February 17, 2023

Open Thread 124

It's time once again for our regular Open Thread. Talk about whatever you want, so long as it isn't Culture War.

Apologies for getting this up late. This is what happens when I play Aurora.

Overhauls are Classes, Aerial Cruise Missiles, Modern Propulsion Part 1, and for 2022 Norway Part 10 and Victory Ships.


  1. February 22, 2023muddywaters said...

    Did fleets often keep a straight course in battle because they were in too tight a formation to do much else without hitting each other?

    It's been suggested here that ships avoided turning in battle, and sometimes even fired easily-dodged torpedoes to make the enemy turn, because early fire control systems couldn't correctly calculate where to aim from a turning ship. However, it seems to me that they'd be even worse at aiming at a turning ship, and hence that this would be a reason to turn more, as turning would hinder your enemy more than you. (At least if the problem is allowing for the ships' velocities. I've seen it suggested that heel/vibration/structural bending can also matter, and those are reasons you don't want to be the one turning.)

    This suggests there may have been a different reason (or reasons) for not turning.

    Early in its history, the line of battle (i.e. a fleet in single line astern, firing to the side) was typically longer than its effective fighting range. Hence, it was an advantage to have the line as tightly spaced as possible, because when a tighter (and hence shorter) line faced a longer and looser line, all of the shorter line was able to fire, while part of the longer line was out of range. (The explicit statements I've seen of this are from 1794 and c.1906. Both these sources note that this is also a reason to prefer fewer larger ships, because they fit in a shorter line than the same total number of guns on more smaller ships.)

    A line that is as tight as possible for going straight, plausibly might be too tight for turning. (Particularly if the flagship was giving the order to turn via actual flags or similarly slow means.) Trying to turn anyway plausibly might mess up the formation enough that they have to stop shooting to avoid friendly fire, or even cause outright collisions.

    This would be less of an issue in smaller battles, and later time periods (better fire control = longer effective ranges, and better communication). A fleet that is already short compared to the range appears to have little reason to make itself even shorter, and hence can instead choose to form a looser line that can maneuver aggressively.

    However, I don't actually know that, and there could have been other reasons. (Including the simple one that if you're trying to get somewhere, a straight line is the shortest way there. This could be a physical place you're trying to attack or defend, or something like "closer so we can use our short-range weapons" or "away because if we stay in this battle we'll lose".)

    This might be related to why intentionally choosing non-broadside angles for their increased armor effectiveness is a thing in games (small virtual fleets), and possibly for cruisers, but mostly not for real battleships.

  2. February 22, 2023ike said...

    (At least if the problem is allowing for the ships’ velocities. I’ve seen it suggested that heel/vibration/structural bending can also matter, and those are reasons you don’t want to be the one turning.)

    Are you overlooking the keep-the-guns-pointed-at-the-enemy part of the problem?

  3. February 23, 2023muddywaters said...

    If you mean the ship's roll and pitch (i.e. aiming accurately from a ship at sea was hard, until it became possible to automate that part), I don't think of that as much worse when you're intentionally turning than when you're going straight, but that sentence was in part intended as that I might be wrong about that.

    (If you mean that turning too far from broadside-on will take some of your guns out of arc, I was probably thinking of that as too obvious to need to be said here. Sorry, I do tend to overdo that.)

  4. February 23, 2023ike said...

    I was talking about the even more obvious one. Turn the turret one arcsecond to the right every time the bow moves one arcsecond to the left. Good luck doing that with the information the turret crew has at its finger tips.

  5. February 23, 2023bean said...


    I'm currently working through Rules of the Game, and it at least incidentally bears on this. One of the major reasons for a lack of maneuvering, at least among the British, was probably the limitations of their signalling setup. Their system was very powerful, but also very slow. George Tryon was the leading advocate for a simpler "follow the leader" system, but his death and subsequent personnel assignments killed that off. I doubt it was a safety measure, as fleets were executing more complicated maneuvers fairly safely (with one obvious exception) decades earlier. The important thing there was apparently that all maneuvers were made at a constant speed, which removed a major variable which had previously made close-range maneuvering too dangerous.

    Re turning and fire-control solutions, I do think that's the correct answer. At least on the British side, the Dryer system would be completely messed up if either ship turned. Pollen's would handle things better, but it didn't see wide service and there were some technical limitations there. I suspect that it may also have been easier to do updates if the other ship maneuvered and you didn't, given that you have a stable baseline to work from in a way you wouldn't if you turned as well. Even during WWII, I believe only the USN could truly maneuver and just not care about the effects on its fire-control solutions. "But it will hinder the enemy more than us" isn't really how someone raised on a diet of scored practice firings is likely to think about tactics.

  6. February 23, 2023ike said...


    Speaking of reading, can you make an amazon-wishlist for us, so we can buy you books?

  7. February 24, 2023AlexT said...

    Isn't it possible that they didn't develop turning FC tools/techniques because there wasn't a real need for them? The idea being that it was vastly more valuable in combat to maintain a fast, cohesive formation.

    Considering battleships had enough firepower to inflict crippling damage to each other in a few salvos (at the right range), actual time under dangerous fire wouldn't be that long, so there would be little time for fancy sailing tricks to dodge shots. But maneuvering would bleed speed and formation integrity, so the fleet would become less able to bring max firepower to bear at the right time (crossing the enemy T etc).

    Thinking here mostly about the big North Sea battles in WW1, but maybe valid later too. Does this make sense?

  8. February 24, 2023muddywaters said...

    @ike: battleship guns do have sights (often visible on the side or top of the turret, though remote sights (directors) were the primary method after c.1915). However, the aiming controls could be awkward.

    As noted above, I think of the aiming part as hard but not much different whether you're intentionally turning or not. However, it's possible that this is where I'm wrong: I've heard that one British director would skip steps at high train rates, and it would explain why the 1940s USN (with RPC) didn't have a problem while others still did.

    It should be possible to distinguish those: do we know whether it was "keep straight" or "keep the enemy on the same relative bearing"? Those are different things if you're on non-parallel courses, even if the enemy doesn't turn.

    @bean: the version I've seen (from Dreadnought Gunnery, which does hint that it might not be generally accepted) is that both the Pollen and the Dreyer started off unable to handle turns, and both gained that ability at around the same time, possibly going through a phase where they could do it in theory but not in practice. (Operators who are (reasonably) afraid for their lives aren't likely to be at their best.)

  9. February 24, 2023muddywaters said...

    The idea being that it was vastly more valuable in combat to maintain a fast, cohesive formation.

    That's basically what I'm suggesting.

  10. February 24, 2023Bernd said...

    or even cause outright collisions.

    It never even occurred to me that ships in the "put rams on everything" era were fighting at ranges closer to those of the age of sail than Jutland. That helps explain that one disaster.

    @bean how's the new Aurora version? DLed, didn't open yet. I hear you can give dark eldar swirlies for dressing like a hot topic catalog now, so automatic GOTY.

  11. February 24, 2023Eric Rall said...

    It never even occurred to me that ships in the “put rams on everything” era were fighting at ranges closer to those of the age of sail than Jutland. That helps explain that one disaster.

    Sort of. AIUI, the "put rams on everything" era ran right up until the start of the Dreadnought era. And while long-range heavy gun were too slow-firing to be effective sole primary armaments for most of that period, the 6" quick-firing guns of the pre-dreadnought era had a theoretical effective range of several miles. The big limiting factors were gunnery and fire control, which were limited partially for technological reasons and partly because of doctrine.

    And doctrine bring us back to your comment: a lot of the doctrine up until the latter part of the Pre-Dreadnought era still assumed that major fleet engagements would be decided at close range, even though in hindsight we see that with good gunnery battles could have been fought and won at significantly longer ranges (not nearly Jutland ranges, but at least a couple thousand yards). As our esteemed host has discussed in the context of ship design, there was rather an extreme lack of major combat experience to inform warship design during the Ironclad and Pre-Dreadnought eras: up until the Spanish-American War and the Russo-Japanese War, the main significant major combat between iron warships apart from the American Civil War was the Battle of Lissa (1866). Both Lissa and the ACW were right at the beginning of the Ironclad era, before longer-ranged guns were available, and both were largely fought in littoral waters which forced closer ranges independently of gunnery. Ramming was used effectively in some ACW naval engagements and was decisive at Lissa, which naval designers and doctrine theorists overgeneralized for far too long for want of other real-world data.

  12. February 24, 2023bean said...


    That's a really good idea. I will try to get to that this weekend.

    @AlexT (first)

    I think it was mostly technical limitations. We live in a world of cheap, accurate gyros. They very much did not.

    As for not needing firepower for very long, that is not the case. Battleships proved themselves to be very durable things, so long as they weren't British battlecruisers. (Hood was unlucky, the others stupid.)


    It has been more years than I like to think about since I read deeply on that stuff. Pollen seems like it should be able to handle both types of turns about equally well, while I would expect Dryer to be more disrupted by an ownship turn. But that's based on high-level reasoning, not experience.


    Battle ranges grew significantly starting in around 1890. I should probably quantify that at some point.

    As for Aurora, 2.2 isn't out yet, I just got back into 2.1.

  13. February 24, 2023bean said...

    Actually, it looks like 2.2 is going to have some major changes to missile combat. This could get very interesting.

  14. February 25, 2023AlexT said...

    Battleships proved themselves to be very durable things

    Pursuing this tangent, has a battleship ever been hit by more than say a dozen heavy caliber rounds (similar to its main battery) without taking serious damage? I.e. not necessarily sunk, but with significantly reduced combat power and structural damage.

    Like, Warspite at Jutland was combat-ineffective from some 15 hits, and on the other side Konig was lucky to escape after a dozen or so, ish? And while Hood at Denmark Strait was bad luck, what happened wasn't inconceivable. And then in her final battle, Bismarck's main battery was effectively disabled in the first 15 minutes. The pattern I'm seing is that BBs were extremely survivable, except against each other.

  15. February 25, 2023muddywaters said...

    @AlexT: most sunk battleships were sunk by something other than shells, but it's plausible that shells were better at disabling ships than at actually sinking them. You previously mentioned an estimate of 10 hits.

    (Note that those 10 hits would often be from many more shots: ~5% hits seems to have been roughly typical.)

  16. February 25, 2023Hugh Fisher said...

    @AlexT, "extremely survivable except against each other" is pretty much why you build battleships in the first place. From the age of sail on the idea of a battleship is that if one shows up, everything smaller has to run away. If you can't send your own battleship in response, you lose control over that patch of sea.

    If you haven't already seen it, YouTube Drachinifel has a nice video "The Architecture of Dreadnoughts" and others which include discussion of the necessary trade-offs. Build your battleship so well armoured that it is largely invulnerable to all enemies? It will be either (or both) so slow that it takes too long to get where it is needed, or so undergunned that it can't damage an enemy battleship of similar size.

    As for examples of surviving, the British battlecruiser Tiger took a lot of heavy hits at Jutland but didn't lose any main guns or speed. While Warspite wasn't in great shape by the end of Jutland, she kept shooting her main guns throughout, quite effectively.

  17. February 26, 2023Fxbdm said...

    What is the advantage of having very large AWACS planes instead of a bunch of smaller ones (Drones!!) that offer much less of a juicy target?

  18. February 26, 2023bean said...

    Bigger radars are significantly better than small ones, so you can't easily replace it with a bunch of smaller platforms.

  19. February 26, 2023Anonymous said...

    OTOH more radars could give you multi-static capability.

  20. February 27, 2023Hugh Fisher said...

    @Fxbdm, @Anonymous the laws of physics say that the more power you send out, the further away you can detect things. The bigger your antenna, the further away you can detect things. (The radar range equation for energy transmitted vs energy received, even for non-stealthy targets, is brutal.)

    A big AWACS plane can cover say a 200km radius circle. (Yeah it's in 3 dimensions, but the altitude range for aircraft is very limited compared to distances.) If your drones have say 20km radius radar range due to being small, you need 100 to cover the same area.

    And the big plane can stay in the air for hours. Drones can't , so you need another 100 to take their places after a while.

    @Anonymous, by multi-static do you mean multi-spectrum radar? Or passive radar? If so, the big plane with a big modern antenna can still detect multiple frequencies at much longer ranges.

  21. February 27, 2023Basil Marte said...

    There are three main considerations.

    1) Power and antenna gain.
    To have a decent range, particularly in a context where adversaries are trying to jam it, radars both need to put out quite a lot of power, and need highly directional ("high-gain") antennas. Assuming no gross errors are made, the latter is a straightforward question of the antenna's size (actually as a multiple of the wavelength used).

    2) The information gained being of moderate usefulness.
    The ideal effect would be to have the radar drones fly immediately next to each other, dish-tip to dish-tip. That improves the antenna gain in the obvious way. If they fly farther apart (even wingtip to wingtip, but the wingspan is larger than the dish) then even in the theoretical best case of signal processing (problem 3.) what you get from a "sparse array" is mostly an interference pattern. You go from "the enemy is on a number between 25 and 35" to "they are on an even number between 26 and 34".

    3) The physical infrastructure of the signal processing.
    The precision requirements on everything are ridiculous if you are trying to treat it as a single dish. As in, millimeter errors in positioning and nanosecond errors in synchronizing turn into ~1° angular errors. Which is not a problem for a solid structure, but is simply not how aircraft fly.

    There are two practical use cases that I know of for radar drones.
    - Near-range radar "escort" for a stealthy aircraft. Have your radar picture and eat it too, as in, not eat an ARM.
    - Medium-range passive (non-emitting) RDF. A modern adaptation of artillery observation balloons looking for muzzle flashes and triangulating.

  22. February 27, 2023Bernd said...

    I'm disappointed the "very large drone array" won't be a thing. Thought that was a clever idea.

  23. February 27, 2023FXBDM said...

    And once more this community comes through with flying colours. I knew I could count on you guys explaining it clearly. Cheers!

  24. February 28, 2023Anonymous said...

    Hugh Fisher:

    And the big plane can stay in the air for hours. Drones can't , so you need another 100 to take their places after a while.

    Global Hawks are very much capable of staying in the air for hours.

    Hugh Fisher:

    @Anonymous, by multi-static do you mean multi-spectrum radar?

    No, multi-static means something else in radars.

    The idea would be to have the transmitter and receiver on different aircraft which could be widely separated so as to still be able to pick up signals reflected at a different angle than straight back.

  25. February 28, 2023Alexander said...

    @Anonymous a Global Hawk isn't that much smaller than a Hawkeye, and in fact has a significantly larger wingspan. I think that's more a point in favour of the 'one large AEW aircraft' plan rather than the 'more smaller drones' option, even if the RQ-4 is actually unmanned. Smaller AEW drones might still have their place though, as most navies have nothing that can get an E-2 into the air, and even the USN won't always have a carrier nearby.

  26. February 28, 2023muddywaters said...

    I think it was mostly technical limitations.

    One thing I like to think about is whether fire control had to happen when it did, or whether there are fun-to-imagine ways it could have gone differently.

    Available precision of manufacturing probably mattered, as is common in the history of technology generally. Precisely made cams etc, as analog computing accumulates error. Rifled guns (known since c.1500, but then impractical), because fire control can only be as good as the weapon's intrinsic precision. For central control systems and particularly directors, having the sights and mount axes aligned to give a tight pattern. (This one was noticed at the time: Fiske invented an electrical-transmission director in 1890, but didn't try to get it adopted because of the alignment issue. Percy Scott, in adopting centralized spotting in 1904, first had to deal with inaccurate sights.)

    I don't know how much of the proto-networking that went into central fire control was invented within the military and how much was pre-existing technology.

    We live in a world of cheap, accurate gyros. They very much did not.

    Dreadnought Gunnery does mention gyro error as a limiting factor, and you've previously suggested that the Dreyer was initially better because plotting range vs time works without a gyro. (The range axis probably needs a plot and rate fit more than the bearing axis does, because optical rangefinders and spotting are noisier than sights. Hence, it's more accurate (if neither ship turns) to average down that noise over some time rather than just using the current rangefinder reading + expected motion during the time of flight. Which raises the possibility that the 1940s USN's special ingredient might be radar??)

    Using a gyro to automatically fire when the ship's roll crosses the target was initially rejected, presumably because it was then worse than doing it manually, but became standard around 1917.

  27. March 01, 2023Anonymous said...


    a Global Hawk isn't that much smaller than a Hawkeye

    It appears to have a bit more than half the mass.

    Though the Reaper is significantly smaller and manages to stay airborne for more than a half a day, the even smaller Predator can do that too but with much less load and at lower altitude.


    and in fact has a significantly larger wingspan.

    Not needing to fit on a carrier does have its advantages.


    as most navies have nothing that can get an E-2 into the air, and even the USN won't always have a carrier nearby.

    The AW101 can do the job for anything big enough to operate it.

  28. March 01, 2023bean said...

    The bigger problem with a Global Hawk/Reaper based AWACS is that it isn't organic to the carrier, and anything which isn't organic can't really be counted on. You can't send one up quickly if you get warning of an incoming raid. It's either there or it isn't.

    As for the AW101, it's good, but it has neither the payload nor the endurance of the E-2.

  29. March 01, 2023Jade Nekotenshi said...

    It does seem to me that there might be some utility in having AW101s, or something like them, distributed around on amphibs and maybe even surface combatants.

  30. March 02, 2023Anonymous said...


    The bigger problem with a Global Hawk/Reaper based AWACS is that it isn't organic to the carrier,

    Neither is an E-3 yet the US has a lot of them (for the moment anyway).


    and anything which isn't organic can't really be counted on.

    Then use the Stingray (related engine to the Global Hawk so should have similar payload capability).

    Jade Nekotenshi:

    It does seem to me that there might be some utility in having AW101s, or something like them, distributed around on amphibs and maybe even surface combatants.

    The average surface combatant probably can't handle it along with the other helicopters it needs, though with Trimarans it might make some sense.

    Though the flattops the crayon eaters get really should have a few.

  31. March 02, 2023bean said...

    Neither is an E-3 yet the US has a lot of them (for the moment anyway).

    Yes, we have E-3s, but there's a reason we also have the E-2. One is for operation from land bases, the other from carriers.

    Then use the Stingray (related engine to the Global Hawk so should have similar payload capability).

    I don't think the MQ-25 is big enough to make a good AWACS platform.

  32. March 02, 2023Jade Nekotenshi said...

    As often as not right now, Ticos and Burkes deploy with one helo despite having two hangars. I'd think at least one or two could ship out with both a 60R and an AEW helo, though there is the slight problem of the US not currently having an AEW helo in service.

    I don't think this is a huge issue right now, but could be if fighting a near-peer: suppose the carrier temporarily loses the ability to launch its Hawkeyes due to a catapult casualty, damage to the flight deck, lack of parts or something like that - or maybe even due to the Hawkeyes in the air having to divert to a land base due to the carrier coming under attack. It seems to me that having some kind of over-the-horizon air search and direction would be better than none. Though I suppose if F-35B can do some of this, then at least LHDs can contribute to that mission somewhat.

    But an enemy ought to be able to detect when a CSG doesn't have AWACS up, and that would certainly be one of the better times to attack, if someone were going to.

  33. March 02, 2023muddywaters said...

    The AW101 is also used as an ASW and cargo helicopter, but I don't know if the same individual helicopter can do both. (The AEW system is referred to as a pod, but I don't know if it's actually removable, and it's probably more weight and drag than you want when not using it.) It's also physically larger than the MH-60R, so might not fit even as the ship's only helicopter.

  34. March 02, 2023Kitplane said...

    Why do Tico's and Burk's deploy with only one helicopter? Seems odd to me.

  35. March 03, 2023Anonymous said...


    I don't think the MQ-25 is big enough to make a good AWACS platform.

    Big enough to be an inflight refueling tanker.

    Jade Nekotenshi:

    Though I suppose if F-35B can do some of this, then at least LHDs can contribute to that mission somewhat.

    Modern fighter radars are good enough to function as mini-AWACS in a pinch.

    Jade Nekotenshi:

    suppose the carrier temporarily loses the ability to launch its Hawkeyes[…]

    Probably better to spend the money preventing that from happening and in the event it does use fighter radar.

    The use case would be to give the crayon eaters AEW capability when their flattops are operating alone.


    The AW101 is also used as an ASW and cargo helicopter, but I don't know if the same individual helicopter can do both.

    I'd be surprised, at the very least the AEW and ASW versions need crew inside the cabin to operate the equipment so don't expect much cargo capacity to be left (though the AEW system might be able to pick up a surfaced sub).


    It's also physically larger than the MH-60R, so might not fit even as the ship's only helicopter.

    LCS-2 is about the only US surface warship I'd have confidence of being able to fit it.

  36. March 03, 2023redRover said...

    Re alternate AEW options - a V-22 with a phased array like that on the wedgetail seems like it would achieve a lot of capability without requiring a full CATOBAR setup, while avoiding the issues of a true helicopter mounted system around range, endurance, and altitude.

    (You would need some engineering for hangar clearance as well as the additional interference from the antenna, both aerodynamic and signals wise, but those seem surmountable for a sufficient budget)

    Drones seem like the value is coming up with new concepts for them - if you try to replace an E-2 or E-3 with a UAV, you likely end up with something that's basically similar. But persistent cheap drones seem like they invite different use cases that augment traditional platform - instead of having a single platform providing overhead coverage of a battlegroup, you deploy a few of them far out to be basically aerial mines or whatever.

    Not that it's a perfect comparison, but I think you also see this with unmanned surface / subsurface vessels - we're not going to replace DDGs or SSNs with UUVs, because by the time you have equivalent performance in terms of speed and sensors and so on, you have a 10k ton platform that you might as well make manned (or at least is comparable to manned platforms in terms of size, cost, and so on). There are no free lunches.

    But for persistent slow speed surveillance or low intensity operations where quantity/presence is more important than quality, they're a better fit. (Or if you have a disposable platform, but naval warfare seems less amenable to that than ground warfare for various reasons)

  37. March 03, 2023redRover said...

    Also, re the AW101 and the DDG - it was designed as a Sea King replacement, and the RN uses them on their various surface combatants. If you poke around enough in the Navy documentation you can find landing envelopes for the H-3 on DDGs and CGs, though interestingly the H-3 had a wider envelope on the FFGs, not sure what drove that.

  38. March 03, 2023Jade Nekotenshi said...

    Yeah, Sea Kings can certainly /land/ on a DDG, but I don't think they can fit in the hangar. The CGs have a bigger, or at least taller, hangar, so they might fit there, and I think there were still Sea Kings in the fleet when the CG-47 class was built.

    My thought on the helo-based AEW is in addition to Hawkeye, though, not instead of. Also, as something to provide an eye in the sky for a surface combatant operating alone or as a SAG, with no flattop in sight.

  39. March 03, 2023Hugh Fisher said...

    I think it is useful to distinguish between AWACS the aircraft type and AWACS the role. I don't see small drones (or helicopters) replacing the AWACS aircraft type, but they can do the role.

    The big AWACS aircraft need to be that size for the antenna and power supply. (And multistatic doesn't help: if you want to detect the reflection of someone else's radar signal at range, you still need a big antenna.) The long radar range means that such aircraft can be not fast, not agile, not stealthy, not covered in defensive gun turrets and countermeasures. A dedicated AWACS aircraft is meant to stay well away from the actual fighting. In the future we may have drone AWACS replacements, but as noted above they still won't be small and cheap.

    If you don't have the really big antenna, your radar detection range comes down and you have to worry about being shot at. It seems to me that this is where AWACS becomes a role that a general purpose combat aircraft can perform, rather than having a dedicated aircraft type. The aircraft with the best radar, or in the best position, concentrates on keeping situational awareness for everyone else up to date. I believe the Iranians often used their F-14s for this in the 1980s, and today it's likely to be the F-35 in a mixed force. In the future I can imagine an R2-D2 style conversation within a drone squadron about who goes active on their radar and datalink transmitter while the others stay passive.

  40. March 04, 2023AlexT said...

    Re airborne radar, how expensive/complicated would a radar-illuminator drone have to be? The concept being to launch it in the general direction of (suspected) enemy forces, and after travelling a set time it starts turning and emitting. Friendlies stay back and pick up reflections. Hopefully the enemy shoots it down with something 10x more expensive. Essentially, a cheap expendable way to increase (quadruple?) the range of a radar of a given size, and keep the precious receiver with all the cool electronics and people passive and far away.

  41. March 04, 2023ike said...

    The other design that it suggests is the ultra-stealthy-long-range-pop-gun-intecptor for anti-AWACS duty.

    @AlexT I feel the power requirements for the illuminator would drive up the size & cost too much.

  42. March 04, 2023DampOctopus said...

    Basil and Hugh have made the important point, about requiring a big antenna to detect a radar reflection at range. To state it a bit more formally: to have the same radar detection range as a single big AWACS, a group of smaller radar aircraft needs to have the same total antenna area and transmission power.

    Using multiple smaller radars has a few advantages, though. It has better resolution, so you can tell the bearing to your target more precisely, making it easier for you to determine the number of aircraft in a close formation, or to spot targets in your Doppler notch by tracking their lateral movement. And a lot of stealth design is about ensuring radar reflections don't go straight back to the transmitter, so if the receiver and transmitter are separated (e.g. there are several of each), stealth is less effective.

    @Basil, the above means that I agree with your first counterargument to the viability of a group of smaller radars, but I disagree with your remaining two points. Contra your second point: the information from a sparse array can be difficult to interpret, but the ambiguity goes away rapidly as you add extra aircraft to the array. To illustrate this by extending your example: you start with "the enemy is on a number between 25 and 35"; with two aircraft you have "the enemy is on an even number between 26 and 34"; and with three aircraft you have "the enemy is on a number between 25 and 35 that is a multiple of 2, 3 and 5" ... which pins it down precisely.

    @Basil, contra your third point, the precision requirements don't seem unreasonable to me. You don't need to keep your aircraft in perfect formation to millimetre-level precision: you just need to keep track of how much they deviate from that formation. Similar problems have already been solved, with suitable calibration: VLBI arrays maintain coherence to about the same precision when one element is halfway to the moon; adaptive optics maintains coherence only across a ~10m telescope mirror, but to sub-micron precision.

    The real killer, I think, is the signal-processing requirements. The required processing has components that scale linearly with the number of elements, quadratically with the number of elements, linearly with the number of elements you could fit in the gaps in your sparse array, and quadratically and possibly cubically with this last number. If you have, say, 1m antennas flying in a formation 1km across, that's a billion-fold increase in the amount of computing you need to do.

  43. March 04, 2023Basil Marte said...

    @AlexT Not at all, but such a "radar starshell" would be mostly useless, because in military radar, basically the emission side is difficult and you get the reception side almost for free.

    If you were trying to detect civilian aircraft (or WW2 bombers), the idea would make complete sense, because ...you have gain on both sides. You can use a good telescope and a wimpy light source just as well as the reverse.

    The problem is that the enemy can (and does) put jammers onto potentially every aircraft. Now whatever you use for illumination has to put more power onto a/the target aircraft than its jammer can produce, otherwise you'll be seeing the jammer, not the actual echo. (If your reception side is good you'll be seeing the jammer very clearly, but that's not exactly useful for most purposes.) And it turns out that generally it's more practical to increase power density by using a larger antenna to concentrate the emitted power into a narrower beam than it is to emit more power.

    Given that, the least unviable idea is to put an EMP warhead on a missile. For the next step up, producing several seconds of illumination, most of the payload mass goes into a rocket-turbine-alternator. The electronics are open-cycle cooled i.e. are boiling off a bath of water. For this to not have terrible range, the initial idea I have is to go for a "wide strake" aerodynamic surface (<1 aspect ratio wing), put the fixed antenna array into the bottom surface, and have the program be: roll 90°, emit while pitching "up", thus sweeping the fan-shaped beam.

  44. March 04, 2023Lambert said...

    What are the challenges involved in powering a phased-array radar? Could you run a jet engine with a big honking alternator for power and regeneratively cool the electronics with JP-5? Since you're not lugging around delicate meatbags and their life-support systems, a UAV could devote a higher mass fraction to heavy-duty cables and antennas.

  45. March 04, 2023Basil Marte said...

    The phase-shifters are expensive (one for each row/column, whichever there are more of).
    Separately, what airframe-relative plane is the antenna in?
    Other than that, more or less yes. Alternator->rectifier->pulse forming network->emitter->antenna. (And I expect more cooling is needed.)

  46. March 05, 2023Anonymous said...


    regeneratively cool the electronics with JP-5

    Like the Nimrod AEW.3?

    No, that's a bad idea.

  47. March 05, 2023Hugh Fisher said...

    Basil Marte More power isn't the only solution to jamming these days. When I did a (non-classified) introduction to radar short course a couple of years ago, that was one of the key points the instructor wanted us to take away. Modern electronics and phased arrays do tricksy things with the outgoing signal so you can extract the reflection even from massive amounts of jamming noise.

    (This does however make multistatic radar harder, because you need precise knowledge about the outgoing signal which is obviously a heck of a lot easier when you sent it.)

    In some ways though this is just moving the power requirement from the transmitter to the ultra high speed parallel processing chips in the receiver. The same course instructor described a typical phased array radar as a massive air conditioner with an antenna attached.

Comments from SlateStarCodex:

Leave a comment

All comments are reviewed before being displayed.

Name (required):

E-mail (required, will not be published):


You can use Markdown in comments!

Enter value: Captcha