2010ApJ...711.1263C

We compare the observed bivariate distribution of masses (M) and ages (τ) of star clusters in the Large Magellanic Cloud (LMC) with the predicted distributions g(M, τ) from three idealized models for the disruption of star clusters: (1) sudden mass-dependent disruption, (2) gradual mass-dependent disruption, and (3) gradual mass-independent disruption. The model with mass-independent disruption provides a good, first-order description of these cluster populations, with g(M, τ) {vprop} M ^βτ^γ, β = -1.8±0.2 and γ = -0.8±0.2, at least for clusters with ages τ ≲ 10⁹ yr and masses M ≳ 10³ M_☉(more specifically, τ ≲ 10⁷(M/10² M_☉)^1.3 yr). This model predicts that the clusters should have a power-law luminosity function, dN/dL {vprop} L^–1.8, in agreement with observations. The first two models, on the other hand, fare poorly when describing the observations, refuting previous claims that mass-dependent disruption of star clusters is observed in the LMC over the studied M-τ domain. Clusters in the SMC can be described by the same g(M, τ) distribution as for the LMC, but with smaller samples and hence larger uncertainties. The successful g(M, τ) model for clusters in the Magellanic Clouds is virtually the same as the one for clusters in the merging Antennae galaxies, but extends the domain of validity to lower masses and to older ages. This indicates that the dominant disruption processes are similar in these very different galaxies over at least τ ≲ 10⁸ yr and possibly τ ≲ 10⁹ yr. The mass functions for young clusters in the LMC are power laws, while that for ancient globular clusters is peaked. We show that the observed shapes of these mass functions are consistent with expectations from the simple evaporation model presented by McLaughlin & Fall.