SOMETIME towards the end of this year, one of the most advanced spacecraft ever built will be allowed to die after only two years in orbit. The Infrared Space Observatory (ISO) is not losing power, nor has it become damaged. In fact, it is working perfectly.
The 拢400 million spacecraft is doomed because it is running out of liquid helium to cool its infrared sensors. In space, the only way to achieve the observatory鈥檚 operating temperature of 1.8 kelvin (-271 掳C) is to plunge its sensors in a huge 鈥渋ce bucket鈥� of 2200 litres of superfluid helium. This gradually boils away, dissipating the spacecraft鈥檚 excess heat into space. When the helium runs out, probably by the end of the year, the temperature of the observatory will rise and the infrared sensors will be blinded by the spacecraft鈥檚 heat.
Although ISO鈥檚 demise was anticipated from the beginning, the European Space Agency, which built and launched it, is determined to find a better way to keep future spacecraft cooler for longer. So in June last year, researchers at ESA鈥檚 European Space Research and Technology Centre (ESTEC) in the Netherlands, began work on a device that cools spacecraft like a refrigerator-a huge improvement on the ice bucket approach that cools ISO. While the outside of a spacecraft, heated by the Sun and its internal equipment, will be at temperatures of around 300 K, the new device will cool its contents to a frigid 0.1 K, much closer to absolute zero than has been possible in space until now.
Advertisement
At these lower temperatures, orbiting observatories will perform better than ever before. In the next few years ESA plans to launch spacecraft that will use superconducting sensors to give unprecedented views of the visible Universe, while others will be equipped with infrared detectors to peer farther into the Universe with greater clarity. But these sensors will only work correctly if they are cooled to within a fraction of a degree of absolute zero. 鈥淚n general, any space mission which requires high-sensitivity detectors will need cooling to low temperatures,鈥� says Jan Tauber, a physicist at ESA.
Tool challenge
But keeping cool in space is much harder than it sounds. Cooling devices that can push the temperature even closer to absolute zero than this have been in use on Earth for many years, but they are wholly unsuited to work in space. Any scientist who runs a cryogenic cooler will tell you that these complex machines need constant care and attention. And while waste heat from terrestrial coolers can easily be dissipated into the atmosphere, dumping it into space as quickly as it is generated is difficult. But the biggest problem is that the final cooling stage only works with the help of gravity-something that is in short supply in space.
Now ESA is so confident that it has solved these problems that it claims its new design will work for up to six years without maintenance. 鈥淚t鈥檚 like building a car that can go a million miles without a service,鈥� says Chris Jewell, the physicist who is coordinating the project.
The new cooler is a complex device. Since no single cooling mechanism can achieve the 300-kelvin temperature drop, ESA鈥檚 chiller has several components, each using a different refrigerating technique. The stages sit within each other like Russian dolls, each cooling and insulating the region within it. The outer stages are mechanical coolers known as Stirling cycle fridges, based on a device conceived in 1816 by Robert Stirling, a Scottish clergyman.
A Stirling cycle fridge works by compressing a gas-helium in this case- inside a cylinder using a piston. This heats up the gas, which is then moved out of the cylinder, cooled by a device that absorbs and stores heat, and passed into a second cylinder where it is allowed to expand and cool, drawing heat from its surroundings. Finally, it is transferred back to the original chamber where the process starts again. The effect is to transfer heat from the second cylinder to the first (see 鈥淭he Stirling cycle鈥�).
Stirling cycle fridges have some attractive properties for space applications. They are extremely efficient and capable of running on the limited power supply available to spacecraft. They are relatively vibration free and able to span huge temperature ranges. These will be no ordinary fridges. 鈥淭he ideas and concepts are very simple, but to make something that will work for years is very different,鈥� says Anna Orlowska, a physicist at the Rutherford Appleton Laboratory (RAL) in Oxfordshire, who is building the Stirling cycle fridge for ESA.
The key to reliability is to eliminate friction. And without friction there will be no need for lubricant, which tends to freeze or evaporate in space. The piston that Orlowska and fellow physicist Tom Bradshaw have designed is so precisely machined that it fits almost perfectly into the cylinder without touching the walls. 鈥淭here鈥檚 a clearance between the piston and the cylinder wall, so there are no rubbing parts,鈥� says Orlowska. The gap is only a few micrometres, so gas leakage is negligible and there is no need for the piston rings which seal conventional cylinders.
The real challenge is to ensure that this tiny gap is maintained while the piston thrusts in and out of the chamber up to 40 times a second. So the RAL team is using a spring system that drives the piston at the required frequency but also aligns it extremely accurately. The system consists of a series of carefully aligned metal discs with a shaft passing through their centre. With the metal discs secure, the shaft is precisely aligned with the cylinder.
Each disc is etched with three spiral cuts so that it acts as a spring. Although the shaft is carefully aligned, this allows it to move backwards and forwards with a resonant frequency close to 40 Hertz. At one end of the shaft is the piston head and at the other a permanent magnet surrounded by a conducting coil, like a loudspeaker drive. Passing an alternating current through the coil repels the magnet, setting the entire shaft oscillating. The system is so accurate that it maintains the tiny gap without ever allowing the piston to touch the cylinder wall.
In theory, a single Stirling cycle fridge could lower the temperature from 300 K to 20 K. But at the lower end of its range, a Stirling cycle fridge鈥檚 ability to do this- its cooling power-becomes extremely poor. A fridge must pump more heat out of its cold chamber than leaks in from the outside. The fridge鈥檚 cooling power is about 100 milliwatts at 20 K, but more than 100 mW of heat seeps in from the outside, says Jean-Michel Lamarre of the Institute of Space Astrophysics in Orsay, near Paris, whose team will combine and test the main components of the cooling system. The problem is rather like preventing a snowflake from melting in a blast furnace.
So Orlowska and Bradshaw鈥檚 design employs a two stage Stirling cycle cooler. The first stage cools to 120 K and at this temperature its cooling power is closer to 1 W, more than enough to cope with the heat leaking in. The second stage takes the temperature down to 20 K.
No vibrations
There is another advantage in having two stages. By arranging the pistons back to back, the vibrations they create tend to cancel out. In addition, a feedback mechanism monitors vibration and changes the motion of the piston drives accordingly to eliminate it.
However, this combined system uses up more power and so creates waste heat that could cause the instruments to overheat, thus defeating the whole object of the exercise. 鈥淲aste heat is an issue for all spacecraft but for us it鈥檚 particularly severe,鈥� points out Lamarre. ESA鈥檚 cooler will produce more than 100 W of excess heat, around five times as much as most instruments, and somehow this must be radiated into space. 鈥淭he solution is to create a large surface area which radiates efficiently,鈥� he says. A perfect radiator at 300 K emits 460 W/m2, and although commercially available paints are far from perfect they can get close to this figure. 鈥淲ith these paints we should be able to handle it,鈥� he says.
While the twin-stage Stirling cycle cooler reduces the temperature to 20 K, the final cooling stage only works below 4 K. To bridge this gap, the ESA team has turned to the principle of a bicycle pump. The next two stages work by pumping helium gas through a tiny nozzle, which cools it by expanding it rapidly, in the same way as high-pressure air inside a bicycle pump gets cold as it is expelled. This is called Joule-Thomson cooling. Two further, similar stages take the temperature down to 4 K and then to 1.6 K.
The long-awaited breakthrough in space refrigeration has been made for the final cooling stage, which is a 鈥渉elium dilution fridge鈥�. This works by mixing two isotopes of liquid helium, helium-3 and helium-4. In every liquid, the atoms and molecules are held together by weak forces known as Van der Waals forces. The cooling occurs because the Van der Waals forces between helium-4 and helium-3 atoms are stronger than those between atoms of the same isotope. As they mix, the liquid absorbs energy in the form of heat from its surroundings to create these stronger bonds. This lowers the temperature of the surroundings to around 0.1 K.
In terrestrial helium dilution fridges, the isotopes are then separated so that the cycle can begin again. Because helium-3 is light it floats to the top of the mixture, where it evaporates more readily than helium-4-a process that relies crucially on gravity. So the biggest problem facing the designers of helium dilution fridges is separating the isotopes without the aid of gravity. Over the years, many physicists have tried and failed. That is, until Alain Benoit took a stab at the problem.
Benoit is a physicist at the Centre for Very Low Temperature Research in Grenoble, France. In 1991 he resolved the problem-effectively by ignoring it. Why bother to recycle the helium isotopes when they are used in such small quantities, he reasoned. 鈥淚 just asked myself: is it possible to make the dilution of the two isotopes without recycling?鈥�
Benoit鈥檚 fridge works by mixing the isotopes and then venting them into space. Over a six-year mission this would use up only a few litres of each isotope, which would be stored onboard. Six years is long enough for most missions.
The cooling power of Benoit鈥檚 fridge is a tiny 100 nanowatts-about a billionth of the power needed to run the entire refrigeration system. A household fridge with the same cooling power would take 20 000 years to chill a litre jug of orange juice from room temperature to a far more refreshing 2 掳C.
For the moment, the entire cooling system is being assembled in parts. At RAL, Orlowska and Bradshaw are testing their Stirling cycle coolers in conjunction with a Joule-Thomson device. 鈥淲e have already demonstrated temperatures of 4 K,鈥� says Bradshaw. And Benoit has completed a helium dilution fridge capable of reaching 0.1 K. The next step is to combine them. 鈥淲e would like to demonstrate the whole system before the end of 1997,鈥� says Benoit.
Next year, ESA will select the instruments that will fly on two missions which are due for launch early next century-the Planck Surveyor (formerly COBRAS/ SAMBA) and FIRST (the far infrared and submillimetre telescope). Both missions will use instruments known as bolometers, which measure tiny temperature increases that occur when they are hit by radiation in the infrared to microwave regions of the electromagnetic spectrum.
Knowing the temperature that can be achieved with the new cooling system will allow the designers to accurately assess the performance of their detectors. If the bolometers on the Planck Surveyor were to run at 0.3 K instead of 0.1 K, for example, their sensitivity would be degraded tenfold, says Tauber. But with bolometers at 0.1 K it will be possible to measure a few tens of thousands of photons.
These missions will see deep into the Universe. The Planck Surveyor, for example, will examine the small-scale ripples in the cosmic background radiation that are the echoes of the big bang. FIRST will open up the last major portion of the spectrum to astronomers by observing the sky at wavelengths between 100 micrometres and 1 millimetre, observing primordial galaxies at the edge of the Universe for the first time.
However unglamorous fridges in space may seem, the secrets of the Universe will remain hidden without them.
* * *
The Stirling cycle
THE Stirling cycle cooler works using the principle that a gas heats up if it is compressed and cools down if it is allowed to expand. But compressing and expanding a gas inside a cylinder using a piston is of little use since the entire cylinder heats up and cools down during each cycle.
So the trick is to move the gas during this process so that the heating takes place at one end of the cylinder and the cooling at the other. In effect, this pumps heat from one end of the cylinder to the other.
Stirling cycle fridges use two pistons: the first compresses and expands gas while the second, called the displacer, moves gas from one chamber to another without altering its pressure.
This process can be made more efficient by removing the heat from the gas before it passes into 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 which absorbs heat. This device is known as a regenerator.
The diagram shows how a Stirling cycle fridge works.
A. The piston is at its lowest point, the displacer at its highest. Above the displacer is the cold region, and between the displacer and piston is the hot region. Helium gas fills both. In this position, the displacer blocks the loop pipe which connects the two regions via the regenerator.
B. The piston moves up, compressing the gas in the hot region and increasing its temperature.
C. The displacer moves down, moving the gas from the hot region to the cold region via the regenerator, which extracts heat and cools the compressed gas.
D. Both piston and displacer move down, expanding the gas in the cool region and reducing its temperature.
E. The displacer, on its way back up, forces the expanded gas back through the loop and regenerator, where it is reheated by the heat stored there from stage C.