Indonesian aircraft missing off Jakarta

[Rananim]

I know that sometimes the simulator does not entirely accurately reflect the way that the aircraft behaves, however at no stage during the many Airspeed Unreliable events I have flown in the simulator, did the Speed Trim System do anything at all that I can remember, and it most certainly did not behave in such a way that presented any problem.

I concur but.....
We're going round in circles here,data is minimal,no engineering input.
It doesnt have to be STS,it could be AP STAB trim.Whats important is that something gave them unwanted trim based on faulty ADIRU data.Now that something can be STS or AP STAB trim.Thats assuming unwanted "incorrect"(ie incorrect for the reality they were facing) trim is the golden nugget here.Talk of elevator feel providing the incorrect sensitivity for the high speed has been sparse here.I prefer the trim theory as it would explain the sudden loss of control more readily IMO.We can only work on deduction here so lets try that.
Previous flight commander reports ALT IAS disagree and STS trimming "wrong way"------->we deduce UAS undiscoverable on takeoff---->we deduce he was PF and he flew manually during UAS event (STS was operating) before disabling AP stab trim and giving control to FO------->Flight returns to Jakarta below RVSM airspace with FO flying manually.
Accident flight....Commander reads and signs tech log----->we deduce he knows of the previous UAS event----->we deduce further he will surely be stop minded if ASI disagree on takeoff----->takeoff is not aborted------>therefore same fault occurs on accident flight as previous flight------>possible static vent problem------->not discoverable until after liftoff------>Capts ALT lags or sticks------>Capts ASI underreads in climb.
We know from previous UAS accidents crews want to throw it back to the autopilot.Understandable but highly undesirable.What if the accident Captain engages AP B(good side) after takeoff in an effort to give himself time to think?What will happen?AP B will use FCC B to control pitch roll up to MCP selected altitude of 5000 right?FCC B is being fed with right side ADIRU data which we are assuming is good.Things dont go well and control is lost....why?AP stab trim is being controlled by FCC A(bad side data).FCC A commands stab trim/STS signal on power up.Result------->overspeed with AP stab trimming nose down.The crew are confused and disoriented(by low AS on Capts side)and distracting warnings, and dont retard thrust,possibly dont retract flaps.They ask ATC for speed readout and return to Jakarta.AP captures 5000 with difficulty and pitch oscillations,soon after sudden loss of control.

[AI23B]

All of these discussions, suppositions theories and explanations are of no relevance in a country where saving face and money is more important than human life. The aeroplane was logged as having a similar IAS fault more than once and yet was signed off as being serviceable each time. It should have been removed from service and manufacturers input sought after the second incident and it's about as simple as that - hindsight? - no common bloody sense.

[GarageYears]

Originally Posted by JulioLS

Because it has a lot further range than audible frequencies, and detectors are readily available.

Next...

Actually, no, the latest research indicates that an 8.8kHz pinger has much greater underwater range...

"Following the two simulation exercises practiced in this study, it is quite clear that the decision made by the aeronautical administrators to adopt the use of a low-frequency ULB in the fuselage of airplanes was justified. Despite the differences between the two methods, both agree on higher pressure values at longer distances for the 8.8 kHz pings, as expected. The numerical model indicates the possibility of detection (RL > 10 dB [12]) at distances of ~8 km for the 8.8 kHz signal, while for the 37.5 kHz signal the detection can be considered not possible since the water depth (3790m) was bigger then the range where RL = 10 dB."

See: https://www.researchgate.net/publica...ropical_waters

Note: The ULBs fitted in Air France Flight 447, which crashed on 1 June 2009, were certified to transmit at 37.5 kHz for a minimum of 30 days at a temperature of 4°C. Investigating the crash, the BEA recommended that FDR ULBs' transmission period be increased to 90 days and for "airplanes performing public transport flights over maritime areas to be equipped with an additional ULB capable of transmitting on a frequency (for example between 8.5 kHz and 9.5 kHz) and for a duration adapted to the pre-localisation of wreckage" (i.e. with increased range).

- GY

[Smott999]

Thanks for this detail Rananim - question

Originally Posted by Rananim

I concur but.....
We're going round in circles here,data is minimal,no engineering input.
It doesnt have to be STS,it could be AP STAB trim.Whats important is that something gave them unwanted trim based on faulty ADIRU data.Now that something can be STS or AP STAB trim.Thats assuming unwanted "incorrect"(ie incorrect for the reality they were facing) trim is the golden nugget here.Talk of elevator feel providing the incorrect sensitivity for the high speed has been sparse here.I prefer the trim theory as it would explain the sudden loss of control more readily IMO.We can only work on deduction here so lets try that.
Previous flight commander reports ALT IAS disagree and STS trimming "wrong way"------->we deduce UAS undiscoverable on takeoff---->we deduce he was PF and he flew manually during UAS event (STS was operating) before disabling AP stab trim and giving control to FO------->Flight returns to Jakarta below RVSM airspace with FO flying manually.
Accident flight....Commander reads and signs tech log----->we deduce he knows of the previous UAS event----->we deduce further he will surely be stop minded if ASI disagree on takeoff----->takeoff is not aborted------>therefore same fault occurs on accident flight as previous flight------>possible static vent problem------->not discoverable until after liftoff------>Capts ALT lags or sticks------>Capts ASI underreads in climb.
We know from previous UAS accidents crews want to throw it back to the autopilot.Understandable but highly undesirable.What if the accident Captain engages AP B(good side) after takeoff in an effort to give himself time to think?What will happen?AP B will use FCC B to control pitch roll up to MCP selected altitude of 5000 right?FCC B is being fed with right side ADIRU data which we are assuming is good.Things dont go well and control is lost....why?AP stab trim is being controlled by FCC A(bad side data).FCC A commands stab trim/STS signal on power up.Result------->overspeed with AP stab trimming nose down.The crew are confused and disoriented(by low AS on Capts side)and distracting warnings, and dont retard thrust,possibly dont retract flaps.They ask ATC for speed readout and return to Jakarta.AP captures 5000 with difficulty and pitch oscillations,soon after sudden loss of control.

....you theorize Captain realizes poss faulty AS on his side, and engages AP B (good side AS) ...yet AP Stab Trim is still being controlled by the (bad side) FCC A.
What should Captain have done to get Stab Trim looking at B?

As a mere SLF freight it seems odd that Cap would recognize faulty airspeed on his A side, and take action to flip the AP to B side....yet still allow the AP Stab Trim (which I infer is separate somehow) to continue to receive input from the faulty A side...?

Perhaps this is a common mistake? All thoughts appreciated and thanks again for the informative forum.

[bobdxb]

Indonesia's National Transportation Safety Committee (KNKT) said on Monday (Nov 5) that the Lion Air jetliner that plunged into the Java Sea last Monday suffered a more devastating crash impact than an AirAsia flight that crashed into the sea in December 2014.

The size of the debris collected so far showed the ill-fated Lion Air dived into the water at high speed and the plane broke in pieces upon impact with the water body, Mr Soerjanto Tjahjono, chief of the National Transportation Safety Committee, told a radio station.

"The plane broke apart upon impact when it hit the water. There has been no signs of material fatigue. There were worries by some that the plane broke apart in mid-air due to material fatigue. We can confirm it is not the case," he told Elshinta radio.

[bobdxb]

Originally Posted by averdung

Sources say that the Indonesian aviation authority's inspectors have already found "faults" in 3 other Lion Air 737s... things like "sensors not sensing" and display errors like the ones squawked before for the accident airplane. Does not look good for the airline, and puts "contributing factors" very much in play.

In order to dispel rumors, "we want to clarify that in the black box there were four flights that experienced problems with the airspeed indicator," agency head Soerjanto Tjahjono said.

[GarageYears]

Originally Posted by Smott999

....you theorize Captain realizes poss faulty AS on his side, and engages AP B (good side AS) ...yet AP Stab Trim is still being controlled by the (bad side) FCC A.
What should Captain have done to get Stab Trim looking at B?

As a mere SLF freight it seems odd that Cap would recognize faulty airspeed on his A side, and take action to flip the AP to B side....yet still allow the AP Stab Trim (which I infer is separate somehow) to continue to receive input from the faulty A side...?

Perhaps this is a common mistake? All thoughts appreciated and thanks again for the informative forum.

If you're referring to the Speed Trim System (STS) it has it's own speed sensing system based on pitots located at the rear of the aircraft.

-GY

[Rananim]

Where has it been verified that the STS was trimming the wrong way, (meaning 'reversed operation', counter to pilot intention)

It wasnt...it was trimming as designed...but to the pilot it appeared to be trimming the wrong way..... makes sense as he had corrupted CAS data.

What should Captain have done to get Stab Trim looking at B?

He cant as far as I know.FCC A commands the signal from power up.After landing,the squat switch on the gear will change it to FCC B for the next flight.The pilot must disengage AP stab trim to stop the unwanted trim.Using electrical trim overrides it but wont disable it.After 5 seconds,STS will trim again.Pulling on the stick will stop the trim but again it will start trimming if stick pressure is released.It has to be disabled in a UAS(or rather this particular UAS scenario).Ive never encountered this problem during UAS in simulator.However,this doesnt mean it cant happen.Too many variables.....ie type of UAS,which FCC is controlling etc.

[Smott999]

Thanks I am not

Originally Posted by GarageYears

If you're referring to the Speed Trim System (STS) it has it's own speed sensing system based on pitots located at the rear of the aircraft.

-GY

I understand that STS has its own separate AS inputs.

however Ran indicated that AP has Stab Trim input as well, and theorized the Cap May have left AP Stab Trim on the faulty side A.
Merely SLF here and perhaps Ran was in error. Ran weigh in as you will thanks all.

[Lifer01]

Specifically, the Speed Trim function does not operate with Autopilot engaged. It’s essentially a “low speed” function.

The AP can, of course, trim the stab though.

[Smott999]

Thanks again Ram

Originally Posted by Rananim

It wasnt...it was trimming as designed...but to the pilot it appeared to be trimming the wrong way..... makes sense as he had corrupted CAS data.


He cant as far as I know.FCC A commands the signal from power up.After landing,the squat switch on the gear will change it to FCC B for the next flight.The pilot must disengage AP stab trim to stop the unwanted trim.Using electrical trim overrides it but wont disable it.After 5 seconds,STS will trim again.Pulling on the stick will stop the trim but again it will start trimming if stick pressure is released.It has to be disabled in a UAS(or rather this particular UAS scenario).Ive never encountered this problem during UAS in simulator.However,this doesnt mean it cant happen.Too many variables.....ie type of UAS,which FCC is controlling etc.

...again, as a SLF it seems a bit odd that there exists the opp to switch AP to B, whilst AP Stab Trim is fixated on A and at best can be disabled. Why not allow AP Stab Trim to use B? I’m sure there’s a good reason! But your theory is quite scary....AP Stab Trim just continues to push down given it’s on faulty too-slow A-side airspeed...?

[MurphyWasRight]

If the crew had never had a 3rd person involved in a sim setting is it a possibility that the unexpected distraction of that additional voice and potential trouble shooting interaction created enough confusion to lead to a sudden upset or undertake a course of action outside of the preflight brief? I'm wondering if the tech might have asked them to reengage the AP at some point for example?

Originally Posted by JulioLS

The data from flight radar (with all the normal caveats) suggests that they never had control of the aircraft.... But your point about the added distraction of an engineer is well made.

Could someone who knows the regulation and actual practice comment on the allowed role of a flying spanner?
Seems sensible to me that on a revenue flight they should have no input unless an emergency is recognized in which case one of course one would use whatever resources seems appropriate.

Attempting to debug a possible intermittent problem while carrying passengers should be way outside the permitted realm.

I am not a pilot but have plenty of experience with 'group debugging' which is almost always less efficient than a single concentrating person, perhaps with some 'spare hands' turning knobs or holding probes but otherwise staying out of the thought process.

[Rananim]

Originally Posted by GarageYears If you're referring to the Speed Trim System (STS) it has it's own speed sensing system based on pitots located at the rear of the aircraft.

Hmm.........no.We really need engineering input to try and get somewhere here but........ FCCs command the STS, AP STAB TRIM functions.

The air data inertial reference unit (ADIRU) sends these signals
to the FCC for speed trim calculations:
* Computed airspeed (CAS)
* Mach
* Inertial vertical speed
* Roll angle
* Angle of attack.

AMM 22

Each ADIRU sends these air data reference (ADR) signals to its
onside FCC:
* Uncorrected altitude
* Baro corrected altitude number 1
* Baro corrected altitude number 2
* Mach
* Computed airspeed
* Maximum allowable airspeed (VMO/MMO)
* True airspeed
* Static pressure
* Total pressure
* Indicated angle of attack.

AMM 22
From above L ADIRU sends CAS to FCC A. FCC A commands the STS signal via AP stab trim motor,just like when it commands AP stab trim

Inputs to ADIRU?

The pitot static system is comprised of three separate pitot tubes and six flush static ports.Two pitot probes and four static ports interface with the air data modules.The remaining auxiliary pitot probe and alternate static ports provide pitot and static pressure to the standby instruments.The air data modules convert pneumatic pressure to electrical signals and send these data to the ADIRUs.Each pitot air data module is connected to its on-side pitot probe;there is no cross-connection.The air data module connected to the Captain's pitot probe sends information to the left ADIRU,while the air data module connected to the First Officer's pitot probe sends information to the right ADIRU.

FCOM 10.20.13
Elevator pitots have no interface with ADIRUs.They feed the elevator feel computer.

FCC A and FCC B
The FCC can supply these commands:
* A/P trim nose up
* A/P trim nose down
* Speed trim nose up
* Speed trim nose down.
If the autopilot is engaged, the FCC can supply an A/P trim
command. If the autopilot is not engaged, the FCC can supply a
speed trim command.

AMM 22

Column Switching Module
When the main electric trim is in operation, switches in the
column switching module open. This does not let the FCC clutch
signal go to the trim actuator. The FCC and the main electric
trim cannot control the trim actuator at the same time.
If the pilot moves the control column forward or aft, signals go to
the two FCCs. These signals do not allow the FCC to supply trim
commands that do not agree with the pilot.

AMM 22

Speed Trim Stability
The speed trim controls the stabilizer to oppose any change of
airspeed. An increase in CAS causes a nose up trim command
to the stabilizer. A decrease in CAS causes a nose down trim
command to the stabilizer.

AMM 22

STS is not there for fun.Its needed for certification.By retrimming it gives you more elevator authority,particularly important in high AoA low CAS
conditions.Nothing absurd about it.

[Kiwiconehead]

Originally Posted by MurphyWasRight

Could someone who knows the regulation and actual practice comment on the allowed role of a flying spanner?
Seems sensible to me that on a revenue flight they should have no input unless an emergency is recognized in which case one of course one would use whatever resources seems appropriate.

Attempting to debug a possible intermittent problem while carrying passengers should be way outside the permitted realm.

I've done both "flying spanner" - which is paxing on the flight so as to have engineering coverage at an outstation, and being an observer for intermittant faults which is really not helpful in most cases anyway.

Flying spanner I have usually sat in the cabin, although on short flights on one contract sat in the jumpseat as the passengers on that contract asked me too many stupid questions when I was in the cabin.

I have done observation flights for things like yaw dampers dropping out on takeoff rolls and the like, but you can't tell much other than "yes the yaw damper dropped out" CDS fault summaries and QAR provide much more data from which you can analyse faults, if you have technical support (thats what the likes of Boeing are for).

I have been asked advice on what to do with failures when in the jumpseat, and my advice is always "whatever the QRH says" then after that is actioned I have discussed the system with crew. But not for troubleshooting, more for crew information.

[MurphyWasRight]

Thanks KiwiC.

I have been asked advice on what to do with failures when in the jumpseat, and my advice is always "whatever the QRH says" then after that is actioned I have discussed the system with crew. But not for troubleshooting, more for crew information

That is what I would hope/expect to be the case.

BTW: When I mentioned not debugging on a revenue flight I meant taking action, not observing and possibly correlating data not recorded such as actual displays etc although as you point out in most cases it probably just confirms the reported fault.
I suspect there may be times when observing the fault first hand may clarify things since the write up could be ambiguous,vague or operates as designed but not as desired/expected.

[gulliBell]

Please excuse my ignorance with the question, but is it possible to fly a complex commercial jet airliner like this when the flight instruments go whacky by pretending it's a Cessna 172, set the power manually, fly an attitude manually, trim it manually, and look outside? Or is that too difficult for pilots of push button planes now?

And in this instance where an air data computer (or whatever gizmo) was replaced as a fix for a defect entered into the technical log, is a maintenance assessment flight required to actually check that the fault has been fixed?

[DaveReidUK]

Originally Posted by gulliBell

And in this instance where an air data computer (or whatever gizmo) was replaced as a fix for a defect entered into the technical log, is a maintenance assessment flight required to actually check that the fault has been fixed?

Nothing was replaced following the previous DPS-CGK flight (for which we have seen the tech log).

As I posted on the day after the crash:

Originally Posted by DaveReidUK

Somebody please correct me if I'm wrong (I belong to the avionics = PFM school), but the rectification action carried out for that entry in the tech log appears to have been to clean the connector plug on the elevator feel computer followed by a ground test, which was successful.

I'd have thought that (a) the EFC might well have been performing correctly anyway (in flight), given that the warning light responds to differences in either hydraulic pressure or pitot dynamic pressure, which were presumably detected by the EFC and (b) a ground test wouldn't necessarily be able to reproduce either of those conditions.

I'd be interested to know what procedure the AMM prescribes for diagnosing and rectifying this issue.

I haven't seen any subsequent posts re what the AMM says.

[TURIN]

As previously posted, I find it hard to believe that a type that has been in service with the operator for nearly a year and a half, flying on a route that gets 6 daily rotations with the type, needed to carry a flying spanner just to sign off the aircraft at the outstation.

I am frequently called out to assist one particuler airline that has been operating a modern type for nearly two years and they still carry a flying spanner to certify the log. Some airlines just don't like other people messing with their kit and also like to 'control' the certifier. (read into that what you will).

[JLWSanDiego]

Great discussion but wondering if a "TechLog" forum should be started for much of the "what ifs" and "how does it work" as was done on AF447 ?

[SLFinAZ]

Originally Posted by gulliBell

Please excuse my ignorance with the question, but is it possible to fly a complex commercial jet airliner like this when the flight instruments go whacky by pretending it's a Cessna 172, set the power manually, fly an attitude manually, trim it manually, and look outside? Or is that too difficult for pilots of push button planes now?

And in this instance where an air data computer (or whatever gizmo) was replaced as a fix for a defect entered into the technical log, is a maintenance assessment flight required to actually check that the fault has been fixed?

Basically if you look at the previous posts specific to both memory items and the QRH that's more or less what your describing, you start by turning off the various levels of automation and then setting pitch and power. So the real question is what exactly was going on? Was there a compounding technical issue or possibly a slow response that caused a structural failure (overspeed with flaps still down for example) or did the crew get distracted for some reason or fail to follow the recommended procedures? The simple truth is that by itself the UAS should not have been enough to cause a loss of control without some other variables at work. This did not happen at cruise to an unsuspecting crew but was a known issue with an engineer onboard and was in fact an expected event if in fact the 4 previous flights all had issues. The limited factual information makes it somewhat baffling till we get the CVR transcript...