2016A&A...589A.130L

Context. Winds of AGB stars are thought to be driven by a combination of pulsation-induced shock waves and radiation pressure on dust. In dynamic atmosphere and wind models, the stellar pulsation is often simulated by prescribing a simple sinusoidal variation in velocity and luminosity at the inner boundary of the model atmosphere.
Aims. We experiment with different forms of the luminosity variation in order to assess the effects on the wind velocity and mass-loss rate, when progressing from the simple sinusoidal recipe towards more realistic descriptions. This will also give an indication of how robust the wind properties derived from the dynamic atmosphere models are.
Methods. Using state-of-the-art dynamical models of C-rich AGB stars, a range of different asymmetric shapes of the luminosity variation and a range of phase shifts of the luminosity variation relative to the radial variation are tested. These tests are performed on two stellar atmosphere models. The first model has dust condensation and, as a consequence, a stellar wind is triggered, while the second model lacks both dust and wind.
Results. The first model with dust and stellar wind is very sensitive to moderate changes in the luminosity variation. There is a complex relationship between the luminosity minimum, and dust condensation: changing the phase corresponding to minimum luminosity can either increase or decrease mass-loss rate and wind velocity. The luminosity maximum dominates the radiative pressure on the dust, which in turn, is important for driving the wind. An earlier occurrence of the maximum, with respect to the propagation of the pulsation-induced shock wave, then increases the wind velocity, while a later occurrence leads to a decrease. These effects of changed luminosity variation are coupled with the dust formation. In contrast there is very little change to the structure of the model without dust.
Conclusions. Changing the luminosity variation, both by introducing a phase shift and by modifying the shape, influences wind velocity and the mass-loss rate. To improve wind models it would probably be desirable to extract boundary conditions from 3D dynamical interior models or stellar pulsation models.