[Mesa-users] Need assistance with double_bh test suite in MESA version 10398 - Chemically Homogeneous Evolution and Radius Discrepancy

[Levi]

Hello all,


I am writing to seek some assistance and expertise regarding a specific issue I have encountered while using the double_bh test suite in MESA version 10398 (complete file attached).


I have set the initial period to 1.1 days with M1 mass at 70 Msun, while M2 mass ranges from 60 to 10 Msun with a 10 Msun interval. Most of the models follow the expected behavior described by Marchant (2016), undergoing chemical homogeneous evolution (CHE) until the central helium is depleted, at which point the evolution ends. However, I have noticed that a subset of models deviate from this expected behavior.


Take the model with M1 = 70 Msun and M2 = 40 Msun as an example. In this model, M2 undergoes CHE, while M1 does not undergo CHE. This system terminates its evolution during the main sequence phase due to L2 overflow. This is because there is a significant drop in its rotational velocity, which is reduced to 0.3 times the critical velocity after the first mass transfer event (refer to subfigure f). Subfigure b depicts the variation in mass of both stars during the mass transfer process, while subfigure e illustrates the specific mass transfer rate. 
My understanding is that during the mass transfer process, as M1 below M2, the increasing orbital separation results in a transfer of angular momentum to the orbital angular momentum. Consequently, the rotation speeds of both stars (M1 and M2) continue to decline. Due to the prolonged duration of this process, the rotation velocity of M1 drops below 0.3 Vcrit. Based on my tests, I have observed that when the rotation velocity of M1 falls below 0.3 Vcrit, the model struggles to complete chemically homogeneous evolution.

Promptly setting M2 as the donor may allow M1 to complete chemically homogeneous evolution. To address this issue, I investigated the MESA code responsible for the donor star selection. I found the following relevant code snippet in the “binary_evolve.f90” file:

if (b%rl_relative_gap_old(b%a_i) > b%rl_relative_gap_old(b%d_i) .and. &
b%rl_relative_gap_old(b%a_i) < -b%implicit_scheme_tolerance .and. &
b%rl_relative_gap_old(b%d_i) < -b%implicit_scheme_tolerance .and. &
abs(b%mtransfer_rate/(Msun/secyer)) < b%mdot_limit_donor_switch) then
switch_donor = .true

Combine the above code, the main reason is that mtransfer_rate is always maintained at a relatively high value (Figure e). According to equation 19 in MESA III (Paxton et al. 2015), the substance transfer rate is closely related to the radius of donor. Subfigure a clearly shows that M1 is indeed in a state of Roche lobe overflow (RLOF) for model numbers between 200 and 660.


Moreover, I noticed that for the M1 = 70 Msun and M2 = 50 Msun model, both stars undergo chemically homogeneous evolution (attachment). They can timely switch donors and maintain sufficient rotation velocity. 


My main concern is why the radius of M1 change in the mass transfer process is so different despite the parameters of these two models being so close, leading to completely different evolutionary results, and how to address this issue.


Best wishes,

Levi

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230725/7f58f8a9/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p1.1_70_40_m.png
Type: image/png
Size: 1481656 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230725/7f58f8a9/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p1.1_70_50_m.png
Type: image/png
Size: 1630756 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230725/7f58f8a9/attachment-0003.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: m1_70_m2_40.zip
Type: application/zip
Size: 16248156 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230725/7f58f8a9/attachment-0001.zip>