The birth and early evolution of star clusters

Star clusters are observed to form in a highly compact state and with low star-formation efficiencies, and only 10 per cent of all clusters appear to survive to middleand old-dynamical age. If the residual gas is expelled on a dynamical time the clusters disrupt. Massive clusters may then feed a hot kinematical stellar component into their host-galaxy’s field population thereby thickening galactic disks, a process that theories of galaxy formation and evolution need to accommodate. If the gas-evacuation time-scale depends on cluster mass, then a power-law embedded-cluster mass function may transform within a few dozen Myr to a mass function with a turnover near 10 M⊙, thereby possibly explaining this universal empirical feature. Discordant empirical evidence on the mass function of star clusters leads to the insight that the physical processes shaping early cluster evolution remain an issue of cutting-edge research. 1. Early cluster evolution The star-formation efficiency (sfe), ǫ ≡ Mecl/(Mecl+Mgas), where Mecl,Mgas are the mass in freshly formed stars and residual gas, respectively, is 0.2<∼ ǫ <∼ 0.4 (Lada & Lada 2003) implying that the physics dominating the star-formation process on scales < 10 pc is stellar feedback. Within this volume, the pre-cluster cloud core contracts under self gravity thereby forming stars ever more vigorously, until feedback energy suffices to halt the process (feedback-termination) (Weidner & Kroupa 2006). This occurs on one to a few crossing times (≈ 10 yr), and since each proto-star needs about 10 yr to accumulate about 95 per cent of its mass (Wuchterl & Tscharnuter 2003), the assumption may be made that the embedded cluster is mostly virialised at feedback-termination. Its stellar velocity dispersion,