As it is well known, and as recently explicitly reiterated, e.g., in Castellani (1999), the widespread use of the Henyey algorithm in producing stellar models assures that different codes with identical input physics must produce quite similar results. In this paper we presented several evidences, as taken from the literature, supporting such an occurrence. Thus evolutionary models are as "good" as the adopted input physics is. In this paper we have however shown that models based on different, but all "reasonable", physics inputs may have sizable differences in the predicted luminosity of He burning low mass stars.
Testing these theoretical predictions on the absolute magnitudes given by the Hipparcos satellite for He burning stars clumping in the field, one finds several indications for preferring the luminosity predicted by the Padua code, as used in Girardi et al. (1998). Even if the argument is still open to further investigation, we feel that the discussion presented in this paper raises - at least - serious doubts on the capability of the most recent "updated input physics" of correctly predicting the luminosity of He burning stars as originated from progenitors which underwent strong electron degeneracy. Therefore giving a serious warning about the uncritical acceptance of theoretical predictions concerning the luminosity of HB stars in galactic globulars and, in turn, all the related quantities.
In this respect, we note that observational evidences concerning the luminosity of globular HB stars are far from being clear. Caputo et al. (1999) have recently discussed RR Lyrae variables in the globular M5 showing that their pulsational properties possibly suggest a luminosity lower than predicted by C99 models. However, Cassisi et al. (1999) and Salaris & Weiss (1998) used HB stars brighter than G98 predictions to derive cluster distance moduli which appear in excellent agreement with some independent evaluations based on Hipparcos parallaxes for field subdwarfs. Here we can only conclude that the matter deserves further investigation, to decide whether or not the mass loss can reconcile "most updated" stellar models with Hipparcos constraints for the clump of He burning stars in the solar neighborhood.
Before closing the paper, let us notice that all the computations discussed in the paper agree in predicting He burning stars systematically brighter when the metallicity is decreased. The theoretical predictions are illustrated, e.g., in Fig. 1 of G98, and Fig. 12 of C99. This has important implications for the method of distance determination based on clump stars, since it indicates that the clump population observed in nearby galaxies may have an intrinsic luminosity different from the local stars sampled by Hipparcos. In the case of the Magellanic Clouds, this difference may amount to 0.2 - 0.3 mag, as inferred by Cole (1998) and G98.
© European Southern Observatory (ESO) 2000
Online publication: January 31, 2000