Faculty Publications


An opacity mechanism for the pulsations of OB stars

Document Type



Atomic processes, Stars: early, stars: variables: cepheids, stars: variables: δ scuti, Type stars: oscillations

Journal/Book/Conference Title

Astrophysical Journal





First Page


Last Page



We propose that the sudden appearance of a tremendous number of same-shell transition iron lines, as the temperature rises above 100,000 K, gives a high sensitivity of the opacity to temperature at the very low densities found in OB giants. This produces k effect pulsations. Opacities need to increase quickly to a factor of 5 above the Cox-Tabor values at temperatures up to 250,000 K. These increases are proposed by Iglesias & Rogers using an improved accounting for all the iron lines and their widths. The reason not all B stars pulsate could be because of a slight primordial deficit in the iron abundance in the surface layer (2 × 106 of the mass). This deficiency reduces the opacity and its sensitivity to temperature and reduces the k effect driving. A slight amount of iron concentration by radiative levitation could cause a star to pulsate even if it did not originally have enough primordial iron to cause this opacity mechanism to operate. Then any slow slight mixing caused by the shear of unstable nonradial pulsations could restabilize the pulsations as actually observed in α Vir and β CMa. Rapid levitation and mixing for the very luminous B stars with their very low density envelopes could even explain the puzzling luminous blue variables that pulsate only a few cycles before they stabilize again. Large amplitude pulsations like those seen in BW Vul would indicate a somewhat larger primordial iron abundance compared to all other B stars, with pulsations limited only at this star's unique large amplitude. Modes found unstable for a typical 12 M⊙, 20,900 L⊙, 22,750 K, 1000 zone stellar model with X = 0.70 and Z = 0.02 in linear nonadiabatic calculations are a radial mode and a few nonradial (g-modes of low degree and order with periods near 0.3 day. Thus multimode behavior is theoretically expected for this pulsation mechanism, and for some B stars only a few nonradial modes (possibly selected by rotation) may survive to observable amplitudes.

Original Publication Date


DOI of published version