Incident: Southwest B38M at Phoenix on Feb 24th 2023, no autopilot trim
avherald.com
avherald.com
I'd of course be curious to read more details and an explanation of the root cause, but it's presumably too soon for that...
[edit] though it looks like they declared an emergency out of an abundance of caution
When you have an abnormal condition that gives you doubt about the aircraft state and priority ATC handling would be helpful, declare an emergency. It’s as simple as that.
https://www.flightradar24.com/blog/squawking-7700-in-flight-...
Closest I ever got was 7600 for a full radio stack loss.
The US Air Force stopped taking deliveries of the KC-46 for a period of time because of sloppy assembly (FOD). One of the ME3 airlines stopped taking delivery of 787s built at the east coast plant because of FOD and other sloppy assembly problems (and Boeing's had to stop delivery a few times on the 787 because of assembly/manufacturing problems). Boeing got into a bit of hot water with the 737NG because they were hand building parts that were supposed to be CNC'd. Stuff like that.
One of the key problems that I vaguely remember is that Boeing did away with a lot of the manual quality control checks in favor of automated checks with no oversight. As Tesla's learned some things are extremely difficult to automate.
In fact, reinforces it. You must check. How do you know the device is calibrated? Is the sensor you're relying on still trust worthy? What happens when your fancy Bluetooth sensor breaks, and you've forgotten how to measure things the old fashioned way?
Now I'm glad Boeing is taking inspiration from Toyota; but American companies tend to lose sight of a key part of the teachings of Demmings: just do it right the first time is wrong. You must build your business around the assumption it will be wrong, (because it always will be), and generate data to be reviewed that can be acted upon to improve processes.
Besides which; remember, the real objective there is cost savings by Quality department downsizing, coupled with remaining Quality Inspector upskilling, which comes with increased cognitive overhead for the inspector, which will manifest in a lower detection rate because you can only be in so many places at once.
T. Decade-on Quality Assurance guy/builder of testing departments.
Companies hate us because we're the one group in the building who can objectively measure how badly everything sucks; yet you can't let up, no matter how badly you're treated because once you stop looking, someone down the line'll pay for it.
And that's just not Quality.
Really needs a new type certification, though FAA and flying public generally don't seem to care.
If by "different handling" we mean "not airworthy", I wouldn't say the design is fine.
https://aviation.stackexchange.com/questions/73132/did-boein...
No MCAS = does not meet the criteria to be employed in Civil Air Transport in the United States.
There were several failures on the system architecture front as well, including a vulnerability to single event upset, where a cosmic ray flipping a single bit in the FC in vommand would trigger an MCAS failure state that the average civil aviation trained pilot was found not to be able to successfully cope with.
It's also in the NTSB and FAA final reports, as well as several House generated investigations.
I spent waaaay too many hours doing that math by hand and looking up crap on Engineering Toolbox to easily forget/forgive that.
The design requirements of the software were wrong. Not the airplane, not the implementation of the software.
I work in a field with increased specialization but we build systems and sometimes similar things come up. There are few who see the end to end system for itself.
I imagine rockets and the space shuttle get that level of scrutiny and that books have been written about it.
That's why the MAX accidents are so notably engineering failures; the industry knew what needed to be engineered, but business shot it down, and forced them to build something fundamentally unsafe in the interests of maximizing profit, and not sweating the product in the box.
With only one you can be in one of 3 states:
S1. The sensor is returning correct readings.
S2. The sensor is returning readings but they are not correct.
S3. The sensor is not returning readings.
To the pilot (or automation) using the sensor there are only two states:
U1 = S1 + S2.
U2 = S3.
S1. Both sensors are returning correct readings.
S2. Both sensors are returning the same incorrect readings.
S3. One sensor is returning correct readings and one sensor is returning incorrect readings.
S4. Both sensors are returning incorrect readings which do not match.
S5. One sensor is returning correct readings and one is not returning readings.
S6. One sensor is returning incorrect readings and one is not returning readings.
S7. Neither sensor is returning readings.
The states the pilot (or automation) sees are
U1 = S1 + S2
U2 = S3 + S4
U3 = S5 + S6
U4 = S7
Here are some numbers.
If the sensor works fine 98% of the time, but 1% of fails in a way that gives false readings, and 1% of the time fails in a way that gives no readings, then when the pilot gets a reading on the single sensor plane there is a 98.99% chance it is right.
On the two sensor plane when the pilot gets a reading that is the same on both sensors there is a 99.99% chance that reading is right. If the pilot only gets a reading from one sensor then it is the same as the single sensor plane case, 98.99%.
Overall, on the two sensor plane if the pilot checks the sensors and accepts the result when the two sensors agree on a reading or when only one sensor gives a reading, they will be getting a correct result 99.97% of the time.
There have been cases where the ONLY thing working on the tail was the trim tab, the yoke cables broke - you can land a plane with only the trim if you really need to.
* WN Delivery Process By Retired on Wednesday, Mar 1st 2023 22:12Z*
On the delivery flights I rode on, after the crew had a meal at the Boeing delivery center and signed the acceptance paperwork. WN dispatch would assign a delivery flight number to one of several "make ready" contract centers. Were various items would be done to bring the aircraft to the current WN configuration. From there it would be released to the line. One of the "make ready" centers is in Phoenix. I suspect in the short time used to turn this aircraft around something was not correctly finished by the "make ready" folks. In the pre-MAX days, the aircraft would be in the contract "make ready" center for a couple of weeks.
Seagate tanked their reputation, but IronWolf Pros seem to work well, too.
[0]: https://www.backblaze.com/blog/backblaze-drive-stats-for-202...
Edit: BTW, HGST is WD, too. :)
None of the design issues have gone away with better software.
None of the management and engineering culture problems have gone away at Boeing. And even if they had done a 180 on both, this particular plane is still compromised.
All they did is the minimum amount of work to be able to fly again.
I gain nothing by flying in those planes where I can find good alternative flights, so why risk it, even if the risk is very low?
I dunno, I would trust it more
And, really, even if the MAX were a stellar plane it still doesn't meet the same crash standards that any newer aircraft (including the A320) has to meet. Your C-series or A320 have seats and floor structures designed to withstand forces of up to 16Gs. The 737? Grandfathered in with 9G seats. It's long past time to retire the 737. It's one thing to keep flying them until the wings fall off, but it's another to sell brand new 737s in 2023.
This isn't criticism or a rant, I'm genuinely curious to know.
I would assume the bigger problem was loss of the electrical trim system. From what I understand, manually trimming a 737 (or any large commercial jet) is a physically demanding task, even with both crew members doing it in tandem. That will add a lot of workload to the pilots that could cause further delay in reacting to new problems as they arise, and there's always the chance that can happen. When a malfunction like this occurs, it is absolutely safest to land as soon as possible so the issue can be assessed on the ground, especially when relevant failure checklists or company policies indicate it.