Low mass stars and brown dwarfs

Ninety percent of stars in our Galaxy are less massive than the Sun. Despite this fact, the properties of stars with low or very low masses are far less certain than those of their more massive counterparts. For instance, establishing an observational mass-luminosity relationship for stars with masses smaller than is still an active field of research (Henry&McCarthy 1993). Similarly, observations are still unable to significantly constrain the lower end of the mass spectrum that 0.3 solar mass is produced by the star formation process.

Even less well understood are sub-stellar objects: the elusive brown dwarfs. It seems that, after a long and eventful search, the question of their existence has recently been settled by clear detections (Rebolo et al., 1995; Nakajima et al., 1995). Of course, this result opens a new field of study for this latest class of cosmic objects which will finally allow theory to be related to observations. Unfortunately, this relation is rather indirect because the classical observables of a brown dwarf (broad-band photometry and spectrum) are determined by its very thin atmosphere, while its physical status depends mostly on the age and mass.

Progress in understanding low-mass stars and brown dwarfs clearly requires a method for determining masses. The first step is to perform radial velocity surveys of large samples of low-mass stars in search of spectroscopic binaries. However, while these surveys provide fundamental statistical results (Duquennoy & Mayor, 1991), they can only yield masses for each component if combined with direct imaging measurements. Duquennoy et al. (1995) have recently discussed the impact of the VLT and VLTI on such a program. In particular, they show that the combination of high precision (~15m/s) radial velocity data with parallaxes provided by Hipparcos and angular separations from VISA would allow the determination of masses of very low mass stars to precisions at the percent level. Even the mass of a suspected 0.03 solar mass brown dwarf companion could be estimated with ~5% accuracy, firmly establishing its sub-stellar nature and allowing one to test evolution scenarios for these still mysterious objects.