91色情片

SOMETIME late next year, American astronauts will open the doors of their
orbiting space shuttle and lob an 87-litre plastic water tank mounted in a
rectangular box out into space. The crew will then retreat to a safe distance
and hit the 鈥渙n鈥� button. Moments later, thrusters along the container鈥檚 sides
will fire up, making it pitch and toss like a buoy on a rough sea. Inside,
spherical slugs of water will slam against the walls of the tank, throwing the
craft into a series of uncontrolled lurches. The mission will last ten days. It
will be a rough ride, but when it is over, the craft鈥檚 designers hope no
spaceship will ever have to endure a trip like it again.

The craft鈥檚 official name is FLEVO, for Facility for Liquids Experimentation
and Validation in Orbit, but most people know it as Sloshsat. Its job is to get
to grips with 鈥渟losh鈥�, a phenomenon that has been the bane of rockets and
satellites for more than 40 years. Slosh is the way that liquids such as rocket
propellant move around inside their tanks in low gravity. It is a crucially
important factor in spacecraft design. But nobody really understands it.

To appreciate just how big a gap in our knowledge this is, think back two
years to NASA鈥檚 Near Earth Asteroid Rendezvous craft, which tumbled out of
control for more than 24 hours after its thrusters shut down. NASA scientists
blame that near disaster on slosh. Propellant slosh has also been blamed for
rockets stalling or exploding in mid-flight. That鈥檚 why scientists need to
understand it. But there is another motive: by learning to control slosh,
scientists could one day harness its energy to help them drive their
spacecraft.

Slosh is one of the occupational hazards of operating in low gravity. Unlike
fuel in a car on Earth, propellants inside a spacecraft refuse to settle
predictably to the bottom of the tank. Without gravity, the molecules coalesce
into slithering, gyrating spheres that can smash into the walls of the tank and
drastically change its centre of mass. The globs can also slip away from the
outlets running to the craft鈥檚 combustion chamber, creating the space equivalent
of air in the fuel line.

鈥淵ou鈥檝e seen a lava lamp with the liquid wax in there?鈥� says Andy Cheng, an
astrophysicist at Johns Hopkins University鈥檚 Applied Physics Laboratory in
Maryland. 鈥淭hat鈥檚 sort of what we鈥檙e talking about.鈥�

Cheng knows better than most the havoc that slosh can wreak. In 1998, he was
the lead researcher on the NEAR mission as it fired its engine for the final leg
of its three-year trip to the asteroid 433 Eros. Just moments into the burn,
NEAR鈥檚 liquid-fuelled engine shut down. The craft pointed its antenna Earthward
in a cry for help, overshot the manoeuvre and started tumbling out of control.
Thirty seconds later, spacecraft trackers at NASA鈥檚 Jet Propulsion Laboratory
reported that they鈥檇 lost radio contact altogether.

Later, NASA engineers deployed their best models to try to work out what had
happened. By their reckoning, the craft should have stabilised within 30 to 40
minutes. They were wrong. In fact, it tumbled for a nerve-racking 27 hours
before its radio signal finally reappeared on NASA鈥檚 Deep Space Network.
鈥淭here鈥檚 something wrong with the models and we don鈥檛 know what,鈥� says
Cheng.

Having come so close to calamity, NEAR controllers decided to navigate with
smaller jolts from the craft鈥檚 thrusters. The spacecraft finally reached orbit
around Eros in February, 13 months behind schedule.

No one is sure exactly what happened but Cheng suspects that NEAR almost
succumbed to some version of propellant slosh. So why did NASA鈥檚 predictions
fail so dismally? According to physicist Jan Vreeburg of the National Aerospace
Laboratory of the Netherlands, one of the scientists who dreamed up the Sloshsat
mission, the problem is that nobody really understands how rocket propellant
behaves in low gravity. 鈥淚f you use even the best estimates, it doesn鈥檛 matter.
The physics is not well understood,鈥� he says.

At the moment, spacecraft try to control slosh by making a series of gentle
鈥渟ettling burns鈥� to calm the propellant prior to a major thrust. But it鈥檚
difficult to predict how much burn you need to persuade the propellant to stop
moving around. And settling burns don鈥檛 always work. 鈥淭he glob still has a lot
of momentum so it bounces off and sloshes around,鈥� says Cheng.

And that鈥檚 not all. If propellants find their way into the wrong place they
can shut down the engine or even make it explode. 鈥淚f the oxidiser isn鈥檛 over
the outlet, you suck gas and you can鈥檛 get the engine going,鈥� says David Chato
of the NASA Glenn Research Center in Cleveland, who manages the NASA side of the
project. 鈥淏ut if there鈥檚 too much oxygen and not enough gas, then when you
finally get it off, you basically have a bomb in your combustion chamber
迟丑谤辞补迟.鈥�

Worse, the problem is getting bigger as satellite electronics become smaller.
鈥淚n the old days, propellant was perhaps 10 per cent of a spacecraft鈥檚 mass,鈥�
says Chato. 鈥淭he fact that liquid was sloshing around was not a big problem. But
these days, 60 to 70 per cent of a spacecraft can be propellant.鈥�

That鈥檚 why researchers are so interested in slosh. The Sloshsat mission is a
vastly more elaborate version of a 1992 European satellite experiment called the
Wet Satellite Model. Wetsat flew freely in space for six minutes after being
released from a sounding rocket shot from Sweden. Its tank was just two
centimetres wide and it had no internal sensors to detect the fluid flow, only
accelerometers to record its overall movements. Even so, the mission gave
mathematicians some real information about the behaviour of fluid in space which
scientists have used successfully to test Sloshsat鈥檚 fluid dynamics model.
鈥淲etsat gave us confidence in our predictive capability,鈥� Vreeburg says.

But the flight was of limited value. The motion of the liquid was restricted
by the tiny tank. 鈥淵ou can鈥檛 build up much velocity in two centimetres,鈥� says
Vreeburg.

Sloshsat, by contrast, will generate data on much more powerful slosh. After
astronauts jettison the craft, the shuttle pilot will retreat to 90
kilometres鈥攆ar enough away to prevent a collision if the satellite careers
out of control, says NASA鈥攁nd keep the orbiter鈥檚 antenna pointing at
Sloshsat for ten days. The astronauts will gather 24 hours鈥� worth of data from
sensors inside Sloshsat鈥檚 tank monitoring the motion of 33.5 litres of water.
They will also record the craft鈥檚 motion using gyroscopes and accelerometers on
the frame.

Where鈥檚 the water?

Designing the internal sensors was a big challenge. Vreeburg鈥檚 team had to
devise detectors that could gauge the location and depth of water along
Sloshsat鈥檚 tank wall without disrupting its flow too much. Mounting the sensors
with glue was a non-starter since molecules would leach into the water and skew
its surface tension. And since most off-the-shelf sensors are made of stainless
steel, they could deposit ions into the water too.

So Vreeburg decided to work from the inside out. He started with a rounded
aluminium cylinder to act as a former for the tank. Then he made about 150 small
rings from thin platinum wire and stuck them on to the aluminium in groups of
seven, each with a central ring surrounded by a hexagonal array.

He coated the aluminum core with polyethylene then reinforced it with epoxy
impregnated aramid fibres, which are even stronger than the fibreglass weave
used in boats, leaving a 15-centimetre porthole exposing the core. 鈥淲e dissolved
the aluminium core with sodium hydroxide and let it come out,鈥� Vreeburg says.
The result was an epoxy-aramid composite tank with the arrays of platinum rings
stuck to its polyethylene lining.

By measuring the capacitance between pairs of rings in the arrays, Vreeburg
says he can work out the depth of the water along the wall of the tank. The
greater the capacitance, the deeper the water.

The measurements rely on the fact that pure water is an electrical insulator.
That means each pair of platinum rings is separated by insulating material and
so acts as a capacitor, and can store electrical charge. The amount of charge it
can store is determined by the amount of insulator between the rings. So when a
charge is generated by pulsing alternating current through the rings, the depth
of the water can be inferred by measuring the capacitance of neighbouring
rings.

But Vreeburg still needed to measure the velocity of the sloshing liquid. He
came up with a tiny cone-shaped glass bead just 0.3 millimetres across that
could be inserted through small holes in the polyethylene tank. Each bead has a
miniature electric coil inside. 鈥淲e will heat this bead with current so it gets
hot. Then we will measure how fast it cools,鈥� Vreeburg says. The faster the bead
cools, the faster the water is flowing over it.

The heads need to protrude about 3 millimetres into the flow, but to minimise
the disturbance, Vreeburg installed them at just 10 places. Again, he used no
glue. Each cone-shaped bead was mounted on a base of solid polyethylene and
heated with a platinum wire to meld it to the polyethylene wall.

The task of analysing Sloshsat data will fall to Arthur Veldman, a
mathematician at the University of Groningen in the Netherlands. He and his
colleagues are experts at applying fluid dynamics models to ships. Now they have
written a mathematical model they believe is good enough to describe the
movements of Sloshsat and its watery payload. 鈥淭ake a bottle of beer,鈥� says
Veldman. 鈥淭hrow the bottle in the air and let it tumble freely. Our code is able
to compute that motion.鈥� He is also working on a computer visualisation that
will show the tank and the fluid in motion inside it.

Veldman and colleagues intend to use the raw data from the mission to refine
their models until they can accurately predict the motion of fluid in Sloshsat鈥檚
tank and the movement of the craft. This will not be easy. The models they are
working on are radically different from the sort they use for Earthbound
applications.

On Earth, the main force acting on a liquid is gravity. When you鈥檙e
predicting the sloshing motion of oil inside a supertanker, for instance, you
don鈥檛 have to worry about things like how quickly the liquid slithers across the
surface of the tank or the strength of the attractive forces in the liquid. But
in space, where gravity is negligible, these 鈥渃apillary forces鈥� take over.

Capillary forces come in two flavours: attraction between the liquid
molecules鈥攕urface tension鈥攁nd adhesion of the molecules to the walls
of the tank, called wetting. These forces oppose each other, and it is this tug
of war that controls the way the liquid interacts with its container.

To visualise this, Veldman says, look carefully inside a tumbler of water.
Water wets glass very well, so it creeps up the inside of the glass. The result
is a very small contact angle between the liquid and the container. Not all
fluids behave like this. Mercury wets glass badly, so its surface bulges
upwards. Most rocket propellants resemble water and wet their containers very
well.

That is why Veldman鈥檚 team will be concentrating on the narrow region where
the fluid inside Sloshsat touches the wall. They figure that if they can model
this accurately, they will be close to understanding how the all-important
capillary forces control the movement of rocket propellant inside the tank.

But there is another complication. On Earth, the balance between wetting and
surface tension is constant for a given liquid on a given surface. But what if
the balance suddenly shifts? The contact angle would change and the liquid鈥檚
behaviour would become highly unpredictable.

Veldman has a hunch that something like this happens inside a spacecraft鈥檚
fuel tanks. He reckons that the contact angles change constantly as the liquid
sloshes about. 鈥淲etting [in microgravity] is physically a badly understood
phenomenon,鈥� Veldman says. 鈥淚s it static or dynamic? Is it changing during the
尘辞迟颈辞苍?鈥�

But Earth-based fluid dynamics cannot cope with this level of complexity.
They assume that the wetting effects stay the same as the liquid sloshes. 鈥淩ight
now, we assume a constant contact angle,鈥� Veldman says. 鈥淏ut we are pretty sure
that when the contact line is moving, this contact angle is different. It is
诲测苍补尘颈肠.鈥�

Exactly what happens remains a mystery. Because no one has come up with a way
of simulating zero gravity in the lab, the measurements have never been made.
That is where the ring sensors come in. They can determine the precise depth of
water along the shifting line where the water touches the wall. From these
readings, mathematicians can easily calculate the contact angle.

If these measurements confirm Veldman鈥檚 hunch, physicists are going to have
to revise their ideas about how liquids behave in microgravity. It would also
mean that fluid models developed for Earth overestimate the velocity and
pressures of sloshing propellants, which might explain why spacecraft like NEAR
sometimes overcompensate and tumble out of control.

Ultimately, scientists would like to run follow-up missions to learn how to
harness the force of propellant slosh. This would mean that a spacecraft could
tap the momentum of its propellant for manoeuvres much as huge tanker ships
adjust their loads to counter wave action. 鈥淚n the future, you could contour the
inside of your tank to take advantage of the forces. It鈥檚 not that much
different than a boat,鈥� Vreeburg says.

If Sloshsat works, it could reduce the odds of spacecraft careering out of
control, stop them from blowing up unexpectedly and even give astronauts a
powerful new lever to help manoeuvre their craft. Not bad for a barrel of
water.