The critical mass for the overstability of massive zero-age main sequence stars

Abstract

The linear stability of massive extreme Pop I stars at the main sequence is investigated with a full nonadiabatic treatment. Contrary to earlier beliefs these stars become vibrationally unstable only beyond a critical mass as high as 440 M© at the main sequence. It has been maintained now for a very long time since the classical work by Ledoux (1941) and by Schwarzschild and Härm (1959) that stars in excess of ~ 100 M© cannot exist on the ground of their overstability alone. According to these results the nuclear energizing of pulsations in the stellar core -the e-mechanism -overcomes the damping and is believed to essentially disrupt the star or to cause strong mass loss which reduces their lifetimes effectively. On the other hand, evidence for the existence of stars with masses well in excess of 100 M 0 has come from observations of the brightest stars in the galaxies of the Local Group (Humphreys, 1982; De Jager, 1980; Massey and Hutchings, 1983) and from the interpretation of Wolf-Rayet stars (Maeder, 1982), Ρ Cygni stars and Hubble-Sandage variables, etc. (Humphreys, 1984). The extreme phenomena like η Carina in our galaxy and R136a in the LMC call for main sequence masses in excess of hundred solar masses. Langer et al., 1985). We have taken special care in using identical input physics both in stellar evolution code and in the pulsation code. Motivated by the apparent discrepancy between theory and observations concerning the upper mass limit, we reconsider in this paper the linear theory for radial pulsations of massive stars. The linear stability analysis is performed with a modified version of the Los Alamos Nonadiabatic Code (LNA). The major modifications introduced to the LNA code are (i) construction of the equilibrium model and (ii) inclusion of a nuclear energy generation term into the energy equation. For our computations the equilibrium model is constructed with the Göttingen stellar evolution code (cf.In the LNA code a first guess to the eigenfrequency is obtained by the use of the quasiadiabatic approximation and a first order expansion about the adiabatic mode. From Dalsgaard and S. Frandsen (eds.), Advances in Helio-and Asteroseismology, 399-402. © 1988 by the I AU. 399 J. Christensen-

Massive Stars in the Local Group: Implications for Stellar Evolution and Star Formation

“…Early estimates of this limit were as low as 60 M (Schwarzschild & Harm 1959). Modern estimates, however, place this limit as high as 440 M (Klapp et al 1987), although this is still based upon the same classical perturbation linearization methods used by Eddington. Recent "nonlinear" analysis suggests that the mass loss from such instabilities would only be comparable to the mass loss of radiatively driven stellar winds (Appenzeller 1987).…”