Read First if flying on a Boeing 737 Max–or maybe not!
November 7, 2018 at 5:09 pm #28046
According to this Gizmodo report the Boeing 737 Max can abruptly dive ‘by mistake’! This may well have caused the Lion Air crash off Indonesia in which everyone died.
I bet if were an Airbus the CAA would immediately ground all such planes until it the fault was proven to be rectified.
April 10, 2019 at 9:01 pm #32528April 16, 2019 at 9:39 am #32707
You can possibly add Boeing 787 planes to the list of what should be avoided. The following was posted on Reddit by an ex-Boeing programmer under his real name. (Later in the thread he challenged Boeing lawyers to come after him as he stated he could prove all the assertions he made)
“To this day, I refuse to fly on a 787. I’m sure that the Dreamliners that came off the assembly line after about a year or so were fine but there’s that first year of production that, as far as I’m concerned, are ticking time bombs. I talked to many engineers who had worked on that program to know just how badly they rushed that initial production.”
That said I’m not really sure how one avoids specific planes if you are going long-haul. Short haul is easier as airline companies tend to standardise, e.g. RyanAir will fly 737Max😱. Long haul is different and you can wind up with pretty much anything. (Many scary tales can be told about flying long haul in the China of 25 years ago) .00April 16, 2019 at 1:44 pm #32713
Yes it was a good post. Did you read the comments from others? I have to say that I was nodding along with all, well with almost all them. I used to say that quality was not a bolt on optional extra, it was the best way to save money and increase margin but few if any were able to hear that message. The company and several others who supplied us with junk, sorry ‘product’ suffered the same fate. At least the pension was OK.
(Unlike my yesterday blood test; the initial rush result came through today with instructions to, ‘sort it out and retest in a week‘.)00April 16, 2019 at 1:57 pm #32715
Although not Boeing our MOD and Thales have similar problems with ‘Watchkeeper’ drones. Of particular note is the way our MOD conceal carp workmanship by covering it up!00April 16, 2019 at 4:43 pm #32727
To be fair, the unstable beast was on a test flight in bad weather to prove, (or otherwise) its capabilities. The fact that it failed suggested to me that the device was not up to the weather conditions. It was yet another case of pitot tubes being of marginal value and carp Thales software not being able to work out when their failed and having a fall back safe operation option. I guess that MOD could be well pleased that a failure mode was found and proved on a test flight rather than on an operational role when lives could have been most. I did find manufacturers got ‘testy’ when faced by a test schedule that was actually designed to test an item under failure conditions. One refused to accept a test plan as part of the tender conditions as ‘they knew it would fail‘ so did not certify it for such conditions. Not very helpful when a number of others accepted the schedule and were happy to cooperate!
I am not sure that army kit that only worked in good weather conditions would be a great asset to anyone really, let alone any army.00April 16, 2019 at 4:57 pm #32729
It rather suggests poor design and/or poor specifications to me – pitot tubes are known to ice up. The remedy is simple, heating elements and at least three pitot tubes so a simple ‘voting’ system can be used. I think the finding against Thales is a scary underscoring of how poor the MOD is in its technical selection process.
“The software algorithms used to identify and disqualify single sensor failure were not always well understood by [Thales] within the UK.”
This is far from being rocket science!
00April 16, 2019 at 9:08 pm #32740
- This reply was modified 3 months ago by Ed P.
Was the Thales device a built to MOD specification device or a production model slightly warmed over to try to edge through MOD testing? That the software was less than stellar and that the design appears to have had ‘issues’ cannot be doubted. Other flying device builders appear to have had pitot tubes issues and caused their craft to ‘fail to fly’. So yes a better system of managing the pitot tubes states along with superior software that does NOT get in a tizzy when, not if it gets doubtful data, does appear to be a requirement. If the MOD failed to make functional tests part of the acceptance then yes they fell down. Certainly the specs I saw when working suggested we would like our purchases to survive extreme conditions in a ‘graceful’ manner. Dropping out of the sky does not fit that with that clause.00April 17, 2019 at 7:17 am #32742
As could be anticipated the US FAA has done a white-wash job. According to them despite previously killing over 500 people and no compulsory additional sensors or fault-voting system the ‘updated software is operationally suitable’. link
Despite this, I certainly will never fly in one of these, and I will avoid airlines such as Ryanair.00April 17, 2019 at 7:51 am #32743
Correction – the two crashes ‘only’ killed 346 people, so I guess the FAA position is understandable!00April 17, 2019 at 11:22 am #32748
Bob WilliamsParticipant@bullstuff2Forumite Points: 3,325
Every aircraft that I ever came across as an AAC technician, whether fixed or rotary winged, was fitted with heater elements in the pitot tube. Even the ‘dinosaur’ that was the Auster. Alan Wood and Drezha will perhaps support the fact that the correct operation of the pitot tube heater is part of pre flight checks, in the RAF, FAA and AAC.
Any new aircraft programme, including drones, is always subject to problems. There are many disasters and failures in the history of flight, from the birth of ballooning in the early 18th century. Much of it is caused by the ‘learning process’ when new conditions and physical limits are encountered. Unfortunately, human error is also a factor, but when aircraft began to be a “product” on sale in quantity for profit, competition and human greed became an issue. This is further complicated when national ‘pride’ caused governments to get involved.
Not every manufacturer and government take the steps and meet the expense of rectifying a problem that causes many deaths. Some will evade the issue entirely, which is the case of Boeing and the US government at this moment. It is to the credit of De Havilland and the British government of the time, that millions were spent to investigate, find and rectify the problem which caused the Comet crashes. However, it took three crashes in less than twelve months, before those steps were taken. At the time, how could aeronautical designers and engineers have known that fitting square passenger windows would cause cracks, leading to explosive decompression and the certain deaths of everyone aboard?
Humans take new paths after a trial and error process, throughout history. But when the profit motive blinds those who decide to ignore the facts, tragedies follow.
“If you think this Universe is bad, you should see some of the others.”
― Philip K. Dick, legendary SF writer.00April 17, 2019 at 11:54 am #32750
Bob, I understand that the rectangular widow shape was part of the issue. The original design called for the windows to be bonded not riveted or bolted in place and that at least in part the change to the specification was a cause of the issue. It was also unfortunate that back in those days investigation were hampered by the debris landing in the sea and the then limited, if any recording capability. The crash count was possibly more understandable given the constraints of the time. The subsequent investigation(s) appear to have been very effective, at the time our air-crash investigators were very highly regarded for much of the work they did, I know and have heard little about them in more recent times so can only hope they remain on the cutting edge and have not been farmed out to some 9 to 5 cost cutting brigade of brigands. I suspect they should have had a field day with any Watchkeeper drone remains had they been found and made availably.00April 19, 2019 at 10:59 am #32804
IEEE has just published a simple guide to what became a massive design and human factors cock-up. link00April 19, 2019 at 2:42 pm #32807
Very thought provoking and sobering. I never expect to risk using a 737 MAX, – or for that matter any other aircraft, but for other machines that is for very different reasons.00April 19, 2019 at 3:04 pm #32809
I did a little research into the latest digital instrument control, but the more I read the less I understand the comments on Thales lack of understanding, or how Boeing could even dream of using the output from just one sensor. In essence modern instrumentation control recognises that all sensors have a lot of spurious noise, and they use a combination of Kalman filters coupled with Bayesian analysis to sort the wheat from the chaff. (Simply put Kalman filters smooth out the readings and Bayesian analysis rejects unlikely readings). Unfortunately it does not matter what fancy corrections you apply if the mechanical pitch sensor gets stuck or jammed up with ice (something that is apparently a known fault with the Boeing sensor), a voting system is then needed in which the system selects the most likely accurate sensor then compares this with knowledge from other instrumentation. e.g. the altimeter!
For those interested there is a primer on Github, and a lot of it is in relatively simple language.00April 19, 2019 at 10:34 pm #32812
Bob WilliamsParticipant@bullstuff2Forumite Points: 3,325
Ed your previous link was very interesting, although there were dissenting voices in response, I believe that the author had the correct (if necessarily long) summation.
For me, the answer to the 737 Max’s problems is simply summed up: using an existing airframe and strapping on larger engines, moving undercarriage forward – changed the aerodynamics which altered flight characteristics in the worst possible ways. The fact also exposed in the article, that the larger, reprofiled engine pods actually created lift, added to the adverse effects. None of this could possibly be compensated for in any way by computers, no matter how sophisticated.
It’s all down to the battle between bean counters and engineers and the engineers always lose. And usually are apportioned the major part of blame when things go wrong.
“If you think this Universe is bad, you should see some of the others.”
― Philip K. Dick, legendary SF writer.00April 20, 2019 at 7:38 am #32814
A correction is probably called for as my posts kept referring to pitot tubes as the problem. In fact it appears that the pitch/angle of attack sensor has been the cause of the 737Max issues and this is rather like a movable vane attached to the aircraft. However as a mechanical element which is exposed to the elements it needs regular service and calibration as do the electrical elements that pick up the actual vane orientation.
Of interest the FAA has just issued a grounding order to a make of small executive jets (Cirrus Vision) for similar Boeing type reasons. link00April 20, 2019 at 8:07 am #32815
Bob, there were several themes that have emerged, the on the cheap use of a single and demonstratively unreliable source of guidance information to the compulsory crash causing MCAS system. That the new nacelle design causes a further unstable source of lift in addition to the thrust centre changes, etc. is held by those on the ‘unhappy side’ to believe that the crate is almost uncontrollable by a human pilot and given the limitations of the presently available generation of computer assistance system uncontrollable by software either. Ed has enumerated various ways that software supported ‘machines’ need to have error sources met and constrained to avoid fatal outcomes, signal noise being a major issue, single point of failure being another. I have not aeronautical operational experience, but every time that single points of failure were mitigated in my spheres of activity, the publicly apparent service failures vanished. Nothing was noticed and no one died, contrast that with the 737 MAX.
The most worrying aspect for me is the disclosure that the lift characteristics resulting from the new design change is highly unstable, being subject to various variables, such as speed and angle of attack. Spoilers can sometimes mask the effects of lift generation, but having the lift component continuously changing sounds like a whole new set of challenges summed up as instability. To have avoided a need for certification when making such gross destabilising changes to the old plane design still sounds horribly unwise to me.
Thank you I would rather not travel at all rather than face unstable design failure and if I must travel I’d rather walk or drive.00April 20, 2019 at 8:11 am #32816
Of interest the FAA has just issued a grounding order to a make of small executive jets (Cirrus Vision) for similar Boeing type reasons. link
That was probably tripped by the realisation that Boeing caused an issue, the link insists on the acceptance of all cookies – including its advertising related ones, so I did not go through with reading the post. I accept your comment about it though.
I suspect a bit of cleaning out of some messy stables by the FAA!00April 20, 2019 at 12:02 pm #32821
I think it is probably reasonable to include a fair quote from the ‘Flying’ web site:
“The agency said, “The noted condition presents an immediate danger to pilots and passengers of Cirrus Design Corporation Model SF50 airplanes because an uncommanded pitch down may be difficult to recover from in some flight regimes with potentially fatal consequences. The before-further-flight compliance time and need to replace the AOA (Angle of Attack) sensors due to the potentially fatal consequences does not allow for prior notice and opportunity to comment for the public.”
It is a pity that the FAA cannot afford to employ the required expertise, otherwise the Boeing passengers may have been protected. I wonder if our CAA have similar problems to the FAA with recruiting and retaining aeronautical/instrument/computer technicians. The Cirrus thing only came to light after three identical near misses which fortunately the pilots were able to survive, despite the ‘bitey dog’ anti-pilot interference systems.
If the latter quote is obscure it refers to an explanation in the IEEE report that ‘modern’ control systems take positive action to ensure that the pilot has a hell of a job to take action contrary to that mandated by the computer control. (Really scary!). The writer referred to this as the ‘bitey dog’ that snaps at the pilot.00April 20, 2019 at 4:12 pm #32823
I share your lack of enthusiasm for the way this has been handled. While the bitey dog might be useful, I am not totally sure I feel comfortable with him as the pilot and the real thing relegated to a watch and fail role as they have now become. Certification should be governed by one set of rules for all, not one for (a) ‘nothing really required’ and one for the rest ‘do everything again and then get recertified’.
Note, (a) appears to be Boeing as the blue eyed one at the moment.00
You must be logged in to reply to this topic.