Lionair 610 crash

Status
Not open for further replies.
What are the first indicators of a runaway trim?

The most dangerous time would probably be takeoff, when it would become noticeable as the aircraft would try to rotate before you put in an input.

Have you done SIMS with runaway trim or improperly set trim scenarios?

Not since I left Boeing.

The "cutout switches" refer to the switches to disable the trim motors?

Yes. Right next to the Captain's knee if I recall correctly. But, the trim would disable if you moved the yoke in the opposite direction, so there was never a time issue. The STS intrigues me...at least as it's being discussed in relation to the 737. Again, we need AV's input.
 
Wow! A lot of reports today. Coming to you from a lovely hotel room so here goes:

-The 737 does not have a stick pusher. Only a stick shaker.

-The STS's purpose is the system trying to return the aircraft to a trimmed speed when the A/P is NOT engaged. This is evident during take off and trying to maintain either a flap bug speed or V2+20 or whatever the commanded speed is. So as the speed increases or decreases from the trimmed speed, the stabiliser is commanded in the direction to return the aircraft to the trimmed speed. But designed to operate during those operations of low GW, aft CoG, high thrust (ie take off). It's like an assist for us. But it only turns the trim wheel slowly to counteract the pitch up.

If you don't trim, the Speed Trim automatically trims (at a normal rate) to compensate. The STS system is disengaged by engaging the A/P or the pilot makes trim inputs from the yoke. As the aircraft returns to the trimmed speed the STS commanded stabiliser movement is removed.

-First indicators of a runaway trim are going to be obvious in the 737. Those big wheels next to your knee start moving either uncommanded or keep going once you stop trimming with increasing pressure on the control column. I have done sims with runaway stabiliser and the memory items are quite clear. Hold the control column (to try and disconnect the trim) and disconnect the A/P and A/T. If the runaway continues after A/P disengaged then use the stab trim cutout switches (to disable the motors), if it still continues then grasp and hold the trim wheel. Then we had a go at trimming manually. Becomes quite a crew co-ordination exercise.

Cut out switches are under the flaps (on the FO's side).
 
Last edited:
Turn business expenses into Business Class! Process $10,000 through pay.com.au to score 20,000 bonus PayRewards Points and join 30k+ savvy business owners enjoying these benefits:

- Pay suppliers who don’t take Amex
- Max out credit card rewards—even on government payments
- Earn & Transfer PayRewards Points to 8+ top airline & hotel partners

AFF Supporters can remove this and all advertisements

Full nose down trim if you don't get it stopped...which would be just like Rostov.

But, there is nothing in this bulletin that we haven't already discussed.
 
Noting that the bulletin is only for the -8 and -9 MAX modrls, does this suggest Boeing introduced additional pitch trim programming in these models.

Definitely appears to be a case where Manual doesn't actually mean Manual
 
so what does this mean? serious?
was the plane been pushed into a downward spiral because of erraneous trim?
 
So this means....?

Are Boeing saying that this activity if it happened is within normal operating parameters and therefore human error? (Sorry not a pilot)
 
This article seems to explain it from a lay person's perspective.

Seems like the issue is that an airflow / angle of attack sensor could have caused the plane to go into a dive. there would have been procedures to correct this, but for some reason this wasn't done.

Bloomberg - Are you a robot?

++++++++++++

A Boeing Co. warning to 737 Max operators around the globe provides the first clues about how bad data from an airflow sensor might have contributed to the deadly crash of an Indonesian airliner last week.

The bulletin and statements by Indonesian investigators suggest that the pilots on the Lion Air 737 Max 8 were battling the plane as its computers commanded a steep dive during its final moments of flight.

Boeing cautioned that the so-called angle-of-attack sensor can provide false readings in limited circumstances -- such as when a plane’s autopilot is switched off -- that cause the 737 Max to pitch nose downward. The sensor malfunction can essentially trick the plane into pointing its nose down to gain the speed it thinks it needs to keep flying.

The Boeing directive doesn’t call for operators to conduct new inspections or take other action. It merely stressed that pilots should follow procedures in the flight manual when encountering erroneous data. Following the protocol should be routine for pilots, though may be more challenging in the heat of the moment when equipment is malfunctioning and alarms are sounding.

American aviation regulators followed by issuing an emergency order Wednesday requiring that airlines follow Boeing’s instructions and add information to pilot manuals showing how to diagnose the problem and respond. Carriers will have three days to update their manuals under the order, issued by the Federal Aviation Administration, according to an emailed statement.

The FAA said the problem “could cause the flight crew to have difficulty controlling the airplane, and lead to excessive nose-down attitude, significant altitude loss, and possible impact with terrain.”

What Boeing’s Post-Crash Warning About 737 Max Means: QuickTake

The sensor on the Lion Air jet been replaced the day before after it failed on a previous trip, the Indonesia National Transportation Safety Committee said in a briefing Wednesday. The malfunction can cause the plane’s computers to erroneously detect a mid-flight stall in airflow, causing the aircraft to abruptly dive to regain the speed it needs to keep flying.

On a previous flight from Bali to Jakarta, the angle-of-attack sensor feeding the captain’s flight displays registered a 20-degree difference from the device on the copilot’s side of the coughpit, the committee said. Pilots on that flight were able to compensate.

An angle-of-attack sensor that had been removed before that previous flight has been brought to the investigators and will be examined in the U.S., the Indonesian officials said.

It’s still possible the FAA may order Boeing to redesign the equipment or software as investigators piece together details of the Oct. 29 crash, which killed 189 people. The agency said it “will take further appropriate actions depending on the results of the investigation.”

800x-1.jpg

Boeing 737 Max 7.

Photographer: Simon Dawson/Bloomberg
Fresh Questions
The new information about Lion Air Flight 610 raises multiple questions investigators will want to examine on the pilots’ actions, how flight crews were trained and whether the maintenance performed on the system was adequate, said Roger Cox, a former NTSB investigator.

“I would definitely be looking at the man-machine interface and how pilots respond,” said Cox, a former airline pilot who flew earlier versions of the 737 and specialized at the NTSB in coughpit actions.

One of the puzzling things about the accident is that the plane was flying in clear skies during daylight, so pilots should have been able to handle the problems they faced with airspeed and erroneous sensors, Cox said. However, in rare instances, accidents have been caused by what investigators call a “startle factor.”

“If you don’t take the appropriate action because you’re surprised, you can make a serious error,” he said.

The pilots union at Southwest Airlines Co., the biggest customer of the 737 Max, hasn’t received any reports from its members of problems with faulty sensor readings, said Jon Weaks, president of the Southwest Airlines Pilots Association.

The airline confirmed that it hasn’t experienced any of the sensor troubles and said its 26 Max remain operational and no schedule disruption is expected.

United Technologies Corp. supplies the angle-of-attack sensors and indicator for the 737 Max, according to Airframers.net. The company didn’t respond to requests for comment. Honeywell International Inc. provides the air data inertial reference unit.

The Lion Air jetliner plunged into the Java Sea minutes after takeoff from Jakarta airport, nosing downward so suddenly that it may have hit speeds of 600 miles an hour before slamming into the water.

Moments earlier, the pilots radioed a request to return to Jakarta, but never turned back toward the airport, according to Indonesia’s safety commission and flight-tracking data. The agency said the pilots were dealing with an erroneous airspeed indication.

800x-1.jpg

Rescue vessels line up during their last day of search operation on Nov. 7.

Photographer: Azwar Ipank/AFP via Getty Images
Boeing, which is headquartered in Chicago, said it is cooperating fully and providing technical assistance as the investigation continues. Boeing’s shares rose 1.5 percent to $372.02 at the close of trading in New York. The stock has risen 26 percent this year, the fourth-largest gain on the 30-member Dow Jones Industrial Average.

Lion Air Jet’s Final Plunge May Have Reached 600 Miles Per Hour

While modern jetliners primarily operate on autopilot, the computerized system can disengage when airspeed indicators malfunction, forcing pilots to manually fly the aircraft. If there is an “uncommanded nose-down stabilizer trim” on the Max, pilots can counteract it by pushing a switch on their control yoke. But the plane’s computers will resume trying to dive as soon as they release the switch, the Boeing bulletin said.

Crashed Lion Air Jet Had Faulty Speed Readings on Last 4 Flights

The Most Max
Lion is the biggest customer for the Boeing 737 Max outside North America

Source: Boeing

Note: Top customers through September

Pilots raise and lower the nose of Boeing jetliners by pushing and pulling on a yoke in the coughpit, which controls panels at the tail known as elevators. In addition, a system known as pitch trim can be changed to prompt nose-up or nose-down movement.

The angle of attack readings are fed into a computer that in some cases will attempt to push down the nose using the pitch trim system. In the early days of the jet age, the pitch trim system was linked to several accidents. If pilots aren’t careful, they can cause severe nose-down trim settings that make it impossible to level a plane.

— With assistance by Kyunghee Park, Anurag Kotoky, Mary Schlangenstein, Justin Bachman, and Natasha Rausch

Bloomberg - Are you a robot?
 
so what does this mean? serious?
was the plane been pushed into a downward spiral because of erraneous trim?

What it "could" mean (early days yet), is that there are software/control issues, that IF mismanaged, could be problematic. My read of the Boeing bulletin is, in layman's terms, "if you get this error, for Christ's sake stick to the script" :)
 
It is starting to sound like a feature added by Boeing to the 737MAX intended to improve safety, may have triggered an event that was not caught correctly by the flight crew. It is implied by the Boeing/FAA bulletin that following the operations manual procedures is the correct response to the suspected failure mode of the AOA sensor and its interaction with the trim system.

Certainly Boeing's bulletin to 737MAX operators suggests that believe there is likely to be a link between faulty sensor, trim system and the actions taken/not taken by the flight crew in this accident.
 
Another piece from airliners

My friend is a 737 driver who forwarded a communique from his chief pilot. It says:

"There is no intention to extend the OMB to the 737NG, however many of you are likely wondering why it does not apply to these variants. The OMB describes a scenario where erroneous Angle of Attack (AOA) data causes the pitch trim system to trim the stabilizer down in increments lasting up to 10 seconds each. The pitch trim system is responding to a pre-programed speed trim schedule required to comply with FAR requirements. While the basic architecture of the 737 MAX variants (designated 737-8/9, as opposed to the NG variants which are designated -800/900) is very similar to the NG, there are some pertinent differences “under the hood.” This is particularly true of the speed trim system, which has expanded authority beyond that available to the NG variants to activate during manual flight – this is because of differences in the longitudinal stability of the 737 MAX. "
 
We're still dealing will a lot of 'ifs' here. Nothing is certain.

The steps involved in dealing with the trim are not complex. If it started running away down, the pilot would naturally pull the yoke back against it, and as pilots are taught to always trim, you'd expect a trimming input in the opposite direction to start immediately. You would probably trim to zero load, which is a normal response.

But, 5 seconds after you cease trimming, it could start to run away again. The same response should be effective.

At some point, preferably when the trim is neutral, you'd need to select the cutout switches to stop the sequence. Thinking about it, selecting those switches at the wrong time could be very bad.

This is all very easy from the comfort of my office. There would likely be multiple, and possibly conflicting, alarms going off. Depending upon what was the basic cause, the airspeed displays could also be corrupted. It should be flyable, but there would be a lot going on.
 
This is all very easy from the comfort of my office. There would likely be multiple, and possibly conflicting, alarms going off. Depending upon what was the basic cause, the airspeed displays could also be corrupted. It should be flyable, but there would be a lot going on.

Could this lead to a question about how well trained this airlines pilots are in HUF? Startle effect and the benefit of “doing nothing” for example.
 
Could this lead to a question about how well trained this airlines pilots are in HUF? Startle effect and the benefit of “doing nothing” for example.

We don't know how well they performed, so we can't really make any comments about how well they were trained. This isn't something in which you can do nothing. If the trim is running away, it needs to be controlled immediately.
 
A couple of people asking what does this all mean and was it a sensor that sent the aircraft into a spiral dive?

My understanding from the AD from Boeing addresses a possible failure mode on the MAX where erroneous AoA signals can cause the stabiliser to drive towards a nose down position. This interaction between sensed AoA and stabiliser position forms part of the Manoeuvre Control Augmentation System (MCAS) which is not installed on the NG (any why I couldn't find any reference to such thing in the manuals earlier). To sum it up, Juddles and NM were on the right track. In any case, the AD is clear on one thing, if an uncommanded nose down trim is applied, then memory items for runaway stabiliser needs to be completed and select the cut out switches.

As a side note, the directive from the company is for crew to review the runaway stabiliser non-normal checklist.
 
When you are down..From the Oz

"A Lion Air plane has crashed into a lamp post at an airport in southwest Sumatra as it taxied to the runway last night, just nine days after one of the discount air carrier’s new Boeing 737 Max 8 aircraft plunged into the Java Sea killing all 189 people on board."

Probably paywalled.

Lion.JPG
 
Last edited:
@AviatorInsight
We now can see how the stabiliser is trimmed but how are yaw and roll trimmed?

Also what actually happens when trimming?. Does it alter the hydraulic pressures to achieve a “zero” stick position?. Say on the 737
 
Boeing states this "erroneous" response is known and can happen under certain circumstances (eg - A/P isolation) and pilots should be able to adequately address the issue. Not specifically thinking about this tragic incident, but that response raises all kinds of questions for me. It would seem this premise is based upon an alternate problem pre-existing that requires a pilot to isolate the A/P, so their concentration would be deflected some what, to whatever that pre-existing problem is. With the A/P isolated, the pilot is not just dealing with whatever problem first existed, but they are now effectively manually flying the plane. Then to add to this, the computers, due to a known but unaddressed (by Boeing) data error, effect an immediate and drastic dive, which the FAA have said "could cause impact with terrain" (do commercial aircraft have terrain avoidance systems?). The pilots are then expected to deal with this by "pushing on a switch", which as soon as they stop pushing, returns the plane to a dive. All of this can apparently be stopped by hitting the "cutout switches" but according to jb747, hitting these at the wrong time could be very bad. So the pilot is trying to deal with an existing problem (of unknown complexity), fly the plane manually, pull out of a computer induced dive, continuously push switches (RSI anyone) and then decide when and how to cancel the dive response. Pilots are human and can possibly become confused by multiple conflicting problems and alarms and it seems to me Boeing have contributed to this possible confusion.

As this is a known and expected problem under certain events, why the heck has Boeing not developed an automated disable of that dive function and have alarms instead. Seems to me, that in this incident, a problematic situation has been turned into a deathly situation for many people because of a known, but unaddressed data fault in certain situations and Boeing are now saying that it's a pilots responsibility, with which to deal. Furthermore, there is nothing yet to suggest that this known error caused the crash and it may be a different cause(s).
Am I misunderstanding the conversation so far?
 
Before the confusion comes the “Startle factor”.

It’s well known in engineering that “unintended consequences” can be more of a problem than the original problem which the engineering was trying to fix.

The aircraft would be in the negative G territory assuming reports of 600mph speed is correct because to get there in seconds the aircraft would have massive acceleration. How easy would it be to “flip a switch” or physically do other manoeuvres like manually turn the Stabilizer trim wheel under negative G conditions. At zero G you are weightless. Then add -1G, -2G?

This one is quite illogical. First the system is designed to protect against stall either caused by erroneous pilot input or otherwise. But then the system relies on correct pilot inputs to fix an erroneous system input. What if both are erroneous?. Or what if the system created a problem that even a competent pilot cannot fix : QF72?.

AF477 was in cruise. It got there by competent piloting using competent equipment. Then the equipment became erroneous, and consequently, the pilots also became erroneous, neither could be saved so the whole system collapsed.

The Boeing SOP reminder assumes competent pilots under the most stressful conditions. Increasingly the human is also relying on competent equipment. This interdependence can be dangerous. Human factors suggest that human competency tends to go to mush under these circumstances.
 
Last edited:
Runaway trim will always be an issue, as long as you have aircraft that have more pitch capability within the trim system, than the elevator. Underslung engines provide such a huge pitching moment, that you need to have a very large trim capability. This system exists on every airliner that I can think of.

Runaway could be caused by something as simple as a sticking switch, or as complex as the autopilot.

Boeing states this "erroneous" response is known and can happen under certain circumstances (eg - A/P isolation) and pilots should be able to adequately address the issue. Not specifically thinking about this tragic incident, but that response raises all kinds of questions for me. It would seem this premise is based upon an alternate problem pre-existing that requires a pilot to isolate the A/P, so their concentration would be deflected some what, to whatever that pre-existing problem is.

Manually flying the aircraft is required with any number of aircraft issues. Autopilots do a pretty bad job of flying when things are going wrong, so you generally need to get the aircraft trimmed out, before attempting to engage an a/p. Any form of air data malfunction will drop the autopilot out, as it's likely to go off following wrong data. That's more or less what QF72 did.


With the A/P isolated, the pilot is not just dealing with whatever problem first existed, but they are now effectively manually flying the plane.

That's one reason why there are two pilots. One flies the aircraft, whilst the other addresses the problems.

Then to add to this, the computers, due to a known but unaddressed (by Boeing) data error, effect an immediate and drastic dive, which the FAA have said "could cause impact with terrain" (do commercial aircraft have terrain avoidance systems?).

It inputs nose down trim, that's not actually the same as initiating a drastic dive, though one will lead to the other.....but not immediately.

The pilots are then expected to deal with this by "pushing on a switch", which as soon as they stop pushing, returns the plane to a dive.

The response of a pilot flying would be to pull against the trim, and to simultaneously start trimming using the normal switches. They are under the thumb of the flying hand, and are used constantly when manually flying. This is not onerous. According to Boeing, the aircraft would start trimming again after a delay of 5 seconds, which you could again stop with a manual trim input (under your thumb).

All of this can apparently be stopped by hitting the "cutout switches" but according to jb747, hitting these at the wrong time could be very bad.

The idea is to select the cutout switches within that 5 second gap. If you do that the aircraft would be approximately in trim. If you did it whilst it was running forward, you'd leave yourself with that trim problem, but no way to fix it. You could make that 5 seconds infinitely long by moving the trim switches every few seconds.

So the pilot is trying to deal with an existing problem (of unknown complexity), fly the plane manually, pull out of a computer induced dive, continuously push switches (RSI anyone) and then decide when and how to cancel the dive response. Pilots are human and can possibly become confused by multiple conflicting problems and alarms and it seems to me Boeing have contributed to this possible confusion.

One flies, the other sorts things out.

As this is a known and expected problem under certain events, why the heck has Boeing not developed an automated disable of that dive function and have alarms instead.

I thought you wanted terrain avoidance a few paragraphs ago. That's an automated system that you can't override, which would cause the aircraft to pull up. Any automated system (and that's what this trimming system is) will have the ability to go wrong. More automation is not ever a fix.
 
Status
Not open for further replies.

Become an AFF member!

Join Australian Frequent Flyer (AFF) for free and unlock insider tips, exclusive deals, and global meetups with 65,000+ frequent flyers.

AFF members can also access our Frequent Flyer Training courses, and upgrade to Fast-track your way to expert traveller status and unlock even more exclusive discounts!

AFF forum abbreviations

Wondering about Y, J or any of the other abbreviations used on our forum?

Check out our guide to common AFF acronyms & abbreviations.
Back
Top