Context. Observations suggest that some massive stars experience violent and eruptive mass loss associated with significant brightening that can not be explained by hydrostatic stellar models. This event seemingly forms dense circumstellar matter (CSM). The mechanism of eruptive mass loss has not been clarified. We focus on the fact that the timescale of nuclear burning gets shorter than the dynamical timescale of the envelope a few years before core collapse for some massive stars. Aims. To reveal the properties of the eruptive mass loss, we investigate its relation to the energy injection at the bottom of the envelope supplied by nuclear burning taking place inside the core. In this study, we do not specify the actual mechanism to transport energy from the site of nuclear burning to the bottom of the envelope. Instead, we parameterize the amount of injected energy and the injection time and try to extract information on these parameters from comparisons with observations. Methods. To this end, we carried out 1-D radiation hydrodynamical simulations for progenitors of red, yellow, and blue supergiants, and Wolf-Rayet stars. We calculated the evolution of the progenitors with a public stellar evolution code. Results. We obtained the light curve associated with the eruption, the amount of ejected mass, and the CSM distribution at the time of core-collapse. Conclusions. The energy injection at the bottom of the envelope of a massive star within a period shorter than the dynamical timescale of the envelope could reproduce some observed optical outbursts prior to the core-collapse and form the CSM, which can power an interaction SN classified as type IIn.

We present the public release of the Complete History of Interaction-Powered Supernovae (CHIPS) code, which is suited to model a variety of transients that arise from interaction with a dense circumstellar medium (CSM). Contrary to existing modelings, which mostly attach the CSM by hand, CHIPS self-consistently simulates both the creation of the CSM from mass eruption of massive stars prior to core collapse, and the subsequent supernova light curve. We demonstrate the performance of CHIPS by presenting examples of the density profiles of the CSM and the light curves. We show that the gross light-curve properties of putative interaction-powered transients (e.g., Type IIn supernovae, rapidly evolving transients and recently discovered fast blue optical transients) can be comprehensively explained with the output of CHIPS.

Observations of Type II supernovae imply that a large fraction of their progenitors experience enhanced mass loss years to decades before core collapse, creating a dense circumstellar medium (CSM). Assuming that the CSM is produced by a single mass eruption event, we analytically model its density profile. We find that a double power-law profile, where the inner (outer) power-law index has a characteristic value of −1.5 (−10 to −12), gives a good fit to the CSM profile obtained using radiation hydrodynamical simulations. With our profile the CSM is well described by just two parameters, the transition radius r* and density at r = r* (alternatively r* and the total CSM mass). We encourage future studies to include this profile, if possible, when modelling emission from interaction-powered transients.

We consider black hole formation in failed supernovae when a dense circumstellar medium (CSM) is present around the massive star progenitor. By utilizing radiation hydrodynamical simulations, we calculate the mass ejection of blue supergiants and Wolf–Rayet stars in the collapsing phase and the radiative shock occurring between the ejecta and the ambient CSM. We find that the resultant emission is redder and dimmer than normal supernovae (bolometric luminosity of 1040– , effective temperature of ∼5 × 103 K, and timescale of 10–100 days) and shows a characteristic power-law decay, which may comprise a fraction of intermediate luminosity red transients (ILRTs) including AT 2017be. In addition to searching for the progenitor star in the archival data, we encourage X-ray follow-up observations of such ILRTs ∼1–10 yr after the collapse, targeting the fallback accretion disk.