LYING in pieces around a Denver machine shop is what looks like a giant piece
of sewer pipe with an oblong loop at one end. In reality, it is a powerful new
type of engine and an environmentalist鈥檚 dream鈥攜et also a throwback to the
early Industrial Revolution.
Much of this invention鈥檚 power lies in its simplicity. The design,
developed over the past three years by scientists at Los Alamos National
Laboratory in New Mexico, is the offspring of an odd union. Half its
intellectual genes come from the Stirling engine: a 180-year-old invention that
can, in theory, be made quieter and cleaner than its internal-combustion
cousins, but has so far been tricky to produce at a competitive price. Its other
parent is the thermoacoustic engine, a modern device that uses heat to make
sound waves, but is notoriously inefficient.
Their singular offspring promises to combine the advantages of both and may
be the first commercially viable engine with no moving parts. As early as next
month, the machine now being completed in Denver will begin to liquefy natural
gas in a remote location where conventional fuels are difficult and expensive to
supply. Soon afterwards, other versions might be pumping water or making
electricity in developing nations. Within a decade, you might have one making
both hot water and electricity in your home.
Advertisement
The inspiration for this miraculous device comes from Robert Stirling, a
19th-century Scottish clergyman, and was born of his concern for the workmen in
his flock. The steam engines that powered the industrial revolution use a fire
under a boiler to heat water and build up a head of high-pressure steam. As the
steam is released, it pushes pistons which in turn drive crankshafts.
But the boilers could not always contain these high pressures and they
occasionally exploded, scalding, maiming or killing workers unlucky enough to be
nearby. Stirling was disturbed by this flaw. In 1816, when he was 26 years old,
he conceived and patented a solution that did away with steam and high
pressures.
In one of its simplest forms, Stirling鈥檚 engine consists of a hollow cylinder
filled with air, with a piston at each end. These pistons are linked together
and in front of each piston there is a heat exchanger. The heat exchanger at one
end鈥攖he hot end鈥攊s used to heat the air, while the other鈥攁t
the cold end鈥攊s used to cool the air.
Heating the hot end expands the air there, driving the nearest piston
outwards. The two pistons then work in unison, sloshing the air to the
cylinder鈥檚 cooler end, where heat is extracted. Finally, the piston at the cold
end compresses the air and the cycle begins again. The hot expanding gas does
more than enough work to drive the linked pistons.
This process can be made more efficient by extracting the heat from the air
before it reaches the cold end and replacing the heat when it returns. This is
done by passing the gas through a porous material with a high heat capacity.
This device, Stirling鈥檚 ingenious creation, is known as a regenerator.
The regenerator insulates the two sides, storing heat from hot gas passing
one way so that the cold gas can pick it up when it returns later in the cycle.
Without the regenerator, the temperatures at the two ends would be nearly equal,
drastically cutting the efficiency of the engine.
Unfortunately, Stirling鈥檚 motors weren鈥檛 as tough as the steam engine, and
the design might have been left as a footnote in engineering history if it
hadn鈥檛 been for Nicolas Carnot, an engineer in Napoleon鈥檚 army. A few years
after Stirling filed his patent, Carnot concocted a mathematical theory that
could be used to calculate how efficient an ideal power plant would be at
converting heat into work. This led to the discovery that Stirling鈥檚 design
could鈥攁t least in theory鈥攂e made to yield nearly perfect
efficiency.
For almost two centuries, scientists have been seduced by the Stirling
engine鈥檚 potential, building prototype after prototype. But they have always
been disappointed. The pistons must fit tight enough to prevent gases escaping
between the piston鈥檚 edges and the cylinder wall. In an internal combustion
engine, a small amount of 鈥渂low-by鈥� can be tolerated, but if it occurs in a
Stirling engine, efficiency plummets. Yet the piston鈥檚 edges must not touch the
cylinder wall; if they do, the edges will wear and leak. The high-tolerance
machining this requires makes Stirling鈥檚 model too expensive to compete with the
internal combustion engine.
Steady rhythms
Among the modern scientists who find Stirling鈥檚 idea maddeningly fascinating
is Peter Ceperley at George Mason University in Virginia. Ceperley is an expert
in wave motion, and that gives him an unusual perspective on the Stirling
engine. In 1979 he realised that, whereas internal combustion engines deliver
their power in brief explosions, a Stirling motor鈥檚 action is much smoother. In
the steady rhythms of its internal motions and pressures, it mimics a travelling
wave.
It dawned on Ceperley that waves鈥攏ot of liquid, but of
sound鈥攃ould form the basis for a Stirling engine. There would be no need
for any troublesome moving parts, because the sound wave would move the gas back
and forth. Gas pistons would replace the metal pistons of ordinary engines and
reproduce in miniature the Stirling cycle of expand, slosh, compress, slosh. So
Ceperley set about constructing acoustic engines from tubing and heat
exchangers. As in the mechanical Stirling engine, the regenerator maintains a
smooth temperature gradient between the hot and cold exchangers.
So how does the engine work? Think of a small blob of gas inside the
regenerator moving back and forth with the sound wave. The temperature gradient
in the regenerator increases the motion of the blob by successively cooling and
warming it. This exaggerates its contraction and expansion, amplifying the wave.
But the idea鈥檚 bang was soon reduced to a whimper. Ceperley鈥檚 prototypes were
plagued by inexplicable energy losses that left them unable to yield any net
power output. Ceperley was stumped.
That鈥檚 when Greg Swift joined the quest. A specialist in low-temperature
physics, he arrived at Los Alamos in 1981 and was intrigued by Ceperley鈥檚 idea,
but soon fell into line with the conventional view that a travelling-wave engine
was impractical.
So Swift and his colleagues spent most of the next 16 years designing and
testing refrigerators that use acoustic engines powered by standing waves.
Unlike a travelling wave, a standing wave stays in one place, oscillating like a
plucked guitar string. The standing-wave engines did produce some net power
output, unlike Ceperley鈥檚, but they were still not efficient enough to be
cost-effective. Nevertheless, Swift and his colleagues were laying the
groundwork that would lead them back to Ceperley鈥檚 concept, and their present
revolutionary design.
For their work in standing-wave engines, the Los Alamos team built a computer
model that showed them how pressure oscillations behave in a closed system. As
the model evolved, the group tested an array of geometries for the loops and
tubing of their engines to determine which would yield greatest efficiency.
The result surprised them. In a standing wave, the peaks of pressure and gas
velocity occur at different times in the cycle. But their model told them that
it would be far more efficient to have pressure and gas velocity in
phase鈥攚hich is exactly what happens in a travelling wave. By flaring and
constricting the tubing at strategic points, the team coaxed each wave鈥檚
pressure and velocity to crest at precisely the same moment. In other words,
they converted the standing waves into travelling waves, and in the process
boosted the engine鈥檚 efficiency dramatically. After almost 20 years, Ceperley
had been proved right: travelling waves could be the basis of an efficient
engine after all. But a few tricks were needed to make it work.
In the new Los Alamos engine, the piping is filled with helium, pressurised
to 30 atmospheres. The hot heat exchanger is held at temperatures up to 700
掳C, the cold one at about 20 掳C. In between is the regenerator, a stack
of about 800 sheets of stainless steel mesh with threads 65 micrometres in
diameter and openings roughly 170 micrometres square.
The loop is attached to a resonator, a piece of steel tubing 4.5-metres long
which acts like an organ pipe, tuning the sound to a specific pitch. The Los
Alamos team chose a length that gives them 80 oscillations per second. It鈥檚
important to keep the frequency low, because the higher it is, the faster gas
oscillates back and forth and the greater the frictional losses.
And there is one more magic ingredient. The Los Alamos model identified the
mysterious energy loss that had dogged Ceperley鈥檚 attempts: the sound waves were
driving a gale of gas around the loop which chilled the hot heat exchanger,
slashing the engine鈥檚 efficiency.
Once they saw the problem, the engineers found a way to eliminate it. They
inserted a 鈥渏et pump鈥濃�攁 disc cut with precisely tapered slots which are
wide at one end and narrow at the other. As a sound wave passes through, it
builds up a back-pressure that neutralises the wind.
When they put all this together, the team found that they had boosted the
engine鈥檚 average efficiency to 30 per cent鈥攁bout the same as an internal
combustion engine, and far better than any other acoustic engine. But they don鈥檛
see their creation as replacing internal combustion engines. For one thing, it
can鈥檛 accelerate or decelerate quickly. Also, says Swift, 鈥渧ehicle propulsion
requires rotating shafts, so the sensible thing is to use an engine that has a
crankshaft. Our religion is to get rid of moving parts.鈥�
So what can you do with a machine that turns heat into sound? Sound waves
can鈥檛 turn a crankshaft, but they can drive a microphone to generate
electricity, or separate impurities from air, or even separate air into its
constituent gases.
And the engine can be run backwards, using sound to pump heat from the cold
end to the hot, cooling without using ozone-eating chemicals. There may be
thermoacoustic air conditioning in your car one day, and acoustic Stirling
engines might be ideal for cooling space-based telescopes and detectors, which
now use liquid helium that runs out in a few years
(快猫短视频, 8 March 1997, p 32).
But the first fruits of this work will come from liquefying natural gas. In
remote locations where oil is pumped, such as offshore or in deserts, the
natural gas that surfaces along with the oil is often burnt away. There is no
practical alternative: it would be too expensive to build a pipeline, or to
provide sufficient electric power to cool and compress the gas, so that it can
be shipped out as a liquid.
In the new year, at a well still to be chosen, Chart Industries of Denver
will join two 12-metre-high versions of the Los Alamos engine together. These
will work together to burn natural gas and generate intense sound waves, which
will produce some 2000 litres of liquified natural gas a day. 鈥淭he engines will
be linked so their vibrations cancel,鈥� says Scott Backhaus, who works with Swift
at Los Alamos.
Fine-tuning
While that test unfolds, Swift and Backhaus are asking their modelling
software 鈥渨hat if?鈥� They have calculated the energy losses caused by individual
components of their engine and are plotting necessary adjustments to boost
efficiency.
鈥淭here are two main components that are bogging us down,鈥� Backhaus says. The
first is the corners in the loop of tubing. 鈥淲hen a flow turns a sharp corner,
it doesn鈥檛 want to stay close to the wall, and that causes a lot of turbulence,鈥�
he says. 鈥淲e can model that and change the interior shape of the tubing to
minimise it.鈥�
The heat exchangers can be improved, too. The team calculates that the most
efficient heat transfer would be provided by a stack of metal plates, each
separated from the next by less than 30 micrometres. This would generate less
drag on the gas than the mesh screens. But the gaps between the plates must vary
by less than 10 per cent over their 20-centimetre breadth. A tricky bit of
engineering.
With refinements like that, the engine鈥檚 efficiency could eventually reach 38
to 40 per cent鈥攔oughly the same as today鈥檚 best car engines. Then the Los
Alamos power plant could become a cost-effective home power station鈥攁
small engine tucked away in the basement, generating both heat and electrical
power. Swift鈥檚 team and a New York engineering company are planning a 1-kilowatt
device that, if scaled up, could meet the energy needs of most households.
鈥淚t鈥檚 a compelling use for Stirling engines, especially one as simple as the
Los Alamos version seems to be,鈥� says Brent van Arsdell, president of the
American Stirling Company in Wichita, Kansas. 鈥淏ecause the same energy does two
things, the net efficiency of the system can be very good, even if the
efficiency of the engine itself isn鈥檛 anything special.鈥� Its biggest advantage
is that you can burn anything to drive the engine鈥攏atural gas, oil,
methane, wood or even household rubbish. Some engineers predict it will
eventually be used for everything from hybrid electric vehicles to gathering
solar power鈥攃onverting sunshine to electricity more efficiently than
today鈥檚 photovoltaic cells.
鈥淎 few people are talking to us about distributed power generation,鈥� says
Backhaus. That is, making electricity not at giant, centralised stations but at
a much larger number of smaller installations. 鈥淵ou could heat the engines with
solar energy, setting them out in the desert, and they could each produce 10 or
20 kilowatts of electricity,鈥� he says. 鈥淵ou also could heat them by burning the
methane seeping out of waste dumps.鈥�
According to Steven Garrett, professor of acoustics at Pennsylvania State
University and a consultant to the Los Alamos group, 鈥淭here are uses for this
that have yet to be imagined.鈥�
