Back to Articles
A Possible Origin of X-rays from O Stars
Feldmeier, A.; Puls, J.; Pauldrach, A.W.A.

Abstract

X-ray spectra of hot, massive stars provide convincing evidence for thermal emissin that extends far out into their stellar winds. Accordingly, strong shocks were proposed as sources of the X-ray emission, where shocks result from line-driven instability. We show from hydrodynamic simulations that the emission from individual shocks which grow out of initially small perturbations may fall one or two orders of magnitude below the observed flux. Instead, we find that mutual collisions of dense shells of gas formed in deep wind regions can lead to shocks with a much stronger emission which almost matches the observed flux.
The model predicts strong variability of the X-ray emission, which is not observed. We propose that - in contrast to the presently asumed spherical symmetry of the wind - the emission stems from a large number of indepndednt, radial cones so that fluctuations average out over the whole emitting volume.


Summary

It has been found that collisions of dense shells of gas can produce material hot enough to emit X-rays. A model of this phenomenon must match both X-ray luminosities and spectra. In addition, it has been found that an appropriate model must have the three following characteristics:

  • Wind absorption, since much of the X-rays will be produced close to the star
  • X-ray emission at large radii
  • Temperature stratisfication to achieve good fits to the data
While the X-rays observed from O and early-B stars are highly variable, they are almost never periodic.

This paper presents an extension of the reverse shock model proposed by Owocki, which was able to duplicate spectral shapes, but was always an order of magnitude or so off in terms of luminosity. The authors perform a 1D line-driven simulation and assume spherical symmetry. The authors find that their methods can produce matching X-ray fluxes. It is admitted that there would be extreme time-variability due to the assumption of spherical symmetry, which is accounted for by assuming that the entire star wouldn't go through a fluctuation at the same time, and that only a spherically-symmetric average could be reasonably assumed.


Back to Articles