Clumping-corrected mass-loss rates of 64 Galactic Wolf-Rayet (WR) stars are
used to study the dependence of mass-loss rates, momentum transfer efficie
ncies and terminal velocities on the basic stellar parameters and chemical
composition. The luminosities of the WR stars have been determined either d
irectly from the masses, using the dependence of L on mass predicted by ste
llar evolution theory, or they were determined from the absolute visual mag
nitudes and the bolometric corrections. For this purpose we improved the re
lation between the bolometric correction and the spectral subclass.
(1) The momentum transfer efficiencies eta (i.e. the ratio between the wind
momentum loss and radiative momentum loss) of WR stars are found to lie in
the range of 1.4 to 17.6, with the mean value of 6.2 for the 64 program st
ars. Such values can probably be explained by radiative driving due to mult
iple scattering of photons in a WR. wind with an ionization stratification.
However, there may be a problem in explaining the driving at low velocitie
s.
(2) We derived the linear regression relations for the dependence of the te
rminal velocity, the momentum transfer efficiency and the mass-loss rates o
n luminosity and chemical composition. We found a tight relation between th
e terminal velocity of the wind and the parameters of the hydrostatic core.
This relation enables the determination of the mass of the WR stars from t
heir observed terminal velocities and chemical composition with an accuracy
of about 0.1 dex. for WN and WC stars. Using evolutionary models of WR sta
rs, the luminosity can then be determined with an accuracy of 0.25 dex or b
etter.
(3) We found that the mass-loss rates ((M)over dot) of WR stars depend stro
ngly on luminosity and also quite strongly on chemical composition. For the
combined sample of WN and WC stars we found that (M)over dot in M. yr(-1)
can be expressed as
(M)over dot similar or equal to 1.0 x 10(-11) (L/L.)(129) y(1.7) Z(0.5) (1)
with an uncertainty of sigma = 0.19 dex
(4) The new mass-loss rates are significantly smaller than adopted in evolu
tionary calculations, by about 0.2 to 0.6 dex, depending on the composition
and on the evolutionary calculations. For H-rich WN stars the new mass-los
s rates are 0.3 dex smaller than adopted in the evolutionary calculations o
f Meynet et al. (1994).
(5) The lower mass-loss rates, derived in this paper compared to previously
adopted values, facilitate the formation of black holes as end points of t
he evolution of massive stars. However they might create a problem in expla
ining the observed WN/WC ratios, unless rotational mixing or mass-loss due
to eruptions is important.