Multiple stellar population mass loss in massive Galactic globular clusters

The degree of mass loss, that is the fraction of stars lost by globular clusters, and specifically by their different populations, is still poorly understood. Many scenarios of the formation of multiple stellar populations, especially the ones involving self-enrichment, assume that the first generation (FG) was more massive at birth than now in order to reproduce the current mass of the second generation (SG). This assumption implies that, during their long-term evolution, clusters lose around 90% of the FG. We tested whether such strong mass loss could take place in a massive globular cluster orbiting the Milky Way at 4 kpc from the centre that is composed of two generations. We performed a series of N-body simulations for 12 Gyr to probe the parameter space of internal cluster properties. We derive that, for an extended FG and a low-mass SG, the cluster loses almost 98% of its initial FG mass and the cluster mass can be as much as 20 times lower after a Hubble time. Furthermore, under these conditions, the derived fraction of SG stars, f(enriched), falls in the range occupied by observed clusters of similar mass (similar to 0.6 - 0.8). In general, the parameters that affect the highest degree of mass loss are the presence or absence of primordial segregation, the depth of the central potential, W-0,W- FG, the initial mass of the SG, M-SG(ini), and the initial half-mass radius of the SG, r(h, SG). Higher M-SG(ini) have not been found to imply higher final f(enriched) due to the deeper cluster potential well which slows down mass loss.