快猫短视频

At the beginning of September, NASA鈥檚 Jet Propulsion Laboratory in Pasadena,
California, enjoyed a formidable reputation for finding its way around the Solar
System. The lab had steered a course through the stars for some of NASA鈥檚 most
successful interplanetary explorers鈥擯ioneer, Mariner and the Mars
Pathfinder to name just a few. All these spacecraft made their way across
millions of kilometres of empty space and arrived at their destinations with
pinpoint accuracy. With that record of success, JPL was confident about its
latest mission: the Mars Climate Orbiter, destined to study the weather on Mars.
But by the end of the month, JPL鈥檚 confidence was shattered and its reputation
in tatters.

On Thursday 23 September, the Mars Climate Orbiter arrived at its
destination. The spacecraft was supposed to enter into a series of elliptical
loops around the Red Planet, using the friction of the atmosphere to slow it
down into a circular orbit. Instead, it plunged deep into the Martian
atmosphere, where aerodynamic forces either ripped it to pieces or burnt it
out.

What went wrong? There was no mechanical failure, no problem with the
spacecraft鈥檚 software or its on-board navigational equipment. In fact, the
spacecraft had done exactly what it was told. Operators at JPL had blundered by
mixing up imperial and metric units, unwittingly sending the spacecraft on a
fatal trajectory. It was already dead before the team realised what they had
done.

The Mars Climate Orbiter was not the first spacecraft to be lost to human
error and is unlikely to be the last. But many spacecraft have survived similar
problems and worse. In fact, the history of space travel is filled with stories
of scientists and engineers battling against the odds to save spacecraft from
almost certain death鈥攁nd triumphing. Survival in space is often a matter
of quick thinking, ingenuity and more than a little luck.

Back in 1998, the Solar and Heliospheric Observatory (SOHO) was undergoing
routine maintenance under the direction of ground controllers at the NASA
Goddard Space Flight Center near Washington DC. SOHO had functioned normally for
two years, watching the Sun from a vantage point roughly 1.5 million kilometres
from Earth. 快猫短视频s fully expected it to survive another three years so that
in 2000 and 2001 it could monitor the peak of the Sun鈥檚 11-year cycle called the
solar maximum. But on 24 June something went dreadfully wrong.

Routine maintenance that day involved recalibrating the spacecraft鈥檚
gyroscopes and resetting its momentum wheels, which speed up and slow down to
allow the spacecraft to change its orientation in space. The wheels are also
used to counter the pressure of sunlight which tends to turn the spacecraft.
Because of this, the wheels gradually build up momentum which has to be dumped
from time to time by slowing them down while firing the spacecraft鈥檚
thrusters鈥攁 procedure that had been performed many times before.

This time, however, things went badly wrong. One of the gyroscopes was
mistakenly left on the wrong setting and the computer鈥檚 onboard fault detection
systems realised something was wrong and tried to adopt the fail-safe mode, by
reorienting the spacecraft so that its solar panels pointed towards the Sun and
its antenna pointed towards the Earth.

What the on-board computer didn鈥檛 know was that another gyroscope had not
switched on because of a software error but it began to use the readings from
this gyroscope anyway. With the careful, comic deliberation that only a computer
can muster, the spacecraft began to compare the zero reading from this gyroscope
with the last good reading. Of course, the difference was huge. The spacecraft
immediately decided it was spinning rapidly and fired its thrusters to counter
this phantom motion. Instead of controlling the motion, this made the spacecraft
begin to spin out of control. Again, the spacecraft鈥檚 fault detection system
realised that something was wrong and tried to right itself a second time.

Back on Earth, controllers noticed that something was wrong but then made
several crucial errors. Nobody on the team noticed that the spacecraft was
beginning to spin dangerously even though that information was clearly available
to them. Nor did anybody realise that the spacecraft was taking readings from a
gyroscope that had not been switched on. Instead, the control team concluded
that another gyroscope鈥攖he only one that was still working
properly鈥攎ust be faulty. Amazingly, they ordered the spacecraft to turn it
off.

SOHO now had no functioning gyroscopes and began to fire its thrusters
wildly. It began to spin far faster than it was ever designed to and eventually
started to tumble end over end. With its solar panels no longer facing the Sun,
it could not generate heat or power. Within minutes SOHO had lost contact with
Earth and began to freeze. A few hours later it was dead.

Now the race began to find out what had gone wrong and whether the spacecraft
could be revived. In weeks, engineers had found the software errors that were
part of the problem and were able to reconstruct the chain of events which had
killed SOHO. They could then make good guesses about the spacecraft鈥檚 condition
and what might be done to resurrect it.

Meanwhile, astronomers had begun to search for the spacecraft using the giant
Arecibo Observatory radio dish in Puerto Rico. They were trying to bounce radio
waves off a spacecraft the size of a trash can lying in empty space some 1.5
million kilometres from Earth. Incredibly, they found their needle in a haystack
by spotting radio waves reflected from its solar panels. By measuring the
variations in these reflections, they were even able to determine SOHO鈥檚
attitude and rate of spin.

International rescue

Francis Vandenbussche is the French space engineer in charge of the recovery
effort. 鈥淚 had never recovered a spacecraft in my life,鈥� he says. So the first
step was to learn from somebody who had. 鈥淚 compiled all the reports I could
find on the recovery of lost spacecraft,鈥� he explains. 鈥淭he closest case was the
Olympus spacecraft鈥� it was similar but not identical. I could not copy
that recovery, but the spirit was there.鈥�

In May 1991, the European Space Agency鈥檚 Olympus satellite went out of
control due to a software error and died. With its solar panels pointing away
from the Sun, its batteries and rocket fuel froze. However, as the spinning
spacecraft slowly turned its face to the Sun, electrical power began to trickle
into its circuits and ground controllers slowly regained control. Three months
after the original failure, the satellite was functioning again. Vandenbussche
headed the rescue team of 160 experts around the world, intending to repeat that
resurrection for SOHO. 鈥淲e treated SOHO like a lost child,鈥� he says.

The team calculated that as SOHO spun slowly, its solar panels would
gradually come back to face the Sun, just as those on Olympus had. Sure enough,
a few months later, controllers began receiving short bursts of signals and were
soon able to interrogate SOHO about its status. The real worry was that the
severe cold had damaged the spacecraft鈥檚 instruments.

Within days, engineers began the slow process of warming up the spacecraft.
First, they thawed a few fuel lines. This allowed thrusters to fire and stop the
spin. Then, week by week, they slowly reactivated the instruments. To their
amazement, nearly all worked perfectly. The most serious problem was the
gyroscopes, which all failed completely soon after the warm-up. However, the
rescue team devised an ingenious scheme to maintain the spacecraft鈥檚 attitude by
tracking the position of the Sun. As long as no other emergencies arise, SOHO
should comfortably complete its mission to monitor the solar maximum over the
next two years.

An altogether different problem threatened the Voyager 2 mission as it
completed its Saturn flyby in 1981. On 25 August, as the spacecraft was
preparing for close encounters with the Saturnian moons Enceladus and Tethys,
the on-board camera鈥檚 scanning platform jammed, preventing the camera from
tracking its targets as they passed by. With a few images lost, it was a minor
disappointment at the end of a spectacularly successful Saturn encounter, but
the threat to later encounters with Uranus and Neptune was severe.

Controllers back on Earth were faced with the task of diagnosing the problem
and preventing its recurrence. The detective work that followed is now regarded
as one of the classic space engineering triumphs of all time. The high-speed
motion of the camera platform was generated by a motor and gears that turned at
up to 170 revolutions per minute. Engineers hypothesised that this rapid
spinning had caused the lubricant on the gears to be flung off. Without proper
lubrication, the metal had worn away sending tiny fragments into the mechanism.
When these fragments accumulated in the teeth of the gear mechanism, it
jammed.

Engineers broke the jam by heating the gearbox in the hope of softening the
lump and then repeatedly switching on the motor to hammer it against the
obstruction. Eventually this strategy worked and the camera platform broke free.
To prevent the problem recurring, controllers came up with a plan for taking
pictures that would never again require the high-speed tracking. Instead they
tracked the camera more slowly and made up the difference by turning the entire
spacecraft. At Uranus in 1986, and again at Neptune in 1989, the camera platform
performed without a hitch.

The journey between Saturn and Uranus took five years, so the engineers had a
long time to perfect their solution. Back at JPL, the Mars Climate Orbiter team
would have given anything for a fraction of that time to solve their problem.
They have another chance to get it right when the Mars Polar Lander reaches the
Red Planet on 3 December. Fingers crossed.