Chengdu J-20 news and analysis Part III

[snne]

It seems to track. Also, I suspect that it's 25 squadrons of 12.




Yes, it's a possibility. That said, China has generally been open as to numbers. J-20 prototypes with series numbers 20XX. The J-36 prototype numbered 36011. Series 36, placeholder 0, 1st modification of design, 1st serial of that mod.

With years of experience tracking the serial numbers of the J-10, J-11, J-16, and J-20, we've found no evidence supporting any of the speculations being made here.

So, what drives this baseless speculation? Bias and a coping mechanism.

Get a grip, folks. Whether China produces more or fewer J-20s doesn’t affect your daily life in any way whatsoever...

 

[lancer21]

Remember these?

View attachment 784668

View attachment 784669

View attachment 784670

View attachment 784671

https://www.sinodefenceforum.com/t/j-20-5th-gen-fighter-thread-viii.9082/post-1126415
https://www.sinodefenceforum.com/t/j-20-5th-gen-fighter-thread-viii.9082/post-1126419
https://www.sinodefenceforum.com/t/j-20-5th-gen-fighter-thread-viii.9082/post-1126421
https://www.sinodefenceforum.com/t/j-20-5th-gen-fighter-thread-viii.9082/post-1126425


While this huge increase in metamaterial production doesn't necessarily mean more annual J-20 production, it is one of the things it strongly indicates...

Fascinating stuff. Let's not forget though that probably the H-20 and J-36/J-50 must feed of this metamaterial pipeline too, along with the UADFs, various UCAVs etc.

 

[MrPhantomMan]

They also from time to time release photos with numbers changed or obscured (which has been a habit since the 70s at least). So I think it's a fair question how many are there. I would suspect more than less though.

 

[Deino]

It seems to track. Also, I suspect that it's 25 squadrons of 12.

Not sure why you think the PLAAF is structured along squadrons? And even more so why 25 squadrons of 12???

It is well known that it operates Brigades consisting of usually 3 subunits aka daduis of usually 8 aircraft, so that nominally a current Air Brigade consists of 8+8+8 = 24 aircraft + twin-seater for training.
Recently it seems however that some brigades have been expanded and have either gained an additional dadui (+8) or the number of aircraft within one dadui has been increased.

 

[DWG]

Furthermore, the F-22 had an annual production of what? Less than 30 jets annually?

Probably best not to argue from F-22 production rates. Production was deep in the 'peace dividend' period and towards a final total of 195, vs a planned total of 750, which would likely have seen something more like 72 per annum..

 

[Blitzo]

The general consensus was at least 195-200 J-20 built by mid 2024. Rumours suggest J-20 production rate increasing from 70 a year to 100 a year around that time, meaning 300 by mid 2025 is consistent with expectations, and broadly matches observed deployments.

Not sure why China repainting and flying a small fleet of J-20s around the country trying to deceive Western observers.is more likely than them just building more.

It's more like 300 by mid 2024, and somewhere around late 2023 they'd increased their production rate to about 100-120 a year more recently.

The current fleet estimate is a floor of 300 minimum, but more likely to have a range between 350-450+ in existence (more likely to be closer to the 450+).

Correct me if I'm wrong, but isn't our official guesstimate over 350 by Q3 2024, and 420-450 airframes by now? AFAICR, from memory, around 50 airframes were estimated to be produced in 2022, and ~70 in 2023.

It's been a couple of months since I last kept track, but annual production has risen to more or less around 100 aircraft per year about a year and a half ago. I haven’t had the opportunity to visit SinoDef often lately, so please correct me if I'm wrong.

@Nx4eu , @Blitzo , @siegecrossbow , @Deino , @SUPER=BUG , et al. (whom I forgot to tag)

Edit: And also, the U.S. military typically procures ~75-85 F-35s per year across the USAF, USMC, and USN. By comparison, the annual procurement of the J-20 by the PLAAF alone exceeds that figure.

Add to that the number of J-35/35A along with various UADF and other CCA types they are going to procure by 2030, and it seems very likely that China will surpass the entire U.S. military’s fleet of LO platforms by the end of this decade.

As with many things China typically does, they first build their industrial might and then rapidly adopt it on an unprecedented scale...

Yes, your first paragraph is correct. I.e.: currently the estimate range is up to 450+ airframes for the confidence interval.

 

[Sferrin]

Probably best not to argue from F-22 production rates. Production was deep in the 'peace dividend' period and towards a final total of 195, vs a planned total of 750, which would likely have seen something more like 72 per annum..

Yep.

 

[NMaude]

Probably best not to argue from F-22 production rates. Production was deep in the 'peace dividend' period and towards a final total of 195, vs a planned total of 750, which would likely have seen something more like 72 per annum..

And the idiots who ordered the premature end of F-22 production are no doubt deeply regretting their extremely shortsighted decision.

 

[overscan (PaulMM)]

It's more like 300 by mid 2024, and somewhere around late 2023 they'd increased their production rate to about 100-120 a year more recently.

The current fleet estimate is a floor of 300 minimum, but more likely to have a range between 350-450+ in existence (more likely to be closer to the 450+).




Yes, your first paragraph is correct. I.e.: currently the estimate range is up to 450+ airframes for the confidence interval.

I meant the Western media consensus (Janes etc) rather than online forum consensus.

I've not really been following deployment of J-20 to units, but 300 is lowest believable value, while 400+ is quite possible.

 

[Blitzo]

I meant the Western media consensus (Janes etc) rather than online forum consensus.

I've not really been following deployment of J-20 to units, but 300 is lowest believable value, while 400+ is quite possible.

Considering the factored in timelag (behind forum consensus by about 1 year, typically) for western media/journalism estimates, that is fair.

Well, what the CB10300 image demonstrates is that we know the lowest possible confirmed number is 300, and that the real number is anywhere from a little higher than that, to much higher than that.
But 300 serves as the lowest floor.

(Essentially the question one asks is -- "how many additional J-20s may have been produced since that specific airframe was produced and inducted?")

 

[Scott Kenny]

Not sure why you think the PLAAF is structured along squadrons?

Because I think in terms of squadrons when we talk planes.

And even more so why 25 squadrons of 12???

12*25=300

It is well known that it operates Brigades consisting of usually 3 subunits aka daduis of usually 8 aircraft, so that nominally a current Air Brigade consists of 8+8+8 = 24 aircraft + twin-seater for training.
Recently it seems however that some brigades have been expanded and have either gained an additional dadui (+8) or the number of aircraft within one dadui has been increased.

Okay, so 12 standard Brigades of 25 birds.

 

[totoro]

In January 2025 Jim Taiclet, LM CEO, said j20 production had increased to over 100 per year. He may have relied on new or on out of date information when he said that, of course.

We had 156th j-20 serial spotted in January 2023. We had rumors of F-22 production run milestone surpassed in spring of 2023.

Several sources seem to Indicate production has met 100-120 ac per year for the last two years.

If indeed 200 were made by summer 2023, it's plausible we are at 400 , give or take, two years later. possibly a few more. And probably reaching over 450 by January 2026.

Thus a possible guesstimate of j-20 production: (figure per whole duration of the given year)
120 in 2025
100 in 2024
80 in 2023
60 in 2022
40 in 2021
30 in 2020
20 in 2019
12 in 2018
8 in 2017
4 in 2016 (as per LRIP rumor)

 

[Deino]

Because I think in terms of squadrons when we talk planes.




12*25=300




Okay, so 12 standard Brigades of 25 birds.

Do I speak Chinese??

NO; You may „think in terms of squadrons when we talk planes“ for another Air Force, but that‘s not valid for the PLAAF. Therefore again, „12*25=300“ is maybe a correct mathematical calculation, but why 300? - this particular 300th J-20 is definitely not the last one within its Brigade - and why this obsession with 12?

We already know there are more than 12 operational brigades (= 1, 2, 4, 5, 8, 9, 19, 41, 55, 56, 97, 111, 172, 176) and why again thsi nonsense with 25 aircraft each?
I explained you the divisor within the PLAAF is most often the factor 4, so a dadui (loosely translatable a s what you call a squadron) is 8 and each brigade consists of three dadui = 3 x 8 = 24 + the co-assigned trainer, in fact for the J-20 units this might simply be another 4.

So please do not try to squeeze the PLAAF into a system that does not work.

 

[Deino]

In January 2025 Jim Taiclet, LM CEO, said j20 production had increased to over 100 per year. He may have relied on new or on out of date information when he said that, of course.

We had 156th j-20 serial spotted in January 2023. We had rumors of F-22 production run milestone surpassed in spring of 2023.

Several sources seem to Indicate production has met 100-120 ac per year for the last two years.

If indeed 200 were made by summer 2023, it's plausible we are at 400 , give or take, two years later. possibly a few more. And probably reaching over 450 by January 2026.

Thus a possible guesstimate of j-20 production: (figure per whole duration of the given year)
120 in 2025
100 in 2024
80 in 2023
60 in 2022
40 in 2021
30 in 2020
20 in 2019
12 in 2018
8 in 2017
4 in 2016 (as per LRIP rumor)

IMO there was a drop of numbers produced per year when they changed (one after another) the production line(s) from the vanilla J-20 to the J-20A/AS.
IMO there is not a production run of 120 in 2025.

 

[Ainen]

IMO there was a drop of numbers produced per year when they changed (one after another) the production line(s) from the vanilla J-20 to the J-20A/AS.
IMO there is not a production run of 120 in 2025.

But was/is it in 2024 or 2025?
And, depending on whether clean transition is viewed as premium(for example, if China views situation as running on the clock), did it fully happen?

 

[totoro]

Who knows what happened during j20 to j-20a production switch.
Some of the existing lines that have made the switch may have had a pause. On the other hand, not all lines may have made the switch at the same time, in the same year or yet in total. Plus, additional lines are also possible.
2025 is not done yet so it's certainly possible 120 won't be met.

Relying mostly on cadence of new images from various new brigades as indication of production cadence is in my opinion not enough, due to fair chances the info is simply outdated, incomplete and trickles in at an uneven pace.

 

[DWG]

It occurs to me that what we have here is a potential target for the German Tank Problem method of analysis.

ETA: Deleted a comment that was just plain mathematically wrong.

 

[siegecrossbow]

There might have been been a slight decrease in production as China ramps up J-20A/S variant in favor of more vanilla J-20. Impossible to get real numbers until we see the serials.

 

[overscan (PaulMM)]

Part of the problem is understanding the serial number format. USSR c/ns had batch and aircraft components, but batch numbers were not necessarily contiguous.

Evidence from construction numbers is sketchy.

Initially, some people thought batch - number in batch, so CB3070 was interpreted as Batch 3, 70th in batch.
Other people suggested 20 per batch, with the number at the end being straight sequential. I've struggled to find known examples for most of these supposed batches with a quick Google search - maybe @Deino can help?

Batch	Pattern	Known examples
J-20 LRIP	XX00xx	XX0011
J-20 Batch 0 (AL-31F)	CB00xx	Any evidence?
J-20 Batch 1 (WS-10C)	CB01xx	Any evidence?
J-20 Batch 2 (WS-10C)	CB02xx	Any evidence?
J-20 Batch 3 (WS-10C)	CB03xx	CB0369, CB0370 (Zhuhai 2022)
J-20 Batch 4 (WS-10C)	CB04xx?	Any evidence?
J-20 Batch 5 (WS-10C)	CB05xxx?	Any evidence?
J-20 Batch 6 (WS-10C)	CB06xxx?	Any evidence?
J-20 Batch 7 (WS-10C)	CB07xxx	CB07156 (Nov/Dec 2023)
J-20 Batch 8 (WS-10C)	CB08xxx	Any evidence?
J-20 Batch 9 (WS-10C)	CB09xxx	Any evidence?
J-20 Batch 10 (WS-10C)	CB10xxx	CB10201 rumoured end of 2023, no proof CB10300 (September 2025)

Its just an assumption that CB10300 means batch 10, 300 total J-20 when we have so few examples to drawn from.

What about J-10 construction numbers?

 

[TwoTribes]

 

[Deino]

Part of the problem is understanding the serial number format. USSR c/ns had batch and aircraft components, but batch numbers were not necessarily contiguous.

Evidence from construction numbers is sketchy.

Initially, some people thought batch - number in batch, so CB3070 was interpreted as Batch 3, 70th in batch.
Other people suggested 20 per batch, with the number at the end being straight sequential. I've struggled to find known examples for most of these supposed batches with a quick Google search - maybe @Deino can help?

Batch	Pattern	Known examples
J-20 LRIP	XX00xx	XX0011
J-20 Batch 0 (AL-31F)	CB00xx	Any evidence?
J-20 Batch 1 (WS-10C)	CB01xx	Any evidence?
J-20 Batch 2 (WS-10C)	CB02xx	Any evidence?
J-20 Batch 3 (WS-10C)	CB03xx	CB0369, CB0370 (Zhuhai 2022)
J-20 Batch 4 (WS-10C)	CB04xx?	Any evidence?
J-20 Batch 5 (WS-10C)	CB05xxx?	Any evidence?
J-20 Batch 6 (WS-10C)	CB06xxx?	Any evidence?
J-20 Batch 7 (WS-10C)	CB07xxx	CB07156 (Nov/Dec 2023)
J-20 Batch 8 (WS-10C)	CB08xxx	Any evidence?
J-20 Batch 9 (WS-10C)	CB09xxx	Any evidence?
J-20 Batch 10 (WS-10C)	CB10xxx	CB10201 rumoured end of 2023, no proof CB10300 (September 2025)

Its just an assumption that CB10300 means batch 10, 300 total J-20 when we have so few examples to drawn from.

What about J-10 construction numbers?

Hmm! Hope this helps a bit with some cns. I know and a theory I'm working on since some time ..



Regarding the J-10C numbers, they follow a different system, which isn't helpful here.

 

[Scott Kenny]

[...] why again thsi nonsense with 25 aircraft each?
I explained you the divisor within the PLAAF is most often the factor 4, so a dadui (loosely translatable a s what you call a squadron) is 8 and each brigade consists of three dadui = 3 x 8 = 24 + the co-assigned trainer, in fact for the J-20 units this might simply be another 4.

3 x 8 = 24, + 1 trainer = 25

 

[Deino]

3 x 8 = 24, + 1 trainer = 25

Once again! + 4 trainer … at least pr my understanding NO PLAAF unit is arranged like you try to portray it and I really don‘t understand why you stick to rthis false theory?

 

[Scott Kenny]

Once again! + 4 trainer … at least pr my understanding NO PLAAF unit is arranged like you try to portray it and I really don‘t understand why you stick to rthis false theory?

You did not include the number of trainers initially.

nominally a current Air Brigade consists of 8+8+8 = 24 aircraft + twin-seater for training.

 

[Deino]

You did not include the number of trainers initially.

Sorry for this, but I explained the PLAAF has always the factor 4 in mind for the smallest active group of fighters!

 

[Scott Kenny]

Sorry for this, but I explained the PLAAF has always the factor 4 in mind for the smallest active group of fighters!

So, just to confirm my dumb ass is understanding, a typical PLAAF Fighter Brigade is 3 groups of 8x fighters, and 4x two-seater of same type? So, 24x J-20s and 4x J-20S (assuming that the -S is a trainer and not a version with a RIO/WSO/"guy in back")?

 

[overscan (PaulMM)]

Obviously, as the J-20S came later, J-20 brigades wouldn't have had any initially.

 

[totoro]

S is unlikely to be a trainer. Otherwise they would have produced it alongside the basic J-20, not 9 years later, after hundreds of basic single seater J20 have already been put into service and had pilots train for them.

As for a typical plaaf fighter brigade - sadly there doesn't seem to be such a thing. J10 units seem to have 28 planes (including 4 trainers), J11B units seem to have 24 (also some B trainers) Old Su-30 units had like 18 planes per unit. Some of the J-7 units had over 30 or possibly even approaching 40 planes. There is just not that much standardization between different plane types.

When it comes to J-20 units, I don't think there is a consensus yet as to its typical size. Most seem to have fewer than 30, some have close to 30, but some seem to have more than 30. It could be not all are fully populated yet, of course. But it could also be there is simply no hard standard.There are even some mixed type units.

I am sure Deino will give a better overview but it's not a simple, straightforward organisation.

 

[Deino]

So, just to confirm my dumb ass is understanding, a typical PLAAF Fighter Brigade is 3 groups of 8x fighters, and 4x two-seater of same type? So, 24x J-20s and 4x J-20S (assuming that the -S is a trainer and not a version with a RIO/WSO/"guy in back")?

Generally spoken yes, but as Paul noted, the J-20S came later and is not a trainer but a dedicated operational type. As such the J-20 units are a bit special and some (the 9th AB for example) have JL-10 as trainers while others more likely have just four more regular J-20.

 

[Scott Kenny]

S is unlikely to be a trainer. Otherwise they would have produced it alongside the basic J-20, not 9 years later, after hundreds of basic single seater J20 have already been put into service and had pilots train for them.

Agreed there, but I am obviously not tracking Chinese hardware well.

I am sure Deino will give a better overview but it's not a simple, straightforward organisation.

Because of course it's not. Heavens forbid that people have an org chart and use the damn thing... *grump, grump*

 

[SspewixzzZ]

Wow guys now we have official static art after 9 years!

Sourse and for more : https://weibo.com/6189120710/Q4OPsy48h

 

[SspewixzzZ]

Wow guys now we have official static art after 9 years!

Sourse and for more : https://weibo.com/6189120710/Q4OPsy48h

2025 Changchun Air Show (or Air Force Open Day) will officially begin on September 19th, but now in the warm-up phase, some great personal photography works have already appeared.
*I really should have taken time off to attend

 

[Kommoe]

2025 Changchun Air Show (or Air Force Open Day) will officially begin on September 19th, but now in the warm-up phase, some great personal photography works have already appeared.
*I really should have taken time off to attend

a buddy of my in Jilin University has a reservation to attend this weekend

 

[TwoTribes]

From Fighterman_FFRC

 

[AGS-1787]

Fairly small sensor on those side windows

 

[TwoTribes]

Front


Detail

 

[snne]

 

[FighterJock]

What is going on there snne? is that a kill marking by any chance? Or is it something else.

 

[lancer21]

That's just the photographer watermark.

 

[Life of Tyo]

View attachment 785123View attachment 785124View attachment 785125

lens Optic MAWS so tiny