Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Boeing 737 Max8 Aircraft Crashes and Investigations [Part 6] 17

Status
Not open for further replies.

Sparweb

Aerospace
May 21, 2003
5,131
This post is the continuation from this series of previous threads:

thread815-445840
thread815-450258
thread815-452000
thread815-454283
thread815-457125

This topic is broken into multiple threads due to the length to be scrolled, and images to load, creating long load times for some users and devices. If you are NEW to this discussion, please read the above threads prior to posting, to avoid rehashing old discussions.

Thank you everyone for your interest! I have learned a lot from the discussion, too.

Some key references:
Ethiopian CAA preliminary report

Indonesian National Transportation Safety Committee preliminary report

A Boeing 737 Technical Site

Washington Post: When Will Boeing 737 Max Fly Again and More Questions

BBC: Boeing to temporarily halt 737 Max production in January
 
Replies continue below

Recommended for you

Understood, but coincidentally there were reports that it may be under powered in the Max. Presumably they've found a workaround or the reports were exaggerated since Boeing hasn't had to commit to switching processors yet, as far as I'm aware.
 
A colleague of mine attempted to control a lumber sorter in a sawmill with an 8080.
Making a couple of outputs a second sounds easy, but there was a critical size.
As the size of the sorter increased there were more sorting bins and more choices, including keeping track of the number of pieces in each bin and diverting to an alternate bin while a bin was being emptied.
Too much information.
The boards arrived at a steady rate.
The 8080 would start missing boards in a large sorter.
Lumber sorters reverted to the old discrete logic gates until faster chips with more than 8 bits were developed.
As tasks are added, there is a real possibility that a task may not be completed in a timely manner.
A missed task in a lumber sorter meant that a board would ride the chain to the end and then drop on the ground.
Not a good thing for an airplane.
Looking at the input from one sensor and generating an output is one thing.
Looking at and comparing the outputs from three sensors, and also checking for rapid changing signals adds much computing time.
This all added to a computer that may be quite busy already.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
A little bit of digging found these gems


Here's the blog is listed


Some of the comments below the article add something to the issue.

A new one



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Great articles LI... I was surprised a financial outfit had noted the processor issue.

Dik
 
From LittleInche's linked article:
THE WALL STREET JOURNAL said:
Boeing Loses MAX Deal to Airbus
Saudi airline says it will buy up to 50 Airbus jets, worth more than $5.5 billion
AND
MOON OF ALABAMA said:
Some decades ago your host programmed special input device drivers for Intel 80286 and alike systems. Their purpose was to record and process data from industrial process sensors, often hundreds at a time. Performance and timing issues required that the 80286 input drivers had be written in low level assembler language. But even with extremely optimized code the system would eventually come to its limits. The delayed procession of data from one sensor would eventually cascade into further delays and in the end the system would fail to record and process anything. The task was simply above its physical limits.



Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Are they saying it could have been too much for the processors to do a more complex and thorough assessment of multiple sensors?
 
Modern processors might prove to be unsuitable for many aircraft. Cosmic radiation goes up with altitude. Modern processor have much smaller transistor size (Moore's Law), and are more susceptible to upset. It's a complex subject, and in the conservative engineering of aerospace there is a tendency to reuse what has proven reliable before. When applications leave earth, and full radiation-hardened processors are needed, the computing power tends to lag decades behind current consumer products. The Curiosity Rover on Mars uses a Rad hardened variation of the PowerPC processor that Apple introduced in products in 1991.
 
Tomfh said:
Are they saying it could have been too much for the processors to do a more complex and thorough assessment of multiple sensors?
Yes. From the days of industrial use of the 8080 processor and the TRASH 80. Tandy RAdio SHack 80.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Seems the FAA are getting slammed for the Air Southwest oversight.

88 planes were brought into the N reg and they managed to sign off 70 planes in one day as compliant.

It normally takes 3-5 people going through the aircraft paperwork 2 days per aircraft up until 10 years old, and an extra day for every 5 years over 10.
 
AH:

Another government agency that has been marginalised.

Dik
 
It's unfortunate that they couldn't migrate the 286 to the 386DX, with internal floating point, which would have simplified the processing.

Feature sizes on the 286 are probably around 200 times bigger than the feature sizes on a current generation Snapdragon processor. However, the throughput is on the order of 4000 times higher, which could allow for triple modular redundancy in the processor software to easily deal with single event upsets, along with running full ECC on the data.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529 Entire Forum list
 
I seem to remember from a few threads back that its the internal floating point which makes it so extremely hard to certify
 
Not sure there's that much difference between the 287 and the 386 internal FPU.

The older FPUs are likely to be substantially less complicated than, say, the Pentium FPU, which had microcode errors.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529 Entire Forum list
 
I think it comes down to the culture at Boeing, from the managers through the board of directors to the owners.
If we can't grandfather it, we won't do it.
So we get a 35 year old processor in a 50 year old airframe.



Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Since the company's board is driving this, they have to start making them personally liable.

Dik
 
I'm sure they'll be fined some absurdly small amount and maybe have to do community service.
 
I agree with both dik and RVAmeche.
Their lost and cancelled sales and payouts to airlines will be the rel punishment.
They have probably paid out over a billion already in settlements to airlines.
From the reports it is not clear if these settlements are for 2019 with more to come or if they are final settlements.
If these settlements are for 2019 alone, the final total may be double or more.
More information on settlements and lost sales at this link.
Did Boeing Just Pay American Airlines More Than $500 Million To Settle 2019 MAX Claims?

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor