1 of 48

An overview of neutron star mergers

NUCLEAR ASTRO SUMMER SCHOOL

(15-MAY-2025)

NSM AND KNE - 15-MAY-2025

1/30

Atul Kedia – NCSU

illustration: NASA Goddard

ATUL KEDIA

POSTDOC @

NORTH CAROLINA STATE UNIVERSITY

ASKEDIA@NCSU.EDU

2 of 48

Lecture outline

  1. Neutron stars (NS) and NS mergers (NSM)
    1. Gravitational Wave
    2. Numerical Relativity simulation�
  2. r-process nucleosynthesis�
  3. Kilonova
    • Observation
    • Simulation
    • Angular dependence

NSM AND KNE - 15-MAY-2025

2/30

Atul Kedia – NCSU

3 of 48

Neutron Star

NSM AND KNE - 15-MAY-2025

3/30

Atul Kedia – NCSU

Yunes, Miller and Yagi, Nature Rev. Phys. (2022)

Neutron Stars:

Compact remnant of Supernovae

Mass ~ Mass of Sun

Size ~ 9-13 km radius

Gravitational pressure balanced by repulsive strong nuclear force, neutron degeneracy pressure

4 of 48

Neutron Star above Athens, Ohio

NSM AND KNE - 15-MAY-2025

4

Atul Kedia – NCSU

Radius ~12km ~ 7 mi

Visualization from https://ns-in-my-city.daniel-wysocki.info/

5 of 48

AT2017gfo / GW170817

NSM AND KNE - 15-MAY-2025

5/30

Atul Kedia – NCSU

6 of 48

Neutron Star merger

NSM AND KNE - 15-MAY-2025

6

Atul Kedia – NCSU

LIGO (Laser Interferometer [for] Gravitational wave observation)

https://phys.org/news/2019-05-ligo-virgo-neutron-star-smash-ups.html

LIGO-Livingston-Louisiana, LIGO-Hanford-Washington, Virgo-Italy, KARGA-Japan, upcoming LIGO-India (2030s)

7 of 48

Timeline of LVK observational runs

7/16

8 of 48

Detection rate�projection

8/16

Credit: Broekgaarden

9 of 48

NSM AND KNE - 15-MAY-2025

9/30

Atul Kedia – NCSU

Future NS observations

*O4 LIGO-Virgo-KAGRA (ongoing)

*3rd Generation GW detectors

More measurements from�NICER: such as PSR-J0437

(Next gen pulse-profiling via Strobe-X)

ATUL KEDIA – NCSU

10 of 48

Sign up for GW Alerts with GraceDB

NSM AND KNE - 15-MAY-2025

10/28

Atul Kedia – NCSU

11 of 48

NSM AND KNE - 15-MAY-2025

11

Atul Kedia – NCSU

Gravitational Wave

Far from the binary neutron stars (or any compact merger, BBH, NS-BH), the weak field regime is valid, and the Einsteins equations reduce to the Linearized Einstein equation

Also see visualization here:

http://www.tapir.caltech.edu/~teviet/Waves/gwave.html

 

 

 

12 of 48

NSM AND KNE - 15-MAY-2025

12

Atul Kedia – NCSU

Flow for Gravitational wave detection

NSM –

Numerical Relativity simulations w. variety of EoSs

Waveform models

(relevant parameters: mass, tidal deformability, mass ratio, spin)

Match with observed GW strain

(Bayesian inference)

Inferred constraints on source properties

(masses, tidal deformability (EoS), mass ratio, spins)

13 of 48

GR: 3+1 (space-time) split and BSSN formalism.

Dynamics: Relativistic hydrodynamics (to describe the flow of matter) + Equations of State (“Close” the equations)

Numerical relativity codes that do this: Einstein Toolkit, Dendro-GR, GR-Athena++, SpECTRE, SPHINCS_BSSN, a few others.

(See Foucart et al. Snowmass white paper, arXiv:2203.08139 [gr-qc])

NSM AND KNE - 15-MAY-2025

13

Atul Kedia – NCSU

 

Numerical Relativity

14 of 48

Neutron Star merger

NSM AND KNE - 15-MAY-2025

14

Atul Kedia – NCSU

Top-Down view

Cross-section Edge-on view

AK et al., PRD (2022)

15 of 48

Impact of the EoS on the Postmerger

NSM AND KNE - 15-MAY-2025

15/30

Atul Kedia – NCSU

AK et al., PRD (2022)

Max density in the system in NR simulation

16 of 48

Impact of the EoS on the remnant

NSM AND KNE - 15-MAY-2025

16/30

Atul Kedia – NCSU

AK et al., PRD (2022)

Dietrich et al. GRG (2020)

EoS softness

17 of 48

Waveform differences

NSM AND KNE - 15-MAY-2025

17/30

Atul Kedia – NCSU

 

18 of 48

Scenarios of NSM: mass dependence

NSM AND KNE - 15-MAY-2025

18/30

Atul Kedia – NCSU

Bartos, Brady, and Marka (2013).

19 of 48

Neutron star merger ejecta contributing to the abundance

NSM AND KNE - 15-MAY-2025

19/30

Atul Kedia – NCSU

 

 

 

20 of 48

NSM AND KNE - 15-MAY-2025

20

Atul Kedia – NCSU

Flow of Nucleosynthesis

Neutron star merger

(EoS)

n-rich Ejecta

(Morphologies, Trajectories)

r-process nucleosynthesis enrichment

Ye (electron fraction)

Electromagnetic emission (Kilonova)

(Morphology of the ejecta)

21 of 48

Stellar chemical composition

NSM AND KNE - 15-MAY-2025

21/30

Atul Kedia – NCSU

  • Fig. from Lippuner (2018)

and r-process

 

 

22 of 48

Some Nomenclature

  •  

NSM AND KNE - 15-MAY-2025

22/30

Atul Kedia – NCSU

23 of 48

 

NSM AND KNE - 15-MAY-2025

23/30

Atul Kedia – NCSU

Calculation done on PRISM (reaction network)

ng -> neutron capture

gn -> photodissociation (neutron removal)

beta -> beta decay

24 of 48

Uncertainties in neutron capture rate

NSM AND KNE - 15-MAY-2025

24/30

Atul Kedia – NCSU

Uncert. for near stability isotopes

Uncert. away from stability

 

N-cap uncertainties propagated to abundances

25 of 48

Neutron star merger ejecta contributing to the kilonova

NSM AND KNE - 15-MAY-2025

25/30

Atul Kedia – NCSU

 

 

 

26 of 48

Kilonova is a multi-physics problem

NSM AND KNE - 15-MAY-2025

26/30

Atul Kedia – NCSU

Equation of state�(merger dynamics)

Bovard+17,�Radice+18�

Ejecta dynamics –�mass distribution, velocity distribution

Breschi+21, Bulla+19,23, Heinzel+21, Kawaguchi+20,21, Kedia+23

Composition – �r-process nucleosynthesis –neutrino, masses, reaction rates

(superheavies, island of stability Mumpower+18, Lund+23,24, Holmbeck+23, Zhu+18)

Thermalization efficiencies alpha, beta, fission fragments, and gamma rays –

Barnes+16,21

Atomic Opacity –

Fontes+ 15, 20, 23, Tanaka+20, Bulla+23

27 of 48

Simulation setup

  • Radiative transfer software using tabulated binned opacities on SuperNu. (Wollaeger et al 2013, 2014)
  • Composition and radioactive heating from r-process elements, nucleosynthetic results from WinNet. (Winteler et al. 2012)
  • Nuclear model
    • Heating rates (Korobkin et al. 2012)
    • Thermalization model of (Barnes et al. (2016))
  • Atomic opacities (Fontes et at. 2020)
  • Reprocessing of light from one component to another.

(Wollaeger et al 2013, 2014, 2018, 2021)

NSM AND KNE - 15-MAY-2025

27/30

Atul Kedia – NCSU

28 of 48

NSM AND KNE - 15-MAY-2025

28/30

Atul Kedia – NCSU

Broadband filters

Visible

29 of 48

Lanthanide curtaining

NSM AND KNE - 15-MAY-2025

29/30

Atul Kedia – NCSU

Wollaeger et al ApJ 2021; See also Korobkin+2021, Kasen+2015

More equatorial

A smoking gun for r-process

Curtaining of low wavelength (“blue”) light by Lanthanides (present in the “red” ejecta) produced by the nucleosynthesis.

Top panel: Ejecta more spherically symmetric and less obstructed by the “red” component. Therefore, there is virtually no angular dependence.

Bottom: “Red” component travels further and obstructs light emitted along the equator. Therefore, Lanthanide curtaining is observed.�Low wavelength light is suppressed substantially along the equator.

High wavelength light passes unaffected (spherically symmetric).

30 of 48

Impact of velocity distribution

NSM AND KNE - 15-MAY-2025

30/30

Atul Kedia – NCSU

31 of 48

AT2017gfo / GW170817

NSM AND KNE - 15-MAY-2025

31/30

Atul Kedia – NCSU

Villar et al ApJL 2017

Abbott et al PRL 2017 , LIGO+Virgo

sGRB ~ 2s

NGC 4993

32 of 48

Multi-Messenger Inference

  • Fig. Credit: Raaijmakers et al, ApJ (2021) ; modified here

NSM AND KNE - 15-MAY-2025

32/30

Atul Kedia – NCSU

Radiative transfer

simulations

Kilonova modelling

BNS merger modelling (Numerical relativity)

Nucleosynthesizing ejecta material

33 of 48

EM v GW ejecta�tension

NSM AND KNE - 15-MAY-2025

33/30

Atul Kedia – NCSU

Kilonova approach

🡨 Assuming(Torus, Peanut morphology )

GW approach

Ejecta properties (Kilonova approach) != Ejecta properties (GW approach)

34 of 48

NICER – Pulse Profile

NSM AND KNE - 15-MAY-2025

34/30

Atul Kedia – NCSU

Lightcurve model

EoS properties: Mass-Radius relation

Riley et al. 2019, 2021

NASA’s Neutron star Interior Composition ExploRer at the International Space Stn.

Credit: NASA Goddard

35 of 48

Equation of State (EoS)

NSM AND KNE - 15-MAY-2025

35/30

Atul Kedia – NCSU

Yunes, Miller and Yagi, Nature Rev. Phys. (2022)

 

Baym, Furusawa, Hatsuda et al. ApJ (2019) 885:42

36 of 48

Bayesian Inference

NSM AND KNE - 15-MAY-2025

36/30

Atul Kedia – NCSU

 

(Bayes’ theorem)

(TOV solver using: RePrimAnd)

37 of 48

Posterior

NSM AND KNE - 15-MAY-2025

37/30

Atul Kedia – NCSU

Prior

10,000 EoSs

38 of 48

Extra :EOS inference

NSM AND KNE - 15-MAY-2025

38/30

Atul Kedia – NCSU

39 of 48

Extra slides

NSM AND KNE - 15-MAY-2025

39/30

Atul Kedia – NCSU

40 of 48

EM v GW ejecta estimates

NSM AND KNE - 15-MAY-2025

40/30

Atul Kedia – NCSU

41 of 48

Ejecta masses and velocities

NSM AND KNE - 15-MAY-2025

41/30

Atul Kedia – NCSU

Darker shade signifies more mass ;

longer arrows indicate faster moving ejecta

42 of 48

NSM AND KNE - 15-MAY-2025

42/30

Atul Kedia – NCSU

NS Prior effect

43 of 48

HyperPipe – (RIFT)�HyperPosterior Pipeline -�Rapid parameter inference on gravitational �wave sources �via Iterative �FiTting

NSM AND KNE - 15-MAY-2025

43/30

Atul Kedia – NCSU

Start

Construct Initial grid

Input Grid

EoS generation for row #1 9.882e+01 …

EoS generation for row #2 1.856e-01 …

EoS generation for row #3, and so on

Join (all Likelihoods above)

Unify

Posterior generation (MC Integrator)

Posterior Samples

Sample Randomizer (Puff-ing)

All sample Likelihoods

iteration i

# lnL sigma_lnL g0 g1 g2 g3

0 0 9.882e-01 1.770e-01 2.720e-02 …

0 0 1.856e-01 1.389e-01 1.916e-02 ….

….

Posterior

# lnL sigma_lnL g0 g1 g2 g3

0 0 2.413e-01 1.055e-01 1.596e-02 …

0 0 1.647e-01 2.853e-01 2.107e-02….

….

e.g.

Internal book-keeping steps

44 of 48

NSM AND KNE - 15-MAY-2025

44/30

Atul Kedia – NCSU

Gaussian iterative fit (bimodal)

Y

Y

Z

Z

X

45 of 48

Three plausible/astrophysical priors for 170817

  • Uniform : m1,m2 uniform; s1z, s2z using the 'z prior', with max magnitude <=0.05
  • Extreme q: as above, but requiring the mass-ratio q<0.6. motivated by kilonova.
  • Positive spin, comparable q: q>0.9 and s_{iz} >0, motivated by galactic pulsars

NSM AND KNE - 15-MAY-2025

45/30

Atul Kedia – NCSU

GW prior effect

46 of 48

NSM AND KNE - 15-MAY-2025

46/30

Atul Kedia – NCSU

  • PRISM: Portable Routines for Integrated nucleoSynthesis Modeling  (Sprouse & Mumpower (NCSU alum))
  • Purpose – Reaction network code for astrophysical abundance yield calculation
  • Execution – for (ZR1,AR1) +...+ (ZRN,ARN) → (ZP1,AP1) +...+ (ZPM,APM)
  • Solves for Y(Zi,Ai,t) from 
  • (P(i) = probability for �each isotope formation)
  • Inputs – 
  • Nuclear data – for reactions rates(Optical Model Potential, Gamma strength function, level densities), decays probabilities. In the form of reaction rates for Alpha decay, beta decay, neutron capture, fission,�
  • Astrophysical data – Tracer particle - Initial abundance, Evolution of the packet in time, temperature, density

Reaction Network

47 of 48

LIGO-Virgo-KAGRA observation rate

NSM AND KNE - 15-MAY-2025

47/30

Atul Kedia – NCSU

Credit: Broekgaarden

48 of 48

NSM AND KNE - 15-MAY-2025

48

Atul Kedia – NCSU

Flow of Physics

Neutron star merger

n-rich Ejecta

r-process nucleosynthesis

Electromagnetic emission (Kilonova)