[Mesa-users] Query regarding binary modelling.

[Mathieu Renzo]

Hi Sohan,

This is a very difficult problem you are tackling!

First of all, from your plot of the donor it seems your system is 
starting mass transfer before the end of the donor's main sequence (case 
A) -- note the absence of a Henyey hook before RLOF. Indeed, your 
inlist_project has 4 days as initial period, which might not be enough 
to get a case B at your Z of 0.001 (set in your inlist extra). This 
suggests you might not be running what you intend to run. For a case B 
you probably want to widen your initial orbit or lower Z even further.

Regarding rotation in the accretor: in my experience the factor limiting 
the timesteps and leading to convergence failures is often 
`min_J_div_delta_J` in controls. I have had some success increasing its 
value. Another trick that has helped me is to limit the increase of a 
timestep after it plummets, so that it grows more gently.
Note that low Z also means weaker winds, which makes it harder for the 
accretor to get rid of incoming material once it reaches critical 
rotation. Another thing you might try is to lower `w_div_wcrit_max`, 
though I see in your inlist you set it to a fairly low 0.5 already, so 
actually this might be too low. I don't know if you still have this 
question, but from the comment in your inlist it seems you might be 
confused about what this parameter is. If the `controls_default.list` is 
not enough, I suggest to do:

grep -Ri "w_div_wcrit" $MESA_DIR/star/p*/*

to find the equations were it enters, which should make it clear.

Unfortunately, robust solutions for your problems do not exist as of yet 
(at least as far as I am aware) and you will have to painfully 
experiment until you find the right trade-off between having your models 
running to the desired end-point and the science you want to do. Maybe 
you can find inspiration in the setup I used for this paper 
<https://arxiv.org/abs/2107.10933> available at this zenodo repository 
<https://zenodo.org/record/4701565>. This worked well for a 25+17Msun 
binary until the end of the main sequence of the accretor star, but I 
have had some success with minor adjustments also at higher masses.

Hope this helps, and good luck!
Mathieu


On 8/29/21 6:36 AM, Sohan Ghodla wrote:
>
> Hi all,
>
> I have been trying to run the |mesa/binary/test_suite/double_bh| model.
> My aim is to study the accretion-induced spin effects on the secondary 
> for a range of secondary masses. The issue is that the *models mostly 
> fail to converge* *during the case b mass transfer* from the donor to 
> accretor.
> I have noticed that this usually happens when the accretor’s 
> |surf_avg_v_rot| becomes identical to |surf_avg_v_crit |shown on the 
> RHS in plot 2 below.
>
> Tweaking variables like |rotational_mdot_boost_fac|, 
> |implicit_mdot_boost| & |implicit_scheme_tolerance| does help if I get 
> lucky but that also starts to fail as soon as I change the 
> |new_omega_div_omega_crit| value. I have added my inlists to this 
> email. The only major modifications here from the test_suite case are:
>
>  1. |do_initial_orbit_sync_1| and |do_initial_orbit_sync_2| have been
>     turned off
>  2. The metallicity is different
>
> Any suggestions will be highly appreciated.
>
> *Donor's Plots:*
>
> image.png
>
>
> *Accretor's plots:*
>
> image.png
>
>
> Thank you,
>
> Sohan Ghodla
>
>
>
> _______________________________________________
> mesa-users at lists.mesastar.org
> https://lists.mesastar.org/mailman/listinfo/mesa-users
>

-- 
Mathieu Renzo <https://users.flatironinstitute.org/~mrenzo/>
Flatiron Research Fellow
Center for Computational Astrophysics
Columbia University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210830/40f64d28/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 106934 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210830/40f64d28/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 98463 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210830/40f64d28/attachment-0001.png>