We present preliminary results of a long-term radial-velocity search for co
mpanions to nearby M dwarfs, started in September 95. The observed sample i
s volume-limited, and defined by the 127 northern (delta > -16 degrees) M d
warfs listed in the Gliese and Jahreiss (CNS3) catalogue with d less than o
r equal to 9 pc and V less than or equal to 15. Observations are obtained w
ith the ELODIE spectrograph on the 1.93-m telescope of the Observatoire de
Haute-Provence. The typical accuracy ranges between 10 m s(-1) (the instrum
ental stability limit) for the brighter stars and 70 m s(-1) at our limitin
g magnitude. We complement the ELODIE velocities with older measurements ex
tracted from the CORAVEL database to extend our time base, albeit obviously
with lower precision. Simultaneously, we perform adaptive optics imaging a
t CFHT and ESO to look for close (a >0.05-0.1 ") visual companions in a lar
ger volume-limited sample. For stellar companions the two techniques togeth
er cover the full separation range, to beyond the limiting distance of the
sample. We will therefore eventually obtain a statistically meaningful inve
ntory of the stellar multiplicity of nearby M-dwarf systems. We also have u
seful sensitivity to giant planets, as illustrated by our recent detection
of a planetary companion to G1 876.
After 2.5 years, we have discovered 12 previously unknown components in thi
s 127 stars sample, plus a companion to an additional star beyond its south
ern declination limit. 7 of these are actually beyond the 9 pc limit, as th
ey belong to systems included in the sample on the basis of CNS3 photometri
c parallaxes which were biased-down by the unrecognized companion. The rema
ining 5 companions are true additions to the 9 pc inventory. More are certa
inly forthcoming, given our present selection bias towards short periods an
d relatively massive companions.
We have derived orbital elements for 7 of the new systems, as well as for s
ome known binaries with previously undetermined orbits. One system, G 203-4
7, associates an M3.5V star with a white dwarf in a rather tight orbit (a(1
) sin i = 15 R.) and represents a Post-Common-Envelope system. Some of the
new M-dwarf binaries will over the next few years provide very precise mass
determinations, and will thus better constrain the still poorly determined
lower main-sequence mass-luminosity relation. The first such results are n
ow being obtained, with some preliminary accuracies that range between 2% a
t 0.4-0.6 M. and 10% at 0.1 M.. We have also discovered the third known det
ached M-dwarf eclipsing binary, and determined its masses with 0.4% accurac
y.