We conducted a search for very short-period transiting objects in the publicly available Kepler data set. Our preliminary survey has revealed four planetary candidates, all with orbital periods less than 12 hr. We have analyzed the data for these candidates using photometric models that include transit light curves, ellipsoidal variations, and secondary eclipses to constrain the candidates' radii, masses, and effective temperatures. Even with masses of only a few Earth masses, the candidates' short periods mean that they may induce stellar radial velocity signals (a few m s(-1)) detectable by currently operating facilities. The origins of such short-period planets are unclear, but we discuss the possibility that they may be the remnants of disrupted hot Jupiters. Whatever their origins, if confirmed as planets, these candidates would be among the shortest-period planets ever discovered. Such planets would be particularly amenable to discovery by the planned TESS mission.

To date, most simulations of the final accretion of the terrestrial planets have assumed that all collisions lead to mergers. Recent hydrodynamic simulations of impacts between planetary mass bodies (Leinhardt, Z.M., Stewart, S.T. [2012]. Astrophys. J. 745,79; Genda, H., Kokubo, E., Ida, S. [2012]. Astrophys J. 744, 137) have parameterized the outcome of planetary collisions in terms of the masses and velocities of the colliding bodies. Using these results, it is now possible to simulate late-stage planetary growth using a more realistic model for collisions. Here, we describe results of eight N-body simulations of terrestrial planet formation that incorporate collisional fragmentation and hit-and-run collisions. The results are compared to simulations using identical initial collisions in which all collisions were assumed to result in mergers (Chambers, J.E. [2001]. Icarus 152, 205-224). The new simulations form 3 to 5 terrestrial planets moving on widely spaced orbits with growth complete by 400 My. The mean time for Earth-like planets to reach half their final mass is 17 My, comparable to the time in simulations without fragmentation. However, the prolonged sweep up of collision fragments lengthens the mean time required for Earth analogues to become fully formed to 159 My. The final planets have somewhat smaller masses m and eccentricities e when fragmentation is included. Masses are particularly reduced in the region now occupied by Mars. The final distributions of m, e and semi-major axis are similar to the terrestrial planets of the Solar System, but the strong concentration of mass in the narrow zone occupied by Earth and Venus is not reproduced. Collisional fragmentation is likely to preferentially eject silicate-rich mantle material leaving a target enriched in iron-rich core material. However, large bodies often reaccrete silicate-rich mantle fragments at a later time, leaving their final composition largely unchanged. The final core mass fractions of all but one planet formed in the simulations lie in the range 0.25-0.37 assuming an initial mass fraction of 0.3. (C) 2013 Elsevier Inc. All rights reserved.

In the core accretion model for giant planet formation, a solid core forms by coagulation of dust grains in a protoplanetary disk and then accretes gas from the disk when the core reaches a critical mass. Both stages must be completed in a few million years before the disk gas disperses. The slowest stage of this process may be oligarchic growth in which a giant-planet core grows by sweeping up smaller, asteroidsize planetesimals. Here, we describe new numerical simulations of oligarchic growth using a particle-in-a-box model. The simulations include several processes that can effect oligarchic growth: (i) planetesimal fragmentation due to mutual collisions, (ii) the modified capture rate of planetesimals due to a core's atmosphere, (iii) drag with the disk gas during encounters with the core (so-called "pebble accretion"), (iv) modification of particle velocities by turbulence and drift caused by gas drag, (v) the presence of a population of mm-to-m size "pebbles" that represent the transition point between disruptive collisions between larger particles, and mergers between dust grains, and (vi) radial drift of small objects due to gas drag. Collisions between planetesimals rapidly generate a population of pebbles. The rate at which a core sweeps up pebbles is controlled by pebble accretion dynamics. Metre-size pebbles lose energy during an encounter with a core due to drag, and settle towards the core, greatly increasing the capture probability during a single encounter. Millimetre-size pebbles are tightly coupled to the gas and most are swept past the core during an encounter rather than being captured. Accretion efficiency per encounter increases with pebble size in this size range. However, radial drift rates also increase with size, so metre-size objects encounter a core on many fewer occasions than mm-size pebbles before they drift out of a region. The net result is that core growth rates vary weakly with pebble size, with the optimal diameter being about 10 cm. The main effect of planetesimal size is to determine the rate of mutual collisions, fragment production and the formation of pebbles. 1-km-diameter planetesimals collide frequently and have low impact strengths, leading to a large surface density of pebbles and rapid core growth via pebble accretion. 100-km-diameter planetesimals produce fewer pebbles, and pebble accretion plays a minor role in this case. The strength of turbulence in the gas determines the scale height of pebbles in the disk, which affects the rate at which they are accreted. For an initial solid surface density of 12 g/cm(2) at 5 AU, with 10-cm diameter pebbles and a disk viscosity parameter alpha = 10(-4), a 10-Earth mass core can form in 3 My for 1-10 km diameter planetesimals. The growth of such a core requires longer than 3 My if planetesimals are 100 km in diameter. (C) 2014 Elsevier Inc. All rights reserved.