Boeing 737 Max8 Aircraft Crashes and Investigations [Part 2] 44

[Alistair_Heaton]

This thread is a continuation of:

thread815-445840


****************************
Another 737 max has crashed during departure in Ethiopia.

To note the data in the picture is intally ground 0 then when airborne is GPS altitude above MSL. The airport is extremely high.

The debris is extremely compact and the fuel burned, they reckon it was 400knts plus when it hit the ground.

Here is the radar24 data pulled from there local site.

It's already being discussed if was another AoA issue with the MCAS system for stall protection.

I will let you make your own conclusions.

 

[Alistair_Heaton]

"Depressingly much of what tends to be credited to pilot errors is actually bad design"

And a bad design in the first iteration is known about but progresses through 50 years with soft fixes through training instead of changing it through grandfather rights because it is cheaper than fixing it because the end user pays for training.

 

[AidanMc]

That's the cancellations for orders started.

That's interesting. I work at a sub tier supplier and our parts end up at Boeing (among others) on the 737s. We happened to have our quarterly all hands meetings yesterday and there was a long discussion about the 737 Max issues. We were told that our customers have contacted us and assured that there would be no drop in purchasing and to continue manufacturing at full steam.

Could be because we're a long lead time product... or just because they want to make everything sound rainbows and sunshine. It's worrying to see cancellations for aircraft though.

 

[itsmoked]

I thank you too Alistair and Spar as you've provided me with a lot more understanding of "ratings".


Changing subject a bit. If the MAX has 2 AoA sensors why didn't the calm one win? Software should've detected one changing radically and flagged it while ignoring it. Or is there only one?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Alistair_Heaton]

Garuda is wanting to cancel 49 airframes.

5 Billion $ order.

Vietjet is also talking about it 25 billion 200 airframes

And Kenya Airways 30 aircraft at 3.56 billion.

If your parts are generic for NG and MAX product lines I can't see a drop in demand. People are not going to stop flying the NG's because of an issue with the MAX. If your product is only used in the MAX which i doubt because you state among others then things might change quiet quickly. I really can;t see it getting released for service any time soon.


"Changing subject a bit. If the MAX has 2 AoA sensors why didn't the calm one win? Software should've detected one changing radically and flagged it while ignoring it. Or is there only one? "

There is two but apparently the MCAS only uses the input for one of them per flight and flips between them.

Normally in my experience with avionics miss matches its left to the pilot to decide which one is given the correct reading. What the machine does is flag up that there is a comparison error between them. Problem with AoA is that is quiet common that they do disagree depending on the manoveur being under taken.

Airsped a difference of 5 knts, Altitude is 60 ft, heading is 6 degrees all trigger comparison errors. When it happens we look at the standby system and say that one is shite twiddle a knob and the display turns yellow to indicate we are in a reduced tolerance mode and manually have to check the standby system and we continue the flight. BUt again we are back to this 3 system redundancy. I can take one AoA vane out of the system but it requires a circuit break pull. We know its happened because we get a caution about icing gear and the low and high speed protection goes nuts. So we get a spurious stick shaker or over speed warning. After we pull it, it then takes the stall protection system out of action and I can't fly in icing conditions any more. And the plane is gorunded as soon as I land.

To note i don't know of any commercial aircraft being flown today apart from test bed aircraft and those ones that fly into hurrancanes that are fitted with more than 2 AoA sensors.

 

[LittleInch]

It smoked - It appears as if the MCAS only takes readings from one sensor (I think the one on the captains side). As I understand it the two halves of the cockpit - Captain and FO have a set of readings from units on their own side of the aircraft. with only two sensors, if there is a difference between them then there are alarms - the issue here is that the alarm for AoA disagree was an extra which the planes didn't have fitted (!!) Apparently these are now going to be part of the basic pack, but the actual AoA readings are still going to be an extra. Someone in Boeing is making a bad call there.

IIRC the Lion air crash reported that the captains unit was the one failing and he had the "stick shaker" active for most of the flight whereas the FO didn't.

The issue with one failing and one not is that the computer doesn't know which one to look at and hence you get all these disagree alarms going off. There is also some evidence that the airspeed signals may have been affected also from the reports from the pilot of Ethiopian airlines, but that isn't clear.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Alistair_Heaton]

It flips between which sensor apparently so the first flight of the day is the No 1 sensor which is the captains side then the next flight the Number 2 on the fo's side. Repeat until there is a full shut down power cycle. Which I suspect was half the problem trying to work out what was going on with the lion air aircraft. One sector the crew was saying everything was fine the next they were saying they were getting trim runaway. Get back to the ground turn the aircraft off, tech come to fix it and turn everything on to test it and everything works fine. Next flight departs and everything good pilots think great they have fixed it..... Second flight the shit hits the fan again because it was the other one that was faulty.

There needs to be a way of killing the MCAS without loosing the electrical trim system as well. That will take the Autopilot away as well.

 

[LittleInch]

Thanks - kind of makes sense that it would alternate inputs.

I thought that an MCAS disable button would be the way to go, but that clearly has issues in that it is a critical part of the flight control system and would disable what was designed as an important safety element in preventing an accidental stall condition. Hence solve one problem and potentially create another one.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Alistair_Heaton]

To be perfectly honest...

Overspeed is usually the problem with pilots.... not stalling....

I would have thought that a logic system turning it off for take off until 10 000 feet would be feasible.

 

[BridgeSmith]

With the AoA sensors being such a cheap but critical part of a system that has the potential to cause the plane to dive into the ground, I would think that mounting maybe 3 of them on each side would be a prudent move. That way, if one fails, there should still be 2 in agreement that it can use. It there aren't at least 2 matching, then the system warns the pilots and bypasses the stall avoidance subroutine.

 

[SwinnyGG]

It's not quite that simple.

The AOA sensor is an aerodynamic device. Similar to a pitot tube, It needs 'clean' air, to function; it also (unlike a pitot tube) needs to be positioned in a specific way relative to the aircraft pitch axis to function correctly.

In other words, you can't just add more next to the existing ones.

Even if you could, the software and processing hardware that interacts with these devices is, by definition, extremely complicated. The amount of traffic between systems and the level of internal code error checking is extreme. This is why third parties provide the systems- design is complicated and expensive.

Adding $2,000 of additional AoA sensors might mean literally tens of millions of dollars in additional engineering time to make them function safely. That's no comfort to the families of those lost on these flights, but it's an economic calculus that is being made either way.

 

[John2025]

Not that the AoA sensors are the only thing that could go wrong with the MCAS, but is there a way to manually switch the sensor that's being used or does it only switch automatically at each take-off?

 

[waross]

A lot of people will not fly on a 737 Max8.
Additionally a lot of not sophisticated but worried people will not want to fly on any 737 or Boeing.
The reasoning being;
"I really don't know the difference so to be safe I won't fly on any Boeing."

Does anyone remember Konica? Not Konica Minolta.
Konica had a fully automatic, dependable and easy to use automatic exposure system in the 60s.
Konica's Hexanon lenses were among the finest lenses in the world. Hexanon lenses were selected by the Japanese government as the standard against which all other lenses were compared.
Then they released the TC-X 1985 – 1987
Built by Cosina for Konica, fully mechanical SLR
First SLR in the world with body and frame completely cast of plastic
First camera in the world that uses DX coding on film cartridge for film speed setting
It was reported that the plastic turned out to be not dimensionally stable.
This may have been public perception rather than the actual reality, but was told to me by a dealer when I questioned why Konica cameras and lenses were no longer stocked.
Konica lost the trust of their dealers and lost their dealer network.
While cameras were not Konica's only business, the model TC-X marked the end of their single lens reflex camera business.
20 years of fine cameras and one bad model ended their camera business.

If the Max8 leads to a re-evaluation of the FAA's grandfathering provisions this could mark the beginning of the end of the 737.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[saikee119]

If the Max8 leads to a re-evaluation of the FAA's grandfathering provisions this could mark the beginning of the end of the 737.

Since FAA does not have the resources it can sublet the new evaluation back to Boeing. Have we not seen it before?

 

[waross]

Alistair;
With respect. I am not challenging you. I am wondering whether my reasoning is valid.
I read with interest your link to planecrashinfo.
I note that mechanical failures account for only 17% of fatal aircraft accidents.
That would make the statistics on the MAX 8 about 6 times worse than your initial estimate.
But the general heading of Mechanical includes;
Engine failure,
Equipment failure,
Structural failure,
Design flaw,
and
Maintenance error.
So design flaw is only a part of the 17%.
In the table:
Notable Accident Causes by Category,
Design flaws account for only 17 out of 212 accidents.
That's down to 8% of total accidents.

The average value for airline transport or part 121 as the FAA calls it is 4.03 fatalities per million flight hours 1998 to 2007.

May we apply the 8% to this figure and call the average number of accidents due to design flaws 4.03 x 8% = 0.322 fatalities per million flight hours?
Now we are over 12 times worse than your initial estimate.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Lou Scannon]

waross, from a philosophical perspective, I would say design factors play a role in EVERY crash. Think of a simple hand launched balsawood toy glider, built with adequate structural safety margin, ample dihedral, positive longitudinal stability, and balanced aerodynamic surface areas. Almost nothing could make it crash. By contrast, the MCAS, for example, based on one faulty input, if left to its own devices is guaranteed to crash the aircraft, and appears to have done so on two occasions against determined human opposition.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[Alistair_Heaton]

Crashes are very rarely down to one factor.

Usually it's something then the human gets involved and due to their actions it becomes worse than it needs to be.

In fact I just had the qrh out this evening with a problem in flight with pressurisation. We spotted it while climbing so leveled off at FL 100 and had enough fuel to get to destination so continued. Got a new fuel plan for going back at fl100 and the technicians replace the rear outflow valve when we got back. A complete none event. If we hadn't spotted it on the way up before the warnings went on then we would have had to descend quickly with the cabin crew doing service hot drinks going every where and trolley s rolling about with the nose pitched down by 20 Deg.

You can't fiddle with the fatalities per million flight hours. It's a straight number unrelated to the other accident figures. Most accidents these days are aircraft going off the runway and nobody killed.

 

[davidbeach]

Crashes are very rarely down to one factor

From what I’ve read about air and rail crashes and what I’ve personally determined about electrical system events I’d change “very rarely” to “never”. Even two factors is astronomically improbable.

 

[waross]

Thank you Alistair.

I agree David. Three contributing factors is more common.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

Hemi;
I was comparing numbers under headings in a table. All crashes versus crashes due to design flaws.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Sparweb]

Thanks SparWeb Would you by any chance have anything to do with the Q400? I am due to move to the CS300 some point this year.

Not directly, but plenty of 100's to 300's cycling through the facility where I work. My resumé has a lot of spicy stuff on it.
Oh I envy you - I'm sure you're very eager to switch to the C-Series jets!
What about the other Bombardier stable-mates, like the RJ's and Globals?
Or are you having so much fun (not) doing emergency descents in the Q400 that you never get the chance?

A quick summary on my thoughts about a few of the recent comments:

I doubt that the grandfathering clauses would be changed by this accident. Restricting that would make many aircraft operations completely uneconomical. By an order of magnitude. Not exaggerating! It would prohibit any aircraft more than 10 years old from flying.

The causes of this accident appear (for now) to be from a bad system safety assessment, and that (for what it's worth) is a very modern process of analysis. It didn't exist in the 70's but definitely a hallmark of modern avionics design now. The failure, we believe, is in the assumptions and conditions used for the analysis. If the audience is willing, I can show you some of these assumptions and how they affect certification of equipment. Takes a bit of time to prepare that kind of stuff, but I'm game if you are.

I take the comments by Chesley Sullenberger seriously. If he's concerned about the FAA, then so am I. My work has some projects that have been stalled by the US government shut-down and a few still are delayed, and I'm not even working at a US company! It must be awful for the aviation companies that are in the USA.

Canada (and EASA) may be taking upon themselves a detailed review of the 737 certification. I hear rumours and rumblings from Transport Canada that a resources are being moved to respond to this need. Last fall I attended a seminar on system safety analysis (only 1 week after the Lion Air crash; how I regret not asking a question at the time). The expert at Transport Canada had a tremendous grasp on the issues and methods to assure reliability of modern aircraft systems. He was even reporting some cutting edge stuff like avionics faults that could be traced to cosmic radiation (facts to back it up!). I have a lot of faith that if people at TC, like this person, are evaluating the Boeing certification, independently of the FAA, that we WILL get to the bottom of these tragedies.

It just can't happen quickly enough.

No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
STF