[Mesa-users] problem of ccsn_llp

[Jared Goldberg]

Hi Kai-an,

Thanks for sending the inlists and models. I've copied your inlist_edep and
ready_for_edep.mod into a clean version of the test suite ccsn_IIp, and
I've reproduced the error.

Here's the short answer:
This is a model that, for whatever reason, is hard to explode. Changing the
infall radius and bomb properties can help as a "stellar engineering"
solution.

Change
*center_R_lower_limit = 5d+7 ! instead of 4d+07*
in inlist_infall, and change



*  max_age_in_seconds = 1 ! formerly 2.5
total_mass_for_inject_extra_ergs_sec = 0.1 ! formerly 0.6   max_timestep =
1d-3*
in inlist_edep
and the explosion will proceed smoothly for this model. Feel free to play
around with these values further, these are just some of the values that
worked for me.


Here's the longer explanation:

There are 2 things that are weird when running with the ready_for_edep, and
I think you have to address both in order to fix the issue. First, notice
the odd dip in the density and blocky spike in the entropy at the beginning
of shock_part1 around zone ~1000. Also look at the abundance spikes in a
single zone around 1050. That region is where MESA is choking.
Another thing that's odd about this run is that when we start inlist_edep,
the total energy of the star is already positive (as a result of that spike
in entropy/internal energy) -- so this problem is "seeded" earlier than
when you put the bomb in, likely in inlist_infall or inlist_end_infall.

[image: image.png]

Notice also that the following warning/error message pops up almost
immediately, and then persists until the run crashes --
 retry: max correction jumped -- give up in solver       6
 retry: max correction jumped -- give up in solver       6
 retry: max correction jumped -- give up in solver       6
 retry: max correction jumped -- give up in solver       6

This indicates the solver is running into problems when checking
tolerances. Perhaps a developer might have a better explanation of the
specific error message, but the effect is that MESA continues to shrink the
timestep in an attempt to find a solution and take a timestep -- which it
shrinks below the minimum allowable timestep, thus resulting in the
termination of the run with the failure message you describe.
We can't be too surprised, as we've inserted a whopping amount of energy in
a narrow region where there might be large composition and
structural gradients, and in a part of the star where we just chopped out
the core, made everything fall in, and then halted that infall at a
somewhat arbitrary radius.


I reran with your final.mod and changed the infall radius to 500 km rather
than 400 km, since sometimes the infall crashing down on and halting at a
deeper inner boundary leads to numerical troubles. This does fix the odd
compositions/properties near the center, but it does not lead immediately
to a successful explosion.

Because the "thermal bomb" technique of exploding a star is already a feat
of stellar engineering and not a first-principles modeling method, and
because results of the supernova explosion and resulting lightcurves are
unaffected by the exact choice of bomb width, duration, etc. (mentioned in
the 4th MESA instrument paper -- but you can validate  this claim more
rigorously if you want to convince yourself), I think it's fine to adjust
the infall radius, bomb width, duration (max age), and even the timestep in
order to get a model to explode. See the upper left panel of Laura Lopez's
XKCD parody of supernova papers ;) :
https://twitter.com/ohdearz/status/1390068388974768128

With the infall radius as 500km, I'm able to get your model to explode more
easily. Here is one set of inlist_edep parameters that works:
  max_age_in_seconds = 1 ! formerly 2.5
  total_mass_for_inject_extra_ergs_sec = 0.1 ! formerly 0.6.
  max_timestep = 1d-3

Here's a new pgstar sceenshot of the early stages of shock_part1 -- things
look much more "normal" (no density drop/temperature spike/entropy spike,
reasonable elemental/proton abundance profiles near the center). AND the
total energy was negative (the star was bound) before we explode it in
inlist_edep.

[image: image.png]

Cheers,
~Jared

On Mon, May 17, 2021 at 5:42 PM 游凱安 via Mesa-users <
mesa-users at lists.mesastar.org> wrote:

> Dear Jared,
>
> below are final.mod of 20m from make_pre_ccsn, part_edep, part_1,
> run_star_extras, ready_for_edep.mod
>
> I have tried various explosion energies: 1.2 foe, 2.4 foe, 3.6 foe, 4.8foe
> but all have failed.
>
> If you need more information, please let me know.
>
> or if you need more information about 13m of make_pre_ccsn, please let me
> know.
>
> Thanks for your help and patience.
>
> ~Kai-An
>
> _______________________________________________
> 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/20210518/44cd0d76/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 60531 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210518/44cd0d76/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 59635 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210518/44cd0d76/attachment-0001.png>