[Heather Jameson (OP)]

Zubin asks:

Could you use two rail guns to collide two plutonium cores for nuclear fission?

What do you think?

[chiralSPO]

Yeah, that would probably work, but what would be gained by doing it that way (and what would be lost... I see two possible scenarios:

1) The two rail guns are in close proximity. This means that the rail guns would likely be destroyed be the resulting nuclear blast. Rail guns are very expensive, especially compared to the alternative of using conventional explosives (which are fairly cheap, and for which the engineering has already been done).

2) The two rail guns are far enough away that they can survive the blast. This could be an interesting way of delivering a nuke without needing a missile or plane. It would be pretty hard to detect a fist-sized chunk of fissile material flying through the air at Mach 4 (and even if noticed, I'm not sure what kind of countermeasures could be deployed. BUT it comes with the problem of getting the two chunks to hit each other square on. Timing, aiming, and accounting for things like wind and humidity would likely be beyond what we are currently capable of. (as far as I understand it, good marksmanship with railguns currently includes hitting battleships from 20 miles away, not hitting supersonic seagulls...)

[Bored chemist]

I read somewhere that the original "gun" design for producing a critical mass was regarded as impractical- the mass of the gun needed would be for to great.
Then someone explained to the gun designer that "It only needs to work once".
Apparently that makes it a bit easier.

It's true two rail guns would be very expensive, but nukes aren't cheap (thankfully). You can't get the plutonium from the supermarket.

[Halc]

A simple hemisphere dropped from a height onto another will suffice.
You put a small hole in one and run a cable through it to guide the one piece onto the other.  Bang!
A variant of this is known as the flagpole bomb.  Quite crude and very low yield since the compressing forces (inertia in this case) are not symmetrical, but it's a crude way of implementing a dirty bomb which will render a region nicely radioactive and maybe level a building or two.

Any two pieces brought together at rail gun speeds will suffer the same low yield from the asymmetry.

[Bored chemist]

A simple hemisphere dropped from a height onto another will suffice.

Really?

" In particular, the gun method cannot be used with plutonium "
from
https://en.wikipedia.org/wiki/Nuclear_chain_reaction#Predetonation

[Halc]

" In particular, the gun method cannot be used with plutonium "
from
https://en.wikipedia.org/wiki/Nuclear_chain_reaction#Predetonation

It cannot be used for a viable efficient weapon.  It would still go bang if you did it, as depicted in the predetonation picture.
I didn't even know that little-boy was a gun-type weapon. Apparently they were a common method for low-yield shells and such.

[Bored chemist]

So, you think the OP's question was "can you use  two rail guns to achieve the same thing you could do with a bit of string?".
Well, you may be right, but that's not how  I interpreted it.
I thought they were seeking to overcome the fact that you need to get things moving very fat to get a proper detonation with plutonium

[Halc]

So, you think the OP's question was "can you use  two rail guns to achieve the same thing you could do with a bit of string?".

Dropping something from a height will in theory get you any speed you like.  If not, you need to drop it into a deeper hole.  The string is only for aiming.  Apparently the technique is enough and the U235 bomb could have worked that way, but the gun does the same job in far less space (the length of the device).

I thought they were seeking to overcome the fact that you need to get things moving very fat to get a proper detonation with plutonium

The OP said 'for nuclear fission' and colliding two plutonium cores would indeed result in their fission, albeit not the 'proper detonation' they typically require.  No amount of speed, gun or not, will suffice.  Plutonium for some reason requires spherical symmetry to get that proper detonation, and hence a pair of rail-guns will not work.  All you would get is that much weaker predetonation that I've been talking about.

[alancalverd]

In summary: yes, but it is a large, complicated, expensive and unreliable way to do it. The ideal explosive device is small, simple, cheap and reliable.

[Bored chemist]

Dropping something from a height will in theory get you any speed you like.

No
On this planet the speed is limited to about 11 km/s.
In practice it's (obviously) much lower.

Apparently the technique is enough and the U235 bomb could have worked that way,

Define "could have worked". In particular, how tall a vacuum tube  would you need?
The velocity they wanted was of the order of 1 km/s

Would they have got planning permission from the Japanese authorities for such a structure?

The OP said 'for nuclear fission'

Yes; they did.
Plutonium undergoes spontaneous fission.
so there's no need for a rail gun (or anything else).

Again, I rather doubt that's what the OP had in mind.

Plutonium for some reason requires spherical symmetry

Got any evidence for that?
In particular, how does one bit of the plutonium know that it is in a spherically symmetrical compression?
The symmetry shows that only the exact (vanishingly small) centre of the bomb can be compressed truly spherically.
For any other bits the loading is "off centre".

If what you said was true, it would be impossible to detonate plutonium, yet we know it is possible, so we know you are wrong.


However, as I said, plutonium does undergo spontaneous fission (strictly- some isotopes do).
The problem with that is that those spontaneous events can trigger the fizzle before the supercritical mass is properly assembled- and that gives you a fizzle.
If you can get the plutonium from "far away" to "assembled" quickly enough you get a proper detonation (the yield's not as good as if you compress it- due to momentum considerations).
You can't do that with a conventional gun.
You might be able to do it with a rail gun

It wouldn't be very practical

The rate of fission in uranium is much lower so you have more time to assemble the supercritical mass before some event triggers the chain reaction

[syhprum]

Dropping anything from a great height will only get it to the escape velocity of the planet upon which you drop it !

[Halc]

Dropping something from a height will in theory get you any speed you like.

No
On this planet the speed is limited to about 11 km/s.

Which is why I said you need a deeper hole (gravity well) if you want more.  I think 11 km/sec is more than the gun in any nuclear device.

In particular, how tall a vacuum tube  would you need?
The velocity they wanted was of the order of 1 km/s

Much taller than the bomb they built, yes.  Not at all practical for the sort of purposes they typically have in mind.

Plutonium undergoes spontaneous fission.
so there's no need for a rail gun (or anything else).

Good response to the OP, yes.  Any amount will produce fission, but a minimum amount is needed to achieve k=1 (criticality), but there's no speed required for it.  A guy with a shovel will suffice.

Again, I rather doubt that's what the OP had in mind.

Well, we must take that into account.  Here's what he said, and here's what we assume he had in mind.  A dropped chunk of material (from say 100m up) isn't going to level a city, but it will definitely still blow stuff up. That was my point there.

Plutonium for some reason requires spherical symmetry

Got any evidence for that?

The gun method doesn't work with it, per your linked page.  The discussion ( https://en.wikipedia.org/wiki/Nuclear_weapon_design#Implosion-type_weapon ) behind that goes into how they shape the charges to get material moving in symmetrically from all sides. A good deal of the technology to get a working bomb in the first place revolved around getting a symmetrical explosion like that.  It's not easy to do.

In particular, how does one bit of the plutonium know that it is in a spherically symmetrical compression?
The symmetry shows that only the exact (vanishingly small) centre of the bomb can be compressed truly spherically.
For any other bits the loading is "off centre".

Yes.  Any ball is spherically symmetrical despite the vast majority of it being "off centre".  The term does not mean that it's all at the center.

You can't do that with a conventional gun.
You might be able to do it with a rail gun

A bunch of guns maybe.  Just one applies momentum/pressure in one dimension instead of from all directions.

The rate of fission in uranium is much lower so you have more time to assemble the supercritical mass before some event triggers the chain reaction

Yes, I learned that from this thread. Hadn't really looked into it before.

[Halc]

On this planet the speed is limited to about 11 km/s.

Dropping anything from a great height will only get it to the escape velocity of the planet upon which you drop it !

A given mass (the mass of Earth say) does not have an escape velocity.  Formula is V = √(2Gm/r) which means any velocity can be achieved (up to c actually) given a low enough r.  So the velocity you're thinking of assumes a fixed r of the surface radius of Earth.

Second point:  If I drop a rock onto Earth from an initially stopped position a light-year up from Earth, it will hit at something like 50 km/sec, ±30 km/sec to account for Earth accelerating back and forth as it orbits, but let's assume we hold it still for this exercise.
I can get it up into something like 600/km/sec by going even higher.  In fact, if you take all the mass away pulling in the wrong direction, I cannot think of a limit at all.  Earth is actually amazing deep in a gravity well, far more than one typically imagines.

B-C's 11 km/sec assumes the Earth is the only mass in existence, and assumes the radius at the current surface, but not if we squish it, which is the 'deeper hole' I spoke of in the part of my comment, full context:

Dropping something from a height will in theory get you any speed you like.  If not, you need to drop it into a deeper hole.

[Bored chemist]

Which is why I said you need a deeper hole (gravity well) if you want more.  I think 11 km/sec is more than the gun in any nuclear device.

I was going to ask what planet you think the OP is on, but I think it may make nor sese to ask what planet you are on.

There is, of course, still a limit; C.

Any ball is spherically symmetrical despite the vast majority of it being "off centre".  The term does not mean that it's all at the center.

Nobody said it did.

But you failed to address the point.
Was that deliberate, or did you not understand it?

[Halc]

I was going to ask what planet you think the OP is on, but I think it may make nor sese to ask what planet you are on.

Read my subsequent post which gives more detail (two reasons) on why 11 km/sec is no speed limit when dropping something onto something with the mass of Earth.

The symmetry shows that only the exact (vanishingly small) centre of the bomb can be compressed truly spherically.
For any other bits the loading is "off centre".

Yes.  Any ball is spherically symmetrical despite the vast majority of it being "off centre".  The term does not mean that it's all at the center.

Nobody said it did.

Well, it appeared that you did in your statement above. If the entire bit of material is compressed in a spherically symmetrical way, then the compression (all of it, not just the center) is spherically symmetrical.  That's not to say it is identical everywhere just like the pressure and temperature at various points on the interior of Earth is not the same despite Earth being nearly spherically symmetrical.
And no, the various bits of mantle in the Earth probably have little way to locally test if the whole collection of bits is spherically symmetrical, so the bits don't 'know'.

But you failed to address the point.
Was that deliberate, or did you not understand it?

I think I did not understand it.  I saw no point other than that to which I responded.  Perhaps you could rephrase it.  You said this:

In particular, how does one bit of the plutonium know that it is in a spherically symmetrical compression?

You're obviously not talking about bits of plutonium actually 'knowing' anything, so that leaves what each bit does, and each bit cares not a hoot if it's in a spherically compressed configuration or not.  It cares that there is enough other plutonium nearby that k is above supercriticality. This is only true for material near the center (but not vanishingly small), and the size of this region is increased by surrounding the sphere with neutron-reflective material which vastly increases the yield. This surrounding material, being a spherical shell of sorts, does not break the symmetry.

The gun method very much does break the symmetry since there are just two parts approaching along a clear axis with no particular pressure/momentum coming from any of the sides.  The reaction is apparently too fast to allow this sort of binary approach.  So the way to compress it is to have it come in from all directions at once.  That's spherical symmetry.  They actually do it with a finite number of discreet chunks (a finite number of guns if you will), but each chunk has a shaped charge designed to behave like a series of much smaller charges evenly arranged, so each 'gun', instead of firing one bullet, instead fires a bunch of bullets converging on a point, which is the center of our sphere.

[Bored chemist]

So, finally you accept that there's no actual requirement for  spherical symmetry- there can't be because there is no way that the metal can know if the compression is spherically symmetric.
I'm not saying that it is self aware, I'm using it as a short hand.
The material behaves in accordance with the situation in which it finds itself.
And, if the compression is large enough, the chain reaction can get underway.
But there is no way in which a "distant" part of the metal can influence the behaviour of another part.
So there is no mechanism for an unsymmetrical compression to cause a different outcome- as long as the density is high (and remains that way for some time- ns to µs.

So, fundamentally, after forcing people to swim in treacle, you seem to accept that this claim you made was wrong.

Plutonium for some reason requires spherical symmetry to get that proper detonation

It needs a critical mass to be assembled very quickly- that's all.

[Halc]

The material behaves in accordance with the situation in which it finds itself.

Yes, and spherically symmetric makes it behave with the biggest bang for the buck.

And, if the compression is large enough, the chain reaction can get underway.

As you've pointed out, the chain reaction goes on even when below critical state.  The compression doesn't matter.  What matters is sufficient material in close proximity.  The pressure is just a side effect, not a cause of the reaction.

But there is no way in which a "distant" part of the metal can influence the behaviour of another part.

Not sure what you mean by this.  It's all about proximity, so sure, if something is meters distant, it plays little role, but if something is nearby (distant, but only a little), then it very much has an influence.

So there is no mechanism for an unsymmetrical compression to cause a different outcome- as long as the density is high (and remains that way for some time- ns to µs.

Sure, I agree with this to a point.  I can arrange the material in a long thing rod with greater overall density and get almost no reaction at all, so I don't totally agree that it's about density.  A long dense rod hasn't the proximity needed, and requires no pressure to maintain its shape.

So, fundamentally, after forcing people to swim in treacle, you seem to accept that this claim you made was wrong.

Is that what this is about? Which claim am I accepting as wrong?

Considerable effort was expended in the Manhattan project attempting to generate a spherically symmetric explosion. A lot of the current science of shaped charges was developed by that project. They blew up countless test objects (metal pipes and spheres mostly) trying to get it right. It matters. Sure, it still bangs if you do it a different way, but with far lower yield.

[Bored chemist]

Which claim am I accepting as wrong?

Plutonium for some reason requires spherical symmetry to get that proper detonation,

In principle the limit to the speed of a coil gun is relativity.
So, with enough money etc you can deliver the "last bit(s)" of the  critical mass quickly enough

Now imagine casting a block of plutonium that's the shape of a ball with a hole in it
Wrap it in something dense- it hardly matters what  Uranium might be a good choice)
Then use 2 coil guns to fire  two cylindrical plugs simultaneously into the two  ends of the hole through the ball. (You need to set this up in a vacuum chamber but... you need that for a "gravity" driven one too.)
When they get there you have an intact ball with more than a critical mass. The two plugs will compress eachother (axially) which helps the efficiency a bit.

If you can "build the ball" quickly enough, it goes bang.
So, notwithstanding what you said, you don't need spherical geometry for it to work.

Sure a sphere might get you more bang for your buck, but that's not the same as saying "Plutonium for some reason requires spherical symmetry".
It doesn't "need" it at all.


The "best" geometry for a gas container is spherical- but most of them are cylindrical.

[Halc]

If you can "build the ball" quickly enough, it goes bang.
So, notwithstanding what you said, you don't need spherical geometry for it to work.

I agree with this.  It can be done with a sphere and plugs of concrete if you do it like that, so yes, if you build it fast enough, it does indeed go bang. I lack the resources to simulate that and see how much energy is needed using that method to get an equivalent yield using the same amount of material as a modern device.

Sure a sphere might get you more bang for your buck

Your description in your prior post describes using a gun to quickly build a sphere. No, a sphere isn't necessary. A cube works fine if you build it fast enough, but it's stupid difficult to build a cube quickly, or to build a sphere with a gun for that matter.

but that's not the same as saying "Plutonium for some reason requires spherical symmetry".
It doesn't "need" it at all.

Your example was hypothetical, not at all practical. A practical weapon requires the symmetry I was talking about. I said 'proper detonation' and it would have been better to have said 'practical detonation'.  I could take one gram of plutonium shaped like a little Batman and if I shoot it at Moscow fast enough, Moscow will be obliterated, but it requires more resources than exist on Earth to shoot it that fast. So can I can claim that critical mass is not necessary?

The "best" geometry for a gas container is spherical- but most of them are cylindrical.

That's because a gas container needs to maximize the volume-pressure to cost ratio, not minimize the average distance between the gas molecules.

[Bored chemist]

stupid difficult to build a cube quickly, or to build a sphere with a gun for that matter.

So "stupid" that they more or less did it.

not minimize the average distance between the gas molecules.

Nobody said it was.
However, the "best"- most efficient shape for a gas vessel is a sphere.
But for practical reasons they don't always use that shape.

The "best" shape for an implosion is a sphere.
But - if you have rail guns, but no explosives, cylindrical symmetry is the best you will get.

I suspect that, with a fast enough gun you can actually get a better bang than compressing a solid sphere..
Imagine a long pipe with thick walls- wrapped with some ballast, and nearly critical, and imagine firing a cylindrical plug into it.
You can make the cylinder arbitrarily long.
If the plug arrives quickly enough, you can set off an arbitrarily large amount of fissile material.
I doubt it's ever practical