Numerical Study of Bar Suppression in Galaxy Models Due to Disc Heating

TL;DR

This study uses N-body simulations to analyze how disc heating and softening parameters influence bar formation and suppression in galaxy models, revealing complex interactions and the impact of numerical heating.

Contribution

It investigates the combined effects of softening parameter and disc mass fraction on bar dynamics, highlighting the role of numerical heating and resolution in galaxy simulations.

Findings

Small softening values cause numerical heating and bar suppression.

Higher disc mass fraction can delay or modify bar formation.

Numerical heating dominates in high-resolution models with small softening.

Abstract

The process of bar formation, evolution and destruction is still a controversial topic regarding galaxy dynamics. Numerical simulations show that these phenomena strongly depend on physical and numerical parameters. In this work, we study the combined influence of the softening parameter, $\epsilon$ and disc mass fraction, $m_{\mathrm{d}}$ on the formation and evolution of bars in isolated disc-halo models via $N$-body simulations with different particle resolutions. Previous studies indicate that the bar strength depends on $m_{\mathrm{d}}$ as $\propto m_{\mathrm{d}}^{-1}$, which is seen as a delay in bar formation. However, the distorsion parameter, $\eta$, which measures the bar's momentum through time, shows that an increase in $m_{\mathrm{d}}$ does not always induce a delay in bar formation. This suggests that $\epsilon$ interact to either enhance or weaken the bar. Moreover, numerical heating dominates in models with small softening values, creating highly accelerated particles at the centre of discs, regardless of $m_{\mathrm{d}}$ or resolution. These enhanced particle accelerations produce chaotic orbits for $\epsilon \leq 5\,$pc, resulting in bar suppression due to collisional dynamics in the centre. In our high resolution models ($N \approx 10^{7}$), small softening values are incapable of reproducing the bar instability. The role of disc mass is as follows: increasing $m_{\mathrm{d}}$ for moderate $\epsilon$ ($\geq 10\,$pc) reduces the amount of drift in the acceleration profile, without affecting the bar's behaviour. Models with lower $m_{\mathrm{d}}$ values coupled with small softening values, have an excess of highly accelerated particles, introducing unwanted effects into otherwise reliable simulations. Finally, we show that the evolution of the disc's vertical acceleration profile is a reliable indicator of numerical heating introduced by $\epsilon$ and the bar.