Next Contents


Mechanical feedback from massive stars is a dominant driver of evolutionary processes in galaxies, and takes place on scales ranging from individual wind-driven bubbles to galactic superwinds. Our understanding of the feedback process is based on the standard evolutionary model for stellar wind- and supernova-driven bubbles (Pikel'ner 1968; Weaver et al. 1977): hot (log T / K ~ 6-7), low-density (n ~ 0.01 cm-3) gas is generated within a double-shock structure, and the pressure of this hot gas, chemically enriched by the stellar products, drives the growth of the thin, radiatively-cooled shell. In the adiabatic model, energy loss from the hot gas is negligible, yielding simple analytic expressions for the shell radius R and expansion velocity v as a function of time t:

Equation 1 (1)

For a stellar wind-driven bubble, the mechanical luminosity L = 1/2 dot{M} vinfty2, where dot{M} and vinfty are the wind mass-loss rate and terminal velocity, respectively. For OB associations, supernovae (SNe) quickly dominate over winds, in which case L = N* E51 / te (e.g., Mac Low & McCray 1988), where N* and E51 are the total number of SNe and SN energy, respectively, and te is the time over which the SNe occur.

The fate of the interior hot gas is crucial to the phase balance and enrichment of the interstellar and intergalactic media. Is stellar feedback indeed the source of the diffuse, hot gas in the ISM? Do galactic superwinds from starbursts eject metal-enriched gas from galaxies? How does mechanical feedback affect the structure of the ISM and porosity, for example, to ionizing radiation? Do superbubbles trigger renewed star formation?

Determining the relevance of the standard, adiabatic model for shell evolution is clearly critical in answering these fundamental questions. Several tests can be applied: 1) Comparing the observed vs predicted dynamics of individual bubbles and superbubbles; 2) Comparing the observed vs predicted statistical properties of superbubble populations, for example size, velocity, and energy distributions; 3) Identifying spatial correlations of superbubbles with the progenitor OB associations and their relics; and 4) Testing the observed vs predicted dynamics and properties of galactic superwinds. The last is presently more difficult and the subject of entire reviews in its own right (e.g., Heckman 2002). Therefore I will discuss here only the first three tests.

Next Contents