Predictions for the hydrogen-free ejecta of pulsational pair-instability supernovae

Present time-domain astronomy efforts will unveil a variety of rare transients. We focus here on pulsational pair-instability evolution, which can result in signatures observable with electromagnetic and gravitational waves. We simulate grids of bare helium stars to characterize the resulting black hole (BH) masses and ejecta composition, velocity, and thermal state. The stars do not react "elastically" to the thermonuclear explosion: there is not a one-to-one correspondence between pair-instability driven ignition and mass ejections, causing ambiguity in what is an observable pulse. In agreement with previous studies, we find that for carbon-oxygen core masses 28Msun< M_CO<30.5Msun the explosions are not strong enough to affect the surface. With increasing mass, they first cause large radial expansion (30.5Msun<M_CO<31.4Msun), and finally, also mass ejection episodes (M_CO>31.4Msun). The lowest mass to be fully disrupted in a pair-instability supernova is M_CO=57Msun. Models with M_CO>121Msun reach the photodisintegration regime, resulting in BHs with M_BH>125Msun. If the pulsating models produce BHs via (weak) explosions, the previously-ejected material might be hit by the blast wave. We characterize the H-free circumstellar material from the pulsational pair-instability of helium cores assuming simply that the ejecta maintain a constant velocity after ejection. Our models produce He-rich ejecta with mass 10^{-3}Msun<M_CSM<40Msun. These ejecta are typically launched at a few thousand \kms and reach distances of ~10^{12}-10^{15} cm before core-collapse. The delays between mass ejection events and the final collapse span a wide and mass-dependent range (from sub-hour to 10^4 years), and the shells ejected can also collide with each other. The range of properties we find suggests a possible connection with (some) type Ibn supernovae.