055 Large Destroyer Thread II

[Iron Man]

On the contrary, I haven't missed your point at all. I wasn't suggesting that a GaN naval system was more sophisticated or harder to develop than one based on GaAs. I am saying that you have omitted the entire lineage preceding Type 346B and thus comparisons with the example land based AESAs in terms of developmental complexity is not a like for like comparison.

You wrote: "because you're building an entire software/hardware architecture around these radars that deals with the simultaneous tracking and engagement of hundreds (thousands?) of targets using 4 linked radar panels, and therefore a naval AESA system that debuts in the same year as a counter-battery radar could easily represent an earlier generation of technological mastery."

My counter argument is why do you think (underlined part) they needed to develop the entire software and hardware architecture for Type 346B to begin with, considering it had an existing software and hardware architecture already in existence with the Type 346/A?

If the Type 346B was a clean sheet design without any iterative preceding variants, then I could entertain your suggestion -- one could make a hypothetical comparison between the requirements of a new naval GaN AESA versus varying new land based GaN AESAs and their requisite performance demands.

Alas, Type 346B is not a clean sheet design, but a follow on variant of Type 346/A, so unless you can demonstrate that the hypothetical application of GaN on Type 346B would offer some sort of new unique challenges in terms of radar development relative to land based radars with GaN, then we should have every reason to accept that they would prioritize applicating a new technology like GaN to high priority domestic radar systems first before allowing it to be applied to other products and export products.

Furthermore, even if Type 346B was a new clean sheet design, it would still make sense that they would apply a new technology like GaN to domestic high priority projects first before allowing it to be applied onto lower priority projects and export cleared products.

Well perhaps not building entirely from scratch in this particular case, but certainly the larger size of the 346B compared to the 346A (by what, “40%”?), means larger radar aperture which leads to increased resolution, signal strength, and beamwidth, which in turns leads to more contacts acquired at greater ranges that will require tracking, more noise generated, better algorithms to improve the SNR, more processing power, more power requirements, more cooling requirements, etc. It’s not merely a matter of adding more T/R modules to the 346A, which BTW is what I think actually happened. I mean you’re basically trying to thread a narrow needle of 2 to 2.5 years time during which GaN for naval applications magically made it onto the 055 but somehow not the 052D. And given how incremental and conservative the PLAN has been historically, I would rather lean more conservatively towards a scaled-up GaAs radar for the 055 that was developed based on the 346A.

Possible, in the sense that it is a reasonable null hypothesis to take. That is the consensus.

It has yet to be definitively confirmed as I wrote in my last post -- however the burden of evidence should be to try and demonstrate that the idea of 055's Type 346B using GaN is an unreasonable one.

Reading between the lines here it sounds like you are claiming 346B on the 055 as the default assumption (null hypothesis) that needs to be argued against, since it is the “consensus”. I do hope you mean something other than “SDF consensus”, since there are no naval construction or radar experts here AFAIK, and consensus here frequently seems to be arrived at by a chorus of fanbois voting with laugh emojis rather than reasoned arguments. A far more reasonable basis for a “consensus” would be something like a big shrimp consensus. Do you know of at least 2 such people who feel it is more likely than not that the 346B uses GaN rather than GaAs, with working links that quote them saying such? I can even buy that kind of (bare bones minimal) consensus. A consensus of the people here? Nah.

I don’t get why people think just because a ship starts construction means the electronics need to be procured before that or even before the hull is completed and launch into water. The radars and other systems can be designed and built during fitting out after the hull is in the water.

Radars aren't just the radars themselves. Every radar system has its own computing requirements, its own control consoles/man-machine interfaces, and its own power and cooling requirements. If you fail to account for these correctly beforehand, any changes made after the ship is already mostly built out and sitting in the water is going to cost exponentially more money to change. Even if the change is just adding 40% more T/R modules, the knock-on effects of this are manifold and something which I already described earlier in this post.

 

[Blitzo]

If I take this quote(?) at face value, then this implies that the PLAAF seems not to have been developing anything from end of the 1970s to the mid 1990s when they became interested again in acquiring an AEW&C platform. If you are just suggesting that no active programs existed but the relevant technology was specifically being pursued as opposed to a more gradual, general maturation of the Chinese MIC, then there is no way to really tell whether that was the case unless someone official/semiofficial comes out and says it. I don’t think this position is necessarily unreasonable, but certainly it is unverifiable by any of us. In any case, regardless of prior history, what we can see evidence of is an initial platform only making its first appearance in 2003, its technological level mostly unknown, with the Chinese MIC having had no prior experience except for the prior KJ-1 which itself was considered obsolete by 1978; I believe there is a failed Y-10 AEW&C program squeezed into somewhere in the 70s as well.

To bring this all back to the present context of the 055 and the 346B, the KJ-2000 was being hailed as some kind of remarkable achievement for the purpose of claiming that the Chinese MIC had advanced enough to be one of the early utilizers of naval GaN technology later on. Given that the US and Russia were flying large AEW&C platforms since the 1970s and Israel since the 1990s, it’s hard to see the KJ-2000 as a major accomplishment without knowing the details of its performance capabilities, or how success in using AESA in the aerial domain translates into substantial gains in the naval domain. In fact the flow of causation is backwards, especially in the case of NRIET/CETC and this radar. Naval advances in radar predate airborne advances; you can easily tell this by noting the introduction dates of GaN-based fighter radars, which are only proliferating across the various leading world militaries in the last several years. I would also suggest that land-based radar advances predate naval advances given as I said previously their significantly lesser complexity to develop, which is why it would make sense to see GaN-based land radars prior to seeing GaN-based naval radars.

This basically just comes down to a difference in perception of where you and others think the PRC radar industry was by the 2000s and what constitutes reasonable extrapolation over the years since.

I’m not sure your personal incredulity should serve as the basis of an argument here, but even if we count first steel cutting and say it must have preceded the “laid down” dates for both the first 052C and the first 052D, that would give you, what, 6 months more time? 9 months more? BTW, now that I think back on it, there was a 055 model that came out of Wuhan in early 2014. That time period should actually serve as the latest date the radar system design could have been frozen, since IIRC that mockup was set up to verify radar non-interference, which means the radar designs involved should have been finalized by then, the only thing left in question being the final details of the superstructure and radar placements still awaiting verification.

Actually I think my personal incredulity can serve as the basis of an argument here, considering the base of knowledge/memory you seem to be operating on.
As for the elapsed time from steel cutting to launch, for 052D it is estimated at about 2 years (2010 steel cutting), for 055 it is estimated at about 3 years (2014 steel cutting).

Well perhaps not building entirely from scratch in this particular case, but certainly the larger size of the 346B compared to the 346A (by what, “40%”?), means larger radar aperture which leads to increased resolution, signal strength, and beamwidth, which in turns leads to more contacts acquired at greater ranges that will require tracking, more noise generated, better algorithms to improve the SNR, more processing power, more power requirements, more cooling requirements, etc. It’s not merely a matter of adding more T/R modules to the 346A, which BTW is what I think actually happened. I mean you’re basically trying to thread a narrow needle of 2 to 2.5 years time during which GaN for naval applications magically made it onto the 055 but somehow not the 052D. And given how incremental and conservative the PLAN has been historically, I would rather lean more conservatively towards a scaled-up GaAs radar for the 055 that was developed based on the 346A.

It makes sense to me why they would apply a new, larger GaN radar to 055 rather than 052D.
- There is a difference of 4-5 years between the equivalent stages of when 055 and 052D started construction, not 2-2.5 years. We cannot have this discussion in a sensible manner if you continue to argue that the 052D started construction (steel cutting) in 2012.
- 055 is a clean sheet, larger hull compared to 052D, which we can very reasonable extrapolate with having much greater power and cooling capacity to support not only a larger radar but a higher power antenna that GaN would enable.

Of course, the Type 346B, if it utilized GaN, would be more capable than Type 346A and would have some software and backend to be iterated and progressed for it. And a 4-5 year difference seems more than enough for that, especially considering the benefits of being able to carry out such measures on a larger clean sheet hull.

Reading between the lines here it sounds like you are claiming 346B on the 055 as the default assumption (null hypothesis) that needs to be argued against, since it is the “consensus”. I do hope you mean something other than “SDF consensus”, since there are no naval construction or radar experts here AFAIK, and consensus here frequently seems to be arrived at by a chorus of fanbois voting with laugh emojis rather than reasoned arguments. A far more reasonable basis for a “consensus” would be something like a big shrimp consensus. Do you know of at least 2 such people who feel it is more likely than not that the 346B uses GaN rather than GaAs, with working links that quote them saying such? I can even buy that kind of (bare bones minimal) consensus. A consensus of the people here? Nah.


Radars aren't just the radars themselves. Every radar system has its own computing requirements, its own control consoles/man-machine interfaces, and its own power and cooling requirements. If you fail to account for these correctly beforehand, any changes made after the ship is already mostly built out and sitting in the water is going to cost exponentially more money to change. Even if the change is just adding 40% more T/R modules, the knock-on effects of this are manifold and something which I already described earlier in this post.

Yes, I am saying that the null hypothesis is that the Type 346B is that it is likely utilizing GaN, which requires evidence to argue against the notion.
You can call it the SDF consensus if you wish and people here as fanboys as well, but the track record consensus here has also been a fair bit more productive than most publicly published "experts" whether they're thinktanks or public non-classified government reports.

If we had big shrimp consensus, then that would be considered definitive confirmation and we wouldn't be having this discussion to start with.


As for the last paragraph, I am very aware of that, as are most people here. That is entirely within accounted for why the idea of Type 346B being GaN is accepted as a reasonable null hypothesis.

 

[Iron Man]

Actually I think my personal incredulity can serve as the basis of an argument here, considering the base of knowledge/memory you seem to be operating on.

I'm not sure what you mean here. Can you extrapolate.

As for the elapsed time from steel cutting to launch, for 052D it is estimated at about 2 years (2010 steel cutting), for 055 it is estimated at about 3 years (2014 steel cutting).

Do you have any linked evidence for this alleged 2 year lead time? I have reviewed the old (closed) 055 thread, and it looks like official start of construction was December 2014, and first appearance of a module was October 2015. So in my estimation the time interval from steeling cutting to at least one visible assembled module is 10 months, which sounds like it's your view of what Wikipedia is referring to when it quotes a "laid down" date. 10 months for a new large class of ships means far less time than that for an essentially established class of ships already being built since more than a decade prior. This is of course all assuming we just blindly go by your definition of start date, which you describe as "steeling cutting". The actual start date has historically been "keel laying", or in the modern era of modular construction, the date of the placement of the first module into the building dock. In the case of the 055, this kind of distinction doesn't really matter at all, since as I already mentioned we have the benefit of a much earlier design-freezing event in the form of the Wuhan mockup which was noted by early 2014. Trying to make a case for electronics changes (or really indecision) that comprise type rather than location after this mockup showed up to me is extremely unreasonable.

It makes sense to me why they would apply a new, larger GaN radar to 055 rather than 052D.
- There is a difference of 4-5 years between the equivalent stages of when 055 and 052D started construction, not 2-2.5 years. We cannot have this discussion in a sensible manner if you continue to argue that the 052D started construction (steel cutting) in 2012.
- 055 is a clean sheet, larger hull compared to 052D, which we can very reasonable extrapolate with having much greater power and cooling capacity to support not only a larger radar but a higher power antenna that GaN would enable.

Of course, the Type 346B, if it utilized GaN, would be more capable than Type 346A and would have some software and backend to be iterated and progressed for it. And a 4-5 year difference seems more than enough for that, especially considering the benefits of being able to carry out such measures on a larger clean sheet hull.

Again I don't agree with your 4-5 year difference, and I also think that a 40% larger GaAs radar already accomplishes significant enough improvement to be worthy of installation in the 055.

Yes, I am saying that the null hypothesis is that the Type 346B is that it is likely utilizing GaN, which requires evidence to argue against the notion.
You can call it the SDF consensus if you wish and people here as fanboys as well, but the track record consensus here has also been a fair bit more productive than most publicly published "experts" whether they're thinktanks or public non-classified government reports.

If we had big shrimp consensus, then that would be considered definitive confirmation and we wouldn't be having this discussion to start with.

In that case my assumption is correct that this consensus is merely the (perceived) agreement of online amateurs with not much greater (if any) Chinese military knowledge than me, and I am content to readily disagree with whatever consensus you think exists here.

As for the last paragraph, I am very aware of that, as are most people here. That is entirely within accounted for why the idea of Type 346B being GaN is accepted as a reasonable null hypothesis.

That was actually a reply to Vincent, who seemed to not be aware of it. It happens to also be accounted for by the idea of a 40% enlarged 346A radar.

 

[totenchan]

Yes, I am saying that the null hypothesis is that the Type 346B is that it is likely utilizing GaN, which requires evidence to argue against the notion.
You can call it the SDF consensus if you wish and people here as fanboys as well, but the track record consensus here has also been a fair bit more

What is the basis for this? If the basis is simply the state of the Chinese radar industry, I don't see how that's a valid reason. I do think its perfectly possible that the Type 346B is using GaN, but without some sort of positive indication that it is, I don't see its possible how you can have a null hypothesis at all - there's just too little information. There's plenty of reasons why the Type 346B could be using GaA, least of which being that even as a GaA, the Type 346B would be probably among the most capable ship based radars through array size alone.

 

[Blitzo]

I'm not sure what you mean here. Can you extrapolate.

You were (and still seem to be) under the impression that the lead 052D cut steel in the same year it was launched, which is a very bizarre assertion.
It would genuinely be amazing and impressive for the PRC shipbuilding if they were able to start construction of an 052D and get it launched six months afterwards.

Do you have any linked evidence for this alleged 2 year lead time? I have reviewed the old (closed) 055 thread, and it looks like official start of construction was December 2014, and first appearance of a module was October 2015. So in my estimation the time interval from steeling cutting to at least one visible assembled module is 10 months, which sounds like it's your view of what Wikipedia is referring to when it quotes a "laid down" date. 10 months for a new large class of ships means far less time than that for an essentially established class of ships already being built since more than a decade prior. This is of course all assuming we just blindly go by your definition of start date, which you describe as "steeling cutting". The actual start date has historically been "keel laying", or in the modern era of modular construction, the date of the placement of the first module into the building dock. In the case of the 055, this kind of distinction doesn't really matter at all, since as I already mentioned we have the benefit of a much earlier design-freezing event in the form of the Wuhan mockup which was noted by early 2014. Trying to make a case for electronics changes (or really indecision) that comprise type rather than location after this mockup showed up to me is extremely unreasonable.

The definition for "start date" of a new ship's construction has been steel cutting/module fabrication for about the last two decades since modular construction became the norm.
In the more distant past, yes, keel laying would be the norm for the start date.

My definition of start date is the one which is common place -- no naval observer would seriously suggest in this day and age that commencement of a ship's construction begins with module laying down. The common starting point is steel cutting/first module fabrication, and it's very strange that you use the word "blindly". If anything the confident tone in which you've written your last few posts in this entire discussion has been inconsistent with your level of knowledge of this domain.

The appearance of the first 055 modules was indeed about October 2015 with a rumoured steel cut in very late 2014, however the first 055 modules sighted were not located in an equivalent position to when the first 052D modules were positively identified in early 2012. The first 055 modules sighted were located in a part of the shipyard that was a staging area prior to final assembly. It would not be until mid to late 2016 that we saw most of the major modules for the lead 055 together (more late 2016 if we include the bow module). This is observable via GE imagery over that time as well as corroborating with photos taken from the ground on that period.

The first 052D modules positively sighted in early 2012 were inside the covered building hall, with the bow extending outwards -- i.e.: the covered building hall is for final assembly of the modules, and extending sufficiently from from the covered building hall that they were visible from the outside.


I'm going to put a bit of effort into this to settle the matter.

The nature of imagery of JNCX and the frequency of concurrent destroyer production means that it is difficult to attribute specific modules to specific hulls when multiple hulls of the same ship are being built. For the lead 052D it was particularly difficult due to sharing the same hullform as the 052Cs that preceded it, making the lead 052D modules more difficult to identify at the time as well as in retrospect, compared to say the first modules for the first 055 hull.

This image was the earliest we have which we can confirm in retrospect to be the lead 052D, from February/March 2012. Circled in green is the lead 052D inside the final assembly area (a covered hall), next to it is the sixth 052C which is slightly ahead of it in assembly at the time this image was taken. As we can see, assembly was along far enough that the bow was externally poking out and visible, but it seems like the superstructure had yet to be installed.




For the lead 055, it is true that we caught external sighting of what was suspected to be parts of its first modules in JNCX's staging area in late 2015.
However, in terms of the hull modules of 055 (circled yellow) actually appearing at the final assembly area equivalent to where the 052Ds were built, that would not occur until mid 2016, and it wouldn't be until December 2016 that we saw the bow for 055 present with the rest of the modules for final assembly, similarly lacking all of its topside superstructures at that time which would only be installed by early/mid 2017 (not depicted in the images below)


Thus, a difference of 4-5 years in terms of equivalent construction stages.

For this matter, the idea of the lead 052D taking six months from steel cutting to launch is very much on you -- but the fact you are basing your argument on Wikipedia (whose mileage greatly varies, as I'm sure we are all aware), makes me wonder why you aren't considering the idea that perhaps the Wikipedia dates/information is simply incorrectly portrayed?

Again I don't agree with your 4-5 year difference, and I also think that a 40% larger GaAs radar already accomplishes significant enough improvement to be worthy of installation in the 055.

If you don't agree with the 4-5 year difference then there is nothing else to say.

As for a 40% larger GaA radar being worthy of installation on 055 as an upgrade, sure that is true. But it's not about what is worthy of being installed on 055 at the time, it is about what sort of material it utilizes for its antennae.

In that case my assumption is correct that this consensus is merely the (perceived) agreement of online amateurs with not much greater (if any) Chinese military knowledge than me, and I am content to readily disagree with whatever consensus you think exists here.

That was actually a reply to Vincent, who seemed to not be aware of it. It happens to also be accounted for by the idea of a 40% enlarged 346A radar.

You are free to continue to disagree with it, and people here will continue operating under that consensus until new information comes to light.

 

[Blitzo]

What is the basis for this? If the basis is simply the state of the Chinese radar industry, I don't see how that's a valid reason. I do think its perfectly possible that the Type 346B is using GaN, but without some sort of positive indication that it is, I don't see its possible how you can have a null hypothesis at all - there's just too little information. There's plenty of reasons why the Type 346B could be using GaA, least of which being that even as a GaA, the Type 346B would be probably among the most capable ship based radars through array size alone.

I certainly agree that it's possible that Type 346B could be using GaA, and I agree that utilizing GaA would continue to make it a very capable ship based on array size alone.
If we did not have evidence that they were already utilizing GaN on a number of different radar types that were publicly displayed and offered for export by 2014-2016, then I would agree a more conservative "we don't know" would be more prudent.


For the record, when I write about 055 in other places I typically just describe it as an AESA, and if one asks me whether I think it uses GaA or GaN, I say we don't know but there's a possibility it utilizes GaN, specifically for the purpose of being conservative.

 

[FriedRiceNSpice]

Whether the first few 055s use GaA or GaN is of little relevance. Whatever type of TR modules the first one or two hulls may be using, it is more probable than not the majority or 055s use GaN modules and whatever the number of hulls still using GaA modules may be, they can be upgraded in the future if the PLAN deems it necessary. It's frankly a waste of time spending pages arguing whether or not the first few 055s use GaN or GaA modules.

 

[Stealthflanker]

Whether the first few 055s use GaA or GaN is of little relevance.

Yeah. Also, one can perfectly make estimates in performance without having to need to actually know whether the High Power Amplifier technology being used is GaAs or GaN.