Previous
abstract 
Next
abstract
Session 80 - Star Clusters in the Milky Way.
Display session, Friday, January 09
Exhibit Hall,
We have fit dynamically evolving Fokker-Planck models to the globular star cluster M15. Using mass-function data from the Hubble Space Telescope observations made by Sosin amp; King (1997) and Piotto et al. (1997), we have determined the global stellar mass function down to a mass of 0.1M_ødot. In addition to the mass function, the star-count and velocity-dispersion profiles, and millisecond pulsar accelerations are also well fit by the model. The Fokker-Planck models are far more successful than are King-Mitchie models in fitting the full set of observational data. The best-fitting model has a mass function with power-law indices of 1.35, 0.7, -1.0, and 0.9 (where 1.35 is the Salpeter index) over the mass intervals of 60 to 1.0 M_ødot, 1.0 to 0.55M_ødot, 0.55 to 0.35M_ødot, and 0.35 to 0.1M_ødot, respectively. The total mass of the cluster is 4\times 10^5 M_ødot, and the half-mass radius is 3.3 parsecs. At no projected radius in the cluster does the local mass function reproduce the input global mass function. At the half-mass radius, the local mass function closely approximates the global mass function for stellar masses below 0.4M_ødot. But for masses above 0.4M_ødot, the global and local mass functions differ considerably. The model contains approximately 7000 nonluminous objects (3% of the total mass) with individual masses of 1.4M_ødot, the majority of which lie within 6^\prime\prime (0.2 pc) of the cluster center. If these objects are neutron stars, the large inferred number suggest that either the upper main sequence mass function was flatter than the Salpeter mass function or that the progenitor stellar mass range for neutron star formation may extend to lower masses in low-metallicity systems than in high metallicity systems.
