Forming disk galaxies in major mergers

II. The central mass concentration problem and a comparison of GADGET3 with GIZMO

Aix-Marseille Univ, CNRS, LAM, Laboratoire d’Astrophysique de Marseille, UMR 7326, 13388 Marseille, France
e-mail: sergey.rodionov@lam.fr

Received: 16 February 2016
Accepted: 28 December 2016

Abstract

Context. In a series of papers, we study the major merger of two disk galaxies in order to establish whether or not such a merger can produce a disk galaxy.

Aims. Our aim here is to describe in detail the technical aspects of our numerical experiments.

Methods. We discuss the initial conditions of our major merger, which consist of two protogalaxies on a collision orbit. We show that such merger simulations can produce a non-realistic central mass concentration, and we propose simple, parametric, active galactic nuclei (AGN)-like feedback as a solution to this problem. Our AGN-like feedback algorithm is very simple: at each time-step we take all particles whose local volume density is above a given threshold value and increase their temperature to a preset value. We also compare the GADGET3 and GIZMO codes, by applying both of them to the same initial conditions.

Results. We show that the evolution of isolated protogalaxies resembles the evolution of disk galaxies, thus arguing that our protogalaxies are well suited for our merger simulations. We demonstrate that the problem with the unphysical central mass concentration in our merger simulations is further aggravated when we increase the resolution. We show that our AGN-like feedback removes this non-physical central mass concentration, and thus allows the formation of realistic bars. Note that our AGN-like feedback mainly affects the central region of a model, without significantly modifying the rest of the galaxy. We demonstrate that, in the context of our kind of simulation, GADGET3 gives results which are very similar to those obtained with the PSPH (density independent SPH) flavor of GIZMO. Moreover, in the examples we tried, the differences between the results of the two flavors of GIZMO – namely PSPH, and MFM (mesh-less algorithm) – are similar to and, in some comparisons, larger than the differences between the results of GADGET3 and PSPH.