Worst possible news for Boeing
September 6th, 2019 at 1:34:56 AM permalink | |
Fleastiff Member since: Oct 27, 2012 Threads: 62 Posts: 7831 | I thought of that too. No need for actual bribes to have been paid though. the expense and delay to Boeing will be enormous. those outsourced coders sure cost Boeing plenty. |
September 6th, 2019 at 11:01:48 AM permalink | |
Pacomartin Member since: Oct 24, 2012 Threads: 1068 Posts: 12569 |
AFAIK there is no indication of bad coding. The coders wrote code to the design, which was a bad design.c |
September 6th, 2019 at 11:43:55 AM permalink | |
Fleastiff Member since: Oct 27, 2012 Threads: 62 Posts: 7831 |
I thought the 'design' elements were: Use only one of three sensors and do not engage in vote pooling. I thought this was from outsourced low-level engineers but that any decent coding shop would never have been so slavishly naïve. i'm drawing o my memory banks for this and they are in a rather depleted state right now. |
September 7th, 2019 at 6:44:21 PM permalink | |
Pacomartin Member since: Oct 24, 2012 Threads: 1068 Posts: 12569 |
The crash of Turkish Airlines 1951 in 2009 (a B737-800) revealed that the autothrottle was responding to one in three sensors measuring altitude. The failure of the altimeters was a common failure, but it usually wasn't serious. However, at least twice before the sensor had failed and the B737's autothrottle had gone into "retard and flare" mode which basically is the proper response for a jet which thinks it is at 18 feet of alitude. In the two previous failures the pilots realized what was happening and turned off the auto-throttle. In the 2009 crash the three pilots did not know the jet was slowing down until it was too late to prevent a stall. The plane stalled and landed tail first short of the runway. The fuselage split into pieces and the three pilots died along with six people in the main cabin. To program in some kind of vote pooling was not a recommendation of the review board. As only 9 people died in 2009, the idea of a jet crashing on it's own in perfect weather did not create the worldwide mass reaction as the MAX crashes. Also the 2009 crash was sort of a slow motion event taking a full 10 minutes. It was much easier to place the bulk of the blame on the pilots. |
September 7th, 2019 at 7:36:22 PM permalink | |
Fleastiff Member since: Oct 27, 2012 Threads: 62 Posts: 7831 |
It is always easy to blame the pilots, particularly if they died. I think it is agreed that relying on one sensor when three are available is poor engineering even if its not an official focus of the investigators. |
September 7th, 2019 at 8:42:30 PM permalink | |
gamerfreak Member since: Feb 19, 2018 Threads: 4 Posts: 527 |
Paco, Has Airbus had any crashes in the last ~30 years that were directly linked to poor design? |
September 8th, 2019 at 1:58:17 AM permalink | |
Fleastiff Member since: Oct 27, 2012 Threads: 62 Posts: 7831 | Guess you have to define poor design. With Airbus the computer flies the plane until such time as the computer decides something is really wrong and dumps the plane into Alternate Law wherein the pilots are suddenly in control irrespective of their alertness or training. |
September 8th, 2019 at 12:35:32 PM permalink | |
Pacomartin Member since: Oct 24, 2012 Threads: 1068 Posts: 12569 |
I think that is impossible to say. I am using the 2009 crash as an example because the jet automatically went into a safety mode based on input from a single sensor. When in fact the sensor was giving bad data, the "automatic safety mode" crashed the plane. As this sequence of events had happened at least twice before and been openly discussed at Boeing, they clearly didn't think it was a poor design, because the pilots had simply become aware of the what was happening and compensated. We had a fatal accident where I worked, but the analysis was that the people had failed to follow "standard operating procedure". But that is almost always the answer. Nevertheless the unit was redesigned so that in the future if someone made the same mistake nobody would get killed. But no one ever said it was a "poor design". |
September 8th, 2019 at 1:00:40 PM permalink | |
Pacomartin Member since: Oct 24, 2012 Threads: 1068 Posts: 12569 | B737 Next Generation orders (NG or Next Generation was designed to counter the Airbus A320 family) 1317 in North America and 545 in the rest of the world from 1993 until end of 2001 1016 in North America and 2402 in the rest of the world from 2002 until end of 2010 Airbus launched the new engine option (neo) program in December 2010 The events of 9/11 forced Boeing to look harder elsewhere for customers. It was no secret that the A321 had more ground clearance than the B737, but Boeing kept the B767 despite having single digit orders many years, and rather than re-engine the B757, they dropped production. They used the factory space for the B757 to increase B737 production. A stretched B737 was developed (B737-300ER) but it's performance specs were far below the B757 and the A321. The A321 outsold the B737-300ER by a large margin. In 2001 it made perfectly good sense to develop a clean sheet design, but once the NGs were all over the world in large numbers, the idea of re-engining the current design started to take hold. |
September 23rd, 2019 at 6:34:54 PM permalink | |
Pacomartin Member since: Oct 24, 2012 Threads: 1068 Posts: 12569 | Some new very disturbing details have emerged about the MCAS design. The original design was due to some wind tunnel tests in 2012 and the MCAS required both a bad angle of attack reading and high G forces. Command was limited to 0.6 degrees from trimmed position. A separate problem appeared in 2016 that did not involve high G forces. So MCAS was modified and the high G force requirement was removed, command was changed to be limited to 2.5 degrees from trimmed position. The decision to remove the high G force requirement should have (at the very least) meant that both AoA sensors should have been used. It seems that the AoA sensors were destroyed by bird strikes. ------------- Early in the process of selling the MAX, according to two people familiar with the discussions, Boeing promised to give Southwest Airlines a substantial rebate for every plane if the MAX required simulator training. One former MAX worker, said the rebate reported to him by managers was $1 million per plane, a figure another Boeing employee indicated is roughly accurate.c |