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.

 

[Wallache]

The NY times is reporting sensors and warning lights that would detect faulty AoA vanes are sold as extras. Neither of the crashed airplanes had them installed:

Link

 

[waross]

Has the FAA been losing credibility for a long time? Or just in response to their delayed response to ground the Max 8?

The whole approvals process and the delegation to Boeing of critical safety analysis.
The FAA apparently allowed Boeing to self approve an apparently flawed system.
I am wondering also if and when we will hear allegations of improper political influence on the FAA, and the FAA yielding to outside influence.

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

 

[BridgeSmith]

I have a few dumb questions. With the plane in the air at 200+mph, is the angle of attack really that different from the angle of the airplane relative to a level line? Couldn't the AoA sensors be roughly checked against the direction of gravity? Or would it be too rough to be useful?

 

[charliealphabravo]

I have a few dumb questions. With the plane in the air at 200+mph, is the angle of attack really that different from the angle of the airplane relative to a level line? Couldn't the AoA sensors be roughly checked against the direction of gravity? Or would it be too rough to be useful?

I was thinking along the same lines. There must be some other condition that could be used to check or limit the MCAS if the AoA don't agree. I know we had the discussion about the usefulness of GPS earlier but suppose you could check the vertical velocity component of the plane. I mean why should MCAS continue to push the nose down if the aircraft is already traveling downward at over 10,000 fpm (ie crashing).

 

[RVAmeche]

The whole approvals process and the delegation to Boeing of critical safety analysis.
The FAA apparently allowed Boeing to self approve an apparently flawed system.
I am wondering also if and when we will hear allegations of improper political influence on the FAA, and the FAA yielding to outside influence.

For sure, I was just wondering if there were other recent instances/incidents that have been building up against them.

 

[Alistair_Heaton]

For an extremely limited set of parameters you could construct a table of predicted AoA for airspeeds and weights. Basically a 2 D model.


As soon as you go 3d turning it falls over. As soon as you add in power changes it falls over.

And the biggest witch of all is.... Mother nature up drafts down drafts windshear.

Still air, flying straight and level you could do it.

You want to hit certain angles of attack for efficiency. The airspeeds these occur at vary with weight.

And when you get up high you start getting limited by transonic flow over bits of the airframe which is related to air temperature.
As you get higher the air density drops so for a given AoA the speed is higher.

Eventually you hit what's called coffin corner where if you go faster you get sonic shock waves forming taking out the rear control surfaces and if you go slower you go higher than the critical angle of attack.

And the pilots do check ruffly by setting an attitude. We teach it as pitch plus power equals performance. So for each phase of flight we know what the picture should be either looking out the window or looking at the attitude indicator. You don't actually need any instruments at all in a light single engine piston aircraft to fly it once you know what your doing. You set power by ear and you know what speed your doing by the picture.

Actually your structural..... its none linear geometry ;-)

 

[Alistair_Heaton]

The direction the aircraft is pointing bears no relationship at all to what angle of attack of the wings to the airflow.


IN aerobatics we are quiet often pointing 60 degs up but the aircraft is not stalled. During a spin we are pointing straight down to the ground and everything is stall apart from the rudder.

Just look at a video of a loop or a barrel roll. At no point in those exercises is the plane stalled.

This is the famous one.

https://youtu.be/Ra_khhzuFlE

 

[itsmoked]

A MAX8 costs $127M why wouldn't they have three AoA sensors voting? How is the price of an AoA sensor even relevant?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Alistair_Heaton]

Because if they added a 3rd sensor in it would come away from the 1960,s type approval. PLus it would have meant that the pilots would have had to do a bit more than a 2 hour CBT which was a huge selling point of the MAX for NG operators.

I don't know of any aircraft with 3 vanes to be honest. They are about 500$ each on my type. They will be cheaper on the max because they will be the same as on the NG so economies of scale.

 

[itsmoked]

I guess I don't really understand the "type ratings". I don't see how an additional sensor would alter a 'plane type' or 'controls system type'. I could see how a joy stick would change things from a yoke, but a redundant sensor?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Alistair_Heaton]

The whole Avionics would have to get reworked and it would be a first. Normally they buy in the avionics system from somewhere else eg Honeywell. Which has its own approval they just need to get it approved for the type. With that the avionics would have to go through a full approval as well.

The whole type thing is a bit of mystery to me as well and what requires a new rating, what just requires differences training and what just needs a 2 hour CBT.

For my type the Q400 was meant to have hot wings and most of the engine control stuff done through buttons in the cockpit and it could have also had auto throttle.

But it was decided that if they changed that then it would be a new full type rating which is 3 weeks ground school and 10 4 hours sessions in the sim followed by 4 hours skills test and either another 4 hours in the sim or 4 circuits in the live aircraft.

So they ditched them and it was 2 sessions in the sim and a company differences exam for anyone with a DH8-100,200 or 300 rating. the 100 is nothing like the 400.

The whole point of this stretching and using old type approvals is to do as little additional approvals and certification as possible. So they try and use as much from the previous as possible and only then test and certify the changes. If a new requirement comes along as long as they don't change the system that had prior approval then they don't have to include it. eg control disconnects on the 737, for the last 30 odd years your meant to be able to split both roll and pitch in case of control jams and have duplicate control runs. No requirement in the 60's so the Max doesn't have them now some 57 years later from the first 737 approval. The 787 has them because its a fresh type approval. I don't think the 747-800 has them either.

 

[Sparweb]

Because if they added a 3rd sensor in it would come away from the 1960's type approval. PLus it would have meant that the pilots would have had to do a bit more than a 2 hour CBT which was a huge selling point of the MAX for NG operators.
I don't know of any aircraft with 3 vanes to be honest. They are about 500$ each on my type. They will be cheaper on the max because they will be the same as on the NG so economies of scale.

By now, this kind of re-certification could be on the table. Boeing is contending with not NEVER delivering hundreds of aircraft. The stakes are high in this game, in the billions.

The cost of the sensor is hardly an issue. Any number of shops could do a hardware installation kit for one of those, for the cost of a few tanks of fuel. Feeding the data in a meaningful way into the MCAS system, then rewriting the software from SCRATCH because it needs a new criticality class would cost millions. That task is beyond the ability of many players in the aerospace industry except those that routinely develop software-driven hardware. I haven't heard yet exactly who developed the MCAS system (I would expect it to be a 3rd party, but it could be Boeing engineers, not making assumptions) then that team working flat out would need more than a year to finish unless they got a LOT of help and had the eye of Sauron on them.

What Alistair is talking about is commonly called "grandfathering" or the continuation of certification rules from a set date in the past. There are good reasons for this practice to continue because 99% of the time changes are minor or don't affect the level of safety in operations. It's that 1% of the time that the change tips over. When doing certification, we regularly do an evaluation of the scope of any change, to determine if it requires a re-evaluation of the certification basis (Transport Canada in my case, but FAA and EASA have equivalents). If the scope is big enough, we could end up using the latest rules, rather than the rules in place when the plane was certified in the past. Even modifiers of aircraft like me do this evaluation, not just the big OEM's. For example, a substantial change would be the adoption of "fly-by-wire" in a previously mechanically controlled aircraft. Other kinds of changes can skirt the edge (but don't usually cross it) like conversion of the interior from passengers to all-cargo, fuselage stretch, or changing from steam-gauges to an all-glass cockpit. If that cargo conversion includes extending the doors, it tips over. If the all-glass cockpit were to reduce the crew workload that you could eliminate one of the pilots, that would definitely tip over an old plane into a re-evaluation under the current rules. That's how the Q400 was grandfathered despite its stretched fuselage.

I've long been convinced that the 737 of today is drastically different from the 737 of 1970. There are some barriers that would be very difficult to cross if the FAA forced Boeing to apply for a new type certificate, using current certification rules to the 737 airframe. Most obvious to a structures guy like me is that after 1970 the required crash loads were changed, but the 737's are all designed to the original levels. In some conditions the floor structure would need to withstand 50% more load to meet today's requirements. There may be a margin of strength already built into the floor structure, so maybe it wouldn't need to be made 50% stronger, but still SHOWING it and fixing the shortcomings are not trivial jobs. This kind of check-and-fix procedure would be required for all points of the design. It would probably not be worth saving, and cheaper to just start with a clean sheet of paper.

For those still reading, here's a mind-bender: The current practice of evaluating the scope of a design change to see if it invalidates the principles in the original design is a rule that came in during the late 1990's. Since the 737 stretches were also done in the mid-1990's they were completed before this new rule came in. If Boeing had tried the NG's after the changed-product rule was introduced, I don't think they would have had such an easy time.


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

 

[Sparweb]

Follow up:
Has anyone read the Wikipedia page about the Boeing 737?
Has the marketing department of Boeing re-written it recently due to all the attention, or has it always been that way?

Some of the language:

The 737 was originally envisioned in 1964...
The primary flight controls are intrinsically safe...
In the event of total hydraulic system failure or double engine failure, they will automatically and seamlessly revert to control via servo tab...

Only a MBA could write this drivel.

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

 

[Alistair_Heaton]

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.

 

[hokie66]

For someone not involved in aircraft in any way except as an occasional passenger, and as someone who has always considered Boeing planes to be better than the others, I just wonder if there is an explanation as to why these two, or three, events occurred, while there were many uneventful flights. What was different? Are some airline pilots that much better trained so that they avoided or dealt with this issue? Or something else?

 

[Alistair_Heaton]

Human factors is a huge part of most fatalities. It is extremely rare with hindsight that if a human involved had taken another course of action then it couldn't have been avoided. Be that human a loader, technician, ground handler or mostly pilot. Everyone concentrates on the stick monkey. But us stick monkeys know fine that its a huge team effort to get an aircraft from A to B safely. Every human is different and training does play a major role and culture of the nationality and company. Some of us can trouble shoot faster than others but there is a limit to what you can expect the human body and brain to process. We are not super human and not all pilots can have the talents of Neil Armstrong. The aircraft design has to be for the lowest acceptable level of operator.


Now we usually look at the fatalities per million flight hours for an aircraft type.


In its day Concorde was considered the safest aircraft type out there because it had zero fatalities. Then one day one crashed and killed everyone. Because it used to fly relatively few hours compared to other types it went from top of the list to technically the worst in the space of 8 mins.


The MAX has been flying less than 2 years and 376 have been delivered. lion air 189 SOB and Ethiopian 157 SOB. Being extremely generous calling it 18 hours per day for 2 years for each airframe gives 5 million flight hours (this is extremely high its more than likely barely half that).


But that gives fatality rate of 69 fatalities per million flight hours. That is ridiculously and unacceptably high even with my completely inaccurate but very forgiving estimated number of flight hours flown already.

Bombardiers CS300 have produced less aircraft in that period the type flys less hours and has a fatality rate of 0.

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.

http://www.planecrashinfo.com/cause.htm

As for if Boeing is better... its historical and company mergers and changes in managment mean that in some ways they are living off the reputation of a completely different company in history. These days there are no bad aircraft that are allowed to fly. They all have there plus points and thier bad. Economics is a major factor in the type a company wants to use. Its the same with the engines as well which are counted seperately to the Hull. In some ways Rolls Royce is living off historical reputation these days as well.

I have never flown an Airbus or a Boeing but wouldn't have a problem if i was told it would be my next type. If i was told it was going to be a MAX I think I would go and get another job thanks very much. 757 with RR engines I think I would have a little sex wee and be prancing around like a 5 year old for days. A380 i would be looking for another job as well but for other reasons.

 

[LittleInch]

Alastair and spar Web. Your contributions are what makes this site worth reading above many others and thankyou for your insights.

Hokie, It seems to be a result of faults in sensors. Their failure rate is known and the more hours that are flown the more likely one will fail. Apparently in the previous lion air flight it took a jump seat pilot who was able to detach probably a bit from the overload on the pilots to see what was going on and recommend a course of action that averted disaster. He wasn't there on the next flight.

Now why that sensor wasn't just replaced and not just "tested" is all down to time cost and possibly low amounts of data going from pilot to maintenence technician as to the fault.

I think the key issue here is that previously a faulty AoA sensor might have kicked off anti stall warnings but not taken control of the aircraft without warning or apparently knowledge of the pilots.

Why was that MCAS required? To keep the flying characteristics of the airplane close to its predecessors after doing some fairly major engine mods which have significantly impacted those characteristics.

It's getting pretty clear that there is no quick fix for this issue but I just can't see Boeing going back to the drawing board. It could break the company.

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

 

[Alistair_Heaton]

In the old days they used to have flight engineers to sort things like this out. All the stick monkeys had to do was keep the thing flying.

The FE ran all the checklists and had a far great depth of knowledge about the systems than pilots do.


But as the designs got better, the quality of the parts got more predictable they got rid of the position. Some may say bring them back but before that we used to have navigators and radio operators. Do we really need them... No.

The accident stats have been decreasing since they have all been removed.

But now the focus is on the Design of the aircraft. If its not fit to be operated by two normal human beings it shouldn't be in the sky.

And I to like this site because it doesn't have the nationalistic willy waving that normally occurs in aviation. Plus although I am a stick monkey I still think like an engineer and its nice to talk about these things with people that think the same way.

 

[Alistair_Heaton]

That's the cancellations for orders started.


I suspect they may end up scrapping the model.

 

[verymadmac]

From my rough understanding (Feel free to correct me, my exposure to FAA is from 10,000km) the rot at the FAA started in the 80's with giving increasing power to the manufactures, and accelerated over the last 10 or more years with chronic under-funding, unstable funding and regulation lag. There also appears to be issues with their ability to take action on unsafe aircraft, at least one AD's on Robinson helicopters blade delamination only came about after a pacific island nation spent a large amount of cash fishing all the blade bits out of a lagoon, the accident investigator was left with the impression that the FAA knew all about the critical issue but couldn't take action without evidence from an outside source, mention was also made about underfunding of NTSB resulting insufficient number of accident investigations being undertaken.

Human factors is a huge part of most fatalities. It is extremely rare with hindsight that if a human involved had taken another course of action then it couldn't have been avoided. Be that human a loader, technician, ground handler or mostly pilot.

Depressingly much of what tends to be credited to pilot errors is actually bad design, the Airbus side sticks is a classic come to mind.

For those that actually would like to see the certification quagmire that the 737 max has become the applicable section starts at page 68, the certification table starts page 71, on other aircraft it can consist of little more that all regulations applicable at amendment XX.

https://rgl.faa.gov/Regulatory_and_...179CDACD213801658625832A006B2E37?OpenDocument

The certification grandfathering drives some odd features, on a cargo converted 737 classic the cargo barrier is at a higher certification amendment than the aft side of the cockpit wall, so after a few attempts the conversion design orginisation put aft facing seats for the supernumerarys on the cockpit wall because ones forward facing ones on the cargo barrier would have been to a higher spec and hence more expense.