Aspects of Scaling in Gravitational Clustering

Abstract

The scaling ansatz of Hamilton et al. effectively extends the idea of self-similar scaling to initial power spectra of any generic shape. Applications of this ansatz have provided a semi-empirical analytical description of gravitational clustering which is extremely useful. This contribution examines the two theoretical ingredients that form the basis of these applications: self-similar evolution and the stable clustering hypothesis. A brief summary of work verifying self-similar scaling for scale free spectra $P(k) \propto k^n$, with $n < -1$ is given. The main results presented here examine the hypothesis that clustering is statistically stable in time on small scales, or equivalently that the mean pair velocity in physical coordinates is zero. The mean pair velocity of particles can be computed accurately from N-body simulations via the pair conservation equation, by using the evolution of the autocorrelation function $ξ(x,t)$. The results thus obtained for scale free spectra with $n = 0, -1, -2$ and for the CDM spectrum are consistent with the stable clustering prediction on the smallest resolved scales, on which the amplitude of $ξ\gsim 200-1000$ for $n = -2$ and $n = 0$, respectively.