Integrated Impact of Hydrodynamic Processes in Massive Stars

1
Integrated Impact of Hydrodynamic Processes in
Massive Stars

• Stellar Hydro Days
• July 26th, 2006
• Patrick Young (LANL/Steward)
• Casey Meakin, David Arnett (Steward)
• LA-UR-05-3961,4652

2
A Sampling of Unresolved Questions

• Evolution and contribution to environment
• f(initial mass)
• f(metallicity)
• Yields from an initial mass function (IMF)
• luminosity
• kinetic energy
• nucleosynthesis
• Progenitors of WDs, LBVs, SNeII, SNeIb/c, GRBs,
  etc.
• How to identify the progenitor of a particular
• object
• Asymmetries in supernova progenitors ?
  asymmetries in explosion, mixing of nuclear
  species

3
A Sampling of Unresolved Questions

• Evolution and contribution to environment
• f(initial mass)
• f(metallicity)
• Yields from an initial mass function (IMF)
• luminosity
• kinetic energy
• nucleosynthesis
• Progenitors of WDs, LBVs, SNeII, SNeIb/c, GRBs,
  etc.
• How to identify the progenitor of a particular
• object
• Asymmetries in supernova progenitors ?
  asymmetries in explosion, mixing of nuclear
  species

4
A Sampling of Unresolved Questions

• Evolution and contribution to environment
• f(initial mass)
• f(metallicity)
• Yields from an initial mass function (IMF)
• luminosity
• kinetic energy
• nucleosynthesis
• Progenitors of WDs, LBVs, SNeII, SNeIb/c, GRBs,
  etc.
• How to identify the progenitor of a particular
• object
• Asymmetries in supernova progenitors ?
  asymmetries in explosion, mixing of nuclear
  species

5
A Sampling of Unresolved Questions

• Progenitors of WDs, SNeII, SNeIb/c, GRBs, etc.
• f(initial mass)
• f(mass loss)
• f(metallicity)
• How to identify the progenitor of a particular
• object
• Yields from an initial mass function (IMF)
• luminosity
• kinetic energy
• nucleosynthesis
• Asymmetries in supernova progenitors ?
  asymmetries in explosion, mixing of nuclear
  species

6
Structure, Physical Processes, Evolution

• Some stellar characteristics which influence
  evolution and final fate
• Mass, core size
• Density profile, entropy gradients
• Composition, neutron excess
• fluid velocities, angular momentum, asymmetries
• energy transport
• Determined primarily by
• Mass Loss
• Nuclear neutrino physics, opacities EOS
• Rotation
• Convection
• Physics of convective boundaries
• Hydrodynamics of stable regions

7
Structure, Physical Processes, Evolution

• Previous slide somewhat misleading
• We tend to separate out physical processes
• Physics of stars, especially hydro, strongly
  coupled
• For example, convection, waves, rotation,
  radiation transport are a single problem
• Stars are not amenable to direct simulation
• multi-D gives snapshots
• analytics treat problems in isolation
• High energy density experiments dont do stable
  hydro
• Combinations of these can allow us to develop
  analytic frameworks that account for the coupling
  seen in more physically complete situations of
  simulation and experiment

8
Stable does not mean static!

• Standard stellar models treat stars as series of
  static states
• Mixing length-like theories evaluate
  thermodynamic criteria to determine extent of
  convection
• Thermodynamic criteria appropriate for predicting
  onset of convection in stratified fluid, but not
  extent of convection once fluid is in motion
• Stable layer is completely static

9
Stable does not mean static!

• Multidimensional simulations of stellar
  convection
• Main sequence 23 M? core convection
  (Meakin/PROMPI)

10
The Convective Boundary

• Three main hydrodynamic regimes in a stratified
  medium
• Region 1 fully convective Brünt-Väisälä
  frequency N2 lt 0
• Unstable - displacements lead to acceleration
• Driven by entropy generation (nuclear burning) or
  entropy loss (surface convection)
• Motion dominated by accelerating plumes

Vorticity
XH
Velocity
11
The Convective Boundary

• Boundary characterized by bulk Richardson number
  Ri b / (?u/?r)2 Ratio of potential energy
  across a layer to energy in shear (?u rms
  turbulent velocity, shear from waves depends on
  context, ?r extent of boundary, b ?N2dr)
• Ri 0.25
• Boundary region. Impact of plumes deposits
  energy through Lagrangian displacement of
  overlying fluid. Internal waves propagate from
  impacts
• Conversion of convective motion to wave motion.
  Shear instabilities, nonlinear waves mix
  efficiently.
• Wave amplitudes ? M2 N2 ? large buoyancy jump
  gives large wave amplitudes even at small Mach
  number

Vorticity
XH
Velocity
12
The Convective Boundary

• Boundary characterized by bulk Richardson number
  Ri b / (?u/?r)2 Ratio of potential energy
  across a layer to energy in shear (?u rms
  turbulent velocity, ?r extent of boundary, b
  ?N2dr)
• Ri 0.25
• Two issues - extent of hydrodynamically unstable
  region larger that thermodynamically unstable
  region
• Entrainment at marginally stable boundary

Vorticity
XH
Velocity
13
The Convective Boundary

• Radiative regions
• Internal waves propagate throughout radiative
  region
• evaluate stability with gradient Ri, buildup of
  waves in cavity or coupling with rotation can
  cause instability
• Radiative damping of waves generates vorticity
  (Kelvins theorem)
• Slow compositional mixing
• Energy transport changes gradients generates an
  effective opacity
• angular momentum transport

Baroclinic generation term
Vorticity
14
Implications for Evolution

• Predictive
• Physics can be incorporated into evolution codes
• Higher radiation pressure -gt less restoring force
• Small effect for small stars
• More important with increasing stellar mass
• Larger convective core masses
• Longer lifetimes
• Higher luminosity, larger radii
• Larger C/O cores at collapse (gt50 larger for 25
  M?)
• Different composition entropy gradients,
  different neutron excess, different yields
• Mixing of processed material through radiative
  regions, deeper dredge-up

15
Observational Tests

• Convective core size
• Indirectly, luminosities, radii of eclipsing
  binaries
• Directly from apsidal motion of binaries
• Standard model cores are systematically
  undersized
• Models with hydrodynamics fit well at all masses

16
Eclipsing Binaries
17
Shell Burning

• See Caseys talk
• Important evolutionary effects
• very different extent of late stage shells
• mixing between shells
• Urca
• entrainment of fuel
• Large perturbations of thermodynamic quantities
• Spatially correlated perturbations

18
Extent of Convection

• Local mixing length convection ignores effect of
  KE on marginally stable layers
• Initial transient input of KE increases extent of
  mixed region 30 for mixing length initial
  models
• New steady state after O abundance readjusted

19
Uncertainties in Nucleosynthesis

• Structure of progenitor
• Asymmetries in progenitor
• Explosion mechanism
• Method of calculating explosion
• Method of calculating nucleosynthesis in
  explosion
• Asymmetries in explosion

20
Structure at Core Collapse

• Comparison of TYCHO models with and without hydro
  mixing
• Models with hydro mixing
• Smoother entropy gradients
• Larger O cores, thus higher ?
• Different abundance profiles
• Multi-D effects
• Coherent perturbations of on large spatial
  scales-gt rippled interfaces, global asymmetries
• Merging of shells
• Wave effects on energy neutrino transport,
  opacities

21
Structure at Core Collapse

• Changed density structure changes collapse
• Large mass accretion rate onto compact remnant
  for much longer time
• Delayed explosion relative to standard stellar
  model
• Weaker explosion
• More fallback
• Relatively low black hole minimum mass
• Stars above 20M? at z? either become weak SNe
  or GRBs (maybe both)

22
Effect on Explosion Changes in 1D Progenitors
Yields from 23 M? with Grevesse Sauval 98 solar
metallicity Core collapse models from TYCHO with
and without hydro mixing 1D collapse from Fryer
1999 Model Explosion Energy Remnant Mass
Ni Mass ?/Fe
Standard 1.65 foe
1.57 M? 0.42 M?
5.74 Hydro weak exp. 0.57 foe
6.01 M? lt0.1 M?
large Hydro strong exp. 3.0 foe
1.64 M? 0.99 M? 6.05
23
Effects of Explosion Asymmetries

• Imposing a small asymmetry during 3D calculation
  of explosion changes mixing
• Yields for otherwise consistent models
• energy (foe) 44Ti (M?) 56Ni
• 23 M? 2.3 1.2x10-5 2.6x10-4
• 23 M? Asymm. 2.3 1.8x10-4 0.019

24
Conclusions

• Stable does not mean static! Hydrodynamics in the
  convective boundary and radiative regions have
  substantial effects on evolution
• ?radiation pressure, ?restoring force, ? effects
  increasingly important at larger masses
• Hydro mixing can be approximated in 1D
  evolutionary models in a predictive way and in
  agreement with observations
• Late burning stages cannot be captured in 1D
• burning shells interact
• perturbations in T,? of perhaps 10 correlated on
  large angular scales near collapse
• rippled composition / thermodynamic boundaries
• different abundance patterns, effects on neutrino
  physics

25
Conclusions

• Core sizes, explosion energies remnant masses
  are substantially different
• Yields can change by orders of magnitude due to
• changes in progenitors
• asymmetries in progenitors
• (not to mention various aspects of the explosion
  calculation itself)
• High resolution 3D simulations and sophisticated
  analytical work are both necessary - physics must
  be generalized to apply to a wide range of
  conditions

26
Next Steps

• Angular momentum transport
• Geneva group has demonstrated importance of
  rotation to stellar evolution
• Analytic - not supplemented with multi-D
  simulations, treated in isolation
• Rotation must interact with g-modes convective
  boundary instabilities - more efficient transport
  of angular momentum
• Talon Charbonnel show that g-modes can induce
  solid body rotation in sun, where wave flux is
  small
• Simulations in progress
• MHD
• You can also get solid-body rotation in sun from
  B fields
• Requires sophisticated MHD
• Simulations in stellar context essentially
  limited to solar convection zone

27
Next Steps

• Nuclear burning - multi-D with larger networks
  post-processing
• Radiation dominated environments - eruptions of
  Luminous Blue Variables and Supernova Impostors
• Inefficient convection on the early pre-Main
  Sequence and late post-Main Sequence
• Binary Evolution
• Connecting star formation to stars

Sandquist Taam
28
O C Burning

• ?conv ?nuc ?thermal
• No direct observational constraints
• Initial 23 M? models from TYCHO with and without
  wave physics included
• 25 element nuclear network for O C burning
• 2D 3D wedges encompassing O or OC shells prior
  to Si ignition

29
Effects of Burning Calculation

• Network size (obviously) important. Any
  calculation requires detailed post-processing
• Duration of burning important. Network and
  freezeout calculations must continue for 10s of
  seconds
• Location of ?-rich freezeout uncertain even at
  high Ye (gt0.4985) final dominant abundances (?,
  56Ni, neutron rich Fe peak) very sensitive to
  initial entropy, composition, thermodynamic
  trajectory

30
Effects of the Explosion Calculation

• Very different neutrino luminosities can produce
  same final kinetic energy
• Yields for otherwise consistent models
• energy (foe) 44Ti (M?) 56Ni
• 23 M? 2.3 1.8x10-4 0.019
• 2.4 3.9x10-4 0.7

31
Effects of the Explosion Energy

• Explosion energy in a simulation is arbitrary
  unless constrained by an observed supernova
• Mechanism dependent changes neutrino energies,
  explosion energies, success of explosion.
• ? vs. ?convection, different equations of state,
  jet-driven models, etc.
• Yields for otherwise consistent models
• energy (foe) 44Ti (M?) 56Ni
• 23 M? binary 1.1 1.2x10-5 2.6x10-4
• 2.0 5.7x10-5 0.055

32
Simulations vs. Constraints
White satisfies constraints, red inconsistent
with constraints, yellow marginal
33
Observational Constraints

• High velocity N-rich, H-poor knots
• Ejecta mass
• Compact remnant mass
• 44Ti and 56Ni
• MTi 1.0 x 10-4 M? from x-rays
• MNi 0.05-0.2 M? from brightness of supernova
• Trends identifiable Ti decreases, Ni increases
  with explosion energy
• BUT yields are very model dependent - multi-D
  effects cutoff time for calculation of burning
  freezeout network size neutron excess, entropy,
  temperature, density evolution can change
  yields by orders of magnitude

34
Effect on Explosion 3D Explosion Calculations

• Initial results from a study of the progenitor of
  Cassiopeia A
• Young (325yr), nearby (3.4 kpc)
• Estimates range from 16 to 60 M? single stars and
  binary scenarios
• Several independent observational constraints
• 3D neutrino-driven explosion calculations a
  range of advanced progenitor models
• What parameter space for the progenitor is
  allowed by each constraint?

35
Structural Perturbations

• Intershell interaction
• Large wave flux from O shell imposes large
  displacements on C shell
• C shell is rippled on large spatial scales
• KE flux of waves may overcome stability of
  intershell region??
• O and C shells may merge??

36
Structural Perturbations

• Angular variation of XO, ?nuc
• Entrained material drawn as streamers into
  burning region
• Composition variations of 10 in 3D
• ?nuc varies by factor of a few in local flashes
  when fresh fuel ingested
• Rapid burning of ingested material may produce
  explosive rather than hydrostatic abundance
  patterns

37
Structural Perturbations

• Significant wave flux outside convective region
• ?T, ?? 0.1-1 for 3D models
• Perturbations reflect Lagrangian displacement of
  material by wave motion
• Perturbations can be correlated instead of random
• Low order modes can impose global asymmetries

38
Extent of Convection

• Comparison with TYCHO models with and without
  hydro mixing
• Predicted outer edge of mixed region in hydro
  mixing model more similar
• Velocity structure more characteristic of waves
  than plumes in additional extent of mixed region

39
White Dwarf Initial-Final Mass Relation

• Procyon Sirius
• Mass radius of binary members known to lt1, L lt
  4
• Cooling ages (Fontaine et al. models)
• Sirius B mass of 5.04 0.29 M? from TYCHO, WD
  1.00 M?
• Procyon B mass of 1.91 0.23 M? from TYCHO, 0.60
  M?
• Consistent with Padova group estimate with well
  calibrated parameterized overshoot
• Most precise determination yet of initial-final
  mass relation for massive white dwarfs (Liebert
  et al. 2005)

Sirius
Procyon
40
Contextual Stellar Evolution

• Make stellar evolution predictive for all stages,
  masses, and compositions
• In order to understand...
• Nucleosynthetic, KE, luminosity yields
• Populations (galaxy evolution)
• Starbursts (first galaxies, interactions)
• Cluster debris disk ages
• Through a strategy of...
• Using hydro simulations, laboratory astrophysics
• Avoiding calibration of parameters
• Stringent tests against observations

My god, it's full of stars!
41
The Nucleosynthetic Yields of Stellar Populations

• Sources
• Supernovae II,Ib/c,Ia LBVs Wolf-Rayets AGB
  stars planetary nebulae novae
• The challenge accurately connecting microphysics
  of nucleosynthesis to macrophysics of stellar
  populations
• Underestimating progenitor mass for a given yield
  by 10 overestimates the number of stars
  contributing that yield by 25 for Salpeter IMF
• Similar problems for kinetic energy yields and
  photon fluxes