Fragmentation of a dynamically condensing radiative layer

TL;DR

This paper investigates the stability of a dynamically condensing radiative gas layer using linear analysis, revealing that all perturbations grow and likely cause fragmentation during condensation.

Contribution

It introduces a self-similar, time-dependent model for the condensing layer and analyzes perturbations with numerical methods, showing instability across all wave numbers.

Findings

All perturbations are unstable with growth rates sufficient for nonlinearity.

Perturbation behavior depends on the product of wave number and cooling length, with different growth regimes.

The layer is expected to fragment into various scales due to instability.

Abstract

In this paper, the stability of a dynamically condensing radiative gas layer is investigated by linear analysis. Our own time-dependent, self-similar solutions describing a dynamical condensing radiative gas layer are used as an unperturbed state. We consider perturbations that are both perpendicular and parallel to the direction of condensation. The transverse wave number of the perturbation is defined by $k$. For $k=0$, it is found that the condensing gas layer is unstable. However, the growth rate is too low to become nonlinear during dynamical condensation. For $k\ne0$, in general, perturbation equations for constant wave number cannot be reduced to an eigenvalue problem due to the unsteady unperturbed state. Therefore, direct numerical integration of the perturbation equations is performed. For comparison, an eigenvalue problem neglecting the time evolution of the unperturbed state is also solved and both results agree well. The gas layer is unstable for all wave numbers, and the growth rate depends a little on wave number. The behaviour of the perturbation is specified by $kL_\mathrm{cool}$ at the centre, where the cooling length, $L_\mathrm{cool}$, represents the length that a sound wave can travel during the cooling time. For $kL_\mathrm{cool}\gg1$, the perturbation grows isobarically. For $kL_\mathrm{cool}\ll1$, the perturbation grows because each part has a different collapse time without interaction. Since the growth rate is sufficiently high, it is not long before the perturbations become nonlinear during the dynamical condensation. Therefore, according to the linear analysis, the cooling layer is expected to split into fragments with various scales.