Pre-main-sequen

e stellar evolution

For spheri ally symmetri a retion, expe t that the stru ture of
the ore + infalling envelope looks like ( f Shahler, Shu & Taam
1980):

Opacity gas
Hydrostatic core

Dust destruction
front

Accretion shock

 A hydrostati ore, initially of low mass due to the non homologous nature of ollapse.

 A retion onto the ore at a rate M_ , via an a retion sho k at

the stellar surfa e.

 An opa ity gap in the inner region of the ow, be ause the stellar

ee tive temperature ex eeds the dust destru tion temperature.

The opa ity gap partially de ouples the problems of protostellar

evolution and infall, as it allows photons produ ed at the sho k to
stream out freely.
Still need to spe ify M_ onto the ore. Assume that this is onstant.
1

Evolution for low masses

For masses M  M , deuterium burning o urs near the enter

for T ex eeding  106 K:

 Provides su ient luminosity to maintain the star onve tively

unstable.

 Conve tion `instantaneously' transports a reted deuterium to

the enter.

 ! steady-state deuterium burning.

Deuterium burning luminosity is,
0
L ' M_ = 12L B
D

10

1
M_
CA
Myr 1

where is the nu lear energy per unit mass available via deuterium
burning:
XQ
;
 [D=H
m
H

where,

 [D=H

= 2  10 5 is the fra tional interstellar abundan e of

deuterium by number.

 X = 0:7.
 Q = 5:5 MeV is the energy available per rea tion.
2

The deuterium burning rea tion,

1

H +2 H !3 He +

yields,

0
 = 4:2  107[D=H 
D

10
1
 A  T A11 8
erg g
1 g m 3 106 K
:

s 1:

Strong temperature dependen e a ts as a thermostat, maintaining

the ore temperature near 106 K.

! a mass-radius relation that is almost linear during the phase of

a tive deuterium burning.

Note: for low mass stars and M_ = 10 5 M yr 1, the a retion

time is shorter than the Kelvin-Helmholtz time.

Numeri al al ulations by Stahler (1988, ApJ,332, 804):

 Approximately linear growth of radius with mass.

 Radii mu h smaller than Hayashi's protostars.
 Dierent intial radii onverge to a single relation.
4

Dierent a retion rates yield dierent tra ks. Lower a retion

rates ! longer hara teristi a retion times relative to the KelvinHelmholtz time ! gravitational ontra tion more important.

Radii of 3 4 R are inferred for the largest T Tauri stars, so

these radii are observationally plausible.

Assume now that the main phase of stellar a retion ends abruptly.
Then,

 During the a retion phase, follow R(M ) tra k essentially set by

deuterium burning riteria in ore. Infall means this phase is not
opti ally visible.

 After a retion eases, opti ally visible stars ontra ts towards

ZAMS along traditional Hayashi tra k.

Boundary between these phases denes a stellar birthline, above

whi h no stars should be found in HR diagram.
Obviously a disk has to remain to explain the properties of Classi al T Tauri stars. However, the a retion rate in CTTS is typi ally
 10 8 Myr 1 (Gullbring et al. 1998), so not unreasonable to
ignore this as a rst approximation.

HR diagram for pre-main-sequen e stars in Taurus (Kenyon &

Hartmann, 1995, ApJS, 101, 117):

Roughly onsistent with the birthline on ept. Note also that the
lassi al T Tauri stars (solid ir les) on average lie further from the
main sequen e than weak-lined T Tauri stars (open ir les).
Dispersion in the lifetime of disks ould be explained via roughly a
fa tor 3 variation in the initial disk mass (Armitage, Clarke & Palla
2002).
7

More massive stars

Be ause the radii are smaller (less luminous) than in the lassi al theory, more massive stars develop radiative stable regions while
a reting.
Low mass

Intermediate mass
Radiative barrier
> core D depletion

fully convective
steady D core
burning

D shell burning
around radiative
core

Shell burning of D drives an in rease in the stellar radius (to burn

D in a shell, need higher temperatures further out than for ore
burning ! greater radii).
For su iently high mass, gravitational ontra tion dominates and
the pre-main-sequen e region is onned to lose to the ZAMS.

Numeri al al ulations by Palla & Stahler (1990, ApJ, 360, L47):

Stars more massive that about 8 M are predi ted to have ongoing
a retion throughout their pre-main-sequen e lifetimes.

HR diagram from Palla & Stahler (1993, ApJ, 418, 414):

10

Weaknesses of the birthline on ept

Some of the assumptions of the birthline on ept are questionable:

 Spheri al symmetry. If a retion o urs via a disk:

(i) The boundary onditions at the stellar surfa e will be dierent in detail (dierent entropy).
(ii) Geometri ally, we should be able to see the photosphere
during the a retion phase for systems that are lose to fa eon. These ould lie above the birthline.

 Constant a retion rate. Reasonable to assume that M_ 

3=G to order of magnitude, but unlikely to be a onstant.
s

 Sudden end to a retion.

Seems unlikely unless there is a

ausal link to some aspe t of stellar evolution (eg jets start up
when deuterium burning ommen es).

Several authors have shown that when reasonable un ertainties in

these parameters are in luded, on lusions derived from pre-mainsequen e models are un ertain for ages below around 1 Myr.
Referen es:

 Barae et al., 2002, A&A, 382, 563.

 Tout, Livio & Bonnell, 1999, MNRAS, 310, 360.

11

For greater ages, observations of e lipsing pre-main-sequen e binaries provide some eviden e for the reliability of model tra ks:

Figure from Palla & Stahler (2001, ApJ, 553, 299).

12