Effects of a Stochastic Initial Mass Function on the Upper Main Sequence Band

Abstract

The theoretical (M_b versus Log T_e) HR diagram for the brigh test galactic OB stars shows an upper boundary for the luminosity, which is characterized by a decreasing luminosity with decreasing effective temperature (Humphreys and Davidson, 1979). The existence of this limit was interpreted by Chiosi et al. (1978) as due to the effect of mass loss by stellar wind on the evolution of most massive stars in core H-burning phase. In fact, evolutionary models calculated at constant mass cover a wider and wider range in effective temperature as the initial mass increases during the main sequence phase. On the contrary, sufficiently high mass-loss rates make the evolutionary sequences of most massive stars (M 60≥M_⊙) shrink toward the zero age main sequence whenever, due to mass loss, CNO processed material is brought to the surface (Chiosi et al., 1978; de Loore et al., 1978; Maeder, 1980). In such a case the main sequence band is found to coincide with the observational one if mass-loss rates of the order of 10^-6 M_⊙/yr during the whole main sequence phase are used. As a consequence of this, the experimental upper boundary for the luminosity has been regarded as a way to constrain the mass-loss rate within the range of observational uncertainty.