Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Philipp Mösta
Einstein fellow @ UC Berkeley [email protected]
NucAstro in the GW astronomy era ECT*, Trento Jun 13, 2017
The most powerful explosions in 3D
Philipp Mösta
Einstein fellow @ UC Berkeley [email protected]
NucAstro in the GW astronomy era ECT*, Trento Jun 13, 2017
Status of core-collapse supernova
simulations
3
~
Extreme core-collapse hyperenergetic superluminous lGRBs
Core-collapse supernovae neutrinos turbulence
Magnetic fields in high-energy astro
(Binary) black holes accretion disks EM counterparts
Binary neutron stars gravitational waves EM counterparts sGRBs
4
New era of transient science
Image: PTF/ZTF/COO Image: LSST
• Current (PTF, DeCAM, ASAS-SN) and upcoming wide-field time domain astronomy (ZTF, LSST, …) -> wealth of data
• adv LIGO / gravitational waves detected
• Computational tools at dawn of new exascale era
5
New era of transient science
Image: PTF/ZTF/COO Image: LSST
Transformative years ahead for our understanding of these events
• Current (PTF, DeCAM, ASAS-SN) and upcoming wide-field time domain astronomy (ZTF, LSST, …) -> wealth of data
• adv LIGO / gravitational waves detected
• Computational tools at dawn of new exascale era
6
Observational Facts
~5 per second in universe ~1 per day observed
large kinetic energies ~1051 erg
SN 1
987A
© A
nglo
-Aus
tral
ian
Obs
erva
tory
Progenitor BSG Sanduleak -69 220a, 18 MSUN
7
~
Astrophysics of core-collapse supernovae
Galaxy evolution/feedbackM82/Chandra/NASA
Heavy element nucleosynthesis
Birth sites of black holes / neutron stars Neutrinos
8
Red Supergiant Betelgeuse D ~200 pc
300 km
800 million km
HST
Central Engine”
- EM waves (optical/UV/X/Gamma): secondary information, late-time probes of engine
Observing core-collapse supernovae
9
Red Supergiant Betelgeuse D ~200 pc
300 km
800 million km
HST
Central Engine”
- Neutrinos - EM waves (optical/UV/X/Gamma):
secondary information, late-time probes of engine
Observing core-collapse supernovae
Observing core-collapse supernovae
10
Red Supergiant Betelgeuse D ~200 pc
300 km
800 million km
HST
Central Engine”
- Gravitational waves - Neutrinos - EM waves (optical/UV/X/Gamma):
secondary information, late-time probes of engine
Core-Collapse Supernova Energetics
11
• Collapse to a neutron star: 3 x 1053 erg = 300 [B]ethe gravitational energy (≈0.15 MSunc2) released. Initially stored in protoneutron star. -> Any explosion mechanism must tap this reservoir.
•1051 erg = 1 B kinetic and internal energy of the ejecta. (Extreme cases: 10 B; hyperenergetic supernova)
•99% of the energy is radiated in neutrinos over tens of seconds in protoneutron star cooling.->Strong evidence from SN 1987A neutrino observations.
SN 1987A: Neutrino Detection
12
13
Gas/plasma dynamicsMagneto-Hydrodynamics
A multiphysics challenge
14
Gas/plasma dynamics
Gravity
Magneto-Hydrodynamics
General Relativity
A multiphysics challenge
Figure: C. Reisswig
ADM 3+1 split of spacetime
3-hyper-surface
• 12 first-order hyperbolic evolution equations • 4 elliptic constraint equations • 4 coordinate gauge degrees of freedom: α, βi
Dynamical gravity / Numerical Relativity
15
Gµ =8G
c4Tµ
Tim
e
Figure: C. Reisswig
ADM 3+1 split of spacetime
3-hyper-surface
Dynamical gravity / Numerical Relativity
16
• Required for black hole collapse and gravitational waves!
Tim
e
17
Gas/plasma dynamics
Nuclear EOS, nuclear reactions & ν interactions
Gravity
Magneto-Hydrodynamics
Nuclear and Neutrino Physics
General Relativity
A multiphysics challenge
18
Gas/plasma dynamics
Nuclear EOS, nuclear reactions & ν interactions
Gravity
Neutrino transport
Magneto-Hydrodynamics
Nuclear and Neutrino Physics
General Relativity
Boltzmann Transport Theory
A multiphysics challenge
19
Gas/plasma dynamics
Nuclear EOS, nuclear reactions & ν interactions
Gravity
Neutrino transportFully
cou
pled
!
Magneto-Hydrodynamics
Nuclear and Neutrino Physics
General Relativity
Boltzmann Transport Theory
All four forces!
A multiphysics challenge
20
Magneto-Hydrodynamics
Nuclear and Neutrino Physics
General Relativity
Boltzmann Transport Theory
Nuclear EOS, nuclear reactions & ν interactions
Neutrino transportFully
cou
pled
!
Additional Complication: Core-Collapse Supernovae are 3D
• rotation • fluid and MHD instabilities, multi-D structure, spatial scales
Need 21st century tools: • cutting edge numerical algorithms • sophisticated open-source software infrastructure • peta/exa scale computers
All four forces!
A multiphysics challengeGas/plasma dynamics
Gravity
Core collapse basics
21
up to
[not drawn to scale]
Fe-group nucleiSi
O/Ne/MG
C
8M . M . 130M
[not drawn to scale]
22
2000km
Protoneutron star r~30km
Iron core
Nuclear equation of state stiffens at nuclear density
Inner core (~0.5 ) -> protoneutron star + shockwave
M
2000km
Core collapse basics
23
2000km
Outer core accretes ontoshock & protoneutron star with O(1) /s
Shock stalls at ~ 100 km
Reviews:Bethe’90 Janka+‘12
Protoneutron star r~30km
Iron core
M
2000km Nuclear equation of state stiffens at nuclear density
Inner core (~0.5 ) -> protoneutron star + shockwave
M
Core collapse basics
accretion
L
shock
The Core-Collapse Supernova Problem
The shock always stalls:
• Dissociation of Fe-group nuclei @ ~8.8 MeV/baryon (~17 B/MSun).
• Neutrino losses initially @ >100 B/s (1 [B]ethe = 1051 ergs).
24
Hans Bethe 1906-2005
~
25
2000km
Core-collapse supernova problem: How to revive the shockwave?
Reviews:Bethe’90 Janka+‘12
accretion
L
shock
Protoneutron star r~30km
Iron core
2000km Nuclear equation of state stiffens at nuclear density
Inner core (~0.5 ) -> protoneutron star + shockwave
M
Core collapse basics
26
2000km
Engine formation?
Reviews:Bethe’90 Janka+‘12
accretion
L
shock
Protoneutron star r~30km
Iron core
2000km Nuclear equation of state stiffens at nuclear density
Inner core (~0.5 ) -> protoneutron star + shockwave
M
Core collapse basics
27
2000km
accretion
L
shock
2000km
Core collapse basicsNeutrino mechanism
28
2000km
accretion
L
shock
2000km
Core collapse basicsNeutrino mechanism
29
2000km
accretion
L
shock
2000km
Core collapse basicsNeutrino mechanism
Theory incomplete!
30
2000km2000km
Roberts+16
3D Volume Visualization of
Entropy
Core collapse basics
Neutrino Mechanism
Ott+ ’08
31
Cooling:
Heating via charged-current absorption:
Bethe & Wilson ’85; also see: Janka ‘01, Janka+ ’07
30 km 60 km 120 km 240 kmNeutrino radiation field:
Heating efficiency:
32
Frontiers in neutrino mechanism
Key for explosion:
compactness
turbulence
SASI
Janka+16
Takiwaki+16
Frontiers in neutrino mechanism
Does rotation help?
34
Frontiers in neutrino mechanism
Current frontiers:
turbulence detailed neutrino transport numerical resolution progenitor burning
Radice+15
Couch+15
Erik’s talk for details!
Discuss!
35
2000km2000km
Obergaulinger+14
Very high magnetic field in non-rotating progenitors can aid explosion (in 2D)
Convection/SASI can amplify field at fixed amplification factor
Most models explode as hydro!
Frontiers in neutrino mechanism
Hypernovae & GRBs
36http://apod.nasa.gov/apod/ap000628.html
Hypernovae & GRBs
37
• But not all stripped-envelope supernovae come with GRBs
• Trace low metallicity environments
• 11 long GRB – core-collapse supernova associations.
• All GRB-SNe are stripped envelope, show outflows v~0.1c
Neutrino mechanism is inefficient; can’t deliver a hypernova
38
Superluminous supernovae
Some events: stripped envelope no interaction
Gal-Yam+12
Elum ~ 1045 erg Erad up to 1052 erg
39
Lunnan+14
Connection between SLSN Ic and lGRBs • prefer star-forming galaxies • low metallicity • large core angular momentum !?
Superluminous supernovae
Greiner+15
Common engine? Magnetar?
40
Superluminous / hyperenergetic supernovae
SLSN Ic SN Ic-bl lGRBs
Common engine? Magnetar?
41
Superluminous / hyperenergetic supernovae
SLSN Ic SN Ic-bl lGRBs
FRBs
Protomagnetar powered explosions
42
Rapid Rotation + B-field amplification
2D: Energetic bipolar explosions Energy in rotation up to 1052 erg
Results in ms-period proto-magnetar
Protomagnetar powered explosions
Obergaulinger+17
2D: Diversity of outcomes depending on rotation rate/progenitor profile
44Octant Symmetry (no odd modes) identical to 2D Full 3D
! 2000 km " ! 2000 km "
3D explosions dynamics very different!PM+ 14
What’s going on here?
45
• m=1 spiral instability • consistent with MHD kink instability;
should hold independent of initial B-field strength
H (Btor )
5 10 15 20 25105
104
103
102
101
1
t tbounce [ms]
r[k
m]
magnetic pressure b2/2
zi = 30 kmzi = 44 kmzi = 59 kmzi = 74 kmet/t , t = 1.4 ms
zi = 30 kmzi = 44 kmzi = 59 kmzi = 74 kmet/t , t = 1.4 ms
fgm
4a
p
Btor
1ms
fgm
4aBz
Btor
5km
PM, Richers+ 14with Sherwood Richers (Caltech)
46
• B-field near proto-NS: Btor >> Bz
• Unstable to MHD screw-pinch kink instability.
• Similar to situation in Tokamak fusion reactors!
Braithwaite+ ’06
Credit: Moser & Bellan, CaltechSarff+13
MHD Kink Instability
3D Volume Visualization of
47
EntropyPM+ 14
dual-lobe ‘slow’ explosion
48
PM+ 17 (in prep.)
180 200 220 240 260 280800
900
1000
1100
1200
1300
1400
t tbounce [ms]
r[k
m]
Continued accretion -> Black hole engine possible!
Implications for long Gamma-Ray Bursts
Jet-driven explosions proposed as site for r-process
49
Neutron-rich nucleosynthesis in supernovae Creating the heaviest elements
• Low electron fraction
• Medium entropy • Low density • High
temperature
Sneden+ 08
50
R-process: First results
PM, Roberts+ 17 (in prep)Halevi, PM, Roberts+ 17 (in prep)
51
Winteler+12
Nishimura+15
R-process in jet-driven supernovae
Robust r-process in 2D simulations for prompt explosions!
52
R-process in jet-driven supernovae
Halevi, PM 17 (in prep)
B = 1013 G
53
R-process in jet-driven supernovae
80 100 120 140 160 180 2001016
1014
1012
1010
108
106
104
102
100
mass number A
Mej
ecta
[M]
L = 0L = 1051 erg/sL = 1052 erg/sL = 1053 erg/s
L = 0L = 1051 erg/sL = 1052 erg/sL = 1053 erg/s
80 100 120 140 160 180 2001016
1014
1012
1010
108
106
104
102
100
mass number A
Mej
ecta
[M]
L = 0L = 1051 erg/sL = 1052 erg/sL = 1053 erg/s
L = 0L = 1051 erg/sL = 1052 erg/sL = 1053 erg/s
PM, Roberts, Halevi+ 17 (in prep)
B = 1012 G / octant
B = 1012 G full 3D
54
Gravitational waves from rotating core collapse
Fuller+15
Richers+17
Gravitational waves from rotating core collapse
Gossan+16
see Jade’s talk later in afternoon
Magnetorotational Mechanism
56Burrows+’07
Caveats:
1) Need high core spin; only in very few progenitor stars?
2) Magnetic field amplification?
57Burrows+’07
Caveats:
1) Need high core spin; only in very few progenitor stars?
• chemically homogenous evolution • binary interactions • but angular momentum loss due to
magnetic braking or mass loss?
Likely rare but can work outBurrows+’07
Magnetorotational Mechanism
58
Caveats:
2) Magnetic field amplification?
• Progenitor magnetic field strength expected no stronger than 10^9/10^10 G • Need > 10^15 G after core bounce • Flux compression and winding only
yield factor 10^3 amplification
At least another factor 10^3 missing!Burrows+’07
Magnetorotational Mechanism
59
One proposed channel: MRI + dynamo
MC Mi Mo
60
Stability criterion:
[Balbus&Hawley 91,98, Akiyama+03, Obergaulinger+09]
What’s the situation in core-collapse?
82 < !2BV + r
d2
dr< 0
Magnetic field amplificationMRI activated locally Akiyama+03, Shibata+06, …
M. Obergaulinger et al.: MRI in core collapse supernovae 259
Fig. 22. Volume rendered magnetic field strength of a model with bz0 = 4 × 1013 G computed in a box of 1 × 2 × 1 km3 with a resolution of 20 m at
t = 21.5 ms (left) and t = 37.2 ms (right), respectively. The coordinate directions are indicated as in Fig. 19.
the weakest initial field, because the theoretical growth rate forthe fastest growing MRI mode is σMRI = 1.14 ms−1 for thesemodels (see Sect. 4.2.1).
To investigate the stability properties of large-scale channelmodes as a function of the box geometry, we simulated mod-els with an initially uniform magnetic field using boxes of dif-ferent size and shape. The models were rotating according toΩ0 = 1900 s−1 and αΩ = −1.25, and their initial magneticfield was bz
0 = 4 × 1013 G when applying velocity dampingboundaries, and bz
0 = 2 × 1013 G otherwise. We varied boththe ratio between the radial and vertical, Lϖ/Lz, and the radialand toroidal size, Lφ/Lϖ, size of the box. The grid resolutionwas 20 m (see Table B.3). Plotting the stress ratios Mmax
ϖφ /Mtermϖφ
(Fig. 23; damping boundaries) and ⟨Mϖφ⟩/Mtermϖφ (Fig. 24; non-
damping boundaries) as a function of the aspect ratio of the com-putational box, provides some indication of the range of Mϖφvalues prevailing during the post-growth phase. The ratios al-low one to distinguish models with a strong variability due tothe dominant re-appearance of channel modes from those mod-els exhibiting a smooth evolution without dominant large-scalecoherent structures.
We find that models with a radial aspect ratio Lϖ/Lz = 1and a toroidal aspect ratio Lφ/Lz ≥ 2 are unstable against par-asitic instabilities, independent of the grid resolution in toroidaldirection. Turbulence develops and leads to a flow structure asshown in Fig. 22. Models having the same radial aspect ratio,but a smaller toroidal one are stable and evolve similarly as ax-isymmetric models, i.e., parasitic instabilities do not grow anda dominant large-scale channel flow develops, which gives riseto a morphology of the type presented in Fig. 22. These find-ings do not depend on how the growth of the MRI ends, i.e.,whether velocity damping is applied and reconnection betweenadjacent channels occurs inside the box, or whether no damp-ing is imposed and reconnection occurs near the surface of thecomputational box.
These results can be understood from the analysis of par-asitic instabilities by Goodman & Xu (1994), who argued thatthree-dimensional flows are unstable against parasitic instabili-ties, but these instabilities can be suppressed by the geometry ofthe computational box. According to their analysis, the growthrate of the parasitic instabilities is highest for modes with halfthe wave number of the unstable MRI modes they are feeding
1log Lϖ / Lz
1
log
Lφ /
Lz
log Mϖφmax / Mϖφ
term
0 0.75 1.50 2.25 3
1log Lϖ / Lz
log Mϖφavge / Mϖφ
term
−0.25 0.25 0.75 1.25 1.75 2.25
Fig. 23. The left panel shows the ratio of the maximum Maxwell stressper unit volume, Mmax
ϖφ , and its value at MRI termination, Mtermϖφ , as a
function of the toroidal and radial aspect ratios, Lφ/Lz and Lϖ/Lz forthe models listed in Table B.3. The right panel shows the ratio of Mϖφaveraged over the saturation phase and the value at MRI termination.Each model is represented by a symbol its color reflecting its maximumMaxwell stress. All models are computed imposing velocity dampingat the radial grid boundaries.
off. Hence, if a channel flow forms at late times with a wave-length equal to the box size in z-direction, Lz, unstable parasiticmodes must have a toroidal wavelength ∼2Lz to grow rapidly.Thus results in the criterion for the channel flow instability wehave found in our simulations.
In accordance with simulations presented recently by Bodoet al. (2008), we find that models with a radial aspect ratioLϖ/Lz ≤ 1 experience a second exponential growth phase asdescribed in Sect. 4.2 (note the large ratios of Mmax
ϖφ and Mtermϖφ
for the corresponding models in Fig. 23), whereas a larger ra-dial aspect ratio appears to favor a less violent post-growth phasewhere coherent channel modes can appear but are disrupted aftera short time. Bodo et al. (2008) obtained this result for simula-tions performed with a toroidal aspect ratio Lφ/Lz = 4.
Obergaulinger+09
Saturation strength? Secondary instabilities, neutrino viscosity and drag Guilet+15, Rembiasz+16, …
Rembiasz+16
Global 2D studies Sawai+13,14, …
Guilet+15
62
First global 3D MHD turbulence simulations
Do MRI and dynamo build up dynamically relevant global field?
• 10 billion grid points (Millenium simulation used 10 billion particles)
• 130 thousand cores on Blue Waters
• 2 weeks wall time
• 60 million compute hours
• 10000 more expensive than any previous simulations
PM+ 15 Nature
63
3D magnetic field structure
dx=500m dx=50mdx=200m dx=100m
PM+ 15 Nature
64
PM+15 Nature
65
Growth at Large Scales
saturation within 60ms
0 5 10
1
2
3
4
5
6
7
t tmap [ms]
E k,m
ag(t
)[1
033er
g]
(2.05 + 0.75 ms1 · (t tmap)) · 1033
5 · 1032 e(ttmap)/t, t = 3.5 msk = 4k = 6k = 8k = 10k = 20k = 50k = 100
(2.05 + 0.75 ms1 · (t tmap)) · 1033
5 · 1032 e(ttmap)/t, t = 3.5 msk = 4k = 6k = 8k = 10k = 20k = 50k = 100
PM+15 Nature
66
Global Field Structure
t=0ms
dx=500m dx=50m
t=10ms t=10msPM+15 Nature
67
Global Field Structure
t=0ms
dx=500m dx=50m
t=10ms t=10ms
Magnetar formation?
PM+15 Nature
68
From simulations to observationsState of the art now:
Current frontier:
Detailed simulations full physics 0.1-1s inner core ~10000km
PM, Tchekhovskoy 17 (in prep)
Full 3D, full physics
Full star
1) Engine model from full-physics simulations 2) Simplified simulations with engine model to shock breakout
69
From simulations to observationsState of the art now:
Current frontier:
Detailed simulations full physics 0.1-1s inner core ~10000km
Squire, PM, Lecoanet 17 (in prep)
1) Engine model from full-physics simulations 2) Simplified simulations with engine model to shock breakout
0 5 10 15 20 25
-20
-10
0
10
20
-5.0×10-7
5.0×10-7
1.5×10-6
2.5×10-6
Daedalus simulationGet mean fields from 3D sim
70
From simulations to observationsState of the art now: Future:
Full-star simulations full physics shock breakout
detailed light curves detailed spectra connect observations and engines map progenitor params
Detailed simulations full physics 0.1-1s inner core ~10000km
Current frontier:
1) Engine model from full-physics simulations 2) Simplified simulations with engine model to shock breakout
71
Summary3D explosion dynamics matter for neutrino-driven
and GRB-SN; details need to be worked out
Very important to work out multimessenger signatures; sensitive to nuclear/neutrino physics
Need subgrid (or similar models) to capture relevant dynamics from high-resolution simulations
3D hybrid simulations to predict observable signatures reliably
72
Thank you!