It is typically assumed that radiation pressure driven winds are accelerated to an asymptotic velocity of V ~ v_esc, where v_esc is the escape velocity from the central source. We note that this is not the case for dusty shells and clouds. Instead, i
f the shell or cloud is initially optically-thick to the UV emission from the source of luminosity L, then there is a significant boost in V that reflects the integral of the momentum absorbed as it is accelerated. For shells reaching a generalized Eddington limit, we show that V ~ (4R_UV L/M_sh c)^1/2, in both point-mass and isothermal-sphere potentials, where R_UV is the radius where the shell becomes optically-thin to UV photons, and M_sh is the mass of the shell. The asymptotic velocity significantly exceeds v_esc for typical parameters, and can explain the ~1000-2000km/s outflows observed from rapidly star-forming galaxies and active galactic nuclei if the surrounding halo has low gas density. Similarly fast outflows from massive stars can be accelerated on few - 10^3 yr timescales. These results carry over to clouds that subtend only a small fraction of the solid angle from the source of radiation and that expand as a consequence of their internal sound speed. We further consider the dynamics of shells that sweep up a dense circumstellar or circumgalactic medium. We calculate the momentum ratio Mdot v/(L/c) in the shell limit and show that it can only significantly exceed ~2 if the effective optical depth of the shell to re-radiated FIR photons is much larger than unity. We discuss simple prescriptions for the properties of galactic outflows for use in large-scale cosmological simulations. We also briefly discuss applications to the dusty ejection episodes of massive stars, the disruption of giant molecular clouds, and AGN.
Binary white dwarf (WD) coalescences driven by gravitational waves or collisions in triple systems are potential progenitors of Type Ia supernovae (SNe Ia). We combine the distribution of 56Ni inferred from observations of SNe Ia with the results of
both sub-Chandrasekhar detonation models and direct collision calculations to estimate what mass WDs should be exploding in each scenario to reproduce the observations. These WD mass distributions are then compared with the observed Galactic WD mass distribution and Monte Carlo simulations of WD-WD binary populations. For collisions, we find that the average mass of the individual components of the WD-WD binary must be peaked at ~0.75Msun, significantly higher than the average WD mass in binaries or in the field of ~0.55-0.60Msun. Thus, if collisions produce a large fraction of SNe Ia, then a mechanism must exist that favors large mass WDs. On the other hand, in an old stellar population, collisions would naturally result in a class of low luminosity SNe Ia, and we suggest these may be related to 1991bg-like events. For sub-Chandrasekhar detonations, we find that the average mass of the exploding WDs must be peaked at ~1.1Msun. This is interestingly similar to the average sum of the masses in WD-WD binaries, but it is not clear (and should be further explored) whether double degenerate mergers would be sufficiently efficient at synthesizing 56Ni to match the observed yields. If not, then actual ~1.1Msun WDs would be needed for sub-Chandrasekhar detonations. Since such high mass WDs are produced relatively quickly in comparison to the age of the environments where SNe Ia are found, this would require either accretion onto lower mass WDs prior to ignition or a long timescale between formation of the ~1.1Msun WD and ignition (such as set by gravitational wave emission or binary interactions).
We use a high-temperature chemical network to derive the molecular abundances in axisymmetric accretion disk models around active galactic nuclei (AGNs) within 100 pc using simple radial and vertical density and temperature distributions motivated by
more detailed physical models. We explore the effects of X-ray irradiation and cosmic ray ionization on the spatial distribution of the molecular abundances of CO, CN, CS, HCN, HCO+, HC3N, C2H, and c-C3H2 using a variety of plausible disk structures. These simple models have molecular regions with a layer of X-ray dominated regions, a midplane without the strong influence of X-rays, and a high-temperature region in the inner portion with moderate X-ray flux where families of polyynes (C$_{rm n}$H$_{2}$) and cyanopolyynes can be enhanced. For the high midplane density disks we explore, we find that cosmic rays produced by supernovae do not significantly affect the regions unless the star formation efficiency significantly exceeds that of the Milky Way. We highlight molecular abundance observations and ratios that may distinguish among theoretical models of the density distribution in AGN disks. Finally, we assess the importance of the shock crossing time and the accretion time relative to the formation time for various chemical species. Vertical column densities are tabulated for a number of molecular species at both the characteristic shock crossing time and steady state. Although we do not attempt to fit any particular system or set of observations, we discuss our models and results in the context of the nearby AGN NGC 1068.
Over a broad range of initial inclinations and eccentricities an appreciable fraction of hierarchical triple star systems with similar masses are essentially unaffected by the Kozai-Lidov mechanism (KM) until the primary in the central binary evolves
into a compact object. Once it does, it may be much less massive than the other components in the ternary, enabling the eccentric Kozai mechanism (EKM): the mutual inclination between the inner and outer binary can flip signs driving the inner binary to very high eccentricity, leading to a close binary or collision. We demonstrate this Mass-loss Induced Eccentric Kozai (MIEK) mechanism by considering an example system and defining an ad-hoc minimal separation between the inner two members at which tidal affects become important. For fixed initial masses and semi-major axes, but uniform distributions of eccentricity and cosine of the mutual inclination, ~10% of systems interact tidally or collide while the primary is on the MS due to the KM or EKM. Those affected by the EKM are not captured by earlier quadrupole-order secular calculations. We show that fully ~30% of systems interact tidally or collide for the first time as the primary swells to AU scales, mostly as a result of the KM. Finally, ~2% of systems interact tidally or collide for the first time after the primary sheds most of its mass and becomes a WD, mostly as a result of the MIEK mechanism. These findings motivate a more detailed study of mass-loss in triple systems and the formation of close NS/WD-MS and NS/WD-NS/WD binaries without an initial common envelope phase.
