[Mesa-users] Issue with solar calibration simplex

[Aaron Dotter]

Hi John,

Please have a read through this paper,

http://adsabs.harvard.edu/abs/2018ApJ...860..131C

it has the answer to your question.  If you want to jump straight to the
answer, see section 2.2.

The general answer is that different surface boundary conditions have
different corresponding (solar-calibrated) alpha_MLT.  These alpha_MLT
values can vary from one to another by more than you're allowing for in
your settings for min_ and max_alpha in your inlist_simplex_search_controls.

Cheers,
Aaron

On Fri, Mar 1, 2019 at 5:55 AM John Smith via Mesa-users <
mesa-users at lists.mesastar.org> wrote:

> Hello All,
>
> I am new to MESA and I have been experimenting with the
> simplex_solar_calibration module. My issue is as follows:
>
> After changing the atmosphere to 'Eddington_grey' in the inlist file
> (default is 'photosphere_tables') and allowing the code to vary over the
> values of FeH, Y, and f_ov, the search converges to values off the desired
> accuracy. For example, for the radius I get R/R_sun = 0.9922189303 --  I
> attach below the inlist files and the output of the search.
>
> Do you have some advise as to what might be going wrong here?
>
> Now, if I allow also the mixing length parameter alpha to vary in the
> search on top of FeH, Y, and f_ov, the chi2 does not appear to converge,
> but the code instead ends up running into higher and higher values -- I had
> to stop it at around chi2 ~ 4000.
>
> I am certainly doing something very wrong here, but I am unable to see
> what might be the issue. I would very much appreciate any help/advice with
> this!
>
> Thank you in advance,
> John
> _______________________________________________
> 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/20190301/c1170685/attachment.html>