News:



  • April 19, 2024, 12:08:09 AM

Login with username, password and session length

Author Topic: 737 Flight Control  (Read 2331 times)

Offline Peter in Fairfax, VA

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1108
737 Flight Control
« on: April 04, 2019, 12:32:46 PM »
Here's a link to the Ethiopian 737 crash analysis preliminary report.

http://peterklosky.com/rpt.pdf


Offline Peter in Fairfax, VA

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1108
Re: 737 Flight Control
« Reply #1 on: April 04, 2019, 12:47:22 PM »
Interesting report.  Appears that, yes, the pilots did know to operate the trim cutout, subsequently taking this action, then acknowledging this to one another in a professional manner, as evidenced by the voice recorder and the flight data recorder.  From my reading of the preliminary report, looks like there is an unexplained issue.

Question:  Since the left flight control obviously bad, with unexpected left stick shaker, etc., why not change the pilot flying to the FO in this scenario?

Offline Perry Rose

  • Go vote, it's so easy dead people do it all the time.
  • 2015
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1662
Re: 737 Flight Control
« Reply #2 on: April 04, 2019, 01:47:50 PM »
Interesting report.  Appears that, yes, the pilots did know to operate the trim cutout, subsequently taking this action, then acknowledging this to one another in a professional manner, as evidenced by the voice recorder and the flight data recorder.  From my reading of the preliminary report, looks like there is an unexplained issue.

Question:  Since the left flight control obviously bad, with unexpected left stick shaker, etc., why not change the pilot flying to the FO in this scenario?

Because I'm the Captain and you're not. It's quite common.
I may be wrong but I doubt it.
I wouldn't take her to a dog fight even if she had a chance to win.
The worst part of growing old is remembering when you were young.

Offline Paul Wood

  • 21 supporter
  • Trade Count: (0)
  • Commander
  • *
  • Posts: 297
Re: 737 Flight Control
« Reply #3 on: April 04, 2019, 03:57:45 PM »
Because I'm the Captain and you're not. It's quite common.

Exactly right! I trained quite a few foreign air crews in my day as a B737 instructor/FAA examiner. It's unbelievable but true.

Paul

Offline Dave Hull

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1908
Re: 737 Flight Control
« Reply #4 on: April 04, 2019, 05:39:02 PM »
Take a look at the first officer's flight hours, and that attitude may not seem so outrageous--he had 361, total. If you are struggling to understand if your settings are wrong or if the plane is broken, I want the experience. On the other hand, if you know that one set of controls is broken and the other is working fine (as postulated by Perry), then it would be easy to decide to hand off the plane. From the preliminary report, that was not clearly the case.

There would appear to be a lot left to explain about how the aircraft performed.

Dave

Offline Howard Rush

  • 22 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 7811
Re: 737 Flight Control
« Reply #5 on: April 04, 2019, 08:42:40 PM »
It woudn't have mattered.  The controls are mechanically tied together. 
The Jive Combat Team
Making combat and stunt great again

Offline Perry Rose

  • Go vote, it's so easy dead people do it all the time.
  • 2015
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1662
Re: 737 Flight Control
« Reply #6 on: April 05, 2019, 05:49:59 AM »
It seems to me that no one was flying the plane. They expected the computers to do it. I was taught, avigate, navigate, communicate. In other words above all fly the plane.
I may be wrong but I doubt it.
I wouldn't take her to a dog fight even if she had a chance to win.
The worst part of growing old is remembering when you were young.

Offline Peter in Fairfax, VA

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1108
Re: 737 Flight Control
« Reply #7 on: April 05, 2019, 06:09:32 AM »
Looking at the flight data recorder, the left alpha vane data diverged, but the right alpha vane was OK.  There was no way to switch to reading the right alpha vane?

Offline Howard Rush

  • 22 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 7811
Re: 737 Flight Control
« Reply #8 on: April 05, 2019, 04:47:15 PM »
There was no way to switch to reading the right alpha vane?

It could be switched automatically if there was a way to know that the right one was giving the correct signal.  Usually, when there are two inputs, the system using them would turn itself off if the two disagree. 
The Jive Combat Team
Making combat and stunt great again

Offline Steve Fitton

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2272
Re: 737 Flight Control
« Reply #9 on: April 05, 2019, 04:52:22 PM »
It could be switched automatically if there was a way to know that the right one was giving the correct signal.  Usually, when there are two inputs, the system using them would turn itself off if the two disagree.

Doesn't MCAS use one or the other AOA sensor, automatically switching from one to the other each flight?  So there is no actual comparing of values in flight.
Steve

Offline Howard Rush

  • 22 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 7811
Re: 737 Flight Control
« Reply #10 on: April 05, 2019, 05:02:01 PM »
Doesn't MCAS use one or the other AOA sensor, automatically switching from one to the other each flight?  So there is no actual comparing of values in flight.


That's what I read.  I can't imagine why they'd do that. 
The Jive Combat Team
Making combat and stunt great again

Offline Dave Hull

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1908
Re: 737 Flight Control
« Reply #11 on: April 05, 2019, 05:15:07 PM »
If you want real fault tolerance with some AI, you need three voting members. For example, the F/A-18 has triple-redundant flight computers. Two are constantly in use. When they disagree, the third computer is polled to see which of the two is in error. If no two of the three can agree, you punch out.

The way to achieve high reliability in complex systems is using redundancy and cross-strapping. That should get you to triple-nines.

If you only have two AOA vane sensors, and they exceed a disagreement threshold, your only choice is to disable anything that uses them and revert to a backup system. I believe this is what Howard is telling us.

Reliability via cross-strapping is an interesting and sometime tricky business. For true redundancy, they cannot depend on common path/common components. This sometimes is not obvious, especially when power forms are involved. For example, if a power fault is detected on one bus and load-shedding occurs, it might disable a sensor being used on a supposedly independent system. It takes a lot of very careful work to achieve real redundancy and fault tolerance.

We will see in the final report whether the Boeing systems that depending on the AOA sensors could be easily disabled by the pilot or not, and if they could be, whether the backup or manual systems were adequate. Many of the reports so far don't make sense, but publishing data directly out of the DFDR gives insight that at least isn't reprocessed by non-experts.....

Offline Steve Thomas

  • Trade Count: (0)
  • Commander
  • ****
  • Posts: 373
Re: 737 Flight Control
« Reply #12 on: April 05, 2019, 06:30:26 PM »
It seems to me that no one was flying the plane. They expected the computers to do it. I was taught, avigate, navigate, communicate. In other words above all fly the plane.

Perry, I’m not sure how you come to that conclusion. For starters, my understanding is that MCAS isn’t even active unless the autopilot’s disengaged. If it introduced erroneous inputs, they had to be hand-flying the aircraft at the time. From my reading of the preliminary report, the autopilot disengaged about a minute after takeoff (presumably due to AoA discrepancy), and they were attempting to control the aircraft manually for the remainder of the flight. They did use the stab trim cutouts at one point, and towards the end the Captain asked the FO to ‘pitch up together’, suggesting that they were both pulling back to try to get the nose up.

It’s very easy to wheel out the ‘aviate, navigate, communicate’ mantra from the comfort of a 1G armchair (and no doubt these pilots were taught the same thing, like just about every other pilot since Orville and Wilbur). Maybe they made mistakes, and maybe you’d have done better. But to suggest that no-one was flying the jet, and they were just relying on the computers, is doing a grave disservice to a couple of fellows who were trying desperately hard to cope with a dire situation.
« Last Edit: April 05, 2019, 10:43:45 PM by Steve Thomas »

Offline Ted Fancher

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2326
Re: 737 Flight Control
« Reply #13 on: April 08, 2019, 06:14:33 PM »
It woudn't have mattered.  The controls are mechanically tied together.

Howard, that was my first thought after a lifetime on Boeings but was uncertain about things developed after I left the NASA post retirement gig.

Much as I hate to admit it in public I found Howard's Boeing stuff I flew amazingly straight forward and unambiguous...as well as close to faultless.  Not, however, suggesting that this doesn't look like a bug that probably needs to be squashed.

Shortly after the second accident I started what turned out to be a too long for prime time discussion post regarding the changes in the potential airline pilot pool based on the historic change in the "aura" of aviation over my nearly eight decades of involvement in aviation.  From every kid with a hair to his name wanting to emulate Lucky Lindy to "who the heck want's to drive a bus with wings" ennui. 

361 logged hours as second (and last) in command of nearly 200 lives in a "complex system based" jetliner is not a headline you'd prefer to share with the flying public. 

To go one step further even high time pilots transitioning to auto-flight system aircraft find themselves stunned at the degree to which initial training on such fleets is predominantly aimed at how to push the "go auto" button shortly after liftoff and then learn which buttons to push in the event the route you installed back at the gate gets changed by ATC and a new one installed.  Very little hand flying accompanied the training on the 757/67 fleet.  After my first couple of simulator training session I (this isn't a joke) I actually asked my instructor how one got out of this chicken (manure) outfit. 

After my first year on the fleet, however, I learned to love flying them (and the 747-400 I flew prior to retirement) and hand flew them a lot as they were delightful to fly.   Plus the autoflight capabilities and the improved situational awareness provided by the instrumentation were/are great assets, particularly in visibility conditions that would have had the crew turning the old 727 for an alternate airport.

If, however, the second in command has only a handful of hours total and most of it is directed at flying by pushbuttons his/her comfort level might well be stressed significantly if the situation demanded a sudden shift to hand flying under dire circumstances.

Sorry...too much BS.  Bottom line, I trust any mechanical problems ultimately determined to have been involved in the accidents will be resolved by the historically amazing people at Boeing. 

Ted



Offline Peter in Fairfax, VA

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1108
Re: 737 Flight Control
« Reply #14 on: April 09, 2019, 06:14:28 AM »
Ted,

Did you read the transcript of the cockpit voice recorder?  Helped me with respecting the pilot's efforts.

Peter

Offline Ted Fancher

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2326
Re: 737 Flight Control
« Reply #15 on: April 09, 2019, 12:50:47 PM »
No, Peter.  I haven't seen it.  Do you have a link? 

By the way, no "personal" disrespect of the crew was intended.  My perspective for the experience comments were simply the result of knowing how much more I understood after a few thousands of hours than I did after the first few hundred; plus an on-line  review of the modest depth of aviation industry/interest in Ethiopia.  For instance, the only "flight training facility" I found in the admittedly short search was the one provided by the airline itself. 

I've little doubt the pilots were bright, well intentioned, and did their best...and until final reports are in, perhaps, found themselves in an un-salvageable situation.  I also read an ASRS report of a US flight crew (airlines and  airports are de-identified in such reports) that encountered an "apparently" similar event and recovered.  Perhaps, of course, the degree of degradation of the MCAS probes might also be a factor re recover-ability.  I've no idea about that.

One final thought about the "big" guppies.  Boeing had a wonderful (from a pilot's perspective) equivalent fleet (in pax count) on the market for many years, the 757/67 fleets which were discontinued.  The need/decision to retrofit the 737 with more powerful and efficient engines whose diameter required them to be mounted on pylons higher on the wing leading edge to provide adequate ground clearance apparently contributed to the need for the MCAS system (based on reports I've read...I'm certainly not an expert).  This "may" not have been necessary to provide the discontinued 57/67 fleet to utilize the more efficient powerplants and provided the additional range for the markets intended to utilize such a fleet.  This might be nostalgia on my part because the 757 was, by a significant margin, my favorite Boeing to fly.

Well, that's embarrassing.  I just reviewed the thread and saw the report link Peter referenced in the very first post.  My only excuse is that I thought this was the same thread as an earlier one re the 737 events and I scrolled immediately to the the end rather than re-read what I thought would be "old" posts.  My  apologies and I will read the report and, if appropriate, amend or remove my prior post. (Text in red to match the color of my face)

Ted

Offline Chris McMillin

  • 22 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 1899
  • AMA 32529
Re: 737 Flight Control
« Reply #16 on: April 10, 2019, 03:42:10 PM »
I didn't read all of that but will, needing to go to school on an airplane to read the accident report takes a few more minutes than I have right now, but it seems that the narrative indicated the trim ran after the dual cutouts were shut off. Not possible on the 727, I never flew any model of 737.
 
If that is true it would make an instrument differences problem into a double problem. Trying to make sense of and correct that was interrupted, within the crew's minds, by having to fight the airplane from becoming untrimmed and therefore hit the ground with intermittent and uncommanded downtrim.
 
I agree with Ted about the broader issue of why does the 737 still exist. It is because of Southwest Airlines ordering them and the majors being in the grips of bankruptcies and individual profit taking from their owners instead of making airplane orders for far superior equipment in the form of the 757 and 767. Available in 180 minute ETOPS too, Ted! Cockpit half the size of the 72, to HNL, Yah!

Chris...


Offline Mike Danford

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 50
Re: 737 Flight Control
« Reply #17 on: April 10, 2019, 06:28:54 PM »
Me three.... Why in the world it still exists escapes me.  You can only stretch a plane yay so far.  Band aid fixes are nearly always problematic.  The 73 was always the SLUF, now it's just a UF.
The 75 is my second favorite airliner of all time, a close second to the A330, notwistanding the buss' miserable automation. 

As an aside, the 757-300, the stretch 75, was a miserable beast.

Tools

Offline Ted Fancher

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2326
Re: 737 Flight Control
« Reply #18 on: April 10, 2019, 10:53:27 PM »
Me three.... Why in the world it still exists escapes me.  You can only stretch a plane yay so far.  Band aid fixes are nearly always problematic.  The 73 was always the SLUF, now it's just a UF.
The 75 is my second favorite airliner of all time, a close second to the A330, notwistanding the buss' miserable automation. 

As an aside, the 757-300, the stretch 75, was a miserable beast.

Tools

Hi Mike,  Your comments about the A330 vice 757 for your "favorite ship" was of great interest to me. 

After I retired from UAL in 2003 I went to work for NASA ASRS as an "expert" (quotes on purpose) analyst of reports submitted by pilots, ATC, etc. (you know the drill, I expect).  In the "cubicles" next to me were a couple of guys who had flown various Airbuses in their careers...I was strictly a Boeing guy (Well, plus a Convair and a few Douglasses, -6 and -7 and -8).  I found myself often conferring with them when I was analyzing reports from 'bus' pilots as the auto flight terminology and functionality was often like sanskrit to me...similar but--well--different from the Boeings; both in terminology and operation...plus of course the miniature "stick" controller vice the ubiquitous wheel of the Boeings.

One of the things I found of great interest was that I had flown airplanes from a "very" young age and--although the planes were different--getting from point A to point B was essentially identical.  You repeatedly tuned a VOR and cranked in an inbound or outbound course and flew from A to B to C to destination.  Nav training was a non event for the most part. 

When we "progressed" to the age of "automated flight" however, all of a sudden it seemed that getting from A to B to C was a different animal given the manufacturer of your airplane and training to prepare for line flying was now more or less "unique" depending on the manufacturer of the bird in which one sat...and the training dedicated to navigation from A to B became the primary rather than "oh, yeah, it's still the same--time for a snooze" course segment.

As I had never experienced an "auto-flight Boeing to Buss or reverse" transition I began to wonder to what degree differences in fleet type nav system operation might relate to nav/system/functional nav errors based on the significant differences in terminology and functionality of the buttons and dials.

I'd be interested in your experience in transitioning from Boeing to Airbus (or vice versa) autoflight terminology and functionality and, more specifically, to your experiences, if any, with the buses ability to "correct" what they perceived to be flight crew input errors.

Oh, yes.  And the fact that in some cases the automation was/is the Captain!

Ted

p.s.  we called 'em "FLUFs"

p.p.s.  Like Chris, I never flew a guppy of any sort.

p.p.p.s.  Your comment about the Buses' "miserable automation" and my confusion with its terminology and ops at ASRS is largely what led me to respond in this manner to your post.

Offline Mike Danford

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 50
Re: 737 Flight Control
« Reply #19 on: April 11, 2019, 07:08:22 AM »
Hey Ted!

First, I'm a die hard Boeing fan, so make the statement with a lot of thought.  I was a Boeing to bus transition, but to be fair, have the most time in the BOEING (tongue in cheek) md-88. 

-edit-  bus refers to Airbus.  A good jet, good company, but known for EXTENSIVE automation that isn't very intuitive.  It really is as if engineers are convinced they can remove decision making from the pilot.  These planes are EXCEPTIONAL when NOTHING is going wrong, whether it be internal or external.

I make the statement purely from a airframe perspective.  That thing LOVES to fly, and it shows.  If I hit the marker at 250 in a mad dog, zero problem.  250 in a Boeing, it's work but totally safe.  180 (vs our usual 170) in a -330, likely going around!  No kidding.  Don't know about the -320 series. 

-edit-  the marker is basically a final approach fix.  In bad weather, in the clouds, you want to arrive here, basically 2000 feet above the ground, 6 1/2 miles from touchdown, or about 3 minutes, configured to land and at your landing speed.  In good visual conditions you generally start to configure to land here.  If traffic is light, approach will let you get here as fast as you want within reason. 

A fully loaded 76-300er going to Africa dribbles off the runway, squats in a turn and will once in a while give you a heart stopping touch of shaker when you retract flaps.  575,000 lb -330 flings into the air, goes right to 30 degrees bank and flies around clean at 220 knots.  Amazing. 

-edit- shaker refers to stick shaker, basically a stall warning feature.  Really gets your attention.  We can test this on the ground with a simple flip switch.  We tell kids to hang onto the controls and then flip this switch and tell them they are shooting the machine gun!  I've shot jet mounted machine guns, it pretty much feels the same!

With auto throttles off, NOT EVEN DEMONSTRATED IN TRAINING!, if you're a little slow and add a little throttle, it goes a little faster, and vice versa.  The 75 and 76, add a little, nothing.  Add a little more, nothing.  Just squeeze em a tad, bam, 10 knots too fast.  Lather rinse repeat! 

-edit- 75 and 76 refer to Boeing 757 and Boeing 767.  Guys from other companies might refer to them as 57 and 67.  Delta had a 767-400 jet, northwest had a 747-400... Neither company had both, and they were referred to as -400s in each company.  After the merger, much confusion!  Lots of little,things like this.  All planes with radios have a place to hang the microphone, which is rarely used, ever.  Shortly after the merger of northwest and delta, you could tell the background of the guy who flew previously by where the microphone was.  At delta we hung it face in between to switches, the cord neatly nestled inbetween.  Northwest guys set it face out, the projecting portion neatly nestled in a vee formed by the intersection of the glare shield and side window...  These little differences are pretty fun to notice.  Much like dialect in a language.

Just a very very gentlemanly airframe.  Trend vectors ROCK on the engine and airspeed displays.  The miserable  methodology of the autopilot is only surpassed by the FMS.  It seems all the intuitive stuff was gobbled up by Boeing and patented. 

-edit- trend vectors are a little arrow off the pointer... Remember, no real Gage's, merely pictures of them on a computer screen.  These little arrows give you an idea of the rate of change the pointer is experiencing... Weird at first, priceless once you figure the little buggers out.  Unique to the bus in my experience.  The 787 might have them, don't know...

The side stick doesn't matter a lick.  I like it, very easy to put your charts in front of you, so pragmatic at the least.  Forget to adjust your armrest and it's like you forgot to fly!  Insofar as the simulator behaving like the plane, well, BOTH are totally computer controlled, so it does!  Spot on!

Have to remember that you are not commanding an aileron position, but a roll rate.  Not an elevator position but a g onset rate...  It's NOT fly by wire, but fly by computer, even when the autopilot is "off".

I have been a HUGE advocate of hand flying since I started in the airlines.  Only in the last 5 years has it become ANYTHING of a priority.  I hate to think of how many airliners would have been saved by a simple modicum of proficient hand flying.  Uh, power plus attitude equals performance... Tell me that doesn't bring back some OLD memories!

I reckon if necessary I can get on the radio and get help with the autopilot, not so with FLYING.  In the Navy we had no minimums at the ship, land or punch out.   Zero zero approaches hand flown to a carrier deck were expected.  Didn't even have a autopilot, or flight director.  Hand flying an airliner in zero zero conditions with how stable they are, and awesome flight directors is flatly boring, and since the northwest merger, we can't even hand fly a cat II.... That's sad.  Unless there were mitigating circumstances, I ALWAYS hand flew cat I approaches, because it's fun!  If tired, really dicey weather, some other issue going on, weak other crewmember, hooked it up and watched.  That happened often enough to stay plenty proficient.

-edit- a flight director is a method of showing the pilot what the autopilot would do, or is doing if engaged.  It shows up on the main gyro.  If you have it on, and the autopilot off, you can follow it manually.  It helps to not over correct, as it is dampened out.  In certain phases of flight, it's usually not so good, you learn when to follow it and not.  Again, once acclimatized, it's usually a great thing. 

Your post United gig sounds pretty neat!  Love to chat about this stuff, Give me a holler any time! 

Mike, 423 883 0721
« Last Edit: April 11, 2019, 04:16:14 PM by Mike Danford »

Offline Fredvon4

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2099
  • Central Texas
Re: 737 Flight Control
« Reply #20 on: April 11, 2019, 09:56:34 AM »
Hey Mike love you post BUT...Sir... please edit and let us laymen in on the JARGON and ACRONYMS...I can guess many becaus I was Army Helicopter maintenance and taught to fly a little bit....but some of this reading would be so much more relevant if I could have several more AHA revelations

fer zample zero zero I get but Cat I or Cat II??
"A good scare teaches more than good advice"

Fred von Gortler IV

Offline Mike Danford

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 50
Re: 737 Flight Control
« Reply #21 on: April 11, 2019, 11:31:08 AM »
Fair enough!  I'll edit when I get a few more moments, but for right now, cat I, II and III.  Each is a category of approach, with lower and lower minimums, or worse and worse conditions can exist and you still fly them.  Basically all forward visibility related.  You do not need much visibility to do cat III, and it requires specially trained pilots and specially certified equipment onboard the plane, and special ground based equipment. 

I've only HAD to rely on that to get in several times in twenty years.  Most if not all companies who can do them, require them to be automated all the way to the ground, meaning auto land is even a requirement. 

It all kinda gets ridiculous in that the difference between a cat II cat III is barely within reaction time to realize you can't see the runway and execute a go around.  It's well recognized that a go around from a cat III at minimums will likely result in a touch and go, which could easily be the safest course of action. 

All of this special training is a little dicey.  Many times it's pretty intensive, which I think it should be, other times it's a self certified watching of a you tube type video....  Such as in the "differences" of very similar variants of an aircraft type, such as is relevant in this thread. 

Like Ted, I've flown quite a number of aircraft types.  The hardest thing to get used to?  Figuring out the communication equipment and cockpit lighting...  Seems minor, until you can't see something on a dark scary night!  If you hear a guy making radio calls on the public address system, likely very new to the equipment. 

Mike


Advertise Here
Tags:
 


Advertise Here