Head-on collisions of binary white dwarf--neutron stars: Simulations in full general relativity

We simulate head-on collisions from rest at large separation of binary white dwarf – neutron stars (WDNSs) in full general relativity. Our study serves as a prelude to our analysis of the circular binary WDNS problem. We focus on compact binaries whose total mass exceeds the maximum mass that a cold degenerate star can support, and our goal is to determine the fate of such systems. A fully general relativistic hydrodynamic computation of a realistic WDNS head-on collision is prohibitive due to the large range of dynamical time scales and length scales involved. For this reason, we construct an equation of state (EOS) which captures the main physical features of NSs while, at the same time, scales down the size of WDs. We call these scaled-down WD models “pseudo-WDs (pWDs)”. Using pWDs, we can study these systems via a sequence of simulations where the size of the pWD gradually increases toward the realistic case. We perform two sets of simulations; One set studies the effects of the NS mass on the final outcome, when the pWD is kept fixed. The other set studies the effect of the pWD compaction on the final outcome, when the pWD mass and the NS are kept fixed. All simulations show that 14%-18% of the initial total rest mass escapes to infinity. All remnant masses still exceed the maximum rest mass that our cold EOS can support (1.92 solar masses), but no case leads to prompt collapse to a black hole. This outcome arises because the final configurations are hot. All cases settle into spherical, quasiequilibrium configurations consisting of a cold NS core surrounded by a hot mantle, resembling Thorne-Zytkow objects. Extrapolating our results to realistic WD compactions, we predict that the likely outcome of a head-on collision of a realistic, massive WDNS system will be the formation of a quasiequilibrium Thorne-Zytkow-like object.

💡 Research Summary

**
The paper presents the first fully general‑relativistic hydrodynamic simulations of head‑on collisions between a white dwarf (WD) and a neutron star (NS) starting from rest at a large separation. Because the physical scales involved differ by several orders of magnitude—WD radii of order 10³ km versus NS radii of ~10 km, dynamical timescales of ~1 s versus ~1 ms, and orbital periods that are ~10⁵ times the NS dynamical time—direct GR simulations would require on the order of 10⁷–10⁸ timesteps even with adaptive mesh refinement. To make the problem tractable, the authors construct a piecewise‑polytropic equation of state (EOS) that reproduces the essential high‑density physics of a neutron star while artificially compressing the low‑density white‑dwarf branch. The resulting “pseudo‑white dwarf” (pWD) models retain the same mass (≈0.98 M☉) but have radii only a few to a few tens of times larger than the NS, allowing the ratio of pWD to NS radii to be varied from 5:1 to 20:1.

Two families of simulations are performed. In the first, the NS mass is varied (1.4, 1.5, 1.6 M☉) while keeping the pWD compaction fixed at a 10:1 radius ratio. In the second, the NS mass is fixed (1.5 M☉) and the pWD compaction is changed (5:1, 10:1, 15:1, 20:1). All runs start with the two stars at rest and let them free‑fall toward each other. The collision generates a strong shock that heats and compresses the material. Approximately 14 %–18 % of the total rest mass is ejected to infinity in every case, while the remainder stays bound.

Crucially, the bound remnant always exceeds the maximum rest mass supported by the cold EOS (1.92 M☉) but does not undergo prompt collapse to a black hole. The reason is that the post‑collision configuration is hot; thermal pressure provides sufficient support to prevent immediate gravitational collapse. The final object is a nearly spherical, quasi‑equilibrium configuration consisting of a cold neutron‑star core surrounded by a hot, massive mantle of white‑dwarf debris. This structure is essentially a Thorne‑Zytkow‑like object (TZlO), a hybrid star first proposed in the 1970s. The authors find that the mass‑loss fraction is remarkably insensitive to the specific pWD compaction or NS mass, indicating a robust outcome.

By extrapolating the results to realistic white‑dwarf compactions (radius ratio ≈ 500:1), the authors argue that an actual WD‑NS head‑on collision would also produce a TZlO rather than a prompt black hole. The hot mantle would gradually cool via neutrino emission and photon radiation over timescales of seconds to hours, while the core remains a stable neutron star. This conclusion has several astrophysical implications. First, it suggests that WD‑NS mergers may not be a dominant source of prompt high‑frequency gravitational‑wave bursts, but the subsequent evolution of the TZlO could generate a distinct, longer‑duration GW signal as the remnant settles. Second, the hot mantle could give rise to observable electromagnetic transients (e.g., soft X‑ray or UV flashes) that might be detectable in coincidence with low‑frequency GW signals from space‑based detectors such as LISA or DECIGO. Finally, the pWD methodology provides a practical framework for tackling other multi‑scale relativistic problems where direct simulation is computationally prohibitive.

In summary, the study demonstrates that, when realistic mass ratios place the total system mass above the cold TOV limit, the thermal pressure generated in a head‑on WD‑NS collision can stave off immediate black‑hole formation, leading instead to a quasi‑stable Thorne‑Zytkow‑like object. The work advances both the numerical techniques for multi‑scale GR hydrodynamics and our theoretical understanding of the possible fates of compact binary mergers involving white dwarfs.