The impact of massive stars on the interstellar medium
In this thesis I present results of numerical simulations carried out with a two-dimensional radiation hydrodynamics code in order to study the impact of massive stars on their surrounding interstellar medium. The evolution of the circumstellar gas is examined for two model stars with different evolutionary tracks and stellar parameters: The first star has an inital mass of 60 solar masses and evolves from a main-sequence O star through luminous blue variable and Wolf-Rayet phases, until it ultimately explodes as a supernova of Type II. The second model star, initially having a mass of 35 solar masses, is supposed to undergo the evolution from the main sequence to the red supergiant and finally the Wolf-Rayet phase, until it also explodes as a supernova. In the 60 solar mass case the interaction of the photoionized HII region with the stellar wind bubble forms a variety of interesting structures like shells, clouds, fingers, and spokes. These results demonstrate that complex structures found in HII regions are not necessarily relics from the time before the gas became ionized but may result from dynamical processes during the course of the HII region evolution. In the 35 solar mass case structure formation in the circumstellar gas during the early main-sequence evolution is much less pronounced because of the lower mechanical wind luminosity of the star. On the other hand, since the shell-like structure of the HII region is largely preserved, effects that rely on this symmetry become more important. I have also analyzed the transfer and deposit of the stellar wind and radiation energy into the circumstellar medium. At the end of the stellar lifetime 0.37 % of the energy released by the 60 solar mass star as Lyman continuum radiation and stellar wind has been transferred to the circumstellar gas. From this fraction 35 % is kinetic energy of bulk motion, 40 % is thermal energy, and the remaining 25 % is ionization energy of hydrogen. The respective values in the 35 solar mass case are 10 %, 36 %, and 54 %, for a total energy transfer efficiency of 1 %. The sweeping up of the slow red supergiant wind by the fast Wolf-Rayet wind in the 35 solar mass case produces remarkable morphological structures and emission signatures, which are compared with existing observations of the Wolf-Rayet bubble S308, whose central star has probably evolved in a manner very similar to the model star. My model reproduces the correct order of magnitude of observed X-ray luminosity, the temperature of the emitting plasma, and the limb brightening of the X-ray intensity profile. This is remarkable, because current analytical and numerical models of Wolf-Rayet bubbles fail to consistently explain these features. A key result is that almost the entire X-ray emission in this stage comes from the shell of red supergiant wind swept up by the shocked Wolf-Rayet wind rather than from the shocked Wolf-Rayet wind itself as hitherto assumed and modeled. This offers a possible solution to what is called the "missing wind problem" of Wolf-Rayet bubbles.