360 lines
18 KiB
Plaintext
360 lines
18 KiB
Plaintext
This note is currently being revised in the light of new information
|
|
supplied by Lindl's ICF paper. 24/11/1995
|
|
|
|
TELLER-ULAM CONSTRUCTION
|
|
|
|
"... it is my judgement in these things that when you see something that
|
|
is technically sweet you go ahead and do it and you argue about what to
|
|
do about it only after you have had your technical success. That is the
|
|
way it was with the atomic bomb. I do not think anyone opposed making it;
|
|
there were some debates about what to do with it after it was made."
|
|
|
|
Robert J. Oppenheimer
|
|
on the H-bomb
|
|
|
|
"Don't bother me with your conscientious scruples. After all, the thing's
|
|
superb physics."
|
|
|
|
Enrico Fermi on the H-bomb
|
|
|
|
The basic problem of the H-bomb is to use the energy and particles
|
|
released in a fission device to firstly compress and secondly heat
|
|
a mass of fusion fuel. Fusion can only occur under temperatures,
|
|
pressures, and densities at, or exceeding, those found at the centre
|
|
of the sun. The latter is the case for a H-bomb since the reactions
|
|
in the bomb occur on a much shorter scale than those in the sun.
|
|
|
|
You have to have extremely fast moving nuclei to overcome
|
|
electrostatic repulsion of the positive proton charges. You need
|
|
about 1 trillion atmospheres (8,000,000,000 tonnes/square inch) or
|
|
about 1 million megabars. This leads to extremely densely packed
|
|
atoms and molecules, which increases the likelihood and frequency
|
|
(rate) of collisions. High compactification of fissile material
|
|
also reduces the mean free path of fast neutrons. To achieve these
|
|
goals, you have to configure the secondary just right. The Teller-
|
|
Ulam multistage configuration does precisely this. It is thought
|
|
that three main concepts are involved in this design.
|
|
|
|
You should think of a H-bomb as a multistage engine, with 3 explosive
|
|
stages. Since the explosions occur so quickly, it seems like only
|
|
one flash occurs, whereas 3 actually do. These correspond to the
|
|
initial fission of the primary, the fusion of the secondary, and the
|
|
fission of the casing or fusion tamper. In the case of a neutron
|
|
bomb, the casing may be made out of a non-fissionable material like
|
|
lead, so you would only get two explosions.
|
|
|
|
Separation of Stages
|
|
|
|
Much detail as to what goes in inside a H-bomb was gained in 1954
|
|
during the Ivy Mike fallout. By a careful analysis of the fallout
|
|
products, you could work out roughly where the energy came from.
|
|
In particular, you looked at the ratio of higher Z radioisotopes in
|
|
the fallout. You tried to find evidence as to whether these products
|
|
had been exposed to unusually high neutron fluxes. Compression of
|
|
the U-235 sparkplug in the secondary would increase the probability of
|
|
multiple neutron exposure. Hence the formation of elements like
|
|
transuranic Einsteinium and Fermium, which were first detected in the
|
|
Ivy Mike fallout. See the references for evidence of massive Li6D compression
|
|
and multiple neutron exposure.
|
|
|
|
The British designed their first H-bomb after examining American supplied
|
|
Russian fallout from the Joe-4 test.
|
|
|
|
Around 50% of the H-bomb energy comes from fusion. The other 50% is
|
|
from fission of the U-238 fusion capsule tamper or weapons casing.
|
|
The fusion-boosted implosion core just serves as a trigger, and gives
|
|
at most a few hundred kT of energy. Ted Taylor has done calculations
|
|
showing it is possible to get into the megaton range for extremely
|
|
efficient fusion-boosted imploders. Tritium gas is injected into the
|
|
core during implosion to achieve boosting.
|
|
|
|
For a given volume of Pu or U, you would find an equivalent volume of
|
|
Li6D to be 25 times less massive, due to differing densities. If you
|
|
fused this amount of Li6D, you would get 3 times as much energy as
|
|
you would fissioning the equivalent amount of Pu or U, taking into
|
|
account the energy released per reaction. Note that although a single
|
|
fission releases more energy than a single fusion event, the fission
|
|
releases the binding energy of 235 nucleons, whereas the fusion does
|
|
the same for five or six nucleons. If you had 235/6 = 40 fusions, you
|
|
would release more energy overall than fission of 235 nucleons. In a H-bomb
|
|
it follows you need about 10x the volume of Li6D than Pu or U, to
|
|
achieve a 50% energy release ratio. In other words, H-bombs have
|
|
a small mass of U or Pu, and a much larger mass of Li6D. In a reaction,
|
|
100% of the material never fuses. With experience, 10% is an outstanding
|
|
result. For a beginner, 1% is a good start.
|
|
|
|
The Failed Classical Super Design
|
|
|
|
Historically, the first theoretical designs for a H-bomb began with the
|
|
classical Super. This was a boosted trigger surrounded by a mass of fusion
|
|
fuel. When the trigger went off, the heat and shockwave were supposed to
|
|
set off an outwardly propagating thermonuclear reaction in the fusion
|
|
material. This didn't work. Calculations by Ulam and von Neumann showed
|
|
that temperatures and pressures weren't high enough to sustain such a
|
|
reaction. It would 'fizzle'. The design was based on what happens in a
|
|
supernova. Here, when material collapses into a neutron star, there is
|
|
an amount of 'bouncing' off the core. When the material is reflected, a
|
|
chain thermonuclear fusion reaction is set off, releasing a good percentage
|
|
of that ever fused by the star over its lifetime.
|
|
|
|
A new idea was called for. This is where Teller, Ulam, and de Hoffmann came
|
|
in. Rough calculations showed that sustained fusion could occur if the Li6D
|
|
mass was separated from the trigger, possibly in the form of a concentric
|
|
cylinder, surrounding a U-235 sparkplug, and surrounded itself by a U-238
|
|
pusher. An ablation layer made up of a low-Z hydride surrounds this pusher.
|
|
It is possible that primary and secondary are at two foci of an ellipsoid.
|
|
|
|
The main unknowns to the public are currently the design of the casing,
|
|
and the shape and size of the secondary, relative to the primary.
|
|
|
|
Compression
|
|
|
|
The problem then is to transfer the energy from the implosion to
|
|
this Li6D cylinder, firstly compressing it, and then heating it.
|
|
Compression must precede heating since hot materials tend to expand
|
|
more than cold ones. This energy transfer is the crucial idea in
|
|
a H-bomb. You must compress the Li6D in under a shake, or else the
|
|
expanding bomb debris will take everything apart before fusion has
|
|
substantially gone underway.
|
|
|
|
The Greenhouse George test showed that a small quantity of D-T could
|
|
be ignited by a fission device.
|
|
|
|
Radiation Coupled Implosion
|
|
|
|
Ed Teller has stated that the transfer of energy from the primary to
|
|
the secondary is primarily via radiation in the form of soft X-rays,
|
|
which travel at light speed. X-rays released by the trigger travel across
|
|
the air gap separating the casing from the trigger, and strike the
|
|
heavy (high-Z) bomb casing. Radiation pressure generated by the X-rays
|
|
is decoupled from the fluid pressure of the fission fragments, which travel
|
|
much more slowly.
|
|
|
|
We can learn a lot from Teller's statement. Mechanical (fluid) pressure isn't
|
|
the transfer mechanism. Nor are hard (MeV) X-rays straight from nuclear
|
|
reactions. Indeed, soft X-rays come from the ionization of a reasonably high-Z
|
|
material. The only place this high-Z material could be is the bomb casing,
|
|
which is responsible for most of the bomb's weight.
|
|
|
|
It is possible that a blackbody radiation mechanism is responsible
|
|
for the tamper implosion.
|
|
|
|
For a few millionths of a second, the insides of the bomb become like
|
|
a blackbody. Since the casing is so massive compared to the rest of
|
|
the components (including the secondary), it expands relatively
|
|
slowly. During the time the vaporised casing expands, a phenomenon known as
|
|
X-ray fluorescence causes the casing ions to generates secondary X-rays.
|
|
Since the casing atoms have been ionised, when the sea of electrons fall back
|
|
into their shells, a uniform emission of secondary soft X-rays is released.
|
|
If the casing is machined just right, it is possible to direct these
|
|
onto the secondary fuel mass from all directions, leading to a very even
|
|
compression. The X-rays act as a photon gas, which equilibriates at light
|
|
speed, much more quickly than a material gas made up of fission particles
|
|
would (this would equilibriates at the speed of sound). The problem of
|
|
the H-bomb is the calculation of the hydrodynamics, not the nuclear physics.
|
|
|
|
It doesn't have to be soft X-rays which cause the fluorescence. Anything with
|
|
enough kinetic energy will do the job - fission fragments or neutrons can do
|
|
it. All that needs to be done is to ionise the casing atoms.
|
|
|
|
What happens is that the secondary X-rays deposit their energy onto the
|
|
ablation layer almost instantaneously and uniformly from all sides. The
|
|
result is instantaneous heating. The surface layer of the fusion target
|
|
is vaporised, forming a surrounding plasma envelope. The layer undergoes
|
|
a blowoff with great force. This causes the inner part of the wrapper
|
|
to compress (Newton's 3rd law) due to rocket recoil. This tamper pushes against
|
|
the secondary Li6D fuel mass, and the mass is compressed to a fraction
|
|
of its original width. If there is an air gap (levitation) between tamper
|
|
and fuel, the tamper can develop more momentum to do the job. This is what
|
|
happens in the levitated cores of fission triggers.
|
|
|
|
Since the ablator is composed of low-Z, light material, the blowoff will
|
|
put a lot of energy into the expanding plasma. This prevents preheating of
|
|
the Li6D fusion fuel before adequate compression is achieved, while still
|
|
allowing for inward momentum coupling. In other words, the impulse is high.
|
|
|
|
By this time, the neutrons from the fission will have reached the sparkplug.
|
|
The fissioning sparkplug ignites the Li6D annular cylinder from the inside,
|
|
while compression occurs on the outside. Burning starts from the inner
|
|
edge of the Li6D and, in under 1 ns, a large fraction of the Li6D is ignited.
|
|
The core reaches 1000-10,000x the original density, igniting at 100 million
|
|
degrees C.
|
|
|
|
The high energy neutrons (> 1 MeV) released by fusion radiate out and
|
|
strike the U-238 atoms of the pusher and expanding casing, causing more
|
|
fission.
|
|
|
|
The casing acts as a heavy gas, whose inertia slows the expansion of the
|
|
explosion. However, it plays no part in confinement of the fusion fuel. The
|
|
compression caused by the imploding tamper does that job. The interatomic
|
|
forces between the casing atoms are negligible.
|
|
|
|
The bomb tamper is crucial in confining the reactions until they develop
|
|
appreciably.
|
|
|
|
To direct energy onto the secondary, you need firstly to
|
|
interact with the casing. All this happens in under 10 shakes.
|
|
|
|
In ICF, a typical fusion sphere consists of layers of: (1) Be or LiH ablator,
|
|
(2) a high Z polymer shield, (3) the main Li6D fuel, (4) the U-238 pusher,
|
|
(5) a void, and (6) a Li6D ignitor.
|
|
|
|
Note that it's not the fission trigger X-rays which cause the blowoff,
|
|
but the secondary X-rays due to the X-ray fluorescence of the high-Z
|
|
heavy bomb casing. The casing acts like a hohlraum target. Nothing is
|
|
reflected as such. Unlike visible light, which is coupled to optical
|
|
bandstates on the surface of metals, X-rays are absorbed due to their
|
|
much higher energy.
|
|
|
|
The X-rays come mainly from the L->K and M->K shell transitions as the
|
|
electrons drop down into the K shell vacancy, and hence lose energy.
|
|
|
|
Another possibility for an X-ray source is bremmstrahlung from deccelerating
|
|
electrons in the ionised plasma.
|
|
|
|
Eventually, the X-rays manage to diffuse through the expanding bomb casing,
|
|
and are released in a huge flux. This causes the initial light burst of a
|
|
nuclear explosion, and is responsible for immediate deaths. Considering this
|
|
light is 1000x brighter than the sun, this is no surprise! The temperature
|
|
soars to over 1000 deg C in microseconds.
|
|
|
|
The mechanism of a H-bomb bears an uncanny relation to indirect drive
|
|
ICF. Implosions driven by this method are relatively insensitive to the
|
|
nature of the primary beams (they could be lasers or ions just as well).
|
|
They are also hydrodynamically more stable. This is important, since the
|
|
fusion fuel mass must be compressed symmetrically and evenly.
|
|
|
|
X-ray - Plasma Interactions
|
|
|
|
This method tends to produce a large volume of target plasma through which
|
|
the X-rays must propagate, however. Although it would be more efficient if
|
|
the plasma were transparent to this radiation, it is not absolutely
|
|
necessary. A diffuse photon gas due to absorption, scattering, and re-
|
|
emission by the target plasma will do.
|
|
|
|
A number of physical effects must be considered. These include:
|
|
|
|
Absorption:
|
|
|
|
- X-ray absorption by target
|
|
- inverse bremsstrahlung (generates collisional low temp
|
|
electrons)
|
|
- parametric instabilities (bremsstrahlung induced
|
|
collisionless hot electrons)
|
|
- resonance absorption (collisionless hot electrons)
|
|
|
|
Hot electrons lead to target expansion, which is not good for compression,
|
|
for it takes more energy to compress a hot gas than a cold one.
|
|
|
|
Other undesirable effects include:
|
|
|
|
- stimulated Brillouin scattering
|
|
- stimulated Raman scattering
|
|
|
|
These also generate preheat and hot electrons in the target.
|
|
|
|
We also need to look at:
|
|
|
|
- thermal conduction (energy absorbed in a critical layer can be
|
|
inihibited from flowing into the ablation region)
|
|
|
|
Conversion Efficiences
|
|
|
|
For planar hohlraums, about 70-80% of the incident energy can be
|
|
converted into X-rays. You get better target coupling at short wavelengths.
|
|
|
|
Other Forms of Compression
|
|
|
|
Instead of radiation, could it be a material shockwave which does
|
|
the compression? Or a combination of both? It is known that at the
|
|
centre of the earth, iron is compressed to 30% its volume, subject to
|
|
about 5 Mbars. So we are way beyond the non-compressible regime, into
|
|
nonlinear effects. In fact, Ulam proposed using shock waves, but this
|
|
would have resulted in less even compression. Compression of the fusion
|
|
fuel can get as high as 1000x solid density, at 100 million degrees C.
|
|
|
|
Ulam is said to have come up with the solution to the energy transfer
|
|
problem when he was looking at ways to improve the efficiency of the
|
|
trigger. The joint Teller-Ulam paper talked about "hydrodynamic lenses
|
|
and radiation mirrors". Could there be some sort of lensing or baffle
|
|
system inside the hohlraum, which focusses radiation onto the Li6D via the
|
|
casing? I find this highly unlikely. Note that the shorter the wavelength,
|
|
the less refracted light gets. It is very hard to bend X-rays, let alone
|
|
gamma rays. Also, wouldn't the lens system vaporise before enough radiation
|
|
was focussed? "Hydrodynamic lenses" is reminiscent of the shaped charges
|
|
used in achieving a spherical shockwave in the trigger implosion.
|
|
|
|
Possible focussing systems include hohlraums shaped like ellipsoids, or
|
|
parabaloids with the primary at the focus. It is very difficult to shape
|
|
the secondary like a cylinder, and get a compression wave travelling just
|
|
before fast neutrons from the sparkplug cause fission - although not
|
|
impossible. Another problem with the cylindrical shape is that compressing
|
|
from the sides is like squeezing a tube of toothpaste. If the compression
|
|
is not fast enough, the material will squirt out the ends.
|
|
|
|
Laser fusion using X-rays to compress pellets of D-T fuel is used in
|
|
Livermore's NOVA. Ten pulsed lasers give a temperature of about 10^8 K, and
|
|
increase particle density by a factor of 10^3. Each pellet is smaller than
|
|
a grain of sand, and absorbs about 200kJ of energy in < 1 ns. Delivered
|
|
power is about 2 x 10^14 W, about 100 times the entire world's electric
|
|
power generating capacity. This is a peaceful example of inertial confinement
|
|
fusion.
|
|
|
|
Neutrons Causing Compression?
|
|
|
|
Neutrons expand out at a slightly greater rate as the fission fragments.
|
|
Can they compress the Li6D in time, before the fragments tear everything
|
|
apart? A shockwave is just a longitudinal compression of the propagation
|
|
medium. Energy is transferred in collisions between the atoms or molecules.
|
|
|
|
If this worked (a classical super design), then the most efficient
|
|
way to capture these fission neutrons would be to surround a fission
|
|
bomb with fusion fuel, and hope to cause an outward propagating shock wave.
|
|
If you didn't surround it, then you'd be wasting lots of neutrons.
|
|
The fact that H-bombs don't look like this (big, fat, and round) is evidence
|
|
against he idea.
|
|
|
|
Other Theories
|
|
|
|
From: merlin <merlin@neuro.usc.edu>
|
|
|
|
The basic idea is the primary is detonated -- neutrons escape in all
|
|
directions -- the secondary could be a hollowed out sphere of U-238
|
|
with a Li6D core -- though usually the secondary is elongated to hold
|
|
more Li6D. The neutrons convert Li6D to TD. They also cause fast
|
|
fissions in the U-238 wrapper around the Li6D -- these fast fissions
|
|
release an enormous amount of energy -- the energy causes the U-238
|
|
to expand (about 2/3 of energy causes expansion outward from center
|
|
of the sphere -- but about 1/3 of energy goes into inward compression
|
|
-- thereby compressing the TD core) -- the shock compression and
|
|
heating of the TD core reaches thermonuclear temperature and pressure
|
|
-- then a recursive reaction begins -- fast neutrons from the TD core
|
|
cause fast fissions in the U-238 wrapper -- fast fissions in the U-238
|
|
wrapper cause additional shock compression and heating of the core --
|
|
if optimum fusion temperature or pressure are exceeded the fusion
|
|
reaction slows down, fewer neutrons are produced, fewer fast fissions
|
|
occur, the U-238 wrapper releases some pressure -- until optimum
|
|
fusion temp and pressure is reached again and the recursive reaction
|
|
stabilizes (at least until you run out of TD to burn). This is why
|
|
in the traditional hydrogen bomb about half of the yield is fusion
|
|
and half of the yield is fission -- the energy has to be balanced in
|
|
order to hold the device together long enough to burn as much of the
|
|
TD fuel as possible. In the neutron bomb you get more waste tritium
|
|
because most of the U-238 mantle has been stripped away -- and the
|
|
device disassembles faster -- with much lower explosive yield.
|
|
|
|
The following diagram is adapted from Matt Kennel's <mbk@lyapunov.UCSD.EDU>:
|
|
|
|
-------------------------------------------------------
|
|
/ | |
|
|
/ oooooo |===========fusion fuel========================
|
|
| oa-bombo --fission spark plug---------------------------
|
|
\ oooooo |==============================================
|
|
\ | |
|
|
-------------------------------------------------------
|
|
<----------><---------------------------...>
|
|
implosion repetition of fusion cells clad in U-238 tampers
|
|
primary
|
|
|
|
1994
|