-
It looks like there were similar problems on previous flights and the PIC of the previous flight eventually turned off the trim system after, I think, the third round of AND (automatic nose down) commands from the computer. So, the PIC on the previous flight appears to have done what needed to be done but the crew on the doomed flight did not.
It's also clear that there were more than one problem with airspeed and altitude as well as AOA readings being wrong.
One truly scary thing is that even the previous flight that landed safely the PIC's stick shaker was on almost the entire flight -- that had to be unnerving. Page 14 of the preliminary report shows some graphs of a few of the relevant data items during the 13 minute flight and you can see the near constant sequencing of the AND commands and the near constant manual nose up commands from the pilots and this continues all the way till the end. I'd like to know if the AND commands would still show up even of the system were turned off, I suspect they would not and that would tend to confirm they never turned it off.
By my estimate reading the charts it took less than 20 seconds for the plane to depart the flight altitude and impact the ground. Also interesting is the fact than during the final approximately 45 seconds the pitch trim position, which was a battle between the computer issued AND commands and the pilots manual pitch up commands, shows the computer continuing to issue the AND commands and the pilots continuing to issue manual nose up commands but in that final 45 seconds it doesn't look like the pilots commands had any effect. So, for about 25 seconds of that final battle the AC pitched down a bit and lost a little altitude, but in the final 20 seconds the AC pitched down more noticeably, the altitude dropped more quickly, and the airspeed increased.
The graph from the previous flight is more compressed in time owing to the much longer flight but it appears the crew fought with the computer for about 20 minutes and then there doesn't seem to be any AND commands for over an hour which kind of suggests that with the trim system turned off there are no AND commands.
So, at this point we have:
1. Numerous instrument issues covering several flights involving altitude, airspeed and AOA indications on the PIC (left) side -- this should have been fixed and confirmed fixed before passengers were allowed to fly on it
2. The previous flight reported much the same problems and issued a PAN-PAN but continued the flight after turning off the trim system -- this should have been a PAN-PAN without the option to cancel
3. The pilots of the doomed flight did not turn off the trim system and lost the fight at the end -- the pilots will get significant blame and that appears warranted
4. The maintenance activities appear to suggest the techs relied on computer self tests which seemed to indicate no problems and its not clear what level of physical inspections were done -- did they do a leak test of the pitot static lines or not
5. How was the AOA transmitter installed and does it have a registration pin or other mechanism to make sure it's installed at the right angle -- it not how did the techs set the angle on installation
Brian
-
3. The pilots of the doomed flight did not turn off the trim system and lost the fight at the end -- the pilots will get significant blame and that appears warranted
Unless:
The airline is shown not to have provided sufficient documentation or training on how to fully manage this new MCAS flight system.
-
Well it looks as if there is some progress in not making the poor pilots the scapegoats....
QuoteIndonesian investigators have said the Lion Air plane that crashed last month killing 189 people was not airworthy and should have been grounded.
https://www.bbc.co.uk/news/world-asia-46121127
Therefore, the pilots should never have been placed in that position in the first place!
It seems to me that this thread has taken an distasteful turn towards establishing the guilt of the pilots, prior to the issuing of the full report. -
2. The previous flight reported much the same problems and issued a PAN-PAN but continued the flight after turning off the trim system -- this should have been a PAN-PAN without the option to cancel
From a systems design standpoint, it is undesirable to create one-way doors IMO. If you make it impossible to cancel an urgency or emergency declaration, you increase the cognitive decision-making hurdle for a human to make that declaration. Aircraft and passengers have been lost from crews not declaring emergencies or otherwise clearly communicating the danger the flight was in. (Avianca 52 is the first one that comes to most people's mind, but there are others.)
There is already some resistance in non-commercial pilot community to declaring an emergency when it seems warranted. (There's another fairly vocal sub-group who argues against this position; I find myself in that group at times.) Pilots fear repercussions, paperwork, an investigation that might turn up something unrelated, etc. It's why the Aviation Safety Reporting System is not run by the FAA and provides immunity to crews, mechanics, and other license-holders, etc. I don't fly professionally, but I can only imagine that the pressure to not lose your job makes rule-following even more high stakes.
IMO, there's good reason for the pilot in command to be the final authority as to the conduct of the flight and would argue against taking that authority and responsibility away in normal ops, abnormal ops, or emergencies.
https://www.eevblog.com/forum/chat/lion-air-crash-jakarta-boeing-737-had-prior-instrument-error/msg1949761/#msg1949761
I want my pilots to be thinking "Aviate, Navigate, Communicate" (in that order of priority). I don't want "Litigate" to be on their minds at all; that can all be done at 0' AGL and 0 knots airspeed. -
2. The previous flight reported much the same problems and issued a PAN-PAN but continued the flight after turning off the trim system -- this should have been a PAN-PAN without the option to cancel
From a systems design standpoint, it is undesirable to create one-way doors IMO. If you make it impossible to cancel an urgency or emergency declaration, you increase the cognitive decision-making hurdle for a human to make that declaration. Aircraft and passengers have been lost from crews not declaring emergencies or otherwise clearly communicating the danger the flight was in. (Avianca 52 is the first one that comes to most people's mind, but there are others.)
There is already some resistance in non-commercial pilot community to declaring an emergency when it seems warranted. (There's another fairly vocal sub-group who argues against this position; I find myself in that group at times.) Pilots fear repercussions, paperwork, an investigation that might turn up something unrelated, etc. It's why the Aviation Safety Reporting System is not run by the FAA and provides immunity to crews, mechanics, and other license-holders, etc. I don't fly professionally, but I can only imagine that the pressure to not lose your job makes rule-following even more high stakes.
IMO, there's good reason for the pilot in command to be the final authority as to the conduct of the flight and would argue against taking that authority and responsibility away in normal ops, abnormal ops, or emergencies.
https://www.eevblog.com/forum/chat/lion-air-crash-jakarta-boeing-737-had-prior-instrument-error/msg1949761/#msg1949761
I want my pilots to be thinking "Aviate, Navigate, Communicate" (in that order of priority). I don't want "Litigate" to be on their minds at all; that can all be done at 0' AGL and 0 knots airspeed.
Yes, that is the dilemma and I know of no easy solution to it. Pilots a reticent to delay or cancel a flight for maintenance issues do to pressure from the airline to stay on schedule so they tend to fly even when conditions should dictate they not fly. You are quite right that having no choice in returning when a PAN-PAN is called might make it less likely they will call in one when deserved -- a kind of catch 22. If this were an isolated event that would be one thing, but given the previous flights with similar issues I think this escalates the problem. We are now seeing people in the aviation community stating publicly that this plane was not air worthy and I would agree with that.
As to the comment that we are being too quick to lay blame on the pilots, well, they should have known to turn of the stab trim system which suggests to me the pilots had an inadequate understanding of the flight controls and that's not acceptable for a commercial pilot, even a junior one. To be sure, the failure to clarify the difference between the 737MAX and previous versions with respect to MCAS has to be seen as a contributing factor here.
Brian -
Two dozen attempts -
It's an automatic system switching in and out, doing retries on its own, and possibly operating with bad noisy sensor data.
You have at least four oscillators. The MCAS software, the sensor(s), the pilot, the co-pilot.
The airplane or the pilot as the ultimate authority is yet another conflict.
You will note from the QRH handbook page I posted earlier that there is nothing there about "turn the system back on and see if it does it again." You don't re-enable a dodgy flight control system. You just don't.
Investigators are saying the plane was not airworthy. It should have been grounded as soon as problems surfaced on the previous flight.
One AOA sensor had been replaced, but still a 20 degree discrepancy.
And nobody here is disagreeing with that.
"Airworthy" has a specific, technical legal meaning, and it is not "plane could not be flown." In highly colloquial language it means "the plane should not be flown." It's a huge difference. For example, if an engine fails in flight, the aircraft is obviously not airworthy, but it can still be flown and landed safely.
I have never said that the airplane was airworthy, or that there is nothing wrong with the MCAS by designed or (lack of) training. I have said that the pilots crashed a plane that unless we learn something new and very strange (like the cutout switches didn't work) could have been landed.
Accidents have multiple causes, and one of those causes is the pilots not following procedures.Really need the CVR to figure out all that happened.
CVRs are very useful, but are only one clue among many that investigators use to determine the cause of accidents. Few GA aircraft have CVRs (or FDRs, for that matter) and yet the US NTSB routinely determines proximate and contributing causes to the vast majority of air crashes in the US. Searchable database is here https://www.ntsb.gov/_layouts/ntsb.aviation/index.aspx and always fascinating reading.
-
We don't know the malfunctions the plane experienced.
An intermittent, something cutting in and out can confuse anyone as to what is working properly or not.
Simply switching off the trim motors leaves the stabilizer actuator where it last was- presumably set at nose down. Maybe it was switched back on to get the motors to move it back, if the sensor data smartened up. Maybe they couldn't manually crank them back in time.
Human psychology is to "blame the victim" and the pilot had 6,000hrs and co-pilot 5,000hrs so they may not have been inept. Boeing certainly is.
-
An intermittent, something cutting in and out can confuse anyone as to what is working properly or not.
These guys weren't anyone, they were airline pilots. Furthermore, as has been discussed here ad nauseum, you don't need to know what is working properly or not to get out of this situation. You need to follow the book procedures. (Aside: it turns out that the QRH procedures for runaway stab are also memory items.)
Plenty of airliners have been lost from crews trying to debug something that should have been switched off or ignored.Simply switching off the trim motors leaves the stabilizer actuator where it last was- presumably set at nose down. Maybe it was switched back on to get the motors to move it back, if the sensor data smartened up. Maybe they couldn't manually crank them back in time.
As has already been shown, there was a lot of time, and there was a significant period of more or less level flight. It will be interesting to know if in that period they could have climbed.Human psychology is to "blame the victim" and the pilot had 6,000hrs and co-pilot 5,000hrs so they may not have been inept.
This is not an argument.Boeing certainly is.
Wait, what?
Human psychology is a lot of things, like a desire to be contrarian, or to defer to experts, or not to defer to experts, or to blame big companies, or not wanting to admit that airlines pilots might be fallible, or that 6000 hours of experience can go by without much in the way of experiences.
I think portraying the pilots as victims "before we know what happened" is also probably premature.
-
Reading the Preliminary Report, previous flight did OK:
"Airspeed unreliable and alt disagree shown after take off. STS was also running to the wrong direction, suspected because of speed difference. Identified that CAPT instrument was unreliable and handover control to FO. Continue NNC of Airspeed Unreliable and ALT disagree."
"... the airline confirmed one of their maintenance engineers was on board of the aircraft during the accident flight. This was an "anticipatory measure" in the event of technical problems with the new aircraft."
A third opinion who may have added confusion or given wrong instructions or flipped a breaker.
I'm of the view Boeing has blood on their hands for the engineering.
One of the many questions put forth to the FAA by The Aviation Herald:
- Why was the MCAS permitted to operate on the base of a single AoA value showing too high angle of attacks? Why does the MCAS not consider the other AoA value?
This is a failure of the software algorithm and fault tree analysis. Can't think of one reason to keep the robot going. AoA also seems to feed corrections to airspeed and altitude.
The forum poster, about knowing to shut off a system you don't even know about:
"As a captain on Boeing 737, I feel betrayed about Boeing's statements about their documentation. Even as of now, the MCAS has not been incorporated into the FCOM, nor into the FCTM. Their press release is a shameless lie." -
-
Reading the Preliminary Report, previous flight did OK:
"Airspeed unreliable and alt disagree shown after take off. STS was also running to the wrong direction, suspected because of speed difference. Identified that CAPT instrument was unreliable and handover control to FO. Continue NNC of Airspeed Unreliable and ALT disagree."
"... the airline confirmed one of their maintenance engineers was on board of the aircraft during the accident flight. This was an "anticipatory measure" in the event of technical problems with the new aircraft."
A third opinion who may have added confusion or given wrong instructions or flipped a breaker.
This is just the thing. It was poor aeronautical decision making to attempt to figure out what is going, and there was no need to figure out what what going on. None. Zip. Nada. What needed to happen was for the pilots to quickly execute three memory items, then assess the situation further after the airplane is under full control.
I mean, look at the situation you're describing. They had a tech on board to help debug a system they knew was misbehaving that directly affected safety of flight, with revenue passengers in the back. What kind of ADM is that? And that decision was made on the ground. And given that they were anticipating stab problems, you'd think the crew would be spring-loaded to execute the runaway trim procedures at the first sign of trouble. But it is obvious they never did this. Maybe we'll find out from the CVR why. If it is because they thought they could figure it out, it will be a textbook case of poor ADM.I'm of the view Boeing has blood on their hands for the engineering.
One of the many questions put forth to the FAA by The Aviation Herald:
- Why was the MCAS permitted to operate on the base of a single AoA value showing too high angle of attacks? Why does the MCAS not consider the other AoA value?
I tend to agree that Boeing probably screwed up. That doesn't mean the pilots didn't screw up. This is not unlike AF447 where a bad pitot heater design allowed the pitot system to fail, which caused the computer to stop providing envelope protection -- but that did not force the pilot to execute a zoom climb and then stall the aircraft.The forum poster, about knowing to shut off a system you don't even know about:
"As a captain on Boeing 737, I feel betrayed about Boeing's statements about their documentation. Even as of now, the MCAS has not been incorporated into the FCOM, nor into the FCTM. Their press release is a shameless lie."
The pilot is being misleading, as you do not have to shut off a system you don't know about. You have to shut off the stab trim system.
I think Boeing's position on this is pretty clear: if you follow the procedure, it doesn't matter. Some pilots are saying it does matter. We'll find out.
-
Update today , the CVR is recovered, buried 8 meters underneath the mud in about 30 meters deep.
https://www.reuters.com/article/us-indonesia-crash-recorder/indonesian-officials-say-crashed-lion-air-jets-cockpit-voice-recorder-found-idUSKCN1P808C?il=0
Local news mentioned they used ROV with side scan sonar, magnetometer to find it.
-
Update today , the CVR is recovered, buried 8 meters underneath the mud in about 30 meters deep.
https://www.reuters.com/article/us-indonesia-crash-recorder/indonesian-officials-say-crashed-lion-air-jets-cockpit-voice-recorder-found-idUSKCN1P808C?il=0
Local news mentioned they used ROV with side scan sonar, magnetometer to find it.
That is good news and hopefully this will clear up the actions the pilots made and provide an insight into the sequence of actions they took. The great question is: why did they not turn of the Stab Trim and did they talk about doing so.
The upside to the crash, if there is an upside, is that there can't be a single 737MAX pilot that does not now know what MCAS is and how and when to power off the Stab Trim.
Brian -
Hopefully the recording data is still ok, as from the look, the outer physical unit looks badly wrecked.
-
Hopefully the recording data is still ok, as from the look, the outer physical unit looks badly wrecked.
WOW, had not looked at the pic before, but damn that thing is busted up. The odd thing is you would expect the damage to be less with an impact with water versus land, but at the speed it must have hit the water it might as well have been land.
Brian -
Well crash landings on water are not actually any better than on land.
At the sort of speeds an impact with water makes it feel pretty much as hard as rock. But with the extra problem of also being very slippery and easy to dig into at the same time, if something actually does poke down into the water it experiences a massive amount of drag that slows it down very fast. So when something like the end of a wing touches the water surface the whole plane will suddenly be jerked into a sharp turn before the wing can't take it anymore and tears apart. Once the plane stops its not air tight anymore so it will just sink like a rock and if you do make it out alive you are now swimming in the middle of the sea with no land in sight while the top layer of water if full of kerosine.
Then again such a noise first crash straight into land would be impossible to survive too. But a more controlled crash landing is certainly more survivable on a open field of land rather than at sea. -
Hopefully the recording data is still ok, as from the look, the outer physical unit looks badly wrecked.
WOW, had not looked at the pic before, but damn that thing is busted up. The odd thing is you would expect the damage to be less with an impact with water versus land, but at the speed it must have hit the water it might as well have been land.
Brian
See for youself for.more.photos ...
https://m.detik.com/news/foto-news/d-4383483/penampakan-cvr-lion-air-pk-lqp-yang-akhirnya-ditemukan -
Data capsule is intact, so should read out perfectly well once taken apart. Broke off the electronics section as designed. Just have to hope the data is not degraded from all the overwrites it has had.
-
... Just have to hope the data is not degraded from all the overwrites it has had.
Do the math... it was a new plane, so not many running hours (flight hours plus ground running). I believe they record at least the last 30 minutes, so 2 writes per hour... Not a problem, even for ordinary consumer flash. -
124 minutes of voice recordings from the CVR transcribed, including the final 15 minutes.
"Lion Air crash investigators will not release the contents of the flight's black box voice recordings until August or September."
Another 9 months before it gets released
NTSB has their investigation on hold due to the US Gov't shutdown. -
Hopefully the recording data is still ok, as from the look, the outer physical unit looks badly wrecked.
WOW, had not looked at the pic before, but damn that thing is busted up. The odd thing is you would expect the damage to be less with an impact with water versus land, but at the speed it must have hit the water it might as well have been land.
Brian
See for youself for.more.photos ...
https://m.detik.com/news/foto-news/d-4383483/penampakan-cvr-lion-air-pk-lqp-yang-akhirnya-ditemukan
Water is practically as hard as concrete when impacted at high speed. An airliner that hits water or ground going ~500mph pretty much shatters into tiny particles. It's something a lot of conspiracy theorists fail to understand, expecting to see recognizable airplane wreckage. -
Another 737 MAX:
-
The pilot is being misleading, as you do not have to shut off a system you don't know about. You have to shut off the stab trim system.
I think Boeing's position on this is pretty clear: if you follow the procedure, it doesn't matter. Some pilots are saying it does matter. We'll find out.
If you can internalize why a system works in the way it does it's generally much easier to work with it than just having a bunch of arcane rules. Now at some point the mental model can become too complex too, but that's clearly not the case here.
Just that if Boeing actually changes the manual at this point it's admitting partial guilt, so they feel they can't. So they need to be forced to and penalized for avoiding taking some responsibility. -
Another 737:
Well, the 737 product line as a whole has possibly the best overall safety record, statistically speaking, of any airliner. (There are 1200 737's in the air at any given moment!!!)
But for two 737 MAX's to crash like this in just a few months, that is very odd indeed.