While it is accepted that the progenitors of (single) W-R stars are massive
O type stars, the exact evolutionary sequence is not well established.
The intermediate phase in which the outer layers are stripped away
by extreme mass-loss is accepted as being either a
red-supergiant (RSG) or an LBV phase for very massive
(60M
) stars,
with direct evolution from the Of to the WN phase (Conti 1976)
only expected for the most massive (
120M
) stars
(Maeder 1982).
Standard evolutionary theory predicts a W-R lifetime
of typically yr, although strongly dependent on
initial mass and metallicity (Maeder & Meynet 1994), with
evolution to the W-R phase only proceeding for higher mass stars at
lower metallicity. In contrast, recent studies for LMC WNL stars
indicate luminosities and mass-loss properties which are
comparable with Galactic stars (Crowther & Smith 1996b).
From spectral morphology and detailed analyses,
Crowther et al. (1995c) proposed the following evolutionary sequences :
OOf
WNLha
WN6--7
WNE
WC
SN.
for very massive (60M
) Galactic stars, while for
less massive (
40M
) stars
OLBV
WN9--11
WN8
WNE
WC
SN.
Confirmation of these sequences have been provided by:
(1) the evolutionary progression from
OfOfpe
WNLha
(Crowther & Bohannan 1996); (2) an intimate relationship
between (less luminous) WN9--11 stars and LBVs, and between (more
luminous) O3If, O3If/WN6 stars and WN6 stars amongst LMC stars
(Crowther & Smith 1996b). While initial mass appears to
be the primary diagnostic, other effects (e.g. stellar rotation and metallicity)
also play a role. The main evolutionary channels between Of and WNL
are shown in Fig 5, illustrating the smooth morphological
sequence to higher wind densities. The sequence from LBVs to WN8 stars
common to Galactic and LMC progenitors is shown in Fig 5(a),
with WN10 stars probably dormant LBVs (true for HDE269582).
Figure 5(b) shows the direct route from late Of to WN9ha,
followed by massive stars exclusively at high metallicity.
The direct evolutionary sequence from early Of to WN6ha followed by
higher mass stars in both the Galaxy
(e.g. Carina, NGC3603) and LMC (30 Dor) is shown in Figure 5(c),
observationally supported by Rauw et al. (1996) and Drissen et al. (1995).
García-Segura et al. (1996) have recently demonstrated that
nebulae are potentially excellent probes for investigating previous
evolutionary histories.
Figure 6: Comparison of the observed chemistries (log C/N versus log C/He
by number)
of the WN/C stars (filled-in circles) with normal WN (HD192163)
and WC (HD165763) stars (open stars) from Crowther et al. (1995e)
with the theoretical predictions of Langer (priv. comm.) for stars of initial
mass 40 (dotted line) and 50
(dashed line)
at Z=0.02 during the W-R phase
While surface chemical enrichment is observed during the
LBV and Of stages, standard evolutionary theory (e.g. Maeder & Meynet 1987)
does not predict significant surface abundance changes until the W-R phase,
nor does it predict W-R stars of sufficiently low luminosity (especially
at low metallicities). However,
recent calculations by Fliegner & Langer (1995) indicate that
the surface compositions of stars more massive than 60
are significantly enriched at an early post-main sequence
phase by incorporating rotational mixing.
Further evidence for the need to incorporate additional mixing
processes in theoretical calculations results from the confirmation by
Crowther et al. (1995e) that WN/C stars are indeed genuine
intermediate WN--WC stars. In the standard theory, convection according
to the Schwarzschild criterion is assumed, which predicts a sharp
discontinuity between WN and WC phases. The
results of Crowther et al. (1995e) for three WN/C stars support
the theoretical predictions of Langer (1991 and priv. comm.) that
slow mixing could be produced through semi-convection, as indicated in
Fig. 6, with a timescale of 10
yr.
Looking to the future, a major advance in the analysis of massive stars will be the utilisation of combined interior and atmospheric models, as introduced by Schaerer et al. (1996), with a great potential for a clearer understanding of W-R and related stars (Schaerer 1996).