Bang!

 

 

The Joy of High Tech

 

by

 

Rodford Edmiston

 

 

Being the occasionally interesting ramblings of a major-league technophile.

 

 

Please note that while I am an engineer (BSCE) and do my research, I am not a professional in this field. Do not take anything here as gospel; check the facts I give. If you find a mistake, please let me know about it.
( Read more... )

This is not intended as a how-to for building nuclear bombs. It is a discussion of some of the more interesting technical aspects of fission and fusion explosive devices. 

 

Producing a fission explosion is not all that difficult. As far as I have been able to find out, every nation which has attempted this succeeded on the first try. The real technical challenges come from extracting the most yield from a given amount of fissile material, and from making very large and very small bombs, in terms of both yield and physical size. There are a number of approaches to these problems. To understand these techniques it is necessary to understand some things about producing fission explosions. 

 

A runaway fission chain reaction is produced by assembling in one place enough fissionable material (the infamous critical mass) under conditions which encourage absorption by the material of neutrons released by the material's decay. To produce an explosion, rather than just a glowing hole in the ground, the fissile mass must reach criticality very, very quickly. (As Douglas Adams put it, in a different context, "Bang the rocks together, guys.") This is not as easy as it sounds. As a device is activated, the energy released produces forces which push it apart, or "disassemble" it. This response is very, very quick. (Material at the heart of the reaction may explosively vaporize while the outer layer is just getting warm.) Assemble your critical mass too slowly, and it blows apart before releasing more than a tiny fraction of its potential nuclear yield. This is known as a squib or fizzle, and while there's no precise definition of how inefficient a bomb must be to qualify, when one happens it is obvious.

 

The two basic methods of assembling a critical mass quickly are the gun and the implosion. The gun is quite simple, and basically involves firing a slug of fissionable material out of a short cannon into the middle of a barely subcritical mass. This configuration is inherently inefficient. The speed of the slug is far slower than the speed of the neutrons given off, and during the approach there's time for several generations of chain reaction. Properly designed shielding can greatly reduce the early neutron exchange. However, even with proper shielding, the target mass - and, to a lesser extent, the slug - will start reacting before the masses physically combine. The device may even disassemble before the slug reaches the target, producing a squib. Because of its properties, plutonium is far more likely than uranium to experience this predetonation; this is especially true for Pu240, a contaminant in reactor-made Pu239. Gun devices are therefore usually uranium devices.

 

The implosion method uses shaped charges usually placed around the outside of a pusher shell, which is most commonly made of aluminum. That, in turn, pushes on the core or, more often, the tamper. This is, in turn, a thick shell of dense material, usually made from unenriched uranium (in which the proportion of U235 to U238 is roughly the same as is found in natural ore) or depleted uranium (nearly pure U238, a byproduct of the enrichment process) placed around the core. The explosives are triggered in a carefully timed sequence, the end goal being to rapidly force a mass of fissionable material inward. As noted, this is usually done indirectly, using a pusher shell and a tamper. The actual mass of fissionable material in a core can be less than what is normally considered a critical mass. Criticality depends not just on the total mass, but the concentration of mass. Which can cause problems. 

 

Heavy metals tend to have peculiar physical properties, and plutonium is no exception. As it cools from molten to room temperature, it goes through *5* phase changes. During two of those 6 phase states it is denser than the final state. Something which could have made life interesting for the folks casting the barely subcritical core for the first bomb, if they hadn't know about it ahead of time.

 

The gun method, as can be imagined, is pretty limited. However, it is also very robust. Most artillery-delivered nuclear warheads are gun designs, as are most bunker-penetrating warheads. That is partly due to this robustness, and partly due to the linear design fitting better in artillery shells. The basic implosion method is inherently more efficient, but there's plenty of room for improvement of both types from their basic configurations. 

 

One of the most important innovations for getting more out of an implosion-triggered fission explosion is what's known as the levitated core. Anyone with experience firing a high-powered rifle can understand how and why it works. The actual core is supported on collapsible studs, so that the pusher/tamper combination slams into the core at high velocity before contact, instead of merely pushing on it. (Note: there's usually a neutron absorbing layer, probably boron, between the explosives and the tamper/pusher.) The explosive-driven shell thus builds up quite a bit of speed and inertia before actually impacting the core. Additionally, by making the core hollow the fissionable material of the core gets a good run and go before slamming together in the center, also helping make the rise to criticality quicker. Because the fissionable material of a hollow core is distributed through a larger volume (due to the hole in the middle) more can be placed in this shell than the normal critical mass.

 

A composite core - a hollow core which in turn has an inner core which is something besides plutonium or uranium - is also an important development. Most composite cores contain an initiator, a neutron source, to give the first generation or so of the chain reaction an extra kick. In early bombs the neutron source was a combination of polonium (specifically, Po210, a plentiful source of alpha particles) and beryllium (which easily has neutrons knocked from its nuclei by alpha particles) in concentric shells. Modern bombs may use a different design for the initiator, such as an electrically operated device for the neutron source.

 

The design of the initiator itself is quite complex. A nickel barrier separates the beryllium and polonium shells until the explosive shockwave collapses the core. The beryllium layer has cone-shaped holes or circumferential angled grooves (pointing outward) which, as the shell collapses, form jets of pressure-liquefied beryllium. These penetrate the nickel to mix with the polonium. (Yes, that's the Munroe effect, just like in an anti-armor shaped charge.) Further complicating the set-up, the initiator itself can be levitated, and may have multiple concentric layers repeating the beryllium-nickel-polonium sequence. 

 

Combining these techniques - a levitated, hollow core with a levitated, hollow initiator inside - greatly increases both yield and efficiency.

 

Another trick is to put more mass around the bomb, to hold it together through inertia for a bit longer while the chain reaction goes through a few more generations. As noted, most nuclear devices have a tamper shell inside or as a part of the pusher, providing concentrated mass for that purpose. Uranium (depleted or unenriched) is normally used, since it is very dense, adding mass with little volume. It is also readily available, since far more depleted uranium (pure or nearly so U238) than enriched (with a higher percentage U235) is produced by the enrichment process. Additionally, the flood of neutrons coming out of the bomb will cause some of the uranium in the tamper to fission, adding more energy. Outside the tamper, the pusher shell and the explosives is a thick aluminum shell, which adds yet more confining mass, as well as actually packaging the bomb inside the casing.

 

The tamper doesn't have to be uranium. The Soviets tested one large fission device with lead replacing the uranium in the tamper. This greatly reduced the yield, of course, but also produced one of the radiologically cleanest fission explosions on record. I've encountered vague references to a low-yield bomb with a gold tamper, to use as a bunker buster where the situation allows accurate placement and minimized fallout is desired, but have seen no hard data on this. 

 

The process of making a bomb go past critical is often described as assembly, because some designs (such as the gun type) actually use explosives to bring two or more subcritical masses together (that is, assemble the critical mass) very quickly. (In The Long Watch Heinlein described two perfectly machined hemispheres of fissionable material slammed together by explosives. I think he was being deliberately wrong. Those flat faces would have trapped gasses between them, even in a near-vacuum, cushioning the assembly of the critical mass and greatly reducing yield.) As described above, explosive compression of a solid or hollow single fissionable mass can also be used. 

 

The secret to a small bomb is that it must also be efficient. This requires squeezing a subcritical mass of fissionable material hard enough fast enough to exceed the critical limits without the device blowing apart before a sizable amount of the material fissions. The lower limit for fission explosions seems to be around 0.01 kilotonne. (At which point one has to wonder "Why bother?") This is a practical limit, set by how much effort the designer wants to put into getting a normally subcritical mass to go supercritical. Interestingly, a typical squib event of a more common yield warhead produces around ten times that much fission energy. Except that the squibs are extremely inefficient and very dirty. The inherent efficiency of small bombs also makes them somewhat radiologically cleaner than less efficient devices.

 

The infamous atomic hand grenade is a myth. The smallest warhead issued by the United States was probably the W-54 Davy Crockett, usually delivered by a recoilless rifle launcher. The warhead was 27.7 cm in diameter, 39.9 cm in length and massed 22.7 kilograms. Its yield was around 20 tonnes (not kilotonnes) TNT equivalent. You just can't make a deployable warhead much smaller or lighter and still have it work. At least, with current methods.

 

From the information I have been able to find, the lowest yield of any device tested was the Buster Able shot. This was known to have a core right at the lower limit of what could be squeezed to criticality, and no-one was surprised when it fizzled. The design had the irreverent name of The Petite Plutonium Fission Bomb, and was designed by Ed Taylor, he of Orion fame. The predicted yield was 0.2 KT. Once the yield from the chemical explosives was subtracted, the fission energy was estimated at the equivalent of under one pound of TNT. This is also the first recorded failure of any nuclear device, the first actual squib known. The core went supercritical and produced a detectable amount of neutrons. Just not many of them...

 

Because a fission bomb depends on rapidly assembling a critical mass of fissionable material there are practical upper limits on how large the yield can be. Beyond a certain point, increases in yield mostly result from improving the efficiency of a bomb, while adding more fissionable material adds more problems. Keeping a large mass of uranium or plutonium subcritical until assembly, then accelerating it to a high-enough velocity quickly enough is a major engineering challenge. An answer to that challenge was developed by Stanislaw Ulam, who came up with the concept of the staged device. Using this, some of the energy released by a small, efficient fission device (the primary) is used to trigger a much larger device (the secondary). The energy generated by the primary is far greater than what any reasonable amount of chemical explosives can provide, giving the designer more to work with.

 

The most efficient configuration of a staged device has the secondary located immediately adjacent to the primary, not around it. Ulam's original idea was to use the explosive shockwave of the primary to provide the compression for the secondary. However, he, Edward Teller and others quickly realized that the electromagnetic radiation from the primary (mostly X-rays) arrived much faster than the physical shockwave. As staged devices were actually developed by the US (the Soviets' first staged devices were concentric designs and therefore less efficient) the radiation vaporizes and ionizes material on the outside of the secondary. More specifically, the X-rays convert to plasma a low molecular weight material - polyethylene or polystyrene - between the secondary and its casing. This radiation causes the vaporized material to re-radiate X-rays. These secondary X-rays (which are softer - that is, of longer wavelength - than the primary X-rays) then vaporize the outside of the secondary. This object is not a sphere but a truncated cone or a cylinder with a truncated cone upper end. This second vaporization produces a reaction effect, forcing the outer layer (which may be a uranium tamper) inward. Even though only a small portion of the energy from the primary actually heads in the direction of the secondary (and this is with design features which help concentrate the energy in that direction), that fraction is vastly greater than the energy of the explosives used to initiate a single-stage device. The result is far greater and quicker compression. 

 

For a staged fission device, with a levitated, composite core, shaped charge implosion for the primary, uranium tampers, and several other tricks, the maximum practical yield is around 200 KT. Fission devices with much larger yields have been built, but these were test articles or "table thumpers" (propaganda demonstrations) and not actual weapons. (For instance, you could put five 200 KT secondaries around a primary and exceed a megatonne, but the device would be too bulky and heavy to deliver by plane or missile.) Another trick is to replace or augment the standard tamper with a shell of U235 or P239. That, however, is very wasteful of bomb-grade material, since the increase in yield over an unenriched tamper is only about double. There could also be criticality problems. 

 

The largest purely fission device detonated by the United States (and possibly by anybody) was Ivy King, with a yield of 500 kilotonnes. This was air-dropped from a B-36, and was largely built up from stockpiled components. However, the core was modified, and contained 60 kilograms of highly enriched uranium surrounded by a natural uranium tamper. To quote one article "While perhaps not the largest deliverable fission bomb possible at this time, it was certainly pushing close to the practical limit." The core contained more that 4 critical masses of enriched uranium, and great pains were taken to safe the device. Replacing the natural uranium of the tamper with bomb-grade uranium or plutonium might have doubled the yield, but was probably not feasible due to the criticality problems this would have added. 

 

One way of getting a little more "Oomph" from fission warheads is known as the boosted bomb. Placing a few grams of a tritium-deuterium mix in a capsule in the core of a fission bomb increases the yield by about 10-20%. The heat and pressure of the nuclear explosion, combined with the high neutron flux, causes a significant amount of fusion between the heavy hydrogen isotope atoms. However, adding too much of this reduces yield because of several factors, including reducing the ability of the fissionable mass to compress radially. The hydrogen isotopes must be in a separate capsule, in spite of the fact that heavy metal hydrides can pack in hydrogen to a density greater than that of liquid hydrogen. (A technical side note, here: the physical process whereby a heavy metal absorbs hydrogen releases heat, which is at least part of what was going on in that "cold fusion" mess, where iridium rods were used.) 

 

The reason for not hydrating the tamper or core of a bomb (that is, of keeping any hydrogen physically isolated from those materials) is that the light atoms - taking up positions within the matrix of heavy atoms - capture or slow too many neutrons, possibly to the point of causing a squib explosion. Several hydrated devices were tested in the Fifties. Most were squibs, and the rest all had greatly reduced yields.

 

Another way, mentioned above, to get light materials to fuse is to place them around the core, in what's known as a layer cake design. This works, but a much smaller proportion of the light atoms fuse; the pressure is less and the confinement time shorter. On the other hand, the mass of fusible material can be much greater than with a boosted bomb, and with work a layer cake can produce a yield of around a megatonne. 

 

The best method we know of for producing a large fusion explosion is an application of Ulam's staged concept. The first true H-bomb (Ivy Mike) used a fission primary to initiate fusion in a container of liquid deuterium. (This container had a shape strongly resembling that of a centerfire pistol cartridge loaded with a jacketed flatpoint bullet.) For added efficiency, a rod of fissionable material - basically a stretched-out A-bomb core - ran along the axis of the Dewar containing the deuterium. When the container imploded, the secondary core - or sparkplug - detonated from the compression and neutron flux, catching the deuterium between two sources of heat and pressure, with neutrons also added from the sparkplug. Fusion then occurred. Actually, more than 75% of Mike's 10.4 megatonnes came from fusion neutrons acting on the secondary's U238 pusher. This arrangement is known as a fission-fusion-fission device, and some variation is used in nearly all current fusion warheads, though the liquid deuterium has been replaced with lithium deuteride or something similar. (Lithium absorbs neutrons to form tritium, which fuses with deuterium much more easily than deuterium can fuse with itself.)

 

As with Mike, most large-yield hydrogen bombs get the biggest part of their yield from the induced fission of the secondary's tamper. Interestingly, the primary alteration to a standard hydrogen bomb to produce a neutron bomb is to replace the uranium in the secondary's tamper with something dense but not fissionable, letting neutrons which would have fissioned the U238 fly free. The test detonation with the largest portion of total yield coming from fusion was probably the Redwing Navajo test shot. 95% of the 4.5 megatonne yield came from fusion, making it remarkably clean.

 

Designing a device which could trigger a fusion explosion without a fission primary is very difficult. Getting the necessary amount and concentration of heat and compression and neutrons by another means would require a quite different system, and probably either something like a sizable homopolar generator or vast banks of capacitors or superconducting coils to supply a short, powerful pulse of current. I don't think anyone has yet succeeded at this.

 

There have also been three-stage devices. These use a small fission device to trigger a larger fission device which triggers either a really big fission or fusion device. There are some advantages to this arrangement, since chemical initiation is easier and simpler with a smaller fission device. However, the high yield of modern small primaries and the efficient way a modern two-stage bomb uses the energy released by the primary, contrasted with the added overall size and complexity due to the extra stage, makes the three-stage arrangement an unnecessary elaboration. 

 

Despite the fears of the Fifties and Sixties, only two nuclear weapons have ever been used against people. Let's hope things stay that way.