[Mesa-users] Switch donor function and accretion rate with implicit winds and rotation

[Zepei Xing]

Dear Mesa users,

I am one of the POSYDON developers. I am writing to describe two issues we encountered regarding mass transfer between two non-degenerate stars when running binary grids with MESA (version 11701) to build POSYDON grids.

The first issue is about identifying the donor and the accretor.  We use the 'contact' mdot_scheme for two main-sequence stars and the 'Kolb' mdot_scheme if one star leaves the main sequence. We have observed that even when setting keep_donor_fixed = .false., if the star 2 leaves the MS and fills the Roche lobe, the star 1 is still regarded as the donor. Consequently, In these cases, the star 2 will expand arbitrarily. We believe it's because the switch_donor function in the set_donor_star subroutine doesn't support the 'Kolb' scheme.

We implemented a temporary fix to allow the code to switch donor regardless of the mdot_scheme. The approach is to calculate the mass transfer rate for both stars and switch donor when the mdot from accretor is higher. This is implemented via use_other_rlo_mdot but it also requires a small change in binary_mdot.f90 in line 85 and 96.

The second issue is about the implicit mass accretion rate calculation. We use implicit method to calculate the accretion rate considering the angular momentum transport of the accreted material and the rotation of the star. In run_binary_support.f90, there are three conditional statements working to avoid repeated calculations of the implicit mdot for the accretor when using the implicit scheme with rotation and implicit winds in line 444, 501, and 603. However, s% was_in_implicit_wind_limit points to the star 2 instead of the accretor as b% s_accretor% was_in_implicit_wind_limit. They lead to troubles for the runs with the star 2 as the donor and these runs are not rare after we implemented the switch donor function. In such binaries, the calculation of the implicit mdot for the accretor will not be skipped, but the explicit mdot will be skipped in binary_mdot.f90, resulting in the stellar winds of the accretor being considered the accretion rate.

Besides, the idea of avoiding repeat calculations of the accretor is not perfect. According to the code in run_binary_support.f90 and binary_mdot.f90, the implicit accretion rate is calculated only in the first iteration and then both explicit and implicit accretion rate will not be updated in the following redos. However, in each redo the mass transfer rate changes and it is possible that the mass transfer rate becomes smaller than the accretion rate.

Please find attached the necessary files, including our input files, zams models and modified binary_mdot.f90, which can reproduce a run that can demonstrate the issue with the implicit mdot. You will see at step 650, the transfer rate is 2.836E-05 while the accretion rate is 7.691E-05. And after step 1120, the reverse mass transfer happens, the accretion rate fluctuates between positive and negative values.



Sincerely,
Zepei Xing[https://res-h3.public.cdn.office.net/assets/mail/file-icon/png/generic_16x16.png]zams_z1.42m3_y0.2511 2.data<https://unigech-my.sharepoint.com/:u:/g/personal/zepei_xing_unige_ch/EfsakDufOHxJp2KwhvHjhtABT-BAXT-z2Xlfo6Q-2Y5ROg>

University of Geneva
Geneva observatory

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inlist_project
Type: application/octet-stream
Size: 2837 bytes
Desc: inlist_project
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inlist2
Type: application/octet-stream
Size: 8604 bytes
Desc: inlist2
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inlist1
Type: application/octet-stream
Size: 8902 bytes
Desc: inlist1
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: run_binary_extras.f
Type: text/x-fortran
Size: 67003 bytes
Desc: run_binary_extras.f
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: run_star_extras.f
Type: text/x-fortran
Size: 70582 bytes
Desc: run_star_extras.f
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: binary_mdot.f90
Type: text/x-fortran
Size: 37509 bytes
Desc: binary_mdot.f90
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230512/bcab7371/attachment-0002.bin>