1. Mitch Begelman & Eric Coughlin JILA, University of Colorado ARE RELATIVISTIC JETS ALWAYS MAGNETIC?

2. Why we assume relativistic jets are propelled by large-scale B-fields: Only way to tap BH spin (?) –OK, good point Rel. electrons cool too rapidly so thermal pressure won’t work –Could use ion pressure if coupling to electrons weak Radiation drag due to aberration limits acceleration by radiation pressure –Only applies at low optical depth (cf. fireball models of GRBs) Need super-Eddington flux for radiative acceleration

3. The Magnetic Flux/Spin Paradigm - strict requirements here as well Magnetic flux threading engine Angular velocity of engine Jet power limited by amount of flux available

4. Tidal Disruption Event A star ventures inside the tidal radius of a black hole and is torn apart

5. Tidal forces …... unbind ~half the debris … throw the other half into highly eccentric orbits Semi-major axis:

6. Energy pumped into the stellar debris by tides … Simulations by Guillochon & Ramirez-Ruiz 2013

10. Common wisdom says that matter falling back in excess of Ṁ E should be blown away: R> (Ṁ/Ṁ)R g : thin Keplerian disk R> (Ṁ/Ṁ)R g : regulates L~L E Shakura & Sunyaev 73 … but this may not always happen

11. Super-Eddington TDE Swift J1644+57 Tchekhovskoy et al. 2014 Swift + Chandra light curves L corrected for beaming Radio “re-brightening” after ~ 4 months l Swift J2058+05 a second case? Cenko et al. 2012 Bloom et al. 2011

12. Swift J1644+57 outburst suggestive of a beamed, relativistic flow = jet

13. Do TDEs have enough magnetic flux? Transient accretion events have access to a fixed amount of flux… Tidal Disruption Event candidate Swift J1644+57: Jet power: L j > 10 45 erg s -1 ~ 100 L E Flux needed:  > 10 30 G-cm 2 Flux available:   ~ 10 25 B 3 (R  /R  ) 2 G-cm 2 Collapsar Gamma-Ray Burst: Jet power: L j > 10 50 erg s -1 ~ 10 11 L E Flux needed:  > 10 28 G-cm 2 Flux available:   ~ 10 25 B 3 (R  /R  ) 2 G-cm 2

14. An alternate approach

15. g eff MRI Powered by dissipation of turbulent B “Empty” funnel

16. g eff MRI Powered by dissipation of turbulent B “Empty” funnel

17. g eff Reconnection MRI Reconnection converts energy to radiation

18. g eff Reconnection MRI Entrainment (by rad’n force) Mass-loading, collimation and acceleration

19. g eff Reconnection MRI Entrainment (by rad’n force) Self-shielding (from drag) Self-shielding from radiation drag

20. Max.  of a radiation-propelled jet: Jet power L j = l L E “Terminal” Lorentz factor   = L j /Ṁ j c 2 –based on available energy Increase   by decreasing Ṁ j c 2 –but if Ṁ too small photons leak out before   is reached (for conical flow; 2/7 instead of 1/4 for paraboloidal) Rees & Meszaros 2005

21. Radiative self-shielding: STATIONARY L j,  j pressure p Drag important if

22. Radiative self-shielding: STATIONARY L j,  j pressure p Drag important if Boundary layer dragged by jet radiation, retarded by radiation from wall

23. Radiative self-shielding: STATIONARY L j,  j pressure p Drag important if Boundary layer dragged by jet radiation, retarded by radiation from wall  BL  1 for rays impinging on boundary layer 

24. Radiative self-shielding: STATIONARY L j,  j pressure p Drag important if BL can be dominated by kinetic energy:

25. Radiative self-shielding: STATIONARY L j,  j pressure p Drag important if Ratio of BL to jet energy:

26. Radiation-driven jet is pressure-confined Spherical envelope with pressure p a ~r -  –  > 2  jet blows up envelope –  > 2  envelope crushes jet Need evacuated funnel held open by rotation … but not too wide a funnel (otherwise radiation can drive circulation or slow wind) WHERE MIGHT WE FIND SUCH FUNNELS?

27. Slim Disk Models of Hyperaccretion Radial pressure force significant Angular momentum below Keplerian H/r ~ few tenths Vertical and radial structure coupled –Can be modeled in 1D –2D models more reliable Only possible if l / l Kep large enough

28. l/l Kep disk opening angle 0.74 0.88 A clue from self-similar slim disk models Gyrentropes: s ( l ) Quasi-Keplerian Disk closes up at l close to l Kep What is going on?

29. Case with mass loss … Assume  scaling for radial transport: Add radial pressure balance (ADIOS scaling)

30. Dynamical conditions don’t allow a bound disk-like flow Flow “closes up” to axis as B  0  Flow becomes “star-like” (with a rotational funnel) Less disk “surface” to lose energy via wind Flow reduces B instead by steepening density/pressure profiles Answer: l is too small to set up a flow with

31. B=0 Less l  steeper   higher accretion L Flow blows up or finds way to vent excess energy  equilibrium with B~0 (B<< GM/R)

32. Summary: Some low- l accretion flows unavoidably produce hyper-Eddington luminosities –TDEs, GRBs, maybe quasi-stars Magnetic flux available might be too small to drive electromagnetic jets with adequate power Radiation pressure is an alternative to driving relativistic jets under these conditions – Can drive the fastest jets:  max ~(L/L E ) 1/4 –Self-shield from drag: boundary layer can carry substantial energy flux –BLs slower (~  j 1/2 ) but wider beaming angle