快猫短视频

FOR as long as most of us can remember, the dream has been there: unlimited,
clean energy from nuclear fusion. No greenhouse gases to worry about and
relatively little radioactive waste. Better still, the fuel for fusion comes
ultimately from sea water鈥攕o no more reliance on Middle East oil or
drilling in the pristine wilderness. The prize is certainly beguiling.

All it takes is some way to make a simple mix of hydrogen isotopes hot enough
for their nuclei to fuse and yield colossal amounts of energy. Yet for almost
half a century, nature has played a game of peekaboo with fusion physicists,
fooling them into thinking they were nearing their goal and making them the butt
of a cruel joke. 鈥淟imitless fusion power is just 40 years away鈥攁nd it
always will be.鈥�

Now a team of researchers at the Joint European Torus (JET) project at
Culham, Oxfordshire鈥攖he international test bed for fusion
machines鈥攎ay have broken the deadlock. At last October鈥檚 meeting of the
American Physical Society in Long Beach, California, they unveiled data showing
that it really is possible to create and hang on to the 100 million 掳C
temperatures needed to bring the power source of the stars down to Earth.

This is a promising development for fusion researchers, whose morale hit a
low point in 1998, following the demise of ITER, the International Thermonuclear
Experimental Reactor, a $10 billion project to build a fusion machine the
size of a 10-storey office block. Calculations by fusion theorists in the US had
raised grave doubts over claims that ITER would achieve ignition. Those doubts,
plus the hefty price tag, led Congress to pull the plug on American
involvement鈥攁nd thus on any hope of building the giant machine.

The death of ITER sparked much hand-wringing among physicists, some of whom
were seriously starting to doubt whether nuclear fusion could ever be tamed.
Perhaps it was to remain the preserve of stars alone. A back-of-the-envelope
calculation hardly gives cause for optimism. From the outside, stars look like
blazing powerhouses, but get close in to a star like the Sun and you find that
it packs barely a watt of fusion power per cubic metre of volume. A viable
fusion reactor must cram in around a million times as much power鈥攏o mean
feat. The more theorists explored the dauntingly complex physics of the plasma
in machines like JET, the less they liked what they saw.

One troublesome question is just what size a fusion reactor should be. Big
machines are better for retaining the heat, simply because being bigger means it
takes longer for the heat to escape. They also give better performance in terms
of the amount of fusion power you get out compared to the heating power you have
to put in.

So bring on the humungous fusion machines, then? Well, not so
fast鈥攖here鈥檚 more to a practical device than just triggering fusion. Any
workable power plant must produce steady amounts of fusion power for years on
end, which means the reactor must be able to withstand years of being blasted by
neutrons from the fusion reactions, and temperatures of 100 million 掳C
(see Diagram).
And all those problems get worse with bigger
machines鈥攂asically because there鈥檚 more fusion power blasting each square
metre of the machine鈥檚 walls.

Build too small a machine, and you can鈥檛 hang on to the high temperatures
long enough. But build one too big, and you run the risk of the plasma trashing
your reactor in no time. Finding the optimal size meant putting a lot of faith
in 鈥渟caling laws鈥濃�攔ules relating size to performance extracted from past
fusion experiments the world over.

And that began to look like a very dodgy strategy. For years fusion
scientists had used these scaling laws as comfort blankets, clinging to their
predictions of how the ultimate fusion machines would behave. 鈥淭he trouble is,
there wasn鈥檛 much real physics in them,鈥� says Steve Cowley of Imperial College,
London. 鈥淭hey were basically just graphs of best-fit lines to data. It gave the
impression that fusion research wasn鈥檛 real science.鈥� Putting the scaling laws
on a more solid foundation was impossible without tackling the daunting
challenge of turbulence鈥攐nce described by Einstein himself as the hardest
problem in the whole of classical physics.

The first detailed model of turbulence in machines like JET and ITER was
dreamed up by William Dorland and Michael Kotschenreuther of the Institute for
Fusion Studies at the University of Texas. The outcome could hardly have been
worse. The model suggested that turbulence-induced heat loss from the plasma in
ITER made ignition unlikely. With some of ITER鈥檚 supporters claiming they were
鈥�99.5 per cent鈥� certain the machine would reach ignition, Dorland and
Kotschenreuther鈥檚 model was as welcome as a ham sandwich at a bar mitzvah.

Some fusion researchers still dismiss the model as fundamentally flawed. But
many others began to have a crisis of confidence about the whole future of
fusion research. Outside the fusion community, some physicists waded in with
finger-wagging criticism of a project on which billions had been spent to no
obvious benefit.

Yet even as Dorland and Kotschenreuther were doing their controversial
calculations, there were glimmerings of a get-out鈥攁 way in which even a
turbulent plasma might hang on to its heat long enough to achieve ignition. It
centred on effects left out of their sums that are now the hottest topic in
fusion research, and making even some hardened sceptics sit up and take notice.
Their promise is belied by their humdrum name: transport barriers.

Creating 100-million 掳C plasma is one thing; keeping it hot is another.
Turbulence within the super-hot plasma has a nasty habit of transporting the
heat out as fast as colossal electric currents and particle beams can shovel it
in. But if you can create a region of low turbulence within the plasma, it acts
like a scarf around a hot-water pipe鈥攁nd the heat stops pouring out of the
machine.

The existence of such heat transport barriers came as a complete surprise to
fusion physicists. 鈥淣o one predicted them,鈥� says Cowley. 鈥淏ut we鈥檙e certainly
all glad that they exist.鈥�

The first transport barrier to reveal itself is known as the
鈥渉igh-confinement mode鈥�, or H-mode, which traps heat at the edge of the plasma.
Experiments at various fusion laboratories have shown that when the amount of
power trapped in the plasma exceeds a threshold, a region of low turbulence
suddenly appears around the edge of the plasma. Exactly why H-mode occurs is
still not fully understood, but its effect is dramatic, doubling the amount of
time the machine can sustain fusion temperatures.

Recent experiments at JET have shown that H-mode isn鈥檛 stable, but repeatedly
collapses, zapping the walls of the machine with huge heat loads. It鈥檚 not all
bad news, though: these sudden releases of energy also allow impurities,
including fusion-killing helium 鈥渁sh鈥�, to escape.

The trick to keeping fusion alive lies in a balancing act, using H-modes to
keep the heat in while allowing their collapse to clean up the plasma.
Researchers at JET have now perfected the trick by 鈥渢ickling鈥� the plasma with
judicious amounts of magnetic and radio-frequency energy. They have also come up
with a way to combat the heat-load problem, using squirts of inert gas that
spread the energy trapped in the edge of the plasma over a wider area, reducing
wear and tear on the machine walls.

But this is not the only good news to come the way of the long-suffering
fusion community. At October鈥檚 meeting of the American Physical Society, Joelle
Mailloux, one of the researchers at JET, presented the best evidence yet for
even more powerful heat-trapping effects鈥攁nd ones that occur in the bulk
of the plasma, not just at the edge: internal transport barriers.

These ITBs are now expected to play a crucial role in the success of any
future power-producing reactors. By carefully controlling conditions inside the
machine, it鈥檚 possible to put a twist in the magnetic field that dramatically
cuts the rate of heat loss. 鈥淭hat gives higher temperatures and plasma densities
even than H-mode,鈥� says Mailloux. 鈥淎nd that should lead to smaller and cheaper
谤别补肠迟辞谤蝉.鈥�

At the APS meeting, Mailloux and her team unveiled the spectacular
improvement in JET鈥檚 performance made possible by ITBs. Their heat-sealing
effect raised the temperature at the centre of the plasma to over 300 million
掳C, well over double the level needed to trigger sustained fusion. It also
increased tenfold the volume of plasma at or above fusion-level
temperatures鈥攁 breathtaking improvement in performance.

But ITBs are more than just supercharged versions of H-mode. 鈥淭heir real
advantage over H-mode is that they produce a large amount of the current in the
plasma,鈥� says Mailloux, 鈥淭hat means much less current has to be supplied from
outside to keep the plasma confined.鈥�

Huge temperatures, bigger plasmas, lower power demands鈥攊s nature
finally giving fusion scientists a break? It sounds too good to be true, and
Mailloux is the first to admit that it鈥檚 hardly plain sailing from here. 鈥淚TBs
are fragile, and they need a large initial input of power,鈥� she says. 鈥淲e don鈥檛
yet have scaling laws for them either.鈥�

Such scaling laws, crucial in creating a case for building power-generating
machines, are expected to benefit from the other source of growing optimism
among fusion scientists: the surge in new theoretical insights. Phenomena like
turbulence are appallingly non-linear, meaning that small effects can turn out
to have big consequences. But solving such equations was simply impossible,
given the limited computer power available in the past. Theorists have had to
work with simplistic models of fusion physics just so they could have equations
they could solve. That doesn鈥檛 mean turbulence can be ignored鈥攁s the
failure of theorists to predict heat-trapping effects shows.

Nowadays, effects such as H-mode and ITBs are being probed using powerful
computers to see if they can be exploited on future commercial fusion power
machines. Theorists like Cowley are also using computer models to study the
potential of different types of fusion machines. Until now, most attention has
focused on tokamaks, the doughnut-shaped 鈥渕agnetic bottle鈥� design conceived by
Soviet scientist Andrei Sakharov and colleagues in the early 1950s.

But these tokamaks aren鈥檛 the only game in town. Back in the 1970s, some
theorists sketched out a modified design they claimed could make fusion easier
to achieve. This is the spherical tokamak鈥攕haped more like a cored apple
than a doughnut鈥攁nd it can cram more plasma into a smaller space, and make
better use of the magnetic fields used to control it. Potentially, this makes it
far more energy-efficient than conventional tokamaks.

That efficiency is described by a number called 鈥渂eta鈥�, which is the ratio of
the heat energy in the plasma to the magnetic field energy needed to hang on to
the plasma. The higher the beta, the more efficient the machine鈥攁nd the
better the prospects of producing economically viable levels of fusion
power.

In 1991, another team at Culham lashed together a prototype from spare parts
lying around the lab. Known as START, it looked like a standard tokamak that had
been squeezed together in the hands of a giant. But by the time it was shut down
in 1998, START had stunned everyone by achieving a beta value of 0.4鈥攎ore
than 10 times as efficient as JET, and still a world record.

Now a successor has been built at Culham. Called MAST (Mega-Amp Spherical
Tokamak), it is already the talk of the fusion community. Using sophisticated
computer techniques, theorists have discovered that the conditions that make
heat-trapping ITBs possible in parts of the plasma in ordinary tokamaks can
exist right through a spherical tokamak鈥攎aking it a good bet for achieving
ignition.

While stressing that it is still early days, even the sceptics are
enthusiastic about spherical tokamaks. 鈥淢AST is a tremendously exciting
experiment,鈥� says Dorland, now at Imperial College, London, the theorist whose
calculations cast doubt on the original ITER project. 鈥淪pherical tokamaks have a
very exciting role to play in fusion.鈥�

And there is good news for ITER. The project is now being revived, although
with a smaller, more sophisticated machine that will be half the cost of the
original leviathan. Theorists are confident that the heat-trapping effects seen
in smaller machines like JET will appear in ITER, which is predicted to generate
around 500 megawatts of power from just a few tens of megawatts of external
heating.

Last November, the first talks were held between the international partners
still involved in the project to decide where to build the scaled-down ITER
machine. There is even talk of the US rejoining the project, which is benefiting
from the renewed sense of optimism about fusion鈥攅ven among former
doubters. They include Richard Hazeltine of the University of Texas, Austin. 鈥淚
was a sceptic about the claims being made a few years ago, but I now share the
optimism,鈥� he says. 鈥淎s well as these heat-trapping effects, we鈥檙e beginning to
understand what鈥檚 happening in fusion. The picture has really changed.鈥�

Advocates of the leaner, fitter ITER admit that getting the backing of the
American fusion community is only part of the task they face. The question now
is whether politicians in the US and elsewhere can be persuaded to forget all
the years of hype and disappointment, and put their faith鈥攁nd yet more
billions of dollars鈥攊nto the quest for this ultimate power source.