magnetorotational processes in core-collapse supernovae

Magnetorotational processes in core-collapse supernovae.

Sergey Moiseenko, Gennady Bisnovatyi-KoganSpace Research Institute,

Moscow, Russia

Outline•Magnetorotational(MR) mechanism of supernova explosion.

•Numerical method (implicit Lagrangian scheme)

•Core-collapse simulations.

•MR supernova with quadrupole field.

•Magneto-Differential-Rotational Instability(MDRI) in MR supernova.

•MR supernova with dipole field, jet formation.

•MR supernova with different core masses and rotation rates.

•MR supernova with different equations of state.

•Mirror symmetry violation of the magnetic field in rotating stars – one-sided jets, kicks.

•Conclusions.

Magnetorotational mechanism for the supernova explosion Bisnovatyi-Kogan (1970)(original article was submitted: September 3, 1969)

Amplification of magnetic fields due to differential rotation, angular momentum transfer by magnetic field. Part of the rotational energy is transformed to the energy of explosion

First 2D calculations: LeBlanck&Wilson (1970) )(original article was submitted: September 25, 1969) ->too large initial magnetic fields. Emag0~Egrav axial jetBisnovatyi-Kogan et al 1976, Meier et al. 1976, Ardeljan et al.1979, Mueller & Hillebrandt 1979, Symbalisty 1984, Ardeljan et al. 2000, Wheeler et al. 2002, 2005, Yamada & Sawai 2004, Kotake et al. 2004, 2005, 2006, Burrows et al.2007, Sawai, Kotake, Yamada 2008, Sawai et al.2013,…

It is popular now!The realistic values of the magnetic field are: Emag<<Egrav ( Emag/Egrav= 10-8-10-12)

Small initial magnetic field -is the main difficulty for the numerical simulations.

The hydrodynamic time scale is much smaller than the magnetic field amplification time scale (if magnetorotational instability is neglected).

Explicit difference schemes can not be applied. (CFL restriction on the time-step).

Implicit schemes should be used.

Basic equations: MHD +self-gravitation, infinite conductivity:

, div 0,

1grad div( ) grad

div ( , ) 0, ( , ), ( , ),

G ,

.

dx d

dt dtdu

pdt

dp F T p P T T

dt

d

dt

u u

H HH H

u

HH u

Axis symmetry ( ) and equatorial symmetry (z=0) are supposed.0

Notations:

Additional condition divH=0

, (r, z), velocity, density, pressure,

magnetic field, gravitational potential, internalenergy,

G gravitationalconstant.

dp

dt t

u x u

H

Boundary conditions

r

poloidal q

0 : rot rot 0,

Equatorial symmetry

0,

0 : or

0,

Outer boundary: 0,

r r

z z

zz

r u u H H

u H

z

Bu

z

P T H

H H

H H

(from Biot-Savart law)

Quadrupole field

Dipole field

Axial symmetry

Presupernova Core CollapseEquations of state take into account degeneracy of electrons and neutrons,

relativity for the electrons, nuclear transitions and nuclear interactions. Temperature effects were taken into account approximately by the

addition of radiation pressure and an ideal gas

.

Neutrino losses were taken into account in the energy equations.

A cool white dwarf was considered at the stability limit with a mass equal to the Chandrasekhar limit.

To obtain the collapse we increase the density at each point by 20% and we also impart uniform rotation on it.

3)(),(

4

0

TTPTPP

.6,2,10

),1()(

)1()419.8(lg)(

0

13/1

13/5

1)1(

00 kforaP

forcbPP

kkbk kc

k

4

0

3( , ) ( ) ( , ),

2 Fe

TT T T

0

20

0 .)(

)( dxx

xP

Equations of state (approximation of tables)

115/2359 ])/101.7(1/[)(103.1),( cgergTTTf

,20,31.664

,207),20(024.5131.664

,7,1

)(

T

TT

T

T.910 TT

Neutrino losses: URCA processes, pair annihilation, photo production of neutrino, plasma neutrino

URCA:

( , ) ( , )F T f T e Approximation of tables from Ivanova, Imshennik, Nadyozhin,1969

, 0,

1, 0,

( , ) ,p

b Fe FeFe

m Fe Fe

E T TT

A T T

Fe –dis-sociation

neutrino diffusion

Operator-difference schemedeveloped by N.V.Ardeljan et al. (Moscow State University)

Method of basic operators (Samarskii) – grid analogs of basic differential operators:

GRAD(scalar) (differential) ~ GRAD(scalar) (grid analog)

DIV(vector) (differential) ~ DIV(vector) (grid analog)

CURL(vector) (differential) ~ CURL(vector) (grid analog)

GRAD(vector) (differential) ~ GRAD(vector) (grid analog)

DIV(tensor) (differential) ~ DIV(tensor) (grid analog)

Implicit scheme. Time step restrictions are weaker for implicit schemes (no CFL condition).

The scheme is Lagrangian=> conservation of angular momentum.

Lagrangian, implicit, triangular grid with rezoning, completely conservative

The stability of the method was explored in Ardeljan & Kosmachevskii Comp. and Math. Modelling 1995, 6, 209 and references therein

Grid reconstructionElementary reconstruction: BD connection is introducedinstead of AC connection. The total number of the knotsand the cells in the grid is not changed.

Addition a knot at the middle of the connection:the knot E is added to the existing knots ABCDon the middle of the BD connection, 2 connectionsAE and EC appear and the total number of cells is increased by 2 cells.

Removal a knot: the knot E is removed from the grid and the total number of the cells is decreased by 2 cells

=>

Interpolation of grid functions on a new grid structure (local): Should be made in conservative way. Conditional minimization of special functionals.

Numerical method testing

The method was tested on the following tests:

1. Collapse of a dust cloud without pressure2. Decomposition of discontinuity problem3. Spherical stationary solution with force-free magnetic field,4. MHD piston problem,

Example of calculation with the triangular grid

Example of the triangular grid

Collapse of rapidly rotating cold protostellar cloud

A&Ass 1996,115, 573

R

Z

0 0.5 10

0.25

0.5

0.75

1

TIME= 0.00001000 ( 0.00000035sec )

RZ

0 0.5 10

0.25

0.5

0.75

1 Density18 5.0704417 4.7324416 4.3944315 4.0564314 3.7184213 3.3804212 3.0424111 2.7044110 2.36649 2.02848 1.690397 1.352396 1.014385 0.6763794 0.3383743 0.04602092 0.005446171 0.000797174

TIME= 0.00001000 ( 0.00000035sec )

Initial state, spherically symmetrical stationary state, initial angular velocity 2.519 (1/sec)

Initial temperature distribution

T

3/2T1.2042 sunM M

int

0.571% 72.7%rot

grav grav

E E

E E

R

Z

0 0.25 0.5 0.750

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Density27186222007820713319418718124116829515535014240411651210356690620.764729.251783.425891.912946.1607.295307.362150.5420.341389

TIME= 4.12450792 ( 0.14246372sec )

R

Z

0 0.005 0.01 0.015 0.020

0.0025

0.005

0.0075

0.01

0.0125

0.015

0.0175

0.02 Density27186222007820713319418718124116829515535014240411651210356690620.764729.251783.425891.912946.1607.295307.362150.5420.341389

TIME= 4.12450792 ( 0.14246372sec )

Maximal compression state

Max. density = 2.5·1014g/cm3

R

Z

0 0.01 0.020

0.005

0.01

0.015

0.02

TIME= 4.12450792 ( 0.14246372sec )

Neutron star formation in the center and formation of the shock wave

«0.01»~10km

R

Z

0 0.2 0.4 0.6 0.80

0.1

0.2

0.3

0.4

0.5

0.6

0.7

TIME= 5.29132543 ( 0.18276651sec )

R

Z

0 0.1 0.2 0.3 0.40

0.1

0.2

0.3

0.4

TIME= 5.29132543 ( 0.18276651sec )

Mixing

Bounce shock wave does not produce SN explosion :(

R

Z

0 0.002 0.004 0.006 0.0080

0.002

0.004

0.006

Angular velocity357.283310.745293.484276.224258.963241.703224.442207.181189.921172.66155.399138.139120.878103.61886.356969.096351.835634.57517.3144

TIME= 4.15163360 ( 0.14340067sec )

Angular velocity (central part of the computational domain). Rotation is differential.

Distribution of the angular velocity

R

Z

0 0.005 0.01 0.015 0.020

0.0025

0.005

0.0075

0.01

0.0125

0.015

0.0175

0.02 Density27186222007820713319418718124116829515535014240411651210356690620.764729.251783.425891.912946.1607.295307.362150.5420.341389

TIME= 4.12450792 ( 0.14246372sec )

R

Angula

rve

locity

0.1 0.2

50

100

150

200

The period of rotation of the young neutron star is about 0.001- 0.003 sec

Rapidly rotating

pre-SN

Crab pulsar P=33ms –

rapid rotation at birth

Core collapse

in binaries

Z

R

O u te rb o u n d ary

Initial magnetic field

,vRc

13

dRJ

H

),( zr HHJ

Biot-Savart law

Initial poloidal magnetic field is divergence free, BUT not balanced. We have to balance it.

Initial magnetic field –quadrupole-like symmetry

R

Z

0 1 2 3 40

0.5

1

1.5

2

2.5

3

3.5TIME= 0.00000000 ( 0.00000000sec )

Ardeljan, Bisnovatyi-Kogan, SM, MNRAS 2005, 359, 333

R

Z

0 0.01 0.020

0.005

0.01

0.015

0.02

TIME= 0.00000779 ( 0.00000027sec )

Toroidal magnetic field amplification.

pink – maximum_1 of Hf^2 blue – maximum_2 of Hf^2Maximal values of Hf=2.5 ·10(16)G

After SN explosion at the border of neutron star H=2 1014G

Temperature and velocity field

Specific angular momentum rV

Time evolution of the energies

time,sec0 0.1 0.2 0.3 0.4

-6.4E+52

-6.3E+52

-6.2E+52

-6.1E+52

-6E+52

-5.9E+52

-5.8E+52

-5.7E+52

-5.6E+52

-5.5E+52

-5.4E+52

Egrav

time,sec0 0.1 0.2 0.3 0.4 0.5

3.6E+52

3.7E+52

3.8E+52

3.9E+52

4E+52

4.1E+52

4.2E+52

Eint

Gravitational energy Internal energy

time,sec0 0.1 0.2 0.3 0.4

1E+52

1.1E+52

1.2E+52

Neutrinolosses

time,sec0 0.1 0.2 0.3 0.4

5E+51

6E+51

7E+51

8E+51

9E+51 Neutrinoluminosity

Neutrino luminosity (ergs/sec)Neutrino losses (ergs)

Time evolution of different types of energies

time,sec0.1 0.2 0.3 0.4 0.5

0

5E+50

1E+51

1.5E+51

2E+51

2.5E+51

3E+51

3.5E+51

4E+51Ekinpol

Erot

Emagpol

Emagtor

Ejected energy and mass

Ejected energy 510.6 10 ergParticle is considered “ejected” –

if its kinetic energy is greater than its potential energy

Ejected mass 0.14M

2.1

time,sec0 0.1 0.2 0.3 0.4

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0.11

0.12

0.13

0.14

Ejected mass/Masssun

time,sec0.1 0.2 0.3 0.4

1E+50

2E+50

3E+50

4E+50

5E+50

6E+50

Ejected energy (ergs)

Magnetorotational supernova in 1D

mag0

grav0explosion

1,

E

Et

Example:2

explosion10 10,t

12 6explosion1 !10 !!0 t

Bisnovaty-Kogan et al. 1976, Ardeljan et al. 1979

time,sec0 0.5 1 1.5

0

5E+50

1E+51

1.5E+51

2E+51

2.5E+51

3E+51

3.5E+51

4E+51

4.5E+51

Erot, (ergs)

time,sec0 0.5 1 1.5

0

1E+50

2E+50

3E+50

4E+50

5E+50Emagtor, (ergs)

time,sec0 0.5 1

0

1E+50

2E+50

3E+50

4E+50

5E+50

6E+50

Ekinpol, (ergs)

time,sec0 0.25 0.5 0.75

0

1E+50

2E+50

3E+50

4E+50

5E+50

6E+50

Emagpol, (ergs)

Magnetorotational explosion for the different 0 2 12

0

10 10mag

grav

E

E

Magnetorotational instability mag. field grows exponentially (Dungey 1958,Velikhov 1959, Chandrasekhar1970,Tayler 1973,

Balbus & Hawley 1991, Spruit 2002, Akiyama et al. 2003…)

Dependence of the explosion time frommag0

grav0

E

E

explosion lo~ ( )gt (for small )

Example:

6explosion10 ~ 6,t

12explosion10 ~ 12.t

Central part of the computational domain . Formation of the MDRI.

R

Z

0.01 0.015 0.020.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

0.011

0.012

0.013

0.014

TIME= 34.83616590 ( 1.20326837sec )TIME= 34.83616590 ( 1.20326837sec )TIME= 34.83616590 ( 1.20326837sec )TIME= 34.83616590 ( 1.20326837sec )

R

Z

0.01 0.015 0.020.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

0.011

0.012

0.013

0.014

TIME= 35.08302173 ( 1.21179496sec )TIME= 35.08302173 ( 1.21179496sec )TIME= 35.08302173 ( 1.21179496sec )TIME= 35.08302173 ( 1.21179496sec )

R

Z

0.01 0.015 0.020.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

0.011

0.012

0.013

0.014

TIME= 35.26651529 ( 1.21813298sec )TIME= 35.26651529 ( 1.21813298sec )TIME= 35.26651529 ( 1.21813298sec )TIME= 35.26651529 ( 1.21813298sec )

R

Z

0.01 0.015 0.020.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

0.011

0.012

0.013

0.014

TIME= 35.38772425 ( 1.22231963sec )TIME= 35.38772425 ( 1.22231963sec )TIME= 35.38772425 ( 1.22231963sec )TIME= 35.38772425 ( 1.22231963sec )

Magneto-Differential-Rotational Instability

Toy model for MDRI in the magnetorotational supernova

; r

dH dH r

dt dr

beginning of the MRI => formation of multiple poloidal differentially rotating

vortexes 0r v

r

dH dH l

dt dl

( )vdl H H

dl

in general we may approximate:

Assuming for the simplicity that is a constant during the first stages of MRI, and taking as a constant we come to the following equation:

ddrr A

H

2

02( )

r

dH H AH H H

dt

0

0

( )0

3 2 1 2( )0

0 1 .

r

r

A H t tr

A H t trr r

H H H

H

e

H HA

e

at the initial stage of the process * :H H const,r

dH r

dr

Initial magnetic field – dipole-like symmetrySM., Ardeljan & Bisnovatyi-Kogan MNRAS 2006, 370, 501

R

Z

0 1 2 3 40

0.5

1

1.5

2

2.5

3

3.5

Magnetorotational explosion for the dipole-like magnetic field

Magnetorotational explosion for the dipole-like magnetic field

Magnetorotational explosion for the dipole-like magnetic field

Ejected energy and mass (dipole)

Ejected energy 510.5 10 erg Particle is considered “ejected” –

if its kinetic energy is greater than its potential energy

Ejected mass 0.14M

time, sec0 0.5 1 1.5

0

1E+50

2E+50

3E+50

4E+50

5E+50

time, sec0 0.5 1 1.5

0

1E+50

2E+50

3E+50

4E+50

5E+50

time, sec0 0.5 1

0.02

0.04

0.06

0.08

0.1

0.12

0.14

Characteristic time of the magnetic field reconnection (rough estimation)

Petcheck mechanism – characteristic reconnection time

Our estimations show: conductivity ~ 8 1020c-1

Magnetic Reynolds number ~1015

Characteristic time of the magnetic field reconnection For the magnetorotational supernova is:

(approximately 10 times larger than characteristic time of magnetorotational supernova explosion). .

Reconnection of the magnetic field does not influence significantly on the supernova explosion.

MR supernova – different core massesBisnovatyi-Kogan, SM, 2008 (in preparation)

Dependence of the MR supernova explosion energy on the core mass and initial angular

momentum

Solid line – initial angular velocity = 3.53s-1

Dashed line - initial angular velocity = 2.52s-1

Recent results(in collaboration with K.Kotake,T.Takiwaki, K.Sato)

Implementation of modified (Shen et al., 1998) equation of state.Approximate treatment of electron captures and neutrino transport. (Kotake et.al.2003) . Neutrino leakage scheme.

Equation for electron fraction

Equation for lepton fraction

l eY Y Y

- neutrino luminosity

Neutrino pressure was taken into account.

B_0=10(9)G B_0=10(12)G

Recent results(in collaboration with K.Kotake,T.Takiwaki, K.Sato)

Recent results(in collaboration with K.Kotake,T.Takiwaki, K.Sato)

B_0=10(9)G B_0=10(12)G

MDRI No MDRI

Toroidal to poloidal magnetic energy relation

MDRI development

Mirror symmetry violation of the magnetic field in rotating stars

Bisnovatyi-Kogan, SM Sov. Astr. 1992, 36, 285

• а. Initial toroidal field

• b. Initial dipole field

• с. Generated toroidal field

• d. Resulting toroidal field

Mirror symmetry violation of the magnetic field in rotating stars

+ =

Resulting toroidal filed is larger in the upper hemisphere.

Violation of mirror symmetry of the magnetic field in magnetorotational explosion leads to: One sided ejections

along the rotational axis.

Rapidly moving radiopulsarss (up to 300 km/s).

In reality we have dipole + quadrupole + other multipoles…

(Lovelace et al. 1992)

Dipole ~

Quadrupole ~

3

1

r

4

1

r

At high magnetic fields neutrino cross-section depends on the magnetic field values.

The magnetorotational supernova explosion is

always asymmetrical.

The pulsar kick velocity can be up to 1000 km/s along rotational axis (Bisnovatyi-Kogan, Astron. & Astroph. Trans., 1993, 3, 287)

Jet, kick and axis of rotation are aligned in MR supernovae.

Evidence for alignment of the rotation and velocity vectors in pulsars

S. Johnston et al. MNRAS, 2005, 364, 1397

“We present strong observational evidence for a relationship between the direction of a pulsar's motion and its rotation axis. We show carefully calibrated polarization data for 25pulsars, 20 of which display linearly polarized emission from the pulse longitude at closest approach to the magnetic pole…we conclude that the velocity vector and the rotation axis are aligned at birth“.

Rapidly moving pulsar VLBI observations W.H.T. Vlemmings et al. MmSAI, 2005, 76, 531

“Determination of pulsar parallaxes and proper motions addresses fundamental astrophysical questions. We have recently finished a VLBI astrometry project to determine the proper motions and parallaxes of 27 pulsars, thereby doubling the total number of pulsar parallaxes. Here we summarize our astrometric technique and present the discovery of a pulsar moving in excess of 1000 kms, PSR B1508+55”.

Cassiopea A- supernova with jets-an example of the magnetorotational supernova

(Hwang et al. ApJL, 2004, 615, L117 )

1 million seconds Chandra servey of Cas A. Second jet was found.

ConclusionsConclusions

• Magnetorotational mechanism (MRM) produces enough energy for the core collapse supernova.

• The MRM is weakly sensitive to the neutrino cooling mechanism.

• MR supernova shape depends on the configuration of the magnetic field and is always asymmetrical.

• MRI is developed in MR supernova explosion.

• One sided jets and rapidly moving pulsars can appear due to MR supernovae.

Thank you!