Star-Massive Black Holes Star-Massive Black Holes EncountersEncounters

J.-P.Luminet

Observatoire de

Paris (LUTH)Tidal Disruption Events

Esa, Madrid 2012

quasarsradiogalaxies

Supermassive black holes (M > 106 MS) in active galactic nuclei

Seyfert

2,7 106 MS < M < 3,5 106 MS)

Massive black holes (M > 106 MS) in quiescent galactic nuclei

Sagittarius A*Sagittarius A*

M15

4000 MS

G1/M31

20 000 MS

Intermediate mass black holes (103 < M < 106 MS)

in globular clusters

The « octopus » model for

AGN

Luminet (1986)

Accretion radius

BH’s gravitational radius

Characteristic distances

Collision radius

Ablation radius

Tidal radius

€

RT ≈ R∗

M•

M∗

⎛

⎝ ⎜

⎞

⎠ ⎟

1/ 3

Hills limit108 MO

Newtonian tidal field

Tidal field = relative acceleration between elements of matter

tidal potential

tidal tensor deviation tensor

• For incompressible homogeneous bodies • static field

(planet-satellite couple in circular orbit)Edouard Roche

Roche limit (1847)

€

RT = 2.45R*

M •

M*

⎛

⎝ ⎜

⎞

⎠ ⎟

1/ 3

• incompressible bodies

Luminet & Carter (1986)

in dynamical field

(asteroid-black hole encounter in parabolic orbit)

Kostic et al. (2009)

0.16 RT ≤ Rp ≤ RT Rp ≤ 0.16 RT

Cigar-like regime Disk-like regime

• compressible bodies

Luminet et al. 1982-1986 : « affine » model

in dynamical field

(star-black hole encounter in parabolic orbit)

Non-disruptive regime

€

RT ≈ 2.9R*

M •

M*

⎛

⎝ ⎜

⎞

⎠ ⎟

1/ 3

Oscillating Riemann ellipsoid => variable star ?

From equilibrium to stellar disruptionFrom equilibrium to stellar disruption

Compare : timescale of varying tidal field / internal timescale of the star

Far from the tidal radius :

Close to and inside the tidal radius :

Star adjusts its equilibium configuration

Strongly dynamical tidal field => Star cannot respond

Parabolic orbit

Tidal radius

Black hole

RTRP

Crucial parameter :

the penetration factor

Carter & Luminet (Nature, 1982)

Slight penetration ( 1) in the tidal radius

« cigar-like » configuration after leaving the tidal

radius

Disruption process in the ellipsoidal model (Luminet & Carter, 1986)

Slight penetration in the tidal radius Disruption process reproduced by

hydrodynamical simulations

« cigar-like » configuration after leaving the tidal radius

« S-like » configuration at the periastron

Ejected matter

Massive black hole

Tidal radius

Accretion of stellar debris (50%) Tidal FlaresLidskii & Ozernoy (1979), Rees (1988)

X-ray flare in RXJ 1242-11

(Komossa & Greiner, 1999)

UV flare in NGC 4252 (Renzini et al. 1995)

Giant X-UV luminous flaresGiant X-UV luminous flares

Detection (ROSAT, Chandra, Galex…) of Detection (ROSAT, Chandra, Galex…) of X-UV flaresX-UV flares from (non active) galactic nucleifrom (non active) galactic nuclei

X-UV-optical luminous flaresX-UV-optical luminous flares

Detection (Chandra, Galex, Pan-starrs…) of Detection (Chandra, Galex, Pan-starrs…) of flaresflares from from (non active) galactic nuclei(non active) galactic nuclei

« Relativistic » Tidal Flares« Relativistic » Tidal Flares

Detection (Swift) of Detection (Swift) of hard-X flareshard-X flares (L (Lxx ~10 ~104747 ergs) ergs)

(Zauderer et al. 2011, Cenko et al. 2012)(Zauderer et al. 2011, Cenko et al. 2012)

Relativistic jets from tidal ejecta ?

(Komossa (Komossa et alet al., Saxton ., Saxton et alet al. , Esquej . , Esquej et al.et al., Gezari , Gezari et alet al., etc.)., etc.)

The Pancake EffectThe Pancake Effect(Carter & Luminet, (Carter & Luminet, NatureNature, 1982), 1982)

For deep penetration ( > 3)

Fixed compressive principal direction of the tidal tensor in the « vertical » direction ==>

The « Rolling Mill » EffectThe « Rolling Mill » Effect

Cartoon version

fixed compressive point after periastron

Deep penetration ( 3) in the tidal radius

Disruption process in the ellipsoidal model (Luminet & Carter, 1986)

Explosive stellar disruption ? Explosive stellar disruption ?

Compression and heating strongly dependent from the Compression and heating strongly dependent from the penetration factorpenetration factor

Maximum values for an ideal gas with polytropic index 5/3 :Maximum values for an ideal gas with polytropic index 5/3 :

Conditions required for Conditions required for explosive thermonuclear explosive thermonuclear reactionsreactions

Detonation of Detonation of metalsmetals (C, N, O, …) by radiative (C, N, O, …) by radiative proton proton captures captures (Luminet & Pichon 1989(Luminet & Pichon 1989))

HeliumHelium detonation by detonation by triple alpha triple alpha (Pichon 1985(Pichon 1985))

Tidally induced supernovae? Tidally induced supernovae?

Luminet, 1986

Rosswog et al., 2008

A special type of GRB ? A special type of GRB ?

Carter, 1992

Lu et al. 2008; Gao et al. 2011

log(M•/M*)

log

(

The pancake triangle (solar type stars)

No disruption

Black hole enters the sta

r

Star enters the Black hole

107 K

108 K

109 K

Shock waves in stellar pancakesShock waves in stellar pancakes(Brassart & Luminet, 2008-2010)(Brassart & Luminet, 2008-2010)

ProblematicsProblematics

• Stellar pancakes calculated in the affine model (Luminet et al. 1983- 1989)

==> Elliptical deformations of the star

• Stellar pancakes calculated by 3D hydro

(Bicknell & Gingold 1983; Laguna et al. 1993; Fulbright 1996)

==> SPH method (uses artificial viscosity to simulate shock waves)

N.B. : Generalized affine model (Ivanov et al. 2001-2003)

• Disagreement : compression of the stellar core less than in the affine model:

max ~ 1.5 for < 10

max ~ const for > 10

• Due to shock waves occuring during the phase of vertical direction ??

• Or bad treatment (SPH) of shock waves ?

New Hydrodynamical Model Euler eqs. + Godunov methods

Principal axes of the tidal tensor

1D approximation

valid from the tidal radius to periastron

= 7

velocity Tidal radius

periastronmax.compression

Phase1:

collapse

= 7

velocity

max.compression

Phase 2:

bounce Subsonic collapse

Supersonic collapse

Subsonic expansion

= 7

densitycollapse

Shock

(after bounce)

Tidal radius

periastronmax.compressionand bounce

= 7temperature

shock

= 15

velocity

shock 1

(before bounce)

shock 2

(after bounce)

= 15

density

shock 1

(before bounce)

shock 2

(after bounce)

maximum central density

maximum central temperature

= 10

Shock-driven maximum temperature

Number of disruption events:

UV/X/gamma-ray flares

€

N ≈10−5 β−1 / year /galaxy

€

N(z < 0.8) ≈104 β−1 / year

Wang & Merritt (2004), Tal (2005) BH binaries : Chen et al (2009)

Short GRB

Long GRB

Modelisation of Gamma-Ray Modelisation of Gamma-Ray BurstsBursts

jet

jet

• -ray transients Swift J1644+57 and Swift

J2058+05 (Zauderer 2011, Cenko 2012)

interpreted as a relativistic outflow from transient accretion onto a 106 Ms black hole

• • GRB 060614GRB 060614 (long (long ~100 s) without SN remnant : ~100 s) without SN remnant : disruption of a WD by an intermediate mass BH ? disruption of a WD by an intermediate mass BH ?

(Lu et al. 2008)(Lu et al. 2008)

• A new class of -ray bursts from stellar disruptions by IMBH ? (Gao et al., 2010) From a total of 328 Swift GRBs with accurate measured durations and without SN association, 25 GRBs satisfying the criteria for GRB060614-type bursts…

Relativistic tidal field

E = specific total energy

L = specific angular momentum

for parabolic orbits :

E = 1

Luminet & Marck, 1985

Precession effect : self-crossing orbitPrecession effect : self-crossing orbit

Double point Double point insideinside the tidal radius: the tidal radius:

Luminet & Marck, 1985

Double point Double point insideinside the tidal radius the tidal radius

Several compressionsSeveral compressions

Brassart & Luminet, 2011

Multi-pancake effect and X/ bursts

• • Multi-peak structure and timescales compatible with Multi-peak structure and timescales compatible with GRB 970815GRB 970815 ( (and and others ?)others ?)

• • White dwarfs / IM BH strong encounters White dwarfs / IM BH strong encounters ((Rosswog et al. 2010, Hass et al. 2012Rosswog et al. 2010, Hass et al. 2012):):

Confirm nuclear ignition (e.g. helium Confirm nuclear ignition (e.g. helium detonation)detonation)

• 3D simulations3D simulations coupling hydrodynamics and nuclear network (Guillochon et al., 2011):

Confirm the occurrence of tidally induced supernovae

Recent hydro models …

• Red giants / BH interactions (Mac Leod et al., 2012):

Tidal stripping of atmosphere

Analogy with stellar collisions

Around a 109 MS black hole, the typical collisional velocities are > 5000 km/s within a distance 0.1 pc from the black hole

= vcoll/v* crushing factor

= RT/Rp

Massive BH (104 - 108 MS)

Supermassive BH > 108 MS

Glob. Clusters, GN, Seyfert Quasars, Giant elliptical …

High velocity head-on collisions, polytropes 5/3

Barnes, 2003Makino, 1999

: disruption / : pancake effect

The EndThe End