N-body model for M51 - I. Multiple encounter versus single passage?

Abstract

A numerical survey of the encounter history of the interacting grand-design system M51 (NGC 5194/5195) is performed with a 3D multiple spherical polar grid code, where both components of the pair are described with self-gravitating star + gas discs embedded in rigid analytical halo potentials. Two classes of models are investigated, (1) nearly parabolic single passages, and (2) bound encounters implying several disc-plane crossings. Both types of models can approximate the general morphology of the M51 system and simultaneously fit the projected velocity difference and separation of the components. In both cases the companion disc-plane crossing responsible for the main spiral structure occurred nearly in the south, about 400-500 Myr ago at a distance of 25-30 kpc, but in opposite directions. In the bound encounter model there is also a more recent crossing, at a distance of 20-25 kpc about 50-100 Myr ago. Our models account for some important kinematical observations of the M51 system not explained by the previous models. Especially, we note that the multiple-encounter model with a recent passage produces significant out-of-plane velocities, which manifest as an S-shaped structure of the major axis rotation curve, and which also explain the high peculiar velocities in the north of the companion. In this model the resulting extended tail is tilted 40°-50° with respect to the inner disc, leading to a velocity field that appears to suggest counter-rotation of the tail with respect to the inner disc. Also some morphological features, like the direction of the tidal extensions from the companion, are better matched by a model with a recent encounter. Importantly, any pre-existing spiral arms are washed out by the tidally triggered spiral arms. The multiple-encounter model assumes a high inclination (i ≈ 85°) orbit, with a low current eccentricity (∈ ≈ 0.2). The possible origin of this type of bound orbital configuration is studied by simulations including the orbital decay via the Chandrasekhar formula for dynamical friction, and also by simulations including a self-consistently modelled live halo for the primary. The gross features of the model, including the tilted far tail, are preserved even when allowing for the effects of several earlier passages. The observed well-defined far tail seems to suggest that the previous passages have been at least 30 per cent more distant than the latest two crossings. According to our limited orbital survey, such an orbital decay can be accounted for, provided that M51 has an extended dark halo containing at least a few times the mass within the visible disc region (total Mhalo/Mdisc ∼ 10).