Evidence of .Ia Supernova Detonations in 3D Hydrodynamical Simulations of Double Degenerate Mergers

Abstract: We report detailed 3D simulations of 1.1 $\mathrm{M_{\odot}}$ Oxygen-Neon (ONe) white dwarfs (WDs) merging with a 0.35 $\mathrm{M_{\odot}}$ helium WD, conducted with the moving-mesh hydrodynamic code AREPO. The simulations utilise self-consistent chemical profiles for the primary WD which were generated by a stellar evolution code incorporating the effects of semi-degenerate carbon burning. We find that a helium detonation is ignited at the base of the helium layer, starting a thermonuclear runaway which encircles the WD and ejects material as a sub-luminous supernovae. Our canonical simulation, (C-120), ejects 0.103 $\mathrm{M_{\odot}}$ of primarily $^{4}\mathrm{He}$, $^{28}\mathrm{Si}$, and $^{32}\mathrm{S}$, after which the primary begins accreting again from the surviving secondary. Our results depend qualitatively on the "inspiral time" simulation parameter, which describes the length of a period of accelerated angular momentum loss. For example, the binary does not survive when inspiral time is too long. We compare the results using our self-consistent chemical profiles to a constant-composition WD structure and find the same explosion pattern when the inspiral time is short. However, we are able to obtain a typical Type Ia supernova (SN Ia) which destroys the primary by using the constant-composition structure and long inspiral. The shock-convergence in this simulation follows the "x-scissor mechanism" described by Gronow et al. in 2020, and causes a secondary detonation due to higher temperatures and higher number density of $^{12}\mathrm{C}$ at the convergence site. These results highlight the potential for unrealistic outcomes when conducting simulations that incorporate unrealistically large enhancements in angular momentum losses and (or) non-realistic chemical structures for the primary WD.