[Mesa-users] Fwd: Trying to model a possible progenitor for type Ib Supernova

Jared Goldberg goldberg at physics.ucsb.edu
Tue Mar 24 16:43:39 EDT 2020 

Oops, forgot to hit reply all to CC mesa-users on this...

---------- Forwarded message ---------
From: Jared Goldberg <goldberg at physics.ucsb.edu>
Date: Tue, Mar 24, 2020 at 1:30 PM
Subject: Re: [Mesa-users] Trying to model a possible progenitor for type Ib
Supernova
To: Amar Aryan <amar at aries.res.in>


Hi Amar,

Unfortunately we can't diagnose the problem getting to core-collapse from a
pgstar plot alone.
Have you tried running with thermohaline mixing turned off at this stage?
It should not be necessary
on such short timescales. Otherwise, what was setting the timestep leading
up to the error? What is
the hydro doing? Is hydro even on? Perhaps someone on this list with more
experience with very stripped
stars can weigh in if this issue looks familiar.

As for how to explode the star, if you have a model at the time of core
collapse, you can excise
the core and set the explosion energy and explode the star using MESA by
following along with the
test_suite case *example_ccsn_IIp*, using your model at core collapse as
the input stellar model in
inlist_infall instead of the default 16M.mod. Then, set the explosion
energy by setting
inject_until_reach_model_with_total_energy = your desired explosion energy
(in ergs) in inlist_edep.
To set the total Nickel mass at the time of explosion, you can set the
value by setting
x_ctrl(12) = your desired Ni mass (in solar masses) in inlist_shock_part5.
I also suggest reading
the rn script to see how the whole process works.

Hope this helps!

Cheers,
~Jared





On Tue, Mar 24, 2020 at 3:01 AM Amar Aryan <amar at aries.res.in> wrote:

> Hi Jared,
>
> First of all, thank you very much for your suggestions. I pray for the
> welfare of the world from this COVID-19 pandemic and hope that you are fine.
>
> As per your suggestions, I increased the mass loss rate to 0.03M_solar per
> year in inlist_remove. This cuts the mass of the star more rapidly and the
> star mass near the core-collapse reaches around 3.4 M_solar, which is my
> requirement. All the inlists runs successfully but the final inlist
> (inlist_to_core_collapse) shows the same timestep error. As per the
> suggestions in Rob Farmers paper, I made the following changes in various
> inlists :
>
> upto *Si burning phase* (i.e. upto inlist_to_si_burn) we set,
>
> varcontrol_target  = *5d-5*
> min_timestep_limit = *1d-12*
> dx_nuc_drop_limit = *1d-3*
> dx_nuc_drop_min_X_limit = *1d-3*
>
> Thereafter, in the final inlist ( inlist_to_core_collpase)
>
> varcontrol_target  = *5d-5*
> min_timestep_limit = *1d-12*
> dx_nuc_drop_limit = *5**d-2*
> dx_nuc_drop_min_X_limit = *5**d-2*
> *_*
> I also tried slightly much higher values of dx_nuc_drop_limit (1d-1) and
> dx_nuc_drop_min_X_limit (1d-1) but this also did not work. I also played
> with various *varcontrol_target* values and various *min_timestep_limit*
> but without any success. Below, I attach the output in final stage just
> before the timestep failure. How should I proceed? Please suggest.
>
>
> STELLA problem is solved. Now I want to know, which file in STELLA exactly
> controls the explosion parameters such as total explosion energy, Nickel
> mass etc?
>
> Thank you very much in advance for any suggestions.
>
> Best Regards,
> Amar
>
> ------------------------------
> *From: *"Jared Goldberg" <goldberg at physics.ucsb.edu>
> *To: *"Amar Aryan" <amar at aries.res.in>
> *Cc: *"mesa-users" <mesa-users at lists.mesastar.org>, "Abhay" <
> abhay at aries.res.in>
> *Sent: *Saturday, February 15, 2020 2:38:06 AM
> *Subject: *Re: [Mesa-users] Trying to model a possible progenitor for
> type Ib Supernova
>
> Hi Amar,
>
>>
>> 1) At the end of evolution, the total star mass is around 6.5 solar
>> masses, so I am not sure if the star has evolved up to the stage of
>> core-collapse. What changes should I make to cut more mass?
>
> See inlist_remove in the example_make_pre_ccsn as an example of how to
> arbitrarily remove mass instantaneously. Alternatively, you could try a
> higher mass-loss rate, or a relaxation routine.
>
>
>> 2) I tried to generate input files for STELLA, so I made some changes in
>> the last inlist file (in my case it is inlist_to_core_collapse ) as
>> suggested in STELLA README file. But at the end of the evolution, the two
>> files (mesa.abn & mesa.hyd) generated are empty. When these files are
>> generated using the second last inlist (in my case inlist_to_si_burn), then
>> these work. So I used so generated mesa.abn and mesa.hyd files in STELLA
>> but with these file STELLA does not seem to converge, putting me in doubt
>> if the star reached the core-collapse stage.
>
>
> It looks like you did get the star to core collapse successfully, nice!
> Regarding STELLA's failure, how did you explode the star?
> With that information we can diagnose better why STELLA isn't working,
> since STELLA will only successfully run if it is given a model which has
> been exploded.
>
> The make_pre_ccsn test case just brings the star to Fe core infall
> ("collapse"), but does not excise the Fe core and does not simulate the
> explosion.
> Rather, something akin to the test_suite case example_ccsn_IIP should be
> done to actually explode the star before handing off to STELLA, using the
> final model produced at core-collapse as your initial model in
> inlist_infall of example_ccsn_IIP.
> Hope this helps. Let me know if you have any other questions!
>
> ~Jared
>
> On Wed, Feb 12, 2020 at 10:01 PM Amar Aryan <amar at aries.res.in> wrote:
>
>> Dear Jared,
>>
>> As per your suggestions, I tried methods prescribed in Rob Farmer's 2016
>> supernova progenitor models paper. I tried many reasonable combinations of
>> parameters in the inlist file. Most of the combinations failed to make the
>> model converge for a star having ZAMS mass of 12 solar mass but there were
>> a few cases when the model converged. I mention them below :
>>
>> I made the following changes in &star_job section
>>
>> ! required_termination_code_string = 'max_model_number' (commented this
>> line)
>> steps_to_take_before_terminate = *240* (uncommented this, *earlier it
>> was* *2*)
>>
>> In the &controls section i commented the follwing lines
>> ! max_number_backups = 1
>> ! max_number_retries = 3
>> !max_model_number = 99999999
>> in &control section section.
>>
>> With the above-mentioned changes, the *model converged!!*  with
>> termination code: fe_core_infall_limit
>> Here I attach a few outputs of  last stages of evolution
>>
>>
>> Here we can clearly see very strong impact of envelope burning in the
>> later phases
>> Now I have a few doubts :
>>
>> 1) At the end of evolution, the total star mass is around 6.5 solar
>> masses, so I am not sure if the star has evolved up to the stage of
>> core-collapse. What changes should I make to cut more mass?
>>
>> 2) I tried to generate input files for STELLA, so I made some changes in
>> the last inlist file (in my case it is *inlist_to_core_collapse *) as
>> suggested in STELLA README file. But at the end of the evolution, the two
>> files (*mesa.abn* & *mesa.hyd*) generated are empty. When these files
>> are generated using the second last inlist (in my case
>> *inlist_to_si_burn*), then these work. So I used so generated mesa.abn
>> and mesa.hyd files in STELLA but with these file STELLA does not seem to
>> converge, putting me in doubt if the star reached the core-collapse stage.
>>
>> Thanks in advance for any suggestions.
>>
>> Amar
>>
>>
>> ------------------------------
>> *From: *"Jared Goldberg" <goldberg at physics.ucsb.edu>
>> *To: *"Amar Aryan" <amar at aries.res.in>
>> *Cc: *"mesa-users" <mesa-users at lists.mesastar.org>, "Abhay" <
>> abhay at aries.res.in>
>> *Sent: *Wednesday, January 22, 2020 2:41:12 AM
>> *Subject: *Re: [Mesa-users] Trying to model a possible progenitor for
>> type Ib Supernova
>>
>> Hi Amar,
>> First of all, I notice you have
>> steps_to_take_before_terminate = 2 ! if not removing anything
>> in the &star_job section of your inlist_remove, as such, the model is
>> only taking 2 timesteps when removing and not removing any hydrogen in the
>> working directory you sent.
>>
>> Commenting out
>> ! required_termination_code_string = 'max_model_number'
>> ! steps_to_take_before_terminate = 2 ! if not removing anything
>> in star_job and
>> ! max_number_backups = 1
>> ! max_number_retries = 3
>> in the controls section allows the model to strip more as it evolves,
>> although the core evolution continues to proceed.
>>
>> When running the inlists with these modifications, I find that the
>> evolution is failing at a similar evolutionary point while the model is
>> still stripping, during core oxygen burning, with complicated shell
>> burning. I find staring at pgstar plots near the crash helpful; see
>> especially the region between 2-2.5 Msun:
>>
>> [image: image.png]
>>
>>
>> [image: image.png]
>> Notice also the little blips in the velocity around 2-2.5 solar masses,
>> where some of that shell burning is taking place.
>>
>>
>> 12 Solar mass stars are challenging to model, with problems often arising
>> during complicated stages of shell burning. If you just want a working
>> model of a type Ib progenitor, an easy fix would be going to a higher
>> progenitor mass (15 - 18Msun, for example), which should make it through
>> these complicated stages of evolution with much less trouble.
>>
>> If you do want to get a 12Mzams progenitor model through this stage of
>> evolution, other solutions may exist, but they are less simple, and I can't
>> as easily point you to "the answer."
>> Turning off Thermohaline mixing (setting thermohaline_coeff = 0 in &
>> controls), for example, gets the model through core O burning. I did not
>> run the model beyond that, so similar issues may still appear in subsequent
>> shell burning episodes. There are great suggestions in Rob Farmer's 2016
>> supernova progenitor models paper (
>> https://ui.adsabs.harvard.edu/abs/2016ApJS..227...22F/abstract) and the
>> related inlists (https://zenodo.org/record/2641723) on mesastar.org.
>> Although the inlists are somewhat out of date for current revisions of
>> MESA, the concepts therein can be very helpful.
>>
>> Also, if you want to tighten the time resolution, perhaps try changing
>> the tolerances in the central abundances (e.g. delta_XNe_cntr_limit,
>> delta_XO_cntr_limit), temperature (delta_lgT_cntr_limit), and density
>> (delta_lgRho_cntr_limit), rather than just changing varcontrol, since these
>> seem to be what set the timesteps in the first place (when looking at the
>> terminal output for dt_limit).
>>
>> Hope this is helpful, and definitely let us know if you have any more
>> questions or if any other issues arise!
>>
>> Cheers,
>> ~Jared
>>
>> On Tue, Jan 21, 2020 at 8:36 AM Amar Aryan via Mesa-users <
>> mesa-users at lists.mesastar.org> wrote:
>>
>>> Hi,
>>> Using mesa version-11701, I am trying to model a possible progenitor for
>>> a type Ib supernova (SN). For this purpose, I am evolving a 12 solar mass
>>> zero-age main-sequence star, considering initial metallicity (Z) of 0.02.
>>> After evolving the model until the exhaustion of helium, I am imposing an
>>> artificial mass-loss rate of 10^(-4) solar masses per year until the total
>>> hydrogen mass of the star goes down to 0.01 solar masses. After the
>>> hydrogen mass reaches the specified limit, I switch off the artificial mass
>>> loss and evolve the model until the onset of core-collapse. For this
>>> purpose, I have used *example_make_pre_ccsn* directory of test_suit
>>> with some changes in the inlist files and in *rn* executable. I am
>>> attaching here my entire working directory.  The evolution starts smoothly
>>> but terminates by giving the following errors :
>>>
>>>
>>>                                                                    dt
>>> 3.1322685072350174D-13
>>>
>>>                                         min_timestep_limit
>>> 9.9999999999999998D-13
>>>
>>>                                                  stopping because of
>>> problems dt < min_timestep_limit
>>>
>>>
>>> I have tried a few solutions myself :
>>>
>>>     a) Varied* min_timestep_limit* in inlist_massive_default in the
>>> range (10^-6 to 10^-14), but these did not work, still got same error.
>>>     b) Then tried with changing *varcontrol_target* in the range (10^-2
>>> to 10^-5) with different min_timestep_limit, but still, we get the same
>>> error.
>>>
>>> I am looking for your guidance and suggestions.
>>> Thank you in advance for any suggestions.
>>>
>>> Regards,
>>> Amar.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> 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/20200324/51f0fa3f/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: grid_cc_012220.png
Type: image/png
Size: 29455 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200324/51f0fa3f/attachment.png>

More information about the Mesa-users mailing list