[Mesa-users] A very short timestep of the evolution of a He star appears during it towards an AGB star

[曾耀田]

Dear Josiah and all MESA users,
Sorry for the late reply. Pretty thanks for your help! My He model indeed shows a density inversion at the very outer layer of the star while the He star starts experiencing its AGB phase. To see how density inversions varied with the different He stars and settings I did a series of tests (i.e., try to set z to a small value of 1d-4, to change the mass of the He from 1.0 to 2.0 with an interval of 0.5 solar masses, and some combinations of these settings.) As your suggestion, I also try to use the "okay_to_reduce_gradT_excess" to the previous He star models that have been tested. All these models evolving faster than before from the bottom of the AGB phase to the top. My main goal of these models is to get the evolutionary tracks during their WD phase. Although these calculations with MLT and MLT++ will end with different final masses due to the enhanced mass loss in MLT++, which makes the evolutionary tracks different, it would not be expected that the structure has substantial changes. I still tested one case (i.e., M = 1.23 Msun) with and without MLT++ to show how different they are during their WD phase. These two calculations do not show substantial changes making them unsuitable to my goal. Now, all my confusion is solved. Thanks again!
With best wishes,
Yaotian


-----Original Messages-----
From:"Josiah Schwab" <jwschwab at ucsc.edu>
Sent Time:2021-05-20 02:42:26 (Thursday)
To: "曾耀田" <zengyaotian at ynao.ac.cn>
Cc: "Jacqueline Goldstein" <jgoldstein at astro.wisc.edu>, "mesa-users at lists.mesastar.org" <mesa-users at lists.mesastar.org>
Subject: Re: [Mesa-users] A very short timestep of the evolution of a He star appears during it towards an AGB star


Hi Yaotain,


I'm re-pasting some of the info from another active thread because I think it applies here too.  I've worked a decent amount on related models (e.g., https://ui.adsabs.harvard.edu/abs/2019ApJ...885...27S/abstract) and to the best of my understanding there aren't any timestep or solver controls that are going to let you easily go through this phase.


This model has a near- (or even maybe even super-) Eddington luminosity (b/c of the He shell burning on a massive degenerate core).  Correspondingly, it has developed an inflated envelope with outer layers that are radiation pressure dominated and have a density inversion.  The steep entropy gradient at the base of this (inefficient) convection zone is usually what is responsible for driving the MESA timestep down.


This is similar to the behavior in massive stars that the MLT++ "hack" described in Section 7 of Paxton et al. (2013) is designed to numerically circumvent (but not physically solve).  You could consider using okay_to_reduce_gradT_excess and related controls, depending on what your goals for the model are.


Josiah
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210608/fc3d98c5/attachment.htm>