FCeng84

Gums - glad to share the level of system understanding that I can. STS and MCAS are both functions implemented within the autopilot and thus use the same path to the stabilizer that the autopilot uses for autotrim when it is engaged and controlling the elevator. In the logic MCAS has priority over STS which to me makes perfect sense as MCAS is there to address less than desirable Cm-alpha characteristics while STS is there to address less than desirable pitching moment vs. speed (would that be Cm-vcas or Cm-u?) characteristics. STS is more of a phugoid thing while MCAS is more related to short period. In my opinion, short period stability characteristics are more important than phugoid stability characteristics. Things can get out of sorts much quicker when related to short period than phugoid.

gums

Salute Vess!

Sounds like the L39 has only a coolie hat switch on top of the stick, heh ?
++++++++++++++++++++++++++++++++++++++=
I must admit that I only flew one jet that "trimmed" all by itself, and that counts 4 others that went the gauntlet from pure cables, pushrods and pulleys, thru hydrualic assisted control surfaces, irreversible hydraulic systems and then a full FBW system with zero mechanical connections to control surfaces. The auto trim was in the VooDoo, which had a most cosmic A/P. Only trim from HAL was when I selected A/P for attitude control, speed, heading, altitude etc. I also had a A/P coupler that flew the attack scenario and a ILS approach without throttle control. Least I could crash if I didn't pay attention to power, heh heh. But that would be my fault.
++++++++++++++++++++++++++++++++
First I saw about auto trim from Hal for the next generation lights was back in the 70's when the Viper FLCS was being discussed in our small community. The non-aero engineer pilots thot the thing had auto-trim. Nope. It trimmed to the commanded gee, and still does 40 years later. We did not have to trim if we slowed down or speeded up. Roll was slightly different due to potemtial rigging of the ailerons and speed could force roll in one direction or the other. This was negligible most of the time because the control law was zero roll rate or whatever you had set.

Even the 'bus is not a pure auto trim, but it's the closest I can find for most flight parameters if auto-throttle is engaged..

I would like to see an official flow chart from Boeing that shows the interaction of STS and MCAS. Inputs and outputs to the stab and cockpit displays when things were awry.

Gums sends...

gums

Salute FC !

Oh great, and that answers my basic question anout priority.

I also understand the phugoid at normal AoA and Aoa changes, but seems the STS implementation is rather harsh, heh heh. I have bad vibes about pilots that do not recognize a need to trim. But if they wanna hold a few pounds of pressure for 3 hours in one of the heavies, that is their call.

Gums sends...

Vessbot

Salute! Yes except the coolie hat part of it on the one I flew was missing, so there was just a little 4-way metal nub that hurt your thumb! But that's a different jet than what I was talking about, (my profile's out of date) the CRJ. There are only electric inputs, and the hierarchy goes something like:

Captain's manual switches
FO's manual switches
Autopilot inputs
"Autotrim" (vague naming award of the year... this moves when the flaps are moving in the first few notches)
Mach trim

Are there any other secret cooks in the kitchen? Who knows!

(Come to think of it, I think the L-39 autotrimmed with flap movement too... but it's been a while)

megan

Sorry gums, a habit I should get out of.

jimtx

I'm confused. Humans have to be protected by MCAS and STS but when the autopilot is used it doesn't care what the stick forces are? It just flys attitude and airspeed and whatever other constraints are input and is quick enough to see an unwanted pitchup that was caused by the Max engines or an unwanted regime that the STS would protect a Human against, whatever that was. I suspect that Gums has already answered the question of whether he would need that help but of course history does say that some of his cohorts failed in that respect. You still haven't answered my question about why steep turns were referenced in the Boeing/FAA literature.

FCeng84

jimtx

FCeng84

Answers to questions from jimtx:

1. Would MCAS activate in a 250 knot 45 degree bank turn? Would it activate in a slightly over bank 35 degree turn at a lower airspeed?
- MCAS activates when AOA exceeds a threshold that is a function of speed/Mach number. MCAS activation is not dependent on bank angle.
- The 737 can be maneuvered to any bank angle at with AOA either above or below the MCAS activation threshold so it is not possible to determine whether or not MCAS would activate at either of the conditions proposed as there is no indication as to the associated AOA or the corresponding normal load factor.

2. Can the autopilot be put in the MCAS envelop where it is not protected?
- At some flight conditions it is possible to command the airplane when via the autopilot to an AOA high enough to put it into a condition where MCAS would activate if the autopilot were to be disconnected. MCAS will not come active with the autopilot engaged.

I hope these are helpful. As always, follow-ups welcome.

gums

Salute FC !

Yeah, AoA seems to be the driver for the pulses, and mach is a function that fades out the MCAS what? a) duration of pulse, b) rate of stab movement, c) amount of stab movement ( which necessarily appears a function of "a" and "b") or all of the above?
The available literature says the mach will limit MCAS cmds according to "some" function/equation until what? 0.68 M ? And then no more MCAS above that, if I read it correctly.
So AoA must not be as big of a factor at higher altitudes and normal cruise mach, huh?
Point being that "q" must be important, and Machinbird's question about linearity of the plane's static stability coefficients when near certain AoA and "q" values must be a player for certification.

Oh well, all very technical, but knowing what the plane is supposed to do and what the "help" provided by Hal is supposed to do is a good thing.

Gums asks...

infrequentflyer789

@FCeng84 I get most of what you've said, makes sense and moreover it matches my understanding of the high-AOA function in the NG, but this bit I don't get.

If I understand the NG functional descriptions that I have acquired, Autopilot and STS take the same signal path which goes through the column switch module and then the console cutout switches, to the actuator. Now, we know MCAS doesn't honor the fwd/aft-column-cutout switches so there must, surely, be either a different signal path or some new bypass/override signal to the column switch module (unless MAX STS doesn't honor the fwd/aft-column switches either?) ?

Unless the column switch module has changed from an actual switch in the path (which is how I understand the NG docs - I may be wrong?) to merely a signal of switch position back to the FCC which then decides in software whether that "cutout switch" is effectively in the signal path or not?

Note: the above, would be a sort of red flag to me, because software is far more flexible than switches and relays, which is great until it goes wrong, by accident or design - it is very easy to change code such that a check/gate/switch point is unintentionally bypassed, and then Murphy decrees it will only be for a set of circumstances/inputs that aren't in the tests. Generally much harder to unintentionally bypass a relay or switch.

The "flaps up" signal from FCC to stab trim actuator must have changed in some way as well, because MCAS is apparently trimming at flaps-down speed when flaps are up?

Appreciate any clarification you can give, also fully understand if it goes beyond the system understanding that you can share - in which case I will remain with my speculation until more is published.

PS: If anyone wants me to post the NG details or diagrams that I am talking about then I can, I didn't want to make this post too large. Plus I get the impression FCeng84 doesn't need them!

FCeng84

The amount of airplane nose down stabilizer that MCAS commands is a function of both Mach number and the amount that AOA has exceeded the MCAS activation threshold (itself a function of speed/Mach). When MCAS commands the stab, it always moves at the MCAS rate of 0.27 deg/sec. The duration of the MCAS stab command is how ever long it takes to move the stabilizer at the MCAS rate the amount of incremental stabilizer commanded by MCAS.

FCeng84

InfrequentFlyer - When MCAS was introduced to address less than desirable Cm-alpha characteristics it was recognized that it would need to be able to apply airplane nose down stabilizer motion while the column was being pulled past its cutout switch activation point. I do not know all of the details, however I do know that some wiring was added to allow the MCAS command to get to the stabilizer even with a large column pull. This "path around the column cutout switch" is not activated for STS commands, rather only for MCAS commands. Note as stated earlier that MCAS commands take priority over STS commands.
MCAS is active only when flaps are up and its commands are a function of AOA regardless of speed. This allows MCAS to be active for flaps up stalls as well as higher speed high AOA maneuvers such as wind up turns.

jimtx

"such as wind up turns" Obviously no airline crew is going to do a windup turn, a test flight procedure, hopefully But they do or used to do steep turns in the simulator which should be a data point on the windup turn flight test info. Never have heard of anyone doing a real (on purpose) steep turn in airline flying. Can you relate that data point to a hazardous AOA and other envelope positions that would make Boeing and the FAA decide to protect the human from it's effects, which most probably they would never see in the airplane and only see it in the simulator if anyone still does steep turns in the simulator. Yet, the autopilot would not need any MCAS help. It would just move the stick back or forward, irrelevant of stick force, to command an attitude that complied with what it wanted the performance instruments to show. But we still have to deal with an autopilot that can be commanded to come close to or enter the straight and level MCAS envelope. MCAS won't help it but the autopilot will move the stick as required, uncaring what the stick force is, to command the required control inputs. But put a human in the loop and we have to help him. What about telling him what to be aware of as @Gums voodoo folks used to do?

infrequentflyer789

Wonder how many people that phrase has killed in the history of aviation engineering.

For instance, obviously no airline crew is going to conduct a high AOA test at lower-than-specified level and go beyond the failure point... (XL888 - AOA sensor fail again).

Whenever there is human in the loop and someone says "obviously" I envisage Mr Murphy materialising dressed as the Joker (Heath Ledger style), and in aviation, strapping himself into the jumpseat.

@FCeng84 - thanks, clarification understood and appreciated.

FCeng84

jimtx - I fully agree with you that wind-up turns are not a normal maneuver. They are fun to experience during flight testing, but I hope no fare paying passenger on a regular transport flight has to ride through one. Wind-up turns are the test maneuver most commonly used to collect data to demonstrate compliance with stick force vs. g requirements as specified in FAR 25. The overall objective is that increases in nose up controller forces as applied by the pilot will be required in order to command higher load factors (i.e., higher AOA at a given speed). This is considered necessary to provide the pilot with an airplane that enables control of load factor at elevated levels if the pilot chooses to command the airplane there and similarly that the airplane will promptly recover to more normal load factors if that is the intent of the flight crew.

In addition to wind-up turns, flight testing also often includes roller coaster maneuvers (wings level pull-ups and push-overs in succession). These are probably more relevant to line operation maneuvers such as vertical maneuvers to change climb/descent angle, avoidance maneuvers prompted by see-and-avoid or TCAS, and tight path control during an emergency descent. Those maneuvers are more dynamic than wind-up turns so the data collected during those handling qualities evaluations does not lend itself very well to showing compliance with the FAR force vs. maneuver requirements. For that compliance, wind-up turn data has been the standard.

CONSO

SLF here- looking at various articles/posts/analysis here and in aviation week, etc, and having spent many decades in aerospace- missile device testing - manufacturing /tooling of military and commercial aircraft, etc as a injun-ear - MY view of the MCAS ( Hal ) system fiasco boils down to the following simplified items

1) AOA mismatch notice/display an optional item- no doubt at a ridiculous cost- so pilots had no notice of that discrepancy
2) AOA single input with no matching- voting allowed to directly affect/move a critical flight control item ( stabilizer )
3) Pilots of 737 prior to Max ‘knew” that push pull of control column in opposition to autopilot would disengage autopilot and allow manual- direct control until autopilot manually reset/engaged
4) description of runaway stabilizer infers continuous movement despite column disconnect or no trim switch.- thus a positive action tripping circuit breaker is needed to stop
5) But MCAS- HAL on 737 MAX ignores both column and trim switch use- and moves stabilizer and then STOPS briefly- ALL dependent on ONE AOA value- with NO notice, NO display, and NO warning
6) Neither instruction manual or training mentions the absolute authority of MCAS- HAL- so pilots believing and trained that pull or push on control column, and/or disconnecting Autopilot gives them ABSOLUTE manual control at ALL times and that trim switch still works( briefly)
7) Pilots are now faced with a WTF conundrum. They are going too fast to lower flaps which stops MCAS-HAL and at a low altitude per training - “runaway” stabilizer seems to stop for a while- and autopilot is off so ????

8) Meanwhile stick shaker rattles, trim wheel starts and stops and ????

And only a few minutes to solve . . . " I'm sorry dave . . "

FCeng84

CONSO - I appreciate your summary list and agree with much of what you have presented. Below I have added my thoughts and a few points of clarification using italic text. The non-italic text below are the summary statements provided by CONSO.

1) AOA mismatch notice/display an optional item- no doubt at a ridiculous cost- so pilots had no notice of that discrepancy
- Agreed - neither did the engineers.

2) AOA single input with no matching- voting allowed to directly affect/move a critical flight control item ( stabilizer )
- Yes. Analysis determined that one increment of MCAS stabilizer motion by itself did not pose a safety issue. No second increment without pilot trim input. Design approach assumed pilot trim input would return airplane to trim prior to any subsequent MCAS activation.

3) Pilots of 737 prior to Max ‘knew” that push pull of control column in opposition to autopilot would disengage autopilot and allow manual- direct control until autopilot manually reset/engaged
- Push / pull on column does disengage autopilot (if engaged) and at all times provides direct control of elevator. Same for all 737 models.
- Autopilot box, however, continues to provide automatic stabilizer control even when autopilot function is disengaged:
- STS on earlier 737 models
- STS and MCAS on 737 Max

4) description of runaway stabilizer infers continuous movement despite column disconnect or no trim switch.- thus a positive action tripping circuit breaker is needed to stop
- Agreed from wording I have seen here in PPRUNE.

5) But MCAS- HAL on 737 MAX ignores both column and trim switch use- and moves stabilizer and then STOPS briefly- ALL dependent on ONE AOA value- with NO notice, NO display, and NO warning
- 737 Max does not ignore column or pilot trim switch. Column is sole control driver for elevator. Pilot trim has priority over automatic control of stabilizer. One difference is that pull column disables automatic airplane nose down STS command while pull column does not disable airplane nose down MCAS command.
- All 737 stabilizer motion involves turning of cockpit stabilizer trim wheels (one at each pilot's inboard knee) with ten revolutions of approximately 8" diameter, one inch wide wheels per degree of stabilizer motion. Wheel turn produces considerable noise as well (ranted not as significant when stick shaker is going.)

6) Neither instruction manual or training mentions the absolute authority of MCAS- HAL- so pilots believing and trained that pull or push on control column, and/or disconnecting Autopilot gives them ABSOLUTE manual control at ALL times and that trim switch still works( briefly)
- I cannot comment on instruction manual or training.
- No argument with this comment other than to restate that pilot stabilizer trim input via wheel mounted thumb switch always stops and overrides automatic stabilizer control whether coming from STS or MCAS functions.

7) Pilots are now faced with a WTF conundrum. They are going too fast to lower flaps which stops MCAS-HAL and at a low altitude per training - “runaway” stabilizer seems to stop for a while- and autopilot is off so ????
- Agreed. Design assumed that pilots would recognize improper stabilizer motion taking them repeatedly away from trim when flying relatively steady condition and take the action of cutting out stabilizer motors. I cannot speak to the associated manual nor training.

8) Meanwhile stick shaker rattles, trim wheel starts and stops and ????
- Agreed - quite the compound mess. Don't forget that they are also seeing air data miscompares (speed and altitude) and havd associated warning lights to potentially take their attention away from improper automatic stabilizer motion.

And only a few minutes to solve . . . " I'm sorry dave . . "
- Must continue to fly the airplane vertical/pitch axis with column and use control wheel mounted stabilizer trim switches to offload column forces while sorting out the situation. HAL never said "I'm sorry Dave", but HAL did continue to give errant nose down stabilizer increments that seem to have eventually overwhelmed Dave.

Let me add that with this and all of my entries on this topic I am not trying to place blame for this tragic accident on any one party. I am merely trying to shed some light on how the system involved functions so that all can better understand what this crew faced.

May all of those 189 souls rest in peace and everyone they left behind find the support and comfort that they need to face life after this horrific loss.

Respectfully submitted - FCEng84

CONSO

......
Appreciate your explanations- like you not trying to blame any ONE.. but perhaps the ' committee ' approach needs revision to incorporate " murphy "

And correct me if the combination of mCAS, stick shaker, control column inputs ALSO increases the control force ( FEEL ) to some max limit. possibly requiring both hands (and planted feet ) to pull back but while in a dive everone becomes nearly weightless held only by harness ?

jimtx

So FAR part 25 has requirements for the ac aerodynamics (?) to provide the pilot with an airplane that enables him to control load factors with a linear stick pull and not have the the surprise of the ac continuing to pull load factor when the pilot is not requesting it? But FAR25 does not care about the actual aerodynamics, in the 737Max, as the autopilot is not protected from the stick force non linearity, rightfully so, as it doesn't care about sick force? So the ac's aerodynamics do not have to be adjusted. The only adjustment needed is to fool the pilot that he has to pull harder to get more g or AOA? And, if the pilot loses this MCAS, Boeing doesn't even tell him to be careful in any flight regime as he will not encounter it in normal operations. Although you alluded to some escape maneuvers that might approach the MCAS envelope. But the current abnormal runaway trim procedure, where MCAS will be disabled, that Boeing advertises as sufficient, does not caution about being careful in a TCAS maneuver or an escape maneuver. Because a Boeing exec said that pilots do not have to know about the nuances of the ac. But of course, nobody is going to get runaway trim and then get a TCAS alert or an escape situation on the same day.