AF447

[greenspinner]

Lemurian,
With the only data we have right now, and as per these data the only concrete things we can say now is that Icing is the most probable cause.
But as I said in my previous threads, it must be really severe Icing condition, like an Iceberg at FL 350

[FlyingCroc]

Iceberg?

A hell of a storm! Check this out on Tim's webpage.
Air France 447 - AFR447 - A detailed meteorological analysis - Satellite and weather data

[Boxkite Montgolfier]

MR8

It is inconceivable to me that experienced pilots would not have searched exhaustively for wx returns at extreme radar range when approaching the notified presence of ITCZ activity. It is reported that a turbulence report was made 10 minutes prior to the ACARs download.
Pitot probe failures, if true, will have have made safe penetration and control very problematic. Thanks to many for a very informed debate, finally!

[MR8]

Wytnucls - I'm sorry, but you are wrong! If you do that in the current scenario, you will STALL.

The 5 degrees nose up is when you are in CLIMB above THRUST RED ALT and above FL100.

The checklist clearly states further:

When at, or above MSA or Circuit Altitude - level off for trouble shooting.

Initial pitch/thrust for FL200 - FL 350 above 190t is 3.5degrees/70.9%N1 - and that is NOT a MEMORY item.

If you would go to 5 degrees, you will fall out of the air before you clear the weather.

MR8

[Wossname]

Hand Flying @ Hi-FL

P51Guy & Parabellum - My questions were in response to Captain Crunch's assertions that seemed to suggest that a) aircrew should not wait years until an enquiry reaches a conclusion, but should act more proactively (his V2 example). Secondly, in his experience, it seems that pilots are less able/experienced in hand flying at altitude, therefore, less prepared to deal with hazardous, low-frequency conditions.

While I realize that sims can only throw so much into the mix, the idea that perhaps increased automation along with enhanced safety envelopes is also bringing a concurrent degradation in 'real' flying skills, is a worrying one.

I'm sure that Airbus don't know what brought down the AF lads, even if they have narrowed down the possibilities, but perhaps they felt it prudent to remind operators and crews of previous directives.

Cap'n Crunch, thanks for your candor.

Regards,

[SaturnV]

MD100,

In the closed thread, Snaproll posted these speeds:
http://www.pprune.org/4969395-post571.html

According to the QRH and based on a weight around 210t:

(speeds are approximate)

Green Dot (minimum clean speed): 245 kts

Turbulence penetration speed: 260 kts

Vls w/ 0.3g buffett margin: 235 kts

Speeds are all indicated so no ISA deviation necessary.

Not much margin.

[c130jbloke]

Recovery op sitrep

Sorry to shunt this one onto a new aspect, but does anybody know if the crash site and / or a floating debris field has been located and confirmed yet ?

I have tried the usual news sites, but I cannot make it out and the area in question is pretty big for a search which is just adding to the mystery of what happened.

Apologies if this is a bone question.

C130JB

[ChristySweet]

Please excuse this layman's intrusion and posting mainstream media but this article has some info on the weather that is pertinent and seems to have been sourced to knowledgeable persons who are on the same page as the discussion here.

Air France jet's flight-control system under scrutiny - Los Angeles Times

>>>>>>>>>>>

"...Meanwhile, new analysis of the weather in the vicinity at the time of the crash appears to cast doubt on earlier reports that the plane encountered severe thunderstorms, lightning and wind gusts. Though there were storms, they were almost certainly less intense than those sometimes encountered above the United States, and lightning was at least 150 miles away, said Greg Forbes, severe-weather expert for the Weather Channel.

Forbes said an examination of weather data for Sunday, including satellite images, indicated updrafts of perhaps 20 mph, far from the initial reports of 100 mph.

"I wouldn't expect it to be enough to break apart the plane," Forbes said.

.....

"Air France executives said the plane had sent out a series of messages indicating technical failures, confirming news reports in Brazil and data that U.S. aviation experts had already gained access to.

A series of serious electronic breakdowns occurred on the Airbus over a four-minute period before the jet plunged into the sea, said Robert Ditchey, an aeronautical engineer, pilot and former airline executive.

The sequence started with an autopilot failure and a loss of the air data inertial reference unit, a system of gyroscopes and electronics that provides information on speed, direction and position. That system has been involved in two previous incidents that caused Airbus jetliners to plunge out of control, though the pilots were able to recover.

The automated messages then indicate that a fault occurred in one of the computers for the major control surfaces on the rear of the plane. Such a failure would have compounded the problems, particularly if the pilots were flying through even moderate turbulence.

The last message indicates that multiple failures were occurring, including pressurization of the cabin. Such a message would have reflected either a loss of the plane's pressurization equipment or a breach of the fuselage, resulting in rapid decompression."

>>>>>>>>>>

[Blacksheep]

From reading the ACARS data and discussions with others in our unit, it certainly does seem that Cb penetration followed by sensor failures and loss of control is the most likely scenario.

Rather worryingly, from a number of comments, it appears that many pilots do not use their Wx radar properly.

Radar is not simply something that you switch on in the hope that significant returns will pop up - the display must be seen to even test and start using it

Spot on!

The radar detects water droplets only. It does not, indeed cannot, detect ice. The turbulence mode only detects turbulence that contains water droplets; if there is no rain in the turbulent air, the turbulence remains invisible. On a dark night, if there is no rain in the top of a Cb it is invisible. The Wx radar must be operated in pitch to scan various altitudes ahead: the beam covers a taller column at greater range than at close range and you can pick up the rain that appears lower down in the column that may indicate the presence of a Cb full of ice towering up above it.

Forgive me, a couple of dumb questions for my own edification. Firstly, if we don't know what caused AF to go down until the report comes out, what steps should crew take to avoid the same fate?

Cbs may sprout up in your face very quickly - while working in equatorial climes for decades I have regularly observed Cbs build up from virtually nothing to monsters towering up to 30,000+ feet in the space of fifteen to twenty minutes. This makes it especially important to actively manipulate the Wx radar tilt when flying in known areas of Cb activity and learn to interpret what you see. Do not imagine that what you see is what you get with Wx radar; the truth is, you have to work it and interpret. Honeywell and Rockwell Collins both produce useful pocket sized pilot guides on the subject.

Contact the Flight Ops Support people through these links and since you're pilots, try asking for a freebie:

Honeywell
Rockwell Collins

[Lost in Saigon]

AS of Friday June 05/09 the official stance seems to be that NONE of the debris found belongs to AF447. It is entirely possible that they are not even in the right area. They were only searching that area because the last ACARS message was received at the approximate time they would have been in that area.

I would imagine that they are now re-thinking their search. It is possible the aircraft flew in some direction for a considerable time before being lost.

[theamrad]

Maybe as a result of our discussion, pilots will start clicking the autopilot off at altitude and get an idea of how it feels in the real airplane on a nice day. It's absurd, imho, for your first experience hand flying at FL350 to be on a dark and stormy night with half of the instruments not working and a flashlight clamped in your teeth.

I concur.

Thanks to the more Airbus avionics orientated guys for trying to make more sense of the ACARS messages for everyone.

If all speed displays are off by more than 16 kts and reliable ADR cannot be identified, switch ADR 1 and 2 off and fly pitch and N1 setting as per the 'Unreliable Speed Indication' paper checklist.

Question concerning Airbus FADEC:
Notwithstanding Safety Concerns warning, and given the earlier posts trying to explain the ACARS messages. I presume with no sign of degraded engine performance and no loss of engine indications, and that PFD flags concerned only airspeed (and possibly alt) – attitude reference was still available. I presume with A/TH disconnected – the throttles allow the setting of a specific n1(from the unreliable speed chart etc), subject to FADEC derived limits??
Also – do the FADEC units derive their main sensed data from their own dedicated sensors – or is some derived externally (ADIRU’s etc)?

p51guy - I think you are mistaken. The pitch/power tables are very rudimentary. When flying close to the coffin corner at altitude and in HEAVY turbulence, you don't stand a chance, not even if you're a stick & rudder ace.

With an unknown airspeed teetering between or exceeding beyond the upper and lower bounds of MMO and min manoeuvring !! And not everyone has the comparitive luxury of even a GPS derived ground speed which isn't dependent on a primary display.

[clivewatson]

Blacksheep - it's clear that you understand the point I was trying to make!

MR8 and greenspinner - thanks for the clarification. Yes, I accept that the radar may have been left in a dimmed condition, but I am surprised that several Airbus drivers have complained that this can easily be missed by crew.

Even if you skip the radar taxi check on a cavok day (which I can understand), there is no excuse for not carrying out a function test when you decide to use the equipment to detect weather (one falcon written off after flight into CB due to crew being unaware that the radar red did not paint).

Even if you skip this test, how on earth do you start to scan for weather if you can't see a screen because it happens to be dimmed? Surely you don't simply set a tilt angle and hope for the best? Please tell me that you don't!

[Blacksheep]

The only way anyone will know for sure what happened is when the FDR and the CVR are brought up from the bottom of the ocean.

They might tell us waht happened but they won't tell us why. Even the best accident investigations rely heavily on data "interpretation" - another word for speculation. For the time being we have ACARS Data that tells us a great deal of what system failures occurred and these data at least permit our speculation to point towards a reasonable conclusion concerning the root cause.

Regarding the difference between Mmo and stall speed at high altitude, Mmo is set low in certification to ensure that the aircraft is kept clear of the edge of the envelope. There is disagreement between the authorities as to where it should properly be set - for example, the B767 Mmo is 0.84 by UKCAA rules and 0.86 by the FAA rules. The book figures are not clearly defined knife-edges.

[Caudillo]

I have no intention of speculating of the possible cause/s of this tragic accident, but I would like to ask a question that relates to the Airbus radar.

Several posters have mentioned that it is possible to have the radar "dimmed" and that this may not be noticed by the crew.

What I don't understand is how it is possible to select, test, tilt and operate the radar if it is dimmed to an extent that it prevents seeing any image or returns. Surely, in order to start any kind of scan using the radar, the image must be visible so that tilt and intensity scans can be seen and set up for the phase of flight.

Radar is not simply something that you switch on in the hope that significant returns will pop up - the display must be seen to even test and start using it - unless Airbus have some fancy system that I have not yet come accross.

Can Airbus pilots please explain the "dimming" problem that several have mentioned

Clive you're quite right. To try to give you an idea of how it works (or doesn't) In front is your ND - Nav Display, which is to the side of the PFD - Primary Flight Display, which contains horizon, altitude, speed, mode annunciators (FMAs) etc. On the ND you can toggle between overlays of weather, of GPWS terrain data, or nothing.

The dimming function doesn't refer to the operation of the weather radar itself, it simply changes the brightness of the selected overlay on your ND. So if you have is pointed right down, and turned to full brightness you be presented with a huge swathe of primary colours, mainly red. Much like the contrast of brightness on your TV you can then adjust the brightness of that colour to what is the most comfortable or least distracting. The actual functions of the weather radar itself are as I'm sure you can imagine, controlled on the weather radar panel elsewhere.

In my own experience I have found the weather radar on my Airbus fallible. A day or two before the Air France accident I flew through an area of numberous isolated CBs. All well defined in daylight in an otherwise clear blue sky and easy to spot and count by eye. Try as I might, scanning up and down with the radar for my own amusement, by my reckoning only 50% produced a return. When it did come, it was very good but for some reason others produced not a sausage. Not a problem on a day like that, but it gave me pause for thought should the same scenario be repeated with the addition of turbulence, night and embedded CBs. I do not think that a weather radar need be inoperative in order for you to fly into something.

[AMF]

backseatjock Article from today's New York Times, which links to above threads.

Investigators are pursuing a theory that excessive air speed -- potentially spurred by ice building up on electronic airspeed sensors -- contributed to the ocean crash of an Air France Airbus A330 amid heavy storms Monday, according to two industry officials familiar with the details............
.........................
If the tubes iced up, the pilots could have quickly seen sharp and rapid drops in their airspeed indicators, according to industry officials.

I'm not sure why anyone has to presuppose iced-up pitot tubes or that the cockpit airpeed indications were erroneous. Sharp and rapid decreases or increases in airspeed while within or above a CB that produces violent atmospheric shearing and/or updrafts/downdrafts are going to occur. All the cockpit indications may have been absolutely correct and shown the true picture of the situation if they were in one of the magnitude the weather picture indicates they may have been; that any aircraft's capablity to counter such very real and extremely dynamic atmospheric forces may not be enough (especially at high altitude and heavy weight where aerodynamic and performance are seriously degraded compared to even 10,000 feet lower).

To me the most telling and glaring ACARS message is the first one...Autopilot Disengament....not the flurry of messages that come a few minutes later. Something caused this. The autopilot's inability to hold altitude or being buffeted past it's pitch or roll limits will cause the AP to disengage. Given the context of where they were and the pilot's own message to his company a few minutes prior regarding turbulence, this has to be considered and from this moment the pilots were hand-flying with various flight control limiting protections against aerodymamic excesses removed.

If an aircraft also enters into an extreme updraft (such as those found inside or above a rapidly developing CB) the Autothrottle would begin to close when the airspeed rapidly rose, or the pilots might disengage it if the ATs aren't reacting as swiftly as the pilots need or want.

And if, while in an extreme updraft, the A/P disengages due to the inability to hold the selected altitude and the pilots do what we're normally trained and told to do in that abnormal situation...attempt to maintain pitch and wings level in order not to exceed AOA and load factor limits and control the speed if with thrust while letting the aircraft ride the updraft to a higher altitude (or to a lower one in the case of a downdraft)...that aircraft is going up.

But if the aircraft is already flying at it's maximum enroute altitude near the top of it's operating envelope for that loaded weight when an extreme updraft encounter begins, it could suddenly find itself higher than it should be...a thousand feet higher?...higher still?.... and therefore in a very perilous situation when it exits the updraft. If it's still being buffeted while at that higher altitude, engines spooled back due to the A/T or pilot's prior attempt to counter the indicated (and very real, not "erroneous") acceleration towards overspeed due to the shear accompanying the updraft, but now suddenly the airspeed is rapidly decaying and excess available engine thrust to stop and reverse decaying trend even more limited becaue of altitude, the situation is downright hairy.

Did this happen to this particular flight? Only the FDR will tell. But what can be said is that it's a very real and possible scenario for any heavily-loaded aircraft near to top of it's current operating envelope if it enters or flies just above a quickly developing CB.

Recovering or attempting to recover from this scenario while descending (and you MUST descend to recover) through the severe or extreme conditions inside or around the CB itself could easily result in a cascade of failures and/or failures of electronic self-monitors and sensors within the aircraft, especially if one or more engines flame out.

That "aerodynamics lost" situation is a KNOWN and ALWAYS-present threat in high-altitude flying that we mitigate through choice of crusing altitudes, routes, and actions to get ourselves away from near the edge of the performance envelope if conditions change or develop that could put us outside what was previously comfortable. It's present even with no thunderstorms within 1000 miles. If you pass into or are pushed by extreme and sudden atmospheric forces (like are found in or around CBs) into an aerodynamically unviable situation/jet upset, very few air data indications in the cockpit will be reading correctly even on a clear, sunny day after that occurs.

There's no need to pre-suppose any pitot or probe heat failures or erroneous readings as a cause for the worst happening, and icing severe enough to overcome working pitot heat would normally be found at lower levels (which would of course hamper recovery, assuming other essential structures and powerplants are still working).

And similarily, there's no need to presume a worst-possible-moment, made-for-a-disaster movie script lightning strike that coincidentally knocks out the one component you really need in a CB-avoidance situation...the radar. What can't be seen can't be avoided, and radar still has limitations even while in good working order, especially when it comes to painting the kind of young, developing cell that creates severe-to-extreme updrafts within itself up to thousands of feet above it's visible (by eye or radar) top.

Once again, mitigating the chance of encountering airborne threats is what we do, but the smaller the eye of the needle you thread while picking your way through CBs...especially in a very dynamic steady-state or developing area...the smaller the chance of mitigating it at all. The small or "soft-looking" as-seen-on-radar area you believe is the best route can quickly close up, especially if you're also seeking to avoid being directly over developing cells where the worst turb can be found. And at FL350 and heavy, any pilot should be avoiding that particular spot for that reason, because if a cell is buildng at 10,000 fpm it's a trap for envelopement. Get closed-in, and it's guaranteed your world is going to be rocked, so perhaps the only way to mitigate the threat is to quit worrying about deviating around individual cells and take the 200 or 300-mile deviation around the whole, cursed area.

To me, combined with the pilot's earlier message, the A/P disengaging first with no coinciding source/power failure messages is the biggest red flag of all. What are the possible reasons for it to do so while it's still powered and functioning? Just because it's not controlling the airplane anymore doesn't mean it isn't functioning, or that it only thinks it's being asked to do something it can't due to erroneous Air Data, iced-up sensor input. It's disengament was probably the exact thing it was designed to do under certain conditions, and unfortunately there's the very real possibility that atmospheric conditions outside the aircraft could have been dictating the show into an entenable position from that point on, or perhaps it was even at that point already.

I truly believe many are overcomplicating this. Not every aircraft accident is an overly-complicated series of many technical-issue, swiss-cheese holes. The Laws of Aerodynamics represent one, rather large one when you're mixing it with CBs.

[Blacksheep]

When it did come, it was very good but for some reason others produced not a sausage.

Most likely there was no rain in them. Weather radar can only detect water droplets. If there's no water droplets in a cloud pattern, there's no return. Wx Radar isn't magic, it has to be used actively and the results interpreted, but sometimes you just don't get the data you expect.

On one occasion I was called to an aircraft departing in heavy tropical rain that returned to the stand with a weather radar defect. The pilot complained that his display was completely red. I turned the range up to 130 miles and it showed the presence of a heavy rain cell extending out to about 60 miles from the coast.

After he'd been gone the regulation thirty minutes I was able to go home for a change of clothes.

[clearedtocross]

ACARS timestamps

Does anyone know if the timestamps shown on the list are transmitted (e.g. originated in the aircraft) or added at the receiver? It might make a whole difference to the timeing of the events.
Generally fault lists are time-stamped at the originator. However, in a low bandwith situation like transmission of data on a HF-link, time and date might not be transmitted as they are not crucial for maintenance purposes. Then, serialized and delayed, the string of events in the list might be stretched or even out of order of occurence.

[blackmail]

c.tocross,

4 minutes.
bm.

[incontrol6]

Re: FADEC question. With A/TH disconnect, FADEC will command N1 based on throttle, per pre-defined power curves for flight condition (altitude, TAT, Mach). ADC is prime input for inlet TAT/Mach, but local engine sensors exist as backup.

[Dysag]

clearedtocross

Your question was already answered in post 61: "ACARS maintenance messages are time stamped according to the event and have nothing to do with the transmission time". You must be getting old.