WHY are two autopilots required for a go-around...

An open discussion of aviation safety related issues.

Moderators: FrankM, el, Dmmoore

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Wed May 04, 2016 7:02 pm

Blah Blah Blah
Blah Blah Blah
Blah Blah Blah
Is the dumbass aggie sitting here reveling in all of you engineers and your big words???....

"Channel", "Signal", "Back drive", "Servo."

Hang on bubba, let's try this:

We'll weld us an some extra tabs on the back of the elevator I think they call them thar things servos and reverse the hot and ground from the autopilot and it'll be sorta like a trim tab...

....what'll be real cool is as it changes how much the elevator bites into the air, it'll move the steerin' wheel back and forth...just like when you jack up the General Lee and grab the front tare and twist it in and out...it'll spin the steerin' wheel.

Of course, if a fella need to do a big pull up to miss bubba's grain bins, when he's a buzzin the home place to impress Bubba's daughter, Daisy Sue that one tab from one autopilot ain't gonna have enough bite...he better have both auto pilots turned on or just do it himself, just like we used to do before our GPS guided tractors if we seen a big rock in the field.

I like this flyin' stuff, pass me a beer.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Thu May 05, 2016 3:08 am

Is the dumbass aggie sitting here reveling in all of you engineers and your big words???....

"Channel", "Signal", "Back drive", "Servo."

Hang on bubba, let's try this:

We'll weld us an some extra tabs on the back of the elevator I think they call them thar things servos and reverse the hot and ground from the autopilot and it'll be sorta like a trim tab...

....what'll be real cool is as it changes how much the elevator bites into the air, it'll move the steerin' wheel back and forth...just like when you jack up the General Lee and grab the front tare and twist it in and out...it'll spin the steerin' wheel.

Of course, if a fella need to do a big pull up to miss bubba's grain bins, when he's a buzzin the home place to impress Bubba's daughter, Daisy Sue that one tab from one autopilot ain't gonna have enough bite...he better have both auto pilots turned on or just do it himself, just like we used to do before our GPS guided tractors if we seen a big rock in the field.

I like this flyin' stuff, pass me a beer.
I am afraid that my English proficiency in nowhere enough to understand that.

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Thu May 05, 2016 6:30 pm

Is the dumbass aggie sitting here reveling in all of you engineers and your big words???....

"Channel", "Signal", "Back drive", "Servo."

[Cowboy red-neck country lingo deleted]
I am afraid that my English proficiency in nowhere enough to understand that.
I THINK I detected that the autopilot works a servo TAB (trim tab if you will).

Thus one tab has less authority than two.

(Therefore, on go-around you really need two tabs giving you your nose-up inputs, otherwise it happens a bit too slow / insufficient.)

(The final stages of a normal, average ILS involves SMALL control inputs...a single autopilot and single tab is more than sufficient, whereas a go-around is a significant pull up)

Like any true trim TAB, if you adjust the TAB and nothing is holding the controls, the yoke will move ("back drive" = the $20 way to say that during your 172 walk around- you move the elevator up and down, the yoke goes in and out (as opposed the in-flight opposite where the yoke moves the elevator, the elevator moves the yoke).

AND SHOUTING- THIS IS ALL ASS HAT, OUTSIDER AGGIE SPECULATION THAT MAY OR MAY NOT BE SOMEWHAT ANALOGOUS TO HOW A 737-834B OPERATES.

I am way over due on Flyboy flaming me for talking out of my ass...which indeed, I do.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Fri May 06, 2016 8:03 am

I THINK I detected that the autopilot works a servo TAB (trim tab if you will).
It doesn't. Then, the rest of the reasoning is flawed.

There is a good reason why NO airplane with hydraulically powered controls have trim tabs, for example: They don't work!!!
Hydraulically powered control surface don't free-float until finding the equilibrium position where the hinge moment is zero, like in you Cessna or my Tomahawk.
They are "rigid" systems that do not respond to external forces: If there is no flow of hydraulic fluid (which is controlled by hydraulic valves), they will not move no matter the external force (unless the external force was such to overcome the hydraulic system, which is never the case otherwise the hydraulic system would not be able to move the control surface against these loads, which is exactly what they are designed and intended to do).
(Therefore, on go-around you really need two tabs giving you your nose-up inputs, otherwise it happens a bit too slow / insufficient.)

(The final stages of a normal, average ILS involves SMALL control inputs...a single autopilot and single tab is more than sufficient, whereas a go-around is a significant pull up)
PP's argument was exactly the opposite: That one single AP channel lacks enough initial* pitch DOWN elevator authority to counteract the pitch up moment created by the underslung engines when they go full coals. (*initial because the elevator trim, or movable stabilizer in the 737, which is also controlled by the AP, will quickly take over, but it will take a couple of seconds for that trim wheel to spin the required number of turns).
Like any true trim TAB, if you adjust the TAB and nothing is holding the controls, the yoke will move ("back drive" = the $20 way to say that during your 172 walk around- you move the elevator up and down, the yoke goes in and out (as opposed the in-flight opposite where the yoke moves the elevator, the elevator moves the yoke).
That doesn't work in the 737 as long as at least one of the hydraulic systems is pressurized, for the reason explained in the beginning: you will not be able to move the elevator with your hands because you have an hydraulic actuator that is full of uncompressible fluid and with its valves closed.

The "back drive" that Flyboy mentioned, which according to him is the same backdrive that PP mentioned, would be something that artificially moves the control columns when the AP moves the elevator, which I presume would be something related with the artificial feel system, yet another thing that is required because the hydraulically powered elevators don't, can't back drive any force (i.e can't provide force and position feedback on the control columns).

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Fri May 06, 2016 8:52 am

Except that this is not the case in the 737 autopilot system, where each and every autopilot channel uses all the hydraulic PCUs that are available (i.e any and each of them gives orders to both crossbars of your planter).
No, it doesn't, at least not in approach mode. The onside autopilot only works the onside servo, which, in turn, works only the onside actuator(s). This is a huge part of Fail-Operational redundancy. "We" wouldn't want bad AP commands going to both servos now, would "we"?
The reason is quite simple: I don't know how this is geared, but the autopilot DOES move the control columnS, which are mechanically linked and they move together, so pushing or pulling one control column moves all control columns, all PCUs, and all elevators out there.
You're close, Gabby, but not quite there. The control columns move not because of the autopilot, but because of the "backdrive" that Bradley referred to (not that you believe him either). In other words, there is a delay involved. Just like there is a delay involved when one side of the elevator commanded by its servo commanded by its AP moves the offside elevator through the torque tube. It's not a huge delay, but it's there.

Hope this helps (I'm not holding by breath though).
I am still trying to understand this...

We are talking here of several mechanisms and devices that are cinematically connected via mechanical links.
Do you believe or know that any delay that can exist in that mechanism would hamper the elevator authority to the point to make it inadequate to effectively control pitch during a go-around.

Does the AP actuates on the same PCUs than the control columns? If yes (what I guess is yes), does the AP actuates on the same input mechanism of the PCUs than the control columns? (I don't know the technical name, but what I mean is the part of the PCUs that receives the control input commands).

If yes, then, because that part of each PCU is mechanically linked to the on-side control column, there is virtually no difference between the AP moving the control column and moving that part of the on-side PCU. The delay that can exist there is the delay of the tension wave through the control cables, which travels at the speed of sound in the cables. The delay would be the same that exists between you moving the control column and the input side of the PCU getting moved by that (the delay in the actual motion of the elevator is a different subject, because it will depend on the flow rate of the hydraulic system through the output side of the PCU, but that is independent of whether the input was manual or by the AP).

(And even less do I understand the delay between the two sides of the elevator when only one side is actuated directly and the other follows trough moved by the torque tube that connects both elevators. Seriously, what can be the delay in a torque tube?)

So IF the above is correct, we have the AP channel on one side actuating on the on-side PCU, but the PCU is connected via the cables to the on-side control column, which in turn is mechanically connected to the off-side control column, which in turn is connected to the off-side PCU via control cables. This chain of mechanisms is not so different from the manual mode, where you also have the control column on one side connected via control cables with the on-side PCU, and mechanically connected to the other control column which in turn is connected to its PCU via cables.

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Fri May 06, 2016 1:38 pm

...A bunch of stuff...
...No doubt, the aggie is wrong in his details...

...BUT...according to Flyboy...

...one autopilot operates one servo apparently on one side of the elevators and the other operates another servo on the other side...

...and for some reason, the one servo is not capable of providing enough 'control input' to do the big pull up as fast as (and/or as much as) it really ought to be done for good operating practice.

Maybe I don't like the logic, but hey, there's compromises all over the place in airplanes. Did I ever tell you about how a 172 can have an elevator jam full up? Got to see it after a practice soft field LANDING and glad I didn't see it after a practice soft field TAKEOFF*. I don't agree with that design either!

*It doesn't totally jam and you can overcome it with a hard shove, but you know...with startle factor and a soft field takeoff putting you in an pretty darn low energy state...
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
flyboy2548m
Posts: 4383
Joined: Sat Feb 02, 2008 12:32 am
Location: Ormond Beach, FL

Re: WHY are two autopilots required for a go-around...

Postby flyboy2548m » Fri May 06, 2016 2:59 pm

I am still trying to understand this...
Try harder, Gabby, it's not that complicated. Honest to goodness, it's not. Or don't try, to use your own words, it's none of my business.
"Lav sinks on 737 Max are too small"

-TeeVee, one of America's finest legal minds.

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Fri May 06, 2016 4:26 pm

...No doubt, the aggie is wrong in his details...
Come on... it's not hair splitting details. Aerodynamic tabs don't work in hydraulically powered controls, the autopilot actuate on the input side of hydraulic controls, and said input side seems to me mechanically interlinked with both control columns and all the relevant hydraulic controls in a continuous cinematic mechanism.
...BUT...according to Flyboy...

...one autopilot operates one servo apparently on one side of the elevators and the other operates another servo on the other side...

...and for some reason, the one servo is not capable of providing enough 'control input' to do the big pull up as fast as (and/or as much as) it really ought to be done for good operating practice.
Okay, that may be the case even if I don't understand why.

Unfortunately, I was not able to find enough information on-line, and the people who seems to know keep doing "obscure" incomplete explanations invoking "spooky" fly-by-cable airplanes and the delay in the transmission of motion in a torque tube. So either they don't know what they are talking about (just like me), or they want to remain mysterious for some reason, or they are making perfectly clear and complete explanations but I am a stupid guy who would not get 2+2=5.

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Fri May 06, 2016 4:51 pm

... it's not hair splitting details...
Ha! Ready to laugh?

Flyboy corrects me..."It's not an actuator, it's a servo."

Dumbass aggie doesn't really know what those $20 words mean.

I go to Wikipedia and read that a "servo" means all sorts of things including tabs and motors and actuators.

https://en.wikipedia.org/wiki/Servo.

Until someone wants to spell this out in simple red-neck language, I'm sticking with my aryloxyphenoxypropanoate herbicides and inhibiting acetyl coenzyme A carboxylase just like cyclohexanediones that only affect monocotyledonous species- it's important to include a crop oil concentrate adjuvant as opposed to a simple nonionic surfactant, and even though they only affect monocots, they can cause solvent burn to sensitive dicotyledonous species...tank mixtures with other dicotyledonous herbicides can result in antagonism of the efficacy, but the whole discussion of whether it's of commercial significance is interesting.

https://www.youtube.com/watch?v=9xhTomq ... LJYQx0EBv8
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
flyboy2548m
Posts: 4383
Joined: Sat Feb 02, 2008 12:32 am
Location: Ormond Beach, FL

Re: WHY are two autopilots required for a go-around...

Postby flyboy2548m » Fri May 06, 2016 5:12 pm

Gabby-boy, remind me again who's the one with confirmation bias around here...
"Lav sinks on 737 Max are too small"

-TeeVee, one of America's finest legal minds.

PurduePilot
Posts: 2130
Joined: Sat Feb 02, 2008 2:02 am

Re: WHY are two autopilots required for a go-around...

Postby PurduePilot » Sat May 07, 2016 5:17 am

Jesus h f***ing Christ...

Fly by cable, when we're talking about big jets, just means "not fly by wire".

You pull on control column. Steel Cable attached to column travels back through fuselage, goes through pulleys and Christmas trees to get split into multiple cables feeding multiple PCU inputs, moves the mechanical hydraulic servo valve on the PCU, and the PCU moves the surface to the commanded positron.

Autopilot servo grabs the cable coming off the control column and moves it whichever way the autoflight function or computer (depending on what we called it on this airplane) to follow the flight path the pilots put in on the mode control panel (MCP). Conveniently, the pilot controls are attached to the same cables, so they move with the autopilot servo input. If the pilot wants to physically overpower the autopilot, he can do that. If both autopilot channel servos are engaged, it takes twice as much force to cause a disconnect since the disconnect is based on displacement.

SO on a fly by cable Boeing, the autopilot servo directly causes the column to move (showing the pilot what Otto is doing to the flight controls). On a fly by wire Boeing, the autoflight function just tells the flight control system computers "pull up this much" and the flight control system does it. But that would not result in the pilot controls moving so there are new components on 777 and 787 called backdrive actuators located beneath the cockpit floor on the bottom of the control column, whose only job is to visually make the pilot controls match the autopilot input to the flight control system.

What were we talking about again?

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Sat May 07, 2016 12:00 pm

Jesus h f***ing Christ...

What were we talking about again?
What are we talking about? How NOT to answer the question.

WHY IN THE HELL IS YOUR ONE SERVO TOO WEAK TO DO A PULL UP?

THIS SEEMINGLY MAKES NO SENSE!

You just ruined everything by saying it's upstream from the hydraulics which is where "the muscle" is.

A weak human is not strong enough to move the actual elevator so he only has to provide a 'modest force' to the control cables for a go around.

BUT YOU GUYS CHOSE TO USE A WIMPY SERVO THAT IS TOO WEAK TO DO WHAT THE HUMAN CAN DO?

...SO LIKE... WHY?

I guess it could be some messed up math that one sevo that is strong enough to go around equals two servos too strong for a pilot to overcome - but I'd hate for someone to have to say such a simple statement.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
flyboy2548m
Posts: 4383
Joined: Sat Feb 02, 2008 12:32 am
Location: Ormond Beach, FL

Re: WHY are two autopilots required for a go-around...

Postby flyboy2548m » Sat May 07, 2016 2:22 pm


What were we talking about again?
Nothing special, Bradley, Gabby is just talking in circles. About a position he doesn't have. Totally free of confirmation bias.
"Lav sinks on 737 Max are too small"

-TeeVee, one of America's finest legal minds.

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Sun May 08, 2016 10:45 am


What were we talking about again?
Nothing special, Bradley, Gabby is just talking in circles. About a position he doesn't have. Totally free of confirmation bias.
Ok, Flyboy. After PP's last post, I have a stronger opinion. I will not call it a position yet, but it's getting closer and closer.

Whatever the reason is for Boeing to have designed the AP in single channel to be ]ok flying an approach in solid IMC down to 200ft (proceduraly) or all the way down to a no-flare hard landing (technically), but not ok to fly a go-around below 2000ft, it is NOT because the single channel autopilot would lack any pitch authority that the dual channel AP or the human pilot would have.

Even an aggie would note it.

Am I sure of the above? No, but if you want to say that I have a position, make it that one.

User avatar
flyboy2548m
Posts: 4383
Joined: Sat Feb 02, 2008 12:32 am
Location: Ormond Beach, FL

Re: WHY are two autopilots required for a go-around...

Postby flyboy2548m » Sun May 08, 2016 1:56 pm


Ok, Flyboy. After PP's last post, I have a stronger opinion. I will not call it a position yet, but it's getting closer and closer.

Whatever the reason is for Boeing to have designed the AP in single channel to be ]ok flying an approach in solid IMC down to 200ft (proceduraly) or all the way down to a no-flare hard landing (technically), but not ok to fly a go-around below 2000ft, it is NOT because the single channel autopilot would lack any pitch authority that the dual channel AP or the human pilot would have, and nobody and nothing will convince me otherwise
Fixed.
"Lav sinks on 737 Max are too small"

-TeeVee, one of America's finest legal minds.

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Sun May 08, 2016 4:38 pm


Ok, Flyboy. After PP's last post, I have a stronger opinion. I will not call it a position yet, but it's getting closer and closer.

Whatever the reason is for Boeing to have designed the AP in single channel to be ]ok flying an approach in solid IMC down to 200ft (proceduraly) or all the way down to a no-flare hard landing (technically), but not ok to fly a go-around below 2000ft, it is NOT because the single channel autopilot would lack any pitch authority that the dual channel AP or the human pilot would have, and nobody and nothing, including relatively clear statements by a knowledgeable Airbus pilot and a young 787 engineer, will convince me otherwise.
Fixed.
More fixed.

And, sinice you are a master at word games, do I need to start a new thread that asks, "WHY was the 737 autopilot system DESIGNED so that one of them is incapable of executing a 'proper' go around below 2000 feet? :mrgreen:
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
Gabriel
Posts: 3663
Joined: Sat Feb 09, 2008 2:55 am
Location: Buenos Aires, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Gabriel » Sun May 08, 2016 6:56 pm


Ok, Flyboy. After PP's last post, I have a stronger opinion. I will not call it a position yet, but it's getting closer and closer.

Whatever the reason is for Boeing to have designed the AP in single channel to be ]ok flying an approach in solid IMC down to 200ft (proceduraly) or all the way down to a no-flare hard landing (technically), but not ok to fly a go-around below 2000ft, it is NOT because the single channel autopilot would lack any pitch authority that the dual channel AP or the human pilot would have, and nobody and nothing will convince me otherwise
Fixed.
On the contrary: It is the manufacturer's employee who is convincing me. Have you read his post?

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Sun May 08, 2016 10:17 pm

Evan, there, shows a diagram, of the autopilot having thingies back in the hydraulic muscle thing next to the elevator.

Nice to know, but why the system was designed to be too weak or slow to pull up from just one side remains.

Also, should one side of the regular control system fail, would the pull up authority then be 'inadequate' from the yoke?

I had a John Deere 6410 drive me nuts because it had a lose connection between the electrical hitch lift control and the actual control. The older Case IH 685 and 695 had manual levers controlling the hitch hydraulics.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

Evan_37
Posts: 8
Joined: Wed May 04, 2016 9:55 pm

Re: WHY are two autopilots required for a go-around...

Postby Evan_37 » Sun May 08, 2016 10:40 pm

Jesus h f***ing Christ...

Fly by cable, when we're talking about big jets, just means "not fly by wire".

You pull on control column. Steel Cable attached to column travels back through fuselage, goes through pulleys and Christmas trees to get split into multiple cables feeding multiple PCU inputs, moves the mechanical hydraulic servo valve on the PCU, and the PCU moves the surface to the commanded positron.

Autopilot servo grabs the cable coming off the control column and moves it whichever way the autoflight function or computer (depending on what we called it on this airplane) to follow the flight path the pilots put in on the mode control panel (MCP). Conveniently, the pilot controls are attached to the same cables, so they move with the autopilot servo input. If the pilot wants to physically overpower the autopilot, he can do that. If both autopilot channel servos are engaged, it takes twice as much force to cause a disconnect since the disconnect is based on displacement.

SO on a fly by cable Boeing, the autopilot servo directly causes the column to move (showing the pilot what Otto is doing to the flight controls). On a fly by wire Boeing, the autoflight function just tells the flight control system computers "pull up this much" and the flight control system does it. But that would not result in the pilot controls moving so there are new components on 777 and 787 called backdrive actuators located beneath the cockpit floor on the bottom of the control column, whose only job is to visually make the pilot controls match the autopilot input to the flight control system.

What were we talking about again?
I don't know what you're talking about but it's not the 737. The 737 autopilot does not "grab the cable coming off the control column", nor do the autopilot servos "directly cause the column to move".

In pitch (elevator), the autopilot commands are DIRECTLY controlling the elevator PCU's via solenoids and transfer valves. The solenoids complete hydraulic power to the transfer valves in the power control units. The transfer valves convert the autopilot electrical signals into hydraulic flow. They consist of a torque motor which moves either a flapper valve or a jet pipe assembly which regulates the hydraulic flow to the PCU output piston. Autopilot does not involve elevator cable inputs except for backdrive.

What you might find interesting is that the 737 under autopilot CWS mode is essentially a crude FBW system. When a pilot applies ample force to the column (above the 'deadzone' detent) while on autopilot, a transducer located below the cockpit on the front control quadrant of the control column sends the manual pitch input signals to the autopilot and the autopilot adjusts the pitch singal to the PCU's; they execute the new command and then maintain that pitch. If you think about it, this is a bit like the A320 sidestick-transducer-FCC-actuator path in manual flight. So, in manual flight on the 737, you have a cable-controlled airplane and in autopilot you have a transducer, wires and solenoid controlled airplane.

Now, none of that affects the single autopilot's ability to achieve full elevator deflection. The systems are designed so that each autopilot is dedicated and dependent on its corresponding hydraulic system. The loss of either a single autopilot or a single hydraulic system will not inhibit the elevator performance. (the loss of both hydraulics with revert to mechanical control, and this is what the balance tabs are for).

A combination of thrust-pitch coupling and nose-up stabilizer trim MIGHT exceed elevator authority under performance-reserve thrust, at least momentarily, but a second autopilot/PCU will not help you here: this is due to the limitations of the control surfaces, not the force applied to them. This is also why you don't slam the thrust levers to full TO/GA unless there is a very good reason. SOP is 85% for routine go-arounds and the go-button single-push will give you about the same reduced go-around thrust.

Bottom line: the reason Boeing restricts single-autopilot operations in critical phases of flight (under 50' RA, transitioning to go-around up to 2000' RA) is fail-passive redundancy, period. Gabriel is correct. It has nothing to do with any elevator cable-grabbing autopilot weakness nonsense.

I just wanted to set that straight.

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Sun May 08, 2016 10:54 pm

...Boeing restricts single-autopilot operations in critical phases of flight (under 50' RA, transitioning to go-around up to 2000' RA) is fail-passive redundancy, period...
2000 feet seems excessive...an order of magnitude or two excessive- The "why" question remains.

Is that the magical altitude where you might successfully recover from a a stall?
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
Not_Karl
Previously banned for not socially distancing
Posts: 4128
Joined: Thu Nov 12, 2009 6:12 pm
Location: Bona Nitogena y otra gaso, Argentina

Re: WHY are two autopilots required for a go-around...

Postby Not_Karl » Sun May 08, 2016 10:55 pm

Evan, there, shows a diagram, of the autopilot having thingies back in the hydraulic muscle thing next to the elevator.
First image is the basic layout of the 737 elevator drive. Second image is PCU installation and the autopilot servo location (NG has two input rods, this one is from a classic). Third image shows the hydraulic and mechanical schematic of the elevator PCU and the autopilot engagement mechanism.

As you can see, the autopilot does not drive the cables, only the servos.
737_elev_1.jpg
737_elev_1.jpg (472.04 KiB) Viewed 6942 times
737_elev_2.jpg
737_elev_2.jpg (304.65 KiB) Viewed 6942 times
737_elev_3.jpg
737_elev_3.jpg (757.8 KiB) Viewed 6942 times
I, from my complete ignorance and asshatness, am still blaming Verbie.
International Ban ALL Aeroplanies Association, founder and president.

"I think, based on the types of aircraft listed, you're pretty much guaranteed a fiery death."
- Contemporary Poet flyboy2548m to a Foffie.

Evan_37
Posts: 8
Joined: Wed May 04, 2016 9:55 pm

Re: WHY are two autopilots required for a go-around...

Postby Evan_37 » Sun May 08, 2016 11:36 pm

...Boeing restricts single-autopilot operations in critical phases of flight (under 50' RA, transitioning to go-around up to 2000' RA) is fail-passive redundancy, period...
2000 feet seems excessive...an order of magnitude or two excessive- The "why" question remains.

Is that the magical altitude where you might successfully recover from a a stall?
Why 2000' RA? I think they just originally set the requirement for the entire operating range of GA autopilot mode. Until a software upgrade in 2001, it wasn't possible to engage GA mode in autopilot above 2000' RA (the 2001 upgrade requires GS capture or flaps out for engagement above 2000' RA). But remember, if the go buttons are pushed during a SINGLE autopilot approach, the A/T (if engaged or armed) will still function in TO/GA and the FD's will still provide GA guidance. I'm sure Boeing felt that was enough blend of automation with a safe provision (aka. the human pilots) for the lack of FD redundancy. FYI: You also cannot engage both autopilots below 800' RA. Why is that? Boeing works in mysterious ways...

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Mon May 09, 2016 1:37 am

...Until a software upgrade in 2001, it wasn't possible to engage GA mode in autopilot above 2000' RA...
That's meaningful.

Seems like deep down, maybe they almost don't like one autopilot doing any go arounds.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.

User avatar
Ancient Mariner
Posts: 3774
Joined: Sun Feb 03, 2008 5:24 pm

Re: WHY are two autopilots required for a go-around...

Postby Ancient Mariner » Mon May 09, 2016 9:54 am

Evan_37:
Bottom line: the reason Boeing restricts single-autopilot operations in critical phases of flight (under 50' RA, transitioning to go-around up to 2000' RA) is fail-passive redundancy, period. Gabriel is correct. It has nothing to do with any elevator cable-grabbing autopilot weakness nonsense.
Thank you, confirms my suspicion and gels with my maritime experience. Except number of axis, same, same.
See, how difficult could it be to answer such a simple question. :mrgreen:
Per

User avatar
3WE
Posts: 8141
Joined: Wed Feb 06, 2008 2:37 pm
Location: Flyover, America

Re: WHY are two autopilots required for a go-around...

Postby 3WE » Mon May 09, 2016 1:27 pm

...because of passive fail only...
Then again, a passive fail at 150 feet on a somewhat critical ILS is ok, but a passive fail at 1500 feet for a fat and dumb pull up is ok?...

Seems like we're back at the beginning.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.


Return to “Aviation Safety Discussion Forum”

Who is online

Users browsing this forum: No registered users and 7 guests