No abstract

Stars formed in clusters can encounter other stars at close distances. In typical open clusters in the Solar neighbourhood containing hundreds or thousands of member stars, ten to twenty per cent of Solar-mass member stars are expected to encounter another star at distances closer than 100 au. These close encounters strongly perturb the planetary systems, directly causing ejection of planets or their capture by the intruding star, as well as exciting the orbits. Using extensive N-body simulations, we study such fly-by encounters between two Solar System analogues, each with four giant planets from Jupiter to Neptune. We quantify the rates of loss and capture immediately after the encounter, e.g., the Neptune analogue is lost in one in four encounters within 100 au, and captured by the flying-by star in one in twelve encounters. We then perform long-term (up to 1 Gyr) simulations investigating the ensuing post-encounter evolution. We show that large numbers of planets are removed from systems due to planet-planet interactions and that captured planets further enhance the system instability. While encounters can initially leave a planetary system containing more planets by inserting additional ones, the long-term instability causes a net reduction in planet number. A captured planet ends up on a retrograde orbit in half of the runs in which it survives for 1Gyr; also, a planet bound to its original host star but flipped during the encounter may survive. Thus, encounters between planetary systems are a channel to create counter-rotating planets, This would happen in around 1% of systems, and such planets are potentially detectable through astrometry or direct imaging.

The Neptunian satellite system is unusual, comprising Triton, a large (∼ 2700 km) moon on a close-in, circular, yet retrograde orbit, flanked by Nereid, the largest irregular satellite (∼300 km) on a highly eccentric orbit. Capture origins have been previously suggested for both moons. Here we explore an alternative in-situ formation model where the two satellites accreted in the circum-Neptunian disk and are imparted irregular and eccentric orbits by a deep planetary encounter with an ice giant (IG), like that predicted in the Nice scenario of early solar system development. We use N -body simulations of an IG approaching Neptune to 20 Neptunian radi (R Nep ), through a belt of circular prograde regular satellites at 10-30 R Nep . We find that half of these primordial satellites remain bound to Neptune and that 0.4-3% are scattered directly onto wide and eccentric orbits resembling that of Nereid. With better matches to the observed orbit, our model has a success rate comparable to or higher than capture of large Nereid-sized irregular satellites from heliocentric orbit. At the same time, the IG encounter injects a large primordial moon onto a retrograde orbit with specific angular momentum similar to Triton's in 0.3-3% of our runs. While less efficient than capture scenarios (Agnor & Hamilton 2006), our model does indicate that an in-situ origin for Triton is dynamically possible. We also simulate the post-encounter collisional and tidal orbital evolution of Triton analogue satellites and find they are decoupled from Nereid on timescales of ∼10 4 years, in agreement withĆuk & Gladman (2005).

No abstract

Most stars form in a clustered environment. Both single and binary stars will sometimes encounter planetary systems in such crowded environments. Encounter rates for binaries may be larger than for single stars, even for binary fractions as low as 10-20 per cent. In this work, we investigate scatterings between a Sun-Jupiter pair and both binary and single stars as in young clusters. We first perform a set of simulations of encounters involving wide ranges of binaries and single stars, finding that wider binaries have larger cross sections for the planet’s ejection. Secondly, we consider such scatterings in a realistic population, drawing parameters for the binaries and single stars from the observed population. The scattering outcomes are diverse, including ejection, capture/exchange and collision. The binaries are more effective than single stars by a factor of several or more in causing the planet’s ejection and collision. Hence, in a cluster, as long as the binary fraction is larger than about 10 per cent, the binaries will dominate the scatterings in terms of these two outcomes. For an open cluster of a stellar density 50 pc−3, a lifetime 100 Myr and a binary fraction 0.5, we estimate that of the order of 1 per cent of the Jupiters are ejected, 0.1 per cent collide with a star, 0.1 per cent change ownership and 10 per cent of the Sun-Jupiter pairs acquire a stellar companion during scatterings. These companions are typically 1000s of au distant and in half of the cases (so 5 per cent of all Sun-Jupiter pairs), they can excite the planet’s orbit through Kozai–Lidov mechanism before stripped by later encounters. Our result suggests that the Solar System may have once had a companion in its birth cluster.