[Mesa-users] Model a star in the HR diagram

Mon Apr 9 11:19:57 EDT 2018

First things first, I realised when I loaded your inlist that you never 
said which version of MESA you're using.  I've run your inlists on r10398 
but I had to remove the controls

       save_pulsation_info_when_terminate = .true.
       save_pulsation_info_filename = '1M.mesa'

as they aren't recognised any more.

For me, each iteration terminates with something like

  have reached logg limit
                                                    logg    2.9915944972778181D+00
                                              logg_limit    3.8700000000000001D+00

save LOGS/profile15.data for model 1181
save photos/x181 for model 1181
save final_profile.data for model 1181
  saved to final.mod
        1181   7.729434   5286.029   1.479705   1.479705   1.500000   1.277855   0.000000   0.000183   0.247000   0.600698   4.915221   2260      0
    5.055249   4.752604   0.811193 -11.055075   0.340691 -99.000000   0.222145   0.999716   0.000009   0.000292   0.399009   0.410405      6      0
  1.4501E+09  20.545906   1.468387 -11.930926   4.568588  -6.981898   0.000000   0.000002   0.000000  2.840E-04  2.932E-04  0.110E-09    varcontrol

  failed to find chi^2 for this run
  failed to get new chi^2 -- try again

The astero module sees the run as "failing" because the logg limit is 
reached before any models have chi^2 computed.  A quick search will show 
you that the logg_limit is set in &astero_search_controls with the line

          sigmas_coeff_for_logg_limit = -5 ! disable by setting to 0

among a set of controls that limits the run to halt the evolution if you 
go too far from the target values.  Note that these are applied even if 
you aren't including that parameter in your chi^2.  So in this case, even 
though you aren't fitting logg, the astero module still sees that you've 
set

       logg_target = 3.97d0
       logg_sigma = 0.02d0

With sigmas_coeff_for_logg_limit = -5, that means logg limit is 
logg_target - 5*logg_sigma, which is the 3.87 that you see in the output.

These constraints are only being applied once the evolutionary track is 
within the range you've specified for age:

       min_age_for_chi2 = 1.45E+009 ! -1 ! (years) only use if > 0
       max_age_for_chi2 = 1.50E+009 ! -1 ! (years) only use if > 0

I don't know what you're trying to do, so you'll have to decided how to 
proceed.  But my guess is you don't want the logg_limit, so set that 
parameter to zero and check the others while you're there.

Finally, pay attention to the output *before* the "failed to find..." 
line.  The last few steps of the evolutionary run often tell you why MESA 
star stopped before astero was able to start evaluating chi^2.

Cheers,
Warrick

------------
Warrick Ball
Postdoc, School of Physics and Astronomy
University of Birmingham, Edgbaston, Birmingham B15 2TT
wball at bison.ph.bham.ac.uk
+44 (0)121 414 4552

On Mon, 9 Apr 2018, gnmandrade at hotmail.com wrote:

> Hi Warrick,
>
>
> Here I send the missing files!
>
> Thanks!
>
>
> Gon\E7alo
>
>
> ________________________________
> De: Warrick Ball <wball at bison.ph.bham.ac.uk>
> Enviado: 9 de abril de 2018 15:38
> Para: gnmandrade at hotmail.com
> Cc: mesa-users at lists.mesastar.org
> Assunto: Re: [Mesa-users] Model a star in the HR diagram
>
> Hi Gon\E7alo,
>
> You only sent the inlist for the astero module, not for MESA star itself
> (i.e. the file with &star_job and &controls).  I can't guarantee that I'm
> running the same thing as you unless you send that inlist too.
>
> W
>
>
> ------------
> Warrick Ball
> Postdoc, School of Physics and Astronomy
> University of Birmingham, Edgbaston, Birmingham B15 2TT
> wball at bison.ph.bham.ac.uk
> +44 (0)121 414 4552
>
> On Mon, 9 Apr 2018, gnmandrade at hotmail.com wrote:
>
>> Hi Warrick,
>>
>>
>> I had already done that!
>>
>> I leave my inlist as an attachment. Thanks for your help!
>>
>>
>> Cheers,
>>
>> Gon\E7alo Andrade
>>
>>
>> ________________________________
>> De: Warrick Ball <wball at bison.ph.bham.ac.uk>
>> Enviado: 9 de abril de 2018 15:21
>> Para: Gon\E7alo Andrade
>> Cc: mesa-users at lists.mesastar.org
>> Assunto: Re: [Mesa-users] Model a star in the HR diagram
>>
>> Hi Gon\E7alo,
>>
>> First, make sure that the \chi^2 has no seismic component by setting
>> `chi2_seismo_fraction` to 0 in `&astero_search_controls`.
>>
>> Second, make sure that you're telling the `astero` module that there
>> aren't any frequencies by setting `nl0`, `nl1`, `nl2` and `nl3` all to 0
>> (same inlist).  I haven't tested but I suspect that if you won't need this
>> after the first step above.
>>
>> I'm hoping that resolves your issues.  Please do share your inlists anyway
>> though, in case that isn't the solution.  Don't hesitate to include
>> inlists with your opening emails.  They're small and it saves us having to
>> ask for them, which we probably will!
>>
>> Cheers,
>> Warrick
>>
>>
>> ------------
>> Warrick Ball
>> Postdoc, School of Physics and Astronomy
>> University of Birmingham, Edgbaston, Birmingham B15 2TT
>> wball at bison.ph.bham.ac.uk
>> +44 (0)121 414 4552
>>
>> On Mon, 9 Apr 2018, mesa-users at lists.mesastar.org wrote:
>>
>>> Dear all,
>>>
>>>
>>>
>>> I am currently using MESA and I need to make a model of a star as close as possible to a given T_eff and log(L).
>>>
>>>
>>> Up to this moment I have been trying to use the solar_calibration example in the test_suite and the example in the Astero folder that is available inside the star directory, but I have encountered two main problems:
>>>
>>>
>>> 1) I have tried to remove the oscillation computation, since I don't need it and it consumes time, by removing all the frequencies in the file "inlist_astero_search_controls" and by commenting the &mode sections in the gyre.in file but, once the models get close to the expected, it begins calculating the frequencies anyway.
>>>
>>>
>>> 2) Ignoring the frequencies problem, I have managed to make small physics changes I needed and the calibration by simplex still functions perfectly fine. However, the moment I change the target T_eff and log(L) and the starting mass, alpha, etc, the code runs but always says it could not find chi^2 in that run, and does not present the results in the output file.
>>>
>>> I have even tried doing a single run with a given mass, alpha, etc, and copying the resulting parameters to the target in the "inlist_astero_search_controls" and I still get the same result.
>>>
>>>
>>> Are there any useful commands I might be missing?
>>>
>>> If needed I can attach the inlists I am using.
>>>
>>>
>>>
>>> I grateful for any help you could give.
>>>
>>>
>>> Best regards,
>>>
>>> Gon\E7alo Andrade
>>>
>>>
>>
>