Evolution of massive stars under new mass-loss rates for RSG: is the mystery of the missing blue gap solved?
Bernardo Salasnich 1,
Alessandro Bressan 2 and
Cesare Chiosi 3,1
Received 18 November 1997 / Accepted 3 September 1998
In this paper we present new models of massive stars based on recent advancements in the theory of diffusive mixing and a new empirical formulation of the mass-loss rates of red supergiant stars. The study is articulated in two main parts. First, by means of a simple diffusive algorithm, we amalgamate the results of complex studies on non local convection (overshooting region) by Xiong (1989) and Grossman (1996), and apply them to model the structure and evolution of massive stars in occurrence of mass loss by stellar winds according to the popular relationship by de Jager et al. (1988). Stars with initial mass in the range 6 to 120 and initial chemical composition [Z=0.008, Y=0.25] and [Z=0.020, Y=0.28] are followed from the zero age main sequence till core He-exhaustion. Particular attention is paid to the 20 star as the prototype of the evolution of massive stars in the luminosity (mass) interval in which both blue and red supergiants occur in the HR diagram (HRD). The models confirm that, in the evolution of a massive star with mass loss, the dimension of the H-exhausted core and the efficiency of intermediate mixing strongly affect the evolution during the subsequent core He-burning phase, the extension of the blue loops in particular. However, despite the new mixing prescription, also these models share the same problems of older models in literature as far as the interpretation of the observational distribution of stars across the HRD is concerned. In the second part, we examine possible causes of the failure and find that the rate of mass loss for the red supergiant stages implied by the de Jager et al. (1988) relationship under-estimates the observational values by a large factor. Revising the whole problem, we adopt the recent formulation by Feast (1991) based on infrared data, and take also into account the possibility that the dust to gas ratio varies with the stellar luminosity. Stellar models are then calculated with the new prescription for the mass-loss rates during the red supergiant stages in addition to the new diffusive algorithm. The models now possess very extended loops in the HRD and are able to match the distribution of stars across the HRD from the earliest to the latest spectral types both in the Milky Way, LMC and SMC. During the loop phase the surface abundance of helium is significantly enhanced with respect to the original value as suggested by observational data for blue supergiant stars. Finally, because the surface H-abundance can be lowered to the limit adopted to start the Wolf-Rayet phase (WNL type), we suggest that a new channel is possible for the formation of low luminosity WNL stars, i.e. by progenitors whose mass can be as low as 20 , that have evolved horizontally across the HRD following the blue-red-blue scheme and suffering large mass loss during the red supergiant stages.
Key words: stars: mass-loss stars: evolution
Send offprint requests to: B. Salasnich
This article contains no SIMBAD objects.
© European Southern Observatory (ESO) 1999
Online publication: December 22, 1998