Indonesian aircraft missing off Jakarta

[FCeng84]

MCAS will not command more than one increment of nose down trim without either having taken out the first one (via the same amount of nose up trim) or having sensed manual pilot electric pitch trim.

[bsieker]

Originally Posted by FCeng84

It is surprising to me that the details of the MCAS functionality have not come to full focus here within PPRUNE yet. MCAS inserts airplane nose down stabilizer as a function of Mach number and how far beyond a trigger AOA value (a function of flight condition) the airplane has gone. When AOA reduces to below the MCAS sctivation threshold MCAS runs the stabilizer airplane nose up the same amount so that it is back where it started. If the pilot commands manual electric trim while MCAS is active, MCAS is “reset” based on the assumption that the pilot has taken over responsibility for pitch trim and will bring the stabilizer to the proper trimmed position.

The only way to get MCAS to insert more than one increment of airplane nose down trim is for all of the following to occur:
1. High AOA is detected leading to one MCAS increment of stabilizer
2. Manual pilot electric trim command is detected thus causing a reset of MCAS
3. Pilot does not fully command stabilizer to its trimmed position
4. High AOA is once again detected leading to another MCAS increment of stabilizer

For an engineering description (going by your username) this seems oddly vague.

• Do all these things have to happen exactly in order? That would mean MCAS would never do more than one trim increment unless the pilot uses the manually-operated electric trim in between? How do we know that?
• Where does the information come from that MCAS undoes its nose-down input when AoA returns below activation threshold? STS does something like that, but I have not seen anything to that effect about MCAS.
• How would MCAS determine the "trimmed position" of the stabilizer? And would MCAS really never operate if the stabilizer were moved to this "trimmed position" (assuming there is an engineering definition)?
• Is there really only an "activation threshold" AoA value (even if dynamically calculated)? Is rate-of-change of AoA not taken into account? I know that many stick-pusher activation systems do take RoC into account.
• It still seems odd that 2. and 3. above would be required for 4. to be considered a sufficient condition for activation.

Originally Posted by FCeng84

MCAS will not command more than one increment of nose down trim without either having taken out the first one (via the same amount of nose up trim) or having sensed manual pilot electric pitch trim.

So you say, but how do you know? And if that is so, why weren't operators told?

Airbus famously has these "FOR INFO" sections in their FCOMs (e. g. ground spoiler activation conditions), which give more in-depth information about the inner workings and algorithms, that is not part of any exam, but very useful to know. Boeing manuals are lacking something similar and are restricted to what is considered "need to know" (on what basis?).

Cheers,
Bernd

[bsieker]

Originally Posted by non_state_actor

The cost of a charter boat and rescue team pales into insignificance against the cost of implementing what you are suggesting. The data expense alone would be to much to bear for most airlines. Additional to cost you have no guarantee it would even work when you need it. Murphy's Law seems to magically prove itself in aviation time after time.

This would not replace the installed FDR/CDR, but complement it, with somewhat lower dependability requirements, rather more on the level of QARs already installed in most aircraft anyway. But in many cases, crucial information would indeed be available much sooner to give clues to the investigation, later to be corroborated (or not, as the unlikely case may be) and possibly complemented (for the last few seconds, perhaps) by the "definitive" FDR/CVR data.

Cheers,
Bernd

[LEOCh]

Originally Posted by Bergerie1

I emphasise again, the MCAS is not an anti-stall device. It neither detects the stall nor acts as any kind of stick pusher. It is there to maintain the correct longitudinal stability before the stall in order to meet the certification requirements.

FAA Part 25 regulations require that there be a minimum stick force per knot when speed is changed from the trim condition. The MCAS is a design solution (a software fix?) applied to the automatics to increase the required force, giving a degree of "artificial stability" by trimming the stabiliser nose down when hand flying. The paragraphs on longitudinal stability are 5.2.2.1.2, 7.2.1.1.4 and 7.2.2.2.3. See here:-

The relevant point in para 7.2.1.1.4 says (and I quote):- The average gradient of the stick force versus speed curves for each test configuration may not be less than 1 lb for each 6 knots for the appropriate speed ranges specified in § 25.175. Therefore, after each curve is drawn, draw a straight line from the intersection of the curve and the required maximum speed to the trim point. Then draw a straight line from the intersection of the curve and the required minimum speed to the trim point. The slope of these lines must be at least 1 lb for each 6 knots. The local slope of the curve must remain stable for this range.

I would have to respectfully disagree that the MCAS is likely there to comply with FAA 25.175, along with the STS. Note the stick force requirement in that section is not at airspeed less than 1.3VSR. It seems more likely that is there to comply with 25.203 (stall characteristics), so the media is not too far off in calling it an "antistall system".

From what is currently out there (including the Leeham news article), the (-MCAS) MAX has objectionable handling characteristics near the stall. At high AoA the nacelles effectively extend the wing forward, moving the neutral point forward also and degrading the static margin and longitudinal stability. However the problem doesn't really fit neatly into the certification regulations, as although the simultaneous pitch up (made worse by likely high power/thrust line effects) is like what an unstable aircraft would do, it is in fact an uncommanded pitch up alongside a transition into a less longitudinally stable (possibly unstable) state. This is likely to result in a stall unless promptly attended to manually or automatically.

Section 25.203 Stall characteristics.

(a) It must be possible to produce and to correct roll and yaw by unreversed use of the aileron and rudder controls, up to the time the airplane is stalled. No abnormal nose-up pitching may occur. The longitudinal control force must be positive up to and throughout the stall. In addition, it must be possible to promptly prevent stalling and to recover from a stall by normal use of the controls.

The abnormal nose-pitch in the regulation is during the stall, a possibility if the tail stalls before the wing (as can occur if COG is behind the neutral point). What the -MCAS MAX does is (likely) a bit different, it pitches up before the stall, and it may (appropriately) pitch down when the stall occurs. It would be hard to argue that by this it is compliant though and no augmentation is necessary. However the degraded longitudinal stability after the uncommanded pitch up would likely also fail 25.203.

Although MCAS it is not an antistall system per se, it effectively functions as one. It should be harder to achieve an intentional 1g stall in a MAX with MCAS operating than a NG.

[bsieker]

Originally Posted by LEOCh

[...]
The abnormal nose-pitch in the regulation is during the stall, a possibility if the tail stalls before the wing (as can occur if COG is behind the neutral point).

During pitch-up and high-AoA maneuvers (and in statically stable / dynamically damped aircraft also in neutral flight), the tailplane exerts a downwards force, even in a rearward CoG situation. If it stalls, the force pushing the tail down will reduce, and the nose-up moment will reduce, so it will not result in (additional) nose-up moment.

(as can occur if COG is behind the neutral point)

That must never occur. You cannot get civil airplanes certified otherwise. If it does (e. g. by massive load-shift), catastrophic things will happen, as the 747 crew experienced in Afghanistan a few years ago.

Bernd

[FCeng84]

Originally Posted by bsieker

• Do all these things have to happen exactly in order? That would mean MCAS would never do more than one trim increment unless the pilot uses the manually-operated electric trim in between? How do we know that?
• Where does the information come from that MCAS undoes its nose-down input when AoA returns below activation threshold? STS does something like that, but I have not seen anything to that effect about MCAS.
• How would MCAS determine the "trimmed position" of the stabilizer? And would MCAS really never operate if the stabilizer were moved to this "trimmed position" (assuming there is an engineering definition)?

So you say, but how do you know? And if that is so, why weren't operators told?

Bernd

Bernd - As for your first two points above, you have it right. MCAS does not know where “trim” is, but it does two things related to trim. First, it will undo what it does in response to high AOA when AOA reduces provided the crew has not activated trim themselves. Second, it makes the assumption that if the crew does make their own trim inputs, they know better and will provide sufficient trim inputs to command the stabilizer to the proper trimmed position.

I cannot speak to the pilot training / education history. My knowledge is of the system operation.

[MickG0105]

Originally Posted by FCeng84

MCAS will not command more than one increment of nose down trim without either having taken out the first one (via the same amount of nose up trim) or having sensed manual pilot electric pitch trim.

Okay, I'm probably not the sharpest tool in the drawer here so please bear with me. Are you saying that Boeing designed a system to address a stall certification issue relating to a pitch up tendency; that system is triggered when the airplane Angle of Attack exceeds a specified threshold; it reacts by winding in one increment of nose down stabiliser trim; having done that if left to its own devices even if the airplane Angle of Attack still exceeds the trigger threshold the system does NOTHING. That strikes me as an odd design. ​​​​​​

[FCeng84]

The amount of nose down stabilizer MCAS puts in is a function of how far beyond the activation AOA MCAS senses that the airplane has gone. You have it correct that MCAS leaves that stabilizer increment in as long as AOA is high. The idea is that the nose down stabilizer increment will help the airplane recover to a more realistic lower AOA and thus should be left in place until AOA is reduced

[bsieker]

Originally Posted by FCeng84

Bernd - As for your first two points above, you have it right. MCAS does not know where “trim” is, but it does two things related to trim. First, it will undo what it does in response to high AOA when AOA reduces provided the crew has not activated trim themselves. Second, it makes the assumption that if the crew does make their own trim inputs, they know better and will provide sufficient trim inputs to command the stabilizer to the proper trimmed position.

Thanks, FCeng84.

I'll replicate your 4 points from before for reference:

1. High AOA is detected leading to one MCAS increment of stabilizer
2. Manual pilot electric trim command is detected thus causing a reset of MCAS
3. Pilot does not fully command stabilizer to its trimmed position
4. High AOA is once again detected leading to another MCAS increment of stabilizer

Still trying to understand what you are saying: so your point 3. above is really just implicit: they probably didn't trim properly because they (4.) encountered high AoA again.

It starts making a bit more sense.

Bernd

[FCeng84]

With AOA reading erroneously high, each time that the crew trimmed nose up the criteria for MCAS to put in another nose down increment were satisfied. For several minutes the pattern of MCAS puts in nose down followed by pilot putting in nose up repeated many times. Why during the last minute the crew responded to MCAS stabilizer increments with quick blips of trim but not nearly enough to re-trim the stabilizer is a mystery. All of that on top of why the crew allowed the system to take the airplane away from trim so many times without recognizing they should disable automatic stabilizer control ....

[bsieker]

Originally Posted by FCeng84

With AOA reading erroneously high, each time that the crew trimmed nose up the criteria for MCAS to put in another nose down increment were satisfied. For several minutes the pattern of MCAS puts in nose down followed by pilot putting in nose up repeated many times. Why during the last minute the crew responded to MCAS stabilizer increments with quick blips of trim but not nearly enough to re-trim the stabilizer is a mystery. All of that on top of why the crew allowed the system to take the airplane away from trim so many times without recognizing they should disable automatic stabilizer control ....

Indeed a mystery. The CVR will hopefully shed some light on it, and provide valuable information on how diagnosing a problem like this can be made easier.

But really, the knowledge that leaving electric trim alone would also have inhibited additional MCAS inputs might have been valuable. I infer that using the mechanical trim wheel will not reset MCAS?

Bernd

[LEOCh]

Hello Bernd, thanks for your reply

Originally Posted by bsieker

During pitch-up and high-AoA maneuvers (and in statically stable / dynamically damped aircraft also in neutral flight), the tailplane exerts a downwards force, even in a rearward CoG situation. If it stalls, the force pushing the tail down will reduce, and the nose-up moment will reduce, so it will not result in (additional) nose-up moment.

I would mostly agree but assert that static longitudinal stability in general requires the horizontal tail (stab+elevator) to be at lower AoA than the wing, but not necessarily in downforce. Drag is generally lowest (and efficiency highest) with COG loaded to a rear limit only slightly in front of the neutral point. In this configuration the tail is generating negligible or zero downforce (and hence minimal drag), but the aircraft still has positive but weak stability. The longitudinal stability can then be augmented for certification/handling (e.g. by STS)

Originally Posted by bsieker

That must never occur. You cannot get civil airplanes certified otherwise. If it does (e. g. by massive load-shift), catastrophic things will happen, as the 747 crew experienced in Afghanistan a few years ago.
Bernd

Quite right, was just pointing that FAA 25.203 describes (and forbids) what can happen to an aircraft loaded to static instability in the stall, the tail operating at higher AoA stalling before the wing causing pitch up. This is not what occurs in the (-MCAS) MAX, but the fact that it pitches up just before the stall due to the nacelles might be construed as similar for certification.

[silverstrata]

Originally Posted by zzuf

Really interesting, can the aircraft be stalled with the MCAS operating?

I think MCAS was a rudimentary attempt to prevent a stall developing.

When at high angles of attack, the 2.5 degrees of forward stab trim would apply control column pressure, and thus urge the pilot to lower the nose (presumably with the stick shaker also operating). And if the high angle of attack persisted, the system would give another 2.5 degrees of forward trim. So far, so good.

However....

2.5 degrees of trim is quite a lot. And if you get the second dose, 5.0 degrees is a helleva lot. And the system can apparently run to full nose down trim, which is simply insane.

Ok, you are now in a 15 or 20 degree nose down dive, to recover from the stall. So how do you pull out of the dive, with full stab trim forward? I recon that is impossible. (I see no evidence in the explanations given, that the MCAS system will automatically rewind - although that might be a sensible option.)

As the description says...

MCAS is implemented on the 737 MAX to enhance pitch characteristics ... at elevated (high) angles of attack. The MCAS function commands nose down stabilizer to enhance pitch characteristics during ... flaps up flight at airspeeds approaching the stall ... If the original elevated AOA condition persists, the MCAS function commands another incremental stabilizer nose down command.

To ‘enhance pitch characteristics’ is lawyer-speak for “get the flippin nose down”. And if the nose is not lowered, you get another dose of nose-down trim - until you are pointing vertically at the ground. I really think that nobody thought this system through....

(And yet Bergerie1 still cannot understand that a device that operates to lower the nose when approaching the stall, is an anti-stall device...! Takes all sorts, I suppose.)

Silver

[MickG0105]

Originally Posted by silverstrata

... And if the high angle of attack persisted, the system would give another 2.5 degrees of forward trim. ...

Well, that's what I thought but that is not the case according to FCeng84:

Originally Posted by FCeng84

MCAS will not command more than one increment of nose down trim without either having taken out the first one (via the same amount of nose up trim) or having sensed manual pilot electric pitch trim.

​​​​​​​

[climber314]

"The silver lining for Boeing is that a Lion Air cancellation would mean less exposure to an airline with dicey credit,” said Hamilton. “Plus, it would open up delivery slots for new orders in competition with the Airbus A320neo.”

According to two sources, Lion Air last year was in financial difficulties and was restructuring leases and considering deferring some orders. Many in the industry believe Boeing helped bail out the airline...

Others in the industry besides Hamilton agree that Lion Air is overstretched.

“They bit off more than they can chew,” said Adam Pilarski, an analyst with consultancy Avitas.

It may also be an opportune moment to cut some of Lion Air’s 188 unfilled MAX orders. Industry sources say it’s been clear for some time that the airline has ordered more planes than it can handle.

https://politicalreporter.net/2018/1...seattle-times/

-------
TEMPO.CO, Jakarta - President Director of Lion Air Edward Sirait said that until now Lion Air has not changed the plan to order around 200 Boeing 737 Max. The planes will still be sent based on the agreement schedule of two companies. "Still, still (according to the original plan)," Edward said at the Office of the Coordinating Ministry of Maritime Affairs, Jakarta, Tuesday, December 4, 2018.

https://bisnis.tempo.co/read/1152302/lion-air-pembelian-boeing-masih-sesuai-rencana/full&view=ok

Something doesn't add up about this whole deal. The Rupiah is down over 40% against the Dollar since 2011, the year the big Boeing deal was signed. Yet Lion Air flies a fleet of brand new jets with a subpar credit rating? Somebody is losing big money here.

Boeing doesn't seem to care about attributing blame to Lion Air. In other cases silence is the norm. Here Boeing commented on an ongoing investigation which is highly unusual. In the big picture the bad P/R may not be worth 188+/- sales with a 4,700+/- backlog on the books.

This doesn't absolve Boeing for not disclosing MCAS, but it just may help to explain how one plane lands and another crashes while facing the same set of issues.

[gums]

Salute!

So I would like to review the bidding, huh?
- stick shaker activates as pilot rotates and lifts off. I can see this happening with an abrupt stick input at the normal speed and maybe a bit more AoA than normal. Stick shaker continues. A caution indication or two is displayed, but altitude is insurance, and plane is climbing.
- MCAS, aka "Hal", sees the high AoA, flaps are up, and speed being way less than the mach washout limit, bumps the stab maybe to the max limit for one "bump" and stick forces go up.
- crew feels this and beeps trim with the stick switch.You can see the increased stick force correlating with the MCAS down trim cmd and resulting stab movement. Stab reverses, stick forces go down and the 5 second timer starts. Hal is waiting, and the high AoA is there.
- Hal again cmds stab to move another increment, and crew feels this, as shown on stick forces, then makes manual up trim inputs.
- The plot shows a gradual increase in down trim position of the stab because everytime crew beeps the trim Hal stops, but stab not trimmed all the way back up to original degrees. So the slope of stab position values is slightly negative.
- Hal waits 5 seconds and the process repeats two dozen times until "the event"
- At "the event", control input force from each pilot diverge, Hal continues to trim nose down and crew briefly beeps trim but mainly pulls back harder and harder.

Is that how most here see it?

Gums asks....

[climber314]

That's the WHAT happened gums. WHY it went down that way is at the bottom of the Java Sea on a CVR at the moment. Ian W has a good point. We should know more by now.

"The financier also told the Seattle Times that Kirana recognizes that he had ordered (too) many jets but lacked the necessary capital."
https://www.aerotelegraph.com/streic...gen-bei-boeing

If this turns out to be "pilot error" it helps Boeing's case. By the same token, Lion Air could face a substantial increase insurance and legal costs. Their financial status is already sketchy by all accounts.

Follow the money.

[Mad (Flt) Scientist]

Originally Posted by zzuf

Really interesting, can the aircraft be stalled with the MCAS operating?

Must be able to do so, otherwise the whole basis of the stall certification would have changed, as it would make the aircraft "envelope protected" if MCAS were able to prevent a stall. AFAIK 737MAX has no such radical change in the certification approach. It seems to be intended to enhance/amend/adjust stall characteristics, not to make a more fundamental design change.

[ATC Watcher]

climber314

it just may help to explain how one plane lands and another crashes while facing the same set of issues.

If I understood the report correctly in one case someone had the reflex to cut out the Stab Trim switches ,and in the second case not.

Lion air is a low cost airline run like a local business in an unregulated country. Anyone who has seen the way Indonesians interpret the road traffic rules will understand what I mean .
But Indonesia is not unique . There are lots of countries out there where dispatching a car, a motorbike or an aircraft with minor defects is not an issue , not dispatching it is one and could eventually lead to someone losing its job.

gums

Is that how most here see it?

Salute, Yes a very good summary I would say ..

[FCeng84]

gums - I mostly agree. The one point I would add is that the gradual net nose down stabilizer motion prior to the last minute is likely that required to maintain pitch trim as speed increases. They were going rather fast already prior to the last minute. It looks like the crew was getting back to near zero column force with each cycle up until that point.