The disruption of a star by the tidal field of a super-massive black hole may provide insights into dormant and otherwise hard-to-study galactic nuclei. State-of-the-art numerical tools have not converged on the importance of the strong relativistic effects in the disruption of stars by potentially spinning black holes. We present a specialised numerical tool to perform global hydrodynamic simulations of stellar tidal disruptions in curved spacetimes. We quantify the role of impact strength and black hole spin onto the stellar structures and mass fallback rates. We adapted the code SPHINCS_BSSN to perform General Relativistic Smoothed-Particle Hydrodynamics (GRSPH) simulations of stellar tidal disruptions in Kerr metric. We coupled the code with a Newtonian self-gravity module, and we added the option to use an entropy evolution formulation to handle numerically challenging situations. Besides describing the implementation and code validation, we present a set of 18 simulations of parabolic tidal disruptions of stellar polytropes to investigate the effect of impact strength and black hole spin. We demonstrate that SPHINCS is capable of performing GRSPH simulations, reproducing benchmark tests to machine precision. Our stellar tidal disruption simulations show that the fallback rates agree with state-of-the-art modelling. Deep events result into structures where self-gravity plays no role, the mass fallback rates peak at lower values, and rise-to-peak timescales decrease with impact strength. In these cases the black hole spin affects noticeable these quantities increasing (decrease) both fallback rate peak and rise-to-peak timescale for prograde (retrograde) spin. Last, fallback rates tend to decay with the characteristic t^-5/3 on long timescales. The results show that SPHINCS can simulate high-resolution stellar tidal disruptions in Kerr metric at a reasonable computational time.

The dynamics and observable signatures of neutron star mergers are governed by physics under the most extreme conditions. They are particularly impacted by the high-density equation of state, which for the most sophisticated models is usually available in the form of tables. Numerical relativity codes usually evolve particularly well-behaved numerical ("conservative") variables, but at the price that the physically interesting ("primitive") variables need to be found at every computational element and at every integration sub-step by means of expensive (and not always successful) root-finding algorithms. We have recently developed the Lagrangian numerical relativity code SPHINCS_BSSN which evolves the spacetime on an adaptive mesh with well tested methods, but the fluid is evolved by means of freely moving particles. Since our evolution equations differ from those of conventional numerical relativity, we need to develop new conservative-to-primitive algorithms if we want to use tabulated equations of state. We present here three such algorithms: a 3D and a 2D Newton-Raphson method and a 1D root-finding algorithm based on Ridders' method. We find the 3D method to be very fast and robust with an average failure fraction in a full-blown neutron star merger simulation (with the DD2 equation of state) well below 1

The discovery of cold structures around Sgr A* has challenged our understanding of the gas dynamics and thermodynamic state of the plasma in its vicinity. This work aims to constrain the conditions for the formation of such structures, namely the cold disc-like structure and the recently discovered G-1-2-3 complex. We conduct hydrodynamic simulations of the observed Wolf-Rayet stars feeding Sgr A*. Our simulations show that the plasma chemical composition is crucial for determining the medium properties. We demonstrate that the formation of a cold disc is possible for chemical compositions that are consistent with observational constraints. However, it is not possible to reproduce all the properties of the observed disc which might suggest the interaction with another structure. Additionally, we present our first results on the hydrodynamic modelling of IRS 16SW as a colliding-wind binary. This is the first step to develop a realistic model on the formation of the G-1-2-3 complex.

The tidal field of a black hole can turn a star into a gas stream whose orbit can precess, especially if the a black hole is rapidly spinning. In this work, we investigate the impact of precession on the light curves of tidal disruption events (TDE). To do so, we perform two-dimensional radiation-hydrodynamic simulations of the interaction of the TDE wind and luminosity with the precessed stream wrapped around the black hole. Our results show that in events with black holes of similar to 10(6) M-circle dot and no orbit-spin inclination, the line of sight has little effect on the light curves, since the stream covers a small fraction of the solid angle as the precession is confined to the orbital plane. In the case of black holes of greater than or similar to 10(7)M(circle dot) and high inclination (i similar to 90 degrees), the light curve peaks can be delayed by similar to 100 days due to presence of the precessed stream blocking the radiation in the early phase of the event. We also discuss our efforts to model self-consistently the hydrodynamic evolution of a tidal stellar stream on curved spacetimes by the presence of a massive black hole.

Solving the Tolman–Oppenheimer–Volkoff (TOV) equations, together with the tidal perturbation equations, for large numbers of equation-of-state (EOS) samples is a major computational bottleneck in Bayesian inference of the dense-matter EOS, and this will become increasingly limiting as next-generation observatories deliver far larger and more precise datasets. We develop neural-network surrogates for the forward TOV mapping that predict neutron star mass, radius, and tidal deformability simultaneously and directly from the EOS parameters and central density. We train and compare two architectures: a conventional feedforward network and a residual network, the latter of which, to our knowledge, has not previously been explored for TOV surrogate modeling. Trained on a piecewise polytropic EOS parameter space, both networks reproduce the numerical solutions to high accuracy, with the coefficient of determination exceeding 0.999 for all three observables, while accelerating the evaluation of stellar observables by roughly two orders of magnitude relative to direct numerical integration. We find that both architectures achieve excellent predictive accuracy at the network sizes considered here, with the residual network providing a modest improvement in accuracy over the feedforward network at the expense of slightly longer inference times. The overall performance differences remain small, indicating that a feedforward network already has sufficient capacity for this mapping while residual connections offer only incremental gains. Nevertheless, the residual architecture provides a robust baseline for future extensions to richer EOS parameterizations or higher-dimensional regression tasks. The resulting surrogates are well-suited to large-scale Bayesian EOS inference and population studies, where repeated TOV evaluations would otherwise dominate the computational cost.

Neutron star mergers hold the key to several grand challenges of contemporary (astro-)physics. In view of the upcoming next generation of ground-based detectors, it is crucial to keep improving theoretical predictions to harvest the full scientific returns from these investments. We introduce here a substantial update of our Lagrangian numerical relativity code SPHINCS_BSSN. Apart from changing our unit system, we add constraint damping terms to the BSSN spacetime evolution equations. We demonstrate that this measure reduces, without noteworthy computational cost, the Hamiltonian constraint violations by more than an order of magnitude. We further implement contributions to thermal energy and pressure that are based on Fermi liquid theory and contain a parametrization of the Dirac effective mass. These terms can be combined with any cold equation of state, and they enhance the physical realism of our simulations and introduce a physics-based concept of a temperature. In a set of merger simulations, we demonstrate good agreement with other temperature-dependent numerical relativity simulations. We find that different parametrizations of the Dirac effective mass can translate into shifts of ∼ 150 Hz in the dominant post-merger gravitational wave peak frequency.

Neutron star mergers are today considered a major production site for rapid neutron capture elements. While the bulk of the matter escapes at fast, but non-relativistic velocities (∼ 0.2 c), a small amount of the dynamically ejected mass reaches mildly relativistic velocities (≳0.6 c). It has been suggested earlier, that in such ejecta parts neutrons may avoid being captured and that their decay could power an early blue precursor to the main kilonova event. Here we study in detail the nucleosynthesis in such fast ejecta with nuclear network calculations along both parametrized and numerical relativity trajectories. We find that the nucleosynthesis can be divided into three channels, in one of which a substantial amount of free neutrons survives when the main r-process has frozen out. We provide a (semi-)analytical model for surviving free neutrons which agrees very well with the network calculations. If the mass fraction of the free neutrons exceeds ∼ 0.05, their β^--decay dominates the nuclear heating rate between ∼ 100 and ∼ 10^4 seconds. This dominance leads to a pronounced kilonova precursor that should for plausible ejecta parameters be visible for ULTRASAT out to ∼200 Mpc. Since at low electron fractions free neutrons can survive even for moderate velocities, mergers with large tidal ejecta, such as asymmetric neutron star mergers or favorable neutron star black hole mergers, may produce particularly bright blue precursors to their subsequent kilonovae.

With the next generation of both electromagnetic and gravitational wave observatories beginning to come online, rapid analysis methods for kilonova data are becoming increasingly important in astronomy. Traditional Bayesian parameter estimation using Markov chain Monte Carlo (MCMC) is time-consuming and relies on explicit likelihood approximations that can break down when modeling uncertainties are significant. We develop a simulation-based inference (SBI) framework for kilonova parameter estimation using density-estimation likelihood-free inference. The framework uses a Gaussian process emulator trained on ∼1300 radiative transfer simulations generated with the POSSIS code. We demonstrate that SBI provides a rapid alternative to MCMC for inference with emulators or approximate likelihoods that is robust to emulator uncertainty and likelihood misspecification. On simulated data, the SBI method accurately recovers injected parameters and produces posterior predictive light curves consistent with the data, but the MCMC posterior recovery suffers from systematic bias caused by likelihood misspecification. When analyzing AT2017gfo, the SBI and MCMC methods yield similar light-curve predictions but different posterior distributions, with a subset of the MCMC posteriors piling up at prior boundaries. The likelihood in the MCMC fails to capture the non-Gaussian, correlated structure of the emulator uncertainty, but SBI learns the posterior directly from forward simulations that include the full predictive distribution. Once trained, the SBI framework generates ∼2×10^4 posterior samples in seconds.

Context. The reported discovery of a cold (similar to 10(4) K) disc-like structure within the central 5 x 10(-3) pc around the super-massive black hole at the centre of the Milk Way, Sagittarius A* (Sgr A*), has challenged our understanding of the gas dynamics and thermodynamic state of the plasma in its immediate vicinity. State-of-the-art simulations do not agree on whether or not such a disc can indeed be a product of the multiple stellar wind interactions of the mass-losing stars in the region. Aims. The aims of this study are to constrain the conditions for the formation of a cold disc as a natural outcome of the system of the mass-losing stars orbiting around Sgr A*, to investigate whether the disc is a transient or long-lasting structure, and to assess the validity of the model through direct comparisons with observations. Methods. We performed a set of hydrodynamic simulations of the observed Wolf-Rayet (WR) stars feeding Sgr A* using the finite- volume adaptive mesh refinement code Ramses. We focus, for the first time, on the impact of the chemical composition of the plasma emanating from the WR stars. Results. The simulations show that the chemical composition of the plasma affects the radiative cooling to a sufficient degree to impact the properties of the medium, such as density and temperature, and, as a consequence, the rate at which the material inflows onto Sgr A*. We demonstrate that the formation of a cold disc from the stellar winds is possible for certain chemical compositions that are consistent with the current observational constraints. However, even in such cases, it is not possible to reproduce the reported properties of the observed disc-like structure, namely its inclination and the fluxes of its hydrogen recombination lines. Conclusions. We conclude that the stellar winds alone are not sufficient to form the cold disc around Sgr A* inferred from observations. Either relevant ingredients are still missing in the model, or the interpretation of the observed data needs to be revised.

Context. Several enigmatic dusty sources have been detected in the central parsec of the Galactic Center. Among them is X7, located at only similar to 0.02 pc from the central supermassive black hole, Sagittarius A* (Sgr A*). Recent observations have shown that X7 is becoming elongated due to the tidal forces of Sgr A*. X7 is expected to be fully disrupted during its pericenter passage around 2035, which might impact the accretion rate of Sgr A*. However, its origin and nature are still unknown. Aims. We investigated the tidal interaction of X7 with Sgr A* in order to constrain its origin. We tested the hypothesis that X7 was produced by one of the observed stars with constrained dynamical properties in the vicinity of Sgr A*. Methods. We employed a set of test-particle simulations to reproduce the observed structure and dynamics of X7. The initial conditions of the models were obtained by extrapolating the observationally constrained orbits of X7 and the known stars into the past, making it possible to find the time and source of origin by minimizing the three-dimensional separation and velocity difference between them. Results. Our results show that ejecta from the star S33/S0-30, launched in similar to 1950, can to a large extent replicate the observed dynamics and structure of X7, provided that it is initially elongated with a velocity gradient across it, and with an initial maximum speed of similar to 600 km s(-1). Conclusions. Our results show that a grazing collision between the star S33/S0-30 and a field object such as a stellar-mass black hole or a Jupiter-mass object is a viable scenario to explain the origin of X7. Despite the uncertainties in the rate of these encounters, recent estimations show that it is plausible for such a scenario to have occurred recently.

We present and explore a new shock-capturing particle hydrodynamics approach. Our starting point is a commonly used discretization of smoothed particle hydrodynamics. We enhance this discretization with Roe's approximate Riemann solver, we identify its dissipative terms, and in these terms, we use slope-limited linear reconstruction. All gradients needed for our method are calculated with linearly reproducing kernels that are constructed to enforce the two lowest-order consistency relations. We scrutinize our reproducing kernel implementation carefully on a "glass-like" particle distribution, and we find that constant and linear functions are recovered to machine precision. We probe our method in a series of challenging 3D benchmark problems ranging from shocks over instabilities to Schulz-Rinne-type vorticity-creating shocks. All of our simulations show excellent agreement with analytic/reference solutions.

The ejection of neutron-rich matter is one of the most important consequences of a neutron star merger. While the bulk of the matter is ejected at fast, but non-relativistic velocities (∼0.2c), a small amount of mildly relativistic dynamic ejecta have been seen in a number of numerical simulations. Such ejecta can have far reaching observational consequences ranging from the shock breakout burst of gamma-rays promptly after the merger, to an early (∼ 1 hour post-merger) blue kilonova precursor signal, to synchrotron emission years after the merger ("kilonova afterglow"). These all potentially carry the imprint of the binary system parameters and the equation of state. By analyzing Lagrangian simulations in full General Relativity, performed with the code SPHINCS_BSSN, we identify two ejection mechanisms for fast ejecta: i) about 30% of the ejecta with v> 0.4c are "sprayed out" from the shear interface between the merging stars and escape along the orbital plane and ii) the remaining ∼ 70% of the fast ejecta result from the central object "bouncing back" after strong, general-relativistic compression. This "bounce component" is ejected in a rather isotropic way and reaches larger velocities (by ∼0.1c) so that its faster parts can catch up with and shock slower parts of the spray ejecta. Even for a case that promptly collapses to a black hole, we find fast ejecta with similar properties to the non-collapsing case, while slower matter parts are swallowed by the forming black hole. We discuss observational implications of these fast ejecta, including shock breakout and kilonova afterglow.

In this book chapter we describe the Lagrangian numerical relativity code . This code evolves spacetimes in full General Relativity by integrating the BSSN equations on structured meshes with a simple dynamical mesh refinement strategy. The fluid is evolved by means of freely moving Lagrangian particles, that are evolved using a modern Smooth Particle Hydrodynamics (SPH) formulation. To robustly and accurately capture shocks, our code uses artificial dissipation terms, but, similar to Finite Volume schemes, we apply a slope-limited reconstruction within the dissipative terms and we use in addition time-dependent dissipation parameters, so that dissipation is only applied where needed. The technically most complicated, but absolutely crucial part of the methodology, is the coupling between the particles and the mesh. For the mapping of the energy-momentum tensor T_μν from the particles to the mesh, we use a sophisticated combination of "Local Regression Estimate" (LRE) method and a "multi-dimensional optimal order detection" (MOOD) approach which we describe in some detail. The mapping of the metric quantities from the grid to the particles is achieved by a quintic Hermite interpolation. Apart from giving an introduction to our numerical methods, we demonstrate the accurate working of our code by presenting a set of representative relativistic hydrodynamics tests. We begin with a relativistic shock tube test, then compare the frequencies of a fully relativistic neutron star with reference values from the literature and, finally, we present full-blown merger simulations of irrotational binary systems, one case where a central remnant survives and another where a black hole forms, and of a binary where only one of the stars is rapidly spinning.

The merger origin long GRB 211211A was a class (re-)defining event. A precursor was identified with a similar to 1 s separation from the main burst, as well as a claimed candidate quasi-periodic oscillation (QPO) with a frequency similar to 20 Hz. Here, we explore the implications of the precursor, assuming the quasi-periodicity is real. The precursor variability time-scale requires relativistic motion with a Lorentz factor Gamma greater than or similar to 80, and implies an engine-driven jetted outflow. The declining amplitude of the consecutive pulses requires an episodic engine with an 'on/off' cycle consistent with the QPO. For a black-hole central engine, the QPO can have its origin in Lense-Thirring precession of the inner disc at similar to 6-9 r(g) (gravitational radii) for a mass M center dot <= 4.5 M-circle dot, and less than or similar to 7 r(g) for M-center dot > 4.5 M-circle dot and dimensionless spin chi similar to 0.3-0.9. Alternatively, at a disc density of similar to 10(8-12 )g cm(-3), the required magnetic field strength for a QPO via magnetohydrodynamic effects will be of the order of B similar to 10(12-14) G. If the central engine is a short-lived magnetar or hypermassive neutron star, then a low-frequency QPO can be produced via instabilities within the disc at a radius of similar to 20-70 km, for a disc density similar to 10(9-12) g cm(-3) and magnetic field greater than or similar to 10(13-14) G. The QPO cannot be coupled to the neutron star spin, as the co-rotation radius is beyond the scale of the disc. Neither engine can be ruled out - however, we favour an origin for the precursor candidate QPO as early jet-disc coupling for a neutron star-black hole merger remnant with mass M-center dot > 4.5 M-circle dot.

The dynamics of a binary neutron stars merger is governed by physics under the most extreme conditions, including strong spacetime curvature, ultra-high matter densities, luminous neutrino emission and the rapid amplification of the initial neutron star magnetic fields. Here we systematically explore how sensitive the magnetic field evolution is to the total mass of the merging binary, to the mass ratio of its components, the stellar spins and to the equation of state. For this purpose, we analyze 16 state-of-the-art GRMHD simulations that employ a subgrid-scale model to account for the unresolved small-scale turbulence. We find that strong and rapid amplification of the magnetic field to volume-averaged values of ∼ 10^16 G in the high-density regions is a very robust outcome of a neutron star merger and this result is only marginally impacted by either mass, mass ratio, spin or equation of state.

With ongoing advancements in nuclear theory and experimentation, together with a growing body of neutron star (NS) observations, a wealth of information on the equation of state (EOS) for matter at extreme densities has become accessible. Here, we utilize a hybrid EOS formulation that combines an empirical parametrization centered around the nuclear saturation density with a generic three-segment piecewise polytrope model at higher densities. We incorporate data derived from chiral effective field theory (χEFT), perturbative quantum chromodynamics (pQCD), and experiments such as PREX-II and CREX. Furthermore, we examine the influence of a total of 70 NS mass measurements up to April 2023, as well as simultaneous mass and radius measurements derived from the x-ray emission from surface hot spots on NSs. Additionally, we consider constraints on tidal properties inferred from the gravitational waves emitted by coalescing NS binaries. To integrate this extensive and varied array of constraints, we utilize a hierarchical Bayesian statistical framework to simultaneously deduce the EOS and the distribution of NS masses. We find that incorporating data from χEFT significantly tightens the constraints on the EOS of NSs near or below the nuclear saturation density. However, constraints derived from pQCD computations and nuclear experiments such as PREX-II and CREX have minimal impact. Taking into account all available data, we derive estimates for key parameters characterizing the EOS of dense nuclear matter. Specifically, we determine the slope (L) and curvature (Ksym) of the symmetry energy to be 54−10+10 and −158−63+73 MeV with 90% credibility, respectively. Additionally, we infer the radius and tidal deformability of an NS with a mass of 1.4 solar masses (M⊙) to be 12.34−0.53+0.43 km and 436−117+109, respectively. Furthermore, we estimate the maximum mass of a nonrotating NS to be 2.22−0.19+0.21M⊙ with 90% credibility.

ABSTRACT Neutron star (NS) mergers where both stars have negligible spins are commonly considered as the most likely ‘standard’ case. In globular clusters, however, the majority of NSs have been spun up to millisecond (ms) periods and, based on observed systems, we estimate that a non-negligible fraction of all double NS mergers ($\sim 4\pm 2\, {{\ \rm per\ cent}}$) contains one component with a spin of a (few) ms. We use the Lagrangian numerical relativity code SPHINCS_BSSN to simulate mergers where one star has no spin and the other has a dimensionless spin parameter of χ = 0.5. Such mergers exhibit several distinct signatures compared to irrotational cases. They form only one, very pronounced spiral arm and they dynamically eject an order of magnitude more mass of unshocked material at the original, very low electron fraction. One can therefore expect particularly bright, red kilonovae. Overall, the spinning case collisions are substantially less violent and they eject smaller amounts of shock-generated semirelativistic material. Therefore, the ejecta produce a weaker blue/ultraviolet kilonova precursor signal, but – since the total amount is larger – brighter kilonova afterglows months after the merger. The spinning cases also have significantly more fallback accretion and thus could power late-time X-ray flares. Since the post-merger remnant loses energy and angular momentum significantly less efficiently to gravitational waves, such systems can delay a potential collapse to a black hole and are therefore candidates for merger-triggered gamma-ray bursts with longer emission time-scales.

A major ingredient for kilonova lightcurves is the radioactive heating rate and its dependence on the electron fraction and velocity of the ejecta and, in principle, on the nuclear mass formula. Heating-rate formulae commonly used as the basis for kilonova models are, strictly speaking, incorrect for electron fractions other than $Y_{e} = 0.04$. Here, we introduce new semi-analytical models for kilonovae with better heating rate prescriptions valid for the full parameter space of kilonova velocities and electron fractions. This new prescription produces, on average, dimmer kilonovae at peak that decay more slowly as compared to previously used prescriptions for otherwise identical kilonova physics. We show the dangers of using inappropriate heating rate estimates by simulating realistic observations and inferring the kilonova parameters via a misspecified heating-rate prescription. While providing great fits to the photometry, an incorrect heating-rate prescription fails to recover the input ejecta masses at greater than $5\sigma$. This bias from an incorrect prescription has disastrous consequences for interpreting kilonovae, their use as additional components in gamma-ray burst afterglows, and understanding their role in cosmic chemical evolution or for multi-messenger constraints on the nuclear equation of state. Given the true heating-rate is uncertain, we estimate there is a $\approx 5$ and $\approx 10\%$ systematic uncertainty in the measured ejecta masses and velocities, respectively. For lanthanide-rich ejecta, the systematic uncertainty could be as high as $\approx 50\%$. This systematic uncertainty limits the precision of any measurements that rely on accurate estimates of kilonova ejecta properties.

Colliding neutron stars: what do gravitational waves and electromagnetic flashes tell us about physics at the extremes? Mergers of two neutron stars are related to many timely astrophysical questions, such as the production of gravitational waves, the densest forms of matter in the Universe and the cosmic origin of heavy elements. This project applies a novel computational methodology to strive for consistent numerical modelling involving a broad range of both physical processes and length/time scales.

We examine which first order phase transitions are consistent with today's astrophysical constraints. In particular, we explore how a well-constrained mass-radius data point would restrict the admissible parameter space and to this end, we employ the most likely candidates of the recent NICER limits of PSR J0030+0451. To systematically vary the stiffness of the equation of state, we employ a parameterizable relativistic mean field equation of state, which is in compliance with results from chiral effective field theory. We model phase transitions via Maxwell constructions and parameterize them by means of the transitional pressure $p_{\rm trans}$ and the jump in energy density $\Delta\epsilon$. This provides us with a generic setup that allows for rather general conclusions to be drawn. We outline some regions in the $p_{\rm trans}$-$\Delta\epsilon$ parameter space that may allow for a phase transition identification in the near future. We also find that a strongly constrained data point, at either exceptionally large or small radii, would reduce the parameter space to such an extent that mass and radius become insufficient indicators of a phase transition.