From supernovae to neutron stars

Transcription

1 From supernovae to neutron stars Yudai Suwa 1,2 1 Yukawa Institute for Theoretical Physics, Kyoto University 2 Max Planck Institute for Astrophysics, Garching

2 Yudai Suwa, Kemer, Turkey 2 /25 Introduction: what is supernova?

3 Supernova before after SN 211fe

4 Yudai Suwa, Kemer, Turkey 4 /25 Supernovae make neutron stars Baade & Zwicky 1934

5 Yudai Suwa, Kemer, Turkey 5 /25 Key observables characterizing supernovae Explosion energy: ~1 51 erg=1 44 J Ejecta mass: ~M = kg measured by fitting SN light curves Ni mass: ~.1M NS mass: ~1-2M measured by binary systems final goal of first-principle (ab initio) simulations

6 Yudai Suwa, Kemer, Turkey 6 /25 Standard scenario of core-collapse supernovae Final phase of stellar evolution Neutrinosphere formation neutrino trapping Neutron star formation (core bounce Fe Neutrinosphere Neutron Star Si O,Ne,Mg C+O HeH ρc~1 11 ρc~114 ρc~1 9 shock stall shock revival Supernova! HOW? NS Si O,Ne,Mg C+O HeH

7 Yudai Suwa, Kemer, Turkey 7 /25 Current paradigm: neutrino-heating mechanism heating region shock cooling region absorption neutron star emission Energy is transferred by neutrinos Most of them are just escaping from the system, but are partially absorbed In gain region, neutrino heating overwhelms neutrino cooling

8 Yudai Suwa, Kemer, Turkey 8 /25 Numerical simulations

9 Yudai Suwa, Kemer, Turkey 9 /25 Physical ingredients In these violent explosions, all known interactions are involving and playing important roles; Strong Weak - nuclear equation of state - structure of neutron stars RNS~1-15 km max(mns)> 2 M - nucleosynthesis - neutrino interactions σν~1-44 cm 2 (Eν/mec 2 ) 2 - ~99% of energy is emitted by ν s - cooling of proto-neutron star - heating of postshock material Electromagnetic - Coulomb collision of p and e - final remnants are pulsars (B~1 12 G) magnetars (B~ G) magnetic fields affect dynamics Gravitational - energy budget EG~3.1x1 53 erg(m/1.4m ) 2 (R/1km) -1 ~.17M c 2 - inducing core collapse - making general relativistic objects (NS/BH)

10 Yudai Suwa, Kemer, Turkey 1/25 What do simulations solve? Numerical Simulations Hydrodynamic equations de dt dρ dt ρ dv dt + ρ v =, = P ρ Φ, + [( e + P ) v ] = ρv Φ + Q E, v v Φ dy e dt = Q N, Φ = 4πGρ, Solve simultaneously Neutrino Boltzmann df cdt + µ f r + + equation [ µ [ µ 2 ( d ln ρ cdt ( d ln ρ + 3v ) + 1 cdt cr r + 3v ) v ] cr cr = j (1 f ) χf + E2 c (hc) 3 [ (1 f ) Rf dµ f E f E ] (1 µ 2 ) f µ R ( 1 f ) ] dµ. ρ: density, v: velocity, P: pressure, Φ: grav. potential, e * : total energy, Ye: elect. frac., Q: neutrino terms f: neut. dist. func, µ: cosθ, E: neut. energy, j: emissivity, χ: absorptivity, R: scatt. kernel

11 Yudai Suwa, Kemer, Turkey 11/25 1D simulations fail to explode Rammp & Janka Liebendörfer+ 1 By including all available physics to simulations, we concluded that the explosion cannot be obtained in 1D! (The exception is an 8.8 M star; Kitaura+ 6) Thompson+ 3 Sumiyoshi+ 5

12 Neutrino-driven explosion in multi-d simulation The Astrophysical Journal, 756:84 (22pp), 212 September 1 The Astrophysical Journal Letters, 767:L6 (7pp), 213 April 1 Mu ller, Janka, & Marek 12 ms 12 ms Bruenn et al. 9 ms We have exploding models driven by neutrino heating with 2D/3D simulations [Suwa+ PASJ, 62, L49 (21); ApJ, 738, 165 (211); ApJ 764, 99 (213); PASJ, 66, L1 (214); arxiv: ] entropy [kb/baryon] Entropy [kb-1] msbetween comparison 1D and 2D 4 2 ms Distance from symmetry axis [km] 2 2 Radial velocity [km s-1] ms ms Brruenn et al. (213) ms ms ms Müller, Janka, Marek (212) Distance along symmetry axis [km] Figure 1. Evolution of the entropy (upper half) and radial velocity (lower half) for B12-WH7, with snapshots at tpb = 12, 9, 15, 2, 3, 4, 6, and 8 ms. Yudai Suwa, Kemer, Turkey 12 /25 The scale in time captureand thethe expansion the baryon supernova shockwave, the color Figure 6. Snapshots of the evolution of model G11, depicting the radial velocity vr grows (left half of to panels) entropyofper s (right half ofbut panels) 115maps ms,remain constant. The radial velocity portion is omitted for the first snapshots. 23 ms, 29 ms, 49 ms, 658 ms, and 92 ms after bounce (from top left totwo bottom right).

13 Dimensionality and numerical simulations grid-based codes only, not completed Dimension 3D Blondin+, 7 Mikami+, 8 Iwakami+, 8 Scheidegger+, 8 Nordhaus+, 1 Hanke+, 12 Couch, 13 Handy+, 14 Only simulations in this region can judge the neutrino-driven explosion scenario Takiwaki, Kotake, & Suwa, 12 Hanke+, 13 Lentz+, 15 Müller, 15 2D (axial-sym.) Yamada & Sato, 94 Blondin & Mezzacappa, 3 Obergaulinger+, 6 Kotake+, 3 Ohnishi+, 6 Murphy+, 8 Takiwaki+, 9 Burrows+, 6 Buras+, 6 Ott+, 8 Suwa+, 1 Sekiguchi+, 11 Müller+, 12 Bruenn+, 13 Obergaulinger+,14 Pan+, 15 1D (spherical-sym.) Rampp & Janka, Liebendörfer+, 1 Thompson+, 3 Sumiyoshi+, 5 O Connor+, 13 Adiabatic cooling only or heat by hand Transport Neutrino Treatment Yudai Suwa, Kemer, Turkey 13/25

14 Yudai Suwa, Kemer, Turkey 14/25 3D simulation with spectral neutrino transfer [Takiwaki, Kotake, Suwa, ApJ, 749, 98 (212); ApJ, 786, 83 (214)] MZAMS=11.2 M 384(r)x128(θ)x256(φ)x2(Eν) K computer T2K-Tsukuba XT4

15 where the entropy production takes place shifts from the gain radius (dotted gray line) to the PNS surface (dotted black line). exp(ln(q) t ωosc ), w as black-dotted line both in our 3D (top (top right) are clos be a rather generic SASI. Note here tha Dimensionality and initial perturbation insensitive to the employed progenitor. In fact, Figure 5 in Marek & Janka (29) shows the transition timescale to be [Takiwaki, Kotake, Suwa, ApJ, 786, 83 (214)] Figure 5. Entropy dispersion (σs ) in spacetime diagrams the text for the definition). The dotted gray line represents t Takiwaki, & Suw radius, while the dotted black line showskotake, the position of black arrow inserted at around 125 ms represents the epo model. the from bottom around 15 ms where for a the 15 entropy M progenitor production takes placeinshifts th to the PNS surface black line). panels of Figuregray 8, line) the saturation levels(dotted of the even modes o The Astrophysical Journal, 749:98 (17pp), 212 April 2 2D Radius [km] Shock radius [km] (ℓ, m ) = (4,), (4,2), (4,4) in 3D are shown to become muc larger than those in 2D (pink line), while the odd mode o (ℓ, m) = (3,) is much the same. insensitive work to thebyemployed progenitor.(1992 In Based on the pioneering Houck & Chevalier Marek (29) shows the transition the linear growth rate & of Janka the SASI in the core-collapse cas was presented by Scheck et al. (28). They pointed out th the cycle efficiency (:Q) which represents how many times th Figure 6. Volume rendering of entropy showing the expands blast morphology in our to 3Dthe (left), 2D (middle), and 1D (ri average radius compared original position per The linear scale is indicated in each panel. unit oscillation frequency (:ωosc ) of the SASI is an importan quantity to characterize the linear growth rate. From Figure 8 1 Q and ωosc in our simulation are approximately estimated t 1 2 be 2 and 25 ms, S11.2-3D respectively. Note that these values2dare i Figure 5. Entropy dispersion (σs ) in spacetime diagrams for the 3D model (see 18 3D the text for the definition). The dotted gray line represents the position of the gain agreement bylow Schec 1.4 with the ones obtained in 2D simulations3d radius, while the dotted black line shows the position of the PNS surface. The 16 et al. (28, e.g., their Figure 17). From the two quantitie black arrow inserted at around 125 ms represents the epoch when the position 14 estimated a the linear growth rate can be straightforwardly 1 where the entropy production takes place shifts from the gain radius (dotted exp(ln(q) t ω ), which is shown in the top panels of Figure osc 12 gray line) to the PNS surface (dotted black line). 1.3 as black-dotted lines. As can be seen, the growth 1 rates observe both in our 3D (top left panel in Figure 8) and 2D simulation 8 (top right) are close to the linear growth rate, which seems t 1 insensitive to the employed progenitor. In fact, Figure 5 in be a rather generic trend for the low-modes 6 (ℓ = 1, 2) of th Marek & Janka (29) shows the transition timescale to be SASI. Note here that the normalized amplitude4 of the shock (th 3D 1D Volume of entropy showing 5 blast Figure rendering the The linear scale is indicated in each panel. Time after bounce [ms] T Figure 7. Time evolution of the 3D model, visualized by mass shell trajectories in thin gray lines (left panel). T that the maximum (top), average (middle), and the minimum (bottom) shock positions are shown, respectively 1 model. 1.3 and 1.4 indicates the mass in unit of M enclosed inside the mass-shell. Right panel showss line) and 3D (red line) models. The 3D low (pink line) corresponds to the low-resolution 3D model, in whic 1.4 model (see Section 2). Radius [km] Figure 6. Volume rendering of entropy showing the blast morphology in our 3D (left), 2D (middle), and 1D (right) models (at t = 178 ms after bounce), respectivel Time after bounce [m The linear scale is indicated in each panel. Note) explosion energy is still too small (~15 erg) compared to observations (~151 erg) Figure 7. Time evolution of the 3D model, visualized by Yudai Suwa, Kemer, Turkey 1 S11.2-3D 2 that the maximum (top), average (middle), and the minim model. 1.3 and 1.4 indicates the mass in unit of M 2D line) and 3D (red line) models. The 3D low (pink line) 15 /25

16 Yudai Suwa, Kemer, Turkey 16/25 Equation of state dependence

17 Yudai Suwa, Kemer, Turkey 17/25 List of SN EOS Courtesy of M. Hempel

18 Yudai Suwa, Kemer, Turkey 18/25 Finite temperature EOSs Lattimer & Swesty (LS) (1991) based on compressible liquid drop model variants with K=18, 22, and 375 MeV H.Shen et al. (1998, 211) relativistic mean field theory (TM1) including hyperon component (~211) Hillebrandt & Wolff (1985) Hartree-Fock calculation G.Shen et al. (21, 211) relativistic mean field theory (NL3, FSUGold) Hempel et al. (212) relativistic mean field theory (TM1, TMA, FSUGold) More recently, Steiner+ (213), Furusawa+ (213), etc. incompressibility K [MeV] symmetry energy J (S) [MeV] slope of symmetry energy L [MeV] LS 18, 22, HShen HW GShen Hempel (NL3) 23. (FSU) 318 (TMA) 23 (FSU) (NL3) (FSU) 3.7 (TMA) 32.6 (FSU) (NL3) 6.5 (FSU) 9 (TMA) 6 (FSU) [Fischer, Hempel, Sagert, Suwa, Schaffner- Bielich, EPJA, 5, 46 (214)]

19 Yudai Suwa, Kemer, Turkey 19/25 Shock radius evolution depending on EOS [Suwa, Takiwaki, Kotake, Fischer, Liebendörfer, Sato, ApJ, 764, 99 (213)] maximum average minimum LS18 and LS375 succeed the explosion HShen EOS fails

20 Radius of neutron star [Suwa, Takiwaki, Kotake, Fischer, Liebendörfer, Sato, ApJ, 764, 99 (213)] Radius of Neutron Star [km] LS18 LS375 Shen Faster contraction is better for the explosion! Time after Bounce [ms] 5 km 1 km 5 km Pressure perturbation Radius Time Radius Yudai Suwa, Kemer, Turkey 2/25

21 Yudai Suwa, Kemer, Turkey 21/25 From supernovae to neutron stars

22 From SN to NS-1 [Suwa, Takiwaki, Kotake, Fischer, Liebendörfer, Sato, ApJ, 764, 99 (213); Suwa, PASJ, 66, L1 (214)] entropy [kb/baryon] ejecta NS NS mass ~1.3 M Progenitor: 11.2 M (Woosley+ 22) Successful explosion! (but still weak with Eexp~1 5 erg) The mass of NS is ~1.3 M The simulation was continued in 1D to follow the PNS cooling phase up to ~7 s p.b. Yudai Suwa, Kemer, Turkey 22/25

23 Yudai Suwa, Kemer, Turkey 23/25 From SN to NS-2 [Suwa, PASJ, 66, L1 (214)] ν (C)NASA ΓxThermal energy = Coulomb energy Z=7 Z=5 Z=26 Crust formation! (Ze) 2 rk B T = Coulomb energy Thermal energy 2

24 Yudai Suwa, Kemer, Turkey 24/25 Implications Crust formation time should depend on EOS (especially symmetry energy?) We may observe crust formation via neutrino luminosity evolution Cross section of neutrino scattering by heavier nuclei or pasta is much larger than that of neutrons and protons Neutrino luminosity may suddenly drop when we have heavier nuclei! Magnetar (large B-field NS) formation competitive process between crust formation and magnetic field escape from NS

25 Yudai Suwa, Kemer, Turkey 25/25 Summary Supernova explosions by neutrino-heating mechanism have become possible Consistent modeling from iron cores to (cold) neutron stars (i.e. until NS crust formation) is doable now related to neutrino observations, magnetar formation, NS pasta, nuclear EOS...

26 Yudai Suwa, Kemer, Turkey 26/25 Announcement A long-term workshop at Yukawa Institute for Theoretical Physics in Kyoto University Nuclear Physics, Compact Stars, and Compact-star Mergers (NPCSM216) Oct. 17 (Mon.) -- Nov. 18 (Fri.), 216 Please join us!