[mesa-users] Sukhbold and Woosley (2014): trying to reproduce results

[Tuguldur Sukhbold]

Dear Etienne,

I am the 1st author of that publication.

As long as you're evolving the same inlist to the same stage in evolution
(with the same setup) you should get the exact same result, assuming that I
did not mess up anything. Notice that even though you're running the same
version of the code (r4930) you're using a different version of OS X and
importantly a different version of the compiler. I calculated all of those
runs on OS X 10.8.4 with a SDK that was released 3 years before yours (it
employs gfortran 4.7.2 from 20120710). I never tested the sensitivity
between different versions of gfortran, but back then I did see appreciable
differences between ifort/gfortran. Rich doesn't carry those older SDKs on
his webpage, but he might be able to help you on that front if you really
want to get the exact same result.

I have just recalculated a 20Msun model from Mov (with f=0.025) and I get
following:
he-core: 6.9195
co-core: 4.7673
exactly the same values listed in Table 2. Your numbers are quite close to
this, and I suspect the ~0.1Msun variation is due to the differences in our
setups.

As for 20Msun of M, all overshoot_f* are not exactly zero but all were
equal to 0.001. This is a typo both in the text of section 4.2 and in Table
2 (f was varied between 0.001-0.025 not 0-0.025).  With f*=0.001 I get:
he-core: 5.8068
co-core: 3.4769
same as listed in Table 2.

With f*=0.0 I get:
he-core: 5.7881
co-core: 2.9659
very close to what you got, again with ~0.1Msun difference.

>From following dropbox link you will be able to download the exact inlists,
terminal logs and final profile snapshots at core Oxygen depletion for all
above 3 calculations:

https://www.dropbox.com/sh/gvyyhygqn2x7525/AAAhs5boaHjziJwLEHdC7cLGa?dl=0

I hope this clears your confusion. Thank you for carefully checking this!

Tuguldur




On Sun, Mar 6, 2016 at 7:17 AM, Etienne Kaiser <e.kaiser at unibas.ch> wrote:

> Dear mesa-users,
>
> for my masterthesis, where I want to see the impact of rotation on massive
> stars using the MESA code, I'm first trying to reproduce some published
> results before continuing forward but I end up with different core masses.
>
> The publication I'm using is from Sukhbold and Woosley (2014): The
> compactness of presupernova stellar cores. This paper can be found in the
> MESA community portal (
> http://mesastar.org/results/the-compactness-of-presupernova-stellar-cores
> ).
> There I'm trying to reproduce the 20M_sol models calculated with MESA, one
> with no overshoot f=0.0 (MESA,M model) and one with overshoot f=0.025
> (MESA,MOV model), as shown in the 'table 2' in this publication.
> I'm using the publicated inlist form the MESA community portal (
> http://mesastar.org/results/the-compactness-of-presupernova-stellar-cores/inlist-for-a-full-star)
> and only added the definition of the helium and carbon cores as defined in
> the publication in 'table 2':
> h1_boundary_limit =2d-1
> he4_boundary_limit = 2d-1
>
> When I'm running the inlist with the MESA revision 4930 I end up with
> different core masses than in the publication:
> f=0.0: he-core: 5.7801 c-core: 2.8611
> f=0.025: he-core: 6.8277 c-core: 4.6885
>
> I tried different things to get to the right core masses but did not
> succeed. Does anybody of you had a similar problem in the past or is there
> some obvious mistake I'm doing? For any hint I would be grateful.
>
> I'm using a MacBook Pro, OS X El Captain, the MESA revision 4930 and
> mesasdk-x86_64-osx-10.10-20150825
>
> Thanks in advance.
> All the best,
> Etienne
>
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20160306/477fb3c2/attachment.html>