An important property of red giants and AGB stars is that they have strong stellar winds. The rates of mass loss due to stellar winds have the values of for the red giants and of for the AGB stars. The latter is otherwise known as superwind (cf. Renzini 1981 and reference therein). The mechanism driving the strong stellar winds from the stars with very low surface temperature and faint radiation is still a matter of debate. For the cool red giants several theoretical models have been suggested to explain the great mass loss. These can be grouped into the following categories according to the powering mechanism: radiation pressure on dust grains (atoms,molecules); driving by wave pressure; driving by shock waves; and finally the pressure of hot gas powers the wind (cf. Castor 1981). For the AGB stars three suggestions can be found for the mechanism powering superwind that are based on the effect of envelope pulsations.
(ii) Pulsational models of AGB stars suggest that superwind might be caused by switching of pulsation from the first overtone to the fundamental mode, perhaps giving rise to sporadic episodes of mass ejection (Wood 1974;Tuchman et al.1979).
(iii) Pulsation-radiation pressure can induce envelope ejection (Wannier 1984; Wilson and Kowalsky 1987; de Jager and Nieuwenhuijzen 1988; Wood and Vassiliadis 1992; Feast 1992; Vassiliadis and Wood 1992).
In all the above investigations the contribution of turbulent pressure is neglected. However, if the turbulent pressure becomes important in the convective envelopes of red giants and AGB stars, the effect of turbulent pressure can cause not only a change in the internal structure, but also dynamical instabilities in the regions close to the surfaces which may affect mass ejection outwards.
The purpose of this paper is to investigate the mechanism driving strong stellar winds from the red giants and the AGB stars. In order to understand the occurrence of dynamical instabilities in the convective envelopes,we present at first a criterion for the dynamical instability. Using this criterion we discuss whether dynamical instabilities can occur or not in the regions close to the surfaces of these stars, which finally affect the mass ejection outwards.
© European Southern Observatory (ESO) 1997