UKC

Boeing

New Topic
This topic has been archived, and won't accept reply postings.
 Pefa 13 Mar 2019

Why did it take Trumps intervention before the FAA did the proper thing ? It makes them look a bit callous does it not?

https://www.rt.com/news/453753-us-faa-boeing-max/?utm_source=browser&ut...

Post edited at 20:17
4
 jkarran 13 Mar 2019
In reply to Pefa:

I doubt trump had anything to do with it, more likely preliminary ideas coming out of a review of atc tapes, radar records and some discussions with their lawyers.

Jk

OP Pefa 13 Mar 2019
In reply to jkarran:

China and SA were first to stop any flights using the Max 8 then it took a few days before the EU joined them and now it takes another couple of days before the US stops them.When anyone can clearly see there looks as if there is an issue with these planes and in the interest of public safety no chances whatsoever should be taken.

Imo it makes Boeing and the FAA look if not very callous then out of touch and why?

Pure Market forces ? Their decision not to ground the planes at the same time as China ie. Immediately looks terrible from a PR point of view. 

Post edited at 21:31
 summo 13 Mar 2019
In reply to Pefa:

I think the US Boeing's already have an update and additional pilot training since the first crash(lion air). It's seems there are automated features which when they kick in, pilots need to over ride or deactivate immediately(according USA aviation man on r4 this morning). 

1
In reply to summo:

> It's seems there are automated features which when they kick in, pilots need to over ride or deactivate immediately

If this is the case then a major oversight by Boeing, I can (vaguely) remember being taught the dangers of automatic control systems in uni over 20 years ago. Some of the examples we studied where aviation based.

 tehmarks 13 Mar 2019
In reply to summo:

From what I can discern, MCAS is an automated system that, up until the Lion Air crash, pilots weren't even aware existed. As I understand it, in certain aircraft configurations (autopilot off, flaps up), at high angles of attack (ie approaching the stall) it commands a pitch down (to avoid a stall). Except it doesn't do this by actuating the primary flight control - the elevator - but by winding in a few degrees of nose down trim in over ten seconds. The pilot manually trimming during that period will cancel the trim, but after some seconds it reactivates. Thus, if the pilot isn't careful, he can end up with an aircraft will full nose-down trim - which will want to point itself at the ground without huge back pressure on the controls.

Of course, this should only occur approaching the stall - but it would seem that if the AoA sensor feeds garbage data to the system in question, it will, well, produce garbage trim commands. It relies on a single sensor (doesn't cross-check values between the two physical AoA sensors), doesn't take into account any other instrumention, nada. So if a single AoA sensor misbehaves, a system which pilots didn't know existed at first will try to pitch the aircraft into the ground. Imagine that happens on the initial climb out?

'cking stupid design, right?

(Thanks mostly to PPRuNe for the MCAS information)

 tehmarks 13 Mar 2019
In reply to tehmarks:

And just to clarify for the non-pilots on the forum:

Trim is a way of counteracting undesirable control forces, so you don't have to continue to hold the controls in one position to maintain the desired attitude. You trim once flying at an attitude and speed you're happy with - you never use trim to actually point the aeroplane in the direction that you want it to go. MCAS is actually flying the aircraft through trim, and is thus completely backwards to one of the fundamental concepts of flying.

If MCAS applied forward elevator, the pilot would be able to counteract it with no increased effort to normal, as the aircraft is in trim. The consequence of MCAS using the trim as a control surface is that for the pilot to counteract it he'll need to pull back much harder on the controls - figures of about 60kg were mentioned on PPRuNe for a 737 flying with full nose-down trim. It's not hard to envisage a situation where it becomes virtually impossible to control the attitude of the aircraft, at least in the short term. Again, not so good on the climb out...

 wintertree 13 Mar 2019
In reply to tehmarks:

> 'cking stupid design, right?

Yup.  But it’s downstream of some other ruddy stupid design decisions as I understand it.  If they’d not built a badly balanced aircraft they wouldn’t have to have bolted on a new safety system, and thus wouldn’t have created the opportunity to muck it up.

 tehmarks 13 Mar 2019
In reply to wintertree:

Well, exactly. Fundamentally they're fixing a physical design flaw with software. Which is a terrible approach.

 jkarran 13 Mar 2019
In reply to Pefa:

Chances are always taken. What airlines and regulators don't like is not understanding them.

Yes market forces are at play, that's life.

No it's not clear there is a design issue, there may be and that needs to be considered carefully, we didn't make commercial aviation as safe as it is by jumping to conclusions.

Bear in mind China and the US are embroiled in a trade war, Boeing is big news in the US and China has an emerging airliner industry. China wasn't hasty purely through an abundance of caution, the stars aligned for them here.

Jk

 jkarran 13 Mar 2019
In reply to wintertree:

> Yup.  But it’s downstream of some other ruddy stupid design decisions as I understand it.  If they’d not built a badly balanced aircraft they wouldn’t have to have bolted on a new safety system, and thus wouldn’t have created the opportunity to muck it up.

I think the reduced static margin and the weird thrust couple are just the consequence of prioritising efficiency and design re-use. Not really unreasonable.

Jk

 jkarran 13 Mar 2019
In reply to tehmarks:

> Well, exactly. Fundamentally they're fixing a physical design flaw with software. Which is a terrible approach.

Debatable. Airliners have long been hybrid systems and there are big performance gains to be had through the judicious application of control.

Jk

 profitofdoom 13 Mar 2019
In reply to tehmarks:

> ....MCAS is actually flying the aircraft through trim, and is thus completely backwards to one of the fundamental concepts of flying. .. for the pilot to counteract it he'll need to pull back much harder on the controls - figures of about 60kg were mentioned on PPRuNe for a 737 flying with full nose-down trim. It's not hard to envisage a situation where it becomes virtually impossible to control the attitude of the aircraft....

And just to clarify even more for non-pilots ha-ha.... OH GREAT, so I'm sitting in my airplane seat just after take-off clutching my teddy bear then the plane suddenly noses down and smashes into the ground at a zillion miles an hour because of a design fault and because the poor pilot can't pull with a force of 60 kg? Thanks a lot Mr. Airline

 Neil Williams 13 Mar 2019
In reply to tehmarks:

> Well, exactly. Fundamentally they're fixing a physical design flaw with software. Which is a terrible approach.


It's an almighty bodge job which was put in to fix a serious problem on the cheap - not something that should be done with aircraft.  But what really beggars belief is not telling pilots about it - whoever made that decision in the airlines where they did has blood on their hands.

Can anyone else think of a relatively recent case where quite a lot of people died because of something done incorrectly on the cheap?  I can.

We really could do with ending the "race to the bottom".

Post edited at 23:54
 tehmarks 13 Mar 2019
In reply to jkarran:

This isn't judicious application of control as far as I can tell, this is a bolt-on to fix some problematic characteristics of flight because the FAA wouldn't otherwise certify the design.

 Neil Williams 13 Mar 2019
In reply to tehmarks:

> This isn't judicious application of control as far as I can tell, this is a bolt-on to fix some problematic characteristics of flight because the FAA wouldn't otherwise certify the design.


Agreed.  I've heard people compare it to the Airbus flight laws but it's nothing like them - in particular they "rein things in" by limiting what can be done rather than directly counteracting what the pilot is attempting to do.

What it certainly is is totally counter to the Boeing flight control philosophy of "the pilot knows best".  In that regard Trump had a bit of a point.

Post edited at 23:55
OP Pefa 14 Mar 2019
In reply to jkarran:

> Yes market forces are at play, that's life.

> No it's not clear there is a design issue, there may be and that needs to be considered carefully, we didn't make commercial aviation as safe as it is by jumping to conclusions.

> Bear in mind China and the US are embroiled in a trade war, Boeing is big news in the US and China has an emerging airliner industry. China wasn't hasty purely through an abundance of caution, the stars aligned for them here.

> Jk

I get that it is going to cost Boeing a lot to pay the carriers compensation whilst their planes are not flying so they will want to minimise that but it makes them look very bad and it makes the FAA look bad to. 

You ask anyone and they would say the chances of two brand new planes with the same new design crashing on the same manoeuvre within 5 months of each other must be astronomical. So I'm sure everyone would say ground them immediately just like the cabin crew unions and pilots unions were advocating but Boeing and the FAA went against this and I think that will put a big dent in their reputation when they should have taken the financial hit and saved themselves any further bother.

Now they have a tec problem with these planes and additionally a bad rep for dismissing everyone's concerns.

Dreadful PR. 

OP Pefa 14 Mar 2019
In reply to wintertree:

> Yup.  But it’s downstream of some other ruddy stupid design decisions as I understand it.  If they’d not built a badly balanced aircraft they wouldn’t have to have bolted on a new safety system, and thus wouldn’t have created the opportunity to muck it up.

Why is this plane badly balanced (Excuse my aeronautical ignorance)? 

Post edited at 00:10
 tehmarks 14 Mar 2019
In reply to Pefa:

Because they've rehashed a 1997 airframe based on a 1984 airframe based on a 1967 airframe and stuck ruddy great new engines on it, among other modifications, which necessitated moving the engine position and height relative to the wing, which it would seem has led to some undesirable aerodynamic consequences.

Post edited at 00:44
In reply to tehmarks:

... namely, to reduce the longitudinal stability of the aircraft. Making it less stable than any previous model of the (very old) 737 design. If I understand it correctly, Boeing 'solved' this problem by adding a software system (the MCAS) which automatically messes about with the trim, unknown to the pilot. If the pilot uses the normal controls, or even the manual trim controls, to try and counteract it ... in a situation that is going wrong ... this wretched system keeps on kicking in, until the forces are too great for the pilot to recover the aircraft by any means. Worse, again if I understand it correctly, this system has been added without all/most pilots knowing anything about it, let alone being trained to deal with it. My big question is, can this wretchedly dangerous automatic system be (easily) turned off by the pilot, if things go badly wrong, so that he can regain control of the aircraft completely manually. It sounds ... from what we've heard so far about pilots not even being made aware of this new MCAS system ... that they cannot, or at least don't know how to, simply because they haven't been told about it. 

It sounds as if this 737 Max is a crap design, or rather umpteenth upgrade/upscaling of an ancient design (dating incredibly from 1964, first flight 1967), that really should be scrapped. Like the whole embarrassing thing of continually upgrading Issigonis's 1959 Austin Mini, and trying to turn it into a car fit for the second decade of the 21st century.

2
 Frank4short 14 Mar 2019
In reply to wintertree:

> Yup.  But it’s downstream of some other ruddy stupid design decisions as I understand it.  If they’d not built a badly balanced aircraft they wouldn’t have to have bolted on a new safety system, and thus wouldn’t have created the opportunity to muck it up.

My understanding from a mate who's an aeronautical engineer is one of Boeing's fundamental design concepts is that their planes are low, as in physically close to the ground, compared to say Airbus. When they increased the size and capacity of the 737 to the Max8/9, they also had to increase the engine size thus leaving them with insufficient ground clearance. Rather than re-design their landing gear to provide additional clearance they moved the engines forward and up on the wing, thus creating an aircraft that is fundamentally out of balance and as such needs additional control equipment, MCAS, to fly stably under certain conditions. However as the authority self certifying their planes they kept quiet about it. 

Essentially it's a botch to cheaply cover a fundamental design flaw, rather than re-designing major parts of the aircraft. 

As to the earlier comments about Trump grounding the fleet. You can be sure as the president who doesn't give a shit he initially talked to the FAA with a view to keeping them flying. It would only be public opinion changed sufficiently with more knowledge that he changed his position. 

2
 wintertree 14 Mar 2019
In reply to jkarran:

> I think the reduced static margin and the weird thrust couple are just the consequence of prioritising efficiency and design re-use. Not really unreasonable.

I work with biotech systems.  Like airplanes, these are complex, interrelated and non-linear systems.  I’ve become really cautious of the “successive problem solving” mentality that stems from design re-use.  Each change and it’s reactive fixes seems minor but over the years they take you far from the happy place where you originally really understood everything.

Rigid Raider 14 Mar 2019
In reply to Pefa:

The three posts above have it; Boeing have fundamentally changed the balance of the aircraft. A friend who is a flying instructor says it already has a reputation for tail strikes. He also says Boeing were in a hurry to get the new model into the market because China is developing an equivalent, so they didn't want to launch it as a new model, rather as a patched up modification. Inexperienced pilots are having difficulty in coping with the anti-stall MCAS system bolted onto the plane and both Lion Air and ET showed erratic rates of climb and both pilots asked to return to the airfield. 

 jkarran 14 Mar 2019
In reply to Pefa:

> You ask anyone and they would say the chances of two brand new planes with the same new design crashing on the same manoeuvre within 5 months of each other must be astronomical. So I'm sure everyone would say ground them immediately just like the cabin crew unions and pilots unions were advocating but Boeing and the FAA went against this and I think that will put a big dent in their reputation when they should have taken the financial hit and saved themselves any further bother.

As animals we're terrible at understanding risk, it's often counter intuitive and our emotions  run away with us.

> Now they have a tec problem with these planes and additionally a bad rep for dismissing everyone's concerns.

Or not. Jumping to conclusions in these matters is a fools game. All we know so far is it crashed and there are reports of unreliable airspeed. That alone shouldn't down any aircraft in VMC but far stranger things have happened.

jk

Post edited at 09:29
 jkarran 14 Mar 2019
In reply to wintertree:

> I work with biotech systems.  Like airplanes, these are complex, interrelated and non-linear systems.  I’ve become really cautious of the “successive problem solving” mentality that stems from design re-use.  Each change and it’s reactive fixes seems minor but over the years they take you far from the happy place where you originally really understood everything.

Absolutely but there is a feedback loop that keeps us learning about our new franken-systems. In truth we never have full understanding of a new design either, we just don't know what we don't know yet.

New design from the ground up each time might be technically better (if imperfect) but we live in a world of competing pressures, it just isn't viable in most cases. Despite this process of evolution building complexity commercial aviation remains exceptionally safe.

jk

In reply to wintertree:

> > I think the reduced static margin and the weird thrust couple are just the consequence of prioritising efficiency and design re-use. Not really unreasonable.

> I work with biotech systems.  Like airplanes, these are complex, interrelated and non-linear systems.  I’ve become really cautious of the “successive problem solving” mentality that stems from design re-use.  Each change and it’s reactive fixes seems minor but over the years they take you far from the happy place where you originally really understood everything.

That's a paragraph that need including in every undergrad STEM design course, and tremendously useful in coding.

 wintertree 14 Mar 2019
In reply to jkarran:

> Absolutely but there is a feedback loop that keeps us learning about our new franken-systems.

Sure but it’s looking like a pretty tragic looking and unnecessary feedback loop this time.  

I would more generally agree with you but then I went and looked at some pictures of the aircraft this morning.

I’m firmly of the irrational camp that something ugly just isn’t supposed to fly.

Intrinsically unstable aircraft kept balanced on a knife edge by computers are great for high agility air superiority fighters.  Not so smart for passenger aircraft.

Rigid Raider 14 Mar 2019
In reply to wintertree:

Spot on. A pilot on PPRuNe has pointed out that the chord of the new wings is much slimmer than the old, for reasons of speed and fuel economy, meaning the aircraft will have lower attitude stability. The new position of the bigger engines and the longer fuselage have just worsened the instability.

 Offwidth 14 Mar 2019
In reply to wintertree:

Good advice.

I also remind people in my control classes how past control problem caused or involved incidents were too often initially blamed on the human element (with clear influences of avoiding liability or impact on share prices). I'd put money on this being a systematic problem with that plane type that Boeing will now need to fix.

1
 Greenbanks 14 Mar 2019
In reply to Neil Williams:

<I've heard people compare it to the Airbus flight laws but it's nothing like them>

For a start, Airbus designed-in the fly-by-wire approach, and have refined this over many years. They now represent state-of-the-art in avionics IT application (still not infallible, but certainly designed pro-actively.

The Boeing v Airbus debate has been on the go for several years (see, for instance Pprune - https://www.pprune.org/questions/542950-boeing-737-vs-airbus-a320-advantage... ).

Some of this relates to the latter's fly-by-wire technology, which has been distrusted (mainly irrationally) by Boeing enthusiasts. What Boeing appear to have done is play catch-up, having realised that e-systems of flight management are here to stay. Unlike Airbus though, they've attempt to cut costs by retro-fitting to an established aircraft design. The results compare very unfavourably with (for example) those aircraft that from design inception had built-in computer-aided management (for instance, the A321 and, in wide body terms, the A350 and Boeing's Dreamliner)

 Neil Williams 14 Mar 2019
In reply to wintertree:

> Intrinsically unstable aircraft kept balanced on a knife edge by computers are great for high agility air superiority fighters.  Not so smart for passenger aircraft.

I would be inclined to agree - or if you are going to have the computer do it, you want something much more like the Airbus control philosophy, whereby the computer does it all and the pilot just pushes the stick the way they want to go.  This is how the Eurofighter works.

The problem here was something that was an utter bodge and kicked in and did something weird right on the edge of the flight envelope - just where you don't want weird stuff being done.

Post edited at 10:54
 jkarran 14 Mar 2019
In reply to wintertree:

> Sure but it’s looking like a pretty tragic looking and unnecessary feedback loop this time. 

Without wishing to sound more callous than usual the learning events usually are tragic in aviation. Whether they're unnecessary is again debatable, things only get fixed once they are both understood and considered too risky to tolerate, that's the inevitable result of the multiple conflicting pressures acting on the industry as it has always been.

> I would more generally agree with you but then I went and looked at some pictures of the aircraft this morning.

I rather like it but I like lumpy bumpy aircraft with odd proportions (747, weird bulged training versions of otherwise sleek fast jets...). Any idea what the radome on the roof is?

> Intrinsically unstable aircraft kept balanced on a knife edge by computers are great for high agility air superiority fighters.  Not so smart for passenger aircraft.

Well it yields a significant drag reduction. If it's done safely I have no conceptual problem with augmented stability in commercial airliners, just as I don't with properly implemented full fly by wire control of an intrinsically stable airframe, both kill you just the same when the computers quit.

jk

Post edited at 11:01
2
 Neil Williams 14 Mar 2019
In reply to Greenbanks:

> Some of this relates to the latter's fly-by-wire technology, which has been distrusted (mainly irrationally) by Boeing enthusiasts. What Boeing appear to have done is play catch-up, having realised that e-systems of flight management are here to stay. Unlike Airbus though, they've attempt to cut costs by retro-fitting to an established aircraft design. The results compare very unfavourably with (for example) those aircraft that from design inception had built-in computer-aided management (for instance, the A321 and, in wide body terms, the A350 and Boeing's Dreamliner)

I'm not sure that's even what they've done.  They haven't added an Airbus-style fly-by-wire system where the computer does the control and the pilot just points it the way he wants to go using a stick that isn't physically connected to the flight control surfaces.  They've added a system to deal - badly - with a situation that arises through errors made in the design, that wouldn't even be there if those errors had not been made, and was just cheaper (but as it turns out deadlier) than fixing those errors properly.

Post edited at 11:00
 wintertree 14 Mar 2019
In reply to jkarran:

> Well it yields a significant drag reduction. If it's done safely I have no conceptual problem with augmented stability in commercial airliners, just as I don't with properly implemented full fly by wire control of an intrinsically stable airframe, both kill you just the same when the computers quit.

It’s not normally about the computers quitting though.

 It’s about this mismatch between a pilot’s understanding and what the computers are doing.  The relevant military pilots are much more heavily trained in flighing in a computer-stability regime, they have ejection mechanisms and they don’t have passengers.

That’s why I’m much less happy finding even conceptual acceptance.  AAIB and the global equivalents’ reports are full of incidents and accidents caused by misunderstanding at the human/computer interface.  Normally there’s time to hopefully resolve it but not if it occurs in the direct flying systems and near takeoff.  

> I rather like it but I like lumpy bumpy aircraft with odd proportions (747, weird bulged training versions of otherwise sleek fast jets...)

I shall not comment.  I think visual design peaked with the Avro-730.  Nice to see a very similar design emerging in Sklyon.  They really nailed it decades ahead of the time with the Avro.

> Any idea what the radome on the roof is?

I haven’t procrastinated that much yet.  Perhaps it’s satellite transponders for internet access.

Post edited at 11:11
 malk 14 Mar 2019
In reply to wintertree:

re human errors i read on  that website that the trim cutoff/MCAS switches go in the opposite direction to the previous 737?

a programming error sounds more likely: https://www.reddit.com/r/flying/comments/b08h03/737_max_mega_thread/eidycdx...

Post edited at 12:48
 malk 14 Mar 2019
In reply to Pefa:

> Imo it makes Boeing and the FAA look if not very callous then out of touch and why?

> Pure Market forces ? Their decision not to ground the planes at the same time as China ie. Immediately looks terrible from a PR point of view. 

UK's CAA was pretty late on the scene (tuesday)-quite callous?

Post edited at 13:04
OP Pefa 14 Mar 2019
In reply to malk:

> UK's CAA was pretty late on the scene (tuesday)-quite callous?

I take points made above that grounding these planes by the authorities and Boeing without having solid proof to justify that action, could result in egg removal from corporate faces as well as big financial losses if the causes of both crashes were not through the negligence of the manufacturer and those who gave it a quality stamp of approval.

The probability of these terrible incidents not being related however seems obviously very low even to those of us with zero knowledge of aeronautics. So for these institutions to tell the world it is safe flies in the face of the evidence, making people lose their trust in not just the quality controls but in the judgement of these institutions.

Yes it makes it look as if they have swapped a responsibility of safety for their customers for a responsibility of covering up to protect their short term share value.

And yes the UK authority (and the EU etc) is included as they should have immediately grounded these planes to reassure a worried public that it put them first and not the corporation. I could be wrong but the British authority must have asked their US counter parts and took their lead rather than be fully independent. 

Personally I enjoy flying and do it quite often (too much lately) but even I and people I know have started to wonder about safety given these two very tragic incidents. Public perception at these times is heightened so correct decisions by those in control are crucial for trust. 

Post edited at 15:56
 Michael Hood 14 Mar 2019
In reply to Pefa:

On a different note how much is this going to cost Boeing if the crashes do turn out to be caused by error of poor design followed by poorer designed software fix.

Direct compensation to victims' families, loss of orders for 737-max and other planes, and drop in share price.

Could it be enough to put Boeing out of business?

I seem to remember that in the old days, critical systems in aircraft were triplicated, but each of the 3 systems were designed by different "independent" teams. Critical safety "decisions" would usually be 3-0 but any disagreement would be settled by majority vote. The thinking was that the different designs would be less likely to have bugs in the same place at the same time.

From (I think) the Boeing 777, they just had one team designing a single system that was triplicated. The thinking being that with all the expertise being in one team, bugs would be much less likely.

Oh great I thought, so now a bug can cause a crash with a 3-0 vote

Roadrunner6 14 Mar 2019
In reply to Pefa:

Trumps got a bee in his bonnet about automation.. yet planes are massively safer.

It sounds like Boeing know what went wrong, not that it helps if airlines haven't sorted it yet.

TBH though I think he was right here. They have the recorders so it should be quick.

Roadrunner6 14 Mar 2019
In reply to Michael Hood:

"Could it be enough to put Boeing out of business?"

Not a chance.. huge military business.

This is 2005, so its probably even more military biased.

Total revenue for 2005: 

in $millions 
Commercial Airplanes : $22,651 

Integrated Defense Systems : $30,791 
IDS Breakdown- 
Aircraft and Weapon Systems : $11,444 
Network Systems : $11,264 
Support Systems : $5,342 
Launch and Orbital Systems : $2,741 

Boeing Capital : $966 

Other(Connexion,Boeing Tech) : $972 

Source : http://www.boeing.com/companyoffices.../annual/05annualreport/05AR_20.pdf

Correction: https://www.washingtonexaminer.com/business/boeing-sales-top-101-billion-am...

This suggests commercial now leads military defense, but still two monster parts.

Post edited at 17:49
OP Pefa 14 Mar 2019
In reply to Roadrunner6:

Are you saying like banks, they are too big to fail?

If people think they are incompetent and dangerously negligent then how do you restore that trust again?

I seen an American article that quoted Boeing in which it stated that when this new software patch is installed the chance of the situations which may have led to these crashes happening again will be less.

Even that statement does not inspire confidence as I don't want less chance of that happening in future I want no chance of that happening in future. 

Pan Ron 14 Mar 2019
In reply to Pefa:

It's two crashes, on a very specific aircraft.  The rest of the 737 line is perfectly fine to fly and the Max is not particularly common.  So if it came to it, it wouldn't be a withdrawal of a large number of aircraft and can no doubt be solved easily by software fixes, bulletins and training - far less stable aircraft have been safely computer controlled for decades.  Confidence in the basic airframe of 99.9% of the 737 fleet is already there.  So I'd imagine Boeing will shrug this off.

In reply to Pan Ron:

Not so sure about this. I think Joe Public is going to take a lot of convincing before wanting to fly on this 737 Max again. As Pefa has said, they're going to have to come up with something better than 'reducing the chances' of it happening again.

1
cb294 14 Mar 2019
In reply to paul_in_cumbria:

This. Many years ago when fluorescent DNA sequencers hit the market we were using some piece of homebrew legacy software  to read the sequencing traces. The algorithm was a bit shit, missing mistakes caused by the chemistry that were easy to spot by eye. 

I was therefore working with a bunch of informatics students who thought they could improve the base calling algorithm for their masters project.

The gave up in desperation, after finding sections in the code that were recognizably compiled from three different higher languages, while the origin of the rest was completely unidentifiable.

Better to bin it and start from scratch...

CB

 jkarran 14 Mar 2019
In reply to Gordon Stainforth:

People have short attention spans, even those that are interested and listening. I'd wager nobody contributing to this thread will read the accident report or be talking about the 737max in 6 months time. They'll be quietly fixed if there is a systematic problem then put back in service without fanfare.

Jk

Pan Ron 14 Mar 2019
In reply to Gordon Stainforth:

One of the problems will be that for the average Joe, all 737s are the same, regardless of whether its a classic, an NG, or a Max.  So they'll have problems.  But then most passengers have no idea what they're flying and as long as the airlines themselves make clear that "we do not operate the MAX" then they may be alright.  Either way, it's not a huge hit to the Boeing coffers and maybe even less disastrous than the fatality-free A380 has turned out to be. 

 Tim Davies 14 Mar 2019
In reply to Pefa:

You can’t build a balanced aeroplane for all regimes of flight, especially one with underslung engines. 

Increase thrust= nose up. Need to trim the stabiliser 

then trim again as speed changes, gear goes down, flaps go out, trim trim trim. 

Pprune is about as reliable a source as UKC at times for accurate information 

 tehmarks 14 Mar 2019
In reply to Tim Davies:

Pitch-power coupling is normal - every pilot who has ever flown an aircraft is familiar with it from the moment they apply power to go-around in a Cessna and suddenly find themselves having to push hard forward to check the strong nose-up response and stop themselves from stalling into Farmer Blogg's field below. The problem isn't so much that, but (as far as I can ascertain) that the 737MAX has a tendency to pitch up approaching the stall, due to a combination of the altered thrust moment and the larger engine nacelle generating lift. This isn't acceptable to the FAA (nor should it be) - control forces shouldn't relax approaching the stall, they should increase. Hence the unrefined bodge that is MCAS.

It would seem that Boeing have assumed that if a pilot interrupts MCAS activation by trimming, they will trim to eliminate control forces and thus if MCAS activates again, it will be starting with a 'clean slate'. If the pilot doesn't trim back to fully trimmed, there'll thus be an incremental build-up of nose-down trim.

It's just an incredibly poorly considered bodge, as far as I can tell from my limited (PPL) experience.

Post edited at 20:23
OP Pefa 14 Mar 2019
In reply to Pan Ron:

> One of the problems will be that for the average Joe, all 737s are the same, regardless of whether its a classic, an NG, or a Max.  So they'll have problems.  But then most passengers have no idea what they're flying and as long as the airlines themselves make clear that "we do not operate the MAX" then they may be alright.  Either way, it's not a huge hit to the Boeing coffers and maybe even less disastrous than the fatality-free A380 has turned out to be. 

You are obviously experienced in this field but don't you think travellers will still be asking in 6 months if their flight is going to be on a 737 Max series? And if so then insist on something else.

It sounds to me as if this software patch is just to cover bigger flaws which doesn't fill me with confidence. 

Roadrunner6 14 Mar 2019
In reply to Pefa:

I think it’s hard to call them incompetent.. thousands and thousands are flying with no issue.

Even if they get rid of the 737 max it’s nothing in the big scheme of their business.

Roadrunner6 14 Mar 2019
In reply to Gordon Stainforth:

The A320s have had issues.. people will forget in a few months.

Pan Ron 14 Mar 2019
In reply to Pefa:

Some passengers may pay attention. But my feeling is that to most they simply count the engines, in which case an A320 range is a "Boeing". Once they stop crashing, like a number of other aircraft early in development, I really think few people will pay attention to the details of aircraft type.

That said I just checked the number of Maxes in service and it's more than I'd thought. But still relatively low.

But I really wouldn't be fearful once the software is fixed. Or even proper training taken place. In this case there are questions about the crew's reaction, especially with a co-pulot who, at (reportedly) 200 hrs, is about as green as they can be. While it's hard to say "fly reputable as airlines" when everyone from Singapore to Air France have committed catastrophic failures, better skilled flightdecks are likely found on more reputable airlines.

Post edited at 23:37
1
 Neil Williams 14 Mar 2019
In reply to Tim Davies:

> You can’t build a balanced aeroplane for all regimes of flight, especially one with underslung engines. 

> Increase thrust= nose up. Need to trim the stabiliser 

> then trim again as speed changes, gear goes down, flaps go out, trim trim trim. 

> Pprune is about as reliable a source as UKC at times for accurate information


Yes, but the need to adjust the trim (in a relatively relaxed manner, and possibly automatically) is rather different from the aircraft becoming actively unstable during relatively extreme manual flight, resulting in an IT system that pushes the nose down hard to avoid it.  That to me is a very poor design.

By comparison with a car, you can easily "get used to" a small pull to one direction (e.g. due to uneven tyre wear or road camber) and compensate for it - but this is more like the power steering fighting with you.

Post edited at 23:50
 Neil Williams 14 Mar 2019
In reply to tehmarks:

> It's just an incredibly poorly considered bodge, as far as I can tell from my limited (PPL) experience.

I've never flown an aircraft, and I would certainly agree that a bodge is precisely what it is.

I expect the bodge will be fixed and tested properly and that will be that, but it is really something that should never have happened.

 dannyboy83 Global Crag Moderator 15 Mar 2019
In reply to Pan Ron:

If anybody is interested, the flight data recordings for Lion Air crash are here:

https://leehamnews.com/2018/11/28/indonesian-authorities-release-preliminar...

What is really terrifying is the fight between the manual trim pilot inputs and the automatic trim MCAS system (it should cut out with manual inputs). Ive just started on the 757, which is far too archaic for these modern software systems, so my knowledge isn't too great about the MCAS. We do practice for trim runaway situations in the sim, but the situation in Lion Air would have presented itself quite differently and must have been pretty confusing.

It will be interesting to see the FDR data from the Ethiopian tragedy to see the similarities.

And it's very, very common to have pilots with just 200 hours flight experience (perhaps not in the US). BA, Ryanair, EasyJet, Flybe, Thomson, Thomas Cook, Jet2 are all big training airlines so will use cadet pilots. But the Captain will be very experienced!

Roadrunner6 15 Mar 2019
In reply to dannyboy83:

I don’t know if it’s true, but I heard the 29 year old was the more experienced with over 8,000 hours of flying

a commercial pilot phoned in to a radio show I was listening to saying that.

In reply to Roadrunner6:

Yes, to clarify, the Captain (Yared Getachew) was 29 years old and had over 8,000 hours flying experience. His co-pilot had about 200 hours experience.

In reply to Gordon Stainforth:

> It sounds as if this 737 Max is a crap design, or rather umpteenth upgrade/upscaling of an ancient design (dating incredibly from 1964, first flight 1967), that really should be scrapped. Like the whole embarrassing thing of continually upgrading Issigonis's 1959 Austin Mini, and trying to turn it into a car fit for the second decade of the 21st century.

Except the mini is the other way round isn't it? They are producing a modern car that pretends to be a tweak to the original design

In reply to Gordon Stainforth:

Reports this morning that a screw-like device found in the wreckage in Ethiopia indicates the plane was configured to dive. This evidence persuaded the US to ground the model and it is now expected that the plane will be grounded for a much longer time. 

Edit - Boeings $600 billion order book for the model is now at risk and A320's could now be the beneficiary, hence the continued hike in Airbus share price.

Post edited at 07:58
 Neil Williams 15 Mar 2019
In reply to DubyaJamesDubya:

> Except the mini is the other way round isn't it? They are producing a modern car that pretends to be a tweak to the original design


The 737 Max is more like the Land Rover Defender - a 30 (debatably 50+) year old design with many tweaks over the years for emissions regulations etc, the most obvious one being the "bonnet bump" on newer models required to add a larger engine.

Now imagine there was a problem with it pulling a bit to the left, so they added an IT system to pull it back to the right that sometimes did that a bit hard...

Post edited at 09:32
Removed User 15 Mar 2019
In reply to tehmarks:

This is key isn't it - an appalling design with a lack of sensor redundancy implemented in firmware even though the multiple sensors required to do precisely that exist.

Pan Ron 15 Mar 2019
In reply to dannyboy83:

> And it's very, very common to have pilots with just 200 hours flight experience (perhaps not in the US). BA, Ryanair, EasyJet, Flybe, Thomson, Thomas Cook, Jet2 are all big training airlines so will use cadet pilots. But the Captain will be very experienced!

That's a fair point.  Not really meaning to cast aspersions on the pilots.  But in a tricky situation, a 200 hr MPL could be a liability. 

The public impression that pilots are reliably highly trained professionals, or aware of the right thing to do in upsets, can be misleading.  From Air France to the Turkish prang at Schipol, Asiana at SFO to the A300 losing a vertical stabiliser out of JFK and a load of other prominent cases, the assumption of doing the right thing under pressure can be misplaced.  I've met all sorts in my minimal flying experience and some who I really wouldn't want to have piloting me.  My father (a 747-200 check captain) was contracted by Boeing, along with a number of other highly experienced skippers from around the world, after his retirement to re-train/certify pilots in a certain Asian flag carrier that was having issues.  The stories of dysfunctional flightdecks, aircraft losing their way over the pacific, and sim-rides going catastrophically wrong, were frequent and alarming.  Even in his own airline, cases of CRM gone awry to the point where both pilots refused to communicate with each other or where one stormed out of the cockpit, were not more frequent than would be expected.

Post edited at 11:15
 profitofdoom 15 Mar 2019
In reply to Pan Ron:

> ...........The stories of dysfunctional flightdecks, aircraft losing their way over the pacific, and sim-rides going catastrophically wrong, were frequent and alarming.....

OH GREAT, now I'm really looking forward to my next flight

 Toerag 15 Mar 2019
In reply to tehmarks:

> From what I can discern, MCAS is an automated system that, up until the Lion Air crash, pilots weren't even aware existed. As I understand it, in certain aircraft configurations (autopilot off, flaps up), at high angles of attack (ie approaching the stall) it commands a pitch down (to avoid a stall). Except it doesn't do this by actuating the primary flight control - the elevator - but by winding in a few degrees of nose down trim in over ten seconds. The pilot manually trimming during that period will cancel the trim, but after some seconds it reactivates. Thus, if the pilot isn't careful, he can end up with an aircraft will full nose-down trim - which will want to point itself at the ground without huge back pressure on the controls.

So the question is, how does the aircraft get into a stall situation in the first place? and if the MCAS only does what a pilot would, how is it a problem? Surely the pilot would see the changes on his instruments or feel them through the seat of his pants?*  Looking at the '6 graphs' article on the BBC it would appear that the rate of climb was all over the shop, would that be noticeable to the pilots? What would a 'normal' graph look like?

*I'm not a pilot, but I've done a stunt flight and understand how g-forces can confuse the pilot in poor visibility, but the vis was fine (people on the ground could see the plane).

> Of course, this should only occur approaching the stall - but it would seem that if the AoA sensor feeds garbage data to the system in question, it will, well, produce garbage trim commands. It relies on a single sensor (doesn't cross-check values between the two physical AoA sensors), doesn't take into account any other instrumention, nada. So if a single AoA sensor misbehaves, a system which pilots didn't know existed at first will try to pitch the aircraft into the ground. Imagine that happens on the initial climb out?

Single sensor? Who on earth allowed a commercial airliner to fly with non-redundant systems?! That is utterly bonkers!

 Michael Hood 15 Mar 2019
In reply to Toerag:

The single sensor issue is key.

Faulty sensor, crap in, crap out, crash

Post edited at 13:22
 NottsRich 15 Mar 2019
In reply to dannyboy83:

> If anybody is interested, the flight data recordings for Lion Air crash are here:

> What is really terrifying is the fight between the manual trim pilot inputs and the automatic trim MCAS system (it should cut out with manual inputs).

From this graph? Yes, scary to imagine what the pilots were fighting against with the continual trim adjustments.

https://leehamnews.com/wp-content/uploads/2018/11/Report-fatal-parameters.p...

Pan Ron 15 Mar 2019
In reply to Toerag:

The graphs are weird. If for no other reason than they appear to have been in a consistent climb in the last minute up to the end.  And the descent is odd - not a normal aspect of a standard instrument or visual departure. Maybe because of the control issues they overshot the cleared altitude so intentionally descended.

 Tim Davies 15 Mar 2019
In reply to Pan Ron:

Can the accident experts that inhabit  ukc calculate the rate of descent that is indicated by the flight data recorder and then compare it to what a “normal” descent should be? 

Lots of cherry picking of single bits of information going on here without an appreciation of the bigger picture. 

OP Pefa 15 Mar 2019
In reply to Pan Ron:

> The graphs are weird. If for no other reason than they appear to have been in a consistent climb in the last minute up to the end.  And the descent is odd - not a normal aspect of a standard instrument or visual departure. Maybe because of the control issues they overshot the cleared altitude so intentionally descended.

Would I be correct in saying the graphs show the Captain making manual trims after each automated trim as if to compensate for the system constantly tilting the nose down?After which he repeatedly needs to tilt the nose up again. Confirming what the experts on here having been saying.

Is the MCAS doing that because of a faulty AoA sensor or a problem with the MCAS or have they physically modified this plane too far out of its safety parameters do you think? 

Pan Ron 15 Mar 2019
In reply to Tim Davies:

The graphs are really too rough to take much from, by me at least.  I have no idea of the workings of this MCAS system so you'll probably get better input from others.

That said, the graphs don't make a lot of sense, but aren't crazy either.  Though I don't get, unless impacting rising terrain, how they crashed into the ground while in a 2.5k fpm climb.  I've flown in and out of Addis a few times and I don't recall a lot of notable high terrain and the picture of the crash site looked flat.  A graph of the aircraft's altitude, airspeed and heading would be more interesting right now. A 2,000 fpm descent is nothing in itself, though depending on what else the aircraft is doing could turn out to be extremely relevent; at 36k ft it would be entirely standard, at 5k ft and 2 minutes after take off is a different matter, and if happening while in a nose-up attitude and at or below stall speed does start to point towards obvious issues. 

Trimming into a turn can be a problem and can lead to excessive nose-up pressure when returning to straight and level, though if that graph is to be believed I'm not convinced that this was the cause.  It was in light of this weirdness that I pointed a bit towards the pilot.  The old adage of aviate-navigate-communicate is drummed in to you a lot during training, and many an aircraft has speared-in due to pilots becoming consumed with technical problem solving and failing to maintain stable flight, especially in instrument meteorological conditions (even if you are having pitch control issues, the priority would be to maintain airspeed and steer clear of anywhere where you were going to connect with the minimum safe altitude).  The rule is to follow your instruments and not your senses as these quickly get disorientated in cloud or darkness.  I can sympathise though when your instruments start giving contradictory messages - there are a lot of ports, vanes and tubes around the nose of an airliner that provide dynamic and static pressure readings which combine to indicate altitude, air speed and mach number, angle of attack and so on.  The way these all correlate tells you what the aircraft is doing and should behave in consistent fashion.  If the paint shop leaves masking tape over, condensation freezes within, or bugs accumulate, you may get readings that are entirely contradictory and confusing.

Really is very hard to say what is going on here.  Plenty of room for speculation but the BBC article isn't particularly informative.  Even the Lion Air data linked to above is just too coarse to really understand easily.  I think the cockpit voice recorder is what will be revealing as the pilots will no doubt be articulating exactly what is confusing them.

Post edited at 15:37
 dannyboy83 Global Crag Moderator 15 Mar 2019
In reply to Tim Davies:

Just quickly looking at the altitude Lion Air graph, it appears to have descended 5000ft in around 40secs - 7500fpm. For context, in normal flight at altitude we'd normally descend anything up to about 2500fpm, and if really necessary with speed brakes out we can get up to 5000fpm. Close to the ground, a normal final descent path on a standard 3 degree slope to runway is about 700fpm.

 dannyboy83 Global Crag Moderator 15 Mar 2019
In reply to Pefa:

I believe you are correct. For some reason (false angle of attack inputs?) there are uncommanded automatic pitch down trim movements.

 dannyboy83 Global Crag Moderator 15 Mar 2019
In reply to Toerag:

I think in theory, this MCAS system is absolutely fine, it's just a simple stall protection system.

Yeh the vertical speed is so erratic, that should never happen. If the autopilot is engaged, it would be absolutely constant, and if hand flying the VS would fluctuate ever so slightly as we're not as precise as the autopilot! It would certainly be very apparent to the pilots.

There are 2 independent AoA on the 737 so redundancy shouldn't be an issue.

 dannyboy83 Global Crag Moderator 15 Mar 2019
In reply to Pan Ron:

I suspect those bbc charts are incomplete.


New Topic
This topic has been archived, and won't accept reply postings.
Loading Notifications...