Lecture #5b-- Winds that Stir the ISM



1) Main sequence stars do not evaporate
The scale height of a dwarf is much less than its radius, which also means that the thermal speed at the surface is much less than the escape speed. This also means that radiative cooling of the surface layers forces the orbits to deviate very far from being virialized, so the star has a kind of "spherical cap" surrounding it that prevents the escape of gas. Thus, dwarfs should not have winds-- but they do. Any they come in two flavors: gas pressure driven (cool stars) and radiatively driven (hot stars)
2) Gas pressure driven winds from cool stars
The defining difference between "cool stars" and "hot stars" is that cool stars have convection zones near their surfaces, whereas hot stars hide their convection deep in their cores. This means that cool stars have a means to twist and stir their surface magnetic fields, causing a release of heat that actually occurs above the surface of the star where the density is very low. Since low density gas does not cool effectively, it can get extremely hot-- allowing the thermal speeds to approach escape velocity. Indeed, the "corona" of the Sun can get as hot as its core where fusion occurs (but the density is too low in the corona for fusion to be an important heat source). So although cool stars cannot evaporate into winds from their surfaces, they can evaporate into winds from the corona-- as long as their is a heat source, there will be a wind. The mass-loss rate is extremely low, and in the main-sequence lifetime of the Sun it will lose less than 1/1000 of its mass (hence the wind is not evolutionarily significant). Wind speeds are typically of order the escape speed or a bit less, so in the range 300 - 800 km/s. 3) Radiatively driven winds from hot stars
Hot stars do not have surface convection, so shouldn't make coronae (though it appears some do find various ways to generate X-ray emitting material via wind stream collisions and shocks). However, they have incredible luminosities, approaching the so-called "Eddington limit" where the radiative force on free electrons approaches the force of gravity on their associated protons. This means the star is partially supported by radiation pressure, so if there was some way to increase the radiative force on the particles above the surface, a wind could be generated even though the temperature is photospheric and the escape speed is highly supersonic. The mechanism that provides the added push is line opacity, whose importance is magnified by Doppler shifs in a supersonic wind. Hence the wind bootstraps itself into being, because it can find the driving it needs to get to high velocity if the act of accelerating it allows it to receive the push necessary to accelerate it. Typically, the luminosity scales like mass cubed, and the mass-loss rate scales like luminosity to the 1.7 power, so the mass-loss rate scales like mass to the 5 power (yes I said the 2 power in the homework, but that's not right, my apologies). This also means that the fractional mass lost over the main-sequence lifetime scales like mass squared (which is what I meant to say in the homework, my mistake).