In young dense clusters repeated collisions between massive stars may lead to the formation of a very massive star (above 100 Msun). In the past the study of the long-term evolution of merger remnants has mostly focussed on collisions between low-mas
s stars (up to about 2 Msun) in the context of blue-straggler formation. The evolution of collision products of more massive stars has not been as thoroughly investigated. In this paper we study the long-term evolution of a number of stellar mergers formed by the head-on collision of a primary star with a mass of 5-40 Msun with a lower mass star at three points in its evolution in order to better understand their evolution. We use smooth particle hydrodynamics (SPH) calculations to model the collision between the stars. The outcome of this calculation is reduced to one dimension and imported into a stellar evolution code. We follow the subsequent evolution of the collision product through the main sequence at least until the onset of helium burning. We find that little hydrogen is mixed into the core of the collision products, in agreement with previous studies of collisions between low-mass stars. For collisions involving evolved stars we find that during the merger the surface nitrogen abundance can be strongly enhanced. The evolution of most of the collision products proceeds analogously to that of normal stars with the same mass, but with a larger radius and luminosity. However, the evolution of collision products that form with a hydrogen depleted core is markedly different from that of normal stars with the same mass. They undergo a long-lived period of hydrogen shell burning close to the main-sequence band in the Hertzsprung-Russell diagram and spend the initial part of core helium burning as compact blue supergiants.
We present Sapporo, a library for performing high-precision gravitational N-body simulations on NVIDIA Graphical Processing Units (GPUs). Our library mimics the GRAPE-6 library, and N-body codes currently running on GRAPE-6 can switch to Sapporo by a
simple relinking of the library. The precision of our library is comparable to that of GRAPE-6, even though internally the GPU hardware is limited to single precision arithmetics. This limitation is effectively overcome by emulating double precision for calculating the distance between particles. The performance loss of this operation is small (< 20%) compared to the advantage of being able to run at high precision. We tested the library using several GRAPE-6-enabled N-body codes, in particular with Starlab and phiGRAPE. We measured peak performance of 800 Gflop/s for running with 10^6 particles on a PC with four commercial G92 architecture GPUs (two GeForce 9800GX2). As a production test, we simulated a 32k Plummer model with equal mass stars well beyond core collapse. The simulation took 41 days, during which the mean performance was 113 Gflop/s. The GPU did not show any problems from running in a production environment for such an extended period of time.
We consider the effect of mass segregation on the observable integrated properties of star clusters. The measurable properties depend on a combination of the dynamical age of the cluster and the physical age of the stars in the cluster. To investigat
e all possible combinations of these two quantities we propose an analytical model for the mass function of segregated star clusters that agrees with the results of N-body simulations, in which any combination can be specified. For a realistic degree of mass segregation and a fixed density profile we find with increasing age an increase in the measured core radii and a central surface brightness that decreases in all filters more rapidly than what is expected from stellar evolution alone. Within a Gyr the measured core radius increases by a factor of two and the central surface density in all filters of a segregated cluster will be overestimated by a similar factor when not taking into account mass segregation in the conversion from light to mass. We find that the $V-I$ colour of mass segregated clusters decreases with radius by about 0.1-0.2 mag, which could be observable. From recent observations of partially resolved extra-galactic clusters a decreasing half-light radius with increasing wavelength was observed, which was attributed to mass segregation. These observations can not be reproduced by our models. We find that the differences between measured radii in different filters are always smaller than 5%.
