“…Hence, having a very accurate modelling of the mass ejection and its origins is of great importance and several studies have already been made to elucidate the ejection mechanism Corresponding author: Elias R. Most emost@itp.uni-frankfurt.de and quantify the various ejection channels. Numerical simulations classify the ejection in terms of matter that is dynamically ejected (Hotokezaka et al 2013;Bauswein et al 2013;Radice et al 2016Radice et al , 2018Palenzuela et al 2015;Lehner et al 2016;Sekiguchi et al 2015Sekiguchi et al , 2016Dietrich & Ujevic 2017;Dietrich et al 2017b,a;Bovard et al 2017;Papenfort et al 2018) during or shortly after the merger of the two stars, and in terms of matter that is ejected secularly (Siegel & Metzger 2017;Fernández et al 2018;Fujibayashi et al 2017), that is, on timescales 100 ms. Of these two channels, the second component is not yet very well understood, mostly due to the lack of long-term three dimensional studies, although notable exceptions exist, starting either from simplified initial conditions (Siegel & Metzger 2017;Fernández et al 2018) or being restricted to two spatial dimensions . In comparison, the dynamically ejected mass component has been explored in far greater detail, using either fully consistent microphysical descriptions at finite temperature and in full general relativity (Radice et al 2016(Radice et al , 2018Lehner et al 2016;Sekiguchi et al 2016;Bovard et al 2017), or in approximations of general relativity (Bauswein et al 2013), or using a simplified microphysics treatment (Dietrich et al 2017b,a;Hotokezaka et al 2013;Ciolfi et al 2017), together with analytical expressions that try to com-bine the abundance of data available (Dietrich & Ujevic 2017).…”