Very close encounters of stars might lead to significant perturbations of their Oort-type clouds, planetesimal belts and planetary systems. We have calculated encounter parameters of Hipparcos stars using HIP2, Pulkovo and CRVAD2 catalogs of radial velocities. It turns out that some stars have encounters within 0.1 pc from each other and might be on an essential collision course up to few thousands AU. We present here examples with accurate astrometric data and stable with respect to errors of radial velocities within $\pm 0.3$ km/s. They include $\beta$ vs. $\gamma$ Virginis, 61 Cygni vs. $\chi^1$ Orionis as well as close encounters involving $\eta$ Bo\"otis, AB Doradus, 61 Urs{\ae} Majoris and others.

This is one of results from our program of massive simulations of close encounters for all nearby stars. Epsilon Eridani is an extremely interesting star having one confirmed planet and multiple asteroid and debris belts. It should have a quite massive Oort cloud as well. Deltorn et al. searched for past Nemesis encounters of $\varepsilon$ Eri. In this paper we show that, according to current astrometric data, an other famous nearby star Luyten 726-8AB (=BL/UV Ceti) will pass at $\lesssim$ 0.93 ly from Epsilon Eridani in $\approx$ 31.5 kyr. So, it will probably pierce through the outer part of the hypothetical Oort cloud of $\epsilon$ Eri. BL/UV Ceti has only about 20 percent of the solar mass. Nevertheless, it could influence directly some long-period comets of Epsilon Eridani. The duration of mutual transit of two star systems within 1 ly from each other is $\gtrsim$ 4.6 kyr. Our simulations show that stellar encounters within 1 ly might be more frequent than previously thought. It could explain Proxima's peculiar trajectory with respect to $\alpha$ Cen AB or even Sedna's trajectory in the solar system.

So far, GJ 710 is the only known star supposed to pass through outskirts of the solar system within 1 ly. We have reexamined the SIMBAD database for additional stellar candidates (from highest ratios of squared parallax to total proper motion) and compared them with new HIP2 parallaxes and known radial velocities. At the moment, the best nominee is double star HD 107914 in the constellation Centaurus at $\approx$ 78.3 pc from the Sun whose principal component is a white (A-type) giant. It does not seem to appear neither in general catalogues of radial velocities available at SIMBAD nor in authoritative Garcia-Sanchez et al. papers on stellar encounters with the solar system. Awaiting for the value $v_r$ of its radial velocity, uknown to the author, we have calculated limits of $|v_r|$ necessary to this star to pass within 1 ly and 1 pc from the Sun in linear approximation. A very accurate value of its total proper motion is also extremely important. In the case of $v_r=-100$ km/s and most "advantageous" HIP2 data, HD 107914 could pass as near as 8380 AU from the Sun in an almost direct collision course with the inner part of the solar system! Inversely, if $v_r$ had a great positive value, then HIP 60503 could be the creator of peculiar trajectories of detached trans-Neptunian objects like Sedna.