If you just want a bad atomic bomb, isn't it enough to put 2 chunks of fissile (enough) material at either end of a very thick steel tube and shoot them together using explosives? The question is how you use $2k to buy or produce the fissile material, because if the BOM was really $2k including "weapons grade" fissile material then the interesting question is how to get the fissile material that cheaply.
If the BOM is $2k excluding the fissile material then... is that really surprising? I thought crude bomb designs were pretty easy to create if you have no interest in maximizing yield or minimizing how dirty the bomb gets?
This would not work with plutonium, which is what Philip’s “Nagasaki-like” (Wikipedia) design apparently used. The gun design does not create the critical configuration fast enough to prevent it ‘fizzling’ - the assembly blowing itself apart before the chain reaction has progressed far enough for a nuclear-scale explosion. Instead, you need an implosion device. This is much more challenging, requiring explosive lenses (precisely-shaped charges), multiple precisely-timed detonators, and an initiator that will produce a burst of neutrons just as the explosive’s shock waves reach the center of the fissile assembly.
I am sure you are right, though, that the bill of materials did not include the fissile material.
Anyone surprised by Phillip’s result was perhaps assuming that it would take something like the Manhattan Project to independently design a bomb.
Yeah, probably the most difficult part of the entire process of making a bomb is enrichment. The uranium in uranium ore is 99.3% (useless) U-238 and only 0.7% (useful) U-235. These are chemically identical, so the only thing you have to go on is the slight difference in weight, thus the _extremely_ difficult methods like centrifuges. (You need to build a giant "cascade" of them refining at different levels.) I'd say it's basically impossible that he figured out a way to refine uranium for $2000.
I've noticed that HN is very critical of young prodigies. It's hilarious that we are doing something similar here but this time with nuclear weapons insteasd of regular science fair fare.
According to Phillips' book, his main innovation was designing the shaped charges that brought the halves of uranium/plutonium together at the right rate.
This he achieved by calling around at Dupont (IIRC) until he found someone with the knowledge, which was willingly shared. (Presumably this critical info was made inaccessible after the paper was graded.)
Freeman Dyson (who worked on the Manhattan Project and later at at IAS Princeton) reviewed Phillips' design, confiding to the course's professor, "Yes, this should work." Phillips' term paper was promptly classified. His only other feedback was to receive an A in the course.
No expert here; apparently this requires more / more enriched uranium. Since enriching uranium seems to be the bottleneck, this is suboptimal. This is why the Hiroshima bomb worked like this, but the second one, in Nagasaki, worked by compressing plutonium with explosives instead.
There's also the question of "compactness". If you want to carry the bomb by plane, say, it has to be light-ish and compact-ish, which this design lends itself less to. If you want to shoot it up in a rocket, this is even more important.
> If you just want a bad atomic bomb, isn't it enough to put 2 chunks of fissile (enough) material at either end of a very thick steel tube and shoot them together using explosives?
Linear (gun-type) fission design was tried at Los Alamos in early days of Manhattan project and was abandoned in favor of spherical implosion device [0].
Nuclear simulations are my hobby, and I have a few pet projects in the area. It is impossible to build a bona fide nuclear device for $2k (it is possible to build a dirty bomb with this budget, but it is not interesting in any way).
To achieve sustainable prompt criticality, you need high-quality starting material with low spontaneous fission, and a fast assembly (microseconds). "Gun-type" assembly is not used anymore even by new nuclear programs, as it is very inefficient (and still requires quite a lot of precision engineering). With implosion assembly, you need to learn quite a lot about shock compression, numerical simulations of detonations, and again it requires some precise engineering. "A nuclear bomb is like a half-ton Swiss watch", as somebody have said.
I believe it is possible to build a nuclear device from scratch, but you need a few years, explosion testing rigs and radiography equipment, a team of engineers and physicists, and a few millions budget. If you don't have starting material, it will be much more expensive.
The only step that takes time and space is creating enough highly enriched uranium for the bomb. You need few hundred or thousand centrifuges running a some time.
I have a friend who is nuclear physicist working for the nuclear proliferation issues in international organization. According to him, every country would do modern boosted fission weapon as their first nuke, and it can take less than year when you have easy access to experts and modern machines. Few subcritical tests (easy to hide) are needed to verify that everything works, numerical simulations do the rest.
Boosted fission weapon is still considered a fission weapon, but it uses fusion to boost the yield. You need a breeder reactor or small particle accelerator to make the necessary tritium but basically anyone can do it. Including North Korea.
More like a few tens of thousands (if not hundreds of thousands) centrifuges. Anyway, states like Iran go for the uranium way only because enrichment plants can be hidden and work underground, unlike plutonium production facilities which require open space and a lot of running water.
Thermonuclear devices require complicated X-ray reflection modeling, ablation modeling, and some other very specialized stuff which barely even have open source libraries for this. There are some, still.
>There are currently no known gun-type weapons in service: advanced nuclear weapon states tended to abandon the design in favor of the implosion-type weapons, boosted fission weapons, and thermonuclear weapons
Because A.Q. Khan and other practitioners have shown that even if you have weapon-grade enriched uranium, it is better to work out a (relatively) crude implosion device, and it will still work better than gun-type.
Not OP. Gun-type assemblies use uranium and are very wasteful in terms of fissile material. State-run nuclear programs have all gone for implosion designs, even North Korea.
Apparently it's a plutonium design, which is interesting because if it's using implosion then the detonation timing and quality of the explosives is commonly thought to be critical (notably the "Fat Man" plutonium bomb was tested before being dropped on Hiroshima, whereas the uranium bomb design wasn't all-up tested before Nagasaki as it was held to be obviously reliable)
I imagine that between the '40s and '70s many of the really hard problems became a lot easier. E.g. using solid-state electronics for the timing, modelling shaped charges as a textbook exercise rather than a research problem, etc. What surprises me is that he managed to get a dirt-cheap initiator.
Maybe he designed a geometry that can achieve the required implosion speeds with slow, easily available explosives ex. AN, KN, AP etc. The size of the device (a container) seems to suggest this.
The actual initiation is the easy part and would have been easy in the 70s, exploding wires that are subjected to a number of kilo Ameperes by some kind of avalanche, solid state switch.
I meant a modulated neutron initiator, not the explosive detonator. India was still using vapour deposition of polonium to make theirs, around the same time. I'm no engineer but that sounds like it costs a lot more than $2000.
Curious... is timing actually that big of a deal when you've already got critical mass? What would happen if you just dropped one half critical mass from 1m above to the other half?
With an implosion type bomb you need to make sure that the shockwave from the explosives hits the plutonium pit all around it almost simultaneously. If there was a portion where the explosive shockwave was delayed then the forces on the rest of the pit are going to tend to "squish" it out towards the portion where the explosion hasn't hit it yet. This asymmetrical compression means now you don't get the ideal compressed sphere so now while it's still going to reach criticality, it's going to be blasted apart with much less of the fissile material being fissioned. A great real world example of why timing is so critical is the first North Korean nuclear bomb test.
They had a yield of less than a kiloton. Obviously way greater than conventional explosives but it should have been an order of magnitude more powerful.
In order for the explosives to form that crucial explosive lense you need detonation across numerous points on the outside simultaneously so you need a detonator that's extremely consistent on timing so a simple resistive detonator with some primary explosives packed around a filament is not going to be effective. The method of choice is an exploding bridgewire design. Basically you have a thin wire and you pass enough current through it to vaporize the wire so quickly that the wire itself generates the initial shockwave as it gets turned into plasma. The rest of the explosive lense is just two different types of explosive with different propagation velocities. The two are combined with a parabolic (or hyperbolic, I forget off the top of my head) face between the two layers so that the explosion that started from a single point goes from convex to concave so it hits the outside of the pit all at the same time.
If you simply dropped a half-critical mass onto another and it was prompt critical then yes, it would explode, but with not nearly as much energy being released as a proper nuclear weapon.
I mean if you are starting from plutonium and assuming it's gonna cost $2k you are lying to yourself. I can only imagine... "Step 1. buy 5 kg of plutonium for $1k".... I think plutonium is 4 mil per kilogram according to google.
isn't getting the weapons grade Uranium (or Plutonium) what is most difficult about making a bomb anyways? My assumption is, once a terrorist group has achieved that, and has transported / smuggled it to their target, it's game over anyways. Harbor towns must be particularly vulnerable to this, considering the amount of shipping containers going through there...
For a simple bomb you want Uranium, Plutonium emits too many neutrons which are likely to cause a small chain reaction while your critical mass is in the process of forming and this predetonation can then blow everything apart before a proper critical mass has been formed. A gun type design where you just smash two subcritical lumps of metal into each other is, as
I remember, nearly impossible to achieve with Plutonium you can realistically produce.
EDIT: Interestingly another commenter says that this actually was a Plutonium design.
No, making a true atomic bomb is extremely difficult science and engineering.
Now, a dirty bomb, which is just a conventional explosive together with radioactive material - designed to spread radioactive material instead of releasing the fission energy into one giant explosion - is easy to create, but not that much more effective than just a conventional bomb. It would trigger a massive cleanup effort, and it would surely give some people cancer later down the line, but it wouldn't be that spectacular.
No, it is not that difficult, at least if you are fine with not having the most efficient weapon. You take a metal pipe, put a lump of Uranium and some explosive at each end and to a first approximation you are done. The USA did not even test their gun type design before dropping it onto Hiroshima because it was so simple that they were confident that it will just work.
Extracting u238 from 235 is a hard process. Using flourine gas, it requires you to turn an ore, to a metal, then to gas and then back to metal. If terrorists can get that done they should quit terrorism and just open a chemical plant
I'm not aware of any terrorist who actually became a commercial pilot, merely those who took a few lessons (1). And I think that's at the upper end of the lengths that terrorists will go to: Past a certain point the low-tech approaches are just way more "bang for the buck", so to speak. And past a certain specialisation of skills, it's way more effective to get that high-paying job and donate to political causes.
It’s not hard to get uranium. You can buy it on Amazon. It is enriched uranium that is hard to come by. I’m guessing the low cost for this project is that he found a way for low enriched uranium to go supercritical.
Enriched Uranium was of course implied but I edited the comment now. Another commenter mentioned that it was a Plutonium design, in which case the Plutonium was almost certainly not part of the bill.
....and you don't mind being lifted out of bed at 3am by the security services in your country, face a lengthy interrogation and a potential prison sentence.
I feel(*) it's not about the critical mass, but the critical density - that's why you need strong compression (either chemical explosives, or xrays from a smaller nuclear bomb) to trigger the explosion (exponential runoff chain reaction).
So this guy figured out how to create the critical density for 2k USD worth of components, which is actually the "hard" part of building a bomb.
You need to compress not the solid chunks of metal, but the gaseous / plasmatic plutonium after a few microseconds past the critical point. The problem is that it emits neutrons too quickly, and the thermal runaway of the critical mass happens too quickly. If you do not hold the plutonium together, it will just fly in all directions before enough of it has reacted.
No static casing us string enough; a dynamic caring made of highly compressed results of a conventional explosion works.
But it's a plutonium design, which according to common wisdom is too fissile for slow assembly but needs both compression and enough inertia to stop it from exploding with a very low yield.
I meant in general, for Plutonium you realistically need an implosion type design, but I think theoretically you could even have a gun type Plutonium bomb, just making clean enough Plutonium is totally uneconomical.
And you still need a looooong barrel and a fast moving plutonium part (see the thin man design on wikipedia and then extrapolate from there) even creating near close enough pure plutonium since economically you won't have pure.
"Parsons’ first order of business was the plutonium gun. Because it needed a muzzle velocity of at least 3,000 feet per second it would have to be 17 feet long. It should weigh no more than a ton, a fifth of the usual weight of a gun that size, which meant it would have to be machined from strong high-alloy steel. It would not require rifling but needed three independently operated primers to make sure it fired. Parsons arranged for the Navy’s gun-design section to engineer it."
Probably not, we are talking really short timescales here. The nuclear explosion is over before you can even see the first crack in the housing of your bomb.
Or you can just buy a (literal) boatload of smoke detectors (americium) and medical waste (curium) and make an americium bomb - or use a neutron gun to transmute your americium into plutonium via curium.
I've been trying occasionally for years to recover a piece of interview where the interviewee was musing on the consequences and handling of someone inventing an atomic bomb well within reach of possession for common citizens.
The comment was roughly that it would be an outstanding technical achievement and probably worthy of major awards, but of course that's not what we'd do. More likely we'd burn the designs, lock the inventor up and throw away the key.
It seems, given this post, that it was not literally accurate, but I still find it a wonderful illustration of how much technology is not morally neutral and can lend itself more or less readily to constructive or destructive purposes.
Really wish I could find it again, or at least attribute properly.
>but of course that's not what we'd do. More likely we'd burn the designs, lock the inventor up and throw away the key.
Which is exactly what we should do (in fact, the inventor shouldn't be locked up, he should be "disappeared"), if we want to avoid horrible destruction.
I really think this kind of thing might be the answer to the Fermi Paradox: Just imagine if any idiot could easily construct not just any bomb, but a doomsday bomb that destroys the whole planet. The planet wouldn't last long, because sooner or later some mentally ill person is going to build one and set it off.
You could deal with them that way, sure. But every other person after them will learn not to divulge it and take the best offer from other countries (submission mentions offers from France and Pakistan for Phillips) and give it to them, because you're not going to reach a worldwide consensus on "let's get rid of every person that develops X theoretical doomsday device" and others will want that edge. This seems so shortsighted it'd do well together with the vitrum flexile story. Just like engineers after WWII, during and after the Cold War, you'd want to give them a comfortable position using their talent for something more productive than developing military-purposes missiles. I'd be more worried about biochemical warfare anyway, I'm surprised there hasn't been any major attack with water supply as a vector.
>This seems so shortsighted it'd do well together with the vitrum flexile story. Just like engineers after WWII, during and after the Cold War, you'd want to give them a comfortable position using their talent for something more productive than developing military-purposes missiles.
This isn't like that situation at all. The WWII scientists could build bombs, but not easily: it isn't easy to get your hands on enriched uranium or plutonium at all. It seems like very few people can understand me here, so let me try to re-iterate: I'm proposing a situation where anyone able of operating a smartphone is able to download the information to build a planet-busting doomsday bomb, and build it in his living room using readily-available, common, everyday tools and materials. Once the knowledge is out there (thanks to the internet, which also didn't exist back then), you can't really squelch it. And you can't restrict the tools or materials needed because they're common. The only thing that was missing was the knowledge of physics needed to put them together.
For a somewhat similar story, look at Harry Turtledove's sci-fi story "The Road Not Taken", where some slightly dim-witted aliens attack Earth using medieval cannon-type technology, and hilariously lose, but were able to get here because they discovered some overlooked but simple branch of physics we somehow missed, which allowed them to easily achieve FTL transit.
You could just offer (with the alternatives maybe making it a "can't refuse") to put the inventor to work doing something more benign and pay them for their skills. You could even let them do side projects like these at the lab so they get to scratch that itch.
If we carefully think about it, someone is going to do it judging from the fact that the person did it purely from textbook based information. I’d rather think it’s better to identify ways to build such things and prevent wide scale adoption in a reasonable way. Locking up people and burying the evidence is a short term solution. Building ways of resilience and to be able to thwart or diffuse such bombs or to make laws which make it harder to extract raw materials seems a good long term solution to think about.
>I’d rather think it’s better to identify ways to build such things and prevent wide scale adoption in a reasonable way.
But how do you do that? Suppose some new physics are discovered, and this makes it trivially easy to build a doomsday bomb using readily-available materials that simply can't be restricted.
>Locking up people and burying the evidence is a short term solution.
Unfortunately, yes: if it gets to my scenario above, it's probably just a matter of time.
>Building ways of resilience and to be able to thwart or diffuse such bombs or to make laws which make it harder to extract raw materials seems a good long term solution to think about.
Again, in my scenario, anyone with half a brain can easily build the doomsday bomb using stuff from the grocery store. We're not talking about plutonium here, we're talking about very common materials. How do you stop the world from being destroyed?
I think I also remember that interview or maybe a YouTube video about it. Something like what would we do if someone discovered that you can put sand and magnets into a microwave with the plate rotating counterclockwise and it goes gigantic boom. Have a look at the Veritasium YouTube channel, I have very vague memories that it might have been there. Or maybe Vsouce or Kurzgesagt.
EDIT: The Vulnerable World Hypothesis [1] with the Easy Nukes thought experiment. And this a TED talk with the author [2] and probably what I watched.
I am afraid that current AI models are the black balls. For example it is now close to impossible to verify anything, even that whether the Russians nuked half of the country or not. (It is somehow doable with trusted chains, but can you really trust them?)
Also I suspect that hacking AI will be much more powerful than any hacking with hand soon.
> More likely we'd burn the designs, lock the inventor up and throw away the key.
Which is depressing since nuclear armed nations dont get invaded or subjugated. Arguably the inventor should be given the noble peace prize every year going forwards.
Well, some sort of Nobel prize may have been mentioned in the awards reasoning, but the premise in this case is more in the spirit of '640 kilotons ought to be enough for anybody'...
Nuclear armed private citizens may not get invaded or subjugated either, yet it's hard to see that the home atomics revolution would make us collectively safer on the whole.
Instead of safety devices ("slowly moved away from creating nuclear things and invented a motorcycle air bag").
Surely that would have a better impact on the world instead of aiding political campaigns with machine learning, ads and data.
Edit: Yeah, reading his Wikipedia page, I'm 100% sure he could have spent his time better.
> The experience he had gained during his campaigns obtaining the voter list from the state and using it for campaign purposes led him [...] to found Aristotle, Inc. [...] a consulting firm for political campaigns [...] It specializes in combining voter lists with personal data from other sources (such as income, gun ownership or church attendance) and data-mining, to assist with micro-targeting of specific voter groups; as of 2007, its database contained detailed information about ca. 175 million U.S. voters and it had about 100 employees.
Another example indicating that The Great Filter is likely technological - I reckon most species have the technological capability for a few rogue individual to wipe them out completely before they have the capability to become truly multiplanetary, and therefore hedged
The real question is every human being have ability to have create these kind of bomb, I can see civilisation can be wiped out in an instance at our current level of mental/psychological/spiritual maturity.
I also read that book and thought it was pretty good (a few small parts were cringy). Tech bros weren't really a thing then, or at any rate I didn't mind it at the time. The linked page, which I hadn't seen before and is about what he is doing now: well THAT seems bro-y. Oh well.
Missing the point. If lone wolf terrorists have access to engineered viruses or atomic bombs, that's bad in a unique way, because there's probably going to be >= 1 unhinged people who will use it, and all you need is 1.
If we're dealing with a world where only states can access the tech, the probability that >= 1 state will be sufficiently unhinged to use it is far lower.
We will eventually need a metaphorical CTRL+Z on tech, probably through government power and maybe world government. The fragility of us as organisms is about constant, but tech is going to keep getting better and cheaper.
Lone wolf terrorists wouldn't result in annihilating ~every city on Earth, but state actors might.
One nuclear bombing is bad. 1,000 nuclear bombings is worse. Risk = probability × outcome.
It's not a choice between only state actors having access vs only lone wolf terrorists. The issue of nuclear disarmament is choosing between states + terrorists (the status quo) vs only terrorists.
Correct when it comes to present day nukes, maybe not if we're talking engineered viruses or some as yet undiscovered weapon. Anyway, I think the point stands that there's significant novel risks posed if individuals of the future are capable of destroying cities if they feel like it, and there will unfortunately need to be a paradigm shift around privacy and government involvement if that day comes.
We don't have a Ctrl-Z, but we have a sort of a pause button. Non-proliferation agreements, IEAE inspections, bans on selling certain products that are key enablers of production of atomic weapons.
With viruses, it's a bit harder, medical work us hard to distinguish from weapons work.
This is again referring to the current state of affairs and therefore misses the point. These controls work for now because such tech is expensive and difficult. A future where that's not the case isn't so unlikely.
Typical government overreach. If you live in a rural area you can't rely on the government for nuclear defense. How are they supposed to defend themselves without nukes of their own?
It's like a $5 meal kit for a "BBQ meat feast" that has the recipe, spices, but doesn't include the meat. So once you've visited the specialist butcher and spent a lot more money, now you have the $5 feast.
If this paper (that none of us can ever see) really included the plutonium in that BOM then it would be interesting to know how he derived a price for it, since it's not really a openly traded commodity.
Funny enough, Pu-238 was on the market at one point (with plenty of restrictions). It was used to power pacemakers. It's still used for spacecraft.
I'm not sure how expensive it was in the '70s but it is very expensive now. American production is supposed to be scaling up, but only to about 1.5 kg a year.
But Pu-239 is presumably the stuff for this hobby project? The one-click price for that is hard to find, even if you bring your own take-out container.
Yes, I was looking for a point of comparison. Pu-239 is probably also quite expensive, as both isotopes are prepared by using a reactor to irradiate stuff. Only way I can see a "$2k" cost as relevant is if it's the cost to someone who's already acquired the nuclear material.
The real nuclear terrorist threat comes from nuclear power plants, they are way less secure than governments want you to believe they are.
As we now know that they go boom once there is no more cooling, which occurs if there is a power outage (with failure of the backup diesel generators) or if there is a leak in the waterpipes.
If the BOM is $2k excluding the fissile material then... is that really surprising? I thought crude bomb designs were pretty easy to create if you have no interest in maximizing yield or minimizing how dirty the bomb gets?