What a long set of weeks it has been! On top of my usual teaching load (a few hours of lecture per week, grading, etc.), I have given two public talks and then flown to Chicago and back for the annual History of Science Society meeting. So I’ve gotten behind on the blog posting, though I have more content than usual for the next few weeks built up in my drafts folder, without time for me to finish it up. During this busy time, by complete coincidence, I also got briefly interviewed for both The Atlantic (on plutonium and nuclear waste) and The New York Times (on the apparent virality of nuclear weapons history).
Louis Slotin and Herb Lehr at the assembly of the Trinity “Gadget.” Source: Los Alamos National Laboratory Archives, photo TR-229.
The Times article had a phrase in it that has generated a few e-mails to me from a confused reader, so I thought it was worth clarifying on here, because it is actually an interesting detail. It is one of those funny phrases that if you knew nothing about the bomb you’d never notice it, and if you knew a good deal about the bomb you’d think it was wrong, but if you know a whole lot more than most people care to know unless they are serious bomb nerds you actually see that it is correct.
Here’s the quote:
First, he glanced at the scientists assembling what they called “the gadget,” a spherical test device five feet in diameter. Then, atop a wooden crate nearby, he noticed a small, blocky object, nondescript except for the role he suddenly realized it played: It was a uranium slug that held the bomb’s fuel. In July 1945, its detonation lit up the New Mexican desert and sent out shock waves that begot a new era.
I’ve added emphasis to the part that may seem confusing. The Trinity “Gadget” and the Fat Man bomb, as everyone knows, were fueled by fission reactions in a sphere of plutonium. The Little Boy bomb dropped on Hiroshima, by contrast, was fueled by enriched uranium. So what’s this reference to a uranium slug inside the Trinity Gadget? Isn’t that wrong?
Detail from the above photo showing the tamper plug cylinder. Inset is a rare glimpse of what the tamper probably looked like, taken from a different Los Alamos photo related to Slotin’s criticality accident. (It is in the middle-right of the linked photo. Yes, I cop to spending time searching the edges of photos like this…) You can see how the tamper plug, rotated, would be inserted into the middle of the tamper sphere.
Perhaps surprisingly — no, it’s not. There was uranium inside both the “Gadget” and Fat Man devices — in the tamper. The tamper was a sphere of uranium that encased the plutonium pit, which itself encased a polonium-beryllium neutron source, Russian-doll style. Here uranium was chosen primarily for its physical rather than its nuclear properties: it was natural, unenriched uranium (“Tuballoy,” in the security jargon of the time), and its purpose was to hold together the core while the core did its best to try and explode. (It also helped reflect neutrons back into the core, which also worked to improve the efficiency.)
The inside of an exploding fission bomb can be considered as a race between two different processes. One is the fission reaction itself, which, as it progresses, rapidly heats the core. This heating of the core, however, causes the core to rapidly expand — the core is trying to blow itself apart. If the core expands beyond a certain radius, the fission chain reaction stops, because the fission neutrons won’t find further plutonium nuclei to react with. If you are a bomb designer, and want your bomb to have a pretty big boom, you want to hold the bomb core together as long as possible, because every 10 nanoseconds or so you can hold it together equals another generation of fission reactions, and each generation releases exponentially more energy than the previous.1
An image that somewhat evokes how bomb designers talk about the dueling conditions inside of the bomb, when they are talking to each other. The “snowplow region” is where the expanding bomb core runs into the tamper and is compressing it from the inside. This is a level of bomb design that I would have normally assumed would be classified but it has been very clearly declassified here, so I guess not. From Glasstone, “Weapons Activities of Los Alamos, Part I” (see footnotes).
So in the Fat Man and Trinity bombs, this is accomplished with a heavy sphere of natural uranium metal. Uranium is heavy and dense, and the process of making plutonium and enriched uranium required the United States to stockpile thousands of tons of it, so the relatively small amount needed for a tamper was easily at-hand. It makes a good substance with which to try and hold an exploding atomic bomb together. The Little Boy bomb, as an aside, used a tungsten tamper, for some reason (maybe to avoid excessive background neutrons, I don’t know).
Now to add one more little bit of detail: we tend to think of the Trinity/Fat Man implosion bombs as just being a set of spheres-inside-spheres. This is a convenient simplification of the actual geometry, which had other factors that influenced it. The tamper, for example, was not just two halves of a hollow sphere that could fit together. Rather, it was more like a solid sphere out of which a central cylinder had been removed. The cylinder was known as the “tamper plug,” and was itself made of two halves that, when assembled, had room for the plutonium pit inside of them.
Why do it this way? Because the scientists and engineers wanted to be able to insert the fissile pit portion into the bomb as one of the final additions. This makes good sense from a safety point of view — they wanted it to be relatively easy to add the final, “nuclear” component of the bomb and to keep it separate from the non-nuclear components (like the high explosives) as long as possible. I don’t want to over-emphasize the “ease” of this operation, because it was not a quick, last-minute action to put the pit inside the bomb. (Some later bomb designs which featured in-flight core insertion were designed to be just this, but this was some years away.) It was still a tetchy, careful operation. But they could assemble the entire rest of the tamper, pusher, and high explosives, then remove one layer of high explosives, remove the top of the pusher, and then lower the tamper plug (with pit) into the center, then replace all of the other parts, hook up the detonators and electrical system, and so on.
A rendering I made in Blender to illustrate the principle here. The pit and initiator are inside of the plug (expanded at right), which is then sealed into a cylinder and inserted into the tamper sphere at the center of the bomb. The tamper is itself embedded in a boron shell which is inside of an aluminum shell which is inside of the explosive lenses which is inside of the casing. This is part of a modeling/visualizing project I’ve been working on for a little while now and will post more on at a future date. The dimensions are roughly correct though there are still many simplified detail (e.g. exactly how the plug fits together — there were uranium screws!).
So when John Coster-Mullen describes, as in the previously-quoted New York Times article, finding a picture of the tamper plug, it’s kind of a cool thing. There’s only one picture that shows it (the one at the beginning of this post), and it is one of those things that you don’t even usually notice about that picture until someone points it out to you. I never noticed it until John pointed it out for me, even though I’d seen the picture many times before. Usually one’s attention is drawn to the Gadget sphere itself, and the people standing around (including Louis Slotin, who would later be killed by playing with a core). It’s kind of surprising it was declassified, since the length of the tamper plug is the diameter of the tamper, and the width of the plug is just a little bigger than the diameter of the plutonium core. The US government usually doesn’t like to reveal, even inadvertently, those kinds of numbers.
There is also one little fact about the natural uranium in the Gadget and Fat Man bomb that is not well appreciated, and I didn’t appreciate well until reading John’s book. (Which I have heard people say is rather expensive for a self-published production, but if you’re a serious Manhattan Project geek it is hard to imagine how you’d get by without a copy of it — it is dense with technical details and anecdotes. It is one of the only books that I don’t often bother to put back in the bookcase because I end up needing to reference it every week or so.)
Neutron cross-sections for the fissioning of uranium and plutonium. The higher the cross-section, the more likely that fission will occur. The indicated “fission neutron energy” means that that is the approximate energy level of neutrons released from fission reactions. So you can see why, in a reactor, those are slowed down by the moderator to increase the likelihood of fissioning. In a bomb, there is no time for slowing things down, so you need fissile material in much higher concentrations. Source: World Nuclear Association.
In talking about which elements are fissile — that is, can sustain a nuclear fission chain reaction — technical people tend to talk about neutron cross sections. This just means, in essence, that the likelihood of a giving elemental isotope (e.g. uranium-235, plutonium-239) undergoing fission when encountering a neutron is related to the energy of that neutron. At the size of neutrons, energy, speed, and temperature all considered to be the same thing. If you look at a neutron cross section chart, like the one above, you will see that uranium-235 has a high likelihood of fissioning from slow neutrons, and a low-but-not-zero likelihood of fissioning from faster neutrons. You will also see that the neutrons released by fission reactions are pretty fast. This is why to sustain a chain reaction in uranium you either need to slow the neutrons down (like in a nuclear reactor, which uses a moderator to do this), or pack in so many U-235 atoms that even the low probability of fissioning from fast neutrons doesn’t mean that a chain reaction won’t happen (like in a nuclear bomb, where you enrich the uranium to be mostly U-235).
Still with me? If you look a little further on the graph, you’ll see that uranium-238 also has a possibility of fissioning, but it is a pretty low one and only even becomes possible with pretty fast neutrons. This is why, in a nutshell, that unenriched uranium can’t power an atomic bomb by itself: it is fissionable but not fissile, because it can’t reliably take fission neutrons and turn them into further fission reactions. But people who have studied how thermonuclear weapons are used know that even uranium-238 can contribute a lot of explosive energy, if it is in the presence of a lot of high-energy neutrons. In a multistage hydrogen bomb, at least 50% of the final explosive energy is derived from the fissioning of U-238, which is made possible by the high-energy neutrons produced from the nuclear fusion stage of the bomb (which itself is set off by an initial fission stage). The neutrons produced by deuterium-tritium fusion are around 14 times more energetic than fission neutrons, so that lets them fission U-238 easily. From the cross-section chart above, you can see that U-238 fissioning can happen from fission neutrons, but only if they happen to be pretty high energy to begin with and stay that way. In practice, neutrons lose energy rather quickly. Still, according to a rather sophisticated analysis of the glassified remains of the Trinity test (“Trinitite”) done a few years back by the scientistsThomas M. Semkow, Pravin P. Parekh, and Douglas K. Haines, a significant portion of the final fissioning output at Trinity (and presumably also Nagasaki) came from the fast fissioning of the tamper, with some of that energy released from the U-238 fissioning.2
For the hardcore bomb geeks, here is a sort of “conclusion table” from the Semkow et al. article. Note that they calculate at least 30% fissioning from uranium, and give some indication the amount of compression of the core, the number of neutrons created, and so on. Their terminology of the “eyeball” is taken from Richard Rhodes, who uses the term in passing in The Making of the Atomic Bomb, and refers to the confined area where the fission chain reaction is taking place.
How significant? Semkow et al. calculate that about 30% of the total yield of the Trinity test came from fissioning of the uranium tamper, which translates to about 6 kilotons of energy. If they had made the tamper out of tungsten (as was the Little Boy tamper), then the total yield of the Gadget would have only been around 14-15 kilotons — not that different from Little Boy (which was ~13-15 kt). And presumably if the Little Boy bomb had used a uranium tamper, assuming that didn’t cause problems with the design (which it probably would have, otherwise they probably would have used one), it would have had the same yield. (This doesn’t mean that Little Boy wasn’t, in fact, horribly inefficient — it got about the same yield but it required 10X the fissile the material to do so!) The total mass of the tamper was around 120 kg of natural uranium, so if it contributed 6 kilotons of yield that means around 350 grams of the tamper underwent fission, and that is about 0.3% of the total mass.3
So the fact that Trinity and Fat Man had uranium inside of them is already kind of interesting, but the fact that a large portion of the blast derived from that uranium is sort of a neat detail. Why don’t we generally learn about this? It isn’t that it is so terribly classified, per se, but it does require a lot of detailed explanation, as evidenced by the length of this post. We tend to abstract the mechanics of the bombs for explaining their conceptual role, and explaining the basic concepts of how they work. I have no problem with this, personally, because hey, let’s be honest, the exact amount of energy derived from different types of fissioning in the bombs is a pretty wonky thing to care about! But every once in awhile you need to understand the wonky things if you want to talk about, say, what that funny little “plug” is in the top-most photograph, and its role in the bomb. I suppose one of the points of the phenomena described by the Times article, where the geek population on the Internet is providing a newfound audience to Manhattan Project details, is that these sorts of wonky aspects are no longer limited to people like John Coster-Mullen, Carey Sublette, or myself. There are some people who might see this focusing on the technical details as missing the broader picture. I don’t happen to think that myself — much of the broader picture is in fact embedded in the technical details, and “new” discussions of technical details are one way of shaking people out of the calcified narratives of the Manhattan Project, something which, as we approach the 70th anniversary of Hiroshima and Nagasaki, seems to me a valuable endeavor.
- Calculating the efficiency of the bomb as a function of how well you can hold it together is apparently the essence of the still mostly-classified Bethe-Feynman formula. It is described qualitatively in Samuel Glasstone, “Weapons Activities of Los Alamos Scientific Laboratory, Part I,” LA-1632 (January 1954), 34-37. My copy of this report comes from the NNSA’s FOIA Reading Room. I downloaded the file in 2009, and sometime since then all of their PDFs have gotten corrupted somehow, and so many of the pages of the PDFs now available on their site are unreadable. For those who are curious, at a technical level, the corruption involved a systematic stripping out of the carriage return (0D) ASCII characters from the PDFs — there are none in any of the files, and there should be several thousand of them. Here is a screenshot from a hex editor showing the corrupted file (on left) versus the uncorrupted one (on the right). There seems to be no easy fix for this problem. I have tried to contact the NNSA about this but have gotten no response. It is one of many troubling incidents revealing, in my view, the very low priority that public release of information, and poor understanding of public-facing information technology, with regards to the present nuclear agencies. [↩]
- Thomas M. Semkow, Pravin P. Parekh, and Douglas K. Haines, “Modeling the Effects of the Trinity Test,” Applied Modeling and Computations in Nuclear Science, ACS Symposium Series (American Chemical Society: Washington, DC, 2006), 142-159. The authors do not estimate the amount of tamper energy to have been released from U-238 fissioning as opposed to U-235 fissioning. [↩]
- A 120 kg tamper of natural uranium ought to contain around 840 grams of U-235 in it, as an aside, which if that all fissioned at once would release around 14 kilotons of energy. The rule of thumb for uranium is that every kilogram which fissions releases about 17 kilotons. [↩]
I’m surprised that the cylinder / slug isn’t instead a set of top & bottom pieces, threaded together around an approx centered radius, so that the seam is centered in the fattest part of the outer cylinder.
I don’t know what kind of reasoning went into that aspect of the geometry. Assuming John’s geometry is correct, perhaps they didn’t want the fastening pieces (which were made out of uranium as well) to be too near the core itself (they were at the extreme ends of the plug), and didn’t trust it to be simply threaded together? But I don’t know.
John’s description of the capsule is quite detailed, and taken largely from interviews. Phillip Morrison describes the capsule as having “funny ends,” and one of those ends could be threaded with an eye-bolt to allow lowering the capsule into the bomb assembly. Apparently the eye-bolt is one of the few parts of the Trinity bomb to still exist. Several hours of work went into making sure it was completely sealed (using gold foil as a gasket) prior to use.
I’d like to make a suggestion.
This is based on a lot of reading of the topic, but I’ve never found a direct quote that states this.
They sliced it like a hot dog bun, instead of in the middle, because they had a great deal of unknowns with reference to jetting.
When a shock wave transits from one medium to another, instabilities occur. As the wave would have gone from one part of the FAT MAN system into the pill shaped cylinder, there would have been an opportunity for the wave to be deformed.
Couple a deformed wave with a slit, even if it was an interrupted slit like would be found if the two halves screwed together, and the potential for a jet forming would have been possibly possible.
This would have ruined the implosion wave, and fizzled the device.
By ensuring the parting lines on the plug, and the ones on the pit didn’t intersect, there would, I think, been less of a chance of failure.
Plus, there are records of issues with trying to thread uranium, and of it being very sensitive to environmental change. Laying two submarine sandwich loaves of Unat and screwing them together would probably be less of an issue than making two halves mate.
Maybe.
Just guessing…
Shawn Hughes
What is the tape on the Gadget covering?
It is keeping sand, dirt, etc. out of the holes that allow the detonators and diagnostic wires to penetrate the casing and get to the explosive lenses on the inside.
This is one of the best posts I’ve read here. I’m not a wonk myself, but it’s just those sort of details that I find most interesting. For example, I had no idea that 30% (!) of the energy from Trinity came from plain old U-238. Amazing.
Dear Alex:
1. as usual, I really enjoyed this post!
2. yes, little boy didn’t used uranium tamper because if it would have, the neutron bckground wouldv’e been too high (high risk of predetonation).
3. tungsten carbide however is almost as good tamper as uranium, because like uranium it is both a good tamper and reflector (dense and capab;e of reflecting neutrons).
4. this “uranium fissionable tamper” is important in two-staged hydrogen bombs, as you have mentioned, but the key aspect of the one-staged hydrogen bomb (sloika/alarm clock model, like Joe-4)- in this type of bombs, the use of this fissionable tamper is the key for the increased yield of those bombs (~400 kt).
Alex, this is really a great article! I had never seen a schematic like the one you created in Blender and included in this post. I did not know that the pit and initiator are sealed into a cylinder. What is the cylinder made of? I did not know that this cylinder is then inserted into the tamper. All schematics I have seen show a spherical plutonium pit surrounded by the tamper. I also did not know about the boron and aluminum shells.
I agree with Ben Johnson that finding out that the U238 contributed 30% of the yield for the plutonium devices and that it would contribute 50% of the energy in a multistage thermonuclear weapon. 50% is astounding.
This article was really great fun! Thanks.
Thanks! As for your questions: The cylinder is just natural uranium (tuballoy). Every part of the tamper is natural uranium — even the screws and hinges that held it together. They wanted it to be a perfectly homogenous unit. The only non-uranium aspects were gold foil that was used to seal all of the edges. I was really surprised by the 30% number also — a very significant fraction.
I hesitate to offer unrequested technical advice (because so frequently it’s something already considered), but the problem you mention in your first footnote looks entirely like DOS-style line endings being converted to Unix-style endings. I’ve seen this with misconfigured FTP servers (and clients, and any other number of tools) that try to transfer binary files in ASCII mode (which tries to ‘help’ you by converting the line endings). If this is the case, then downloading the files on a different system might work, or a tool to convert line endings might fix them (unix2dos is the first one Google suggested to me).
Hi Stephen: The problems seem to be in the files/servers themselves. I have tried downloading them on a multitude of different systems with different browsers, etc. No dice. I have tried manually converting the line endings but nothing I could come up with could distinguish legit conversions and non-legit ones. Anyway I will see about trying to force a binary download, but I suspect something is wrong on the NNSA’s end of things. (Theirs is the only website that I have seen this happen on.)
I saved some of the NNSA files before they were corrupted and printed them out. They are in storage along with most of my other files relating to missile and nuclear research conducted around the world. You can view my older nuclear bibliographies on the FAS website and if anyone is interested I can provide newer information.
Hi, I’m just random lurker. Since my background is IT I think I can try to help but:
can you post a sample link to a document that doesn’t open correctly? I tried several random links there and all seems to work just fine.
Here’s the version on the NNSA’s website of the same file. When I open it, on any computer, many of the pages are unreadable (which renders differently depending on the PDF viewer). When I compare it to the version of the file I downloaded in 2009, the problem seems to be that the current files have had certain characters stripped from them.
From a quick compare-test I’ve run on those two files (which are, in short, sets of BMP images in grayscale) I can say that the one from NNSA is damaged. I cannot say, if those changes you discovered were done in purpose or by accident (i.e. some hardware failure or OS problem). As I said – those are just BMP format scans wrapped up in PDF format. Modifying random single value would come up as damaged single image in whole set (here: the pdf file) and in effect some pages would be blank or filled with garbage.
The next step would be to compare modification dates in all pdfs server-side (of course I am not able to do it, I do not have maintenance access to those servers) – we could see then at least when and which files were accessed for writing.
I will try to dig up more from those pdfs later, maybe there is some way to restore some data? I do not know but will try to find out. If it helps…
I have looked into this pretty extensively and I don’t think there’s any way on the receiving end to fix these. Something is wrong in either how the files are stored or transmitted that results in their their 0D hex character codes being stripped out. This produces bad checksums (it throws off the byte counts), random errors, etc. It didn’t use to always be this way with their files but for the last few years it has been. I’ve tried getting in touch with NNSA on the issue but never could find a way to contact them that got a response. I do not suspect it is malicious — I suspect “mere” incompetence.
[…] limits of Hansen’s knowledge about Fat Man in 1988. An obvious one is that it is missing the aluminum pusher which sits in between the tamper and the high explosives. There are other issues relating to the […]
I found it fascinating to see those pictures of Slotin’s lab. Perhaps the thing that surprised me the most is just how crude it was. It looks more like a place someone might fix a car in than build a nuclear weapon. I almost wonder if the lack of space to move around the assembly table may have helped lead to the accident.
The screwdriver was a pretty rough tool to move such a thing. Best to fit the tool to the job and the was iffy here at best. As I remember (an iffy subject) it may have simply been that Slotin slipped, causing the accident.
On the other hand, Slotin was a bit reckless around radiation. While working on a project submerged in the water of a reactor at Oak Ridge, Slotin found he had a problem with the experiment. He wanted to get to it to fix it, requiring a shut down, but the plant was scheduled to run for some time. Rather than wait the week or so that required, Slotin slipped in for a midnight dip into the pool, where he went into the tank, corrected the issue and emerged.
So either possibly related — What was he thinking then?!! — or totally irrelevant, his experience was a reminder this stuff could hurt you — and not off in the theoretical future.
In the close up picture I can’t see anything that looks like a seam in the cylinder. Was the entire cylinder wrapped in the gold foil? It certainly doesn’t look gold in the pic. It is possible, of course, that the seam is simply in the lower, shadowed area of the cylinder.
Very interesting text. In my opinion, this is an engineering solution to a technical problem: how to insert a plutonium sphere in a nuclear device such as Trinity or the Nagasaki bomb without dismantling the whole thing. The British used the same approach on their Hurricane device. What is new information is that the tamper contributed to the overall power of the nuclear explosion. In a thermonuclear device, this is to be expected because of the extremely high level of radiation from the secondary. In a nuclear device, that is quite surprising. Thank you for helping people understand more about nuclear devices.
Hello
Might I ask a question concerning tungsten tamper in plutonium bomb?
You suggest that it is possible, only the the yield will be smaller without extra fissioning.
But nuclearweaponsarchive suggest possible problem: Rayleigh-Taylor instability , which may arise if tamper is less dense than fissile core. Does it prevent making plutonium bomb with tungsten tamper? If no, what is the penalty for using tungsten?
For what it’s worth, it probably wouldn’t be too hard to reconstruct those PDFs with an appropriate dictionary. You would look for two words that were stuck together with no whitespace characters in-between. I suspect you could get a coder in India, Russia (oh, the irony) or some other places to code it for $50-100.
And yes, I find it very sad how bad “we” are at protecting priceless data. As I understand it there’s a fair amount of irreplaceable science data that NASA has lost because the tapes it was stored on have degraded (not to mention the difficulty of finding equipment to read them). I can only hope that more recent data won’t suffer the same fate since online and near-line storage are much more common than offline storage today.
[…] The classic example of this, of course, is the implosion bomb design. The bare sphere critical mass of plutonium-239 is 10 kg. The Nagasaki bomb contained 6.2 kg of plutonium as its fuel. At normal, room-temperature densities, a solid sphere of 6.2 kg of plutonium is not critical. Increase its density by 2.5X through the careful application of high explosives, however, and suddenly that is at least one critical mass of plutonium. Even this is something of an oversimplification, because it’s not just the density that matters: the allotropic (chemical) phase of plutonium, for example, affects its critical mass conditions (and plutonium is notorious for having an unusual number of these phases), and the Nagasaki bomb also included many other useful features meant to help the reaction along like a neutron initiator (which gave it a little shot of about 100 neutrons to start things off), and a heavy, natural-uranium tamper. […]