Cusps in the center of galaxies: a real conflict with observations or a numerical artefact of cosmological simulations?

TL;DR

This paper investigates whether the cuspy dark matter profiles in galaxy centers seen in simulations are real or artifacts, revealing that numerical effects may cause these profiles and questioning the core-cusp problem's implications.

Contribution

It demonstrates that standard N-body simulations may produce artificial cusps due to numerical effects, challenging previous interpretations of the core-cusp problem.

Findings

Cuspy profiles are stable but show unphysical variations in particle integrals.

Unphysical Fokker-Planck streams are present in the cusp region.

Numerical effects may explain the cuspy profiles in simulations.

Abstract

Galaxy observations and N-body cosmological simulations produce conflicting dark matter halo density profiles for galaxy central regions. While simulations suggest a cuspy and universal density profile (UDP) of this region, the majority of observations favor variable profiles with a core in the center. In this paper, we investigate the convergency of standard N-body simulations, especially in the cusp region, following the approach proposed by \protect{\citep{13}}. We simulate the well known Hernquist model using the SPH code Gadget-3 and consider the full array of dynamical parameters of the particles. We find that, although the cuspy profile is stable, all integrals of motion characterizing individual particles suffer strong unphysical variations along the whole halo, revealing an effective interaction between the test bodies. This result casts doubts on the reliability of the velocity distribution function obtained in the simulations. Moreover, we find unphysical Fokker-Planck streams of particles in the cusp region. The same streams should appear in cosmological N-body simulations, being strong enough to change the shape of the cusp or even to create it. Our analysis, based on the Hernquist model and the standard SPH code, strongly suggests that the UDPs generally found by the cosmological N-body simulations may be a consequence of numerical effects. A much better understanding of the N-body simulation convergency is necessary before a 'core-cusp problem' can properly be used to question the validity of the CDM model.