We overview the GRAPE-6 project, a follow-up of the teraflops GRAPE-4 project. GRAPE-6 will be completed by 1999-2000 and its planned peak speed is 200 Tflops. Its architecture will be largely similar to that of GRAPE-4, which is a specialized hardware to calculate the gravitational interaction between particles. The improvement of the speed will mainly come from the advance in the silicon semiconductor technology. GRAPE-6 will enable us to directly simulate the evolution of star clusters with up to 1 million stars.

In the dynamical study of globular star clusters, five types of dynamical models are in current use. The following list includes recent highlights, some of which are developed in these proceedings by other authors.

We have developed a massively parallel special-purpose computer system for N-body simulations, GRAPE-4 (GRAvity-PipE 4). The GRAPE-4 system is designed for high-accuracy simulations of dense stellar systems. The GRAPE-4 calculates gravitational forces, their derivatives in time and potential energies. It has a hardware for prediction of positions and velocities, which is used for the individual timestep scheme. We integrated 1692 chips of 640 megaflops performance to achieve the peak speed of 1.08 teraflops.

I present a summary of results from various HST photometric studies of the dense central regions of Galactic globular clusters that my collaborators and I have carried out over the last 6 years. The dataset includes short exposures of 47 Tuc, M15, M3 and M13 obtained with the aberrated Planetary Camera-I (PC-I) and F555W (“V”) and F785LP (“I”) filters, as well as post-refurbishment Wide Field Planetary Camera 2 (WFPC2) snapshots of the post core collapse clusters M15, M30, and NGC 6624 in F336W (“U”), F439W (“B”), and V. Recently, a very deep, doubly oversampled PC-I U image of the core of 47 Tuc, and accompanying B and V images, have also been analyzed. In addition, we have carried out extensive checks of incompleteness and photometric error with the help of multiband image simulations that mimic the relevant characteristics of the HST PC-I and WFPC2 images: empirical point spread function, crowding effects based on a realistic density profile and stellar luminosity function (LF), noise, undersampling, A/D saturation, etc..

There was a time, more than a century ago, when Astronomy & Astrophysics was not a European Journal yet, but still an American one. In its Vol. XII, Bailey (1893) published, from photographic observations of two globular clusters (GCs), ω Centauri and 47 Tucanae, what were probably the first extensive star counts, which represent the oldest observational constraint for the study of the structure of GCs. Bailey’s counts, together with some new material concerning other clusters were used by Pickering (1897) in the first important comparisons between observed and theoretical profiles in order to study the radial distribution of stars in clusters.

Binary stars in a globular cluster (hereafter, GC) may be primordial (i.e. formed along with the cluster), or the result of cluster dynamics. “Dynamical” binaries can result from conservative three-body encounters (e.g. Spitzer, 1987) if a third star can carry away enough kinetic energy to leave two others bound, or from dissipative two-body encounters, if two stars happen to pass within a few stellar radii of one other (Fabian, Pringle, & Rees, 1975). Such non-primordial systems are likely to be found primarily in evolved GC cores, both because conditions are more favorable for making them there, and because of mass segregation. Knowledge of the formation process allows reasonable estimates to be made of their mass and energy distributions. The initial spatial, mass, and energy distributions of primordial binaries, on the other hand, are largely unknown.

Dynamical models of star clusters are maturing in the sense that effects other than simple point particle dynamics are taken into account. We summarize the relevance of, and prospects for, this new generation of N-body models.

We summarize recent work on the structure and dynamics of the Galactic bar and inner disk. Current work focusses on constructing a quantitative model which integrates NIR photometry, source count observations, gas kinematics, stellar dynamical observations, and microlensing. Some avenues for future research are discussed.

We briefly describe the performance and accuracy of the Marseille GRAPE-3 systems. We first give the description of the configurations. We then outline the modifications necessary to optimize the tree code for use on a GRAPE system, and discuss performance and accuracy issues. We summarize the results of several science projects executed with our hardware and software.

Kuhn & Miller (1989, KM hereafter) proposed that a small stellar system in orbit around a large galaxy could be heated through resonant coupling between the orbital motion and internal modes of oscillation of the satellite. They further argued that the apparently large mass-to-light (M/L) ratios of dwarf spheroidal (dSph) galaxies around the Milky Way (e.g., Armandroff, Olszewski & Pryor 1995) could be illusory because the satellites are far from virial equilibrium as a consequence of such resonant heating. We note that if the true M/Ls are small, some dSphs have much more kinetic energy than is required to unbind them; a resonant pumping mechanism which adds a modest amount of energy each cycle could not achieve this (Pryor 1996).

The Keck 10-m telescope has begun to produce many new findings in a field of stellar spectroscopy with the help of the High Resolution Spectrograph (HIRES). The other large telescopes such as the Subaru, VLT, and Gemini are now being constructed and expected to perform high-resolution spectroscopy in a near future within several years or so. These highresolution spectroscopic studies will be carried out based on studies which have been made mainly with existing telescopes with aperture smaller than 4 m.

I can compute arbitrary-abundance models and spectra at high resolution using millions of atomic and diatomic molecular lines. Examples are given for Sakurai’s object and for a λBoo star. I am continuing to improve the input line data.

For the past seven years a large, international team of astrophysicists and atomic physicists has been engaged in the systematic exploration and analysis of the extraordinary spectrum of one chemically-peculiar star, chi Lupi (B9.5pHgMn + A2 Vm). The observational data consist of 345 Ã of the star’s ultraviolet spectrum acquired with the Goddard High Resolution Spectrograph (GHRS) on board the Hubble Space Telescope (HST) at a resolving power of approximately 90,000 and an average S/N ratio of about 75 in the continuum. These data were supplemented with new optical-wavelength echelle spectra obtained at the AAT and at KPNO[l]. We have named this study “The x Lupi Pathfinder Project”, because it is in many respects similar to the initial exploration of a strange new land – never before has the ultraviolet spectrum of an ultra-sharp-lined star been observed at this level of detail and accuracy.

The purpose of this paper is to define a set of “standard”, “reference” or “comparison” stars, according to their use, of spectral type A0V in order to be used for further studies in the domain of peculiar stars as well as normal stars of this Teff and luminosity range.

The quest for accurate fundamental parameters (M, R, Teff log g, v sin i) of CP stars has been a lasting problem ever since the nature and origin of these stars has been discussed. The main reason for the difficulty lies in the anomalous colours of these stars, linked with their abundance anomalies which are responsible for complicated blocking and backwarming effects. In the following, we review the Teff determinations proposed so far, and show how the Hipparcos results improve the radius and log g determinations of nearby Ap stars.

The NLTE abundance analyses concerning late-B through A stars carried out before 1990 have been well summarized by Takada-Hidai (1991), which may be consulted for details; about 10 elements had been studied in NLTE by that time.

In an attempt to make the present summary comprehensible yet sufficiently concise I will risk some over-simplification: further details are given in Seaton (1997, 1998 — Papers I and II).

In a star let Fv dv be the outward flux of radiant and let an atom of element k present an effective area of σv(k) for absorption of radiation. Then the momentum absorbed per atom and per-unit area and unit time is G(k) = (1/c) ∫ Fvσv(k) dv which is just the force acting on the atom. The radiative acceleration is grad(k) = G(k)/M(k) where M(k) is the atom’s mass.

In the light of recent results from helioseismology, the abundance anomalies seen at the surface of Bp, Ap, Am and Fm stars appear as the measurable superficial effect of a process that has much more general impact.

We present a brief review about recent progresses concerning the study of diffusion processes in CP stars. The most spectacular of them concerns the calculation of radiative accelerations in stellar envelopes for which an accuracy better than 30% can now be reached for a large number of ions. This improvement is mainly due to huge and accurate atomic and opacity data bases available since the beginning of the 90’s. Developments of efficient computational methods have been carried out to take advantage of these new data. These progresses have, in turn, led to a better understanding of how the element stratification is building up with time. A computation of self-consistent stellar evolution models, including time-dependent diffusion, can now be within the scope of the next few years. However, the progresses previously mentioned do not apply for stellar atmospheres and upper layers of envelopes.