Timescales of major mergers from simulations of isolated binary galaxy collisions

A six-dimensional parameter space based on high-resolution numerical simulations of isolated binary galaxy collisions has been constructed to investigate the dynamical friction timescales, τ<SUB>mer</SUB>, for major mergers. Our experiments follow the gravitational encounters between ̃600 pairs of similarly massive late- and early-type galaxies with orbital parameters that meet the predictions of the Λ-cold dark matter (ΛCDM) cosmology. We analyse the performance of different schemes for tracking the secular evolution of mergers, finding that the product of the intergalactic distance and velocity is best suited to identify the time of coalescence. In contrast, a widely used merger-time estimator such as the exhaustion of the orbital spin is shown to systematically underpredict τ<SUB>mer</SUB>, resulting in relative errors that can reach 60% for nearly radial encounters. We find that the internal spins of the progenitors can lead to total variations in the merger times larger than 30% in highly circular encounters, whereas only the spin of the principal halo is capable of modulating the strength of the interaction prevailing throughout a merger. The comparison of our simulated merger times with predictions from different variants of a well-known fitting formula has revealed an only partially satisfactory agreement, which has led us to recalculate the values of the coefficients of these expressions to obtain relations that fit major mergers perfectly. The observed biases between data and predictions, which do not only apply to the present work, are inconsistent with expectations from differences in the degree of idealisation of the collisions, their metric, spin-related biases, or the simulation set-up. This indicates a certain lack of accuracy of the dynamical friction modelling, arising perhaps from a still incomplete identification of the parameters governing orbital decay.