[Mesa-users] Evolving very massive stars

[Rob Farmer]

Hi,

Its great to see your inlist as well as your inlist having many different
timestep controls set.  The terminal output should tell you what is
limiting the timestep for your model, have a look back over the output from
around when the model started having difficulties.  Have a look at the
information here
http://mesa.sourceforge.net/controls_defaults.html#timestep_controls for
how to intercept the timestep controls.

Rob

On Tue, 21 Apr 2020 at 07:53, nakauchi--- via Mesa-users <
mesa-users at lists.mesastar.org> wrote:

> Dear mesa-users,
>
> I am computing evolution of very massive stars of ~ 1000Msun with various
> metallicities, Z/Zsun = 1e-1, 1e-2, 1e-3, 1e-4, and 0. I hope evolution is
> followed until He depletion, without resting on the MLT++ option, if
> possible. I neglect the wind mass-loss.
>
> In the case of Z/Z_sun = 1e-1, evolution can be followed until He is
> almost depleted at the center. But in the other metallicities, the
> computation stops by numerical difficulty during H or He buring. The
> numerical convergence becomes worse suddenly and the timestep decreases
> rapidly to the limiting value, min_timestep_limit. This happens when the
> star reaches the Hayashi line (effective temperature ~ 4800 K).
>
> I am wondering which inlist parameter plays a key role in making converged
> models until later-burning stage. I mainly changed  mesh_delta_coeff and
> varcontrol_target, but this does not resolve the problem. Does anyone have
> any idea from your experience?
>
> The inlist file is attached below. The HR diagram and evolution of central
> abundance are shown in fig_hrd and fig_abund, respectively. I am using
> mesa-r12115.
> Thanks in advance for any insights.
>
> Best regards,
> Daisuke
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200430/85775ec6/attachment.htm>