BUILDING and launching satellites is an expensive, time-consuming business. A telecommunications satellite the size and weight of a family car usually takes at least five years to design and build. It would also set you back around 拢250 million 鈥� a figure that is even beyond the budget of many small countries.
But this isn鈥檛 the only way to send a satellite into space. Since 1979, space scientists at the University of Surrey in Guildford have sent 10 satellites into low orbits around the Earth. Each weighed only a few tens of kilograms, cost less than 拢2 million and took roughly twelve months to design, build and launch. So far, the satellites have carried small experiments into orbit and allowed countries such as South Korea, Chile and Portugal to start up space programmes at a fraction of the usual cost.
Now Surrey Satellite Technology Ltd, the company set up by the university to manage the project, is using its experience with low-cost, small satellites to design ones with built-in rocket motors that are bigger and better than before but still cheaper than conventional satellites. Small countries with limited space budgets will be able to launch relatively advanced remote-sensing satellites and even operate the constellations of satellites necessary for continuous digital communications. SSTL believes it can keep costs below $5 million for each vehicle and hopes to launch its first attempt next spring.
Advertisement
SSTL has turned satellite design on its head. In the past, satellites have been built like fighter aircraft with the emphasis on performance rather than cost, says Jeff Ward, who is in charge of R&D at the company. 鈥淭here is nothing special about satellites that requires this level of performance,鈥� he points out. Instead, SSTL wants to build satellites where the emphasis is on cost-effectiveness. This means re-examining every aspect of their design, from the computers that control them to the way they fine-tune their position in orbit.
A built-in rocket motor is the most effective way of adjusting a satellite鈥檚 orbit. But in the past, SSTL has foregone such luxuries to minimise costs. Now it says a motor is necessary. One reason why small satellites are good value is that they can hitchhike on the back of larger missions, which bear the brunt of the launch costs. But the growth in the number of small satellites means that suitable launch opportunities are becoming scarce. Furthermore, this limited number of missions provides access to only a few orbits. 鈥淲ith a built-in motor, a satellite can manoeuvre into a different orbit and this will allow us to make the most of any future lifts,鈥� says Ward.
But conventional rocket motors are expensive because they are designed to achieve extremely high levels of performance. For example, Royal Ordnance, an engineering firm based in Aylesbury, manufacturers motors to such high specifications that the amount of thrust each produces can be guaranteed to within one per cent. This allows the amount of fuel needed for a mission to be calculated exactly. Ward believes such levels of performance are unnecessary. He claims it would be cheaper to build a motor to lower specifications and carry enough fuel to cover any variations.
Fuel also causes other problems. The conventional motors suitable for use with small satellites use hydrazine, an explosive compound of nitrogen and hydrogen that is also extremely poisonous. Each motor must be fitted with several expensive valves to prevent leaks, and anybody handling hydrazine must wear a protective suit incorporating life support systems, just in case.
But the explosive potential of hydrazine makes the people in charge of the primary missions nervous. We have to convince them that our satellites represent no risk to their missions,鈥� explains Ward. The slightest possibility of an explosion makes this job difficult. The alternative is to use Russian launch facilities where owners of primary missions have less control over who rides with them.
To guarantee launch opportunities in the West, SSTL is developing its own rocket motor which uses polyethylene for fuel. This hard plastic solid is neither poisonous nor explosive. So the fuel requires few special handling facilities and will cause less concern among other satellite operators on the same flight. SSTL has built and tested a prototype, but is now deciding whether the new thruster will be cheaper to build and operate than hydrazine motors.
Staying cool
Rocket motors generate heat regardless of the fuel they use, and the temperature of the satellite is critical. For instance, batteries only operate effectively at between 0 掳C and 70 掳C and most on-board electronics need to be protected from extreme temperatures. During firing, however, the nozzle in a rocket motor can reach several thousand degrees. To prevent conduction through the metal structure of the craft, special insulating material will be placed at strategic joints between the motor and the rest of the satellite, says Craig Underwood, a physicist at the University of Surrey who specialises in satellite design. The heat from the rocket nozzle can also radiate through space, so the parts of the craft in sight of the nozzle must be covered with a material that reflects this radiation. The material must prevent the components becoming too hot but it must also be a poor radiator of heat so that the craft does not cool below its operating temperature when the motor is switched off.
These materials are expensive. For example, the insulating material costs around 拢5000 for a spool the size of a roll of sticky tape. Underwood says that with careful design, these thermal materials will add less than 拢10 000 to the total cost of each mission.
Another problem is that built-in rocket motors make missions complex. The length and timing of each firing is critical and the position and direction in which the craft is pointing must be carefully monitored before and after each 鈥渂urn鈥�. Conventional satellites take extremely accurate attitude measurements before firing. But such accuracy is expensive and, according to Ward, unnecessary. However, the satellite will be able to make highly accurate measurements of its position in space using the US Global Positioning System, which is cheap to use. And instead of using long burns, SSTL鈥檚 satellites will move in a series of short steps to compensate for the poorer performance of the motor. After each burn the position will be measured and corrections made for any errors. The disadvantage is that a series of short burns uses more fuel than a single, long burn. But according to Ward, this kind of trade-off between fuel economy and expensive hardware is what makes SSTLs satellites more cost-effective.
Satellites with built-in rocket motors have several advantages. For example, they can maintain specific orbits. This allows a remote-sensing satellite to photograph exactly the same part of the Earth鈥檚 surface once a week, say, so that the images can be compared. A fixed orbit also ensures that telecommunications satellites pass over the same point at regular intervals.
SSTL鈥檚 rocket motor will eventually lift its satellites to higher altitudes than ever before. But the height of the orbit determines the amount of radiation the satellite is exposed to. This is important because radiation can interfere with the electronic circuitry that controls the satellite and damage conventional microchips beyond repair. Most satellites in high orbits use microchips that are designed to cope with the radiation but these are roughly one hundred times more expensive than conventional chips says Ward. In the past, SSTL has reduced costs by using conventional chips to control its satellites and keeping them below 1500 kilometres where radiation levels are relatively low.
Ward admits that there has always been a risk that the circuitry would be irreparably damaged or fail at a crucial moment. 鈥淚t will always be possible to reduce this risk but only at a price,鈥� he points out. 鈥淏ut you can never reduce it to zero, no matter how much you spend.
Look at the Hubble Space Telescope.鈥� So far, the strategy has paid off: not one of SSTL鈥檚 satellites has been written off by radiation damage.
Radiation risk
With this success behind them, the team is looking for ways to operate their satellites at higher altitudes without resorting to large-scale use of radiation-hardened microchips. This could be difficult. At these altitudes, satellites begin to pass through the Van Allen radiation belts, which consist of energetic particles trapped by the Earth鈥檚 magnetic field. Satellites in these orbits have to cope with radiation levels up to a thousand times higher than at 800 kilometres.
One strategy to cope with increased levels of radiation is to reduce the length of the missions, says Ward. 鈥淭he components that last for ten years at low altitude may last for only two or three at a higher altitude,鈥� he points out. One way to protect electronic components is by shielding them behind sheets of aluminium, but this adds weight and cannot guard against the most energetic forms of radiation. Another is to encode information in such a way that an on-board computer can spot errors and correct them. But errors are not the only problem that radiation can cause. Energetic particles passing through a conventional chip can flick electronic switches in a way that short-circuits the power and melts the chip. According to Underwood, this can be avoided if the power can be turned off quickly enough, although a back-up chip must be available to take over during the power cut.
Even satellites protected in these ways may not be able to cope with highly elliptical orbits that take them through the Van Allen radiation belts several times a day. 鈥淚n these circumstances, we could decide that the most cost-effective option is to use a radiation-hardened microchip,鈥� says Underwood.
The long-term goal is to send a satellite into lunar orbit, says Ward. 鈥淲e first thought we could do it when we realised that a Moonshot requires only 15 per cent more fuel than placing a satellite in a geostationary orbit at 36 000 kilometres,鈥� he says. Such a satellite could reach the Moon for as little as 拢10 million: a price well within the budgets of many smaller countries and organisations. Radiation is not such a concern on lunar missions because the craft would only have to pass through the Van Allen radiation belts once, after which radiation levels drop by about a factor of ten. During transit, all but essential circuits and their back-ups could be switched off.
Interest in lunar missions is increasing. Last year, the US space probe Clementine spent two months photographing the lunar surface at the relatively cheap price of roughly $80 million. And last month, NASA agreed to spend $60 million on another mission to orbit the Moon in 1997. (By contrast, the Apollo programme would have cost more than $100 billion at today鈥檚 prices). International Space Enterprises, based in San Diego California, wants to send a lunar rover to the Moon鈥檚 surface equipped with a camera that would allow people on Earth to shoot video footage of the surface. It claims to be negotiating with several potential customers in the theme park, entertainment and film industries.
SSTL has its own ideas. The Millennium Fund, set up to finance celebrations at the turn of the century with proceeds from the National Lottery, is looking for ways to honour British technology. 鈥淎 lunar mission could be one option,鈥� says a confident Ward. 鈥淏ut I wouldn鈥檛 be surprised if we reached the Moon before then.鈥�