Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF 447 Search to resume (part2)

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

AF 447 Search to resume (part2)

Old 11th May 2011, 02:57
  #1121 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Hi,
Originally Posted by Oldengmkr
Perhaps AF 447 experienced a similar AoA law event as discussed in these excerpts from the cited AAIB and it precipitated a deep stall.
The use of AoA protection in direct Law, basically is to prevent a "low speed stall" to develop; what makes you think then, that by preventing an aircraft to reach a so called alpha Max would precipitate a "deep stall"?

A "deep stall" is a result of the loss of any elevator control due to the turbulent airflow generated by the wings. Not sure that this symptom could be applied to something else than a T-tail airframe.
But, more importantly, what you do understand from what was at stake in this report of A330-A340 Airprox event seems very far from AF447 case. The main issue was about the proximity of those two aircraft in turbulent weather. TCAS issues would be related with this 1st June 2009 catastrophe?

Originally Posted by Oldengmkr
As best as I can determine, this AoA law is still in use today. If a similar AoA law incident occurred to AF 447 it’s severity may have been compounded by the reduced crew on the flight deck at the time.
What was this AoA law incident reported?
There was no "reduced crew on the flight deck" either that we are aware of! Pilots were supposed to be at their post as per SOP.

S~
Olivier
takata is offline  
Old 11th May 2011, 03:10
  #1122 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 77
Posts: 1,458
Turbine D
In my opinion, big turbofan engines today are no more or no less "finicky" than their modern day fighter engine counterparts. I should add one other comment: Most of this technology evolved from the military J79 engine, the difference being the J79 produced 17,000 pounds of thrust and the F110-129 produces 30,000 pounds of thrust!
Thanks for that Turbine D. Can that be interpreted that you consider it improbable that AF447's engines would become stalled, even at high Angle of Attack? I know that engine technology has come a long way since Davies described an aircraft flaming out both engines while locked in a deep stall.
When you see what the F-22 can do, I am totally amazed how the engines handle major power changes in all sorts of outlandish attitudes and AOA states. The aircraft types I once flew would be wreckage in short order if they attempted those maneuvers. Right Gums?
Machinbird is offline  
Old 11th May 2011, 03:52
  #1123 (permalink)  
Thread Starter
 
Join Date: Jun 2009
Location: florida
Age: 77
Posts: 1,128
Deep stalls and T-tails

Salute!

I take issue about the definition used by takata:

A "deep stall" is a result of the loss of any elevator control due to the turbulent airflow generated by the wings. Not sure that this symptom could be applied to something else than a T-tail airframe.
As we have seen in my first post regarding this:

http://www.pprune.org/6432295-post772.html

Some "deep stalls" have nothing to do with turbulent airflow over the horizontal tail control surfaces. I even included a pitch moment chart to show how this happens in the Viper. I also flew the VooDoo, which did have a T-tail. We didn't have "turbulent" flow, we had the vortex from the main wing strike our high tail and the increased AoA "pitched" the nose up and then we tumbled!! Because we had normal static stability, it was not difficult to quickly recover using the drag chute, but we also were not in a deep stall following the violent pitch-up. We had an AoA protect feature that shook the stick and then applied 28 pounds of forward stick if you kept pulling. However, you could "beat the system" with pitch rate at fairly slow speeds.

It could be the c.g. I tellya!!! If that plane had a c.g. further back than it normally used, the thing could easily settle into a fairly stable deep stall if its pitch attitude was steep and then it ran outta air molecules for the tail to work with - just like the Viper. Further, as with the Viper, the reversionary control laws allow lots more forward tail movement command using the control stick than aft stick commands.

The data plots are gonna be real interesting, and I am willing to bet folks will blame the crew.

respectfully,

Gums sends...
gums is offline  
Old 11th May 2011, 04:02
  #1124 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Originally Posted by Machinbird
Thanks for that Turbine D. Can that be interpreted that you consider it improbable that AF447's engines would become stalled, even at high Angle of Attack? I know that engine technology has come a long way since Davies described an aircraft flaming out both engines while locked in a deep stall.
If Turbine D say so then! (about military jets).
This GE J79 was introduced in 1955. But since, many events are still recorded during standard operation in tropical atmosphere. I'm not even talking about an "abonormal attitude" situation like AF447 was supposed to be in such a "free falling like a leaf thru the storm" scenario.

I don't think that you really bothered to read this 2007 (short) paper I already posted (and there is much more on this subject, proof that this issue is taken very seriously and for a while). Now, this paper is only about "ice crystal" above 22,000 feet related events and their statistics are far from being complete:

Since 1990
, there have been at least 100 jet engine power-loss events on both commuter and large transport airplanes, mostly at altitudes higher than 22,000 feet, the highest altitude where airframe icing is expected to exist.

"Power loss" is defined as engine instability such as a surge, stall, flame*out, or rollback that results in a sub-idle operating condition. High-altitude ice crystals are believed to have caused most or all of these events.

S~
Olivier
takata is offline  
Old 11th May 2011, 04:38
  #1125 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Originally Posted by gums
It could be the c.g. I tellya!!! If that plane had a c.g. further back than it normally used, the thing could easily settle into a fairly stable deep stall if its pitch attitude was steep and then it ran outta air molecules for the tail to work with - just like the Viper. Further, as with the Viper, the reversionary control laws allow lots more forward tail movement command using the control stick than aft stick commands.
I'm not going to be nit-picking about such a definition because the full loss of elevator control, in any case - not only cg related, could result at the same situation.
Nonetheless, I'm sticking with the impact description provided by the BEA and I can't see why such a description would imply a mandatory "deep stalled" attitude ; the vertical vector may also be due to inertia. There is no description of an excessive pitch rate at sea level (alpha Max should be well above 35° if she had some thrust remaining). Without much more details provided, it may look also like a CFIT attitude to me.

Originally Posted by gums
The data plots are gonna be real interesting, and I am willing to bet folks will blame the crew.
It would certainly be easy (too easy?) to find something that they did which would have been a contributory factor to the end result. As a matter of fact, we still have no idea today to what they were really (exactly) confronted.

S~
Olivier
takata is offline  
Old 11th May 2011, 05:12
  #1126 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 77
Posts: 1,458
Takata
I'm sticking with the impact description provided by the BEA and I can't see why such a description would imply a mandatory "deep stalled" attitude ; the vertical vector may also be due to inertia.
Ok, So they were gliding down with inadequate power to fly?. Is that what you suspect? If so what kind of airspeed would they be indicating? Approximately what flight path angle relative to the surface.? (Or what would you do in the circumstances?)
Probably due to engine core icing I assume. But the engines didn't "phone home"? Wouldn't they be expected to?

And I too accept the BEA impact description-as far as it goes.
Machinbird is offline  
Old 11th May 2011, 05:19
  #1127 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
auv-ee, sliced and diced sound does help. I'd sincerely hope a submarine or towed search vehicle would not use human ears* (which can dig signals out of noise if well trained). A simple 100Hz to 200Hz FFT, of the sort RR NDB showed is about as good as can be done. And it's probably close to what was done. If there is very much chirp on the signal it might not hit the filters as well. And if they were partially buried their 1700 meter range might get as low as a few hundred meters. 6dB loss would cut range to 50%, and put the towed pinger detector just barely at the edge of its range. 6dB loss is not much depending on how the pinger couples sound to the water.

Note that pingers like watches use are resonant chunks of piezoelectric material that are excited directly as part of the oscillator circuit. Burial in mud might do bad things to the waveform and frequency in such a modulated oscillator

* In such a case as this with a 37kHz or so ping, well outside human hearing range normally speaking, a simple frequency converter circuit could be used just as is used in typical radios. One of the radios I have works down to about 1kHz to 16kHz depending on the filter in use. (I, er, um, doctored my R-390A just a wee bit. {^_-}) That could be coupled into the ultrasonics from the sonar detector and convert them to human hearing range very nicely. In such a case the human ears might have detected something the FFT techniques would have trouble with. That's a long-shot "might", mind you. (Note that human ears actually work on a "bazillions of resonant circuits" basis. This leads to some unfortunate hearing pathologies as well as fascinating abilities to sort out sounds from noise. Two ears also helps.)

{^_^}
JD-EE is offline  
Old 11th May 2011, 05:31
  #1128 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
Originally Posted by AlphaZuluRomeo
Quote:
Originally Posted by JD-EE
The submarine had no chance of reception unless it's crush depth is WAY below that of anything I could find on the Internet a couple years ago.

Hum, does this statement also apply to submarines equipped with towed array sonars (which can be towed lower than the ship's depth)?
It would apply if the tow cannot get under about 2500 to 3000 meters. That presumes a slant range of 1700 meters to the sensor and a search path with some overlap. It'd need more passes at 2500' than at 3000' depths. The deeper it goes the less overlap is needed.

I'm idly wondering why a submarine would have a towed head that could go down that deep. I can't think of a military usage. (Of course, in this field my imagination is limited. {^_-})

{^_^}
JD-EE is offline  
Old 11th May 2011, 06:10
  #1129 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
GY, that's 160.5 dB relative to what at what distance?

If it is 160.5 dB relative to hearing threshold at a distance of 0.17 meter then at 1700 meters it would be suffering square law attenuation of 10000^2 to 1 or 80dB. Then you have water losses.

All I have for solid information is that they seem to be described as detectable to 1700 meters distance. Thus my question about the submarine detecting it. What is the ambient noise level at around 36 kHz to 39 kHz? What is the ambient noise spectral characteristic in that range? (Gaussian, lots of tones,, chirps, and whatnot.) 160.5 dB relative to what? ( 10 dB is a 10 : 1 power ratio. It is a ratio so it has no units of its own. And I suspect in this case it is not relative to a jet engine at 1 meter. {^_-} I also suspect it is relative to 1 micro Pascal sound pressure level. (Yes, relative to 1 micro Pascal at 1 meter.) Nor do I know the receiver sensitivity. (Whales can hear very weak signals in water, far weaker than human ears can hear in air.)

I do understand, if I am correct, that military sonars run at power levels that can cause cavitation in the water. I understand 10kW is not out of the question for power levels.

And I understand that at 1.5 kHz the MDS is about 30 dB to 50dB relative to 1 micro Pascal. I've no idea what transducers can do these days with regards to efficiency changing the pressure waves in water into electricity.

I do suspect the 1700 meters is awfully conservative. There'd only be a 65dB loss in that distance for distance squared loss. Add another 7dB for absorption and you get 72 dB loss or about 88dB for signal level at 1700 meters. I thought that had some from one of the BEA papers. But, I am getting old and soft on the memory.

Something just came to mind. I bet water is a somewhat "dispersive" medium that would tend to smear the ping's power out over time. I'm not sure of the magnitude of this effect, though.

{^_^} (Please excuse the stream of conscious here. I was doing a lot of "looking up" and learning while writing this.)
JD-EE is offline  
Old 11th May 2011, 06:21
  #1130 (permalink)  
 
Join Date: Oct 2009
Location: berlin
Posts: 145
shadoku : with ~50 supposed unbuckled recovered at sea, there was perhaps 100 of them like that in the plane. With a sudden pitch, that makes about 6 tons moving backward (nose up) or forward (nose down) in the plane. Doesn't make a big unbalance?
CVR ( bordtime LKP+100s ) "this is captain calling: PAX please open the seatbelts and wake to the front as far as you could, we are in a stall recovery, we need your whight in the front, we will explain the details later....."
grity is online now  
Old 11th May 2011, 06:31
  #1131 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
EDLB, non-complex crystal oscillators are in the +/- 50 to +/- 100 parts per million tolerance range, typically. They tend to be fragile for shocks. And they chirp significantly as the oscillator circuit turns on. (It would have to power cycle to save energy.)

I am not sure what the medium would do to long pings with multipath and otehr effects.

ACLS65, I have long ruminated on the concept of ping back when pinged. That, however, requires a complex receiver to avoid false triggering. That means more things to fail even if it's square mm of silicon surface. And it continuously sucks power without producing a useful output.

An RC oscillator with an RC timer for both the 1 second period and the 10ms transmit period power levels can be kept very low with most of the power used being devoted to providing an output at all times with very few parts to break or die of old age. I have not discarded the transponder concept in my mind, yet. But it doesn't seem to be as big a win as one might want or expect at first glance.

{^_^}
JD-EE is offline  
Old 11th May 2011, 06:39
  #1132 (permalink)  
 
Join Date: May 2011
Location: london
Age: 37
Posts: 6
Hi,

Thanks very much for this comment. It help me to think about my ideals.

Tks again and pls keep posting.
If you want to get more materials that related to this topic, you can visit: Resume samples 2011

Best regards.
hambim336 is offline  
Old 11th May 2011, 06:48
  #1133 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
Originally Posted by RR NDB
And the 2nd Solution at page 10:

Installation of low frequency ULB (between 8.5 and 9.5 kHz) attached to the plane

"Made my day"
All that for only doubling the range? The higher loss at 37.5kHz compared to the loss at 8.5-9.5 kHz is about what you get with slightly more than doubling your distance from the transducer.

A slightly different strategy might prove to be a winner. If you know the frequency within 1 Hz and it is transmitted continuously you can search with sub 1Hz FFT windows. That can purchase a good factor of 10 in detection range.

(Mumble mode) Continuous transmission at 1/100th power means you need a 1Hz bandwidth to get equal detectability for "very roughly" the same battery life. By then Dopper becomes a problem with a moving receiver. Oops! It seems there may not be a good strategy for improving detectability. But watching it on a waterfall spectrum display would allow one to eyeball finding it very nicely with 1Hz or slightly smaller windows.

{^_^}

{o.o}
JD-EE is offline  
Old 11th May 2011, 06:53
  #1134 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
Dave Senior - might "The sentence would then translate as "Bloated and damaged by the bite of an unknown shark"." be better rendered as "Bloated and damaged by the bite of a shark(?)." The intent being "damaged by an unknown critter in a manner that looks a bit like shark bite ."

{o.o}
JD-EE is offline  
Old 11th May 2011, 08:11
  #1135 (permalink)  
 
Join Date: Jul 2007
Location: My Stringy Brane
Posts: 379
I have updated the image to reflect recently corrected data:



I have extended the radial south from LKP for reference -- only the DFDR can (hopefully) reveal AF447's true path.

Interestingly, the impact point lies within a triangle drawn between the LKP, pollution spot, and first recovered bodies.

Last edited by Machaca; 11th May 2011 at 08:29.
Machaca is offline  
Old 11th May 2011, 08:23
  #1136 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 77
Posts: 1,458
Quote:
Originally Posted by Machinbird
There the aircraft pitches up into a high speed stall, freezing the RTL unit at the pre-existing correct value, since the high AOA triggers ADR disagree results similar to the Perpignan A-320.

Takata
I'm really unable to follow you here: any "high speed stall" won't freeze the RTLU, only wrong air data will.
Also, if the aircraft would wrongly pitch up, the end risk would be a low speed stall if both wrong airspeed retained were in reality lower than the system would think... not a high speed one.
Next, if your pitots are really clogged for inducing the system in error, you'll need to clean them again before they would record some wrong values due to any "stalled/spin" attitude (and now trigger unreliable airspeed, freezing RTLU at Mach .80, as mm43 quoted above).
Moreover, Perpignan is a really bad case of reference as this whole sequence was way too short as to even close a correlation window and to valid or reject any airspeed discrepancies [meaning, you can have many intermittent UAS alerts popping up in the cockpit without having a real/solid UAS event being triggered at the maintenance computer level]
Hi Takata, sorry if I went over the concept too quickly.
A dynamic loss of control can occur due to rate limiting in the control system. This can be recognized from the actual control position traces by control surfaces moving in one direction and then immediately reversing direction before achieving the initial commanded position. The resultant position trace does not have a normal rounded "top", but instead forms a tri-angular top (or bottom). This can be caused by a number of possibilities however hydraulic demand rate exceeding supply is one potential cause. Occasional "peaking" of the trace does not prove rate limiting and rate limiting need not cause loss of control. If only small control motions are required to maintain control (low gain behavior), then generally nothing happens. But if large control motions are being commanded, and if a phase lag results that is large enough, then the resulting control motion can actually drive increasing oscillation. That is the theory in layman's terms. The actual theory uses transfer functions and complex mathematics and involves math that I haven't used in almost 50 years. I have posted links to papers on related subjects in the not too distant past and can again if requested.
What a dynamic loss of control means is that a beautifully flying aircraft can put on its Mr.Hyde face and act very nastily essentially with little warning.
Flight conditions can be a trigger for these oscillations. With hydraulic rate limiting, a power reduction could be a triggering event. If the aircraft begins pitching, the flight control laws system cannot react quickly enough to damp or limit the motion of the aircraft. The hydraulic system is not defective and likely would not set a warning (depending on its design). But the aircraft could tromp all over its normal limits. If the aircraft then stalled, it would not be surprising to lose the 3 ADR channels.
From page 86 of BEA accident report on A-320, D-AXLA
The Captain controlled a left roll movement, caused by the stall. The aeroplane’s high angle of attack and the roll movements generated asymmetry, and a speed variation between ADR 1 and 2 appeared. This increasing divergence caused a rejection of the three ADRs by the FAC then the ELAC
If this can happen in an A320, why cannot the same thing occur in the A330?
As soon as the 3 ADRs are rejected, the RTL will be left at last value. This could likely occur in just a few seconds after the beginning of loss of control. Note: no freezing of pitot tubes is required to generate this situation!

BTW, I am not trying to discredit Airbus. I have flown on A330 aircraft several times as SLF since the loss of AF447, and am not particularly concerned. The situation to create this type problem would not be common.
At this point in my life, I probably am a bit of an aviation dinosaur, but I am not opposed to the new technology, nor do I wish for the "good old days" of aviation. I merely think that Airbus may have been a bit presumptuous when they didn't give the pilots the ready ability to outvote the computers.
Machinbird is offline  
Old 11th May 2011, 08:58
  #1137 (permalink)  
 
Join Date: Aug 2005
Location: EDLB
Posts: 173
There are plenty ways to PLL a RC oscillator with a Xtal reference even with low power. The timing can be done with a relativ low frequency low power device like in a GSM phone. It is also not a problem to make it failsafe, when the PLL fails for whatever reason it still transmitts with the RC frequency. The PLL loop can be made very slow and digital, which means that the XTAL oscillator is only needed for a small part of time. And if the xtal fails a relativ robust silicon RC oscillator can be used instead. With todays technology it is not very difficult, its more a question how much engineering you are willing to pay for.
EDLB is offline  
Old 11th May 2011, 09:24
  #1138 (permalink)  
 
Join Date: Jul 2009
Location: UK
Posts: 133
Originally Posted by takata View Post
...
But, more importantly, what you do understand from what was at stake in this report of A330-A340 Airprox event seems very far from AF447 case. The main issue was about the proximity of those two aircraft in turbulent weather. TCAS issues would be related with this 1st June 2009 catastrophe?


What was this AoA law incident reported?
...
It is strange to me that the violent manoeuvre by the A340 is assumed possible but very rare so the main recommendation was to not closely stack planes vertically just in case a lower one decides to climb without permission.

possible similarities are:-

Event triggered by encounter with turbulence affecting indicated speed causing a control system mode change from autopilot.

Manual action was required by pilot to recover aircraft - within 30s the aircraft was too high and unable to maintain speed even with (4 in this case) engines at full thrust.

Stretching a point maybe but if pitots were frozen on the way up - you have the classic simultaneous overspeed and stall warnings.
sensor_validation is offline  
Old 11th May 2011, 09:58
  #1139 (permalink)  
 
Join Date: Dec 2007
Location: London Under EGLL(LHR) 27R ILS
Age: 26
Posts: 490
Black boxes in Paris tommorow morning?

Reference: Aviation Today


The flight recorders from Air France Flight 447 will be transported to the Bureau d’Enquetes et d’Analyses (BEA) by plane on May 12, the agency announced May 10. The transport will take place after the La Capricieuse, a French Navy patrol boat, reaches the port of Cayenne with the flight recorders. The boat is scheduled to reach the port May 11.

The transport comes in wake of finding both the FDR which was located on May 1 and the Cockpit Voice Recorder (CVR) located May 3, helping investigators determine the events that caused the June 1, 2009, crash of the Airbus A330 that killed 228 people. Flight 447 was en route from Rio de Janerio, Brazil to Paris-Charles de Gaulle airport when it disappeared from radar and plunged into the Atlantic Ocean.

La Capricieuse reached the search vessel Ile de Sein on May 7. Recovery of airplane parts, including one engine and computers from the avionics bay, has continued over the past few days, according to a May 9 BEA briefing.

Jean-Paul Troadec, BEA director, Alain Bouillard, Investigator-In-Charge, Christophe Menez, head of the engineering department, and Philippe Vinogradoff, special representative to the AF Flight 447 families of the victims will be present at a May 12 press briefing. Held on the BEA's premises at Le Bourget, the briefing will discuss FDR findings.
HeathrowAirport is offline  
Old 11th May 2011, 10:00
  #1140 (permalink)  
 
Join Date: Jun 2009
Location: madrid
Posts: 1
Those flight paths posted by senor_validation are incredible!

Flight Paths of Flight AF 447 and of the flights that crossed the zone around the same time

I am an avid follower of all your posts, and in awe of the knowledge and expertise you all have. I could not be more removed from the airline industry (consultant to the banking industry) and thank you for all of your time and diligence on uncovering the reasons for this tragedy.

Thank you!
NatBat 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.