Thursday, March 19, 2015

Third-Generation Nuclear Weapons

I ran into this stuff while working on something else, and thought it was interesting enough to be worth sharing.   I may or may not do a longer article on this at some point; getting more information on this is likely to be very difficult, given that the US government holds this stuff pretty close to its chest (as it should).

So what, exactly, is a third-generation nuclear weapon?

That's a bit subjective, to be honest.   Most sources describe the first generation as the atomic bomb, and the second generation as the hydrogen bomb,[Fe] though I've also seen first generation used to refer to the bomber-delivered, bulky, high-yield H-bombs of the '50s and early '60s, and second generation for the missile-delivered, miniaturized, mid-yield weapons of the late '60s on.[Su]   In either case, a third-generation weapon is a bomb that has a "tailored effect"; it's optimized to produce most of its energy in one particular form or direction.   Some sources include "salted" nuclear weapons and neutron bombs as third generation weapons, two concepts that go back to the '50s.   But, for the most part, a "third-generation" weapon is one with some degree of directionality: it emits most of its energy in a narrow cone rather then spread out over all directions.

For this article, I'm going to limit myself to this later definition, since there's already enough information out there about salted bombs and neutron bombs.   These directional devices are also often referred to as Nuclear Directed Energy Weapons, or NDEWs.


What was Casaba-Howitzer?   I don't entirely know.   (You're going to be hearing that a lot in this essay.)   It's not even entirely clear if it was a single project, called Casaba-Howitzer, or two separate projects, Casaba and Howitzer.   On the net it's usually referred to as a single program, probably because that's how it's described in George Dyson's excellent book Project Orion
.[Dy]   Dyson conducted extensive interviews with people who worked directly on these programs back in the '60s, and includes a quote from one of them referring explicitly to "Casaba-Howitzer", so I'm reluctant to contradict him.   But all my other sources refer to these as two separate programs.[Fe][St]   So I dunno.

Update (09/29/15): Scott Lowther says he has a General Dynamics bibliography on Casaba-Howitzer that specifically refers to it by that name, so I guess that settles that issue.   See here, in the comments.

What I do know is that Casaba and/or Howitzer was an outgrowth of Project Orion, the attempt in the late '50s to build a spaceship that would be driven by detonating nuclear bombs.   Now, an ordinary nuclear bomb produces a spherical pulse of energy, so a lot of that energy is wasted since it doesn't hit your ship's pusher plate.   What Orion and Casaba-Howitzer did was use a radiation case to direct a large portion of that energy to a pancake of propellant.[Dy]   As Freeman Dyson describes it:
"If you have something that starts in the shape of a pancake and you heat it up to a very high temperature it will expand more sideways along the axis, and less at the edges.   The pressure gradient is highest along the axis, so then after a while, since the velocity is highest along the axis, it becomes cigar-shaped.   So you get inversion, something that begins like a pancake becomes like a cigar, and something that begins as a cigar becomes a pancake, if you just let it expand freely."[Dy]
So the Orion pulse propulsion units and Casaba-Howitzer warheads would both produce a spear of plasma.   The Orion spear would be somewhat wider and less focused, and used to push a spaceship.   Casaba-Howitzer would be narrow, and used destructively, as an anti-ballistic missile warhead or anti-satellite weapon.[Dy]

Besides Orion, Casaba-Howitzer may have also been connected to a program called GLIPAR, or Guide Line Identification Program for Anti-missile Research, run by the Advanced Research Projects Agency.   GLIPAR contracted a variety of groups to investigate any and all ideas, no matter how seemingly ridiculous - even things like antigravity or magnetic "shields".   This apparently included some kind of "nuclear cannon" or "nuclear howitzer" concept, though details are completely absent.[ARPA]

Beyond that, we know that the idea was originally suggested by Moe Scharff at Livermore Laboratory.   We know that there was at least one nuclear test related to Casaba-Howitzer, though not how many there were, what their code names were, or when they happened
.[Dy]   It apparently did not work as well as they hoped - a later report, which I believe is referring to Casaba-Howitzer though I'm not certain, mentions that they hoped to achieve a plasma velocity of 50 km/s, but "only a fraction of... [that] was achieved."   The theoretical maximum velocity for this concept is apparently 1000 km/s, but it looks like no one ever came close to getting that.[Fe]

Shaped Nuclear Charges

This is an idea that has shown up in a number of civilian publications on nuclear strategy and technology, but which I have yet to see in an official government document.   This is not the same thing as Casaba-Howitzer, which used a radiation case to channel the energy output of a spherical nuclear explosion in a particular direction.   A shaped nuclear charge would modify the design of the nuclear explosive itself, to control how the fusion reaction propagates within the fusion secondary, to ensure that as much as possible of the energy from it is released in the desired direction.[Wi][Fe]   To quote one source:

"Both conventional and thermonuclear shaped charges tailor an explosive burn-wave using a detonation front that releases energy along a prescribed path.   Both can produce jets of molten metal having velocities greatly in excess of the detonation velocity.

"For thermonuclear fuels such as deuterium plus tritium, the burn-wave can be directed by placing hollow bubbles or inert solids in the path of the detonation front in order to alter its velocity.   Of course, ignition of a thermonuclear burn in a warhead requires a fission trigger to achieve the necessary compression and temperature... but even with such a (nondirected) trigger, the overall directivity of a thermonuclear shaped charge can still be significant...   Using a convergent conical thermonuclear burn-wave with a suitable liner, one could theoretically create a jet traveling at 10,000 kilometers per second, or 3 percent of the speed of light."[Fe]
Another source claims that 10% of the speed of light is the theoretical limit for the jet speed of a thermonuclear shaped-charged explosive.[Wi]


This is the most famous of the third-generation weapons, by far, and I'm only going to summarize its very complicated story here since it's recounted well elsewhere.   EXCALIBUR used the output of a nuclear explosion to power an X-ray laser.   It was intended for use as an anti-ballistic missile weapon, though it would also make a fine anti-satellite weapon.   Since each bomb could energize many separate X-ray emitters - about fifty seems to be the usual number, though some accounts go as high as 100,000 - yes, seriously - anyway, suffice it to say that EXCALIBUR seemed like a real game-changing technology for Reagan's Strategic Defense Initiative.

Besides the basic X-ray laser, further improvements were also planned.   The goal for the initial EXCALIBUR weapon was to produce X-ray beams one million times brighter then the X-rays emitted by a non-directional bomb.   Once basic EXCALIBUR was developed, further upgrades would lead to EXCALIBUR Plus, which would increase the brightness by a further thousand-fold.   The final evolution would be Super EXCALIBUR, which would use X-ray lenses to focus the beams, giving a further thousand-fold increase in brightness over EXCALIBUR Plus.

The original EXCALIBUR was going to be based on "pop-up" submarine-launched ballistic missiles, fired into the path of an oncoming swarm of Soviet ICBM's.   However, in order to work, EXCALIBUR would need to fire before the enemy missiles discharged their warheads, since even the mighty EXCALIBUR could not cope with the thousands of decoy balloons the Soviets might include.   And since the X-rays produced by EXCALIBUR could not penetrate very far through air, the Soviets could foil the system by building "fast-burn" boosters that would complete their acceleration and discharge their warheads before they'd left the protective shield of the atmosphere.

EXCALIBUR Plus was supposed to defeat these fast-burn boosters, by being powerful enough to burn its way through the atmosphere.   However, to do this, it would need to be based in orbit, where it could shoot down through a relatively thin layer of air, rather then popped up on a ballistic missile, where it would have to make a slant shot through the upper atmosphere.   This orbital basing would violate the Outer Space Treaty.   Even worse, the devices would have to be placed in a relatively low orbit, meaning that relatively few would be in a position to fire at any given time, and they would be vulnerable to enemy attack.

Super EXCALIBUR solved this problem too.   It would still be based in orbit, but it was powerful enough to destroy ICBM's from geostationary orbit, where all the bomb-sats would be in position to fire at the same time, and where they would have some protection against attack from sheer distance.

At any rate, none of this ended up happening.   The full story is still known only to those inside the nuclear weapons complex, but it appears that sensor errors in early EXCALIBUR development tests led to a drastic over-estimation of how well the prototypes were working.  New tests, using corrected sensors, took a lot of the wind out of the project's sails - not in the sense of "this won't work", but in the sense of "this is going to be a lot harder then we thought."   Plus, Reagan wanted a purely non-nuclear SDI system if he could get it - he wanted to make nuclear weapons "impotent and obsolete", not just pit nukes against nukes.  For a long time, SDIO's leadership insisted that they were only spending money on EXCALIBUR to determine if the Russians could use it against us, we definitely have no intention of building it, no sir.   Then the Cold War ended, and nuclear testing stopped, and that was that.[Br2]

The known EXCALIBUR nuclear tests were:

DIABLO HAWK, September 13, 1978, LLNL.   Test apparatus failed.

DAUPHIN, November 14, 1980, LLNL.   First probable X-ray lasing success.

CABRA, March 26 1983, LLNL.   Sensor failure.

ROMANO, December 16 1983, LLNL.   First strong X-ray lasing evidence.

CORREO, August 2 1984, LANL.   X-ray laser fails.

COTTAGE, March 23 1985, LLNL.   First focusing attempt.

GOLDSTONE, December 28 1985, LLNL.   First use of corrected sensors shows laser is dimmer then previously thought.

LABQUARK, September 30 1986, LLNL.   More focusing tests.

DELAMAR, April 18 1987, LLNL.   More focusing tests.   Not successful.

KERNVILLE, February 15 1988, LLNL.   Data gathering on basic X-ray laser.[Br2]


PROMETHEUS was Los Alamos lab's competitor to EXCALIBUR, and probably descended from Casaba-Howitzer, and possibly classified work on shaped-charge nuclear explosives.   It used a nuclear explosive to propel a number of small kinetic impactors to about 100 km/sec, often referred to as a "nuclear shotgun."   It was primarily intended as an anti-ballistic missile weapon.   A report from the time describes its mechanism as follows:

"One research engineer familiar with the project described the device as operating much like a rifle, using a polystyrene-filled barrel to help couple a plate to the 'gunpowder-like' blast of a directed nuclear charge.   After the impulse from the explosion generates an intense shock wave, the plate 'fractionates' into millions of tiny particles.   Of course, these would vaporize if in direct contact with the bomb, but as configured, the pellets reportedly achieved speeds of 100 kilometers per second without vaporization."[Fe]
This is sourced to a report from SPARTA, Inc., which unfortunately I have been unable to obtain.   Another source reports that PROMETHEUS projectiles could "penetrate fifteen inches of aluminum if you can keep them together."[Ul]   Besides using it directly against warheads, SDI wanted to use it to distinguish decoys from real warheads: PROMETHEUS could be used to hit incoming objects with dust clouds, causing measurable changes in their trajectory, allowing defender's to calculate the mass of the target and therefore whether it's a heavy warhead or a light decoy.   At least one nuclear test was conducted as part of the PROMETHEUS program: CHAMITA, fired on August 17 1985, which reportedly propelled a 1 kg tungsten/molybdenum target to 70 km per second.[Fe]

Optical-Wavelength Lasers

This project was similar to EXCALIBUR, but produced a visual wavelength laser instead of an X-ray.   
This reduced its range substantially, but made it much easier to attack targets in the atmosphere, since the atmosphere doesn't absorb visual light as well as X-rays.   It may have been code-named PERSEID[Fe2] and run at Los Alamos[Ha], but the evidence is inconclusive; PERSEID may have been a separate, conventionally-powered excimer laser project.

Microwave Bombs

It's possible to use nuclear weapons to produce a high-intensity, directional blast of microwaves - an EMP, in other words.   This is not the same thing as the EMP produce by a "conventional" high-altitude nuclear explosion - probably.   Unfortunately, details about this program are still classified, so all we have to go on are unconfirmed leaks and speculation by civilians, but the best guess is that this would work on a very diffirent principle: using a small nuclear explosive to "pump" a microwave generator.[Fe]

Before anything else, let me say that we're getting into an area of physics I don't really understand as well as I should.   Electromagnetism is not my thing.   So I may make some mistakes here, and if I do, please point them out in the comments so I can fix them.

Civilian speculation about how a microwave bomb would work focused on two concepts: the electron plasma oscillator and the magnetic flux compressor.   The electron plasma oscillator would use the X- and gamma-rays generated by a nuclear explosion to create electrical currents.   To quote the source:

"The device could be surrounded by a cylindrical waveguide structure, possibly built up from many concentric metallic cylinders to serve three purposes: they could act as reflex diodes, emitting an intense pulse of electrons by Compton scattering and the photoelectric effect; they could provide a cavity structure in which the fields could 'ring' at a resonant frequency; and they could serve as a microwave horn antenna to direct a beam.   Since this diode-waveguide-antenna structure would remain intact for only a short time before it was blown apart, it would have to exploit the near speed-of-light velocities of the gammas, electrons, and microwaves, to generate a beam quickly."[Fe]

[Fe] estimates that about 0.001% of the weapon yield could be converted into directional microwaves using this process, at a frequency in the tens of gigahertz, leading to the conclusion that such a weapon would only be effective at 300 to 1,000 km range in space, or less then 200 km against targets within the atmosphere.

The magnetic flux compressor works in a similar fashion to a conventional explosive flux compressor, only using a nuclear explosive.   This produces an enormous pulse of electricity, which is fed into a microwave antenna.   [Fe] estimates that 5% or more of the nuclear yield might be transformed into electric current, and that "a few percent" of that energy might in turn be transformed into microwaves before the system blows itself apart, yielding about 0.1% of the nuclear yield in directional microwaves.

Incidentally, besides the usual uses of EMP, there was evidently speculation at Lawrence Livermore that a microwave bomb might perhaps be designed to produce microwaves at a frequency that would interfere with the functioning of the human nervous system - a "brain bomb".   I've no evidence the brain bomb ever went beyond speculation, though.[Br]

We do know the EMP project was primarily based at Sandia lab, though there was probably also involvement by either Los Alamos or Livermore to design the physics package itself - I suspect Livermore, since they were working on an unspecified nuclear weapon to "hold at risk... strategic relocatable targets" at the time[Ty], and that was apparently the primary aim of the microwave bomb project.[Fe]  There was probably at least one nuclear test as part of the microwave bomb program, though this is not certain - we know that at least one concept besides PROMETHEUS and EXCALIBUR had a test, and the microwave bomb is the most likely candidate.[Fe]

Particle Beams

There are periodic mentions that the Department of Energy investigated using nuclear weapons to produce particle beams[Fe], but no other information is available.

The End of the Cold War

These concepts were all still-born, never deployed.   Research was already being scaled back by the end of the Reagan administration, and appears to have been finally terminated after the end of nuclear testing in 1992.   Although nuclear weapons research continues at LANL and LLNL using subcritical tests and computer simulations, developing third-generation weapons would need live-fire nuclear tests, and probably a lot of them.

That said, we'll see how relations between the US, China, Russia, and eventually India shape out over the next few decades.   I would not be overly surprised to see nuclear weapons testing resume at some point - and, if it does, expect to see renewed interest in these concepts.   Making something like Super EXCALIBUR work would be enormously challenging - but if you succeed, it would be an absolute game-changer in nuclear weapons technology.

Works Cited:

[ARPA]: The Advanced Research Projects Agency, 1958-1974.

[Br]: Broad, William J.   Star Warriors: A Penetrating Look into the Lives of the Young Scientists Behind Our Space Age Weaponry.   Simon and Schuster, 1985.

[Br2]: Broad, William J.   Teller's War: The Top-Secret Story Behind the Star Wars Deception.   Simon and Schuster, 1992.

[Dy]: Dyson, George.   Project Orion: The True Story of the Atomic Spaceship.   Henry Holt and Company, 2002.

[Fe]: Fenstermacher, Dan L.   "The Effects of Nuclear Test-Ban Regimes on Third-Generation-Weapons Innovation."   Science & Global Security, Vol. 1 (1990), pp. 187-223.

[Fe2]: Fenstermacher, Dan L.   "Arms Race: The Next Generation."   Bulletin of the Atomic Scientists, March 1991, pp. 29-34.

[Fi]: Fitzgerald, Mary C.   "The Russian Image of Future War."   Comparative Strategy, Vol. 13 No. 2 (April-June 1994), pp. 167-180.

[Ha]: Hartford, Jr., Allen.   Chemical and Laser Sciences Division, Annual Report 1989.   LA-11833-PR.

[St]: Strobel, Warren.   "SDI's 'Nuclear Shotgun' on Pentagon's Fast Track."   Washington Times, April 22 1987, pp. 1, 16A.

[Su]: Subramanian, R. R.   "Third Generation Nuclear Weapons."   Strategic Analysis, Vol. 6 No. 9 (1982), pp. 567-569.

[Ty]: Tyler, James V.   "Nuclear Weapon Development."   Energy & Technology Review, Sept. 1986, pp. 19-25.

[Ul]: Ulsamer, Edgar.   "Strategic Connections in Space."   Air Force Magazine, Vol. 70, August 1987, pp. 78-84.

[Wi]: Winterberg, Friedwardt.   The Physical Principles of Thermonuclear Explosive Devices.  Fusion Energy Foundation, 1981.


  1. I wanted to pick your mind a little on a subject which has intrigued me.
    "will survive except for a direct hit" is a phrase that has always intrigued me.
    More directly related to hardened facilities like Cheyenne Mountain and your excellent article on the DUCC. With 2,000 feet of granite above it - why wouldn't NORAD headquarters survive a "direct hit" Is it the 'spalling' (which they suddenly seemed worried about in the middle of construction? Or am I simply not grasping the enormity of the force applied; such as the difference in striking the earth next to an anthill with a sledgehammer - as opposed to a 'direct hit' on the anthill.

    It all just seems a bit depressing that these nightmarish, doomsday structures all became obsolete cannon fodder after a few years of missile guidance improvement (CEP turning the 'Crystal Palace' into 'Bingo!')

    1. You know, I have asked the same question, and I've never gotten a good answer. My back-of-the-envelope numbers suggest that, if NORAD has a decent facility hardness, it ought to be able to survive at least one hit, though they could dig it out with repeated strikes. But there seems to be general agreement that it would only take one strike to collapse it. My hypothesis has been that it's because of how NORAD was built, as large, open caverns, rather then as a harder tunnel design, so that the roof is more vulnerable to spalling then something like DUCC would be. From the timeline, I suspect they started construction on NORAD before they had a solid understanding of how ground shock works. But I don't really know for sure.

  2. Yes- it seems that in the middle of the project they became very concerned about the 'brittleness' of the rock and started installing all of those long screws into the rock face to 'hold it together'. Contemporary photos show a lot of heavy-duty woven steel (like chain-link fencing) covering the exposed rock surfaces.

    On the same sort of topic; I remember reading somewhere about studies for 'super-hardened' ICBM silos. There was some speculation about the almost comical question of if the surrounding overfill was blown away at what angle of 'lean' would the structure and missile still be functional? (brings to mind a Roadrunner- Wiley Coyote cartoon!).

    1. Rock bolts and nets are a pretty standard hardening mechanism.

      At some point I'm going to do an article on the MX super-hardened basing proposals - I even have a name for it, "Ace in the Hole". It was a fascinating set of ideas. In the most-studied version they planned to bury tunnel-boring machines along with the missiles, because they figured they could bury them deep enough that they'd survive anything the Russians threw at them, but they couldn't do the same for the tunnel exits.

  3. "or less then 200 km against targets within the atmosphere."
    Fenstermacher's paper states that microwave nuke is useless under an ALTITUDE of 200 km, not a range of 200 km. And that's due to atmospheric breakdown in dense air from the high intensity of the source.

    1. Sorry for how long this sat in moderation; math has been keeping me busy. I've only re-skimmed Fenstermacher's paper, so I may be missing something, but it looks to me like the atmospheric breakdown in dense air is an issue for the electron plasma oscillator design, but that the magnetic flux compressor could still be effective?

  4. Didn't the original notion for the Orion project come from an underground test that blew the massive steel lid off the shaft ?? IIRC, some estimates hold it reached 'escape velocity'...

    1. You're thinking of Project Thunderwell! Thunderwell was supposed to be a sort of Jules Verne cannon using a hydrogen bomb instead of gunpowder. Didn't get very far - I'm not sure if it was ever even a formal lab project, or just an idea they kicked around. As far as I can recall, Thunderwell was separate from Orion - I believe I've seen speculation about Orion-style spacecraft dating to before the PASCAL nuclear test which inspired it - but I may be misremembering as it's been a long time since I read George Dyson's book.

      I believe the general consensus is that the lid DID reach escape velocity - but probably VERY briefly, as atmospheric friction likely vaporized it before it could get very far.