COMAC C919

T

tamsen_ikard

Junior Member
Registered Member
latenlazy said:
Unless these industry experts have worked in a product development process before they don’t know what they’re talking about with regards to making major changes to important components. You don’t in fact need to redo the whole test stack every time you swap out an old component for a new component. The amount of time you need to retest a component swap in fact depends on 1) the maturity of the component itself, 2) the maturity of the overall system the component belongs in, 3) the level of risk involved in either the component or its swap. Testing a new engine is of course always going to be a very big deal, but outside of that most other non critical systems only need to demonstrate equivalent fulfillment of requirements. You also don’t need to swap to all domestic all at once. In fact the parts replacement process has both lower test burden and lower risk if you swap out components gradually as part of your continuous product improvement process, since it will allow you to put those components out into field to be tested in real operations without stacking risks of having everything new tested all at once. This is almost certainly what will happen with C919 indigenization, and in fact this process has already started.

Outside of a new engine and the avionics pretty much every foreign part in the C919 is low complexity and low risk, and only needs to demonstrate equivalent engineering performance requirements. The domestic avionics however can effectively be an indigenous copy of the foreign avionics forked so that it can be sustained domestically. The engines meanwhile will already have their own test and certification process so would only need to fulfill verification requirements at the integration testing level. While these things will take time they are also not the kind of full decade long reset of development that the sources cited in that article claims. Just because someone is a consultant in an industry doesn’t mean they actually know how product development in that industry works.

EDIT: And to the point I'm making here, after double checking the professional background of the consultant they're quoting, that dude has never seen time on a factory floor or an R&D lab.
Click to expand...

What you have said has given me assurance that it might not take that long as claimed in this article. But one of your assumptions could be wrong in the current geopolitical environment.

You said "You also don’t need to swap to all domestic all at once. In fact the parts replacement process has both lower test burden and lower risk if you swap out components gradually as part of your continuous product improvement process"

That kind of gradual replacement is only possible if China is allowed to continue to use those foreign components and actually make C919 in a large scale manner and have a viable business. Then they can slowly replace component while continue to manufacture planes and see them run in a commercial setting.

But the way US has been going crazy sanctioning China every other day, do you think US will allow China time to gradually replace components while developing an industry?

If this crazy pace of sanctions continue, there is a strong chance US will essentially stop all components export with a single order like putting COMAC on the entity list.

That essentially halts C919 production and forces COMAC to replace everything at once, which could require full redesign and recertification. That's why I said in the beginning that C919 project could be on borrowed time as US could essentially halt it with one single blow.
 
L

latenlazy

Brigadier
tamsen_ikard said:
Well, I'm not making my comment out of a hat. Its something I have read many times in news articles. Example, this one from 2022:

Please, Log in or Register to view URLs content!

Sash Tusa, aerospace and defence analyst at research firm Agency Partners says in this article,

“To take a C919 and turn it into a Chinese-only aircraft would require redesign, testing of every single certificate, most important of which is the engines. I’ll see you in the late 2030s,”

This is a statement made by an industry expert, and it is a reasonable one. Making major changes to important components like engines and flight safety system will essentially need a redesign and they will have to do re-certification tests, which takes many-many years.
Click to expand...
As a follow up to help clarify *why* what I described about continuous improvement engineering works the way it does despite the very high demands on safety and reliability standards, the *engineering requirements* that are spec'ed for any and all components in a plane is what defines their safety and reliability performance. When you are undergoing safety and certification testing you are not only testing your parts, but testing the qualification of those *requirements* that define how those parts were engineered to perform. Furthermore, even when you hand off the parts development to a third party, you *own* the information on how those parts work and how they're designed, if you didn't define the design of those parts outright to your 3rd party supplier, because this is an *essential* part of the due diligence you need to do for the integrity of your own product. You cannot stick a 3rd party part into your product which you don't understand or have transparency on, so as part of any components supply deal the 3rd party has to oblige to share enough details about what their components do and how they're put together (and in fact there is no point in trying to "hide" this information because once it's in your customers hands they could choose to dismantle and study it anyways).

What this all means in summation is that given that you know how all the parts in the plane already work and what engineering requirements they have been designed and built toward, anytime you need to swap out a part from one supplier to another you only need to verify that the new parts meet the same requirements as the old part, without treating the new part as a whole new black box that requires re-test of the *whole* system. The system's performance integrity is defined by the collection of requirements already verified and validated through certification and operations, so the only unknown you need to confirm with the new part in your testing process is "does my new part meet the same performance requirements as my old part".
 
Last edited:
L

latenlazy

Brigadier
tamsen_ikard said:
What you have said has given me assurance that it might not take that long as claimed in this article. But one of your assumptions could be wrong in the current geopolitical environment.

You said "You also don’t need to swap to all domestic all at once. In fact the parts replacement process has both lower test burden and lower risk if you swap out components gradually as part of your continuous product improvement process"

That kind of gradual replacement is only possible if China is allowed to continue to use those foreign components and actually make C919 in a large scale manner and have a viable business. Then they can slowly replace component while continue to manufacture planes and see them run in a commercial setting.

But the way US has been going crazy sanctioning China every other day, do you think US will allow China time to gradually replace components while developing an industry?

If this crazy pace of sanctions continue, there is a strong chance US will essentially stop all components export with a single order like putting COMAC on the entity list.

That essentially halts C919 production and forces COMAC to replace everything at once, which could require full redesign and recertification. That's why I said in the beginning that C919 project could be on borrowed time as US could essentially halt it with one single blow.
Click to expand...
Read my follow up.

If there is a ban of parts then COMAC has no obligation to maintain the integrity of any supplier agreements on IP protection and they could easily reverse engineer the parts without much if any difficulty. Most of these parts are not special beyond the assurances they provide to foreign certification agencies, and the make and workings of those parts are already known simply by necessity of the supplier integration process in your production overall development.
 
T

tamsen_ikard

Junior Member
Registered Member
latenlazy said:
Read my follow up.

If there is a ban of parts then COMAC has no obligation to maintain the integrity of any supplier agreements on IP protection and they could easily reverse engineer the parts without much if any difficulty. Most of these parts are not special beyond the assurances they provide to foreign certification agencies, and the make and workings of those parts are already known simply by necessity of the supplier integration process in your production overall development.
Click to expand...

If there is a ban lets say, this year, which could very well happen in the current geopolitical environment, how long do you think it could take to reverse engineer all these components? Even if its easy to do, it could still take 3-4 years. And all of those years, C919 production will have to be stopped. After all the reverse engineered and self developed components are made and tested, then comes integration into the overall system, which is also a difficult task which could take 1-2 years. Again, I'm making reasonable guesses which are not over the top. Thats already 6-8 years to just develop and integrate a fully indigenous C919. Then comes the safety and recertification tests. How long will that take? atleast 3-4 years.

So, it is a reasonable guess that if US sanctions COMAC then C919 project will be essentially be stopped from full commercial production for about a decade and even longer.

This is what is worrying, China's situation is extremely precarious right now when it comes to having a homegrown commercial aviation industry
 
L

latenlazy

Brigadier
tamsen_ikard said:
If there is a ban lets say, this year, which could very well happen in the current geopolitical environment, how long do you think it could take to reverse engineer all these components? Even if its easy to do, it could still take 3-4 years. And all of those years, C919 production will have to be stopped. After all the reverse engineered and self developed components are made and tested, then comes integration into the overall system, which is also a difficult task which could take 1-2 years. Again, I'm making reasonable guesses which are not over the top. Thats already 6-8 years to just develop and integrate a fully indigenous C919. Then comes the safety and recertification tests. How long will that take? atleast 3-4 years.

So, it is a reasonable guess that if US sanctions COMAC then C919 project will be essentially be stopped from full commercial production for about a decade and even longer.

This is what is worrying, China's situation is extremely precarious right now when it comes to having a homegrown commercial aviation industry
Click to expand...
Most of these "foreign" parts are *made* in China. That is part of the joint venture agreement. If a ban happened China would still own production of those parts. But even if you had to do a full reverse engineering, given that you've already validated your requirements and know how these parts work, and in fact own their design because of the earlier product integration process, you would only need to spend maybe 1-2 years tops doing reverse engineering and you could do integration testing *concurrently*. Reverse engineering for parts you already have the designs for is *not* a difficult or slow task. All the time consuming testing you had to do for certification is a one time thing, because it is effectively about testing the overall integrity and performance of the system's combined *requirements*. You *would* need to reproduce the whole time consuming stack of testing if you made major alterations to the design, but *not* if you're just swapping to alternative suppliers for the exact same part designed to the exact same requirements and specifications.

The only exception to this is the engine, which China doesn’t own at all and is effectively its own full stack modularized product attached to the plane.
 
tphuang

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
tamsen_ikard said:
If there is a ban lets say, this year, which could very well happen in the current geopolitical environment, how long do you think it could take to reverse engineer all these components? Even if its easy to do, it could still take 3-4 years. And all of those years, C919 production will have to be stopped. After all the reverse engineered and self developed components are made and tested, then comes integration into the overall system, which is also a difficult task which could take 1-2 years. Again, I'm making reasonable guesses which are not over the top. Thats already 6-8 years to just develop and integrate a fully indigenous C919. Then comes the safety and recertification tests. How long will that take? atleast 3-4 years.

So, it is a reasonable guess that if US sanctions COMAC then C919 project will be essentially be stopped from full commercial production for about a decade and even longer.

This is what is worrying, China's situation is extremely precarious right now when it comes to having a homegrown commercial aviation industry
Click to expand...
Again, that's not going to happen if you actually bother to look up earlier this thread to see where they are on all the various indigenization efforts.

They could've gone at this with like 90% domestic parts, but they chose not to because it would've slowed down the project. Another reason they used this much foreign subsystem is hoping to get certified by EASA.

But now that it's been certified and ready for revenue service any day now, they got more important immediate task at hand, which is to figure out how to work with airlines to provide the best after service support and to help airlines raise availability of C919 to a commercially acceptable level of over 99%. Domestic substitution is ongoing, but it's far from certain that C919 will actually be a successful airline project. So when you look at a commercial airliner, this is something you should be focusing on.
 
G

gadgetcool5

Senior Member
Registered Member
tphuang said:
Again, that's not going to happen if you actually bother to look up earlier this thread to see where they are on all the various indigenization efforts.

They could've gone at this with like 90% domestic parts, but they chose not to because it would've slowed down the project. Another reason they used this much foreign subsystem is hoping to get certified by EASA.

But now that it's been certified and ready for revenue service any day now, they got more important immediate task at hand, which is to figure out how to work with airlines to provide the best after service support and to help airlines raise availability of C919 to a commercially acceptable level of over 99%. Domestic substitution is ongoing, but it's far from certain that C919 will actually be a successful airline project. So when you look at a commercial airliner, this is something you should be focusing on.
Click to expand...
What, in your view, would be the definition of a successful airline project? What would define the difference between success or failure for the C919?
 
You must log in or register to reply here.
Top