F 
R

4
- 
9 
9

---

Home | Update | LATEST ISSUE | Gallery | FR Profile | Datafiles | FR 4/99

DEEP SPACE 1 ON ITS WAY

Faster, better, cheaper - Deep Space 1, which was launched in October of 1998, ist the first project of NASA's New Millennium program. Unlike other NASA missions into deep space, the New Millennium missions focus on flight testing new technologies rather than scientific research.

NASA itself calls DS1 a high-risk mission. A total of twelve innovative technologies are on board the probe. Many of them are vital for the operation of the craft and do not have a back-up except of a few sub-level options. Some of the technologies had to be validated within a few hours from launch. Among them the new generation solar arrays which supposedly generate 15 to 20 percent more power than other systems of the same size.

The two arrays each consist of four 1,13 by 1,6 meter large modules. A total of 720 Fresnel lenses made of Silicon concentrate are focusing the sunlight on the 3600 solar cells. The system generates about 2400 Watts and a voltage of approximately 100 V. The power abates as the spacecraft ages and the distance to the sun increases. This technology is called SCARLET (Solar Concentrator Array with refractive Linear Element Technology) and its utilization on Deep Space 1 is the first application of the system in space.

The properly working power supply is especially essential for the space probe's advanced propulsion system. Deep Space 1 is powered by an ion engine, the first time that this technology is used on a deep space mission. Americans, Russians and Europeans have been working on ion engines since the late fifties. Unlike conventional rocket engines, which burn chemical fuels and the generated gases expand in a nozzle, the ion engine is generating thrust through the ionization of fuel atoms. These are then accelerated out of the engine via a pair of metal grids which are charged positive and negative respectively with up to 1280 volts electric potential.

DS1 carries noble gas Xenon as fuel which is 4-1/2 times as heavy as air. The thrust stream of the Xenon ions reaches a speed of up to 100,000 km/h. However, due to the very low amount of fuel mass that moves through the system, this ion engine produces a maximum of one fiftieth of a pound of thrust at full throttle. This is approximately as much as a sheet of paper would weigh in one's hand and is not enough to launch a spacecraft from Earth and overcome gravity.

However, once the spacecraft is beyond the gravity of Earth, the ion engine has a distinct advantage over conventional engines which can generate a lot of thrust over a short period of time. The ion engine has a much longer standing time, making it the ideal propulsion for long lasting deep space missions. According to NASA, an ion engine can generate about ten times as much thrust as a conventional engine for a given amount of fuel. For the primary mission of DS1 (up to September of 1999), NASA expects the ion engine to accelerate the probe by 13000 km/h.

However, when the ion engine was started for the first time in November of 1998, it first did not look like this goal could be reached. After running at low power settings for about four and a half minutes, the engine shut itself off. NASA engineers now think that this was due to a contamination of the metal grids at the rear of the engine. Two weeks later, on 25 November, the engine started up again and ran without problems.

On 2 December, the ion engine had thrusted for a continous 190 hours. This is longer than any other conventional propulsion system has done before. At the beginning of January, the DS1 ion engine had accumulated a total of 850 hours.

Still, the engine was stopped several times in December to allow the validation of other new technologies on board the craft without interence by the ion engine. Among these technologies is autonomous optical navigation system (AutoNav). With its help, DS1 is supposed to rendezvous with the asteroid 1992 KD in July of 1999.

Deep Space 1's AutoNav is supposed to take many navigation duties from the ground controllers on Earth. AutoNav is supposed to take images from known asteroids and compare their position with the stars in the background. For that purpose, the system has stored the orbits of 250 asteroids and the positions of 250,000 stars in its memory. AutoNav will then direct necessary trajectory changes through changes of the ion engine thrust profile or by activating the DS1 hydrazine thrusters.

With the help of AutoNav, DS1 is supposed to come as close as 10 kilometers to 1992 KD. According to NASA this would be the closest approach to a solar system body yet. If all system operate nominally, NASA even thinks about closing in to 5 kilometers.

Deep Space 1 will take one even closer step to autonomous flight operations. On board the probe is an experimental computer software, called the Remote Agent, which is supposed to make its own decisions and come up with a plan of activities to reach a given goal. The computer used is a RAD 6000 which was also utilized on the Mars Pathfinder mission. The Remote Agent software, which some compare to the artificial intelligence of the HAL 9000 computer featured in the science fiction movie "2001 - a space odyssey", is comprised of three modules:

- the Planer, which schedules a certain chronology of events and activities to reach a goal set by ground control,

- the Executive, which makes decisions based on the plan of the first module, taking into account tthe current state of the spacecraft, its limitations and the mission objective,

- the Watcher, the so called mode identification and reconfiguration module (MIR), which is the Remote Agent's fault protection, constantly looking out for glitches and, if it finds one, immediately informing the Executive.

The remote agent software with its somewhat artificial intelligence is schedule to be tested sometime between March and May of 1999. A first experiment, lasting only a few hours, will only involve the Executive and the MIR. A second experiment one week later will activate all three modules for a duration of six days. The Planer will then have come up with its own concept for accomplishing a thrust maneuver with the ion engine. During each of the experiments, the ground controllers will incorporate a failure to see how the Remote Agent copes with it.

In designing spacecraft that can operate and accomplish mission goals autonomously, NASA sees one way of reducing the costs for mission operations. According to estimates by project managers, the use of the Remote Agent software could lead to savings in mission costs of up to 60 percent.

From page 36 of FLUG REVUE 4/99