AGB Star Yields for Everyone
John Lattanzio1
and
Amanda Karakas2
1: Centre for Stellar and Planetary Astrophysics,
Monash University,Australia
2: Origins Institute, Department of Physics &
Astronomy, McMaster University , Canada
Structure of Talk
•
•
Reminder of AGB stars and vital role
Outline of what we want
•
•
•
•
State of the art models
Synthetic AGB models
•
•
•
What we can do
What we cannot yet do
Strengths and weaknesses
Best yields available now (and soon)
Distribution
Public Speaking for Scientists
Time
Trivia Question 1
•
Who said:
“To the extent that its possible, it’s the
isotopes that keep the theorists honest.”
•
Dave Arnett (U Arizona)
Not this one…
S-process elements
Massive stars produce most of the galactic magnesium,
which is primarily 24Mg at low Z
But 3 - 6 Msun AGB stars can produce large amounts of
the heavy magnesium isotopes
(Y. Fenner, A. Karakas, B. Gibson, J. Lattanzio)
AGB stars are needed to recover the observed
at low metallicity
25,26M/24Mg
Limongi et al. (2002) calculations generate more
Woosley & Weaver (1995)
25,26Mg
(Y. Fenner, A. Karakas, B. Gibson, J. Lattanzio, PASA, 2003)
ratios
than
GCE of
19F
SN, WR
and AGB
– Renda et al (2003)
SN & WR
SN only
x = Milky Way
= LMC
= w Cen
Sodium
12C
23Na
14N
22Ne
Note: some
23Na
22Ne
23Na
23Na
is primary and some is secondary!
AGB star nucleosynthesis
•
•
•
•
Note the interplay of H and He burning…
Note the dependence on mixing regions…
Note that the return depends on mass-loss
Note that these are not the most precise
inputs to the models…
Trivia Question 2
•
What is the original definition of a SHIP?
•
A three masted sailing vessel with
square-rigging on all three masts
Barque
Brig
What do we want?
•
•
Quantitative predictions of surface
abundances over time for all M and
compositions
Is this too much to ask???
Something Like This – but as f(time)
Something Like This – but as f(time)
Trivia Question 3
•
Which famous (Italian) stellar astronomer
said:
“We provide evolutionary tracks, yields and
physics from the pre-main sequence
through to the thermally pulsing AGB. Its
certainly wrong, but its freely available.”
•
Alessandro Chieffi
So why can’t we do it?
•
•
Well we can, but…
Each star takes a lot of cpu.
•
•
Typical AGB star (with my code) takes
between 2 days and 4 weeks depending
on the mass, mass-loss, and
nucleosynthesis
There are many uncertainties…
Main Problems
•
Mass-loss
•
•
•
Dredge-Up
•
•
•
Rate determines when AGB ends
Hence huge effect on yields
Remains hard to do properly
Surely its hydrodynamic…
Convective borders and overshoot?
How to Calculate Dredge-Up?
•
•
•
Depends on how you apply
Schwarzschild …. some apply it
naively and get NO dredge-up!
We search for a neutral point… a bit
like overshooting…
Depends on how you implement it…
Convection: mostly affects dredge-Up
Calculating Dredge-Up
No dredge-up
Results depend on
how you mix
and iterate
Deep dredge-up
Convection: Dredge-Up
•
•
•
•
•
Its not simple
Its not one-dimensional
Its not static
Its not mixing-length
But it is very very hard to calculate…
Convective Boundaries
Eg Herwig and O…
Trivia Question 4
•
Who said:
“The ultimate goal of nuclear physics
is to produce dust.”
•
Erwin Sedlmayr
(Has his own brewery
in Berlin…)
The Best on Offer?
•
•
•
•
•
•
Best are full, detailed models
Star by star…
From ZAMS to end of AGB…
With the best you can do for all the
physics…
One star per node on large cluster
Elements considered usually stop
before the s-process…
Amanda Karakas’ Thesis
•
Three compositions
•
•
•
Many masses
•
•
•
Solar, LMC and SMC (Z=0.02, 0.008, 0.004)
Since then has added Z=0.0001
M=1.5 to 7
Various bits published but not all 
Contact Amanda for yields…
Other Full Model Yields
•
•
Herwig (2004) Z=0.0001, M=2 to 6
Simon Campbell (in Fenner et al 2004)
•
•
•
•
•
Globular Cluster chemical evolution
M=1.25 to 6.5
[Fe/H]=-1.4 and mix from Big Bang and SN
Ie “Z” = 0.0017 (Z=1-X-Y is not proportional to Fe)
Ventura et al (2002)
•
•
•
Different theory of convection (FST, not MLT)
M=3 to 6 (HBB range)
Z=0.0002 to 0.01 (5 values)
Serious Advantage of Full Models
•
•
•
•
The interplay between HBB and TDU
Which ends first?
Unknown until we made some full
detailed models go all the way to end
of AGB
Has large effect on final envelope
composition
Trivia Question 5
•
Who were described by Ned Kelly as:
“…big, ugly, fat-necked, wombat-headed,
fat-bellied, magpie-legged, narrow-hipped,
splay-footed sons of Irish bailifs or English
landlords…”
•
The Victorian Police Force
The Other/Older Approach:
Synthetic AGB Models
•
•
•
AGB stars are complicated
AGB stars take a long time to compute
But Paczynski noted that the surface L
varied very nearly linearly with core mass
•
…at least for lower masses…
Core-mass v Luminosity
L/L = 59250(Mc/M – 0.5)
Other relations can be fitted…
•
•
•
•
•
Mass in intershell
convection zone
Duration of intershell
convective zone
Interpulse period
Etc
Entire population can be
synthesized very quickly!
Synthetic AGB Models
•
•
•
•
But only as good as the inputs
And things are not simple really
Eg dredge-up law does depend on the
envelope mass as well as the core and
composition
There is “memory” of the earlier life of
the star, at some level
The Biggest Problems?
•
•
•
•
•
Not easy to know
when to include
Depends on M and Z
Some turn it “on” at
specific M and Z
But what T? What
rho? How much
mass is burning?
Very hard…
•
•
•
Can fit the dredge-up
parameter l as a
function of everything
But that has to come
from detailed models
and is not known…
Beware of this!
Advantages
•
•
•
•
•
•
Well suited to yields!
Quick!
Can change reaction rates quickly
Can investigate mass-loss dependence
In principle…very powerful
But limited by the inputs of course
Synthetic Yields
•
Long history, starting with
•
•
•
•
Renzini and Voli 1981
DO NOT USE RENZINI & VOLI (1981)!
Many better yields around today!
Who?
•
•
•
•
Van den Hoek & Groenewegen (1997)
Marigo (2001)
Forestini and Charbonnel (1997)
Izzard et al (2004)
Comparison of synthetic
with detailed yields: 4He
Black = Karakas (full)
Blue = Izzard et al (syn)
Red = Marigo (syn)
Green = Forestini &
Charbonnel (syn)
Magenta = vdHG (syn)
Cyan = Ventura (full)
12C
13C
14N
Z=0.02
Z=0.02
Z=0.02
Z=0.008
Z=0.008
Z=0.008
Z=0.004
Z=0.004
Z=0.004
New Synthetic Yields from Izzard
•
•
•
Rob Izzard worked very hard on this
Aim is to include binaries
They go to places that single stars do not
•
•
•
So simple fits will not work all the time
Good fits to (our) detailed models
Real nucleosynthesis in most cases
•
Fit T and rho and then do real burning
Izzard Yields
Synthetic = red
Detailed = green
Izzard Yields
Massive stars too
Red and green =
full models
Cyan and magenta =
synthetic
Add binary population as well
•
Huge number of
extra parameters
to do with binaries
Includes
•
•
•
•
mass accretion
mass-loss
mass transfer
H
4.3
Population Synthesis
H
4He
4He
12C
12C
13C
13C
14N
14N
15N
15N
16O
16O
17O
17O
27Al
27Al
32S
32S
36Ar
36Ar
40Ca
40Ca
Trivia Question 6
•
•
Which WWII General said:
“I can take care of the Germans, but I am not sure I
can beat Montgomery and Eisenhower.”
George Patton
What is the Best Now?
•
•
Full models are probably best
Monash group
•
•
•
Karakas, Campbell, Lattanzio etc
Ventura et al 2002 (limited mass range)
Herwig 2004 (Z=0.0001 only, and
overshoot)
Synthetic if you must….
•
•
•
Izzard et al 2004
Beware: gap for M=7 to 12 = SUPER-AGB
Online yield calculator:
•
•
http://www.ciqua.org/binaryyields/index.php
For single stars set second at M=0.1 and
very wide 
Izzard Online Yield Calculator
Izzard Online Yield Calculator
Izzard Online Yield Calculator
Izzard Online Yield Calculator
Trivia Question 7
•
Which geologist/astronomer said:
“The beauty of science is that nature will
tell you when you are wrong. So will
your colleagues, but they may not
always be right!”
•
Jerry Wasserburg
Accepting Crafoord Prize
in Geology, 1986,
from King of Sweden
What is in the pipeline?
•
•
•
Project with Stan Woosley
He runs Z=0 supernovae
Their yields are the initial composition for
my AGB models and his massive models
The yields from that run (plus earlier, plus
Big bang…) are the next generation
After a few generations just go to scaled
solar
Full, detailed, consistent models
•
•
•
•
With all their uncertainties 
Publication and distribution
•
•
•
•
Karakas: contact her (copy of thesis)
Campbell: contact him (copy of thesis)
All will appear on webpages eventually…
Izzard for calculating your own…
•
Contact him for a copy of his thesis
The End?
More appropriate for a theorist