Formation and pre-MS evolution of massive stars with growing accretion

Abstract

We briefly describe the three existing scenarios for forming massive stars and emphasize that the arguments often used to reject the accretion scenario for massive stars are misleading. It is usually not accounted for the fact that the turbulent pressure associated to large turbulent velocities in clouds necessarily imply relatively high accretion rates for massive stars. We show the basic difference between the formation of low and high mass stars based on the values of the free fall time and of the Kelvin-Helmoltz timescale, and define the concept of birthline for massive stars. Due to D-burning, the radius and location of the birthline in the HR diagram, as well as the lifetimes are very sensitive to the accretion rate \dot M_accr. If a form \dot M_accr=A (M/M_sun)^phi is adopted, the observations in the HR diagram and the lifetimes support a value of A=10^{-5} M_sun/yr and a value of phi>1. Remarkably, such a law is consistent with the relation found by Churchwell (1998) and Henning et al. (2000) between the outflow rates and the luminosities of ultra--compact HII regions, if we assume that a fraction 0.15 to 0.3 of the global inflow is accreted. The above relation implies high \dot M_accr=10^{-3} M_sun/yr for the most massive stars. The physical possibility of such high \dot M_accr is supported by current numerical models. Finally, we give simple analytical arguments in favour of the growth of \dot M_accr with the already accreted mass. We also suggest that due to Bondi-Hoyle accretion, the formation of binary stars is largely favoured among massive stars in the accretion scenario.