overthewing, that's correct. Although I don't know if the FO who was PF at 0200 had been PF since Rio.

welsh wingman, the crew apparently flew together from Paris to Rio. There was a long layover scheduled, and it was also Pentecost Sunday weekend, so the wife of the junior FO traveled to Rio for a short holiday.

Hi,

takata
Sorry to return on the work bench .. but ...
Seem's the "commandant de bord" (captain) was not aware of the Air France rules concerning the longhaul crews :eek: .. as it's written in the BEA report interim N°3 that the captain asked to the F/O Mr NoName 32 years old if he was a "fully qualified co-pilot"
Again I repeat .. it's a odd question coming from an AF captain.

You must be talking about the SOP part of the FCOM which is adapted to the need and will of each Airline.
But thanks GOD, most still apply commun sense.



A33Zab : Can anyone explain this?

http://i45.servimg.com/u/f45/11/75/17/84/wiz_pi10.gif

Very bizarre - Not a word on it in the last interim report ... !?

For the last minute (and probably much more ?) before and up to the UAS, there seemed to have continuous and repetitive 'automatic' switching between V/S mode and some other vertical mode (?) Was it a kind of reversion transparant to the crew ?

Selected v/s was nothing else than -5000 ft/min ...

What is the consequence on the initial climb ?
What is the consequence on the LOC ?

That thing is not minor, and must be addressed - Anything to do with the WRG message ?

SaturnV
 

Thanks for clarification re: the long layover (long enough for F/O 32's wife to holiday). Don't like using (deceased) pilots names, but sometimes it just causes less confusion in the end.

Think I'm inferring from the following:

(BEA Interim Report)

(BEA Synthesis note)

Therefore I'm assuming that the pilot who went to rest was the one who was called back around 2am, ie F/O 37, and F/O 32 had been flying since Rio.

Hi Old Engineer,
Yes! but you really did such a great job at missing close to everything!
:p
A33Zab did ask about the meaning of Vertical speed selected graph... (I suppose, did you Zab?). Which graph depicts something unrelated with... (I don't know what your post was talking about - seems imput logic?).

Vertical speed selected is a manual setting for: vertical speed ! (surprisingly)
This function is used by autopilot only.
This graph is showing the VSS value after reseting to current flight level due to intermintent returns of flight directors (as they were never turned off).
http://images.ibsrv.net/ibsrv/res/sr...s/embarass.gif

overthewing
 

Affirmative. It also explains why F/O 32 was the obvious choice to carry on as the PF and become the relief pilot (given the pre-AF447 position as set out by takata).

MountainWest@1546
 

MW

Thanks for your comments. In addition to that, I realize now that I completely misinterpreted this graph as representing the actions of the aircraft. After a lot of careful reading, I see that the graph indicates, intermittantly, the settings within certain of the computers. These settings continued to have effect continuously, as I now understand it.

So I think my post will not be useful to the discussion and am going to delete it. I'll plead to being my own French translator. :)

Although, above @ #1556 CONF iture has posted this graph again, with the comment "Selected v/s was nothing else than -5000 ft/min ...". And he closes with "That thing is not minor, and must be addressed - Anything to do with the WRG message ?".

I'm noting that this -5000 fpm appears to be without flight deck input, so I'd say the graph is saying something critical to us.

The english report is out.
Link can be found a few pages back.

How did the crew get to where they were at 0208?

0 h 30 mn The crew received information from the [AF] OCC on the presence of a convective zone linked to the ITCZ between SALPU and TASIL

1 h 35 mn 46 No response from AF 447 to ATLANTICO's request for the estimated time at TASIL

1 h 35 mn 49 Copilot’s ND scale changes from 320 NM to 160 NM.

[Between 1 h 35 mn 53 and 1 h 36 mn 14, ATLANTICO asked again three times for the estimated time at TASIL with no response from the crew.]

1 hr 50 mn 35 The radar display mode changes from WXR ONLY to WXR+TURB. The Captain’s ND scale changes from 160 NM to 40 NM.

[A little later [after changing the scale] he [the Captain] mentioned the appearance of Saint-Elmo’s fire and said that “it’s going to be turbulent” when he went to take a rest.]

2 hr 00 mn 17 Copilot’s ND scale changes from 160 NM to 80 NM.

2 hr 01 mn 46 Captain leaves the cockpit

2 hr 08 mn (approximately) The copilot in the left seat moved the weather radar gain control to maximum, after noticing that he was in calibrated mode.

2 hr 09 mn 53 Copilot’s ND scale changes from 80 NM to 40 NM.

"The signal corresponding to the “fasten seat belts” information was not heard on the recording."
_____________________________

The following English translation appears to be an extract or summary of AF procedures on use of weather radar.

In CAL mode without adjusting tilt, the BEA report notes that ice crystals will be hard to detect. The report also notes that "An oceanic cumulonimbus reflects radar waves less than a continental cumulonimbus cloud of the same size and height."

The change in gain from CAL to MAX on the PNF display was not made until 0208. There was apparently no change in tilt over this period.

The Graph, totally misinterpreted
 

takata,

Saw your comment in the "reply" window on recent comments. It's gone, but it made me smile. I'll never be a French speaker until I understand that they omit all information that would make themselves clear to outsiders. :)

Sorry to add to your workload of keeping things straight.

excitation,

Yes, I knew that; just thought the French would be closer in meaning.

No takata, we're not talking after but before !

What was the issue on the FDs before ... ?

Let me see if I can confuse the issue further. The Vs 'Zipper' coincides with the THS Zipper I noted earlier. Roughly one second control inputted POSITIONS, of VARYING value, approximating 1.5 degrees deflection, EACH WAY. Yes CONFiture, this is in autopilot, nine secs continuous prior disconnect.

Tell me it's in the FCS, and I'll be on my way. For you must know, just because bear notices something does NOT mean he 'gets' it. Also, I don't believe the ship is this.....erm, AGILE. Let's ask gums.

control stick feedback
 

Since the late 60's the only feedback to the control stick/yoke has been artificial. Exceptions were the planes with mechanical linkage to the control surface actuators, and even those had little "direct" feedback. We called it "boost". We also saw yaw and pitch dampers that the clever engineers provided us to "help" keep the pointy end forward and reduce pilot-induced-oscillations, etc. The VooDoo and Phantom and others had bellows that used pneumatic pressure from an "orifice" someplace to stiffen the stick. We also had bobweights to provide a gee indication on the stick. Real easy, and kept us from commanding more gee than we wanted using only a direct hydraulic valve. OTOH, roll was usually a simple spring mechanism that made it harder to command left or right the further you deflected the stick/yoke. No dynamic pressure feedback, just a spring doofer.

For the non-pinball wizard pilots, some type of force feedback for the stick/yoke would be nice. Unfortunately, airframe vibration or buffet is different. You have to have "touch", and you cannot teach "touch". Helped many young pilots learn to fly a new jet ( three of them), but I could never "teach" "touch".

"Can't you feel that?"

"No sir."

"lemme do it and follow me thru"

etc.

Many of we old dinosaurs would like an electronic system that "feels" like the old planes. This is possible, maybe even preferred. But in the end the pilot has to trust the instruments ( unreliable and confusing as they were with AF447) and somteimes rely upon some "touch" as to how much should I pull/push, and is it a mach buffet or stall onset, or a big piece just fell off ( the 'bus incident when tail came off due to excessive commands by pilot)?

There is no substitute for airmanship, dammit. If the public wishes to ride in an airplane with no human up front, then let's try it for a few months. We could program the jet to takeoff, cruise and land, while avoiding bad weather and making radio calls to appropriate agencies, etc. And when a situation arises that the cosmic engineers had not thought of, then......

Hi overthewing,
Yes, everything looks right, except for take-off from Rio: I don't know if the F/O was the same PF.

The first fully automatic trans Atlantic flight was from Newfoundland to Brize Norton, UK, in September 1947 by a C54.

The idea has not caught on yet ! The flight was from runway to runway, only. Humans were still required on the ground, from and to the terminals.

English interim report location:

FLIGHT AF 447

Liberté, égalité, fraternité
 

takata


This is where I believe the Air France way of doing things is fundamentally flawed. Most other airlines have a distinct chain of command, Captain > Senior First Officer > Junior First Officer. In this case the older, more experienced first officer should have been "In Charge" while the Captain was resting. [This would not prevent the Junior FO controlling the flight and making tactical decisions but may have encouraged the Senior FO to be more assertive when things were not going so well.]

This may have resulted in a better outcome !

Indeed.....

Airspeed from groundspeed
 

Hi takata,
Airspeed from groundspeed, V/S, windspeed and temperature is shown on this graph: CASfromGS, It is based on wind 45 kt @125 deg, ISA +11 deg C, DFDR data for v/s, altitude and HDG.

The blue dots are Mach from ADR, pink is Mach from GS, wind and v/s, and the green dots show the corresponding CAS.

DozyWannabee, I note that takata's recent note about the plane aimed down when the PF requested NU and your comments about not seeing a conflict in that suggests another basic rule for "strange situations" such as the sudden drop to ALT 2 with no airspeed indication: Don't Play With The Throttles, Bunky.

The moment arm with the engines and their thrust do exactly the opposite of what you think you want. If you are thinking overspeed pull up. But don't drop the throttle. That will nose you down. If you are thinking stall, don't goose the throttle. That will tilt you up and slow you down. So don't play with the throttle.

But, for that matter, don't change anything except roll. Maintain pitch, don't touch the silly throttle, and keep the wings more or less level but fanaticism about it doesn't pay. Watch the altimeter. If it start changing rapidly aim the plane "slightly" to compensate. But don't play with the throttle except for VERY small nudges.

Roll control is problematic in the stall. When you're stalled, unstall first, then level the wings.

HN39 - I am addressing what to do upon entering ALT 2 with no airspeed indication. Once in a stall first course on the table is admit you are stalled. Second course is nose down with an urgency depending on rate of drop of altitude. Third course is airspeed. Fourth course is a gradual pull up once airspeed is high enough. Fifth course is the dessert of moving back to planned course.

If no airspeed continues use AoA vane as a clue whether you are stalled or not. When it "agrees" with pitch you're good to pull up - gradually. No AoA report in the cockpit? Um, Er, Ah, Boss - you're gonna die?

Thread drift
 

John has updated the very first post in this thread to include some essential links for all newbies here. Thx, great!

Can I also include some excellent weather analysis here: Air France 447 - AFR447 - A detailed meteorological analysis - Satellite and weather data by Tim Vasquez

JD-EE; Agreed, almost. If you have AoA or next best, valid stall warning on/off as a substitute of it, I would put leveling the wings third. If you let yourself be guided by AoA, you don't really need to watch airspeed for recovery and can leave it to course five.

van Horck, it would be nice if Tim Vasquez updated his graphics and analysis again, given the additional information in the Third Interim Report. e.g., the turbulence trace, and the sound of ice crystals.

The contrast between what the crew flying AF447 did at SALPU and ORARO and what the crew of AF 459 (35 minutes behind) did is striking. In essence, the crew of AF 447 thought they could ride near or at the top of the clouds, and never gave a second thought to the presence of Cbs, even though they were specifically cautioned about such.

For the lawyers seeking damages, their inattention and complacency is the equivalent of finding a vein of gold with the first dig of a shovel.

Reading between the lines, even the BEA seems at a loss to explain it. It will be interesting to see if the BEA can determine how many of the captain's and PF's rotations to South America occurred on flights through an active ITCZ.

Questions: Selected Vertical Speed.
 

CONF iture:
Exact! and the spikes @ 02:10:16 'CLOSING of the CAS monitoring window & ADR rejections'?

http://i474.photobucket.com/albums/r...LTUNSTABLE.jpg

Hi Old Engineer,
Before AP disconnecting (at 0210:05) the graph is showing a VSS of zero, not -5,000 ft. (one need to read the discontinuous green line) and the red line (real V/S) is not altered in any way by it. It is also showing a different active mode when autoflight is still ON.

The cyclical drops (regular pikes down to MIN value) is certainly due to this autoflight mode (hence normal). In fact, I did not comment on those regular pikes as nobody here knows how those values are sampled. The same channel may be used, when AP is ON, for recording another flight parameter at regular intervals. It looks for me to behave this way.

After FMGECs faults (at 0210:05), the mode change is obvious, next VSS value increase/decrease is now linked with FD resetings. But, before this point, the VSS values (including MIN) are not linked with FD as there is another graph for FDs showing nothing anormal before UAS.

SaturnV

I've asked Tim, but he was too busy to come up with an update, he made the document then, to help us in the days when no plane had been found.

I will mail him and ask him again I promise, but he is a busy man......


Update 14:50 Amsterdam time I've emailed Tim and will come back on this if relevant

Gums and JD-EE
 

Asymeteric Thrust

Other old(er) Dinosaurs than Gums might recall that the HP42's top two engines needed careful handling on the ground. The tail wheel could lift, unexpectedly.

JD-EE reminds us that asymetric thrust is not only what we practice when one engine is shut down. TWO (or more) Underslung engines have a pitch effect.

( Would that be equal to 4 units of NU trim ? I have no way of telling as my armchair is uncalibrated. I am sure that the answer is "...It depends...")

Takata noticed a small improvement in the recovery from the stall when the power was reduced, all too briefly - and then, sadly, it was restored.

What would I have done as PF ? I have not had the luxury of time to think, to change my mind from climbing with full thrust, to pitching nose down. Closing down BOTH engines to flight idle might be the right thing to do - but would FEEL very wrong. Descending at night through a Cb without any clear Airspeed...I know that my Penetration speed should be between 240 and 260 kts, but how can I tell... I must try to keep my wings fairly level if I can...

Perhaps.... it is as well that my armchair does not feel overloaded, too!

maybe, maybe not.

AF pilot unions have always refused the "distinct chain of command, Captain > Senior First Officer > Junior First Officer" as this would mean less trained/less qualified (read : less paid) "junior F/O".

The underlying idea being : total qualification of the crew is better with one captain and two fully qualified F/O's.

Now, of course, the two F/O's being "equal" means CRM problems can arise as we see here.
Obviously the PNF had a, slightly, better understanding of the mess, but at no time he ordered "my plane" (at the time, F/O's did not get any LHS-flying training).

One last comment, the older/more experienced F/O is not always the best choice for PIC.
Never flown with aging, not very concerned F/O's ?

vanHorck, thank you, and thank Tim as well for all his past effort.

The assessment that I would be most interested in is whether the loud noise that can be heard on the CVR of ice crystals hitting the plane indicates they had penetrated the Cb itself, or if not, how close would one need to be to a Cb to get that amount of ice crystals.

I don't recall that the BEA gave sufficient data to correlate the time when the sound of ice crystals can be heard on the CVR and the PNF turns on the anti-ice.

Hi A33Zab,
Well, not sure. In fact, I'd say that at around 0210:16, speeds could have been coherent again for a very short time and FDs were back, disapearing almost immediately. In between, a new set VSS was sampled, taking the actual value from the real V/S at this exact point.
CAS monitoring doesn't alter the VSS parameter (see 0210:05-0210:15 and subsequent values staying where they were while V/S increased or decreased). Its value freezes to the current V/S when FDs kicks off.

What about Captain / Senior First Officer / First Officer ?

Surely this means a special training for LHS qualification including ground school, sim training, line training with checkout as Senior First Officer.

And yes, pay is inbetween First Officers (not called Junior..) and Captains.

I don't see something wrong with that.

Hi Confiture,
Thank you for attracting my attention on this point. It seemed normal to me and no, BEA would not comment something "normal" without good reasons to do so.

Nonetheless, the fact that FDs were not turned off attracted also their attention in order to find what kind of indications could the bars have displayed on PFDs. There is a discussion on it and they are trying to reconstruct everything displayed for analysis. So, I don't think that there is any sort of conspiracy about that, don't you agree?

I really believe that BEA investigators are trying their best as to verify everything, but it takes time, as you know as well, to investigate every single bit of data. You should drop off this attitude of conspiracy as it won't help families or pilots to continue this way by discrediting them constantly.

Is it possible that the F/O who was PNF in the left seat was also reluctant to take control because of a lack of dexterity with the LHS sidestick? As a F/O, primarily sitting in the right seat, his experience and dexterity would with using his right hand. Indeed, I have heard mentioned elsewhere that when two F/Os are on the flight deck, the PF is always supposed to be the one in the right seat.

Hi SeenItAll,
Very probable, added to the fact that he wasn't sure of his own understanding of events. Very sad in fact as he seems to be always on the verge of doing the right thing.

Certainly this is material for reflection, and better understanding of the moments before, and when all of it started.....

Here is a set of graphs, that may help visualize certain parameters at the moments before the AP disconnect, and passing from Normal Law to Alternate Law - this is the moment of the AP and A/THR disconnects.

A red vertical line marks the moment of the Normal to Alternate Law change.

I notice a Pitch Nose Down, right before the Law change.
I can also see now, that the slight Altitude loss (if the BEA graph is accurate) was right after the Law change. At the time of the previous post, I thought it is before the Law change

One can visualize the increased degree of turbulence based on the areas marked on the 3 graphs of acceleration right before the Law change. There is an increase in up-down-left-right-forward-backward motion. The Vertical Speed around that time also show and increase, and decrease - reflecting the Vertical Acceleration changes.


http://farm7.static.flickr.com/6128/...7ddffc2b_b.jpg

Quote:

---

Originally Posted by xcitation (Post 6619901)

Very important point. So...

AP disconnects
a/c at 0 deg pitch => will descend
pilot stick back 3/4 causing nose up pitch 11 deg.
climb rate goes up.
after 4 seconds first stall warning
after 15 seconds altitude has not changed (downdraft/reduced thrust?)

So PF was correct to stick back but he over does it.
Stall alarm sounds, he eases off on the stick pushes it fwd briefly, alarm stops.
Then he resumes stick back to maintain altitude.
Now the THS starts helping PF to pitch nose up.
Now he is gaining altitude with help from THS.

Report #3, Page 111, Longitudinal parameters

Notice the insidious influence the auto THS (cyan) has on elevator position. At some points he is stick forwards (red line, 02:12:17) yet elevators stay around -30 deg (purple). Clearly this would add to the confusion of inputs.

---

SaturnV, “… how close would one need to be to a Cb to get that amount of ice crystals.” # 1584.
Some research and flight tests suggest that significant amounts of ice crystals can be found up to 10-15 nm from a Cb (measured from the edge of the red zone).
From experience around and in Cbs with ice crystals, it is most unlikely IMHO that AF 447 entered a Cb core, although the aircraft may have passed through or over the tops of smaller cells.
Ice crystals can be heard – a sort of swishing sound; this is often accompanied with a very wet windscreen (screen heating), which adds to the confusion. Furthermore the ice crystal cloud is very tenuous and might not even be classified as IMC in some circumstances.
Louder, harsher noises come from hard hail often found in the same situations, particularly over the core; this can be heard clearly and might on occasions sound like rain.

The best advice is to stay well clear of Cbs, the avoidance distance (from the edge of the red zone) being proportional to the size of the core. I have encountered signs of outflow cloud up to 60nm in one Caribbean cell and 100nm from a huge cell in Africa; basically any anvil cloud can have a high concentration of crystals.

Just because modern aircraft have high tech radars which automatically detect the severity of ‘this, that and everything else’, and in multicolour, does not imply that technology understands the effect of the conditions on a particular aircraft type, or on a on a type thought not to be susceptible, but now operating with different performance parameters, e.g. engine power, TAT.
Ice crystals caught the industry napping several years ago with engine problems – the engine design had changed, not the weather. At the time there were signs of blocked ‘tubes’ and overpowered heaters (TAT) but no connection was made with pitots.

Removing, at this point, the "turbulence" effect (the good clue to see its effect is to look at simulated longitudinal axis analysis tables provided), nonetheless you are right about the up/down caused.

Anyway, change of law is due to unreliable airspeed.
The pitch down (at longer term) moment is also due to the reduction of thrust at 0209:58 (Mach 0.80 selected), decrease is about 15% in few seconds.
Altitude is "indicated altitude", as well as speeds, and UAS is causing a loss of "indicated" altitude (but ISIS is barely not affected, about 100 ft). It is obvious when you compare pitch vs V/S curves and altitude altogether: the former increased but altitude indicated change is delayed by 10-15 seconds.