Ethiopian ET302 Crash.

Status
Not open for further replies.
Originally Posted by oil_film_movies
The lesson learned here should be that news media reports of loose interpretations of ASRS complaints is not reliable.
Take the FAA's Elwell's direct word for it. For him to be lying is not likely, a lot of integrity there.


Most media people are not qualified to report on the details of this issue.


Originally Posted by oil_film_movies
Sometimes with all this publicity the flight control law block diagrams (transfer functions, tables, logic, etc.) will be leaked or even presented. Nice if the source code was leaked as well!
Then I could analyze it to find out what they are doing on a detailed level.
One guy years ago came to our avionics company with a bunch of block diagrams from a company I won't name..... not right to do unless the times calls for it.
In cases of public safety and with all the political confusion like this one,..


Ans so could a number of people here.

All of the aircraft and avionics companies I worked for had you sign a Non-Disclosure agreement that basically said you may not take data out of the engineering dept. or disclose it to any outside entity, with the exception of a request from the FAA, NTSB, or any other government agency and only then after consultation with the Legal Dept.

Bringing proprietary data from one company to another opens up both the individual and the receiving company to complicated legal action.

Originally Posted by oil_film_movies
Boeing should make public the details or someone should leak it.


So Airbus or competing companies can copy it?
 
Last edited:
Originally Posted by MolaKule
Originally Posted by oil_film_movies
Boeing should make public the details or someone should leak it.
So Airbus or competing companies can copy it?
In ordinary times, without this PR nightmare and without two crashes hanging over Boeing's head, block diagrams and/or source code should not be shared, either illegally or presented at tech conferences.
In this case, I would have no problem seeing them leak out. Boeing needs to show airlines how this is mechanized in software inside, and prove they aren't hiding anything else. Not ordinary times right now.

About control law block diagrams: They aren't as high-tech or unique or private as you might think. I've seen and worked on control laws from space probes to airliners to fighter planes, about a dozen different kinds of vehicles, and they all use the same elements. Some airlines get peaks at them. ... Also, many times the control laws are shared with flight simulator training companies for implementation, which means they pretty much open it up at times. NASA gets a lot of them, and they notoriously leak them to China.
 
Maybe Astro14 or anyone else here who's a commercial airline pilot can shed some light on this.

How much pilot training do commercial pilots get per year, and how much emphasis is put on emergency situations/conditions?

With airlines ran in other countries ... who dictates what training material is covered and how often pilots are trained? The airlines, plane manufactures or some government agency of that country? Or some of all involved?

What about in the USA ... is it the airlines, plane manufactures or some oversight by the FAA or similar organization (or a combo of those) in terms of pilot training material and training frequency?
 
Originally Posted by ZeeOSix
Maybe Astro14 or anyone else here who's a commercial airline pilot can shed some light on this.

How much pilot training do commercial pilots get per year, and how much emphasis is put on emergency situations/conditions?

With airlines ran in other countries ... who dictates what training material is covered and how often pilots are trained? The airlines, plane manufactures or some government agency of that country? Or some of all involved?

What about in the USA ... is it the airlines, plane manufactures or some oversight by the FAA or similar organization (or a combo of those) in terms of pilot training material and training frequency?




Definitely not an expert so I hope Astro will confirm this but from what I understand, each country or region has their own aviation oversight commission like the FAA. So standards may vary. However, if that airline flies routes to and from the US, they must follow the standards set by the FAA. There are and have been airlines that were banned from flying to the US until those standards were met.
 
Originally Posted by oil_film_movies

...About control law block diagrams: They aren't as high-tech or unique or private as you might think. I've seen and worked on control laws from space probes to airliners to fighter planes, about a dozen different kinds of vehicles, and they all use the same elements...Also, many times the control laws are shared with flight simulator training companies for implementation..


For those not familiar with the "Control Laws" Block Diagrams here is a short explanation and some simplistic examples:

In terms of aircraft Stability and Control, engineers express the various Inputs (sensor inputs) to a Control System and Outputs to control the surfaces in terms of System Transfer Functions, which describes the inputs and how the control system Output will manipulate the control surfaces or,

Output = Input X Transfer Functions.

These Inputs, System Transfer Functions, and Outputs are further expressed in terms of partial differential equations that are derived from the airplane's flight coefficients as determined (usually) by data from wind tunnel testing of scale models, or simulation and from sensor characteristics. Those partial differential equation coefficient-expressions are then coded (in software) into say the Flight Control Computer (FCC) for automatic control of Pitch, Yaw, and Roll.

For example, we program into the FCC that we will fly a heading of 100 degrees (ESE) and at an altitude of 34,000 feet.

During flight the gyro or other navigational sensor says we drifted 2 degrees south or to a heading 102 degrees. The transfer function will then smoothly deflect the rudder trim tab or ailerons to put us back on course. The same happens when altitude is gained or lost except in this case the elevator trim tabs will deflect up or down to correct for altitude.

A simple example of one such single Block Diagram is shown below from a text on Airplane Flight Dynamics and Automatic Flight Controls. Keep in mind there are many of these transfer functions coded within the FCC;



System Transfer Functiion.webp
 
Last edited:
The fourth generation of Boeing 737 family is a huge step forward in terms of fuel efficiency, emissions reduction and noise reduction for the single aisle market. Compromises have been made to achieve these advances and Boeing introduced the MCAS system on this airframe to increase safety, not reduce it.

That being said, based on the reports I've read and public articles that have been written it appears to me that there is a product introduction training issue with this new technology. In simple terms MCAS (maneuvering characteristics augmentation system) is in place to detect and ultimately prevent the aircraft from going into a stall condition. The two crashes that have occurred both happened very shortly after take-off when the conditions monitored were likely detected by the MCAS as a pending stall. Air speed (low), attitude (angle of attack high), and flight control settings (flaps up) set the intervention into action.

Boeing will adjust the parameters that MCAS monitors and likely make it less sensitive and I am hoping that the benefits of this feature are covered more in depth at the specific airline training areas. It is a new feature that I feel hasn't been emphasized enough with regard to training.
 
This is getting crazier. Seattle engineers who worked on this thing are now facing more angry hornets buzzing around: Internal Boeing execs, the FBI, grand juries, a Senate hearing, Canadian & European authorities, FAA duck-and-cover officials, .... who else will gang up on this group? And all because they didn't add enough AoA redundancy management as most flight control engineers would have done, a simple mistake followed by a storm.

Originally Posted by MolaKule
Originally Posted by oil_film_movies
...About control law block diagrams: They aren't as high-tech or unique or private as you might think.
For those not familiar with the "Control Laws" Block Diagrams here is a short explanation and some simplistic examples: {see the previous post for diagrams etc.}
True there are a lot of linear differential equation Laplace transfer functions (& some funky Z-transforms) you find in fundamental control systems courses and books or tech papers that are in real aircraft. .....Over the years, the industry has added and experimented with a lot of non-linear functional mappings from input-output, and that's really the parts companies try to keep 'secret' but can't because they end up having to share a lot of detailed inside algorithmic info & source code to simulator companies, the FAA, NASA, etc., some of whose personnel sell it to China and/or competing companies, other world governments, almost immediately upon implementation. (It's like KFC having to disclose their spice recipe to other companies & entities, expecting that audience to keep it a secret.. it's out.) I think Tesla, for example, has an easier time keeping their software algorithms secret since they don't have to share it with outside entities.

Originally Posted by FowVay
... conditions monitored were likely detected by the MCAS as a pending stall. Air speed (low), attitude (angle of attack high), and flight control settings (flaps up) set the intervention into action.
Air speed was never the issue. .... One AoA vane was way off, a single one. .... MCAS then used a single bad AoA sensor to add electric pitch down trim, that's the issue. Nothing else with the systems. ( Bad design to rely on only 1 sensor to run the electric motors moving the horizontal tailplanes for pitch down, big issues there.)
Then, the pilots didn't recognize the "runaway" trim was happening, and/or didn't remember (or ever know!) how to simply turn the electric pitch trim motors OFF, two easy switches to flip, right in front of them.

Originally Posted by FowVay
.... I feel hasn't been emphasized enough with regard to training.
Boeing & the FAA previously felt any runaway trim situation would be handled in the usual way for this airplane type, turning pitch trim motors OFF, an easy thing to do according to their understanding of human factors at the time.
 
Last edited by a moderator:
Originally Posted by FowVay
The fourth generation of Boeing 737 family is a huge step forward in terms of fuel efficiency, emissions reduction and noise reduction for the single aisle market. Compromises have been made to achieve these advances and Boeing introduced the MCAS system on this airframe to increase safety, not reduce it.

That being said, based on the reports I've read and public articles that have been written it appears to me that there is a product introduction training issue with this new technology. In simple terms MCAS (maneuvering characteristics augmentation system) is in place to detect and ultimately prevent the aircraft from going into a stall condition. The two crashes that have occurred both happened very shortly after take-off when the conditions monitored were likely detected by the MCAS as a pending stall. Air speed (low), attitude (angle of attack high), and flight control settings (flaps up) set the intervention into action.

Boeing will adjust the parameters that MCAS monitors and likely make it less sensitive and I am hoping that the benefits of this feature are covered more in depth at the specific airline training areas. It is a new feature that I feel hasn't been emphasized enough with regard to training.



thumbsup2.gif
Good Summary.
 
Originally Posted by ZeeOSix
Originally Posted by KD0AXS
Here's some more good info from Juan Browne.

https://www.youtube.com/watch?v=9Ts_AjU89Qk&feature=youtu.be


He mentioned that the MCAS system has been in the 737 from the very beginning of that planes existence, so what did Boeing change when they applied it to the new 737 MAX series of planes?


If Mr. Brown is implying that a B737-100 incorporated any sort of MCAS than this gentleman is simply wrong. The Max series (4th Generation) is the first aircraft to incorporate this system. MCAS was incorporated because the design and dynamics of the Max necessitated it. The Max 10 is going to have a few other items specific to that particular airframe that no other 737 has had also.
 
Originally Posted by FowVay
Originally Posted by ZeeOSix
Originally Posted by KD0AXS
Here's some more good info from Juan Browne.

https://www.youtube.com/watch?v=9Ts_AjU89Qk&feature=youtu.be

He mentioned that the MCAS system has been in the 737 from the very beginning of that planes existence, so what did Boeing change when they applied it to the new 737 MAX series of planes?

If Mr. Brown is implying that a B737-100 incorporated any sort of MCAS than this gentleman is simply wrong. The Max series (4th Generation) is the first aircraft to incorporate this system. MCAS was incorporated because the design and dynamics of the Max necessitated it. The Max 10 is going to have a few other items specific to that particular airframe that no other 737 has had also.


I went back and watched the video again ... and he's saying the STAB TRIM cut-out switches are located in the same place they have been since the 737 was made ... guess I inferred that meant they also had MCAS, but apparently not.
 
Originally Posted by oil_film_movies
...True there are a lot of linear differential equation Laplace transfer functions (& some funky Z-transforms) you find in fundamental control systems courses and books or tech papers that are in real aircraft. .....Over the years, the industry has added and experimented with a lot of non-linear functional mappings from input-output, and that's really the parts companies try to keep 'secret' but can't because they end up having to share a lot of detailed inside algorithmic info & source code to simulator companies...



So are you saying the partial differential equations used in those Transfer Functions do not take into account any non-linear behavior?

There is nothing funcky about Z-transforms if you understand them.
 
Last edited:
I heard on the news today that the pilots had received specialized training on the mcas. So wonder if we will ever know the truth if the black box shows, or switches were found turned off or on. And how do these switches actually do their jobs? Are they a direct connection carrying many amps to that jack-screw motor or do they go through a solid state or mechanical relay? I would like to see actual construction photos showing the purported control cables.
Oh and what other systems do the computers automatically just take command of on these planes? A case of typical modern engineering, having one input for attitude control, there are many other inputs available, like airspeed, rate of climb, altitude, heading and the conditions shown with an artificial horizon. So I guess computers just can't handle all the inputs a human can??
 
Last edited:
This video is from a 737 MAX pilot and was made last November after the Lion Air crash. He talks about MCAS and what he says in this video sounds pretty spot on to what's happened in both of these crashes from what we know so far.

One thing he said is that the yoke force bias also changes when the aircraft thinks there is a stall condition to help the pilot take control in a stall condition. The yoke becomes easier to push forward (nose down) and harder to pull back (nose up). So it sounds like when the pilots were fighting against MCAS and trying to pull the nose back up, the yoke would have actually become more resistant in a pull back direction which is opposite what you'd want in the case when fighting a faulty MCAS control system that's continually trying to put the nose down.
 
Originally Posted by ZeeOSix
Originally Posted by FowVay
Originally Posted by ZeeOSix
Originally Posted by KD0AXS
Here's some more good info from Juan Browne.

https://www.youtube.com/watch?v=9Ts_AjU89Qk&feature=youtu.be

He mentioned that the MCAS system has been in the 737 from the very beginning of that planes existence, so what did Boeing change when they applied it to the new 737 MAX series of planes?

If Mr. Brown is implying that a B737-100 incorporated any sort of MCAS than this gentleman is simply wrong. The Max series (4th Generation) is the first aircraft to incorporate this system. MCAS was incorporated because the design and dynamics of the Max necessitated it. The Max 10 is going to have a few other items specific to that particular airframe that no other 737 has had also.


I went back and watched the video again ... and he's saying the STAB TRIM cut-out switches are located in the same place they have been since the 737 was made ... guess I inferred that meant they also had MCAS, but apparently not.


Only watched the vid once and obviously after reading Astro's and others posts...but you are right (second time around) the vid says that the cutout was there in case things went awry and the system progressively and repeatedly sent you in the wrong direction.

This is akin (IMO) to my analogy in other threads about the drunk in the passenger seat grabbing and jerking the wheel to avoid a potentially non existent issue, and ending inthe ditch (Oz law is that the drunk gets charged DUI for assuming control of the vehicle...so should the programmer).

Electronic turbine overspeeds have three redundant circuits, and two of the three polled being in overspeed push a turbine park event (trip)...multiple sensors with single polling is a serious control system logic fault in any industrial application...parking a turbine and parking a plane at speed are seriously different consequences.
 
BTW...and it seems that we are doing it everywhere...we are currently relearning stuff that we learned 40 60 years ago...and subesquently got lulled into some level of complacency with respect to problems that were put to bed two generations ago.

Got into debate today as to whether an Aluminium flame trap on a methanol tank corroding was a fundamental design flaw (material compatibility) or preventative maintenance....only us old farts even knew that the two don't mix....it had dropped out of common knowledge, as we hadn't had to deal with the consequences for a geneeration of engineers at least.
 
Originally Posted by Exhaustgases
I heard on the news today that the pilots had received specialized training on the mcas. So wonder if we will ever know the truth if the black box shows, or switches were found turned off or on. And how do these switches actually do their jobs? Are they a direct connection carrying many amps to that jack-screw motor or do they go through a solid state or mechanical relay? I would like to see actual construction photos showing the purported control cables.
Oh and what other systems do the computers automatically just take command of on these planes? A case of typical modern engineering, having one input for attitude control, there are many other inputs available, like airspeed, rate of climb, altitude, heading and the conditions shown with an artificial horizon. So I guess computers just can't handle all the inputs a human can??


For many pilots, the training for the MAX was a video on an iPad*. They were already qualified on the 737, so, this was only "differences" training. There have been some very contentious meetings between pilot groups (particularly SWA and AA) and Boeing executives. The iPad/computer video simply was insufficient and did not describe MCAS operation.

As with all mishaps, there are a multitude of causes. I said it in an earlier post, but Boeing bears some responsibility for installing this system and failing to tell pilots about it.

I imagine that there is now plenty of discussion on the MCAS in the MAX in 737 circles.

As far as the MCAS/trim failures in the US, my source was an experienced 737 Captain at another (not my own) airline. A friend from my first squadron. I trust his information, he's pretty high up in the union at that airline and has flown only 737 for the last 24 years. He knows his stuff.

Stab trim runaway is a training event. So is unreliable airspeed. We see those malfunctions in the simulator. The Stab Trim malfunction is easy in a Boeing - throw the switches. Runaway stops. And an MCAS failure would be stopped the same way.

Unreliable airspeed is more challenging. You have to be able to understand what's accurate, and what's not. For example, AF-447 was pulling up at cruise altitude to 10 degrees of pitch, in a rapid climb, while airspeed was increasing and the power was coming back. Only airplane I've ever flown that can accelerate like that had afterburners. Airliners at cruise simply can't climb, even at full thrust, at 10 degrees nose up. So, right there, an experienced pilot would know that at least one of the parameters couldn't be true. In their case, it was airspeed, it showed increasing, when in fact, it was decreasing, due to the iced up probe.

But, as I said in our discussion on that event, the airplane was telling the pilots that the airspeed was excessive (beyond redline) and the correct response was to pull up and slow down to avoid damage. That's how they were trained. When the airplane stopped climbing, it's because it was deeply stalled, but the instruments didn't show that. As it fell from the sky, they applied stall recovery procedures, again, as they were trained to do.

Those procedures only work when the airplane is in normal law. And they never noticed, among the cascade of failures and warnings, the one message that said, "Alternate Law". Or, if they did see it, failed to understand what that meant for stall recovery, perhaps they were never taught that.

Or, perhaps, sometime after midnight, 1,000 miles over the ocean, when they were startled by all the warnings, including multiple sirens and horns, red lights, yellow annunciations and messages on various displays, they just couldn't remember that bit as they struggled to figure out what was happening and how to control the airplane.

Engineers are great. Good at math. But what they build doesn't always work, or fail, the way they anticipate. Each warning, or logic tree, makes perfect sense when it is viewed by itself, in a well-lit office, between the hours of 9 to 5, when you're at zero knots and 1 G, perhaps while sipping on your coffee.

It's another thing entirely when the failure and warning is presented in difficult operating circumstances, like night, terrain, or over water, and in conjunction with other warnings, other failures, in a sudden surprise, with your life and others riding on the outcome, where there may not be a simple answer.

I have to smile at those who think that because they have flown a simple airplane, that they understand flying. Sure.

If you can ride a moped, you're ready for SuperBike or MotoGP racing. The basics of riding are the same, right?

Not much different, really.... Moped, and race bike? You can ride a moped on a quiet street, it's just like being in a race at 180 MPH on a 200 HP machine...

The speed, complexity, handling and operational environment of an airliner bear zero resemblance to bumping around the pattern in a Cessna. Sure, they're both technically flying, but they have little in common.



*Pilot training takes many forms.

We have video training 3 times a year, about 6 hours of coursework in each session on "hot topics" or reviews of systems, procedures, etc. For example, we will get a winter operations (de-icing procedures, contaminated runway operations, holdover time calculation) review in the fall, along with a review of systems, current operational focus areas. About a dozen videos each, three times per year.

We get in the simulator every 9 months. It's either a two day, or a three day event. During each training event, we will work with Flight Attendants on operating doors and emergency equipment and cover our responses to things like passenger misconduct, suspect device on board, preparing the cabin for evacuation. We get a systems review and test. We get briefed on current operational focus areas, then review normal and emergency procedures and then fly the simulator in a 4 hour session. They are very long days, about 12 hours each, with some classroom and some flying each day.

New information comes to us via Flight Manual changes, Operations Manual Changes, Bulletins, Ops Alerts, email messages, Fleet newsletters, messages in our crew log-in screen, and, of course NOTAM and other external sources. It can be daunting to stay abreast of all the changes, and all the updates.

I can see the frustration of 737 pilots.

Buried in that avalanche of different information sources, was a video on your new airplane, and the video didn't explain that there was a new system in the MAX that would trim the elevator down in certain circumstances. In checking the Flight Manual, to try and consolidate your updated understanding, if you didn't find anything on MCAS (as was true at several airlines), then you, as a pilot, have every right to be upset with how the information was disseminated.
 
As I explained in Post #5046176, there is a long process in the development of aircraft.

In defense of engineers, the hardware and software engineers develop hardware and software according to the Requirements given to them. They don't develop those components in a vacuum.

Whichever group within Boeing developed the MCAS, they did it from Requirements that were handed down to them from above. (From a guy who worked for McDonnell Douglas, Boeing, and Avionics companies from 1981 to 2013).

I would have to guess that somewhere in the process of developing or updating the Requirements, the MCAS system may not have been thoroughly vetted by Pilots and Human Factors, otherwise there would have been better info dissemination and descriptions regarding the MCAS and its effect on flight.
 
Last edited:
Apparently the Boeing 737 MAX for Lion Air and Ethiopian Air lacked an extra cost option safety features(an angle of attack indicator and a disagree light) to warning of the system failures.

So now kindly Boeing corporation is going to offer these features for "free" in software and retrofit.
 
Status
Not open for further replies.
Back
Top Bottom