How does the debris from a stellar tidal disruption join an - PowerPoint PPT Presentation

How does the debris from a stellar tidal disruption join an accretion flow? Roseanne M. Cheng 1 Hotaka Shiokawa 1 , Julian H. Krolik 1 , Tsvi Piran 2 , Scott C. Noble 3 1 Department of Physics and Astronomy, Johns Hopkins University, 2 Racah Institute of Physics, The Hebrew University of Jerusalem, 3 Department of Physics and Engineering Physics, University of Tulsa January 22, 2015 Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 1 / 12

Tidal disruption events: traditional model of t − 5 / 3 Geometry of disk: ∼ 2 R T � 1 / 3 � M BH R T = R ∗ M ∗ 2 R T ( M BH / M ∗ ) 1 / 3 1 a min = Accretion timescale τ acc ≪ τ 0 = P orb ( a min ) Gas is accreted after a few orbits because debris streams will focus at periastron and precess due to general relativity (shocks → O ( cR g / R T )). Rees (1988) Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 2 / 12

Classic theoretical expectations Rate of first return of debris to periastron (Phinney, 1989) � t � − 5 / 3 M return = dM d ǫ dt = 1 M ∗ ˙ d ǫ 3 τ 0 τ 0 IF quick entry into disk at R T , THEN M acc = ˙ ˙ M return ∝ t − 5 / 3 AND IF efficient radiation, THEN Mc 2 ∝ t − 5 / 3 L = ǫ ˙ BUT M BH � 10 7 M ⊙ L peak / L Edd ≫ 1 , (Ulmer, 1997) IF photosphere near ISCO, THEN T ∼ 10 5 K Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 3 / 12

Additional challenges to the t − 5 / 3 -paradigm Observational difficulties: → Radiation should be primarily in the UV (Rees, 1988) − − → Bolometric (Lodato & Rossi, 2011) and extinction corrections to lightcurve Inconsistencies with classical theory and observed candidates: (Cenko et al., 2012; Gezari et al., 2012; Chornock et al., 2014; Holoien et al., 2014; Arcavi et al., 2014; Vinko et al., 2015) Peak luminosities are lower than classical expectation: L obs ∼ 10 43 − 10 44 erg/s L peak ∼ 5 × 10 46 erg/s ← → Temperature lower than classical expectation: T obs ∼ 10 4 K T ∼ 10 5 K ← → How do we explain the TDE candidates with jets? (observed in hard X-ray) Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 4 / 12

Detailed calculations of the accretion process are necessary Current state of the theory → shocks at periastron are not efficient enough to circularize material − τ acc ≮ τ 0 (Kochanek, 1994; Guillochon et al., 2014) Mechanism by which tidal debris settles into accretion flow unknown Our approach: simulate encounter from disruption to formation of accretion disk use general relativistic hydrodynamics (GRHD) Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 5 / 12

Tidal disruption computed in local and global frame Simulate disruption of star by a Schwarzschild black hole and evolution of debris streams with general relativistic hydrodynamics G Local : (initial data) τ = x 0 relativistic (2PN) calculation in FNC frame λ 2 x 2 λ 0 = u following the star (Cheng & Evans, 2013) Global : P local data as initial conditions for simulation λ 1 in black hole frame with Harm3d full GRHD ( Shiokawa, Krolik, Cheng, Piran, Noble, 2014 ) x 1 Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 6 / 12

Initial conditions: White dwarf vs Intermediate mass black hole Dynamic lengthscales/timescales depend on mass ratio 400 relativistic Newtonian 200 Choose parameters for modest computational expense 0 − 200 M WD = 0 . 64 M ⊙ y [ R g ] − 400 = 500 M ⊙ M BH GM BH / c 2 − 600 R g = = R T = 107 R g R p − 800 a min = 495 R g − 1000 − 1200 − 800 − 600 − 400 − 200 0 200 400 600 800 x [ R g ] Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 7 / 12

Initial conditions: stellar debris in global frame Swing in apsidal angle due to relativistic effects during disruption Small PN effects (GR apsidal precession in stellar orbit and GR corrections to tidal stress) lead to strong shocks near orbital apocenters expected for main sequence star encounters (Cheng & Bogdanovi´ c,2014) Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 8 / 12

Results: shock formation (Shiokawa et al., 2014) Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 9 / 12

Results: Mass inflow rate (Shiokawa et al., 2014) Accretion rate simulated in Ballistic Return Rate Accretion Rate : α =0.1 black hole frame for τ < 12 Accretion Rate : α =0.01 6 10 Accretion rate extrapolated (analytic accretion theory) M [M ∗ /(GM BH /c 3 )] 7 10 from simulation for τ > 12 8 10 ˙ M peak is 10% of ˙ classical expectation 9 10 later, flatter peak τ peak ∼ 3 − 8 10 10 1 10 100 τ Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 10 / 12

Heating rate in accretion disk formation Heating rate calculated from shocks in simulation Scaled to main sequence star disruption by M BH = 10 6 M ⊙ Initially nozzle then apocenter shocks Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 11 / 12

Conclusions characteristic scale at which the tidal streams merge to form an accretion flow a min ≫ R T in addition to shock at nozzle, find existence of outer shocks → largely due to relativistic effects − accumulation of mass into accretion flow requires ∼ 5 τ 0 further time delay due to a larger disk than expected, which has a significantly longer inflow time ˙ M peak is 10% of classical expectation Expect significant departures from classical expectations for the lightcurve associated with tidal disruptions Cheng (JHU) Stellar Tidal Disruptions January 22, 2015 12 / 12