User avatar
bikerthai
Posts: 2965
Joined: Wed Apr 28, 2010 1:45 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 5:51 pm

Stitch wrote:
And again, why are we speculating that Boeing will have to "restrict" the flight testing envelope? The failure occurred well beyond what the flight test articles will ever experience. There is no reason to restrict them in any way.


Concur. Passing the ultimate load test is only needed to qualify the plane for passenger service. It does not limit Boeing from flight testing within the vast majority of the test parameters.

However within the current environment of hypersensitivity who knows what'll be in the mind of upper management?

bt
Intelligent seeks knowledge. Enlightened seeks wisdom.
 
StTim
Posts: 3446
Joined: Thu Aug 08, 2013 7:39 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 6:20 pm

Having 6 test aircraft may make some elements of testing quicker but most of the frames will have little or no test equipment loaded. The heavilly instrumented frames carry most of the initial load.
 
mjoelnir
Posts: 8515
Joined: Sun Feb 03, 2013 11:06 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 7:22 pm

All the last certification campaigns used more than one aircraft. The last big campaign A350-900 used 5 frames. A350-1000 used 3 frames. A330-900 used 2 frames. Non finished inside of 12 month.
The 787-8 using 6 frames for the test campaign, used 22 month from first flight to EIS.
The A350-900 used 5 frames and used 19 month from first flight to EIS
The A350-1000, 3 frames and 15 month
The A330-900, 2 frames and 14 month.

The 777-9 is somewhere between normal derivative and clean sheet, so I do not see how even 6 frames should lead to a less than 12 month campaign.
 
VV
Posts: 947
Joined: Sat Feb 13, 2016 1:03 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 7:31 pm

Opus99 wrote:
The fact that there are 6 test aircrafts does that make the certification process any shorter?


In some way it would shorten the process, yes.

If everything goes perfectly then a certification at the end of 2020 is possible, but usually things do not happen perfectly according to the plan.

The timing is very tight and I do not see much buffer in the schedule.
A 2020 certification is possible but very challenging.
 
morrisond
Posts: 1394
Joined: Thu Jan 07, 2010 12:22 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 7:49 pm

However - hasn't the certification process already begun? They have been able to do a lot of the ground work - some of which I would assume normally takes place after first flight.
 
mjoelnir
Posts: 8515
Joined: Sun Feb 03, 2013 11:06 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 9:02 pm

morrisond wrote:
However - hasn't the certification process already begun? They have been able to do a lot of the ground work - some of which I would assume normally takes place after first flight.


Why do you assume that? The A330-900 was also waiting for engines, do you thing they did no groundwork?
 
jagraham
Posts: 924
Joined: Sun Jul 17, 2016 11:10 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 10:01 pm

bikerthai wrote:
estorilm wrote:
Like I said, maybe a stupid question..


Not really a stupid question, because the answers are somewhat complex.

estorilm wrote:
Likewise, the wings can't be too "happy" about performing this test again.


If this was a metal wing, then truly, the components on the wing would have gone past yield (metal deformed). But the thing is, that if you re-load the metal, the stress strain curve will go back up the same linear slope until it reaches yield and ultimate failure again, just at a different deflection point. As the regulation does not specify where in the deflection point you have to meet, just the actual load, then in theory, you can retest a metal wing. '

For a composite wing, not sure if there is any yield, it will break when it breaks. All you can do is check for delamination. If there is no delamination, then you can probably test it again. Note that even if the wing is composite, there are many metal parts. A thorough check with the stress notes must be done to see if the condition as noted above for the metal wing have been reached before you give a second try.

Now, if the decompression is caused by deformation of the fuselage, then it is a more complex problem to solve. They may decide to stiffen up the door frame . That may be done with a simple doubler patch for the test frames but may require a re-design. The re-design may not go into effect until later frames, so the first few frames will have to fly with flight restrictions, either weight limits, internal cabin pressure limits or both.

bt


Or doubler plates at the point of failure. If the failure is the fuselage or door frame. It could be the lock or latch, which would probably require a beefed up part. But with such fixes applied, there should be no need for test limitations.
 
VV
Posts: 947
Joined: Sat Feb 13, 2016 1:03 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 10:26 pm

morrisond wrote:
However - hasn't the certification process already begun? They have been able to do a lot of the ground work - some of which I would assume normally takes place after first flight.


The certification process started already when the aircraft was launched.

To agree on the certification basis is part of the certification process. I guess.
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 11:32 pm

StTim wrote:
Having 6 test aircraft may make some elements of testing quicker but most of the frames will have little or no test equipment loaded. The heavilly instrumented frames carry most of the initial load.


Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.

This is why you need private sector experts from various domains to come in and analyze a problem together. You get enormous insights and efficiencies this way. It has been proven over and over and over again with the movements toward Lean Manufacturing, Scrum Delivery, Prince/II project management, etc.. If a process has been iterated on 50+ times, there's an incredibly good chance that a clean-sheet re-architecting of the process will yield efficiencies undreamt of by the operators and bystanders. AMD just pulled this off with their Ryzen processors after basically a decade of failed iterations of the same product.

Boeing needs a clean-sheet 737 replacement; and if a 6-craft fleet takes a year of daily flights to complete a smooth round of clean-sheet certification, so does the FAA.
 
mattcawby
Posts: 109
Joined: Sun Mar 05, 2006 12:20 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 02, 2019 11:49 pm

Opus99 wrote:
The engine re-installation has begun on WH002. The engines c-shells are on ground and ready for installation.


The thrust reversers have been on the ground since the engines were removed. There are no engines yet on WH001 or WH002
paineairport.com
 
iamlucky13
Posts: 1072
Joined: Wed Aug 08, 2007 12:35 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 1:13 am

patrickjp93 wrote:
StTim wrote:
Having 6 test aircraft may make some elements of testing quicker but most of the frames will have little or no test equipment loaded. The heavilly instrumented frames carry most of the initial load.


Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.

This is why you need private sector experts from various domains to come in and analyze a problem together. You get enormous insights and efficiencies this way. It has been proven over and over and over again with the movements toward Lean Manufacturing, Scrum Delivery, Prince/II project management, etc.. If a process has been iterated on 50+ times, there's an incredibly good chance that a clean-sheet re-architecting of the process will yield efficiencies undreamt of by the operators and bystanders. AMD just pulled this off with their Ryzen processors after basically a decade of failed iterations of the same product.

Boeing needs a clean-sheet 737 replacement; and if a 6-craft fleet takes a year of daily flights to complete a smooth round of clean-sheet certification, so does the FAA.


I often wonder if it's worth responding to such posts, because the aspects being misunderstood are not obvious, well beyond the personal or professional experience of most people, and interwoven with genuine shortcomings that can be criticized. But, might as well give it a shot:

The process is definitely bloated and bureaucratic. A big part of the challenge is it got that way is because more people died when it wasn't, and nobody has come up with a way to significantly streamline it without potentially reducing the safety. It's more or less a standing accusation that the 737 MAX issues are an example of that.

It's not for lack of trying, either. It costs Boeing and Airbus both a huge amount of money to certify an aircraft. They likewise spend a lot of money on hiring and training to analyze and adapt their processes to be more efficient. A lot of people get brought in from different industries as part of that effort. I live in the Seattle area and have been regularly getting to know people doing "lean this" and "sprint development that" for years, and have been exposed in various ways to the supply chain.

Those sorts of methodologies work a lot differently when managing teams of a few dozen or hundreds of people working on non-safety critical software, or building cars with a few tens of thousands of parts with little to no traceability beyond the batch level, than it is with teams of 10,000+ people creating a safety-critical product with a mix of mechanical, electrical, software systems; design margins significantly smaller than in most industries (eg - 1.5x typical in aerospace vs. 3x typical in construction); and several million parts, the least significant of which are typically traceable at the batch level, thousands of which are individually serialized, with every discrepancy documented by serial number and tracked for the life of the aircraft.

All that said, when I was working in test engineering for vehicles, far simpler, far less regulated products, our production prototype test phase was 4-6 months. That's after having gone through about 18 months of prior testing including mule prototyping of all major new hardware, then an alpha build of the intended design with minimal discrepancies, not on the production line, before making the production prototype on an approximately representative production line.

The processes aren't directly comparable because it's easier to build earlier stage test ground vehicles than airliners, but certifying an aircraft in 12-18 months actually sounds pretty good to me.
 
OldAeroGuy
Posts: 3886
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 1:27 am

patrickjp93 wrote:
Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.


I'm afraid you have little idea of Federal Aviation Regulations (FAR's) and the testing required to satisfy them.

Let's correct one thing. The FAA/EASA do very little actual flying. Most testing is done by OEM pilots who are delegated by the Cert Agencies. In fact Cert Agency personnel will not set foot an airplane until enough testing has been done to show that it is safe and reliable.

And very little tax money is expended, test costs are borne by the OEM's.

Read through AC25-7D. It outlines the certification flight testing required to satisfy FAR Part 25. Then try to devise a test program that can demonstrate all conditions in a couple of months. Don't forget the remote testing to find icing conditions or limit cross winds. Iceland is a good bet for both but only if it's the correct time of year.

Then you'll also need to do the Community Noise testing (Part 36), ETOPS demos, Frequency and Reliability at airports around the world... The List goes on and on.

In addition, a large portion of the testing does not involve cert work. Fuel mileage is not required for Part 25 Cert. It is very important for success in the market place though. A significant portion of flight test is dedicated to proving that the airframe/engine are meeting their cruise performance drag and SFC design criteria.

And of course sales demos and trips to the Paris/London/Dubai air shows that have to be built into the schedule.

Plus there are always a few surprises, like the 777-9 cargo door, that crop up and may require re-test.

The OEM's have been developing, certifying and delivering commercial airplanes for 60 years. Testing costs time and money, they run very lean test operations to minimize the effort. The typical 12 - 18 month test programs don't have slack time and no repeat testing that isn't required to correct a problem.

As a final note, the 787 Program planned to have a 6 month test program. It didn't turn out well.

Please do a little research before making grand pronouncements. Knowledge gained in one field may not be transferable to another.
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 1:30 am

iamlucky13 wrote:
patrickjp93 wrote:
StTim wrote:
Having 6 test aircraft may make some elements of testing quicker but most of the frames will have little or no test equipment loaded. The heavilly instrumented frames carry most of the initial load.


Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.

This is why you need private sector experts from various domains to come in and analyze a problem together. You get enormous insights and efficiencies this way. It has been proven over and over and over again with the movements toward Lean Manufacturing, Scrum Delivery, Prince/II project management, etc.. If a process has been iterated on 50+ times, there's an incredibly good chance that a clean-sheet re-architecting of the process will yield efficiencies undreamt of by the operators and bystanders. AMD just pulled this off with their Ryzen processors after basically a decade of failed iterations of the same product.

Boeing needs a clean-sheet 737 replacement; and if a 6-craft fleet takes a year of daily flights to complete a smooth round of clean-sheet certification, so does the FAA.


I often wonder if it's worth responding to such posts, because the aspects being misunderstood are not obvious, well beyond the personal or professional experience of most people, and interwoven with genuine shortcomings that can be criticized. But, might as well give it a shot:

The process is definitely bloated and bureaucratic. A big part of the challenge is it got that way is because more people died when it wasn't, and nobody has come up with a way to significantly streamline it without potentially reducing the safety. It's more or less a standing accusation that the 737 MAX issues are an example of that.


Look, I am NOT attempting to exonerate Boeing of a proven over-par (sorry, taking a stand against sub-par as a derogatory term, as someone who outright sucks at golf) technical solution when they'd already taken the proven tech through 10 years of U.S. military service, but the idea that no one has streamlined anything without reducing safety is a Straw Man at best and Ad Hominem at worst. If you think I'm grandstanding on my ability, that's one thing, however, everything I'm only just trying to unpack about FAA process SO FAR I have either strictly bound within proofs of contradiction/contraposition or I have explicitly stated I'd like to see the data before rendering a final conclusion. There has never been a particularly effective government except the ones as minimal as; to take a finger in the wind, the ones of Lichtenstein, Switzerland, and the U.S.A., if we take a measurement of wealth growth for the citizens per dollar spent on government. That being said, to take an almost psychopathic position in addressing the problem, you can't arrive at the best solution unless you address ALL facets of the problem and proposed solution architectures.

Yes, we get much more fuzzy-feely when regulators take a long, long time to approve things, but you have to have people like me keeping that emotional, irrational instinct in check, because in the end it doesn't improve anything to effectively repeat tests of the N-Xth degree when there's no justification.

And I'm one of those people who painstakingly proves, in absolute mathematical terms, that such tests are useless when the others are executed. Unless you yourself intend to get down in the mud with the people you disagree with, you have no right to disagree with them. That is the foremost essence of debate and integrity.
 
OldAeroGuy
Posts: 3886
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 3:49 am

patrickjp93 wrote:
Unless you yourself intend to get down in the mud with the people you disagree with, you have no right to disagree with them. That is the foremost essence of debate and integrity.


Have you read and understood AC25-7D? If you haven't, you're not down in the mud yet.
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
VV
Posts: 947
Joined: Sat Feb 13, 2016 1:03 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 10:20 am

mattcawby wrote:
Opus99 wrote:
The engine re-installation has begun on WH002. The engines c-shells are on ground and ready for installation.


The thrust reversers have been on the ground since the engines were removed. There are no engines yet on WH001 or WH002


Do you know if the tweets about GE9X redelivery to Boeing were true.
 
User avatar
atcsundevil
Moderator
Topic Author
Posts: 3586
Joined: Sat Mar 20, 2010 12:22 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 2:37 pm

Please stick to discussing the topic. This is the 777X Testing Thread. Other discussion belongs in the appropriate threads.

✈️ atcsundevil
 
JameshSEA
Posts: 4
Joined: Thu Oct 03, 2019 12:48 am

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 4:01 pm

Boeing has already built a couple 77X air frames, including one for stress tests and one air frame that is set to be delivered to Emirates.
 
TropicalSky
Posts: 265
Joined: Fri May 05, 2017 1:37 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 4:56 pm

Any update on if the GE9X is back in KPAE as yet?
 
Scotron12
Posts: 141
Joined: Fri Mar 01, 2019 2:13 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 5:00 pm

VV wrote:
mattcawby wrote:
Opus99 wrote:
The engine re-installation has begun on WH002. The engines c-shells are on ground and ready for installation.


The thrust reversers have been on the ground since the engines were removed. There are no engines yet on WH001 or WH002


Do you know if the tweets about GE9X redelivery to Boeing were true.


Difficult to say. Guess they arrived in the dead of night and are currently hiding somewhere.

One would think someone would have witnessed the engines arrival...why all the secrecy if true??
 
User avatar
kanban
Posts: 3956
Joined: Sun Jan 06, 2008 1:00 am

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 5:12 pm

patrickjp93 wrote:
StTim wrote:
You have test cases, versions of test cases, test conditions happening individually, test conditions happening concurrently. Time taken to set up test cases.

Repeat test cases after fixes.

Finally certification flying.

12 months minimum.

And I'm saying it's likely bogus that it takes 12 months MINIMUM by necessity rather than simply by poor management, especially if you have multiple test craft and enough test pilots on staff. Assuming the FAA isn't stalling between test flights to crunch data (because this CAN and SHOULD be done concurrently with the flying and is largely automatable), then if it takes 12 months with 1 craft, then 6 with 2, 4 with 3. And my job as a software tester was not just to create a comprehensive suite of tests, but make sure it was a minimal comprehensive suite, IE absolutely no redundant tests, because the improvements in processing seed needed to be in production ASAP or millions of dollars could be lost. Half the time I was writing proposals for removing test cases or otherwise consolidating them because they were orthogonal and combinable in entirely predictable ways. If you think there are zero such cases with the FAA's test regimen, I'd like to sell you the Verrazano Bridge.

There's really no point in testing overspeed tolerance at 2, 3, and 4% parameters individually. Just go right up to the edge before the predicted redline where turbulent and supersonic flow cause control loss, validate that you're still handling as normal, then go past your red line, validate that the control is lost in the EXPECTED manner, and move on to the next set of tests. Incremental tests like that are completely redundant and results highly predictable (can be interpolated with 8+ 9s of accuracy if you have the base and far tail cases in hard, reliable data).

We all know how inefficient government agencies are, and I'm not saying play fast and loose. I'm saying I'd be very interested to see how much of the testing is purely redundant, because that's your tax dollars at waste doing nothing useful. The stress test frame is a dedicated one performing all your longterm studies of wear and tear.


one of the purposes of stepped testing is to determine if there is a failure where it occurs. test to the max first doesn't tell you anything more that it failed and then you have to figure out the sequential steps to identify the exact point the failure began. something that's difficult if you've destroyed the test subject..

Also the testing is Boeings cost.. there are no tax dollars involved.
 
User avatar
Revelation
Posts: 21414
Joined: Wed Feb 09, 2005 9:37 pm

Re: Boeing 777X Testing Thread - 2019

Thu Oct 03, 2019 5:19 pm

Scotron12 wrote:
VV wrote:
mattcawby wrote:

The thrust reversers have been on the ground since the engines were removed. There are no engines yet on WH001 or WH002


Do you know if the tweets about GE9X redelivery to Boeing were true.


Difficult to say. Guess they arrived in the dead of night and are currently hiding somewhere.

One would think someone would have witnessed the engines arrival...why all the secrecy if true??


Absence of evidence ≠ evidence of absence...
Wake up to find out that you are the eyes of the world
The heart has its beaches, its homeland and thoughts of its own
Wake now, discover that you are the song that the morning brings
The heart has its seasons, its evenings and songs of its own
 
User avatar
zkojq
Posts: 3884
Joined: Fri Sep 02, 2011 12:42 am

Re: Boeing 777X Testing Thread - 2019

Mon Oct 07, 2019 3:34 pm

Maybe it's early to ask, but is there a publically available 777x production list anywhere. Some a.net folks did a fantastic job of these for the 787 and A350.

musman9853 wrote:
it is a bit behind schedule but people were claiming that this engine issue would take forever to get fixed and there would be absolutely no way it would fly anywhere near on time.

Although not a disaster compared to other recent certification efforts, nine months late is hardly timely.

patrickjp93 wrote:
Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.


Yes, after the MAX FUBAR I'm sure that the FAA will be happy to expedite the process and take all the shortcuts necessary during certification to make sure that Boeing meets the deadlines they promised their customers. :roll:
First to fly the 787-9
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Mon Oct 07, 2019 3:58 pm

zkojq wrote:
Maybe it's early to ask, but is there a publically available 777x production list anywhere. Some a.net folks did a fantastic job of these for the 787 and A350.

musman9853 wrote:
it is a bit behind schedule but people were claiming that this engine issue would take forever to get fixed and there would be absolutely no way it would fly anywhere near on time.

Although not a disaster compared to other recent certification efforts, nine months late is hardly timely.

patrickjp93 wrote:
Then fire the cruft staff of the FAA and buy more test equipment and pilots or analysts. That's a terrible excuse. The FAA has been certifying craft for 60 years now. Much of this is highly understood and predictable. If you can have 6 craft in the air 7 days a week for a year and not have certification complete under smooth circumstances, the certification process is bloated in ways which would probably infuriate even the most cautious tax payer.


Yes, after the MAX FUBAR I'm sure that the FAA will be happy to expedite the process and take all the shortcuts necessary during certification to make sure that Boeing meets the deadlines they promised their customers. :roll:


I didn't advocate taking shortcuts or cutting corners. If it CAN be parallelized (for a reasonable price given the gains), then parallelize it! Manufacturing processes and algorithms are not so different.
 
trijetsonly
Posts: 689
Joined: Tue Jul 21, 2009 10:38 pm

Re: Boeing 777X Testing Thread - 2019

Mon Oct 07, 2019 4:15 pm

zkojq wrote:
Maybe it's early to ask, but is there a publically available 777x production list anywhere. Some a.net folks did a fantastic job of these for the 787 and A350.


There has been one here:
https://docs.google.com/spreadsheets/d/1xPZP2NmigprVBd5dklYcFOGraAn4C8Z4W0z2TPD-eD0/edit#gid=5

But it seems like it hasn't been updated in a while.
Happy Landings
 
OldAeroGuy
Posts: 3886
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 4:03 pm

patrickjp93 wrote:
I didn't advocate taking shortcuts or cutting corners. If it CAN be parallelized (for a reasonable price given the gains), then parallelize it! Manufacturing processes and algorithms are not so different.


What leads you to believe this isn't already being done?

Have you read AC25-7D yet?
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 5:21 pm

OldAeroGuy wrote:
patrickjp93 wrote:
I didn't advocate taking shortcuts or cutting corners. If it CAN be parallelized (for a reasonable price given the gains), then parallelize it! Manufacturing processes and algorithms are not so different.


What leads you to believe this isn't already being done?

Have you read AC25-7D yet?

It's clearly not being done effectively. At most 2 test craft flying simultaneously.

Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.
 
User avatar
ssteve
Posts: 1367
Joined: Fri Dec 02, 2011 8:32 am

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 5:29 pm

patrickjp93 wrote:
Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


2 devices can give you the throughput of 5 if the schedule allows higher utilization. Also, there's only one engine manufacturer here. They might just rather have your additional three planes more or less immediately ready to deliver.
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 6:10 pm

ssteve wrote:
patrickjp93 wrote:
Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


2 devices can give you the throughput of 5 if the schedule allows higher utilization. Also, there's only one engine manufacturer here. They might just rather have your additional three planes more or less immediately ready to deliver.


If the utilization for each day isn't at least 8 air hours (takes time to offload testing dar), then there is PLENTY of room for fixing that too.
 
OldAeroGuy
Posts: 3886
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:15 pm

patrickjp93 wrote:
OldAeroGuy wrote:
patrickjp93 wrote:
I didn't advocate taking shortcuts or cutting corners. If it CAN be parallelized (for a reasonable price given the gains), then parallelize it! Manufacturing processes and algorithms are not so different.


What leads you to believe this isn't already being done?

Have you read AC25-7D yet?

It's clearly not being done effectively. At most 2 test craft flying simultaneously.

Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


You are a quick study.

Which test points do you think are redundant?
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
fcogafa
Posts: 1180
Joined: Fri May 16, 2008 4:37 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:22 pm

Revelation wrote:
Scotron12 wrote:
VV wrote:

Do you know if the tweets about GE9X redelivery to Boeing were true.


Difficult to say. Guess they arrived in the dead of night and are currently hiding somewhere.

One would think someone would have witnessed the engines arrival...why all the secrecy if true??


Absence of evidence ≠ evidence of absence...


I stood and watched 2 large 'packages' being unloaded from the An124 at PAE last Saturday, around 1800 local in daylight. They both had covers with 'GE9X' written in large letters on them. Of course this may have been a carefully crafted fake by Boeing....
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:33 pm

OldAeroGuy wrote:
patrickjp93 wrote:
OldAeroGuy wrote:

What leads you to believe this isn't already being done?

Have you read AC25-7D yet?

It's clearly not being done effectively. At most 2 test craft flying simultaneously.

Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


You are a quick study.

Which test points do you think are redundant?


Not quick at all. When I was a teenager I read Harry Potter & the Order of the Phoenix in just 14 hours. I've had over a week to read a document half as long by word count.

Let's take static testing as a starting point. As you go (incrementally) toward the 150% target, de-lamination, cracking, or structural failures will happen at any given percentage, so running so many incremental tests up to the 150% target is pointless. The static test frame can easily be rigged up with enough sensors to detect any component failure along the way. Start with pressure and climb from 0 to 150 until it fails or holds. Then on the wing and other structural flexes, climb 1% at a time up to 150 in 1 hour until a failure is detected or the 150 is held for the requisite time. Going by the 777X schedule, that's 3 days saved already. That's easy though and subject to simple inductive proofs. It's the flight testing where so many variables are measured in each test that take time to prove are redundant.

Brake testing can also be truncated. They already test on runways far, far longer than any plane is going to overrun. You can modify the thrust reverse test portion to incrementally climb from 0 to 100% reverse (failure happens when it happens and is detectable at exact point of failure) in one test and be perfectly fine. The runways for testing are long enough there's no risk even in a catastrophic failure. The same then goes for the maximum brake wear test. You don't need the leadup steps once the thrust reverser results are validated. If the manual brakes fail at that point, they need a fundamental redesign regardless of all previous tests, and the reversers, spoilers, and flaps will keep the plane from overrunning the runway (we can even add contingency bungees in the event of multi-system failure in the last 1/3 of the test runway to catch the weight if needed). So in total I have somewhere between 6 and 11 days of savable time using the 777X's testing schedule so far.
Last edited by patrickjp93 on Tue Oct 08, 2019 7:47 pm, edited 1 time in total.
 
iamlucky13
Posts: 1072
Joined: Wed Aug 08, 2007 12:35 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:39 pm

patrickjp93 wrote:
iamlucky13 wrote:
The process is definitely bloated and bureaucratic. A big part of the challenge is it got that way is because more people died when it wasn't, and nobody has come up with a way to significantly streamline it without potentially reducing the safety.

the idea that no one has streamlined anything without reducing safety is a Straw Man at best and Ad Hominem at worst.


I agree, hence why I did not say that. The "significantly" in my statement was a critical part of it. Streamlining is happening all the time, but it's hard to identify major changes that can be made to the certification process to improve it, because requirements also tend not to get added arbitrarily without objections. Granted, there are some exceptions - the change to require copilots to have an air transport license (including 1500 flight hours) instead of just a commercial pilot license (250 hours) does not seem to address the issues behind the accident that led to that change, for example.

patrickjp93 wrote:
If you think I'm grandstanding on my ability, that's one thing,


I make no comment of your ability. However, I have a pretty strong suspicion you have not worked on aircraft certification before and are not familiar with the level of analysis, planning, and documentation involved in identifying necessary test cases. I suspect you are likewise unfamiliar with the level of effort that goes into dispensing with unnecessary test cases, and restricting to ground or subsystem tests those cases that are necessary but can be addressed without including them in flight testing plans.

patrickjp93 wrote:
There has never been a particularly effective government except the ones as minimal as; to take a finger in the wind, the ones of Lichtenstein, Switzerland, and the U.S.A.


The two European countries you mention are EASA members, which is currently pushing for more stringent requirements.

patrickjp93 wrote:
Yes, we get much more fuzzy-feely when regulators take a long, long time to approve things, but you have to have people like me keeping that emotional, irrational instinct in check, because in the end it doesn't improve anything to effectively repeat tests of the N-Xth degree when there's no justification.


I agree here, too, but again suspect you are speaking from conjecture, not experience in this industry.

patrickjp93 wrote:
And I'm one of those people who painstakingly proves, in absolute mathematical terms, that such tests are useless when the others are executed.


That is a valued skill in aerospace, and it is likely that you would be able to find work in the industry if you desired. I will not say that the aerospace industry is free from knee jerk "this is how we've always done it" or emotional "this seems more thorough" responses, but in the long run, data wins out if you stick to your case.
 
User avatar
PW100
Posts: 3817
Joined: Mon Jan 28, 2002 9:17 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:41 pm

patrickjp93 wrote:
ssteve wrote:
patrickjp93 wrote:
Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


2 devices can give you the throughput of 5 if the schedule allows higher utilization. Also, there's only one engine manufacturer here. They might just rather have your additional three planes more or less immediately ready to deliver.


If the utilization for each day isn't at least 8 air hours (takes time to offload testing dar), then there is PLENTY of room for fixing that too.

That is what Boeing thought on the 787 test program when the intial delays started to come out: compress test schedule using six frames. They found out the hard way that that is a rather high risk proposition.
Immigration officer: "What's the purpose of your visit to the USA?" Spotter: "Shooting airliners with my Canon!"
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 7:53 pm

PW100 wrote:
patrickjp93 wrote:
ssteve wrote:

2 devices can give you the throughput of 5 if the schedule allows higher utilization. Also, there's only one engine manufacturer here. They might just rather have your additional three planes more or less immediately ready to deliver.


If the utilization for each day isn't at least 8 air hours (takes time to offload testing dar), then there is PLENTY of room for fixing that too.

That is what Boeing thought on the 787 test program when the intial delays started to come out: compress test schedule using six frames. They found out the hard way that that is a rather high risk proposition.


It was high risk because of poor engineering and construction. You have to remember these variables, corollaries, and absolutes are independent. If I build an A350 right now exactly to Airbus' specs and quality with my own equivalent parts from other OEMs with the same quality process, I have an equal quality frame that WILL pass certification. So beginning with the assumption that I am an honest OEM who has painstakingly crafted my craft, there is zero risk in parallelizing the certification. The risk only comes if my product sucks, but then again I'm a psychopathic engineer and not a corner-cutting middle manager, so it wouldn't. Luckily, the management from Boeing that was cutting corners with the 787 and 737 MAX is largely gone already (hence no reports of firings from Boeing).
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:00 pm

iamlucky13 wrote:
patrickjp93 wrote:
iamlucky13 wrote:
The process is definitely bloated and bureaucratic. A big part of the challenge is it got that way is because more people died when it wasn't, and nobody has come up with a way to significantly streamline it without potentially reducing the safety.

the idea that no one has streamlined anything without reducing safety is a Straw Man at best and Ad Hominem at worst.


I agree, hence why I did not say that. The "significantly" in my statement was a critical part of it. Streamlining is happening all the time, but it's hard to identify major changes that can be made to the certification process to improve it, because requirements also tend not to get added arbitrarily without objections. Granted, there are some exceptions - the change to require copilots to have an air transport license (including 1500 flight hours) instead of just a commercial pilot license (250 hours) does not seem to address the issues behind the accident that led to that change, for example.

patrickjp93 wrote:
If you think I'm grandstanding on my ability, that's one thing,


I make no comment of your ability. However, I have a pretty strong suspicion you have not worked on aircraft certification before and are not familiar with the level of analysis, planning, and documentation involved in identifying necessary test cases. I suspect you are likewise unfamiliar with the level of effort that goes into dispensing with unnecessary test cases, and restricting to ground or subsystem tests those cases that are necessary but can be addressed without including them in flight testing plans.

patrickjp93 wrote:
There has never been a particularly effective government except the ones as minimal as; to take a finger in the wind, the ones of Lichtenstein, Switzerland, and the U.S.A.


The two European countries you mention are EASA members, which is currently pushing for more stringent requirements.

patrickjp93 wrote:
Yes, we get much more fuzzy-feely when regulators take a long, long time to approve things, but you have to have people like me keeping that emotional, irrational instinct in check, because in the end it doesn't improve anything to effectively repeat tests of the N-Xth degree when there's no justification.


I agree here, too, but again suspect you are speaking from conjecture, not experience in this industry.

patrickjp93 wrote:
And I'm one of those people who painstakingly proves, in absolute mathematical terms, that such tests are useless when the others are executed.


That is a valued skill in aerospace, and it is likely that you would be able to find work in the industry if you desired. I will not say that the aerospace industry is free from knee jerk "this is how we've always done it" or emotional "this seems more thorough" responses, but in the long run, data wins out if you stick to your case.


I have experience developing mainframe and SORCO software that passes far more conservative regulation than aerospace products of just about any kind outside military. Aerospace doesn't pay enough to be worth my time.

Separate the field from the job. If my job is to tell you how to effectively (and correctly) reduce your testing overhead, I don't give a damn what IS being tested insofar as the field. I care about what I can prove (mathematically, not subjectively) what is redundant or subsumable. I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein. If you're too vested in your own ideas to listen to facts (hence my use of mathematical proofs), then I can't help you. Call me a psychopath, but you're either right or wrong, and the world turns regardless.
 
iamlucky13
Posts: 1072
Joined: Wed Aug 08, 2007 12:35 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:19 pm

patrickjp93 wrote:
iamlucky13 wrote:
patrickjp93 wrote:

the idea that no one has streamlined anything without reducing safety is a Straw Man at best and Ad Hominem at worst.


I agree, hence why I did not say that. The "significantly" in my statement was a critical part of it. Streamlining is happening all the time, but it's hard to identify major changes that can be made to the certification process to improve it, because requirements also tend not to get added arbitrarily without objections. Granted, there are some exceptions - the change to require copilots to have an air transport license (including 1500 flight hours) instead of just a commercial pilot license (250 hours) does not seem to address the issues behind the accident that led to that change, for example.

patrickjp93 wrote:
If you think I'm grandstanding on my ability, that's one thing,


I make no comment of your ability. However, I have a pretty strong suspicion you have not worked on aircraft certification before and are not familiar with the level of analysis, planning, and documentation involved in identifying necessary test cases. I suspect you are likewise unfamiliar with the level of effort that goes into dispensing with unnecessary test cases, and restricting to ground or subsystem tests those cases that are necessary but can be addressed without including them in flight testing plans.

patrickjp93 wrote:
There has never been a particularly effective government except the ones as minimal as; to take a finger in the wind, the ones of Lichtenstein, Switzerland, and the U.S.A.


The two European countries you mention are EASA members, which is currently pushing for more stringent requirements.

patrickjp93 wrote:
Yes, we get much more fuzzy-feely when regulators take a long, long time to approve things, but you have to have people like me keeping that emotional, irrational instinct in check, because in the end it doesn't improve anything to effectively repeat tests of the N-Xth degree when there's no justification.


I agree here, too, but again suspect you are speaking from conjecture, not experience in this industry.

patrickjp93 wrote:
And I'm one of those people who painstakingly proves, in absolute mathematical terms, that such tests are useless when the others are executed.


That is a valued skill in aerospace, and it is likely that you would be able to find work in the industry if you desired. I will not say that the aerospace industry is free from knee jerk "this is how we've always done it" or emotional "this seems more thorough" responses, but in the long run, data wins out if you stick to your case.


I have experience developing mainframe and SORCO software that passes far more conservative regulation than aerospace products of just about any kind outside military. Aerospace doesn't pay enough to be worth my time.

Separate the field from the job. If my job is to tell you how to effectively (and correctly) reduce your testing overhead, I don't give a damn what IS being tested insofar as the field. I care about what I can prove (mathematically, not subjectively) what is redundant or subsumable. I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein. If you're too vested in your own ideas to listen to facts (hence my use of mathematical proofs), then I can't help you. Call me a psychopath, but you're either right or wrong, and the world turns regardless.


You can not fully separate the field from the job because the field is relevant to the requirements to be satisfied. I'm sure from your own study, you know some of the requirements. I'm sure from my exposure to the volume of them that you don't know anywhere near all of the requirements, so you don't have the information to disprove all the related test criteria.

I'm not dismissing mathematical proof, when it is offered.

I don't know how much you're worth to software companies, but if you can cut the 777X testing program to 4 months, you're worth tens if not hundreds of millions of dollars in aerospace.
 
StTim
Posts: 3446
Joined: Thu Aug 08, 2013 7:39 am

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:19 pm

patrickjp93 wrote:
OldAeroGuy wrote:
patrickjp93 wrote:
It's clearly not being done effectively. At most 2 test craft flying simultaneously.

Yup, not that complicated, and yes, there are redundant test points, and you can parallelize this across 5 craft at low risk and get it knocked out in 4 months even as is.


You are a quick study.

Which test points do you think are redundant?


Not quick at all. When I was a teenager I read Harry Potter & the Order of the Phoenix in just 14 hours. I've had over a week to read a document half as long by word count.

Let's take static testing as a starting point. As you go (incrementally) toward the 150% target, de-lamination, cracking, or structural failures will happen at any given percentage, so running so many incremental tests up to the 150% target is pointless. The static test frame can easily be rigged up with enough sensors to detect any component failure along the way. Start with pressure and climb from 0 to 150 until it fails or holds. Then on the wing and other structural flexes, climb 1% at a time up to 150 in 1 hour until a failure is detected or the 150 is held for the requisite time. Going by the 777X schedule, that's 3 days saved already. That's easy though and subject to simple inductive proofs. It's the flight testing where so many variables are measured in each test that take time to prove are redundant.

Brake testing can also be truncated. They already test on runways far, far longer than any plane is going to overrun. You can modify the thrust reverse test portion to incrementally climb from 0 to 100% reverse (failure happens when it happens and is detectable at exact point of failure) in one test and be perfectly fine. The runways for testing are long enough there's no risk even in a catastrophic failure. The same then goes for the maximum brake wear test. You don't need the leadup steps once the thrust reverser results are validated. If the manual brakes fail at that point, they need a fundamental redesign regardless of all previous tests, and the reversers, spoilers, and flaps will keep the plane from overrunning the runway (we can even add contingency bungees in the event of multi-system failure in the last 1/3 of the test runway to catch the weight if needed). So in total I have somewhere between 6 and 11 days of savable time using the 777X's testing schedule so far.


I suggest you take your proposals to Boeing (or Airbus) and ask for a mega salary. I mean if you can save them 8 months development time so easily I am sure the will snap you up. That would save then 10’s if not 100’s of millions.

Until then I suggest that extremely qualified engineers have planned the schedule in the best way they can - given what they know.
 
User avatar
bikerthai
Posts: 2965
Joined: Wed Apr 28, 2010 1:45 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:33 pm

patrickjp93 wrote:
I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein.


:rotfl:

As a 30+ years veteran in Aerospace Engineer design, I just want to say that a B.S. in Physics, and M.S. in Computer Science and umpteenth years of software design experience means this much :talktothehand: in aerospace structure design. Heck, even a B.S. or M.S. in Aerospace only count for maybe 5% of knowledge needed to design an effective aerospace structure. Not to say that those with the above degrees can not do well in the field, but more likely most of the lesson one learns is on the job training.

So before expounding on how much more you know about aerospace design than those who are in the field, be warn that the more you sound high and mighty, the more silly you'll come across.

I've had my run in with a few young up-and-coming know-it-all engineers in my time. And I'm still cleaning up their mistakes :sarcastic: as well as my own :bomb: :fight:

bt
Intelligent seeks knowledge. Enlightened seeks wisdom.
 
User avatar
glideslope
Posts: 1553
Joined: Sun May 30, 2004 8:06 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:42 pm

bikerthai wrote:
patrickjp93 wrote:
I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein.


:rotfl:

As a 30+ years veteran in Aerospace Engineer design, I just want to say that a B.S. in Physics, and M.S. in Computer Science and umpteenth years of software design experience means this much :talktothehand: in aerospace structure design. Heck, even a B.S. or M.S. in Aerospace only count for maybe 5% of knowledge needed to design an effective aerospace structure. Not to say that those with the above degrees can not do well in the field, but more likely most of the lesson one learns is on the job training.

So before expounding on how much more you know about aerospace design than those who are in the field, be warn that the more you sound high and mighty, the more silly you'll come across.

I've had my run in with a few young up-and-coming know-it-all engineers in my time. And I'm still cleaning up their mistakes :sarcastic: as well as my own :bomb: :fight:

bt


Beautiful. :checkeredflag:
To know your Enemy, you must become your Enemy.” Sun Tzu
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:42 pm

iamlucky13 wrote:
patrickjp93 wrote:
iamlucky13 wrote:

I agree, hence why I did not say that. The "significantly" in my statement was a critical part of it. Streamlining is happening all the time, but it's hard to identify major changes that can be made to the certification process to improve it, because requirements also tend not to get added arbitrarily without objections. Granted, there are some exceptions - the change to require copilots to have an air transport license (including 1500 flight hours) instead of just a commercial pilot license (250 hours) does not seem to address the issues behind the accident that led to that change, for example.

As of now I know a great deal of them, but it's in exploring the interconnectedness of variables that it gets messy. That said, if complexity ever deterred a good mathematician or engineer, we wouldn't have planes today at all.

And no, it's not proving to be all that hard just yet. Taking my more liberal 11-day estimate, that's 40 million USD saved by either major OEM, or effectively 40 million USD in taxpayer money.

Experience requirements are always a fickle, touchy subject. Experience is only 1 contributing facet of skill, and it's a double-edged sword in that repeated experiences form bias and can actually reduce how well any given person can move from narrow to broad focus and stay objective. I call this the "Old Guard Effect". It's not even THWADI, which is more specific. "The Old Guard" effect as I phrase it is where people become attuned to framing thoughts and arguments specific ways because that's what has always worked for them within their structures. It's also why the best people in their fields don't tend to keep a single position more than three years. Varied experience, less bias, more eclectic source material to work with, more ability to collaborate with people from other disciplines or viewpoints.

I make no comment of your ability. However, I have a pretty strong suspicion you have not worked on aircraft certification before and are not familiar with the level of analysis, planning, and documentation involved in identifying necessary test cases. I suspect you are likewise unfamiliar with the level of effort that goes into dispensing with unnecessary test cases, and restricting to ground or subsystem tests those cases that are necessary but can be addressed without including them in flight testing plans.

The two European countries you mention are EASA members, which is currently pushing for more stringent requirements.

People push for things for all kinds of reasons. People want their safety guaranteed and their governments to be quick to act and accountable. It takes cool, good heads slowing down the knee-jerk reactions for results to come out well at all, let alone optimal.

I agree here, too, but again suspect you are speaking from conjecture, not experience in this industry.

I have experience in a much more stringent industry, in case you've never dealt with financial regulators.



That is a valued skill in aerospace, and it is likely that you would be able to find work in the industry if you desired. I will not say that the aerospace industry is free from knee jerk "this is how we've always done it" or emotional "this seems more thorough" responses, but in the long run, data wins out if you stick to your case.


I have experience developing mainframe and SORCO software that passes far more conservative regulation than aerospace products of just about any kind outside military. Aerospace doesn't pay enough to be worth my time.

patrckjp93 wrote:
Separate the field from the job. If my job is to tell you how to effectively (and correctly) reduce your testing overhead, I don't give a damn what IS being tested insofar as the field. I care about what I can prove (mathematically, not subjectively) what is redundant or subsumable. I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein. If you're too vested in your own ideas to listen to facts (hence my use of mathematical proofs), then I can't help you. Call me a psychopath, but you're either right or wrong, and the world turns regardless.


You can not fully separate the field from the job because the field is relevant to the requirements to be satisfied. I'm sure from your own study, you know some of the requirements. I'm sure from my exposure to the volume of them that you don't know anywhere near all of the requirements, so you don't have the information to disprove all the related test criteria.

You have to do your damn best at separating the two, or objectivity and fact have no meaning.

I'm not trying to disprove all the criteria. That's a fool's errand. As I've said before my goal is to show redundancy, duplication, and therefore show how to optimize the process within its contextual bounds. Changing the context itself is a whole other ballgame.

I'm not dismissing mathematical proof, when it is offered.

I don't know how much you're worth to software companies, but if you can cut the 777X testing program to 4 months, you're worth tens if not hundreds of millions of dollars in aerospace.


Just because I can mathematically prove a truth doesn't mean I can convince dimwitted politicians to ignore the lobbyists from the competition who have a vested interest in keeping certification prices high. That's for glad-handing managers who get paid double what even the best Boeing engineer gets paid do when they otherwise are sticking their noses where they don't belong.
 
jagraham
Posts: 924
Joined: Sun Jul 17, 2016 11:10 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:45 pm

When it comes to test programs, they have to brief, fly, debrief, and analyze. That's why the test flying averages 4hr per day per plane until the end.
Also, some tests require other tests to be completed successfully. Opening up the flight envelope, for example, is something that cannot be rushed.
In the end, Boeing is putting 6 airplanes on the program. Yes, only 2 will be opening up the envelope, but that's something that will not reduce much time wise even if they use all six planes.
And flight test does not catch everything. It could not catch lithium battery issues on the 787. It did not catch MCAS on the 737MAX. They just don't deliberately break things (other than turn off one engine for engine out tests) and see what happens in flight. The whole test methodology would need to change to catch such things. And it would take 10x as long because an entire matrix of combinations of broken things would have to be tested. I have not seen a good answer to that type of things.
 
User avatar
glideslope
Posts: 1553
Joined: Sun May 30, 2004 8:06 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:47 pm

TropicalSky wrote:
Any update on if the GE9X is back in KPAE as yet?


No engines for flight testing are back. GE knows the issue. My understanding is they are still working with outside contractors on a new coating that will meet wear targets due to the increased heat signatures being uncovered. IMO, we will not see a 779 fly until after the new year.
To know your Enemy, you must become your Enemy.” Sun Tzu
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:48 pm

bikerthai wrote:
patrickjp93 wrote:
I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein.


:rotfl:

As a 30+ years veteran in Aerospace Engineer design, I just want to say that a B.S. in Physics, and M.S. in Computer Science and umpteenth years of software design experience means this much :talktothehand: in aerospace structure design. Heck, even a B.S. or M.S. in Aerospace only count for maybe 5% of knowledge needed to design an effective aerospace structure. Not to say that those with the above degrees can not do well in the field, but more likely most of the lesson one learns is on the job training.

So before expounding on how much more you know about aerospace design than those who are in the field, be warn that the more you sound high and mighty, the more silly you'll come across.

I've had my run in with a few young up-and-coming know-it-all engineers in my time. And I'm still cleaning up their mistakes :sarcastic: as well as my own :bomb: :fight:

bt


And leave it to A.netters to take what I say out of context. I'm not attempting to design an aircraft. I'm attempting to optimize a testing suite and remove duplicate test cases. That it involves aircraft or anything else that has behaviors is irrelevant to the core task.

That said, I've contributed to a number of open source C/C++ mathematical libraries (Boost and LibC++ in LLVM) used by Boeing and Airbus, so at least up to that point, I'm familiar enough with the math to probably smack plenty of engineers sideways. If you think PDEQ mathematics themselves are hard, try building a solver for them in binary that has zero rounding error out to 12 decimal places using just 64 bits that doesn't perform like garbage. Oh, and try to build the compiler and profiler tools to be able to optimize the use in a holistic program rather than just in a standalone program.

Engineers of today are not half as good as they appear. Behind all of them are good computer scientists.
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Tue Oct 08, 2019 8:57 pm

jagraham wrote:
When it comes to test programs, they have to brief, fly, debrief, and analyze. That's why the test flying averages 4hr per day per plane until the end.
Also, some tests require other tests to be completed successfully. Opening up the flight envelope, for example, is something that cannot be rushed.
In the end, Boeing is putting 6 airplanes on the program. Yes, only 2 will be opening up the envelope, but that's something that will not reduce much time wise even if they use all six planes.
And flight test does not catch everything. It could not catch lithium battery issues on the 787. It did not catch MCAS on the 737MAX. They just don't deliberately break things (other than turn off one engine for engine out tests) and see what happens in flight. The whole test methodology would need to change to catch such things. And it would take 10x as long because an entire matrix of combinations of broken things would have to be tested. I have not seen a good answer to that type of things.


If you wanted to catch everything, the 737 NG would still be in certification and the A320 CEO would be too. And you can certainly expand the flight envelope at a faster pace than currently allowed without issue. As for time, flight envelope testing is over 1/2 the battle of flight testing, so yes, speeding it up would be a prime target.

As for batteries, well, once again, proof that governments are hardly effective or efficient. The FAA came up with new rules for folding wingtips even though that is far better understood than Lithium Cobalt battery isolation, storage, and cooling under continuous charge and discharge load at 41,000 feet... Truth is, even now, if shrapnel struck the battery casings, they'd go up in flames in an instant and be a problem. The research into Carbon-Saltwater batteries is in full swing in part because of the problems with Lithium Polymer battery types.

Exhaustive testing is pointless. This is exactly why you need to draw information from the software world. Exhaustive testing is redundant when cases overlap or when combinations of events could not possibly happen (which is mathematically provable). There are known logic frameworks made in Answer-Set Prolog and other languages which were purpose-built to eliminate redundant test matrix items in astonishingly short order. The only problem with AS-P (Not ASP, stupid Microsoft...) is that the language is a royal pain to work with designed by academics instead of practical people.
 
OldAeroGuy
Posts: 3886
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 12:08 am

patrickjp93 wrote:
Not quick at all. When I was a teenager I read Harry Potter & the Order of the Phoenix in just 14 hours. I've had over a week to read a document half as long by word count.


Very admirable. I'm not sure of the relevance though unless it involves testing quidditch capable brooms.

The real question is your understanding of FAR Part 25 Certification Flight Test as described by AC25-7D.

patrickjp93 wrote:
Let's take static testing as a starting point. As you go (incrementally) toward the 150% target, de-lamination, cracking, or structural failures will happen at any given percentage, so running so many incremental tests up to the 150% target is pointless. The static test frame can easily be rigged up with enough sensors to detect any component failure along the way. Start with pressure and climb from 0 to 150 until it fails or holds. Then on the wing and other structural flexes, climb 1% at a time up to 150 in 1 hour until a failure is detected or the 150 is held for the requisite time. Going by the 777X schedule, that's 3 days saved already. That's easy though and subject to simple inductive proofs. It's the flight testing where so many variables are measured in each test that take time to prove are redundant.


Except the the structural load test is not on the flight test critical path. It needs to be complete before final Certification but it isn't needed to start flight test work. Zero days saved.

patrickjp93 wrote:
Brake testing can also be truncated. They already test on runways far, far longer than any plane is going to overrun. You can modify the thrust reverse test portion to incrementally climb from 0 to 100% reverse (failure happens when it happens and is detectable at exact point of failure) in one test and be perfectly fine. The runways for testing are long enough there's no risk even in a catastrophic failure. The same then goes for the maximum brake wear test. You don't need the leadup steps once the thrust reverser results are validated. If the manual brakes fail at that point, they need a fundamental redesign regardless of all previous tests, and the reversers, spoilers, and flaps will keep the plane from overrunning the runway (we can even add contingency bungees in the event of multi-system failure in the last 1/3 of the test runway to catch the weight if needed).


Exactly what are you testing for here? I don't know of any testing that requires an incremental investigation of reverse thrust. Please cite the relevant paragraph or page number in AC25-7D. Adding this testing would add about 5 days to the 777-9 test program.

patrickjp93 wrote:
So in total I have somewhere between 6 and 11 days of savable time using the 777X's testing schedule so far.


Actually, you've probably added 5 days by introducing unnecessary testing. Still looking for the duplicate testing you're going to eliminate.
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
User avatar
keesje
Posts: 13178
Joined: Thu Apr 12, 2001 2:08 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 12:20 am

5 Years back, when FAA allowed Boeing grandfathering design and requirements from the 77W there was already discussion.

The U.S. regulator has shown "a surprising amount of flexibility" allowing significantly updated aircraft with new engines and wings to be grandfathered, said Hans Weber, president of Tecop International Inc., a consulting firm that specializes in aircraft certification.

https://www.wsj.com/articles/boeing-777 ... 1398916634

Since then, it became visible shortcuts were taken on the MAX certification, with regard to sensor redundancy, emergency warning systems, fail save design and human interfaces. And that is for the 737 MAX, which seems an easy supplemental certification of the 737NG, compared to the all new 777X.

"It appears that Boeing took unfair advantage of a relaxed regulatory environment and, for the sake of speed and cost, cut corners in the certification process."

I wonder if EASA will require the 777x as a new aircraft instead of a derivative, because it has new engines, wings, tail, fuselage and cockpit. The blown out door might be traced back to grandfathered door design, placed in an fuselage with different load requirements.

We are watching the same Boeing, the same FAA and "streamlined" certification process on the 777X as on the 737MAX. Thinking the 777X is a totally seperate topic & nothings changed seems grossly optimistic. I expect more to be on the table than the GE9X and flying door; the way the 777X was certified. Starting with the special conditions, criteria and exceptions and what drove FAA to allow them, safety?

https://www.aerotime.aero/clement.charpentreau/24012-boeing-reorganizes-safety-organization-before-congress-hearing
"Never mistake motion for action." Ernest Hemingway
 
jagraham
Posts: 924
Joined: Sun Jul 17, 2016 11:10 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 1:35 am

patrickjp93 wrote:
jagraham wrote:
When it comes to test programs, they have to brief, fly, debrief, and analyze. That's why the test flying averages 4hr per day per plane until the end.
Also, some tests require other tests to be completed successfully. Opening up the flight envelope, for example, is something that cannot be rushed.
In the end, Boeing is putting 6 airplanes on the program. Yes, only 2 will be opening up the envelope, but that's something that will not reduce much time wise even if they use all six planes.
And flight test does not catch everything. It could not catch lithium battery issues on the 787. It did not catch MCAS on the 737MAX. They just don't deliberately break things (other than turn off one engine for engine out tests) and see what happens in flight. The whole test methodology would need to change to catch such things. And it would take 10x as long because an entire matrix of combinations of broken things would have to be tested. I have not seen a good answer to that type of things.


If you wanted to catch everything, the 737 NG would still be in certification and the A320 CEO would be too. And you can certainly expand the flight envelope at a faster pace than currently allowed without issue. As for time, flight envelope testing is over 1/2 the battle of flight testing, so yes, speeding it up would be a prime target.

As for batteries, well, once again, proof that governments are hardly effective or efficient. The FAA came up with new rules for folding wingtips even though that is far better understood than Lithium Cobalt battery isolation, storage, and cooling under continuous charge and discharge load at 41,000 feet... Truth is, even now, if shrapnel struck the battery casings, they'd go up in flames in an instant and be a problem. The research into Carbon-Saltwater batteries is in full swing in part because of the problems with Lithium Polymer battery types.

Exhaustive testing is pointless. This is exactly why you need to draw information from the software world. Exhaustive testing is redundant when cases overlap or when combinations of events could not possibly happen (which is mathematically provable). There are known logic frameworks made in Answer-Set Prolog and other languages which were purpose-built to eliminate redundant test matrix items in astonishingly short order. The only problem with AS-P (Not ASP, stupid Microsoft...) is that the language is a royal pain to work with designed by academics instead of practical people.


Exhaustive testing is not pointless. See SJ-30. Or Challenger 600.

In any case, because of MCAS, the testing regimes will become more rigorous, not less. Regardless of how wasteful any of us on a.net think they are.

In the new housings, the batteries are not nearly as vulnerable as you say. A small hole will do what the vent does, with the unfortunate side effect of having the byproducts vent into the avionics bay instead of out of the plane. It would probably take a hole of 10% of the surface area of the case before the intrinsic containment / oxygen starvation that the case provides would fail. A lot more than a bit of shrapnel. Not to mention that the batteries are in the tail and the shrapnel generating engines are on the wings. Perhaps the APU could be a problem, but it is behind the compartment containing the batteries and, like the engines, would primarily eject shrapnel in the plane of rotation. If the engines or the APU or any secondary explosion were to impale the batteries enough to cause containment to fail, that plane has a lot more problems than batteries on fire.
 
JayinKitsap
Posts: 1479
Joined: Sat Nov 26, 2005 9:55 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 4:02 am

bikerthai wrote:
patrickjp93 wrote:
I don't need experience in aerospace to be able to read testing frameworks and--using my B.S. in Physics, M.Sc. in Computer Science, and years of experience building software and tuning test suites--find the unnecessary pieces therein.


:rotfl:

As a 30+ years veteran in Aerospace Engineer design, I just want to say that a B.S. in Physics, and M.S. in Computer Science and umpteenth years of software design experience means this much :talktothehand: in aerospace structure design. Heck, even a B.S. or M.S. in Aerospace only count for maybe 5% of knowledge needed to design an effective aerospace structure. Not to say that those with the above degrees can not do well in the field, but more likely most of the lesson one learns is on the job training.

So before expounding on how much more you know about aerospace design than those who are in the field, be warn that the more you sound high and mighty, the more silly you'll come across.

I've had my run in with a few young up-and-coming know-it-all engineers in my time. And I'm still cleaning up their mistakes :sarcastic: as well as my own :bomb: :fight:

bt


In total agreement

As a 40+ year veteran in Building Structures and Facility engineering I have seen quite large divergences between digital models and the real world. Material properties often miss tempers, weld stresses, bonding issues, and the assumed fixity at the boundaries. With fluid flow, models often miss local turbulence and similar issues. How many test programs have issues with flutter? Yes, a lot of testing in Aerospace could be changed as we have moved to the digital world, but other testing should be added.

Engines are very high tech fields with tons of testing. But several recent engine programs have had problems, should certification have caught these, it is hoped but reality strikes. Nearly every certification program hits problems and delays, because things are not quite what the models say.
 
patrickjp93
Posts: 389
Joined: Thu Aug 22, 2019 12:00 pm

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 10:30 am

jagraham wrote:
patrickjp93 wrote:
jagraham wrote:
When it comes to test programs, they have to brief, fly, debrief, and analyze. That's why the test flying averages 4hr per day per plane until the end.
Also, some tests require other tests to be completed successfully. Opening up the flight envelope, for example, is something that cannot be rushed.
In the end, Boeing is putting 6 airplanes on the program. Yes, only 2 will be opening up the envelope, but that's something that will not reduce much time wise even if they use all six planes.
And flight test does not catch everything. It could not catch lithium battery issues on the 787. It did not catch MCAS on the 737MAX. They just don't deliberately break things (other than turn off one engine for engine out tests) and see what happens in flight. The whole test methodology would need to change to catch such things. And it would take 10x as long because an entire matrix of combinations of broken things would have to be tested. I have not seen a good answer to that type of things.


If you wanted to catch everything, the 737 NG would still be in certification and the A320 CEO would be too. And you can certainly expand the flight envelope at a faster pace than currently allowed without issue. As for time, flight envelope testing is over 1/2 the battle of flight testing, so yes, speeding it up would be a prime target.

As for batteries, well, once again, proof that governments are hardly effective or efficient. The FAA came up with new rules for folding wingtips even though that is far better understood than Lithium Cobalt battery isolation, storage, and cooling under continuous charge and discharge load at 41,000 feet... Truth is, even now, if shrapnel struck the battery casings, they'd go up in flames in an instant and be a problem. The research into Carbon-Saltwater batteries is in full swing in part because of the problems with Lithium Polymer battery types.

Exhaustive testing is pointless. This is exactly why you need to draw information from the software world. Exhaustive testing is redundant when cases overlap or when combinations of events could not possibly happen (which is mathematically provable). There are known logic frameworks made in Answer-Set Prolog and other languages which were purpose-built to eliminate redundant test matrix items in astonishingly short order. The only problem with AS-P (Not ASP, stupid Microsoft...) is that the language is a royal pain to work with designed by academics instead of practical people.


Exhaustive testing is not pointless. See SJ-30. Or Challenger 600.

In any case, because of MCAS, the testing regimes will become more rigorous, not less. Regardless of how wasteful any of us on a.net think they are.

In the new housings, the batteries are not nearly as vulnerable as you say. A small hole will do what the vent does, with the unfortunate side effect of having the byproducts vent into the avionics bay instead of out of the plane. It would probably take a hole of 10% of the surface area of the case before the intrinsic containment / oxygen starvation that the case provides would fail. A lot more than a bit of shrapnel. Not to mention that the batteries are in the tail and the shrapnel generating engines are on the wings. Perhaps the APU could be a problem, but it is behind the compartment containing the batteries and, like the engines, would primarily eject shrapnel in the plane of rotation. If the engines or the APU or any secondary explosion were to impale the batteries enough to cause containment to fail, that plane has a lot more problems than batteries on fire.


I'll try to hunt down the youtube video, but there's an ex-Boeing engineer who shoots even the revised casing with a 9mm round, and the LiCo battery incinerates itself incredibly quickly even hooked up to the supposed safety systems. He's been openly sounding the alarm on that faulty engineering for half a decade.

Edit: correction: an ex engineer of the company that ended up building Boeing's battery chargers. https://youtu.be/UqIzcuNpXP0?t=495 Battery section goes for roughly 6 minutes.
 
texl1649
Posts: 1065
Joined: Thu Aug 02, 2007 5:38 am

Re: Boeing 777X Testing Thread - 2019

Wed Oct 09, 2019 1:03 pm

The batteries haven't been an issue on the 777 ever, and I'm unaware of any testing being an issue on the 779. Back to the discussion at hand, I can't imagine any non-conspiracy/Harry Potter reason Boeing/GE would be flying an AN-124 from Columbus to Paine with large packages labeled GE9X unless those engines were actually being delivered for the flight test program.

Surely they are being or have recently been mounted again, in a hanger, to the first flight test article, for re-integration. Naturally as it's no longer the two dry months a year in Seattle this has to happen indoors, hence not any pictures yet.

Popular Searches On Airliners.net

Top Photos of Last:   24 Hours  •  48 Hours  •  7 Days  •  30 Days  •  180 Days  •  365 Days  •  All Time

Military Aircraft Every type from fighters to helicopters from air forces around the globe

Classic Airliners Props and jets from the good old days

Flight Decks Views from inside the cockpit

Aircraft Cabins Passenger cabin shots showing seat arrangements as well as cargo aircraft interior

Cargo Aircraft Pictures of great freighter aircraft

Government Aircraft Aircraft flying government officials

Helicopters Our large helicopter section. Both military and civil versions

Blimps / Airships Everything from the Goodyear blimp to the Zeppelin

Night Photos Beautiful shots taken while the sun is below the horizon

Accidents Accident, incident and crash related photos

Air to Air Photos taken by airborne photographers of airborne aircraft

Special Paint Schemes Aircraft painted in beautiful and original liveries

Airport Overviews Airport overviews from the air or ground

Tails and Winglets Tail and Winglet closeups with beautiful airline logos