NS Military Realism Consultancy Thread Mk X Purps Safe Space

[Puzikas]

Google Translate deliveres a decent enough translation here

You dont miss anything important from it.

[United Earthlings]

Of the four-boat Trident fleet in the Royal Navy, only one is expected to be on a deterrence patrol at any given time, so the British are paying for four submarines and 64 missiles just to keep one submarine and 16 missiles ready.

For Future Reference,

IIRC, the Royal Navy Trident Fleet cycle goes like this: 1 on patrol, 1 standing up, 1 in reserve, 1 in maintenance. Also, IIRC the British never actually bought 64 missiles, but just enough to support two boats on deployment in addition since the 1990s the boats on patrol haven't even gone to see with all 16 missiles with the standard patrols again IIRC being at sea with 8 missiles max and furthermore with each missile only carrying around 4 to 6 warheads each.

China identified three major weaknesses in the USNs defenses: The current AEGIS/SPY-1/VLS triplet is not well suited for BMD

I don't think that's a bug on the US Navy's part, but a feature as it allows the USN to focus its priorities. Furthermore, with this shift in focus, you have the US Army and Air Force taking up the slack for BMD defense with their Patriots, THAADs and GDIs deployed around the Pacific Rim.

So, while yes the US Navy may have a weakness when it comes to BMD defense, when combined with Joint Operations doctrine as is pretty standard now for the US military, the weakness sort of balances itself out.

[Austrasien]

Greater capabilities aside the SPY-1D(V) radar is quite obsolete, but it simply cannot be replaced because the USN is unable to shepherd a replacement through the acquisition labyrinth. Even if no new capabilities were desired an AESA would eliminate the extremely complex "plumbing" the old PESA arrays require and improve their reliability by eliminating multiple single points of failure. The same reason the USAF leapfrogged straight to AESA technology.

An AESA variant, the SPY-1E, even exists and was developed a decade ago. But the change in the weight distribution (more in the radar panels, less below deck) is apparently problematic for the Burke stability because of the small size of the ship and high elevation of the panels, which is itself a symptom of the USNs inability to replace the Burke with a new hull that meets modern requirements.

[Gallia-]

What would a modern hull replacing Burke look like?

Just Zumwalt or sth else?

Also I guess this means my idea of "Gayla" sticking with PESAs until it can develop something super silly like air-cooled AESAs is a bit kooky. ):

But AESAs are so $$$. How will Industrialwar cope?

[Austrasien]

A space defense cruiser really. The Burke is fine for what it does, air defense, but the USN currently has no concrete plan for anything but more, better Burkes.

If the Strike Cruiser had ever been a thing its hull would have been a good starting place.

[Gallia-]

Interesting.

Which Strike Cruiser though? Weren't there multiple concepts?

Or do you just mean "=>20kt surface combatant"?

[Atomic Utopia]

The number and types of algorithms used for target recognition, counter-countermeasure, clutter suppression and the like is already very, very large. It is a field unto itself.

Guided weapons though generally do not learn. They usually have very short operating lives, and because the current paradigm is "wooden" rounds or as close to it as possible, generally do not receive any new data after being manufactured. Even if machine learning is employed it would be in R&D to develop or refine algorithms for this or that purpose that would be inserted into the weapon in a fixed form. When a missile is fired in combat it must work, there is no longer any opportunity to refine its behavior.

Well the training would be done during the development of the missile, where the neural net would be trained on target classification (spot the real target from the fakes) and other such things based upon training data extracted from combat information/spies who monitor enemy tests of their defenses. Then the program would be tested on data that was put aside for testing purposes, seen to be accurate enough, and the missiles programmed to use that version of the neural net to target enemies. Plus, you would simply have the missile transmit the data back to the computers on the plane, so new data would be provided to augment the training of missiles that are currently in production, or perhaps even missiles already produced, allowing for missiles to get more accurate with every intercept.

Though I can definitely see how that would be a problem with a missile computer, I had not anticipated how simple those things are, so the processing power would probably be too little. Which is odd considering how common ML is these days for civilian applications as a completely mundane tool. But regardless, do you have any links to articles on ECM/ECCM ML applications?

[Gallia-]

Phones and search engines are not particularly disposable items, contrary to what you might think.

Missiles are entirely disposable, single-use, and have a lifetime of minutes.

Machine learning for past threats it not terribly useful because it can't anticipate future threats.

[NeuPolska]

>Develop AI that will learn from its mistakes as well as any accessible information and constantly seek to improve/perfect itself to be the best possible war machine
>Connect said AI to internet and allow it to access everything including secret military files
>???
>Profit
(and bow down before our cybernetic overlords)

[Austrasien]

Well the training would be done during the development of the missile, where the neural net would be trained on target classification (spot the real target from the fakes) and other such things based upon training data extracted from combat information/spies who monitor enemy tests of their defenses. Then the program would be tested on data that was put aside for testing purposes, seen to be accurate enough, and the missiles programmed to use that version of the neural net to target enemies.

Things like this are already used.

But it is not actually practical to refine an algorithm through repeated testing. Missile tests are quite expensive and they become more expensive with each new generation of missile. Most of the testing has to be done with scene generators that are useful but only to a point. Simulated scenes do not perfectly match real scenes.

Plus, you would simply have the missile transmit the data back to the computers on the plane, so new data would be provided to augment the training of missiles that are currently in production, or perhaps even missiles already produced, allowing for missiles to get more accurate with every intercept.

Collecting detailed data from missiles engagements could be useful. But all the conflicts in the world since the end of the Gulf War have resulted in less than 50 air-to-air kills. So acquiring the data in useful amounts would actually be quite difficult. A dataset of air-to-air kills would include almost no 5th generation aircraft for example or many other highly important targets and advanced countermeasures would also be almost completely absent.

Given the very large number of parameters in a guided weapons engagement it would take an extremely large number of engagements to generate useful data. Clutter suppression algorithms based on real testing at army ranges was a major part of the Longbow radars development; and they worked. But they promptly stopped working when the Longbow radar was taken somewhere other than were the data had been derived from, even in apparently similar terrain (clutter suppression algorithms developed at desert ranges in the US did not work in Iraq). Creating a generalized model of radar interference from terrain (a common cause of guided weapons failure) proved to be impossible in practice and in theory, terrain as far as we know is truly random, and this has not changed substantially.

Which doesn't mean it is impossible. Animals natural senses work rather well at picking out difficult targets in adverse conditions and they were also created through an iterative process of adaptation. But their complexity is far beyond anything we can create right now and their precise functioning poorly understood.

The Longbow Stationary Target Indicator Technical History is worth reading.

[Gallia-]

>implying biological systems are more advanced than man-made

not very techno-optimist of you, viky!

i for one can't wait to have my eyeballs replaced by huge camera lenses that see in all spectra and have new colors flood my vision!

like those of grils who flee from cyber cock and clamshell exoskeleton. ):

[Onekawa-Nukanor]

Is there any point in time when it could be considered the time battleship be pcame obsolete because of the rise of the carrier?

I want to set a RP in the 1920s and 30s and wanted to know how far back I have to go where I don't have to worry about aircraft carriers ruining my sexy battleship and battlecruiser squadrons.

[Neo-Pontic Empire]

Is there any point in time when it could be considered the time battleship be pcame obsolete because of the rise of the carrier?

I want to set a RP in the 1920s and 30s and wanted to know how far back I have to go where I don't have to worry about aircraft carriers ruining my sexy battleship and battlecruiser squadrons.

I mean battleships on the open sea were fairly safe even during the late war, the Japanese only lost four battleships that I can think of entirely or primarily to aircraft when outside of port (Yamato, Musashi, Hyuga, and Fuso) and even then they were all lost in battles where their opponent had near air superiority (which your opponents could theoretically have, just listing stuff), more capital ships, and more support ships which were often more advanced. Case and point is Layte Gulf where the Americans had an 8:1 advantage in fleet carriers as well as 8 light and 18 escort carriers vs the Japanese 3 light carriers as well as a 12:9 advantage in battleships and a combined 300:67 advantage in total number of ships. Even then the Japanese only lost 3 battleships, 1 of them primarily to ship fire, relatively lighter losses than their carriers or cruisers.

Plus in the 1920s and 1930s naval aviation was a lot more primitive you shouldn't be seeing hordes of 300+ planes heading your way unless we are talking about NS numbers with populations in the tend of billions in which case your fleet probably has hundreds of destroyers and 50+ cruisers as escorts. That being said just because battleships are survivable doesn't mean they are useful, a floating cube 1 mile in each direction and 5 feet thick is hard to sink, doesn't make it useful. I would say if you are fighting a carrier centric force battleships start to lose out to carriers in terms of combat effectiveness in the late 30s early 40s.

[Laritaia]

>implying biological systems are more advanced than man-made

not very techno-optimist of you, viky!

i for one can't wait to have my eyeballs replaced by huge camera lenses that see in all spectra and have new colors flood my vision!

like those of grils who flee from cyber cock and clamshell exoskeleton. ):

Kat irl

[Gallia-]

blocked in my contrey

[Crookfur]

Is there any point in time when it could be considered the time battleship be pcame obsolete because of the rise of the carrier?

I want to set a RP in the 1920s and 30s and wanted to know how far back I have to go where I don't have to worry about aircraft carriers ruining my sexy battleship and battlecruiser squadrons.

You are pribably fine up.until the mid to late 30s. Probably the real turning point was taranto. Whilst it wasn't an underway engagement it's the event that woke everyone up to the power of naval aviation.

[Federated Kingdom of Prussia]

Is it true that interwar and WWII US Army had an obsession with reliability in its equipment because it had some pretty big problems with reliability in WWI? And can someone go into more detail on this?

[Austrasien]

Is there any point in time when it could be considered the time battleship be pcame obsolete because of the rise of the carrier?

WWI.

The aircraft carrier did not so much make the battleship obsolete as filled the void left by the decline of fleet actions, which battleships were supposed to dominate.

[Gallia-]

(...) even a very high-technology tank might not be able to rescue American hostages in a hotel in San Salvador or Manila.

kugelpanzers for eagle claw ):

[Scandinavian Nations]

What would a modern hull replacing Burke look like?
Just Zumwalt or sth else?

A slightly lengthened Burke.
Hull evolution is generally over. Mature technology. Like how home and office PCs have settled on being a roughly 8"x18"x18" box, laptops on a pair of hinged 14"x11" boxes.

Sure, every once in a while you'll see Apple try a cylindrical PC and Microsoft try a weird hinge with CPU in the screen box. There's your Zumwalt.

[Gallia-]

"Mature technology" sounds like a cop-out excuse for "diminishing intellectual input" or perhaps "laziness" given that the USN/NAMEs involved in CVN-78 didn't even bother to fix the design flaws of the Nimitz hull. :\

[Laritaia]

"Mature technology" sounds like a cop-out excuse for "diminishing intellectual input" or perhaps "laziness" given that the USN didn't even bother to fix the design flaws of the Nimitz hull. :\

i thought they did fix the list issue

[Gallia-]

"Mature technology" sounds like a cop-out excuse for "diminishing intellectual input" or perhaps "laziness" given that the USN didn't even bother to fix the design flaws of the Nimitz hull. :\

i thought they did fix the list issue

Maybe?

They never made the island out of super fiberglass or whatever they were considering to fix the list, though. Newport News said they wouldn't be able to make composite panels big enough or round enough to replace the island and piping that the Navy wanted. So they used a slightly lighter steel for the island. I'm not sure what steel they went with, but they had the choice of a 700 tons or 175 tons mass saving. They'd need to shave about 400 tons or so to correct the list.

[Laritaia]

so drunj carrier remains drunj

[Gallia-]

so drunj carrier remains drunj

https://www.onr.navy.mil/en/Media-Cente ... ion-CVN-78

They went with HSLA 115, which shaved 100-200 tons from the topside (the specific estimate was 175 tons) mass and reduced the list. I'm not sure if there are other, cumulative changes that total to a complete list elimination, though. It's probably still slightly drunj, but you already snickered about that a while ago TBF. HSLA 115 is a heat treated form of HSLA 100, which is used in the Zumwalt.

I guess HSLA 65 was too difficult.

Unfortunately it seems out civilization has pretty much reached intellectual inflection point and I guess from now on our outputs will be increasingly reduced until we're happy to make a twenty years out of date "modern" fighter, like Europe is. Time to pack it up and move to PRC if you want to be where the real engineering is I guess. Oh well.