A new interpretation of luminous blue stars

A major revision of current theoretical ideas about the brightest blue stars must be made if Carson's (1976) radiative opacities are adopted in stellar models. Unlike earlier opacities, these exhibit a large 'bump' due to CNO ionization, which leads to very strong central condensation, convective instability, and pulsational instability in hot diffuse stellar envelopes. Despite a number of theoretical uncertainties, the new picture of the structure of very luminous stars is reasonably successful in accounting for a variety of previously unexplained observations. The stellar models for the phase of core hydrogen burning predict large radii and rather cool effective temperatures for O stars and a spreading out of the main-sequence band in the H-R diagram toward luminous cool supergiants for masses higher than about 20 solar masses. In massive X-ray binary systems, circular orbits and supergiant-like visual companions are expected to be quite common. Long-period variability is predicted to exist for massive blue supergiants of luminosity class Ia. The models for helium stars predict large radii and rather cool effective temperatures for Wolf-Rayet stars, as well as multimodal pulsational instability and, possibly, surface turbulence for these stars.