Lion Air Crash

[Josip Vrandecic -Mes]

Recent expert opinions :

**LINK**

Edited By Josip Vrandecic -Mes on 18/11/2018 09:59:56

[Jimbo565]

Jo

Link is not working for me !!

JIm

[Paul C.]

Working ok for me, but I am in Spain on holiday if that makes any difference

Paul.

[Stearman65]

Posted by Jim Carss on 18/11/2018 11:03:16:

Jo

Link is not working for me !!

JIm

Works for me!

[Jimbo565]

Problem at my end,using explorer,had to right click and open ok.

[Colin Leighfield]

It seems appalling in this day and age that this problem wasn’t foreseen. With all of the computer simulated testing it’s hard to believe that this predictable scenario wasn’t identified. I’m a great fan of Boeing but if this is correct, and it seems to be, they really need to be flogged over this one.

[ken anderson.]

worked OK for me Jo, poor people losing their lives through something like that is really unforgivable....

ken anderson...ne...1..... what a shame dept.

[Paul Marsh]

Sounds like a repeat of Quantas Flight 72, luckily, the aircraft managed to land ok, similar problem. incorrect information going into the flight computer.

Problem there is millions of lines of code, and many conditions where they get tested before the aircraft enters service, but a section of code can cause loss of an aircraft and lives to something so small.

[Josip Vrandecic -Mes]

@ HI Ken and Jim , Ken please to help our generous friend Jim(I'm fond of him ) about the link ,and thanks so much.

@ Hello dear Colin, I'm afraid this is repeated at Boeing....

[Colin Leighfield]

Very sad Jo.

[Jon H]

I think it might be premature to dump all of the blame on boeing. We do not know what training these specific pilots had and if it was sufficient. If it wasnt and it didnt include this information then was that boeing's omission or the airline?

We also do not know if there was another problem, or if in the heat of the moment they were simply unable to diagnose the fault and then think about the proper corrective action as they would have had to deal with the sizeable stream of error codes from the flight computers, and then consult the manual for each eventuality, all while trying to fly the plane and i doubt that is particularly easy.

[Piers Bowlan]

As Jon says, premature! Boeing have so far answered criticism that they did not supply pilots with information regarding the stall protection system fitted on the B737 Max. Boeing say that the failure that the Lion Air aircraft experienced was covered by a check list item, which was also a required 'pilot memory item'. In which case they should have know what to do when faced with an aircraft that was not responding correctly to their control inputs. The fact that the aircraft had a history of similar faults in the days leading up to the accident might also suggest that the company in question do not have a robust maintenance regime in place? As the pilots would have read the Tech Log entries prior to their fateful flight they should have briefed each other for possible air data failure scenarios prior to take-off. It is easy for me to say that but it would have been good airmanship at any rate. I am not criticising Boeing, the pilots or the companies maintenance procedures but pointing out that when accidents occur it is usually the result of a whole series of events. Hopefully when the full FAA report is published the results will be disseminated throughout the industry. I hate the expression 'and lessons learned' but here it seems appropriate!

[Josip Vrandecic -Mes]

Posted by Jon - Laser Engines on 19/11/2018 08:43:47:

''I think it might be premature to dump all of the blame on boeing. We do not know what training these specific pilots had and if it was sufficient. If it wasnt and it didnt include this information then was that boeing's omission or the airline?

We also do not know if there was another problem, or if in the heat of the moment they were simply unable to diagnose the fault and then think about the proper corrective action as they would have had to deal with the sizeable stream of error codes from the flight computers, and then consult the manual for each eventuality, all while trying to fly the plane and i doubt that is particularly easy.''

Hi Jon , absolutely not easy to fly, especially on new passenger planes, it might be a bit,call in a negative context to name of this company, but it has already proven that new aviation products went to exploitation without the expert's human attention and tracking by the manufacturer (profit) ... And I I'm a fan of Boeing, and I flew to 727, which I still consider to be the best thing that happened to me ... but I forgot I was still with my head... in the analogue era.

Regards

[Martyn K]

Force Nose down anti stall has been around for years. The reason for the Trident crash at Staines in 1972 was because the pilot over-rode the anti stall pitch down after retracting the LE slats too soon.

Martyn

[Colin Leighfield]

Martyn, that’s right but the Trident didn’t go in nose-down. It stalled level nose high and it still was flat when it hit the deck. A number of passengers were still alive when rescuers arrived but died very shortly afterwards. This whole thing with tail blanketed nose up stall really kicked off with the BAC111prototype crash in 963 that killed Mike Lithgow and others. In stall tests at different angles of attack it got into a tail blanketing position in deep stall and the elevators were ineffective in correcting it. It stalled flat all the way to the deck and the Trident did the same thing. After the 111 crash stick-shakers were fitted to warn the pilot before the critical angle of attack was reached. The Trident would have had that system but because the pilot had retracted the leading edge flaps at too low an air speed the whole sequence probably occurred too quickly before he realised what he had done and he was at low altitude anyway.

The 737 crash is totally different. Boeing installed this new nose down corrective system that is completely new to anything fitted previously but it is being claimed that they didn’t inform adequately, so experienced 737 pilots were taken by surprise when it functioned and didn’t know the corrective information. It sounds like more than conjecture and it appears that Boeing isn’t denying it either. We will know eventually but it sounds like something that wasn’t caused by defective systems or pilot error in the usual sense.

[Jon H]

Some useful analysis here. gives a good overall explanation of the factors involved.

[Nigel R]

"Problem there is millions of lines of code, and many conditions where they get tested before the aircraft enters service, but a section of code can cause loss of an aircraft and lives to something so small."

Are you familiar with 178B/C DAL A software certification?

Edited By Nigel R on 19/11/2018 14:01:25

[Colin Leighfield]

Thank you Jon, excellent. We have to wait and see now. Reading back I might owe Martyn an apology for misreading what he was saying. I thought at first he was implying that the Trident went in nose-down, but I see now that he wasn’t. He was spot on in saying that the pilot over-riding the nose down pitch correction was part of the sequence of events beginning when he retracted the leading edge flaps too early. I jumped in too quickly.

[Geoff S]

Thanks, Jon. Those Mentour videos are always interesting (even the dog, which sometimes features) and gives non-patronising information accessible even to those unfamiliar with flying. It just shows how complicated flying big aircraft is.

I just hope they get to the bottom of this quickly and make effective steps to stop anything like this happening again.

 

As for the millions of lines of code. I've reviewed code at work for software that's not totally flight critical but affects engine testing and it really is the most boring job I've ever done. I loved writing software using both low (assembler) and relatively high ('C' ) level languages but going through someone else's code is a different matter all together.

However, even an apparently trivial change done incorrectly can have completely apparently random effects, so checks are necessary. Mind you the Appolo programs were probably full of bugs and they worked OK (I don't think the Appolo 13 problem was s/w related).

Geoff

Edited By Geoff Sleath on 19/11/2018 16:11:36

[Jon H]

I think the biggest thing to remember about these accidents is that the chain of events needed for the accident to occur is considerably longer than most members of the public, and especially the media, actually realise.

When i was at uni i went to lecture which detailed the full investigation into the causes of the Concorde crash in France. It must have taken nearly 2 hours to get through the various contributory factors that finally lead to the crash. It was nowhere near as simple as just running over a piece of metal and bursting the tyre. The shape of the metal, the tyre design, the speed, fuel load at that time, the point of impact...and that was just to make it leak. There was then a huge list of events that caused to it actually catch fire.

If you could go back in time and present the accident report before the crash, i doubt anyone outside aviation would have believed it was even possible.

[Piers Bowlan]

Posted by Jon - Laser Engines on 19/11/2018 16:44:24:

I think the biggest thing to remember about these accidents is that the chain of events needed for the accident to occur is considerably longer than most members of the public, and especially the media, actually realise.

When i was at uni i went to lecture which detailed the full investigation into the causes of the Concorde crash in France. It must have taken nearly 2 hours to get through the various contributory factors that finally lead to the crash. It was nowhere near as simple as just running over a piece of metal and bursting the tyre. The shape of the metal, the tyre design, the speed, fuel load at that time, the point of impact...and that was just to make it leak. There was then a huge list of events that caused to it actually catch fire.

If you could go back in time and present the accident report before the crash, i doubt anyone outside aviation would have believed it was even possible.

I am glad you highlighted the Air France Concorde crash Jon. I watched a documentary about this and the reporter could only get his head around one simple fact throughout the entire one hour program; they ran over a piece of metal! A missed opportunity to cover many of the very complex factors covered in the accident report. I watched a similar documentary about Air France 447 and this didn't even scratch the surface regarding the chain of events leading to the disaster.

For those that are interested, this is a video regarding the B737NG Max QRH (Quick Reference Handbook) memory items. If you fast forward to 7 minutes it covers 'Stab Trim Runaway'. According to the current theory, this is what the crew had to deal with, with the aircraft becoming progressively more nose heavy and out of trim as a consequence of the CMAS system getting corrupted information from the faulty A of A vane. It is debatable whether the crew having knowledge of CMAS would have been of benefit in solving the problem. In any event, they would have had information overload with numerous EICAS messages and warnings not to mention the stall warning horn blaring repeatedly. Very distracting to say the least.

What is surprising is that, according to the theory, a single failure of one A of A vane (there are two) could cause the CMAS to be activated, erroneously causing the stab trim to trim forward repeatedly (nose down) with disastrous consequences. There is probably a lot more to this than has been suggested so far. Desperately sad.

[Nigel R]

A single point of failure causing this behaviour would be surprising on this airplane. A sensor disagree type situation should be specified to inhibit behaviour depending on valid sensor information.

I'd await the AAIB findings. They're really quite good at this sort of thing.

[Colin Leighfield]

One of the surprising things about the Concorde disaster was the lack of attention paid to the fact that the reason the debris was thrown up (apart from the fact that it shouldn’t have been there, although it was at the edge of the runway, not in a position that would normally allow it to be run into), was that the plane was tracking to one side instead of running down the centre line. It was fresh out of service and this was the first flight afterwards. Later it was found that in reassembling a wheel bogie a spacer had been left out of one of the wheel hub axles. As a result a wheel was dragging and causing the Concorde to track to one side. If it had been on track the debris wouldn’t have been caught and thrown up. If I remember what I read correctly this maintenance error also meant that the pilot had to lift off at a lower than normal airspeed, added to which the plane was above normal take-off weight for some reason.

The combination of circumstances then meant that the air speed was below normal and the plane consequently at higher than normal alpha, increasing drag at the same time as engine thrust was being reduced by ingestion of flame from the fire. The only chance the pilot had was to complete a circuit and land which he apparently was trying to do. However this combination of negatives prevented him from gaining any height while trying to turn and we all saw what happened. A lot of money was then spent providing protection to the fuel tanks against such events, but not too long afterwards Air France pulled the plug and the manufacturers weren’t prepared to continue supporting Concorde for BAe use only, so that sadly was it. In truth for such an advanced and ground-breaking design it had an amazing safety record, which makes the disaster even more poignant when you look at root causes.

[Gordon Tarling]

Whilst that Mentour video makes a lot of sense, it doesn't begin to explain just why increasing airspeed and decreasing pitch on the artificial horizon(s) were seemingly ignored. I think there's a lot more convolutions to unfold before we really know what happened.

[Jon H]

I agree with the above posts. For the failure of a single vane to confuse the system to that degree is a problem, however were the pilots expected to intervene? was it man or machine that failed? my suspicion is that it was a mixture of the two as is so often the case.

As for concorde, just off the top of my head i think the chain went something along the lines of:

DC10 thrust reverser wear strip fails
Wear strip replaced with non spec material and not in compliance with correct procedure
Strip falls off on runway some time later
Strip hit by another aircraft bending into a certain shape and leaving it exactly where concorde will hit it just as it reaches V1

Concorde itself was over weight due to excessive fuel and with slight tail heavy c/g
They hit the strip at V1 meaning they had to take off and could not stop on the runway.
The shape of the strip and design of the tyre meant it broke into large fragments and didnt just go flat
A fragment hit just the right spot on the wing to set up a hydrodynamic shockwave in the excessively full tank (there was no air gap due to excess fuelling)
This ruptured the tank causing a massive leak
Fuel entered the intakes of engine 1 and 2 causing surges and loss of power
The aircraft yawed due to asymmetric thrust.
Yaw was exacerbated by a component missing from the port undercarriage.
The undercarriage hit several runway edge lights kicking debris into engines 1 and 2
The fuel trail behind the aircraft was too rich to ignite even with the afterburner. However, the leak was ahead of the undercarriage bay and the turbulence in the bay was enough to 'lean' the mixture and make it combustible.
It also just so happened that high voltage cables in the undercarriage bay had been cut by tyre debris and this started the fire.
The aircraft left the ground much slower than it should have
The fire damaged the undercarriage so it would not retract
The drag of the undercarriage combined with excessive angle of attack due to low speed, excessive weight and only 3 working engines prevented acceleration.
Fire damaged the wing and control surfaced
Control was lost.

That list of variables is just what i can recall from memory 10 years after the lecture. There was also a great deal of trouble in the cockpit due to confusion as they were receiving many warnings that seemed contradictory. Engine 1 lost power and recovered on at least 2 occasions and engine 2 was shutdown due to the fire outside the nacelle giving false fire warnings. I think the chap spent 20 minutes just explaining the engine instruments and the information they were giving. As the pilots cant see the wings from the cockpit they couldnt understand what was happening.

The lecture was extremely interesting and well worth it. There was also a moment of reflection to consider the victims of the crash.