What Are Type Ia Supernovae?

1
What Are Type Ia Supernovae?

• Jens C. Niemeyer
• Max-Planck-Institut für Astrophysik
• Based on collaborations with
• W. Hillebrandt (MPA Garching)
• S.E. Woosley (UC Santa Cruz)
• M. Reinecke (MPA Garching)
• B. Leibundgut (ESO Garching)
• and others

2
SN 1994D
3
Type Ia SupernovaeFacts ...

• General properties of SNe Ia
• Very homogeneous class of events, only small (and
  correlated) variations.
• Rise time 15 - 20 days
• Decay time many months
• No hydrogen is seen in the spectra !
• Early spectra Si, Ca, Mg, ...(abs.)
• Late spectra Fe, Ni, (emiss.)
• Very high velocities (10000 km/s)
• SN Ia are found in all types of galaxies,
  including ellipticals
• Progenitor systems must have long lifetimes

Courtesy of the Supernova Cosmology Project
4
and Theory

• Standard model (Hoyle Fowler 1960)
• SNe Ia are thermonuclear explosions of CO white
  dwarf stars.
• Evolution to criticality
• Accretion from a binary companion leads to growth
  of the WD to the critical Chandrasekhar mass
• ( 1.4 solar masses).
• After 1000 years of slow thermonuclear
  cooking, a violent explosion is triggered at or
  near the center
• complete incineration within less than
  two seconds, no compact remnant!

5
Deflagrations and Detonations

• Deflagrations (Flames)
• Subsonic burning fronts, propagating by heat
  conduction. Laminar flame speed and flame width
  (Timmes Woosley 1992)
• SL 0.001 c , ? 1 ?m.1 cm

Detonations Supersonic burning fronts,
propagating by shock heating. Detonation width
and speed SD us 0.1 c , ?D 100 ?
-In principle, both modes of propagation are
allowed in the supernova. Details of the
ignition process decide which mode is realized.
-Both modes are hydrodynamically unstable in
multiple dimensions!
6
Anatomy of an Explosion IPrompt Detonation

• Ignition

• Prompt Detonation (Arnett 1969 Hansen Wheeler
  1969)
• Supersonic propagation doesnt allow the star to
  expand prior to being burned. Almost no
  production of intermediate mass elements.

Ruled out by observations!
7
Anatomy of an Explosion IIPure Turbulent
Deflagration

• Ignition

• Deflagration Phase (many classic references)
• Burning propagates as a subsonic flame. The
  Rayleigh-Taylor instability (buoyancy!) produces
  rising bubbles. Shear flows at the bubble walls
  produce turbulence (Kelvin-Helmholtz
  instability).
• Turbulent combustion
• Turbulence increases the flame surface and hence
  the speed. Under certain conditions, the laminar
  speed becomes irrelevant (J.N. Hillebrandt
  1995)
• The turbulent flame speed is equal to the speed
  of the fastest turbulent eddies ! (first observed
  by Damköhler 1940)

8
Anatomy of an Explosion IIIDelayed Detonation
(Khokhlov 1991 Woosley Weaver 1994)
Deflagration-Detonation-Transition (DDT)
(Zeldovich et al. 1970 Khokhlov 1991 Woosley
Weaver 1994) DDT may be possible as a result of
local flame quenching and fast turbulent mixing
(Khokhlov 1997, J.N. Woosley 1997). Advantages
for 1D SN Ia models (papers by Nomoto, Höflich,
Thielemann,) Detonation sweeps up unburned
CO, gives additional kick. Transition density
is convenient tuning parameter. Problems Doesnt
work if flames cant be quenched (J.N. 1999)
. May not be needed (new 3D results).

• Deflagration Phase

9
Bottom Lines IFlames and Detonations

• Understanding turbulent combustion is crucial
  for understanding SN Ia explosions
• Most important
• Effective turbulent burning speed (may be
  independent of microphysics!)
• Robustness of flames with regard to turbulent
  quenching (DDT!)
• Delayed Detonations
• Allow good fits of 1D simulations to
  observations. Get rid of unburned material.
  Provide convenient fitting parameter for SN Ia
  family (transition density). BUT
• Physics of DDT indicates very low probability.
• May not be needed to explain explosion strength.

10
ZoologyCurrently Discussed Explosion Models
Type Ia Supernova
Explosion Models
Merging
Chandrasekhar
Sub-Chandrasekhar
White Dwarfs
Mass Models
Mass Models
Prompt
Initial
Detonation
Deflagration
Pure (Fast)
Slow Turbulent
Turbulent Deflagration
Deflagration DDT
Delayed Detonation
Pulsational Detonation
11
Multidimensional Simulations of SN IaWhere Are
We?

• Warnings
• Turbulence is a key element of all Chandrasekhar
  mass models.
• No 2D (not to mention 3D) simulation to date
  reaches the fully turbulent regime!
• This is what we really simulate

12
Large Eddy Simulations of Exploding White Dwarfs

• 2D simulation of an exploding white dwarf
  (Reinecke, Hillebrandt J.N. 1998)
• Uses a flame capturing/tracking scheme based on a
  level set method The turbulent flame speed is
  given by
• ST (2 q)1/2
• 
  
• where q is the turbulent subgrid energy obtained
  from a subgrid-scale model for the unresolved
  turbulence (J.N. Hillebrandt 1995). Large
  scales solved by higher-order Godunov scheme
  (PPM) (Colella Woodward 1984).

768x768 simulation
Movie made by M. Reinecke
13
Everything is in the details
512x512
1024x1024
256x256
Reinecke 2001 Global energy release is almost
independent of resolution by virtue of the
subgrid-model.
14
(No Transcript)
15
3D LES of Turbulent Deflagration A Healthy
Explosion?
3D, resolution study
2D vs. 3D (central ignition)
explosion
explosion
re-collapse
re-collapse
Reinecke 2001 - Systematically higher energy
release in 3D (consistent with Khokhlov 2001). -
Weaker dependence on the initial conditions than
in 2D.
16
Problems of the Chandrasekhar Mass Paradigm

• What are the progenitors?
• Any single scenario has difficulties explaining
  SN Ia occurrence in oldest and youngest host
  populations simultaneously (Howell 2001).
• Where is the hydrogen?
• If binary companion is H donor there should be
  some trace of H in the spectra. None has been
  found so far (e.g. Cumming et al. 1996).
• Where is the low-velocity C and O (or Si, Ca)?
• In multi-D deflagration models some unburned
  material always remains near the center. Can a
  delayed detonation get rid of all of it (Khokhlov
  2001)? Or maybe fully developed turbulence?
• Correlation of SN Ia subtype with host population
• Subluminous events only occur in old pop.s
  (Howell 2001), overluminous ones in young stellar
  systems (Hamuy et al. 1996). Why?
• Nickel masses
• SN 1991bg-like objects produce only 0.1 solar
  masses of Ni. This amount doesnt even unbind a
  Chandrasekhar mass WD

17
ZoologyCurrently Discussed Explosion Models
Type Ia Supernova
collapse to neutron star?
high-velocity nickel?
Explosion Models
Merging
Chandrasekhar
Sub-Chandrasekhar
White Dwarfs
Mass Models
Mass Models
Prompt
Initial
Detonation
Deflagration
Pure (Fast)
Slow Turbulent
Turbulent Deflagration
Deflagration DDT
Delayed Detonation
Pulsational Detonation
18
Usually Discussed Systematic Effects

1. Supernova evolution
2. Sample evolution
3. Grey dust
4. Lensing effects

need to know what they are!
19
Supernova Evolution
High Redshift
Low Redshift
20
Sample Evolution
high-mass progenitors
low-mass progenitors
High Redshift
Low Redshift
21
SummaryWhat Can We Expect?

• If all SNe Ia are Chandrasekhar mass events
• Little SN evolution.
• No sample evolution.
• If they are either all sub-Chandras or all
  mergers
• Possibly substantial SN evolution.
• No sample evolution.
• If they are a mix of some or all of the above
• Sample and SN evolution.