Secondary infall: Theory versus simulations

The applicability of the highly idealized secondary infall model to ''realistic'' initial conditions is investigated. The collapse of protohalos seeded by 3 sigma density perturbations to an Einstein-de Sitter universe is studied here for a variety of scale-free power spectra with spectral indices ranging from n = 1 to -2. Initial conditions are set by the constrained realization algorithm and the dynamical evolution is calculated both analytically and numerically. The analytic calculation is based on the simple secondary infall model where spherical symmetry is assumed. A full numerical simulation is performed by a Tree N-body code where no symmetry is assumed. A hybrid calculation has been performed by using a monopole term code, where no symmetry is imposed on the particles but the force is approximated by the monopole term only. The main purpose of using such code is to suppress off-center mergers. In all cases studied here the rotation curves calculated by the two numerical codes are in agreement over most of the mass of the halos, excluding the very inner region, and these are compared with the analytically calculated ones. The main result obtained here, which reinforces the findings of many N-body experiments, is that the collapse proceeds ''gently'' and not via violent relaxation. There is a strong correlation of the final energy of individual particles with the initial one. In particular, we find a preservation of the ranking of particles according to their binding energy. In cases where the analytic model predicts nonincreasing rotation curves its predictions are confirmed by the simulations. Otherwise, sensitive dependence on initial conditions is found, and the analytic model fails completely. In the cosmological context power spectra with n greater than or equal to - 1 yields (in the mean) nonincreasing rotation curves, and in such cases the secondary infall model is expected to be a useful tool in calculating the final virialized structure of collapsing halos.