[Mesa-users] Numerical issues during Helium flash

[Théo-Courty]

Hi,

Thanks both of you for your suggestions! Changing energy_eqn_option to « eps_grav » seems to have fixed the issue.

Best,

Théo

> Le 12 févr. 2026 à 15:54, Radek Smolec <smolec at camk.edu.pl> a écrit :
> 
> Hi Theo,
> 
> Also, please try to uncomment this control in your inlist:
>       !convergence_ignore_equL_residuals = .true.
> so the control is .true. this should help going through Helium flash smoothly. Relaxing the convergence criterion at this stage of evolution is very helpful.
> Best wishes,
> Radek
> 
> czw., 12 lut 2026 o 15:47 Ana Antonini via Mesa-users <mesa-users at lists.mesastar.org <mailto:mesa-users at lists.mesastar.org>> napisał(a):
>> Hi Theo, 
>> 
>> Without extra information, I'd say you can try adding some extra mesh and timestep controls to regions/quantities of interest, like "delta_lgL_He_limit",  and see if it helps.
>> Changing to the energy_eqn_option to "eps_grav" might also be helpful.
>> Another thing that guides me when starting a new project with MESA is finding out if there are any works on the subject that use MESA, and seeing what kinds of controls and physics those works are adopting. 
>> 
>> We could say more if you provide us with more information. For example, what is MESA reporting as the reason for bringing down the timesteps? 
>> Perhaps also a little of the terminal output when timesteps start to decrease a lot. 
>> 
>> Good luck on your mesa journey!
>> Best,
>> Ana Antonini
>> 
>> 
>> Em qui., 12 de fev. de 2026, 9:52 AM, Théo-Courty via Mesa-users <mesa-users at lists.mesastar.org <mailto:mesa-users at lists.mesastar.org>> escreveu:
>>> Dear Mesa users
>>> 
>>> I’m a PhD student fairly new to Mesa. I’m currently trying to evolve stars with different initial masses from ZAMS to WDCS. When I try to evolve lower mass star (1-2 Msun), I come across numerical difficulties during He Flash phase: the timestep becomes very small (log dt < -8 yr) and the simulation then takes forever (> ~10-20h) or can even stop due to timestep limit. My inlists are based on the test suite make_co_wd and I’m using version r25.12.1. I’m looking for advices/workarounds to ease this phase so the simulations can run in a reasonable amount of time. I’m attaching my inlists to this message. Thanks a lot!
>>> 
>>> Best,
>>> 
>>> Théo
>>> 
>>> _______________________________________________
>>> mesa-users at lists.mesastar.org <mailto:mesa-users at lists.mesastar.org>
>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>> 
>> _______________________________________________
>> mesa-users at lists.mesastar.org <mailto:mesa-users at lists.mesastar.org>
>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20260213/85e6d153/attachment.htm>