Simulation of Nuclear Weapon Secondary Explosion

Anonymous Coward writes "Here is what all those DOE supercomputers have been crunching: On April 30, the Crestone project team at the Laboratory successfully completed the first three-dimensional simulation of a nuclear weapon secondary explosion. The total processor time was 2.01 million hours... The details are at Los Alamos National Laboratory." The secondary explosion in today's modern weapons occurs when a fission device explodes and compresses a light isotope (often tritium) until it creates a fusion reaction. This increases the total yield by a factor of perhaps 100-1000.

That's just great

[11223]

That's just great - now does the cluster go back to Seti@Home crunching?

Personally, I think that they need to get their priorities straight. After we make contact with the little grey men with bug eyes, they'll give us all the computer technology we need to do these simulations.

at the bottom of the same page

[chowda]

HA! look for your self.. almost the last thing on the page...

Lost: set of house/car keys
Found: two black keys
Pool memberships available at prorated rates
Sportmen's Club hold Co-Ed Introductory Shooting class

Los Alamos?!?

[crovax]

Well if was done at Los Alamos then this is an achievment for all of mand kind ;-)
-----
If my facts are wrong then tell me. I don't mind.

2 million hours?

[Chairboy]

I don't get it, my new 3dfx card can render nuclear explosions at 66 fps! Maybe they should update their drivers...

: )

Results...

[puppet10]

Not to be released on the 11 o'clock news (hopefully).

Re:at the bottom of the same page

[YIAAL]

Good grief, it's really there! Ah, such a juxtaposition of the sublime and the quotidian. Or something like that.

Practical?

[No Such Agency]

So does this mean that there will never again be an excuse for nuclear weapons testing? If there's a complete computer simulation of an A-bomb or H-bomb, is it easy or hard to plug in different parameters such as "below 50m of rock of crushing constant x", or "under 100m seawater"? I guess it depends on wheither you'd have to run the whole thing again from scratch or not. I'm not really sure wheither this was a "model" they constructed, or an "algorithm for making future models which won't take so long to process"...

Maybe they should give this to the French military as a public service to the rest of the world.

Nukes

[HotThought]

Does this simulation simulate the thermal and radiative effects on a large scale as well, or simply what is going down on the atomic level? I would assume only the atomic level. -A

not as many physical tests

[peter303]

As in airplane design, you don't need to build every model and test them, but you have to do some of them to verify the computations.

How much faster would it have been....

[whimmel]

..if they used a new dual-CPU G4 PowerMac.

Benefits of doing this simulation

[gyozadude]

We can acquire much beneficial information from doing this 3D thermo-nuclear simulation. A solution to a secondary explosion shares a lot of similarities to the implosion of a nuclear fusion pellet as proposed for future Inertial Confinement Fusion (ICF) power reactors. The fundamental physics such as radiation, electromagnetism, and rapid condensation are probably identical. Since the National Ignition Facility (NIF) [a big laser at Livermore designed to ignite a fusion pellet] is still under construction, there isn't a lot of ICF pellet data that folks can use to verify the computational models. However, there certainly should be lots of data that can be compared against for bomb physics. The resulting benefit for all human kind is that we can do less testing and more computer modelling if the models prove they can simulate the physics correctly. Many of the models can also apply to less extreme conditions such as next-gen propulsion systems, hypersonic structural design and atmospheric re-entry. I guess I'd try to educate myself more about the fundamental science and benefits before I'd pass cynical judgment upon the use of this much CPU/hardware. Our future, non-CO2 producing power supply may depend on this research.

Re:How much faster would it have been....

[Valar]

Looks like someone is proud of their ability to add negatives (Program+G4=-Program). G4s might be resonable fast if run without that slothy OS layer. But for most purposes, they run about as fast as a p3. Also, real supercomputers have faster memory paging, and MUCH more bus bandwidth, which in the long term becomes very important for this kind of computation. p.s.-- please excuse spelling/grammar, i'm operating on like 3 hours of sleep)

Re:Nukes

[Detritus]

I think it is on a large (non-atomic) scale. You have soft x-rays from the primary flowing down the radiation channel, ablating materials, being reflected/absorbed, compressing the secondary, which has a Pu-239 fission "sparkplug" in its center. During all of this, the materials in the secondary and bomb jacket are being transmuted to other elements. There are a lot of complicated things going on in a very short amount of time. I would love to see a simulation of the process, but then they would have to shoot me :-).

Re:Practical?

[tcomeau]

So does this mean that there will never again be an excuse for nuclear weapons testing?

That's the hope, if not the plan. This simulation is described as part of the "stockpile stewardship" program. Instead of grabbing a warhead off the shelf or out of the silo, sticking it down the hole, and setting it off to see if the expected shelf life matches reality, they fiddle with the initial conditions and "set off" a virtual device.

DOE has in fact discussed sharing the software, and perhaps access to the hardware, with the French, though I haven't seen any press on this recently.

There is still a case where you would need to light up a new device: If somebody came up with a new, safer, approach to weapons design. But that hasn't happened in quite some time. Most recent testing (both ours and others) has been around safety issues: Will the device that has been sitting quietly for a decade or three still behave as you expect?

Or so I'm told.

tc>

Secondary Explosion

[Anonymous Coward]

The secondary explosion in today's modern weapons occurs when a fission device explodes and compresses a light isotope (often tritium) until it creates a fusion reaction. This increases the total yield by a factor of perhaps 100-1000.

The fusion reaction does not dramatically multiply the yield by itself, but instead serves to recompress the core causing a subsequent, much more dramatic fission explosion. A thermonuclear device undergoes a fission, fusion, fission sequence. Its not clear from the article just what was simulated here.

Very cool

[AshPattern]

The guy in the front in the picture looks like Gordan Freeman.