Building fission bombs

Octopus graffiti, with mustache and glasses

As recommended by a fellow attendee at the unofficial summer ‘grill thrill’ barbecues, I am currently reading Richard Rhodes’ Dark Sun: The Making of the Hydrogen Bomb. While the book can be detailed to the point of exhaustion sometimes, it does contain a lot of interesting information, on everything from atomic bomb design to the differences in governmental structure and operation in the United States and Soviet Union.

One thing the book has definitely done is diminish my concerns about terrorists building nuclear weapons. Even the ‘simple’ gun-type configuration uranium bomb is a lot more complicated that many of the diagrams and descriptions I have seen would make you believe. A plutonium implosion device is far more complex still. Getting from a sufficient quantity of fissile material to a working bomb is an extremely complex undertaking, requiring a lot of equipment and expertise. It also requires a lot of exotic materials and manufacturing processes. It is certainly easier now than it was for the Russians in 1949 (largely because more information is available), but the degree may not be as great as most people think. Because of espionnage, the Russians actually had the plans for the American bombs while they were building their own. Even under intense pressure from Stalin and Beria and with considerable resources (including access to industrial facilities and thousands of forced labourers), it took the Soviets four years to copy them. That makes it seem unlikely that terrorists without significant support from a state, access to industrial facilities, and high degrees of technical knowledge could emulate them.

Another interesting topic covered in the book is the hasty abandonment of Los Alamos at the end of the war. It would make interesting reading for those who saw the advent of atomic weapons as an immediate sea change in warfare. As it happened, there was apparently a long period after the war where no usable weapons were assembled and available, and the teams of people who would be required to make them so were dispersed around the United States, doing other things. The first bombs definitely weren’t designed with simplicity or shelf-life as a top priority. As a consequence, most of the deterrent effect of the bombs in the immediate post-war period was based around faulty information.

I will write a full review of the book when I have finished it.

[Update: 12 April 2010] My full review was online quite a while ago: Dark Sun: The Making of the Hydrogen Bomb.

Author: Milan

In the spring of 2005, I graduated from the University of British Columbia with a degree in International Relations and a general focus in the area of environmental politics. In the fall of 2005, I began reading for an M.Phil in IR at Wadham College, Oxford. Outside school, I am very interested in photography, writing, and the outdoors. I am writing this blog to keep in touch with friends and family around the world, provide a more personal view of graduate student life in Oxford, and pass on some lessons I've learned here.

27 thoughts on “Building fission bombs”

  1. RDS-1, the first Soviet nuclear weapon, was an implosion-type plutonium device. It seems likely that they could have built a gun-type uranium bomb more quickly.


    John Mueller
    Department of Political Science
    Ohio State University
    January 1, 2008

    A terrorist atomic bomb is commonly held to be the single most serious threat to the national security of the United States. Assessed in appropriate context, that could actually be seen to be a rather cheering conclusion because the likelihood that a terrorist group will come up with an atomic bomb seems to be vanishingly small. Moreover, the degree to which al-Qaeda–the chief demon group and one of the few terrorist groups to see value in striking the United States–has sought, or is capable of, obtaining such a weapon seems to have been substantially exaggerated.

  3. Others contend the crudest type of bomb would be “simple and robust” and “very simple” to detonate (Bunn and Wier 2006, 140). Younger disagrees:

    “Another challenge…is how to choose the right tolerances. “Just put a slug uranium into a gun barrel and shoot it into another slug of uranium” is one deception of how easy it is to make a nuclear explosive. However, if the gap between the barrel and the slug is too tight, then the slug may stick as it is accelerated down the barrel. If the gap is too big, then other more complex, issues may arise. All of these problems can be solved by experimentation, but this experimentation requires a level of technical resources that, until recently, few countries had. How do you measure the progress of an explosive detonation without destroying the equipment doing the measurement? How do you perform precision measurements on something that only lasts a fraction of a millionth of a second? (2007, 89)”

  4. It seems likely that they could have built a gun-type uranium bomb more quickly.

    It is interesting that scientists under such intense pressure opted for the more complex (though much more efficient) plutonium implosion design. I will let you know if I can find an explanation why in the book.

  5. The security concern is about a terrorist organization using a new weapon, not building one. It takes a considerable amount of time and resources to successfully construct a nuclear weapon, but the inability of small organizations to construct such weapons doesn’t preclude use.

    This is the concern over nuclear proliferation. Depending on which news network you fancy, Iran’s nuclear weapons program has either come to a complete halt, is over a decade away, two years away, four months away, or has already constructed a nuclear weapon. As messy as the political situation becomes when new countries develop nuclear deterrents, the more frightening aspect is the whether those countries will sell or supply information, resources, or nuclear weapons to smaller organizations that might be considerably more willing to use them.

    I’ve never seen the possibility of a terrorist attack using a nuclear weapon presented as if the organizations are trying to develop their own nuclear technology. The ambiguous “assemble” or “get their hands on” might be used, but the main concern is with small groups ability to purchase or steal nuclear weapons from legitimate states, or for states to carry out covert attacks against their ideological rivals through the use of a third party.

    “the likelihood that a terrorist group will come up with an atomic bomb seems to be vanishingly small.” Hopefully this is true, but I question the reasoning behind this claim. I don’t claim an expertise in this field; however, anecdotally proliferation seems to be on the rise. Indian and Pakistan have nuclear weapons. An Indian scientist admitted to selling nuclear research. North Korea is believed to have several rudimentary nuclear weapons. Iran may or may not be in the process of developing them. The only reason I could imagine as supporting this claim is that poor security of the U.S.S.R’s nuclear stockpile has historically been feared method of armament. The fact that Russia is no longer a threat is reassuring although we are still unaware of what (if anything) went missing from their stockpile.

    Anyways, I’m not trying to be a fear-monger. I didn’t believe Iraq was on the verge of a nuclear weapon, and don’t believe Iran has or is close to having them. (I tend to accept reports the IEA gives, over the CIA) However, I’m hoping you will clarify your statement, because I would be interested in knowing why there is less opportunity for terrorists to get their hands on nuclear weapons. It was my understanding the risk will trend upward with proliferation, which is only going to become more likely as technological prowess and globalization continue. Complexity only seems to discount the – every bright yet disgruntled student will build their own fusion weapon – scenario presented by the Outer Limits.

  6. One possible answer to the question I asked about the Soviets is that the implosion-type device requires a lot less fissile material.

  7. I finished reading the Mueller article. He makes an effective, although extremely limited case from complexity. The basic premise of the article is logically sound. By multiplying the probabilities of many consecutive goals needed to achieve a successful nuclear detonation will give us an overall probability. However, he even admits that there is a drastic difference in outcome depending on how you assign the probabilities for completing each stage. I’m not going to take him to task for that, because I happen to agree that even liberal estimations lead to a very low probability over the course of the twenty-step list. The major and almost fatal flaw is that he assumes terrorists will have to pass through most of the steps. He only says three things about the possibility of terrorists procuring a complete weapon, which skips to step eighteen; (1) Historically countries have refused to arm other countries with nuclear weapons, the singular example being China’s refusal to arm North Korea. (2) Even if countries could trust others (states, terrorist groups), they will refuse to arm them because of the possibility that the world might find out they did so. (3) Most people don’t think this is likely.

    Mueller’s treatment of Russia is great. Security has increased. There was an accounting for fissile material. Nuclear bombs deteriorate, so even if terrorists had fissile material, or even most of a bomb, the fact that they haven’t used it yet is actually a valid reason for doubting they currently have the capacity. Anything that falls under the basic rubric of the multistage technical challenge, Mueller covers in detail, explains very well, and makes a convincing case. Unfortunately, he is so completely focused on deriding anyone who believes this is possible, that he becomes flippant. I agree with the broad swath of his argument, we should resist the alarmists and take a serious look at the issue, and there are probably programs more deserving of funding than trying to protect against low probability threats. But his case suffers when he tries to inflate the odds in his favour to the maximum. Crossing international boarders with fissile material is a hurdle that carries a legitimate probability of failure, but Mueller wants us to believe that being a tourist, unable to speak the language and being unfamiliar with the roads and driving regulations of the state the group is purchasing the fissile material from, constitutes a serious probability of failure. How many people really believe that a terrorist group is going to fully inform the law transgressing, but ultimately good hearted smuggler, as to the nature of the material he is to transport before leaving the goods in his sole care?

    There are other areas where Mueller also gets sloppy. Several times he appeals to history and common knowledge. The obvious failure being that the danger from this case lies in what the public doesn’t know. He doesn’t really stop to consider whether governments would public declare that they had lost, sold, or given nuclear material or weapons to another country or organization. I’m not going to launch into some conspiracy theory, but to use history – so far as we know, no state has decided to give, or has lost a nuclear weapon – as an absolute predictor of future behavior questionable.

    Oddly enough, he tries to use the fact that Pakistani scientists were punished for selling nuclear secrets in an argument about how the risks of getting caught would insure that scientists won’t sell nuclear secrets.

    Additionally, he is very focused on America and the probability it will suffer a nuclear terrorist attack. Most of the paper is about Al-Qaeda, which is indirect focus on America. He offers valid reasons for doing so, Al-Qaeda has mentioned wanting to procure nuclear weapons, and most experts believe they are among the most advanced terrorist organization out there. So I don’t condemn Mueller focus, but it is unfortunate, because it neglects analysis of organizations that might have strong relations and shared politics with other states, to the point where those states might be willing to arm them, and are less likely to be caught. In this light, (3) isn’t a real argument. (2) This might be true, but the criminal cost/benefit analysis usually focuses on the likelihood of getting caught. (1) isn’t even true, since Canada was armed, and is predicated on what the public knows, and ignores the sympathies governments might have for some “terrorist/freedom fighter” movements.

    The real flaw behind Mueller’s dismissal of the possibility of procuring a complete bomb comes from a failure to address the fundamental risk of proliferation. Technical knowledge is cumulative. Every bit of knowledge released makes the process easier. Mueller makes a compelling case about the technical limitations of trying to procure or steal and assemble the parts of a nuclear weapon from scratch. He also makes a nice psychological argument; terrorists are not likely to want to invest in the process if they investigate the challenges, but he offers almost no treatment of the possibility of getting a fully assembled bomb, and hasn’t provided a reason to believe that proliferation doesn’t increase the risk to greater than “vanishingly small”.

    I did enjoy being called an alarmist crackpot though.

  8. One possible answer to the question I asked about the Soviets is that the implosion-type device requires a lot less fissile material.

    I think a lack of U235 is the reason the Soviets went with plutonium and, by extension, an implosion-type device.

    According to Rhodes’ book, U235 was just becoming available to Soviet bomb-makers in March of 1950. By contrast, they tested their first plutonium bomb on 29 August 1949.

    To get U235, you need either centrifuges or a gaseous diffusion plant. By contrast, a big pile of graphite and uranium can be used to breed plutonium, though the effort required to isolate it after is industrial in scale.

  9. “Later scholarship has also shown that the decisive brake on early Soviet development was not problems in weapons design but, as in the Manhattan Project, the difficulty in procuring fissile materials, especially since the Soviet Union had no uranium deposits known when it began its program (unlike the United States).”

    From Wikipedia, the free encyclopedia

  10. The recent questions directed towards Pakistan completely undermine Mueller’s analysis. I think he has to take the risk of terrorist organizations obtaining complete weapons more seriously.

  11. From what I have seen and read, the Taliban is no threat to the Pakistani state, despite the proximity of Swat to the capital.

    That being said, the presence of nuclear weapons in unstable states is definitely worrisome.

  12. I agree. I’m not particularly concerned about Pakistan losing a nuclear weapon right now, but still think the largest probability of terrorist organizations procuring nuclear weapons is through acquiring complete ones, which invalidates Mueller’s analysis of the steps required to build one.

  13. To differing degrees, I think all existing nuclear weapons have control systems to prevent their use by unauthorized personnel. Of course, it is almost impossible to find out exactly what those systems consist of.

    If they were very well designed, they could force even a sophisticated thief to disassemble the bomb and rebuild it, in order to get a working device.

    For instance, if the exploding bridgewire detonators attached to the high explosive shell around the pit (or primary) have fuses with random-length delays, detonating the bomb successfully would require knowledge of what all the delays were. If the detonators do not fire in a very precisely coordinated way, the result is likely to be the liquefaction of the plutonium core, followed by it being forced out of the casing as a fountain of liquid metal. That is nasty, but much less so than a detonation (especially for a themonuclear bomb). Here is more on bomb design.

    I wonder which states are actually most likely to lose control of their bombs: relatively unsophisticated powers who only have a few, or large powers with hundreds or thousands. The former have fewer overall resources, but they also derive more ‘value’ per bomb.

  14. Hacking Nuclear Command and Control

    “The International Commission on Nuclear Non-proliferation and Disarmament (ICNND) has released an unclassified report exploring the possibility of cyber terrorists launching nuclear weapons. Ominous exploits include unreliable early warning sensors, unsecure nuclear weapons storage, transportation blunders, breaches in the chain of command, and the use of Windows on nuclear submarines. A traditional large-scale terrorist attack, such as the 2008 Mumbai attacks, could be combined with computer network operations in an attempt to start a nuclear war. Amidst the confusion of the traditional attack, communications could be disrupted, false declarations of war could be issued on both sides, and early warning sensors could be spoofed. Adding to this is the short time frame in which a retaliatory nuclear response must be decided upon, in some cases as little as 15 minutes. The amount of firepower that could be unleashed in these 15 minutes would be equivalent to approximately 100,000 Hiroshima bombs.”

  15. Iran: The Challenge of Independent Enrichment
    February 26, 2009

    Even as it conducts pre-commissioning testing at the Bushehr nuclear power plant, Iran continues to struggle with its efforts to independently master the nuclear fuel cycle. A series of recent reports suggest that not only is Iran still experiencing very real problems, but that efforts have slowed considerably.

  16. Nuclear Weapons: Terrorism and the Nonstate Actor
    July 1, 2008

    The threat of nuclear terrorism has become a fact of life in the nuclear age, brought home by the 9/11 attacks and the prospect of a nonstate actor getting hold of a deliverable nuke. While such a scenario indeed seems ominous, it is not as likely as many fear. The greater danger is the availability of highly enriched uranium (HEU), though increased security worldwide of HEU stocks is making it less likely that sufficient quantities will fall into the wrong hands. Still, eternal vigilance is required to mitigate the threat, however feasible the threat might be.

  17. “Indeed, the most rudimentary nuclear device can wreak immediate and extensive devastation. The good news is that, unlike RDDs, a nuclear device is neither easily fabricated nor easily acquired. STRATFOR has discussed the challenges facing state actors that want to embark on such a path. A nuclear weapons program represents a profound and comprehensive commitment of national resources, requiring not just a single facility but a complete industrial base. And the challenges are compounded dramatically when the nation-state seeks to hide the pursuit from the international community.

    It is not a matter of simply recruiting or kidnapping a few experts. A nuclear weapons program requires the long-term ability to establish and maintain facilities — often very electricity-intensive — and conduct years of experimentation. Immense quantities of materials must be acquired from abroad — everything from raw precursors to fissile material to high-grade industrial components — often subject to intense international scrutiny.

    It took North Korea extensive Soviet assistance with both civilian and military nuclear technology and more than 50 years to get to the point where it could test a device that fizzled. Though subject to international sanctions, Pyongyang was able to accomplish this with facilities that were never bombed and the industrial resources of the entire nation. Ultimately, even nonstate actors that control territory – such as Hezbollah and the Revolutionary Armed Forces of Colombia — are regularly disrupted by opposing state actors and would be stretched to coordinate such an effort (holding hostages for years is not the same as sustaining and powering a complex and fixed nuclear development facility for a decade or more). Given the current state of world affairs, it is simply not possible for a nonstate actor to successfully fabricate a nuclear device from scratch.”

  18. Nuclear Weapons: Devices and Deliverable Warheads
    June 17, 2008

    On July 16, 1945, at a remote testing range in southern New Mexico, the United States detonated the world’s first atomic bomb. Developing the device was probably the most complex and expensive exercise in applied physics in human history. Even today, weaponizing the atom remains one of the most challenging endeavors a country can engage in — and one few ultimately choose.

  19. “Mohamed ElBaradei, the outgoing head of the International Atomic Energy Agency, the UN’s nuclear watchdog, said that Iran would not produce a nuclear weapon any time soon. “In many ways, I think the threat has been hyped,” he told the Bulletin of the Atomic Scientists. But in its latest report the agency chided Iran for concealing military aspects of its nuclear programme, which the Islamic Republic says is only civilian.”

  20. “The bottom line is that a nuclear device is the only element of the CBRN threat that can be relied upon to create mass casualties and guarantee the success of a strategic strike. However, a nuclear device is also by far the hardest of the CBRN weapons to obtain or manufacture and therefore the least likely to be used. Given the pressure that al Qaeda and its regional franchise groups are under in the post-9/11 world, it is simply not possible for them to begin a weapons program intended to design and build a nuclear device. Unlike countries such as North Korea and Iran, jihadists simply do not have the resources or the secure territory on which to build such facilities. Even with money and secure facilities, it is still a long and difficult endeavor to create a nuclear weapons program — as is evident in the efforts of North Korea and Iran. This means that jihadists would be forced to obtain an entire nuclear device from a country that did have a nuclear weapons program, or fissile material such as highly enriched uranium (enriched to 80 percent or higher of the fissile isotope U-235) that they could use to build a crude, gun-type nuclear weapon.

    Indeed, we know from al Qaeda defectors like Jamal al-Fadl that al Qaeda attempted to obtain fissile material as long ago as 1994. The organization was duped by some of the scammers who were roaming the globe attempting to sell bogus material following the collapse of the Soviet Union. Several U.S. government agencies were duped in similar scams.”

  21. Well, this is out of nowhere, but I thought it might be of interest to you (or perhaps you already knew) that Canada possessed nuclear weapons during the cold war:

    The primary weapon of Canada’s primary fighter jet of the era was an unguided rocket with a 1.5kt nuclear warhead. Wikipedia has a good article about the weapon. Technically, the weapons were owned by the US for use by Canadian aircraft.

  22. There were also CIM-10 Bomarc missiles with nuclear warheads deployed in Canada, between 1963 and 1971.

    At present, I am reading Richard Rhodes’ The Making of the Atomic Bomb. I will post a review when I finish it.

  23. Additional information on the challenges of building gun-type uranium 235 bombs is included on p.463 of the paperback of Richard Rhodes’ The Making of the Atomic Bomb.

    In particular, he discusses problems with predetonation, the difficulty of getting adequate muzzle velocity (see also p.467), and the large volume of material required for non-implosion designs.

    For information on implosion-based plutonium bombs, see p.467.

Leave a Reply

Your email address will not be published. Required fields are marked *