A quick query for the A320 experts - is manual THS operation available in all flight modes or only when auto-trim is disabled?

Incidentally, an elevator with sufficient authority to override pitch excursions at extreme stabilizer angles would need to be very large indeed and would make the aircraft very sensitive without some very clever stick force gradient!

It is available in all flight laws. If the A/P is on, using the manual trim will disengage it.

Agreed. But a look at the graph, after the first stall, you'd notice that up to around 120 kts, the elevator authority was enough to check both the stabilizer and the engine thrust moments, which is the reason why I wrote earlier that, had the stabiliser position been changed and/or the thrust reduced, it would have been a by-the-book stall recovery. (That's why it is quite important to know why the THS didn't move : jammed or un-used ?)

Thanks for taking the time Dysag. I take it from that that there are no common components that could have caused both problems. Cheers.
I'm not sure about this though
because if this is as good as it sounds then 2 of 3 AOA sensors wouldn't have frozen at the exact same time.
Framer

Thanks for the clarification. Presumably manual pitch trim overrides auto-trim?

This indeed makes the lack of THS movement all the more mysterious; was it frozen, being incorrectly commanded to the extreme pitch limit as a consequence of multiple AoA detection system failures - or did the crew simply make no attempt to use the manual pitch trim?

Perhaps the CVR will clarify this - does the FDR record manual pitch trim pilot inputs?

Regarding 'stuck' AoA sensors, it is unlikely that the sensors themselves would be made by different companies, even though the computer architecture might be. Thinks of it as 'same mouse, different computer' for a simile.

I've had faulty AoA sensors on another aircraft type; on one occasion we experienced stick-shaker at rotate, but on another occasion the stick pusher operated at 300 ft during a training flapless approach when the stall protection system falsly detected a rapid AoA increase, triggering stall protection way before the critical AoA would have been reached (as it is designed to do).

The 737 QRH states "Elevator control is sufficient to safely land the airplane regardless of stabilizer position"
BEagle said,Can I take it that when Boeing wrote that they were assuming the aircraft was constantly operated within normal pitch limits, and from what BEagle has said that all airliners currently in operation would be in a similar boat to the A320 with respect to unusual attitudes and extreme THS angles?
Framer

A bad batch of components does happen, albeit maybe not from new, but definitely after return to shop for fault rectification or modification. However, sensors contaminated with paint happens a lot more often.

Re. the Boeing 737 reference, was that extract taken from a 'jammed stabilizer' abnormal procedure?

Presumably there is an associated speed limit?

BEagle it is from the STABILIZER TRIM INOPERATIVE checklist and there is no speed reference....Actually it is leading you down a path to a F15 landning so I guess that is a speed reference really.

Appreciate the informative chat from TP on the phone (my AB AND social advisor:hmm:). Back to banging my 'drum' about 'training', except here I see more a mindset than a training issue.

We have an aeroplane that 'cannot stall' - not even the doorman can do it. It did and it appears the crew were taken unawares. At 45:12, ?stalled? we see a classic input of left stick on a stalled wing (Ab initio Stalling I exercise) producing the expected roll to the right (struggling with the FDR trace here on aileron position, incidentally - stick left, aileron left (red) up??). ?Is it possible for the FBW to reverse the aileron response or is the trace wrongly annotated - or was there an 'override'??

The danger, as we progress further into FBW and software is that we will lose the basic skills of handling. The system is fantastic - in normal use - but has apparent serious flaws in the abnormal.

Now, back to the CAS query from my post (currently 907). I believe Lemurian replied assuring me that those readings were false and due to 'unusual manoeuvres. His reply has been lost in the body count. I SUSPECT I probably have more experience of 'unusual manoueuvres' in swept wing a/c than he, and have NEVER seen that size of CAS fluctuation, not even with a wing mounted pitot and certainly not with a c/l set. Also I see only an increasing pitch (admittedly 'extreme') and changes in in roll. I am becoming convinced that several bits of the system were not telling the truth here, which renders the 'Titanic' principle worthless. Anyone else expect to see 3x200kt speed excursions in 10 seconds in those manoevres?


EDITED to say I've realised I misread the aileron traces -doh!

BOAC
Good to see you are finally reading the report! Re the CAS the report states:
CAS invalid below 40 kt. I can't explain the spikes but I'll bet Sir Fred's pension that the actual a/c airspeed did not follow the graph!

Re that aileron trace: would have been much easier upside down!
Red - left aileron - up on the trace = down on the wing, blue - left aileron goes the other way. No reversal, logical because the roll control is Direct throughout from 15:45:15.

Training - on type rating training I am always very careful about stalling and direct law handling, there is a fair amount in the course.. but it was probably that doorman who wrote in the FCTM, and I quote...

They might just ask him to rewrite that bit...
TP

BOAC
What I wrote was that these readings are in all probability the result of the pitot/static sensors being in a vastly disturbed airflow...nothing more and nothing less...Could we call them "transients" ?
As for our respective experiences...as the Arabs say : Maalesh ! At least over you I have the advantage as both the 737 and the 320 are among my ratings.
But you're right, I've never ever flown any of my jets in (as opposed to into ) a stall...old and not bold, alas !

Beagle
Yes, man trim overrides auto-trim.
As for the un-moved THS, yes we need more data...

Additionally, since the data is CAS and downstream of processing by the ADCs, it is possible [speculation] that the data are erroneous due to inappropriate corrections being applied/extrapolated outside the expected range of sensor operation.

Spikes in CAS graph are due to sloppy plotting
 

Since the report says that CAS data between 15:45:42 and 15:45:53 was "NCD" (no computed data), another name for "invalid" data, there was no value that could justifiably have been plotted. The values that we see are artifacts, and have no relation to the actual air speed.

Connecting lines should not have been drawn, and arguably even the invalid data points should not have been plotted. They have no significance whatsoever.

In fact such lines should never be drawn, unless the expected rate-of-change is so low that linear interpolation will provide a sufficient approximation.


Bernd

Hello, Bernd.
Have you started a WB-graph of this accident as I think that there are just a few - but important - facts that we should know ?

Smack on the wrist from teacher - I have amended my interpretation of the aileron traces.:ugh: I had assumed the CAS was 'invalid' only due to going below its lower limit.

Yes, TP - it is time for the AB folk to acknowledge the need for upset recovery.


......and MFS - that last paragraph popped my CB:)

If the guys in the cockpit were getting the same CAS readings........................

Since the AoA vanes APPEAR to have behaved normally for the first part of the flight, can anyone tell if they were being properly heated or are we looking at an insidious icing problem rather than the 'painting' issue which might just be no more than a valid 'cautionary warning'?

PJ2, your last post was interesting ... but it disappeared (?)


I read you 5/5 TP, and that's why I do ask the BEA :

Where is the load factor graph ???

From time 15 45 57 the elevators were in the amazing full down position when the Capt (We don't know for the FO ...) was actually pulling on its side stick.

As the speed was building up, I suspect the elevators were just trying to counteract the increasing effect of the inappropriate THS position, and so in order to respect the load factor limitation ... :hmm:

That Airbus is a really complex machine .............






For Safety Concerns,
AeroPeru and Birgenair were both at night time with no external visual reference.

A320 Probe Heat
 

One pushbutton (pb) on the overhead panel controls all probe (pitot, AOA, TAT) and window heat, static ports. Normally in AUTO - the AOA probes would have been heated when one engine running or a/c in flight. Can be selected ON - all heat then on. There is no off selection possible it's either AUTO or ON.
There are three independent Probe Heat Computers (PHC) which control and monitor, and provide warnings for each individual sensor. They are grouped into CAPT,F/O, and STBY probes - so a heating failure/overheat to two AOA probes would show as two ECAM warnings provide by two PHC's.
There is no mention of ECAM warnings in the report.

CONF
Do you think the BEA would have mentioned Dual Input on the sidesticks? It is very relevant to any AB FBW event.
TP

You know, TP, in that kind of event, I would almost like to see some kind of "dual input" or even better "priority right" as the guy on the right seat has no idea of the kind of input is made by the left seater ... so he does not know why the situation is deteriorating ...

1. Is my Captain suddenly incapacitated ?
2. Is my Capt in a bad day ?
3. Is its sidestick suddenly u/s ?
4. Is it anything else ?
5. Should I intervene ?

I personally don't think I would have had the nerves to remain in the stand by mode ...

The FO Side Stick Position is not number 1 on my most wanted parameters list but as the BEA already gave us not less than 40 parameters showing absolutely no kind of status change ... why not ?

CONF iture -

I asked this once but was criticized for doing so by some "expert" here. Not knowing about FBW or Airbus aircraft, I asked if there was any kind of 'priority' in the side sticks. You mentioned 'incapacitation' as an example. If one pilot has his body weight against one side stick, and the other pilot attempts to move his, what happens?

Thank you.
That correlates to my experiences in the SIM.
My factory AI manuals do not make any mention of the fact that a HS trimmed full stop cannot be controlled by elevator. Neither was that covered during my training.

Prio lockout
 

Depress the red disconnect pushbutton on the joystick for 40 secs and the other joystick becomes locked out.

Maxfli -
Thank you.....but.....40 seconds?! That seems like a very long time in a possible critical situation.

Sidestick priority
 

DC-ATE,

as soon as you press the button, your stick has priority. It was used "in anger" at least once to recover from an upset after takeoff when incorrect re-wiring of a stick after maintenance caused reverse roll commands.

If you press and hold the button for 40 seconds, the priority status becomes latched, and your stick keeps priority even when you release the button. Useful, e. g., as mentioned, when the other pilot becomes incapacitated, and his arm still leans against the stick, deflecting it.

If the take-over button (which doubles as A/P-disconnect button) is not pressed, both stick inputs are summed, limited to the amount of a single stick's maximum deflection.


Bernd

DC-ATE, you've misunderstood.

As I understand the concept:

Press and hold the 'on side' take-over button and control will be instantly transferred to that side. Keep it pressed for 40 sec and the other side will then be deactivated until reset, allowing the other side to maintain control permanently.

Thus if an incapacitated CM1 falls against the sidestick and keeps it displaced, as soon as CM2 presses his/her take-over button he/she takes control. After 40 sec he/she can release everything as the CM1 sidestick then has no effect.

Correct?

Not quite, as
...no, you don't need it to be held. depess the button and "Priority Left...(or Right...) will be announced.

BEagle - correct.
Lemurian - you do have to hold the priority button. If you release within 40 seconds the other sidestick becomes live again.
Tree said
You are right, no mention, and the current course no longer trains jammed THS as a separate item, it is included in the dual hydraulic failure training. However with a jammed THS you would only use CONF 3 for landing - as FCOM 3 says "Do not select configuration FULL so as not to degrade the handling qualities".

And in the FCTM when discussing HYD G+Y failure it says.. (not suggesting this failure happened, but it's good background)

"(1) The stabilizer is lost. In alternate law, the autotrim function is provided through the elevators. At landing gear extension, switching to direct law, the autotrim function is lost. However, the mean elevator position at that time is memorized, and becomes the reference for centered sidestick position. This is why, in order to ensure proper centered sidestick position for approach and landing, the procedure requires to wait for stabilization at VAPP, before landing gear extension.
If this procedure is missed, the flare and pitch control in case of go-around may be difficult. " (My bolding).


The information is there but you have to dig for it and understand the implications. So as this was effectively a go-around in CONF FULL instead of 3 - the pitch control was difficult...
Minor note - THS deflection was 11.2°, full deflection is 13.5°, so not quite at the stop.
TP

TP, I think Lemurian was referring to 'normal' take-over, whereas as you correctly surmised, my query concerned a take-over against an undesirable input from the other CM.

By keeping the button pressed, the CM who has taken over will maintain control - but needs to hold it for 40 sec to 'isolate' the other CM's sidestick.

Is that it?

:ouch: :uhoh:
Thanks. Shows how much I've used it !
I've only used it to confirm the "I have control ..." bit. I've never needed to de-activate the F/O's sidestick.

BEagle
Perfect !

I have to say it's fascinating watching current, type rated pilots educating each other about more and more subtle interactions, reversions and sub-modes. I wish I knew that much about the aircraft I fly; *gets books out*.

I've just read the initial report and one section really hit home:
I remember being called off standby as P2 for an air test, very early on in my flying career. Great! I thought, that looks like fun... I won't bore with all the details but we did end up in an emergency descent with a door warning and both packs and one engine failed at one point (never got it relit). I had a fair amount of difficulty keeping up with the schedule and assumed the other pilot & engineers knew what they were doing. With the benefit of hindsight, I'm not sure that they did, especially as we had about 30 hangers-on in the back along for a 'jolly'. :ooh: I really don't need to add that I had no training in and no adequate briefing for what we were attempting.

When you see:
followed by:
you can almost feel the time/commercial pressure building up. You end up in the absurd situation that you attempt something as a group that you would have never attempted as an individual. Two line pilots with an "observer" going where test pilots fear to tread. I'm very sympathetic as I've been there.

The precise technical details of what actually went wrong, near the end, will come out eventually. What stands out, I feel, is that when test pilots "test" something, they are prepared for that test to fail: indeed, they are *expecting* it to fail in a multitude of ways, having extensively researched the subject. For a normal line crew to be put into a situation like this is, at the very least, regrettable - in the extreme it could be argued it was a form of corporate manslaughter...

FullWings
Have a look at the CRM book, and it's certainly not commercial pressure...
That the cockpit set-up was a bit abnormal is saying the least as you've also noticed that (page 43) the ANZ pilot actively participated in the tests (Granted, just the "systems" test on the ECAM ) but still, it could have been done by either pilot or an engineer...
Who really was in charge ?

That has been the argument of a few of us for some time, but we cannot, with the data we have, rule out the possibility of a technical failure that prevented them from recovering from their predicament.

Are we reading the same report? At 340ft. having been in clean config for 7 seconds at speed of 263kt and 15 degrees bank angle and with full up elevator command and power at climb it was still uncontrollable!


From 15 h 45 min 57 s, the Captain’s longitudinal input was nose up, the elevator was at 14.5° nose down. At 15 h 45 min 58 s, the flaps and slats were selected to position 1, then to position 0 two seconds later. At 15 h 46 min 00 s, the stall warning stopped and was followed by a CRC warning that corresponded to a Master Warning, which stopped two seconds later.
At 15 h 46 min 01 s, the pitch reached a maximum of 51° nose down. The bank angle was 45° to the right, the speed was 183 kt and the altitude about 1,620 ft. From this moment on, the Captain’s longitudinal input was to the rear stop. At 15 h 46 min 02 s, the thrust control levers were pulled back towards a position close to IDLE (about 6°). The EPR on both engines dropped towards 1.2.
At 15 h 46 min 02 s, the thrust control levers were placed on CLIMB. The EPR on both engines increased towards 1.25.
At 15 h 46 min 04 s, a GPWS TERRAIN TERRAIN warning was recorded.
At 15 h 46 min 05 s, another CRC warning (Master Warning) was set off.
The recordings stopped at 15 h 46 min 06.8 s. The last recorded values were a pitch of 14° nose down, a bank angle of 15° to the right, a speed of 263 kt and an altitude of 340 ft.

So in the last 5.8 seconds the a/c pitched from 51 degrees down to 14 degrees nose down.....it increased 80kts from 183kts to 263kts.
Does a 37 degree pitch change in 5.8 seconds with an average speed of 223 kts equate to a g loading similar to the limiting figure of 2.5g that some have mentioned?
Regards Framer

framer;
Perhaps what you're really asking is, was the aircraft's response limiting the ability to increase the 'g' load beyond the limit, possibly to avoid striking the water? How close the aircraft was to a tertiary stall would be one of the open questions of course.

Another point of view is, this is an interesting question because I believe it may touch upon a basic misconception of fbw and the 320 automation. The assumption may possibly be, the fbw automation will, through the flight controls, prevent a 2.5g load in all circumstances, (because that is the control law and that is what the system is designed to do).

At this point, we have an aircraft that is entirely out of control with, for reasons we do not fully understand yet, a THS angle of just above 11deg NU and a CAS of 263kts.

The key is in understanding that what the airplane was doing in pitch as a result of the 11+deg NU THS setting was, in the last 6 seconds or so, beyond the capability of the flight controls to counter regardless of the operating control laws and the flight control orders. This may be one interpretation of the L&R Elevator position parameter which, in the last 9 seconds or so is 12.5deg ND, perhaps, (we don't know) trying to counter the rising pitch and increasing 'g' load. At those speeds and pitch rates, the 'g' load would have been quite high I should think.

Both an acceleration rate and a pitch change in the time stated is obviously possible because that's what it did.

We do not know yet why the THS remained at it's last setting. From what I understand about the admittedly-rapidly changing events and circumstances, there were a few "windows" of possible response by the ELACs and SECs to begin reducing the THS setting. For whatever reason, that did not happen.

Clearly the airplane was responding, during the rapid, last dive, to the trim setting because the data shows that the nose was beginning to rise again, just as the aircraft struck the water.

It was mentioned in an earlier post that we don't know that the crew did not try to manually change the THS setting. I agree - we don't know that from the data that has been available to us all. We don't even know if the THS was not in some way jammed at it's last position and the crew tried very hard to move the manual trim wheel on the pedestal. The data shows no movement of the THS. We do know however, that, assuming a normally-operating THS, that it is manually, mechanically possible to change the trim setting. There is not enough data to conclude anything further.

Still thinking about this, however...

So, Bernd, these lines have been automatically drawn (?) and have not been manually erased (?), or have been manually drawn (?)

Thanks PJ,
I'm still not sure of the answer though...are you saying that it can't be calculated from the info available?
My line of thinking is that the THS and the aft side-stick are combining to try and pitch the a/c nose up. The attitude is pitching up at a rate of about 6 degrees per second (in the last 6 secs) . Is the reason for the pitch rate being limited to six degrees/sec because of aerodynamic forces or does it relate to the onboard computers limiting the rate (through elevator) with max g loading in mind?
PS I am aware that I'm out of my depth here with respect to Airbus systems knowledge and accident investigation knowledge, I appreciate the answers supplied by those of you more kowledgable than I.

framer;

Well I think first of all those are good questions to ask and would be challenging for even Airbus pilots to answer because these are flight regimes with which most would notbe familiar. Lest some take from that statement the false notion that there are pilots out there "who do not understand their airplane sufficiently, these control laws are at the very edges if not beyond the normal flight envelope and are in test-pilot and engineer territory where no airline pilot has a right to be.

The actual calculation of the 'g' load could be done quite easily - I imagine that it's a matter of inserting the knowns (TAS, pitch rate) into the equation but I am not a mathematician or engineer so will leave the rest of the answer for someone who is qualified to respond.

As stated, the elevators were already fully down to the stops. We don't know whether they were effective in reducing the 'g' load or whether the high speed and high THS angle made rendered their deflection less, or even not effective - that is a question for another Airbus engineer. Is there something to learn from the last eight seconds or so? The first people to answer that question will be the Investigators.

To the point many have brought up here however, is there something to learn regarding C* Laws (traditional name for fbw laws) or are we in territory so far outside the envelope that engineers cannot reasonably be expected to design a machine that is crash-proof. We must keep in mind that despite vast misconceptions on the part of some, especially those who do not fly the 320, automation, fbw and control laws do not make the airplane crash-proof nor was that ever the claim, nor would it reasonably be the expectation of flight crews.

The next question that is naturally to be asked however is, let us say that the control law did do as it was designed and limited the 'g' load to 2.5g's. What then, in this particular accident? Given a higher altitude, was the aircraft entering an increasing fugoid response which would have eventually exceeded either the airplane's or the crew's ability to regain control? We simply don't know.

Control laws are designed to take maximum advantage of the airfoil while keeping the aircraft out of the stall regime, (although high 'g' loads will pass the flight control computers to "alternate law, restricted or unrestricted"), and/or keep the structure within certification design limits. Must these same laws also cater to those circumstances where there may be insufficient altitude to recover?

That is about the limit of what I can say to someone who doesn't fly the airplane and it is getting close to the limits of my knowledge as well. Best leave the question open for others more capable than I to have a go at.

1) 2.5g if that's it, is a bit conservative I'd think... if you couldn't pull the 3g safely in such a situation, you'd be limited whatever the control situation..

2) but I don't think airframe limit loads at those airspeeds are an over-riding consideration. Recovery - even if you go to ultimate & bend it a bit (.1% strain from memory) is more important.

3) Why isn't there a... 4) Yes, I suspect the trim setting is crucial to all this, it should revert (or be set manually back to) somewhere near trim for Vref + a bit

So shoot me... :ouch:

Ref: See Python comment in Schiphol therad for what a programmer would do, not my post, but I get the idea... I'd also revert back to the user - 'Cannot resolve this condition - here are your options... 1) Take total control 2) No other options

The Flight Computer has lost it, why revert to any mode other than... 'You have control'

Thanks PJ,
I guess it's a case of hurry up and wait now.
I think the entire CVR transcript will help with some of the human factors
unknowns as well.
Regards Framer.

The problem with the CVRs is they are only available to the accident investigators. And while that is perhaps appropriate because of loss of life of the crew, that really limits the info to a very few. If the information was available to other pilots (only) outside of the investigation, we might get more meaningful information.

HarryMann;

With respect to these details, I wouldn't take them as hard-and-fast limits/restrictions. I'm not a test pilot and have only had the need-to-know course a long time ago so these areas are far less familiar to me than normal ops on the 320. With the passage of time, one gets very good at flying this airplane and learns a trick or three but where these pilots were is well beyond airline experience. I'm just trying to think a bit outside the box with what I think are some possible but not definitive responses to the question asked. I recommend (when reading my post, anyway), a "hmm, interesting..." stance with a pause to keep the question open rather than hardening conclusions into 4 dot points. It's not as though I don't agree with your points...I just can't comment on them because I don't know enough about these 320 flight regimes.

DC-ATE;
At least in Canada, and possibly the US, it is against the law to make the CVR available to anyone other than the IIC (Investigator In Charge) and his/her investigating team.

There is no benefit nor reason to provide the CVR to "pilots" even though they may be involved in flight safety work for their association, (you don't mention this qualification but I didn't think you'd mean that the CVR would be handed over to just any pilots). Training, experience in flight safety work is needed to interpret all data including the CVR.

Providing a transcript is another matter and while not protected in law, is still not encouraged though especially in the US, it is done all the time, ostensibly for CNN's and Faux News' ratings.

The investigators almost always draw correct inferences and conclusions in the process. It just takes time, as you likely know.