Jumbo

Where did I ever assumed point 1 ) ??
As to 2) and 3)............... When an aircraft like the 737........be it 800 slams ( stalls ) into the ground by ( reasonable ) good weather one can assume point 2) is the correct answer also implying the obvious ........hence point 3) !!
That simple.
I frankly don't give a damn as to why the A/TH retard when they do. I simply disconnect and burry the damn things foward into the dashboard. I'm sure engineering would come up with some very interesting explanation afterward..........til then....I don't want fall off the sky.

Surely this thread has now reached its conclusion?

We have read all the arguments and know many of the facts that led to the accident. Some of us have seen the wreckage (myself included) and it is a sad sight I can tell you.

Why don't we now wait for the final report before starting these discussions again? I can see it starting into a Boeing v Airbus moving throttles v non-moving ones type of argument!

We know that the aircraft was not flown or monitored properly. Let us get the full facts before the blame is apportioned.

- well done! I think that is known as a 'non sequitur'?

I've actually read the whole thread, and what a thread it has been! I believe I posted this info and others have pointed out that the right radio alitimeter feeds FCC-B and that the left radio alitimeter feeds the autothrottle computer (when no failure is present).

The autothrottle computer on the 737NG is programmed to retard even when not in autoland. Conditions are to be:

On G/S
Flaps > 12.5 degrees
27' or less of radio altitude

Agreed and I have stated this before as well. The intent of my comment about failure signals is the reliablity of the radio altimeter transceiver. Rananim asked if the #1 rad alt HAD sent out a fail, would this have occured.

Graybeard said:
Agreed, but I will let you folks debate the action of pulling CB's in flight.

Wizofoz said:
Agreed and maybe this should be SOP when noting erroneous radio altitude below 2500'.

737AvEng
 

Are you looking at wiring diagrams for the Turkish aircraft involved in the crash? Maintenance Manuals, Wiring Diagram Manuals and Schematic Manuals are issued by Boeing against specific aircraft. At best you are guessing. You or I don't know the specifications for that particular aircraft.
If there were any other critical maintenance related failures on that aircraft, I'm sure they would already have been highlighted in the preliminairy report.
While it is unforunate that the rad alt didn't completely fail, there is more than enough data being displayed in that cockpit to give the crew an idea of how the approach is progressing. I just can't get my head around how 3 people in that cockpit couldn't pick up on something to tell them things weren't right...........

The Dutch have started moving the crashed aircraft into a hangar where it will obviously be reconstructed for further investigation:

Dutch remove crashed Turkish airlines plane : Europe World

Time for a plea from the pilot's chair. With over 2000 posts now here, and probably around 100 actually relevant to the crash on THIS forum, can I point out to all those working feverishly with pin allocations, circuit diagrams, system logic flows etc, PLEASE NOTE it is not relevant to this thread. Certainly of interest to some (me included) and no doubt worth posting SOMEWHERE on PPrune - maybe the engineers' forum?

One more time:-

I don't think that I (hopefully a reasonably competent 737 pilot) would have noticed the -8 radalt indication, especially if training an F/O

It is more than likely that the GPWS 'gear' warning (the crew having then confirmed barometric height as 2000 above airfield) would have been dismissed as one of those annoying little glitches that all software produces on many occasions, and thus put into the 'of interest' box. Thus I would not even have given a moment's thought to touching any CBs - even if I had noticed.

I, in what may have been a 'hurried' approach, would not have considered the closed throttles (for around 80 of the reported 100 seconds) to be unexpected.

I would almost certainly have called (or flown) a g/round at 500' radio, and would indeed have been attentively reviewing 'progress' from 1000' down as required by both commonsense and SOPs.

At that point I would hopefully have diverged from what happened. If not, I would hope to have seen (or flown) a 'correct' recovery attempt from the ensuing stall warning.

So, posts of 'why didn't they notice for 100 seconds', 'which pin on the radalt circuit board does what', 'was it Smiths or xxx equipment?', 'why didn't they pull the CB', 'see even with moving throttles.....' etc etc have no place being expanded or repeated HERE. Also until we have some of the CVR details, comments about 'heirarchy', 'ex military' and 'cultural issues' have no place either. We just do not know what was said or done.

This appears to be a major failure of the 'pink' software and that is where the focus should lie for pilots. Some of us find it difficult to believe as well. I'm sure the training fraternity are looking closely at the way we train and the standards we set, and yes, the manufacturer will be looking closely too.

NO
 

"Maybe it's not all that easy to scream " speed " when a turkish captain is screwing the pooch." sorry but don't agree with that...

:sad::E:{:suspect::ok:

Any Danish speakers here to help? I have tried a web translator on the page linked to by transoratia at post 1972 which APPEARS to say that the CVR may be showing distraction in conversation at the time of the stall warning, but the translator is not up to the job. This might just be the first data coming out from the CVR.

Many people writing many things about what they would have done with , and why this electron did this or that. I confess to not reading the last 80 pages in toto, but as this was an -800 there is a speed tape. If they were extending flaps they must have been looking at it to know when to select each flap setting. Why did they stop looking at it. In the latter stages of the approach, at final flap setting, the speed must have decayed into a very large 'amber band'. This is a great visual clue that something is not correct. It then starts to flash and there is an FMC message, long before S.S speed. Where were 3 pairs of eyes looking? So many clues missed. Why the A/T retarded, etc. etc. is not the primary concern of the PF. Why the speed is in the amber band is not a primary concern. IT IS, so do something about it and ask why later.
Am I missing soemthing?

Quite so, Rainboe.

As in the A320 accident off Perpignan, from what one can deduce.

- not in my understanding!

There is nothing new in the article. It's a rehash of the DSB press-release mixed with the usual sensationalism.

Not Danish, but understand most of it. I find no proof of actual CVR findings in the article nor about pilots having a conversation. It basically says the pilots were inattentive (uopmærksomme) during the fateful seconds.
The gist of the article is about Late Logging. It is established that several of the pilots who flew the aircraft on the previous flights noticed the malfunctioning RA but did not write it up.

Where has this been established? So far as I can see, the only clue that the RADALT was faulty came from the FDR, after the event.

OK OK ... sorry. The article quotes the AEI secretary general who says the DSB has spoken with Turkish pilots who, according to the article, have said so. In other words, AEI claims maintenance had not been made aware of the fault.

Guys - the Radalt is definitely contributory but not causal.

The reason I ask about the article is
from Transoratia

Well, I disagree with Transoratia's interpretation, see above. Perhaps he was slightly casual with his choice of words, just as I was about the non-reporting of the previous malfunctions.

Thanks Juud - impressive linguistics as always:) Is that all of the quote - looking to see where transoratia gets that bit from?

NB - I prefer not to comment on AEI..................

Oops again, missed the first paragraph of the following three. Must be the Monday phenomenon :ouch:

"The preliminary investigation by DSB has already established that this altimeter sent an erroneous signal to the autopilot, which steered the aircraft towards the runway, while the two pilots were engaged in a conversation".

"The erroneous signal appeared while the aircraft was still at 700 metres altitude, and said that the aircraft suddenly was 1.3 metres above ground. This caused the autopilot to reduce thrust to idle, even if the aircraft was too high to land".

"During the following precious seconds, as the pilots were inattentive, the speed of the aircraft reduced until it reached stall speed, and fell down. The pilots were able to notice that the situation was critical, and commanded maximum thrust, but too late".


You can make your own conclusions. My conclusion is that there is nothing new regarding the accident flight, which was not known before. The only new information is regarding the previous flights.

Thanks SNF

TRANSORATIA - if you are on frequency - can you tidy this up? Where was that bit? You have put a slightly different interpretation on the article.

Rainboe:

So a minor element of an uncertified configuration of the automation malfunctioned triggering an uncommanded mode the crew were not made aware of - given the TL's were probably at idle anyway in the early part of the descent on G/S, except by a subtle change in mode annunciation that could be easily missed.

I don't excuse the crews apparent inattention to airspeed, thrust setting or mode annunciation during the approach. But neither should we excuse the triggering effect caused by the malfunction of a relatively minor piece of equipment, nor its effect on recovery from the ensuing stall. All as a result of the way Boeing chose to wire the equipment to provide an undocumented and uncertified configuration of the automation. A configuration which crews are not normally trained on and judging by many of the comments from NG qualified pilots on this thread, were until this accident largely unaware of.

I doubt the investigative or regulatory agencies will either.

- MU - you need to clear your mind of the confusion being generated here. We are in danger of another wild goose chase here (no Hudson pun intended). This autothrottle function is neither undocumented nor uncertified and actually NOTHING really to do with autoland. It is a function which autoland indeed uses (but does not require) and is available whenever the A/T is engaged. All crews will be aware of the function, some may not be aware that RA1 provides it. I come back to the basic premise that we have thrashed here, the lack of 'awareness' of the a/c speed. The malfunctioning A/T would have had no effect on the stall recovery had that been flown according to the manufacturer's requirements. PLEASE FORGET AUTOLAND here and ANYONE who mentions it (except me...:).)

I'm not quite sure what was uncertified here. The crew was flying an approach with an autopilot and the autothrottle engaged. If things had gone properly, I am sure that the pilot would have disconnected the autopilot at some point and finished the landing. That would be quite normal and certainly legal.
I am not sure what some of the people here are talking about with a single autopilot autoland. On the Classic B747, the rad alt would trip the autopilot off at 150ft. RA. On the -400s, the autopilot can fly the aircraft down to 5ft. RA before it trips off. If left alone, I am quite sure the aircraft would fly itself into the ground but surely you can't call this autoland.
It also has to be remembered that the logic on the aircraft has to cover many other possible conditions, not just the setup they had that day. Personally, I think if anyone was to study the systems thoroughly, they would be quite satisfied with the design in place.

engaged in conversation
 

I recall a plane crash in which the cvr showed the pilots were discussing swimming pools in the backyards of the houses below...they stalled this particular transport jet...it didn't have autothrottles.

so, let's drop the autothrottles and at least consider inattention to basic flying skills as the cause of the crash???????

I was a little hesitant a few days ago saying that they would look at what the previous crews had entered in the tech log.
Apparently it was not written in the tech log.
Still the basic airman ship and allowing the airspeed to go 40 knots below VREF is
the root problem here.
This R/A was a simple failure, monitoring the airspeed was a major malfunction.

The article can be read by anyone. I have no interest to make something up SnowFalcon2.
In the article, there is a lot written that has already been hashed and rehashed here. I am too lazy to repeat all that.
I have just picked out of the article what I believed was new information (and I have read all 2000 posts).

The sentence "Det er allerede fastslået ved den hollandske havarikommisions indledende undersøgelser, at denne højdemåler sendte et forkert signal til autopiloten, som selv styrede flyet ned mod landingsbanen, mens de to piloter var optaget af en samtale." translates to:
"It has already been established by the Dutch accident commission's (DSB) initial investigations that this altimeter has sent a wrong signal to the autopilot (!) which steered the plane down towards the runway while the pilots were busy with a conversation."
I take that to be the interpretation of the journalist and not necessarily the exact situation on the flight deck.
However, I take exception to the statement that I am making things up or translate incorrectly.
Link: Groft sjusk med fly-logbøger var årsag til flystyrt i Amsterdam | Ingeniøren

I'd be interested to know if a timeline has been communicated when more official information will be released. Anyone?

There was no non-reporting. Is anyone suggesting that THY crew were interviewed after the event and said 'Oh yeah, we remember the RADALT glitching but we were in a hurry to get home and didn't write it up'. Of course not. RADALTs, by their very nature, do un-lock now and again.

The fault on previous flights (and this) was only seen after the FDR was analysed.

If the quoted part re. conversation in the article is true, why is this comment not in the official initial press release document? As it is not there, where did this journo pick this info up?

Thanks to all the Danish translators - I think we can put this aspect to bed now. (On the other hand, a few could disagree and take this on another hundred or so posts) Reading between the lines it would appear that round about the time the false signal first arrived from RA1 the crew were 'chatting' about something - possibly even that - who knows? It does not appear to suggest they were similarly 'chatting' as it approached the stall.

Thats the way it usually happens, don't write the tech log item, it will ground the aircraft etc.
When in reality it always bites the next crew on the back side..
Turkish airlines has no excuse for this, they have backup aircraft not like some charter ones there that are banned in the EU and have to fly or go broke.
But its still not correct no matter how you look at it.
This airplane had a simple fault, something caused these pilots to stall the airplane, and it was not a simple R/A failure.

MU300A (#2009)

Let me write it once more, hopefully for the last time.

The RETARD function serves a purpose even when not on a Dual Channel approach:

The crew is on a Single Channel approach at some stage PF disengages the A/P but not the A/T. Instead PF deselects the speed mode with the A/T engaged.

Because Boeing doesn't want Min Speed Reversion to pop up during the flare and ruin the landing, the A/T are designed to RETARD at 24/27 feet (chose the right one)

Any questions?

The nose pitches further and further up as the trimmer clacks away....the airspeed drops to Vref -40 and not one of three pilots notices something strange?

What was happening in the cockpit that three pilots did not notice this?

It must have been something pretty damn engrossing.

No matter what happened with the RA and the A/T and all that happened before, bottom line is that these boys were distracted very badly by something at a critical time.

What was it?

Well done ant. Said it like it is. It does serve a purpose. Proper operation of the system depends on proper piloting. You cannot have such a system that will work for ANY poor pilot. It's amazing there are people who cannot understand the system, yet feel they are knowledgeable enough to castigate the designer!

Nicholas49- way out. Nonsense. Nobody has suggested it, you have absolutely no evidence for mad airspeed theories!

JG1, thank you, take a ticket and stand in the corner- you are the 47th. person to post that and the 547th not to have read the thread history!

We really have reached the limit of what can be discussed based on what evidence has been released. It's a zoo, with every shade of opinion and experience (and inexperience) posting the same queries and remarks, and EVERYBODY knows how to design an autoflight system better than Boeing! I'm impressed (not).

ant1

Thanks for the explanation.

A question.

How many pilots would normally be relying on the RETARD function to close the throttles on this type of approach? In other words, in 99.9% of approaches of the type flown by the accident crew, wouldn't the throttles already be closed by the time the RETARD function initiates? Just not at 2000' as in this case, but over the runway threshold?

You're not following the thread. You would disconnect a single channel approach when visual hundreds of feet up- it would not have happened. You then either turn OFF the A/T (and lose automatic G/A thrust capability) or put A/T to standby ('Speed Off') so if you hit the TOGA switches, the A/T will reactivate and give you G/A thrust! Depending on your airline SOP. If you are not visual by Cat1 limit, you would GA. Well above 27'!

It's flogged to death now!

I don't use the speed off method but if someone does then he either:

- Disengages the A/T and retard manually or
- Lets the A/T retard by themselves (maybe with a little help if needed)

So if you go speed off till touchdown then the A/T have to retard otherwise even if the PF retarded them they might go back forward.

And regarding the RA, I've already said I'd be happier with a RA comparator that could disable the retard feature or even disengage the A/T altogether, specially if the single RA dependence is proved to be a factor in this one.