What’s Happening at JPL

Deep Space 1:

For decades all space enthusiasts have known that chemical rockets are not the best way to travel fast and far once you leave the strong gravity of the Earth. The basic equation for a simple rocket operating under ideal conditions in space is

M(0)/M(1) = e(v/c)

where v is the final velocity, c is the exhaust velocity, M(0) is the initial mass including propellant, and M(1) is the final or empty mass, including everything else --engines, structure, fuel tanks, computers, and payload. The number e = 2.7182818... is a fundamental mathematical constant rather like pi, called "the base of the natural system of logarithms" (fun, huh?!).

Then, approximately:

e**1 = 2.718; e**2 = 7.389; e**3 = 20.09; e**4 = 54.60; ... etc ....

This means that if you want a rocket to go as fast as its exhaust, you just have to make M(0)/M(1) (the "mass ratio") equal 2.718, which is not difficult; but if you want to go (say) 4c, then you are in serious trouble, because the mass ratio required is over 50. This is not reasonable from the point of view of practical engineering. Notice in particular how fast things fall apart as (v/c) gets large. For normal chemical rockets, c is in the range of about 2.5-4.7 km/s. Of course you can put a little rocket on top of a big one, etc., which is quite feasible from a strictly engineering point of view. This strategy is often necessary, but in the end it just moves the issue to one of economics, as the entire vehicle quickly becomes enormous if the final velocity is large. To appreciate the depressing implications, the reader is invited to try out the above formula to calculate the minimum initial mass of a chemical rocket vehicle for a mission requiring a total v, or "mission velocity", of say 60 km/s, with a total payload of 100 kgm. It is apparent that to really break out of this situation (at least with a rocket of some kind), we simply must have higher exhaust velocities. Modern rocket engines already convert a very large fraction of the chemical energy in the fuels into kinetic energy of the jet. Chemistry offers little hope of improving c significantly by finding better propellants: we are stuck with combinations of the atoms we already know, and there does not seem to be any practical way to beat liquid hydrogen and oxygen by much.

One approach which has long fascinated space enthusiasts is the ion drive, which replaces the chemical rocket exhaust with a beam of charged particles, accelerated by electric fields. Since laboratory particle beams are routinely accelerated to velocities up to nearly the speed of light, this approach completely removes the fundamental limitation afflicting chemical rockets. Needless to say there are other practical problems, mostly involving the mass and efficiency of the power conversion system. Due to such issues it seems very difficult to obtain enough thrust from an ion drive to give high acceleration. Even 0.001g (where g is the acceleration of gravity on Earth, about 9.81 m/s;) is quite challenging. However, low accelerations, continued for days and weeks, are quite enough in the weightless, frictionless environment of space. For example, to achieve the 10 km/s or so needed to escape the Solar System from the Earth's orbit would require less than 2 weeks at 0.001g. If continued, this same acceleration would be enough to fly by Pluto in a year. Combined with nuclear reactors for power, ion drives promise to make travel over the entire Solar System practical within humanly reasonable times. The "Catch 22" problem in the past has been that no specific mission NASA has flown heretofore has required very high velocities, so the Agency has not developed ion drives for flight propulsion, given the financial pressure of approved missions. On the other hand, no mission could realistically be proposed which required an ion drive, since the technology has never been adequately demonstrated.

Deep Space 1, the first of NASA's "New Millennium" program (technology development missions intended to embody Administrator Goldin's "Faster, Better, Cheaper" principles), has, as its main goal, to break this impasse both for the ion drive, and also to develop some 11 other technologies needed for future NASA missions. DS1's main propulsion is a 30 cm dia, xenon ion thruster, with a maximum exhaust velocity c of about 30 km/s, more than 6 times higher than that of the best H-O2-fueled chemical rockets. The thruster has a maximum thrust of just 92 mN at 2.3 kW operating power.

30 cm Xe Ion Thruster for Deep Space 1 (NASA image)

In the figure, the rear part of the chamber contains Xe gas at low density. Electrons are injected into the chamber with sufficient energy (few eV) to ionize Xe atoms by impact. The Xe ions then migrate to the holes at the right, where they are accelerated to 1280 V between the grids. Finally, a cloud of lower- energy electrons is injected into the beam to allow the ions to recombine and prevent buildup of a large (-) charge on the spacecraft. Earlier this year the engine completed an 8000 hour (about 1 year) qualifying test in a thermal vacuum chamber at JPL. The DS1 mission was originally planned for launch on 21 July, but in April it was announced that it would be necessary to delay the launch. It has been desired to fly a comet and asteroid rendezvous mission which is as realistic and representative as possible in order to be certain the technology demonstration is truly relevant to real needs of future science missions. Thus DS1 will return some real science data.

The new mission plan, just released, has launch (on a smaller model of the Delta launch vehicle) on 15 Oct. 1998. The first destination will be 1992 KD, a near-earth asteroid, which the new orbit will fly past on 28 July, 1999. The diameter of 1992 KD is believed to be about 3 km, and its orbit extends from within the orbit of Mars out to about 3 AU; aside from these facts, little is currently known of its nature. The primary technology demonstration goals of the mission are scheduled to be completed a year after launch, after which DS1 will be placed on a trajectory to encounter the short-period comet Borrelly. Besides the ion drive, other technologies being developed include a light-weight solar concentrator array, autonomous optical navigation, and a Miniature Integrated Camera and Imaging spectrometer (MICAS). The science instruments include MICAS and a miniature space plasma integrated ion and electron spectrometer experiment. Together the instruments will allow detailed characterization of 1992 KD and its interaction with the solar wind, as well as diagnosis of the operation of the ion drive and its effects on the local plasma environment of the spacecraft.

Further details may be found at http://nmp.jpl.nasa.gov/ds1

2MASS Results -- A New Class of Stars:

At the summer meeting of the American Astronomical Society in San Diego in June it was announced that 2MASS, the 2 Micron All Sky Survey which we discussed here a few months ago, has identified a new class of cool stars, call class "L", which extends the familiar OBAFGKM spectral sequence. According to this classification, going back in its roots to the beginning of the century, O stars are hottest, pouring out most of their energy in the ultraviolet, while M stars are the coolest and by far the most numerous, ranging from swarms of low-mass M dwarfs, main sequence stars with as little as 1/10,000th the luminosity of the Sun, to rare red supergiants like Betelgeus. The letter divisions are subdivided further into 10 categories 0 (hottest) to 9 (coolest) so that in this system the Sun is classified G2V, a G star near the warm end (subgroup 2) of the class. The "V" means it is a dwarf, or main-sequence star, still stably burning hydrogen as most stars do for most of their lifetimes. Heretofore the coolest, dimmest main sequence stars were M9V. The L stars are so intrinsically dim that they have been difficult to identify by normal optical observations, despite being very probably even more common than M dwarfs. A typical object of this class might have a temperature of 2000 K (about 1/3 that of the Sun), and a size about like Jupiter. Maybe half of them are likely not stars at all, strictly speaking, but rather "brown dwarfs" -- objects below the mass limit (about 8% of the Sun's mass) needed to ignite hydrogen burning and reach the main sequence.

The 2MASS observations, yielding brightness measurements in the J (1.2micron) H((1.56micron;) and Ks (2.2micron;) bands, do not in themselves give the detailed spectral information needed to determine the spectral class, but require follow-up observations with a spectroscope on a large telescope. 2MASS does, however, yield a large number of candidates based on the 3-band infrared colours, and it has been found by the limited spectroscopy done to date that these colour-selected candidates are rich in L stars and brown dwarfs. Much of this detailed follow-up work has been done by Dr. J. Davy Kirkpatrick of IPAC, who presented the results.

About 20 L stars have been found in roughly 1% of the sky for which the analysis has been completed, and of these, about 6 appear to be brown dwarfs. The results suggest that the L stars are intrinsically about as numerous as all other types combined. If so, Proxima Centauri may conceivably be displaced as the nearest known body outside the Solar System.

Bill Wheaton
Waw@ipac.caltech.edu