[Mesa-users] Issues Evolving AM CVn Binary

[AbdelBassit Senhadji]

Hi Josh,

Here are couple of controls you can use:

&controls
   max_timestep = 1d4 ! Adjust based on your system's needs
   min_timestep = 1d-6
   varcontrol_target = 1d-4
   mesh_delta_coeff = 0.5
   mesh_delta_coeff_for_highT = 0.3
   newton_iterations_limit = 20
   use_other_energy_eqn = .true. ! Use a different energy equation solver if applicable
   max_surface_cell_dq = 1d-3 ! Ensure a fine mesh near the surface if mass transfer occurs there
   use_accumulate_factor = .true. ! Helps with mass loss
   max_years_for_timestep = 1d-3 ! Limit the maximum timestep to avoid large jumps
   min_years_for_timestep = 1d-8
/

  *   max_timestep and min_timestep: Ensure that these are set to reasonable values for the evolutionary phase you are studying. Reducing max_timestep during phases of rapid changes can help with convergence.
  *   mesh_delta_coeff and mesh_delta_coeff_for_highT: These control the grid resolution. Increasing the resolution by lowering these coefficients can sometimes help, though it may slow down the calculations.
  *   varcontrol_target: This parameter controls the allowed changes in model variables per timestep. Tightening this tolerance (by setting a lower value) can improve stability.
  *   newton_iterations_limit: Increasing this limit allows more iterations per timestep, which can help the solver converge.


I also suggest to have a look at the opacity tables and the equation of state (EOS) used are appropriate for a helium-rich donor star. Sometimes convergence issues can stem from improper handling of the physical conditions in such stars.

Hope these suggestions will help!

Abdel
________________________________
De : Mesa-users <mesa-users-bounces at lists.mesastar.org> de la part de Joshua Aiken via Mesa-users <mesa-users at lists.mesastar.org>
Envoyé : 14 juin 2024 19:59
À : mesa-users at lists.mesastar.org <mesa-users at lists.mesastar.org>
Objet : [Mesa-users] Issues Evolving AM CVn Binary

Hi everyone,

I am currently trying to model the evolution of a system that starts as a (lowish mass) ZAMS donor star and a white dwarf companion (a cataclysmic variable (CV) system), that through mass transfer erodes the H content of the donor star, turning the system into an AM CVn system (a H-deficient/He-rich donor onto a white dwarf accretor, with an orbital period less than 1 hour). For the most part, the evolution works as expected, however, during the AM CVn phase, MESA struggles to converge and I haven't been able to figure out why (and how to fix the issue).

I am using MESA r211211 and I have provided the inlists that I am using for a specific case (M1=1.10Msun, M2=1.2Msun, P=2.88days) and also included the history file and plot of the period-donor mass evolution for that case. The convergence issues start when mass transfer begins during the AM CVn phase (when the period is below ~1Hr). The provided case specifically ends due to the redo limit, but in other similar cases the timestep becomes really small and the evolution still doesn't go anywhere. The mass transfer rate is only ~10^-7 Msun/yr which is completely expected for an AM CVn.  Has anyone experienced similar issues when trying to transfer mass from a donor composed of almost entirely helium? or have any ideas on what the problem might be/things to try?

Any help on getting these evolutions to progress further through the AM CVn phase (ideally until the donor mass is about 0.02 Msun) would be greatly appreciated.

Thanks,
Josh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20240615/c1318fbc/attachment.htm>