Konstantin Yakunin Joint Institute for Computational Sciences Oak Ridge National Laboratory. 3/29/17 Particle Physics and Astro-Cosmology Seminar, UTK

Size: px

Start display at page:

Download "Konstantin Yakunin Joint Institute for Computational Sciences Oak Ridge National Laboratory. 3/29/17 Particle Physics and Astro-Cosmology Seminar, UTK"

Bryan Grant
5 years ago
Views:

1 Konstantin Yakunin Joint Institute for Computational Sciences Oak Ridge National Laboratory 3/29/17 Particle Physics and Astro-Cosmology Seminar, UTK 1

2 Credit: NASA/Dana Berry, Sky Works Digital Credit: LSC Credit: LSC Credit: Emil Ivanov

3 The first GW signals were detected on 14 September 2015 and 26 December 2015 GW Duration 0.2 s Distance 440 ± 160 MPc M1 = 36 and M2 = 29 Frequency: Hz SNR = 24 (σ =5.1) GW Duration 1.0 s Distance 440 ± 180 MPc M1 = 14.2 and M2 = 7.5 Frequency: Hz SNR = 13 (σ =5.0)

4 CCSNe Observation Formal MOU!

5 Credit: StudyBlue Strongest GW signal: Rotating progenitor Non-rotating progenitor

SASI Explosion Energy versus Progenitor Mass

Progenitors; 256 x 256 Spatial Resolution 0.

Convection Explosion Energy [B] 0.6 0.4 0.

progenitor W-H 20 solar mass progenitor W-H 25

6 SASI Explosion Energy versus Progenitor Mass Wossley-Heger 12, 15, 20, 25 Solar Mass Nonrotating Progenitors; 256 x 256 Spatial Resolution 0.8 Core Bounce PNS Instabilities Neutrino-Driven Convection Explosion Energy [B] W-H 12 solar mass progenitor W-H 15 solar mass progenitor W-H 20 solar mass progenitor W-H 25 solar mass progenitor Time from bounce [s] Explosion 3/29/17 6

Non-rotating or slowly rotating progenitors Long

signal (< 50 ms), high amplitude at bounce

7 Non-rotating or slowly rotating progenitors Long signal (> 1 sec), low, moderate amplitude Bruenn et al. 2016, Ap.J. 818, 123 Rapidly rotating progenitors Short signal (< 50 ms), high amplitude at bounce Richers et al. 200=17, arxiv: Burrows et al. 2007, Ap.J. 664, 416 3/29/17 7

Slowly Rotating Prompt convection Neutrino-driven convection & SASI PNS convection Rapidly Rotating Bounce/ringdown of millisecond PNS low T/ W

8 Slowly Rotating Prompt convection Neutrino-driven convection & SASI PNS convection Rapidly Rotating Bounce/ringdown of millisecond PNS low T/ W instabilities Rapidly rotating progenitor ~ 1% of expected CCSNe Rotation profile is parameterized by central angular velocity and a differential rotation

9 Self-Consistent Supernova Model

Distance from symmetry axis [km] Distance from symmetry axis [km] 1000 500 0 500 1000 1000 500 0 500 1000 Bruenn et al.

10 Distance from symmetry axis [km] Distance from symmetry axis [km] Bruenn et al. ApJ, 818, 123 (2016) D Explosion Models B12-WH B15-WH an explosion energy of B. Using the IIP-analytical model of Popov (1993) and the simulations of Litvinova & Nadezhin B20-WH07 (1985), Tsvetkov et al. (2006) obtain an B25-WH07 explosion energy of B. Takáts & Vinkó (2006), from the Nadyozhin (2003) formulae, derive an explosion energy of B. Using the semi-analytical model of Zampieri 500 et al. (2003), Pastorello et al. (2009) derive an explosion energy of B. The radiation-hydrodynamics modeling of Utrobin & Chugai (2008) gives an explosion energy of ± 0.03 B for SN 2005cs. SN 2009kr SN 2009kr was discovered 2009 November in the spiral galaxy NGC 1832, and identified as either a Type IIL (Elias-Rosa et al. 2010) or a transitional event between the Type IIL and the Type IIP (Fraser et al. 2010) Archival imaging of the supernova site reveals a yellow supergiant progenitor with estimated M ZAMS of M (Elias-Rosa et al. 2010) and M (Fraser et al. 2010) The latter authors criticized the former for not comparing Distance along symmetry axis the [km] measured luminosity of the Distance progenitor along symmetry withaxis models [km] at the end of core helium burning, as recommended by Smartt (2009). Taking the latter estimate of the progenitor M ZAMS and v p,50d = 4960 ± 280 km s 1 (Poznanski 2013), we infer an explosion energy of B using the Dessart et al. (2010) grid. No estimates of M Ni56 have been published. Energy Explosion [B] Energy [B] E + = Energy sum over positive energy zones E + ov = E + + Overburden E + ov, rec = E + ov + Nuclear recombination B12-WH07 B15-WH07 B20-WH07 B25-WH07 Available nuclear recombination energy ( E rec ) Time After Bounce [ms] Axisymmetric Core-Collapse Supernova Simulations 29 Figure Panel a: Diagnostic energy, E + (dashed lines), E ov + (dash-dotted lines) including binding energy of the overburden on and off the grid, and E ov,rec + (solid lines) including estimated gain from nuclear recombination, E rec, 0.1 plotted versus time after bounce for all model using colors in Figure 1. Panel b: Estimate of potentially recoverable SN 1993J nuclear recombination energy E rec as described in the text SN 2004et SN 1987A 56 Ni Mass [M ] this0.06 negative contribution as the binding energy of the overburden. The original SN 2004Aprogenitor binding energies are plotted in Figure 13, and the fractionssn of2012aw these progenitors mapped to 0.04 the grids at the initiation of our simulations are indicated by the region interior to the tick marks. The off-grid overburden binding 0.02 SN 2005cs SN 2004dj energies for the progenitors used in our simulations (vertical ticks in Figure 13) are , , , and B, respectively, for the 12, 15, 20, and 25 M progenitors. These binding 10 energies do 15 not change20appreciably25 during the course of our simulations. ZAMS Progenitor TheMass overburden [M ] energy that we consider is the total energy of all negative energy zones Figure 28. Inferred ejected on the grid that lie above 56 Ni mass, M the innermost Ni56, for Type IIP supernovae positive energy zones with M plus the Ni56 estimates described in Section 4.2 whose progenitors have been observed ontotal archival energy images. of the Progenitor off-grid material. The overburdencorrected diagnostic energy, E ov + M ZAMS E + estimates are those used in Figure 27. Ejected M Ni56 of our models are shown + overburden by the filledenergy, red circles. is plotted As in theinpreceding Figure graph, 12a (dash-dotted the error bars onlines) our simulation and given results at are the unknown. time of this report in Table 2. It is delayed in growth relative to E +, and reaches positivity at about 350, 380, 530, context and 650ofms theafter observations bounce, respectively, rather than to forsuggest B12-WH07, that these B15- models, WH07, with B20-WH07, their many andapproximations, B25-WH07. are a complete rep- Binding Energy [B] Binding Energy [B] Figu velop mass the c evol -ri tion ener boun and free part othe neut of tion

11 Structure of GW Signal from 2D model Yakunin et al. PRD, 92, (2015) Prompt convection (30 ms) SASI and active accretion on PNS (228 ms) Explosion and shock expansion (780 ms)

12 3/29/17 Yakunin et al PRD

13 Results obtained with the CHIMERA GR multiphysics supernova code with state-of-the-art neutrino interactions. 3/29/17 13 Yakunin et al PRD

14 Simulations: signal characteristics (f, A, etc); physical mechanism producing GW signals, bank of waveforms Signal Search: search algorithms based on the most reliable parts of waveforms, proposal of detector design to observe physical properties of supernovae via GW signals 3/29/17 14

15 SNR = 41 SNR = 6 3/29/17 15 Thanks to Marek Szczepanczyk

17 5

18 Evolution of ground based detectors Third generation: From 10 to 1Hz 10 x lower thermal noise 10 x times lower quantum (shot) noise

19 The Search for GWs from CCSNe The image part with relationship ID rid2 was not found in the file. GW emission from core-collapse detectable out to ~100kpc Extreme post-corecollapse GW emission models detectable out to ~10-15 Mpc 3] Gill et al aligo CCSNe Detection Distance Waveform cwb Distance 50% hrss FC Distance (Mpc) GRB Distance (Mpc) LB LB LB LB LB Piro Piro Piro Piro G CCSNe Detection Distance Waveform cwb Distance 50% hrss FC Distance (Mpc) GRB Distance (Mpc) Muller1-N Muller1-L Muller1-W N/A 1 Yak Yak Yak Yak

20 CCSNe Rate within 20 Mpc Rate of CCSNe per Century M81 Local Group Group Li 2011 Galaxy Conversion Cappellaro 1996 Galaxy Conversion Virgo Cluster 0 3] Gill et al Distance (Mpc)

21 SNEWS: SuperNova Early Warning System LVD Daya Bay snews.bnl.gov Super-K HALO KamLAND IceCube Borexino

22 Neutrino Detectors Expect time of flight delay for massive neutrinos Distance reach of detectors SK will see ~1 event from Andromeda; HK will get a ~dozen

23 Summary of supernova neutrino detectors Detector Type Location Mass (kton) 10 kpc Status Super-K Water Japan Running Galactic sensitivity Extragalactic LVD Scintillator Italy Running KamLAND Scintillator Japan Running Borexino Scintillator Italy Running IceCube Long string South Pole (600) (10 6 ) Running Baksan Scintillator Russia Running HALO Lead Canada Running Daya Bay Scintillator China Running NOνA Scintillator USA Running MicroBooNE Liquid argon USA Running SNO+ Scintillator Canada Under construction DUNE Liquid argon USA Future Hyper-K Water Japan ,000 Future JUNO Scintillator China Future PINGU Long string South pole (600) (10 6 ) Future plus reactor experiments, DM experiments...

24 LIGO-CCSN Collaboration CCSNe Theory CCSNe Data Analysis LIGO Burst Data Analysis

25 2D From 2D to 3D Ray by Ray structure 3D 512(r) x 256(θ) à 256 processors Angular resolution < 1 540(r) x 180(θ) x 180(ɸ) à processors Angular resolution ~ 2 Efficiency of the code: 100 ms/month à 100 ms/week

26 2D vs 3D Supernova Explosion Lentz et al. ApJL, 807, L31 (2015) Shock radius [km] Shock Radius C15-3D C15-2D 700 C15-1D 650 Mean shock radius 600 Minimum/maximum Time [ms] Figure 1. Mean (solid) shock radius for models C15-3D (green), C15-2 Diagnostic energy [B] a) C15-3D C15-2D Explosion Energy Time [ms]

27 Comparisons use same time window (from 3D) and temporal resolution (from 2D). Results obtained with the CHIMERA GR multiphysics supernova code with state-of-the-art neutrino interactions. 3/29/17 27 Yakunin et al. 2017, arxiv: v1

28 3/29/17 Yakunin et al. 2017, arxiv: v1 28

29 Most reliable part of signal Most reliable part of signal in frequency domain 3/29/17 Andresen et al s20 A+[cm] A, s11.2 A+[cm] A, T ]

30 Dimmelmeieret al. PRD, ,2008 Kotake et al. PRD, , 2003 Schreidegger et al. A&A, 2010 Richerset al. arxiv: /29/17 30

A possible bounce signal Emission process

E GW [Mc 2 ] Core Bounce 10 300 3x10-21 ~10-8

3x10-21 ~10-12 SASI/ND convection 450 700

31 A possible bounce signal Emission process Duration (ms) f peak [Hz] Typical h at 10 kpc E GW [Mc 2 ] Core Bounce x10-21 ~10-8 Prompt convection x10-21 ~10-12 SASI/ND convection x x10-9 Δt/100ms Explosion 3/29/17 > x x

32 Simulations help to improve data analysis and increase chances for detection! o We are able to perform realistic 3D simulations and produce reliable waveforms. o Waveforms from 2D simulations have similar characteristics as 3D ones. Thus, 2D simulations can be used to create a bank of waveforms. Now, even realistic 2D simulations are computationally inexpensive. o It would be good to summarize the main characteristics of GW signals into a table in any publication that presents new waveforms o To produce more realistic waveforms we have to perform realistic CCSN simulations with slow-rotating progenitor (bounce signal + neutrino-driven explosion signal) 3/29/17 32

33 Blondin Mauney Casanova Chu Endeve Hix Landfield Lentz Lingerfelt Messer Mezzacappa Roberts Yakunin Bruenn Marronetti Funded by 3/29/17 Harris 33

Core-collapse Supernove through Cosmic Time...

Core-collapse Supernove through Cosmic Time... Eric J Lentz University of Tennessee, Knoxville S. Bruenn (FAU), W. R.Hix (ORNL/UTK), O. E. B. Messer (ORNL), A. Mezzacappa (UTK), J. Blondin (NCSU), E. Endeve