[MODEST] some news items

[Piet Hut]

Dear Colleagues:

Here is some assorted news about our MODEST activities:

Next two meetings:
MODEST-2, Amsterdam, December 16-17, 2002
MODEST-3, Australia, July 9-11, 2003

Our 10-author review article, written immediately following MODEST-1,
has now been formally accepted for publication by New Astronomy.
We have made some minor modifications, mostly additional references,
and you can find the final version as the revision under the old URL:
http://arXiv.org/abs/astro-ph/0207318 .

Three weeks ago, we had a few lively discussion days here in Princeton
when Peter Eggleton, Luda Kiseleva and Onno Pols came over for a visit
at the Institute for Advanced Study.  Of the more-or-less locals,
Steve McMillan, Jamie Lombardi, and Jarrod Hurley also used the
opportunity to join us.  Some highlights:
-- Peter told us about his interest in symbiotic stars, and we
   discussed possible improvements in ways of handling those in our
   large-scale simulations.
-- Onno and Jamie started to work on ways to use Jamie's SPH-inspired
   star addition program to add two of Onno's stars, in order to see
   whether they will be able to get the merger product in good enough
   shape to be acceptable as a new starting condition for Onno's version
   of the Eggleton code.
-- Jarrod started working on adapting his stellar evolution fitting
   functions to the interface conventions proposed in our review paper,
   so that he can combine them with Jamie's triptych setup.

Attay Kovetz and Dina Prialnik told me a few weeks ago that they have
made progress with their code, which is now capable of reaching the
double-shell source phase unattended (going through MS, helium core
flash, HB and up to the AGB).  Such a calculations now takes a few
minutes on a lap-top, and only about 500 time steps.

Don VandenBerg and I exchanged some email recently about modest-related
topics.  Here are some of his specific conclusions:

   I have realized that, as efficient as my stellar evolution code is, one
   that employs tabular physics (equation of state, opacities, nuclear
   reactions) is the only way to go to have a truly speedy program.  Just
   a few months ago, Pavel Denissenkov came to U.Vic. to be my research
   associate (i.e., I am funding his position) for a couple of years.  He
   uses an old version of a Paczynski code (which could be easily brought
   up to date, but hasn't yet) and it is amazingly fast.  He told me that
   (using a 1.8 GHz PC), he followed the evolution of a 1 M_Sun star nearly
   all the way to the He flash in about 15 min.  (This is using a Lagrangian
   code, requiring 10-15,000 individual stellar models.)  I don't personally
   believe that one should use a Lagrangian code for RGB evolution: the
   technique developed by Peter Eggleton is the ONLY way to go.  Using my
   code I can also evolve to the He flash in about 15 cpu minutes (using
   a Lagrangian code for the main-sequence evolution, and the Eggleton
   technique for the RGB evolution), taking about 700 models.  I expect
   that this time could be reduced by at least a factor of 10 if I employed
   tabular physics - which is starting to get into an interesting regime
   for MODEST work.  But one will never be competitive with such analytic
   formulae as those produced by Hurley et al (2000, MNRAS, 315, 543),
   which must surely be the best way to go for non-interacting stars.

All this seems to point in the direction that including full stellar
evolution in star cluster simulations is becoming really feasible, as
far as the hardware is concerned.  In a long simulation of a few weeks
duration, there are 10^4 units of `a few minutes', so in principle we
could easily handle a direction computation of those stars that are
merger products, among the 10^5 to 3.10^5 stars that we can simulate
for a Hubble time (the other stars can be dealt with through fitting
formulae and/or table lookup).

Any comments, anyone?

Piet

---------------------------------------------------------------------------
  Piet Hut                              email: piet at ias.edu
  Institute for Advanced Study          tel.:  (609)734-8075
  1 Einstein Drive                      fax:   (609)951-4489
  Princeton, NJ 08540-0631              web:   http://www.ids.ias.edu/~piet
---------------------------------------------------------------------------