trait that we lock on to the first problem reported and try to solve that problem even in the face of conflicting evidence. The issue of human/machine interface is one that affects all manufacturers not just of aircraft, but of any complex machinery - especially one that enables something requiring as much safety and redundancy as air travel.
DozyWannabe is offline  
Old 28th Jun 2009, 23:03
  #2464 (permalink)  
Join Date: Apr 2009
Location: Petaluma
Posts: 330

Points taken. I will point out however, that AB finds it necessary to restrict Rudder deflection, 6 degrees at >350knots. Someone thinks it's important, though it is indeed a poor foot rest. Boeing cues 'Rudder with Caution' when unprotected. With a fully powered Rudder and no mechanical defeat beyond 3 degrees at high mach, well you see the problem. On a dynamic FD, it may at some point be necessary to Rudder, and inputs could possibly be non scalable due to movement of the cockpit.
Without feedback or instruments, one is left with?? No cues to gauge control?
Will Fraser is offline  
Old 28th Jun 2009, 23:11
  #2465 (permalink)  
Join Date: Jan 2008
Location: Herts, UK
Posts: 743
Originally Posted by DozyWannabe
It seems to be a human trait that we lock on to the first problem reported and try to solve that problem even in the face of conflicting evidence.
Classic and well documented behaviour, pointed at by the Investigation into the 100 mile glide dead-stick A330 fuel-leak incident, amongst others...

Originally Posted by WillFraser
On a dynamic FD, it may at some point be necessary to Rudder, and inputs could possibly be non scalable due to movement of the cockpit. Without feedback or instruments, one is left with?? No cues to gauge control?
... and an old turn & slip BF instrument would be helpful at that point?

Only asking...
HarryMann is offline  
Old 28th Jun 2009, 23:14
  #2466 (permalink)  
Join Date: Jul 2002
Location: UK
Posts: 3,182
Will, the point is you're really not supposed to use the rudder in cruise in the first place.
DozyWannabe is offline  
Old 29th Jun 2009, 00:31
  #2467 (permalink)  
Join Date: Jun 2006
Location: Terra Firma
Posts: 170
We must protect the aircraft and it's passengers from pilot-error through clever automation.

So goes the aircraft current design philosophy. But of course this will never stop HUMAN-error. All it does is shift the primary cause of errors away from pilots towards engineers. Is it possible for engineers to write software that covers all possible airborne contingencies? Who is better placed to recover the aircraft safely when (not if) it all goes pear-shaped?

There has to be a balance between automation and allowing the pilot to have full manual control of the aircraft when he needs to. I don't think when have that balance at the moment. And of course the pilot needs to have the skills, training and tools to manually fly the aircraft on 'dark and stormy' nights.
Bleve is offline  
Old 29th Jun 2009, 01:36
  #2468 (permalink)  
Join Date: Nov 2000
Location: Sol, sector ZZ9 plural Z alpha
Posts: 80

Can I just bring the point home that there is nothing that can be deduced about the on-board position, integrity or activities of the Captain, or for that matter the FA at this stage. After some 3000 posts its saddening to see the degree to which imaginations start to override fact. At the moment all we know is that these poor souls were found dead in the sea a considerable time after and considerable distance away from the incident. We dont know anything else.

I dont mean to have a go, some great analysis going on in this thread, but let's not start writing anyone's obituaries just yet.

Clear_Prop is offline  
Old 29th Jun 2009, 01:41
  #2469 (permalink)  
Join Date: Jan 2008
Location: Herts, UK
Posts: 743 we are inevitably going down that road anyway, nobody seems to have asked these questions.

Where would the c.g. be at that point of the cruise, if left to it's own devices?
What is the effect on pitch controllability when reverting to Alt2 or Direct mode of the above c.g. - from the pilot's perspective?

Ditto yaw stability?
HarryMann is offline  
Old 29th Jun 2009, 05:49
  #2470 (permalink)  
Join Date: Jun 2008
Location: ANC
Posts: 10
aft CG

"The tail 'trim' tank would probably be full, to achieve better fuel economy.
The handling would of course be affected in Pitch and Yaw to some extent"

CG would, very likely, be at 38% of MAC in cruise, however up to now I do not know if during the fight certification tests they relay stall this very airplane either in normal forward CG or never at all and use computer accepted stall limits and from there on wards. As per the FCTM A330-340 it is implied that there is NOT a Stall Speed, but this is my perception of the words on the book.

An event on a MD-11 with aft CG that was caught by severe turb in cruise lost some 15,000 feet during a FLAT stall as the airplane recovered the pitch control as the air got thicker during the descent and the elevator responded after thousands of feet lost. It is not the same acft nor the same company as we all know but it comes to my mind if the aft critical CG for saving fuel can also kill.

I know flight certification is costly as they fly some zillion hours before the passengers embark but much of what is on the manual is computer derived.
4krew is offline  
Old 29th Jun 2009, 06:32
  #2471 (permalink)  
Join Date: Aug 2005
Location: fairly close to the colonial capitol
Age: 52
Posts: 1,686

One correction, the A330 limits rudder travel to 4 degrees above 350 CAS. If I am incorrect, someone please say so. I am not rated on type.

To your point of protections being offered, I agree there is an understanding that the rudder could be needed at cruising speeds and there is reason for limitations, but the cases for this need are rare cases indeed.

Yaw dampers take on most of the task of yaw control at cruise. When these units go offline, it is best not to try to even out a yaw oscillation since one will almost always be behind the curve.

The only other time I would think the use of that control surface is called for at speed would be during an unusual attitude recovery maneuver. In this case, travel limiters or force feedback systems are there more to prevent dangerous pilot-induced yaw than to protect the vertical stabilizer.
vapilot2004 is offline  
Old 29th Jun 2009, 06:44
  #2472 (permalink)  
Join Date: Jun 2009
Location: among +1 to -8
Posts: 7
Can I just bring the point home that there is nothing that can be deduced about the on-board position, integrity or activities of the Captain, or for that matter the FA at this stage.
Snotty finger waving emoticon aside, and not to put a too ghoulish point on this, but is there one documented case anywhere that someone can cite in which a belted in left-seater or right-seater was found identifiable, not in the seat, after a large airliner crash?
quaeler is offline  
Old 29th Jun 2009, 06:46
  #2473 (permalink)  
Join Date: Apr 2009
Location: Petaluma
Posts: 330

PJ2 quoted P,R,Y limits a/p; ELAC quoted 3 Degrees deflection per side; I had thought 4 as well, I believe 6 total is correct.

I agree with all you say; this is not a standard event. At some point from stable cruise to bits in the Ocean there is wide latitude for conjecture. Something caused the VS to separate rather cleanly from the Fuselage. It may be more apparent the sequence/event after the briefing July 2. I won't exclude that the deflection seen on the Rudder at discovery and recovery is not the deflection that caused the separation, iow, full deflection to starboard. It took massive hydraulic power to put it there, and at sever, with no hydraulics or command to return it to neutral (or other), it may be there just as it was. The damage to the bottom of the Rudder may have been air induced, that portion, in the airflow, would be taking the brunt of the air.

Aside, We haven't addressed the engines yet, any water event that could damage all three pitots/statics could cause loss of power in one or both powerplants. If only one, there is a use for Rudder correction as well.
Will Fraser is offline  
Old 29th Jun 2009, 06:52
  #2474 (permalink)  
Join Date: Mar 2003
Location: BC
Age: 72
Posts: 2,452
Where would the c.g. be at that point of the cruise, if left to it's own devices?
What is the effect on pitch controllability when reverting to Alt2 or Direct mode of the above c.g. - from the pilot's perspective?
Representative cg fore-aft limits would be 17% to 39% MAC. At 210k kg, the cg would nominally be 29% to 33%...slightly aft; the tail would have about 4500kg of fuel at that point.

Pitch control in alternate law 2 (which is what they would have been in) is load factor demand law similar to normal law. Yaw remains in alternate law. They would have been in Roll direct law, which can achieve a roll rate of about 25deg/sec. In smooth air, the aircraft would be very controllable, in other words.

In very turbulent air and certainly in any inadvertent entry into a cell, my sense is, one would have to mindful of large control inputs, above all trying to maintain a level pitch attitude, letting the altitude and even the speed wander rather than chasing it with power (and certainly not pitch) but respecting the stall speed. In extreme conditions, an overspeed is not as serious as a stall - even a 50kt+ overspeed when the choices are really limited.

The weight of the fuel in the wings is actually a slight bonus, helping to reduce the bending moment. The outboard tanks are burnt at the end of the flight.

Settling the rudder limiter question:

Rudder and pedal deflection is limited as a function of the speed.

Each channel of the limiter is controlled and monitored by its associated SEC.

In case of double SEC failure the max rudder deflection remains at the value reached before failure then max deflection is available when the slats are extended.

Rudder at discovery and recovery is not the deflection that caused the separation, iow, full deflection to starboard. It took massive hydraulic power to put it there, and at sever, with no hydraulics or command to return it to neutral (or other), it may be there just as it was.
Help me out - trying to understand the two statements - "Rudder at discovery and recovery is not the deflection that caused the separation" and then, "it took massive hydraulic power to put it there" - put it where? the position we see it in now or at the moment of failure?
Right now, the rudder would be free to move and it's position is a function of the recovery operation. I suspect the damage on the bottom of the elevator is mechanical, not high-speed air-induced.

Last edited by PJ2; 29th Jun 2009 at 07:28.
PJ2 is offline  
Old 29th Jun 2009, 06:57
  #2475 (permalink)  
Join Date: Aug 2005
Location: fairly close to the colonial capitol
Age: 52
Posts: 1,686

I'm in as much doubt as you on the limitation figures. I will go with the experts on the A330. **

**PJ2 has kindly provided us with an edited response after a mere 31 minutes of this reply - thus my own edit. Graphics have been added and more info. Thank you P.

Engine out - good point there. The yaw dampers would attempt to take care of the initial asymmetrical thrust moments and after a manual application, trim would take over from there. At cruising speeds, the deflection angles would indeed be small and I was caught out on that one. Good find and correction of my points offered.

A bit late here and I am in no condition to be at the helm of a car let alone a yoke or sidestick on an airplane.

Last edited by vapilot2004; 29th Jun 2009 at 09:14. Reason: PJ2 edited post above - part of questions addressed
vapilot2004 is offline  
Old 29th Jun 2009, 07:40
  #2476 (permalink)  
Join Date: Jul 2000
Location: Thailand
Posts: 942
Effective use of simulators

Notwithstanding the outcome of any definitive conclusion, if there ever is one in this instance, the questions raised about the degree of automation and slavish adherence to same have been well reasoned and enlightening. As the Boeing aircraft become seemingly weekly more like the equivalent Airbus product (Cat II, Single engined 30 flap Auto-landings now possible on B737 NG) it seems to me that the training philosophy will go the same way for Boeing as it currently is for AI.

Mention of the apparent waste of time spent on LOFT exercises and the fixation on that apparent, 'worst case scenario' of an engine failure just after V1 has me nodding my head in agreement. There are several instances which can lead to the loss of an aircraft if a catastrophic event occurs, as shown by the posts here in this thread.

Some very simple faults have led to devestating crashes, as mentioned here, the Tristar in the Everglades and so on. If we can destroy an aircraft when a light bulb blows then how much more likely is it that we will lose one when the Flight Instruments suffer a total failure or we are overwhelmed by contradictory messages and signals from the ever more complex computers on board?

Time spent starting engines, manually closing start valves, taxying, climbing to FL 150 with a Hydraulic failure in the climb followed by a call from Ops to say we have a bomb on board so please return to base asap does nothing to enhance the pilot's ability to fly the aircraft. All of these events can be covered on a line check, and should be from today onwards. Training Captains of the world, unite and demand that sims be used for something useful, not the present folly and waste bent on ticking boxes determined by some retired pilot working for the various Ministires and Aviation Authorities.

Simulator time should be exclusively set aside for the very scenarios mentioned in this thread. These are the events which are killing people, not B or Yellow Hydraulic system failures or passengers with make believe heart attacks.

High altitude upset, stall recovery, unusual attitude recovery, total loss of Flight Instruments and /or total loss of Nav guidance/IRSs, flight with total loss of electrics and so on should be our first priority, however unlikely they might be according to the statisticians. I am sure this list is far from exhaustive, please add to it as you think fit.

The statisticians don't fly the aircraft, we do.

Btw, how many times has a modern engine failed a few knots after V1?
rubik101 is offline  
Old 29th Jun 2009, 07:52
  #2477 (permalink)  
Join Date: Oct 2003
Location: Canberra Australia
Posts: 1,300
Pitot Icing

Seems we have a recurring problem with some pitots at high altitude when penetrating the tops of CBs. That problem could well be that when we are penetrating CiroCu/CiroStratus rich in ice crystals at way sub zero OATs we may be converting some of those crystals to water with a heated probe. I would expect that some ice crystals will enter the heated pitot simply because of their inertia.The melted crystals may then be entering the tubing connecting the pitot and refreezing to form a blockage particularly if the drain holes of the pitots are too small or already blocked or partially blocked with crystals.

As a TP I have not had any experience in the flight testing of pitot systems except to determine position errors PEs for pitot static systems. These tests are usually done at as close to sea level as possible and more often are done using a trailing static and a test pitot on a projecting boom.

Does anyone know whether the offending pitots and attendant tubings have been adequately tested under the worst conditions to be expected at high altitude.
Milt is offline  
Old 29th Jun 2009, 08:39
  #2478 (permalink)  
Join Date: Aug 2005
Location: London, UK
Posts: 194
a quick q on stalls in jet aircraft;
other than the stick-shaker, do you get any similar clues to being close to stall as you do in a light aircraft; eg loss of authority, vibration, quietness; I am assuming the control setup and cockpit/engine noise mask all these clues, but just curious.
oversteer is offline  
Old 29th Jun 2009, 09:03
  #2479 (permalink)  
Join Date: May 2002
Location: In a nice house
Posts: 981
This is a long thread!

If a pitot breaks off completely it can generate a stall warning. In poor visibility/ thunderstorm/ night it could cause the instant stall recovery actions. Couple this with the incorrect speed readout and this could cause things to go wrong quite quickly.

Simulator training (recurrent). I completely agree that most companies have this honed to the core these days, with just the standard OPC/LPC items covered plus other "scenarios" which often times are unrealistic, to get the "ticks in the boxes". However, this is changing. In future it will be possible to have much more crew specific training - so if for example, you feel you have a weak area, or it is shown in your training file, training will be targetted towards improving that area of knowledge/ skill. This is a huge leap forward and hopefully most airlines will be using this system soon.

I don't agree with all the comments going on about how Airbus is this giant computer that can over-ride a pilot. Yes its systems are complex but it has probably saved plenty of lives too. I think most Airbus pilots are very respectful of the complexity and get it. Boeings can and have crashed due to errors of the crew, for example, thinking the autothrust is in when its not, and stalling the aircraft. Now and Airbus wouldn't let you do that.... So to say all Airbuses are rubbish because they have protections and all Boeings are great because they don't have the same level of protections is rather too simplistic.
Airbus Girl is offline  
Old 29th Jun 2009, 09:49
  #2480 (permalink)  
Join Date: Jun 2001
Location: East of the Sun & West of the Moon
Posts: 286
Quote: Will Fraser

Histrionics aside, may I ask for clarification on one of your recent posts re: 447, flight computers, and hand flying? With possible corruption of sensors, 330 flight computers seem to produce a reasonably consistent chain of output that may 'confuse', or inhibit manual flying which results as a trip out of the a/p and consequent degradation to Alternate Law or Alternate Law 2. Reference Qantas and NWA, the computer displays 'Stall Stall'. Whether or not that cue is to be taken seriously seems to depend on decisions made by the PF. Do you consider that a hindrance in recovery? Or, could the PF, acting on cue, push the nose down, even add power, though the a/c is anything but stalled?

When a failure of an element of the flight control system occurs or data from a source becomes unreliable most of the "consistent chain of output" is intended to, and for the most part does, reduce rather create confusion or inhibit manual flying of the aircraft. The basic purpose of a reversion to Alternate Law and if necessary (the two are not always coincident) the disconnection of the autopilot is to simplify the pilot's means of controlling the aircraft by removing the autopilot and/or some of the more refined protections of Normal Law that might be adversely affected by the failure while the crew ascertains the source of the problem and configues the systems to best minimize the effect.

Some failures by their nature create greater problems than others. In any airplane a loss of valid airspeed or angle of attack data creates the potential for spurious stall warnings as have been recorded in the QANTAS, Air Caraibes and NW reports. This does carry with it the potential for some confusion, and it's precisely for that reason that diagnosis of validity and flight control response to such warnings following such a corruption of data are handed back to the pilot. In all the instances that I'm aware of with Airbus aircraft the crew was able to effectively determine whether the warning was valid and maintain positive control of the aircraft. This is what you would expect from a competent crew. In simple terms, if one moment you are flying along "fat, dumb & happy" and the next your airspeed starts to rapidly fluctuate or becomes grossly inaccurate without equivalent changes in pitch and power, it's a reasonable diagnosis that a stall warning occuring at the same time should not be considered valid unless other basic aircraft parameters have undergone very significant change. So, yes an invalid stall warning can be a source of some confusion which the crew needs to resolve, but no, it should not be a serious hinderance to maintaining basic aircraft stability as long fundamentals such as attitude, power, and trim are observed and managed by the pilot.

Also, re: Rudder Travel Limiter. If due to corrupt AS data the ship loses Rudder protection, and possibly all limit to Rudder travel, would you think that a Rudder input might put beyond-fail stress on the VS? To wit, as AS data is lost, could the Rudder Limit reach a status of Low Speed restrictions as determined by the computer? In other words, could the Rudder's travel be deflected enough to convince the sensors in the tail in concert with corrupt a/s that low speed had actually been obtained allowing this deflection to be the new parameter, though the a/c is still travelling at M.80?
Actually, as has been pointed out here several times in discussions regarding AA587 it is possible to create stresses beyond the designed limits even with the Rudder Travel Limiter fully functioning, so yes it's certainly possible for this to occur with RTL failed regardless of the deflection limit at the time of failure.

I think the question your driving at is: Since the RTL system on the A330 autmatically freezes the maximum rudder deflection value based on airspeed at time of failure is it possilbe for it to be conned into setting an erroneous max. deflection value prior to failure if the airspeed information becomes inaccurate? The simple answer is that I don't know for sure. The more refined answer is that such a situation, if it is possible at all, would have to be considered extremely remote. For it to occur there would have to be not just a simultaneous failure of the three different airspeed inputs, but those failures would have to occur at near identical rates and yet still allow the measured airspeed values to remain sufficiently consistent with the other air data inputs so as to not cause a diagnosed failure of each affected ADR. Any substantial differentials between the airspeeds or failures of the ADRs would cause the system to consider itself failed and lock the max. rudder deflection value at that point.

Even then, based on my experience with the aircraft I suspect that the designers may have considered such a case and included a test for it in the limiter's logic. Though I'm sure there are some very remote failure cases that likely haven't been considered, I think many here with a layman's or even a pilot's general understanding of the flight control systems fail to give sufficent credit to the designers/engineers for the degree to which they have gone to try and analyze and account for each possible failure mode. It doesn't make the system perfect, but it is a lot more so than you'd judge from the theories you are reading here, and in most cases it is a significant improvement on the previous generation's accepted standards of safety (i.e. no limit freeze function, just a warning to use rudder with care).

Quote: Will Fraser

PJ2 quoted P,R,Y limits a/p; ELAC quoted 3 Degrees deflection per side; I had thought 4 as well, I believe 6 total is correct.
This is not correct. My statement to you in response to your question was:

That value is dependent on speed, which according to my manuals at M.80 and FL350 (272 KIAS) is approx. 6 degrees (not 8).
You had stated 8 degrees of deflection which was incorrect. I referred to the same chart that PJ2 has recently posted and provided you with the figure corresponding to M.80 at FL350. You can make your own estimate if you like. If the aircraft's speed had been reduced to 260 KIAS (the turbulent air penetration speed below transition) then the limit might have been slightly, but not significantly, higher.


Last edited by ELAC; 29th Jun 2009 at 10:01.
ELAC is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.