How Safe is the Boeing 737 Max's MCAS System?

[PeterDonis]

From my understanding, adjusting the trim manually does disable MCAS. But it can reengage after 5 seconds or so if all the same logic like AoA and airspeed are true.

Yes, that's my understanding as well. And of course in the presence of faulty AoA sensor data which is not detected as faulty, this will cause MCAS to keep on adding nose down trim again and again, creating the "tug of war" situation.

To really disable it apparently, I've pointed out in a previous post, you actually need to correct the trim electrically first, then disconnect it using cutouts, then use manual trim for the rest of the flight. And then in a post I pointed out a pilot commenting that disabling those cutouts is what a pilot would do first in a nanosecond and that they don't disable MCAS.

I'm not sure this is true. The various comments online don't all appear consistent. But the preliminary report from Indonesia on the Lion Air incident includes a description of key events on the previous flight of that aircraft; on that flight, it appears that the pilots noticed that automatic nose down trim was being put in, used the cutout switch, and there were no further trim problems for the rest of the flight. That flight was the night before; in the morning, the flight that crashed, there was, as far as I can tell, a different flight crew, who apparently did not use the cutout switch at all, but kept trying to put in nose up trim manually to counter the repeated automatic nose down trim.

 

[cyboman]

Yes, that's my understanding as well. And of course in the presence of faulty AoA sensor data which is not detected as faulty, this will cause MCAS to keep on adding nose down trim again and again, creating the "tug of war" situation.
I'm not sure this is true. The various comments online don't all appear consistent. But the preliminary report from Indonesia on the Lion Air incident includes a description of key events on the previous flight of that aircraft; on that flight, it appears that the pilots noticed that automatic nose down trim was being put in, used the cutout switch, and there were no further trim problems for the rest of the flight. That flight was the night before; in the morning, the flight that crashed, there was, as far as I can tell, a different flight crew, who apparently did not use the cutout switch at all, but kept trying to put in nose up trim manually to counter the repeated automatic nose down trim.

Yes it is confusing there is conflicting reports, opinions and statements. From the actual air authority bulletins is where I got the need to adjust the trim electrically first before disabling as if you proceed without doing that, I guess it's suggested the pilot may not be able to recover the trim manually.

 

[PeterDonis]

from what I could gather you made it clear that MCAS does not cause the aircraft to pitch down.

I've already explained repeatedly what I meant. I'm not going to explain it again. Once more: I'm not disagreeing with you about the aerodynamics of the plane, or about the fact that MCAS adjusts the horizontal stabilizer (since that's what the trim system in general does). I'm just emphasizing different aspects than you are.

Also you say the purpose is not to pitch the nose down. It's to adjust trim. Huh? It adjusts trim TO pitch the nose down. Not pitch the nose up or keep it neutral or provide stick feel. It says exactly that in the graph I pasted.

I've already explained repeatedly my understanding of the purpose of the trim system. The graph you pasted is not a technical publication, nor is it a pilot's manual; I would not rely on its exact wording any more than I would rely on a pop science article's exact wording to learn actual science.

Try talking to pilots and asking them if they think the purpose of the trim system is to pitch the nose down. Or read the "See How It Flies" articles I linked to, which explain in detail how pilots use the trim system and what they think it's for.

 

[jim hardy]

worth a look
http://www.avioesemusicas.com/wp-content/uploads/2018/10/TBC-19-Uncommanded-Nose-Down-Stab-Trim-Due-to-AOA.pdf
it describes the tug-of-war that can result

 

[PeterDonis]

From the actual air authority bulletins is where I got the need to adjust the trim electrically first before disabling as if you proceed without doing that, I guess it's suggested the pilot may not be able to recover the trim manually.

The air authority bulletins are presumably going to give Boeing's recommended procedures, which, as I think we've already noted, do not say to immediately use the trim cutout switch. But I don't know what kind of failure condition those procedures assume. If the procedures are not designed with the failure condition of a faulty AoA sensor in mind, they might not work well for that failure condition. It certainly seems to me that, if the pilot believes the AoA sensor data is faulty, the best response would be to immediately cut out all automatic trim adjustments. It looks like the previous evening's Lion Air flight crew did that and it worked, allowing them to complete the flight safely.

 

[cyboman]

I've already explained repeatedly what I meant. I'm not going to explain it again. Once more: I'm not disagreeing with you about the aerodynamics of the plane, or about the fact that MCAS adjusts the horizontal stabilizer (since that's what the trim system in general does). I'm just emphasizing different aspects than you are.
I've already explained repeatedly my understanding of the purpose of the trim system. The graph you pasted is not a technical publication, nor is it a pilot's manual; I would not rely on its exact wording any more than I would rely on a pop science article's exact wording to learn actual science.

Try talking to pilots and asking them if they think the purpose of the trim system is to pitch the nose down. Or read the "See How It Flies" articles I linked to, which explain in detail how pilots use the trim system and what they think it's for.

I understand what the trim system is used for. That's not equivalent to what MCAS is used for. And I don't think trim is defined primarily as a force feedback system. It eliminates control forces as an effect of changed aerodynamics to the stab.

It's not just the graph, I've quoted comments from pilots and various other sources. I can find more that will literally say: The purpose of MCAS is to counteract the positive pitch attitude caused by new engines on the craft, it does this by sending nose down trim commands. Or something like that said 100 different ways.

Again I see you actually replied to this comment. I would hypothesize the answer to your question here is no:

Albrecht
March 15, 2019 at 3:03 pm
“Artificial feel” trim has been present in 737s since ages. MCAS is a new, additional module not for stick feel, but to add envelope protection at low speed and high AOA. It was required due to the aerodynamic instability caused by the larger, more forward engine nacelles.

REPLY
Peter Donis
March 15, 2019 at 8:58 pm
Isn’t this equivalent to saying that MCAS is adjusting the “artificial feel” trim to compensate for the effect of the new engine nacelles? As I understand it, Boeing’s stated purpose for doing this was to make the plane feel like previous 737s, to avoid having to retrain pilots.

https://airfactsjournal.com/2019/03/can-boeing-trust-pilots/

 

[PeterDonis]

worth a look

This seems to confirm that the cutout switches disable MCAS; but it does say "both" cutout switches. So perhaps if only one of the two switches (I assume they mean there is one for the pilot and one for the copilot) is thrown, MCAS is not disabled; only throwing both does.

 

[PeterDonis]

I understand what the trim system is used for. That's not equivalent to what MCAS is used for.

Huh? The MCAS system is a part of the trim system.

I've quoted comments from pilots and various other sources. I can find more that will literally say: The purpose of MCAS is to counteract the positive pitch attitude caused by new engines on the craft, it does this by sending nose down trim commands.

Sure. But "sending nose down trim commands" is not the same as "pitching the nose down". It's not even the same as "commanding the nose to pitch down". You do that using the yoke, not the trim system. Ask a pilot. Or read the "See How It Flies" articles.

 

[jim hardy]

eyewitness reports of smoke and debris from the plane on its way down are another item in my list of unresolved observations.

The Ethiopian Airlines plane that crashed killing 157 people was making a strange rattling noise and trailed smoke and debris as it swerved above a field of panicked cows before hitting earth, according to witnesses.

Flight 302 took off from the Ethiopian capital on Sunday morning bound for Nairobi with passengers from more than 30 countries. All on board the Boeing 737 MAX 8 died.

The pilot had requested permission to return, saying he was having problems - but it was too late.

Half a dozen witnesses interviewed by Reuters in the farmland where the plane came down reported smoke billowing out behind, while four of them also described a loud sound.

"It was a loud rattling sound. Like straining and shaking metal," said Turn Buzuna, a 26-year-old housewife and farmer who lives about 300 meters (328 yards) from the crash site.

old jim

 

[PeterDonis]

I see you actually replied to this comment. I would hypothesize the answer to your question here is no:

We'll see if anyone replies. I would be interested in the answer.

 

[cyboman]

Huh? The MCAS system is a part of the trim system.
Sure. But "sending nose down trim commands" is not the same as "pitching the nose down". It's not even the same as "commanding the nose to pitch down". You do that using the yoke, not the trim system. Ask a pilot. Or read the "See How It Flies" articles.

MCAS adjusts trim TO pitch the nose down. You're just making semantic arguments that don't really say anything in my opinion. The commands MCAS sends results in the nose pitching down. I can say it a dozen different ways. I don't think you want to admit that for some reason. Your assertion that MCAS is primarily to provide column force feedback is completely incorrect in my opinion.

 

[cyboman]

We'll see if anyone replies. I would be interested in the answer.

I would assume you disagree with Albrecht who states MCAS is not for stick feel.

 

[cyboman]

eyewitness reports of smoke and debris from the plane on its way down are another item in my list of unresolved observations.

old jim

Just as a side note I read somewhere other witnesses said they didn't see smoke. So there's some conflicting reports among the witnesses.

Though totally possible, it seems unlikely such a catastrophic mechanical failure causing fire / smoke would happen in a plane that new.

 

[PeterDonis]

You're just making semantic arguments that don't really say anything in my opinion.

No, I'm disagreeing with a particular categorical statement you keep making:

The commands MCAS sends results in the nose pitching down.

This is not true categorically, it's only true with a particular qualifier: if the pilot is not applying force to the yoke to compensate. And as I've already explained, in normal operation, any time the MCAS is sending nose down trim commands, the pilot will be applying force to the yoke to compensate, so the result of the MCAS trim commands will not be to pitch the nose down, it will be to increase the feedback force on the yoke that the pilot feels.

You have not bothered to respond to my repeated statements along these lines even once.

 

[PeterDonis]

I would assume you disagree with Albrecht who states MCAS is not for stick feel.

That's why I replied to him. Because I've seen plenty of other pilots (including the one in the article OCR linked to, which I mentioned before) say that it is. Not to mention Boeing saying that the reason they put MCAS in was to make the 737 MAX feel like previous 737s so pilots wouldn't have to be retrained. I want to see how he responds to that.

 

[cyboman]

No, I'm disagreeing with a particular categorical statement you keep making:
This is not true categorically, it's only true with a particular qualifier: if the pilot is not applying force to the yoke to compensate. And as I've already explained, in normal operation, any time the MCAS is sending nose down trim commands, the pilot will be applying force to the yoke to compensate, so the result of the MCAS trim commands will not be to pitch the nose down, it will be to increase the feedback force on the yoke that the pilot feels.

You have not bothered to respond to my repeated statements along these lines even once.

I have been engaging with this very assertion over and over. You're basically saying that if the system is acting without any additional input, it will result in the nose pitching down. And saying it's not suppose to operate that way, it's commands for nose down trim are supposed to be overridden by pilot yoke input. At least that's what it sounds like you're saying but I'm sure you'll clarify further somehow. In any case, I don't agree, the simple way to refute this is the engines are known to create a positive pitch attitude force. MCAS is designed to counteract that. You can get into the details of the use cases of how MCAS operates as it seems you do. But the primary function of MCAS is to correct that pitch up effect of the engines not provide force feedback or stick feel.

 

[cyboman]

That's why I replied to him. Because I've seen plenty of other pilots (including the one in the article OCR linked to, which I mentioned before) say that it is. Not to mention Boeing saying that the reason they put MCAS in was to make the 737 MAX feel like previous 737s so pilots wouldn't have to be retrained. I want to see how he responds to that.

You think Boeing saying "make it feel like previous 737s" scientifically proves MCAS is primarily an artificial feel system and isn't meant to correct pitch attitude?

 

[jim hardy]

That's the reason behind it, to make the yoke feel similar to earlier planes when performing maneuvers involving steep pitch and bank..

Without it the plane would respond differently than earlier ones.

It's a closed loop : pilot - yoke - control surfaces - pitch - back to pilot
and in a closed loop you can observe it anywhere and you'll have a signal coming in and a signal going out

@PeterDonis is i think examining the loop at the node pilot-yoke
@cyboman i think is examining it at the node pitch-back to pilot
MCAS adds a computed signal at the node yoke-control surfaces so pilot receives same pitch to yoke response as in earlier planes and they don't have to re-certify plane or pilot. ..

automatic control is confusing that way. You can start anyplace in the loop and write a transfer function.
it's hard to express in words but the math is just that way.
You can't open the loop to analyze it. I'd say that's the crux of your disagreement.old jim

 

[cyboman]

That's the reason behind it, to make the yoke feel similar to earlier planes when performing maneuvers involving steep pitch and bank..

Without it the plane would respond differently than earlier ones.

It's a closed loop - pilot - yoke - control surfaces - pitch - back to pilot
and in a closed loop you can observe it anywhere and you'll have a signal coming in and a signal going out

@PeterDonis is i think examining the loop at the node pilot-yoke
@cyboman i think is examining it at the node pitch-back to pilot
MCAS adds a computed signal at the node yoke-control surfaces.

automatic control is confusing that way. You can start anyplace in the loop and write a transfer function.
it's hard to express in words but the math is just that way.old jim

That's very interesting. I really like that loop diagram. I think I need to understand it better.

This I think helps to illustrate my point: MCAS is working after the yoke, it's pathway is to the control systems, then the feedback eventually get's back to the pilot via the yoke. But the MCAS system is not operating on the yoke and it's fundamental role is not force feedback.

So shouldn't it be: pilot - yoke - control surfaces - pitch - yoke - back to pilot

And isn't pitch part of control surfaces?

pilot - yoke - (MCAS) - control surfaces - yoke - back to pilot

 

[jim hardy]

So shouldn't it be: pilot - yoke - control surfaces - pitch - yoke - back to pilot

i don't see a second yoke in the loop.

But indeed MCAS is in the loop where you drew it.

 

[OCR]

Great find there.

Thank you...

Would that wake effect get worse with more airspeed ? As they found when WW2 fighter planes got close to mach 1?

I'm not sure... but here is the basics of the phenomenon you refer to...

Mach tuck - Wikipedia

...experienced pilots would get into a PIO.

And the acronym PIO, explained by a very short video... .

Pilot induced-Oscillation

I don’t recommend flying the Bleriot any higher than you are willing to jump!

If you’re faced with a forced landing, fly the thing as far into the crash as possible... fly it until the last piece stops moving.

.

 

[jim hardy]

then the feedback eventually get's back to the pilot via the yoke

pitch feedback gets back to pilot via horizon , either the real one or artificial one (or AOA gage if there is one)
and probably his "seat of the pants" feel.

 

[cyboman]

i don't see a second yoke in the loop.

But indeed MCAS is in the loop where you drew it.

I see, I suppose I was thinking about where the feedback occurs on the yoke. So the pilot pulls up on the yoke, it adjusts the control surfaces, that effects plane pitch attitude, those forces then are felt on the yoke, and to the pilot.

 

[cyboman]

No, I'm disagreeing with a particular categorical statement you keep making:
This is not true categorically, it's only true with a particular qualifier: if the pilot is not applying force to the yoke to compensate. And as I've already explained, in normal operation, any time the MCAS is sending nose down trim commands, the pilot will be applying force to the yoke to compensate, so the result of the MCAS trim commands will not be to pitch the nose down, it will be to increase the feedback force on the yoke that the pilot feels.

You have not bothered to respond to my repeated statements along these lines even once.

Further, by sending nose down trim commands to counteract the positive pitch attitude caused by the new engines, MCAS has the effect of providing a feel closer to previous models. But it's not the other way around as you would seem to suggest. It doesn't provide force feedback so the pilot can correctly deal with the positive pitch attitude from the engines using the yoke. If that were the case it wouldn't be messing with the stab.

 

[jim hardy]

Bob Hoover is my hero. Saw his Shrike show in Homestead 1976.

Somewhere i have a photo of his yellow P51 upside down, gear and flaps extended, maybe fifty feet over the shrubs at end of Homestead General's runway 36.
He continued the roll and touched down perfectly.

I guess it just 'felt good' .

sorry for digression - 'old firehorse' instinct...

old jim

 

[OCR]

Done enough troubleshooting to know that the apparent conflicts don't resolve until you arrive at complete understanding.

.

.

...

 

[cyboman]

Bob Hoover is my hero. Saw his Shrike show in Homestead 1976.

Somewhere i have a photo of his yellow P51 upside down, gear and flaps extended, maybe fifty feet over the shrubs at end of Homestead General's runway 36.
He continued the roll and touched down perfectly.

I guess it just 'felt good' .

sorry for digression - 'old firehorse' instinct...

old jim

Haha, well I'm going to have to read about Hoover now.

 

[OCR]

Somewhere i have a photo of his yellow P51 upside down, gear and flaps extended

.

 

[cyboman]

View attachment 240379.

That's one damn fine looking aircraft. I bet it's a blast to fly.
Looks like it's got a nice powerplant:
Powerplant: 1 × Packard V-1650-7 liquid-cooled V-12, with a 2 stage intercooled supercharger, 1,490 hp (1,111 kW) at 3,000 rpm;[118] 1,720 hp (1,280 kW) at WEP

 

[PeterDonis]

I have been engaging with this very assertion over and over. You're basically saying that if the system is acting without any additional input, it will result in the nose pitching down. And saying it's not suppose to operate that way, it's commands for nose down trim are supposed to be overridden by pilot yoke input.

Yes, because the system only operates in manual flight, and the definition of manual flight is that the pilot is supposed to control the plane's pitch attitude with yoke input.

I don't agree, the simple way to refute this is the engines are known to create a positive pitch attitude force.

The engines create a pitch up moment which depends on angle of attack and airspeed, yes. But I don't see how that refutes what I was saying. See below.

the primary function of MCAS is to correct that pitch up effect of the engines

Correct it how? By pitching the nose down? No, that would be ridiculous. If the pilot wants a 10 degree pitch up attitude, the pilot wants a 10 degree pitch up attitude. We're talking about manual flight, remember. It's not the job of MCAS, or any automated system, to decide what the plane's pitch attitude should be. It's the pilot's job. The job of the MCAS is to correct for the pitch up effect of the engines so that the pilot can do his job the same way he did on previous 737s. That's the manufacturer's stated intent.

Or look at it another way. When you say "the pitch up effect of the engines", what does that mean? Does it mean that, if the pilot pulls back on the yoke to hold the plane at 10 degrees pitch up (we're assuming straight and level flight to start with), the engines somehow intervene and move the plane to, say, 15 degrees pitch up instead? No, of course not. The pilot is commanding a particular pitch attitude with the yoke. The difference the engines make is that, in the absence of MCAS, the force the pilot needs to exert to pull back on the yoke to hold a 10 degree pitch up attitude is less than it was on previous 737s. (Note: I've picked the 10 degree number at random, I don't know if it's actually within the range where MCAS will operate. I've just assumed that it is for purposes of this example.) With MCAS operating, the force the pilot needs to exert to pull back on the yoke to hold a 10 degree pitch up attitude is increased, because MCAS adds nose down trim. The same effect could be achieved in the absence of MCAS by the pilot manually adding nose down trim. But either way, the effect of MCAS is not to pitch the nose down to compensate for the effect of the engines, because the effect of the engines was not to pitch the nose up in the first place. The pilot did that.

You think Boeing saying "make it feel like previous 737s" scientifically proves MCAS is primarily an artificial feel system and isn't meant to correct pitch attitude?

I think the manufacturer's stated intended purpose for the system carries a lot more weight than the statements of random people on the Internet, if we're trying to figure out what the intended purpose of the system is.

 

[PeterDonis]

It's a closed loop : pilot - yoke - control surfaces - pitch - back to pilot
and in a closed loop you can observe it anywhere and you'll have a signal coming in and a signal going out

@PeterDonis is i think examining the loop at the node pilot-yoke
@cyboman i think is examining it at the node pitch-back to pilot
MCAS adds a computed signal at the node yoke-control surfaces so pilot receives same pitch to yoke response as in earlier planes and they don't have to re-certify plane or pilot. ..

I think this is an illuminating way to look at it, particularly the point about being able to look at any node in the loop and have a signal coming in and a signal going out. So there isn't really a "start" or "end" of the loop: it's a loop, running continuously.

I would just add a couple of comments:

First, the trim system in general changes the yoke - control surfaces transfer function. (In the 737, it does this by changing the angle of the horizontal stabilizer. In other airplanes it might do it with trim tabs. But it's functionally the same as far as your closed loop flow chart is concerned.) MCAS is just a special case of this. In one of the previous links, it is noted that there are multiple automatic trim adjustment functions on the 737, to compensate for changes in speed and other flight parameters. All of them have the same effect in terms of the flow chart: they change the yoke - control surfaces transfer function.

Second, the loop as you've written it really tracks two separate things: the actual pitch attitude of the plane, and the force the pilot has to exert on the yoke. Both of those responses are changed by the trim system, and the locus of the change in both cases is the yoke - control surfaces transfer function. To make this more explicit, I'll rewrite the loop in two ways, one for the pitch attitude and one for the force:

(Pitch attitude) pilot moves yoke - yoke movement moves control surfaces - control surface movement changes pitch - pilot sees changed pitch attitude

(Yoke force) pilot exerts force on yoke - yoke force moves control surfaces - control surface movement changes feedback force on yoke - pilot feels changed feedback force

Both of these loops are "running" at the same time, and the pilot's actions in the two loops are not independent: either one determines the other. So from the pilot's point of view, he is expecting two kinds of feedback from every yoke action: a change in the plane's pitch attitude, and a change in the feedback force on the yoke. What MCAS, or trim adjustment generally, does is change the relationship between those two pieces of feedback. That relationship is part of what pilots are talking about when they talk about how it "feels" to fly the plane.

 

[cyboman]

The job of the MCAS is to correct for the pitch up effect of the engines so that the pilot can do his job the same way he did on previous 737s.

Completely, agree. I've been pretty clear on that point. I don't think that was your clear position before. It was more aligned with providing stick feel or column force feedback.

I think the manufacturer's stated intended purpose for the system carries a lot more weight than the statements of random people on the Internet, if we're trying to figure out what the intended purpose of the system is.

I would note that we might want to take any statement from Boeing with a grain of salt as they are in a complicated situation. In light of what's happened, any statements regarding their systems are going to be very carefully worded and I'm going to guess their going to keep the details pretty close to their chest as much as they can.

If the primary purpose is to make it feel like earlier versions, that doesn't refute what I'm claiming. I'm claiming how it gets there, and it does that by nose down trim. That effects the forces on the column and the feel, but it effects the aerodynamics of the aircraft first, which however you want to describe it, acts to increase negative pitch attitude. It doesn't provide force feedback for the pilot alone as it's function. The reason it doesn't feel like earlier versions is the engines (mostly). So you can say it's meant to change the feel because of the new engines or you can say it's meant to deal with the new engines because they change the feel. It's the same thing.

I think we've reached the useful limit of our disagreement and it makes sense to not waste any more energy on it. I think we've both made our positions clear. If we come across any additional info that clarifies the primary function of MCAS and other details than maybe we can revisit.

 

[PeterDonis]

shouldn't it be: pilot - yoke - control surfaces - pitch - yoke - back to pilot

No; I think the (valid) issue you are raising here is better addressed by viewing the loop as two concurrent control loops, as described in the post I made just now in response to @jim hardy . One loop deals with the change in pitch attitude, the other deals with the change in feedback force.

 

[PeterDonis]

I don't think that was your clear position before.

It has always been my position, but evidently it wasn't clear to you before. I'm glad it's clear now and that we have agreement on this point.

I would note that we might want to take any statement from Boeing with a grain of salt as they are in a complicated situation.

As far as I can tell, their statement that the purpose of MCAS was to make the plane feel like previous 737s was made when they went to the FAA to get the 737 MAX grandfathered under the certification of previous 737s. That seems like pretty strong evidence to me that that was Boeing's intent, particularly in light of the fact that they did not give pilots any details about MCAS or how it worked. They must have believed that MCAS would, in fact, make the 737 MAX feel similar enough to previous 737s that pilots wouldn't notice, or at least wouldn't have any issues. And the FAA must have agreed, or it wouldn't have approved the certification.

In light of what's happened, any statements regarding their systems are going to be very carefully worded and I'm going to guess their going to keep the details pretty close to their chest as much as they can.

They certainly are now, yes. But I'm looking at what they said about MCAS before there were any incidents; see above. I think what they said then is good information about what the intent of MCAS was.

 

[PeterDonis]

which however you want to describe it, acts to increase negative pitch attitude

I'm sorry, but the fact that you continue to insist on saying "increase negative pitch attitude" is what makes your statement plain wrong, instead of just being a different choice of wording. Your statement is equivalent to claiming that adding nose down trim increases negative pitch attitude. That's just wrong. I've already described why. In fact I did so in the very post (post #170) where I made the statement about what the job of MCAS was that you agreed with.

If you had said "MCAS compensates for the effect of the new engines by adding nose down trim", that would be fine and I would never have had a problem with it. But you cannot make the blanket statement that this increases negative pitch attitude. Adding nose down trim can increase negative pitch attitude, under the circumstances I've already described. But that does not mean it always does increase negative pitch attitude, which is what you are claiming. You cannot just equate "adds nose down trim" with "increases negative pitch attitude". That's wrong.