Ethiopia crash

[KevinAu 441]

9 minutes ago, skelsey said:

Correct, but I'm not sure an EFATO is a good analogy to the MCAS-related incidents for the simple reason that EFATO, the challenges it presents, recognition and recovery are very well known and, indeed, trained for every single time the crew visit the sim.

MCAS, on the other hand, was a problem that the manufacturer had failed to tell pilots existed and was insidious in its appearance -- don't forget that the 737 does a lot of 'uncommanded' trimming (STS) in any case so pilots would be well used to the trim wheels whirring away in unexpected (and, to the pilot, apparently counter-productive) directions around this stage of flight.

If the manufacturer had said "look, we've put this new system in to the aircraft, this is what it does and where the inputs come from and here's some possible failure modes that you need to be conscious of" in their training package then that would be one thing. But they didn't -- and don't forget also that with so few 737 MAX simulators in existence most crews will never have had experience of non-normals in a MAX sim -- only in an NG, where the root cause, an AoA failure, would present in a completely different manner.

To extend the EFATO analogy -- what would your assessment be if an aircraft exhibited totally different flight characteristics and displayed compeltely different indications to another variant when an engine failed on takeoff but crews hadn't been told that or ever seen it in the sim? Would it be reasonable to expect them to diagnose the problem as quickly and perform to the same standard?

But they did have the bulletin about MCAS failure and procedure already. They should have been aware of it. It came out shortly after lionair. Knowing what happened to lionair, I would think that self preservation instincts would have at least prompted them to look at it some time within the last few months, if not put it somewhere near the front of their minds.

[Bob Scott 7,720]

2 minutes ago, skelsey said:

Correct, but I'm not sure an EFATO is a good analogy to the MCAS-related incidents for the simple reason that EFATO, the challenges it presents, recognition and recovery are very well known and, indeed, trained for every single time the crew visit the sim.

MCAS, on the other hand, was a problem that the manufacturer had failed to tell pilots existed and was insidious in its appearance -- don't forget that the 737 does a lot of 'uncommanded' trimming (STS) in any case so pilots would be well used to the trim wheels whirring away in unexpected (and, to the pilot, apparently counter-productive) directions around this stage of flight.

If the manufacturer had said "look, we've put this new system in to the aircraft, this is what it does and where the inputs come from and here's some possible failure modes that you need to be conscious of" in their training package then that would be one thing. But they didn't -- and don't forget also that with so few 737 MAX simulators in existence most crews will never have had experience of non-normals in a MAX sim -- only in an NG, where the root cause, an AoA failure, would present in a completely different manner.

To extend the EFATO analogy -- what would your assessment be if an aircraft exhibited totally different flight characteristics and displayed compeltely different indications to another variant when an engine failed on takeoff but crews hadn't been told that or ever seen it in the sim? Would it be reasonable to expect them to diagnose the problem as quickly and perform to the same standard?

From the released portions of the CVR transcript, these guys knew they had a trim problem.  There's nothing insidious at all about the A/P dropping off and then grabbing a handful of airplane to find a grossly-out-of-trim condition--that's exactly how a runaway trim most commonly presents itself, and I guarantee you it's taught and practiced in the simulator.  Once you recognize that it has happened, you maintain control of the acft while executing the memory items and get the PNF into the checklist.  If the trim runaway had instead been due to a bad switch, chafed wiring, a stuck relay, or faulty inputs from the autopilot, the problem is handled the same way without regard to the cause of the runaway.  Yes, it would have been good to know some of the other tell-tale signs, but I truly don't believe this particular problem was foreseen before these accidents.  As we used to say in my USAF flying days, every note, caution, and warning in the Dash-1 (the FM) should have someone's name on it, because most of those scraps of wisdom were discovered by somebody the hard way.

 

[HighBypass 2,339]

2 hours ago, w6kd said:

....and the declining hands-on proficiency of crews that log thousands of hours of autopilot time, but less and less actual hands-on flying, sure looks suspect to me as well....

Children of the Magenta Line Part II - The Sequel...

I wonder if Captain Warren Vanderburgh (RIP sir..) is metaphorically turning in his grave?

[Murmur 3,941]

I wonder how many pitch-trim runaways happened in the, say, last 10 years, and how many of them ended up in fatal crashes?

Edited April 10, 2019 by Murmur

[KevinAu 441]

Hard to say when they don’t end in a crash. I know of a couple, one recently here and the one at eagle years ago.

[skelsey 765]

23 hours ago, KevinAu said:

But they did have the bulletin about MCAS failure and procedure already. They should have been aware of it. It came out shortly after lionair. Knowing what happened to lionair, I would think that self preservation instincts would have at least prompted them to look at it some time within the last few months, if not put it somewhere near the front of their minds.

Indeed, and procedural steps aside the FO did cut the stab out so perhaps he did have that thought in his mind.

As to why they might have turned it back on - here's a demonstration of just how difficult flying the aeroplane might have been in that scenario and, critically, how it may well have been impossible to manually trim nose up (and the transcript suggests that the Captain was struggling to keep the nose up before the switches were, we assume, turned back on)

https://giphy.com/gifs/TgbF6sloqHQxGcxJTe 

On 4/10/2019 at 1:51 PM, KevinAu said:

if he snapped out of the shock of it and reached backed into the first step of his trained abnormal procedure, ie declaring the problem, calling for memory items, etc, whatever steps Ethiopian had tried to hammer into him in the sim, then he would have found a safe space in his mind to step forward one step at a time along a path to handle the situation, utilize his fo, and comply with the procedure and not turn the trim back on. He didn’t do that. He never got past the shock of the noise and handling problems. 

Quite, and I think this is a key issue. Pilots may well be expected to keep calm and deal with problems but at the end of the day they are still humans. How then do we prevent other crews from experiencing the same issues? Is training as is currently delivered sufficient to convey the level of distraction and confusion that may arise in a situation like this? How regularly do crews practice a stab trim runaway at low level in the sim? Do the sims accurately reproduce the level of effort required to manually trim the aircraft?

RTO/EFATO drills are rehearsed before every flight - what about other drills that may require very prompt action and memory items (like a stab trim runaway)?

How do we help prevent that complacency from a systemic point of view? 

Another factor is that the 737 is not like a modern EICAS aircraft where the crew will get a helpful >STAB TRM UNSCHD (etc) message to tell them exactly what to do. Again, a legacy of a 50 year old design and the need for commonality. This may have been OK when the 737 was a very analogue aircraft, but is it still OK when you start layering more and more complex software and automation in to the aircraft?

Aside from the crew, I can think of a number of things which Boeing could have done to have given this situation a better chance of a positive outcome:

- Provided crews with a proper description of MCAS (just look at the confusion and lack of clarity or reference material at the start of this thread alone about what exactly MCAS was for).

- Given the MAX EICAS to provide clearer indications of problems (not done for commercial reasons as I imagine that would have required new certification/training)

- Provided a more comprehensive differences course (again, not done for commercial reasons)

- Provided a better STAB TRIM RUNAWAY checklist for the MAX more suited to MCAS runaways (for instance, including a step to attempt to trim electrically to neutral before cutting out the stab) - but would this have fitted with the commonality requirement? Probably not and if so then commercial factors trump safety again. 

- Incorporated the level of redundancy for the MCAS system that they are doing as we speak (i.e. not relying on a single sensor with zero input error checking to automatically drive probably the most powerful control on the aircraft) from the start but again for commercial reasons it was 'necessary' to rush the MAX out the door.

Boeing made a lot of choices when it came to the MAX and at the moment it feels as though too many of those choices were motivated by commercial requirements instead of putting safety first. As they say, "if you think safety is expensive try an accident" -- and Boeing are finding that out the hard way.

[John_Cillis 2,406]

2 hours ago, Murmur said:

I wonder how many pitch-trim runaways happened in the, say, last 10 years, and how many of them ended up in fatal crashes?

One must also remember the jackscrew crash off the coast of California so many years ago, mechanical parts and electrical parts, and software can break down.  Even the simplest flying machine, the trike, which is just weight shift and lately just strutted with little to no guy wires, is subject to failure of what is called affectionately the OMG!!! bolt.  I do not know anyone it has happened to but I know it can happen, hence a parachute backup, sans wing.

John

[cowpatz 243]

11 hours ago, KevinAu said:

Yeah, I have no idea what you’re getting at. We’ve already been going over that fdr readout for the last several pages of this thread. There’s nothing new in what you posted.

I'm picking you just looked at the 2D FDR plots and thought "nothing new here".

In your posts you mention runaway trim. This was not runaway trim. This was MCAS doing what it is supposed to do.

Take a look at at the HF (Human factor) aspect and put yourself in the window seat for a moment......

On rotate or shortly after the stick shaker goes off and never stops until the end.The trim wheel would be doing its thing and clattering away as it does.
At some point the stick shaker goes off. It is startling, noisey and jarring. Most pilots would never experience this out on the line with the last time probably in the simulator on initial conversion or a subsequent focused training detail.
The airspeed is normal but in the "red bricks" The startle factor (straight after flap retraction and so close to the ground) would be huge.  Initial thoughts would be focused on stall recovery. EICAS  messages pop up that are more suggestive of a Pitot Static issue than anything else.

IAS DISAGREE
ALT DISAGREE
AOA DISAGREE (If fitted?)
FEEL DIFF PRESS

If AOA DISAGREE was not fitted, the initial assessment would likely be that something probably took out a Pitot tube or the input data was corrupt. It takes time to discount this.as a cause.
What does IAS DISAGREE, ALT DISAGREE or FEEL DIFF PRESS have to do with an angle of attack system?

This is all closely followed by EGPWS callouts....more pucker factor. Attempts to engage the AP initially didn't work.

The AP finally engages and the aircraft accelerates and flap is retracted.
Then all hell breaks loose. The AP disengages, the nose pitches down and the EGPWS activates (Stick shaker still going). The pitching down could perhaps be misinterpreted as a stall or just the AP letting loose - out of trim?. Now above flap speed so can't re-select flap. Meantime MCAS keeps doing it's job trimming AND (The trim wheel may not be heard moving due to the stick shaker, EGPWS and cross cockpit communication (Diagnosis/Checklists?). Any pilots response here would be to maintain attitude and power. At this point there is an awareness that the force required to maintain pitch is high and ANU trim is applied. It interrupts MCAS, and the AND trim momentarily, but the control pressure is alarmingly higher than normal. The trim is released (abnormally high column loading requiring a re-grip or just to check it has stopped?), MCAS re-engages and trims down 3 degrees in 7 secs requiring near full aft control column displacement. More EGPWS warnings and that damned stick shaker noise and feedback. This is whole new territory for any pilot and so close to the ground. Some ANU trim is applied and the STAB moves but not enough to relieve the high loading before it stops. At some point on from here the F/O places the STAB TRIM SWITCHES into CUTOUT. An attempt at manual trimming is tried but is fruitless (Way to much control loading for it to work). Things appear to have temporarily stabilised but there is still a huge control loading given the speed. Starting to tire the STAB TRIM switches are returned to norm (There is no evidence to suggest that they were positioned back to NORM but assuming that they were we can see from the FDR data that some ANU was available. Unlike the MCAS recovery procedure the RUNAWAY STAB procedure does not state that the trim cutout switches need to stay in CUTOUT. To add to the confusion the Overspeed clacker sounds as well as the stick shaker still doing it's thing. Exhausted, the MCAS reactivates and AND trim causes a rapid change in pitch and slight negative G loading. The control column is all the way back and it still pitches over. The rest of it would have been shear terror.

A couple of take outs that you probably didn't note:

"The stick-shaker is both mentally and physically jarring by design to alert the pilot in no uncertain terms to the pending disaster. To have this activated the entire flight would have been exhausting. "

"If the pilots did not reactivate the trim in an ill-fated last attempt to save the plane, then the plane itself overrode a supposedly fail-safe cutoff switch, and the problem runs far deeper than the MCAS system."

"The mistake seems to have made in the MCAS system design, and the rush to call it’s errors “stab trim runway” when the actual MCAS trim situation differs in critical ways."

"The question is rightfully asked whether these pilots should ever have been put in the position to require extreme piloting skill to return an aircraft safely to the ground." 

2 trained airline crews flew into the ground, with a third narrowly escaping the same fate by virtue of a 3rd crew member (who wasn't loaded up and had time to fortunately determine a cause).

There is no simulator training available for MCAS activation.

All very easy to criticise the actions of a crew, after the fact and based only on cold hard FDR information.

To fully appreciate the events the Human Factor aspect has to be looked at, both in the aircraft system designs and the actions or apparent inactions of the flight crew. One of these has proven to be flawed.

 

 

 

 

 

[KevinAu 441]

13 hours ago, cowpatz said:

I'm picking you just looked at the 2D FDR plots and thought "nothing new here".

Once again, I will post for you what they were given before http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/fe8237743be9b8968625835b004fc051/$FILE/2018-23-51_Correction.pdf

Pages 6 and 7 was posted into their manuals. All those terrifying bells and whistles you mentioned should have rang a bell, if they had read the emergency notice.

Edited April 11, 2019 by n4gix
PLEASE DO NOT QUOTE THE ENTIRE MESSAGE YOU ARE REPLYING TO!

[Wink207 78]

15 minutes ago, KevinAu said:

Once again, I will post for you what they were given before http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/fe8237743be9b8968625835b004fc051/$FILE/2018-23-51_Correction.pdf

Pages 6 and 7 was posted into their manuals. All those terrifying bells and whistles you mentioned should have rang a bell, if they had read the emergency notice.

My goodness. It can't be much plainer than that. In light of the Lion crash, I think they might have briefed this before each trip, if not each flight.

[norman s 15]

It would help anyone who is following this topic if someone can explain how the information contained in the FAA document is then shared with the other regulatory bodies and airlines around the world.

Did it get caught up in the US shutdown and Christmas holidays?

[Biggles2010 343]

Reading page 6 of this document, there could hardly be a clearer statement of why this aircraft should never have been certified for passenger operation with only one aoa sensor.

There was some indication previously, that dual sensors, with a comparator system, were available, but only as an option, at great extra cost. Even an mcas failure warning light is an extra cost option, at $80,000.  Boeing knew airlines would go for the lowest possible cost, regardless of safety, if they could legally get away with it.

I think comments about the competance of the pilots is premature, until all the facts are known and assessed by expert crash investigators. However I do agree with various comments about the limited amount, and type, of hands on flying aquired by many of todays commercial pilots. In particular, they do not get experience in exploring the edges of the flight envelope, apart from occasionally in the simulator. Those who trained in the RAF or USAF in the 60's and 70's will remember what it was like, with much higher risks of mechanical failure, and few of the modern stability enhancements. The skills we aquired may not be so relevant in todays airliners, where everything is more remote, but it did give a great awareness of aircraft handling in difficult situations. The Air France 447 incident comes to mind. But with the current incidents, the real questions must be why Boeing chose not to tell anyone about the mcas system until after the first fatal accident.

 

jbee 

[KevinAu 441]

No, that AD was issued before the shutdown. It’s a letter sent to all known operators. As you can see, it requires flight crews to insert pages into their flight manuals. So they had to have put hands on a sheet of paper or acknowledged an ipad manual update, that alerted them to this issue.

[norman s 15]

Thanks Kevin.

In the summary section the following words appear stating that this AD  was sent to all known US operators and owners of this aircraft.

So I ask again how do we find out how this information was relayed to the worldwide operators.

[KevinAu 441]

34 minutes ago, norman s said:

Thanks Kevin.

In the summary section the following words appear stating that this AD  was sent to all known US operators and owners of this aircraft.

So I ask again how do we find out how this information was relayed to the worldwide operators.

I dont’t know. Good question. You can just subscribe to them online. They would have been under jurisdiction their own country’s aviation department, so it might have to go through them somehow. But Boeing also has a hand in this since it is a change to the flight manual. So I imagine it would be communicated via the product support that Boeing would be providing to Ethiopean even if the FAA doesn’t include them on the mailing list.