[Mesa-users] mesh grid controls for radial comparison in two separately evolved solar models.

[Earl Bellinger]

Dear Josh,

I am not sure I fully understand your difficulties, but I can try to help.

In my view, the important part of calibrating a solar model is not that the
mesh on which the evolutionary calculations take place has some
particularly fine spatial resolution. Normally one can re-mesh
(interpolate) a solar model after the fact without much trouble. The usual
guideline is that one should choose a resolution such that the results do
not change when going to an even finer resolution, and this may even
already be achieved with the default settings. If you want additional
resolution at a certain location within the model (e.g. at the base of the
convection zone for your overshooting models), you can also adjust some
controls to zoom in on that spot.

Rather, the important part is rather that the total solar radius and
luminosity at the solar age have a certain precision, as these quantities
are extremely well-determined. For more details, see for example JCD's
recent Living Review on Solar Structure and Evolution:
https://arxiv.org/abs/2007.06488

Hence to calibrate a solar model one should change the initial Y and
alpha_MLT such that the surface R=L=1 (within 10^-5 as you say, or perhaps
even better) at the solar age (and also the initial Z if diffusion is
included). And indeed, this can be readily achieved in MESA.

Does this help?

Best wishes,
Earl

*Earl Patrick Bellinger *
Postdoctoral Research Fellow
Max Planck Institute for Astrophysics
Stellar Astrophysics Centre
https://earlbellinger.com


On Wed, Jan 12, 2022 at 3:48 PM Clarke, Josh via Mesa-users <
mesa-users at lists.mesastar.org> wrote:

> Hello MESA users,
>
>
>
> I have come across a bit of a conundrum. I am currently trying to
> replicate a test that a collaborator of mine used with their own code.
>
> I am modifying the temperature gradients in overshooting regions,
> comparing it’s inclusion in a solar-like star of 1Msun to a model of the
> same star with no modification.
>
> Both stars are evolved from a zero age main-sequence model. The aim is to
> compare the profiles of these two models once they terminate at a set age,
> modifying the mixing_length_alpha of the modified model so that it reaches
> solar structure at the terminal age.
>
>
>
> The difficulty is that the changes in the structural parameters I expect
> to see are very small. In order for the linear interpolation to be accurate
> I have been advised to get the compared mesh points to have a relative
> radial difference of 1x10-5 Rsun. My concern is that since MESA is
> lagrangian in nature I will be unable to achieve this degree of relative
> accuracy. I was wondering if any of you smart people know anyway to control
> the mesh so that this kind of radial comparison is possible? Or if it’s
> just realistically unfeasible? Please feel free to correct me if I have
> made any horrendous assumptions.
>
>
>
> My main models have been run with mesh_delta_coeff = 1d-1 but I’ve also
> run models with 5d-1 and 1d0 to decrease simulation run time. mesh_min_dlnR
> = 1d-14.
>
> All other mesh grid controls are default. I am using Gold tolerances with:
>
> gold_tol_residual_norm1 = 1d-11
>
> gold_tol_max_residual1 = 1d-9
>
> gold_iter_for_resid_tol2 = 10
>
> gold_tol_residual_norm2 = 1d-10
>
> gold_tol_max_residual2 = 1d-8
>
> gold_iter_for_resid_tol3 = 15
>
> gold_tol_residual_norm3 = 1d-6
>
> gold_tol_max_residual3 = 1d-4
>
>
>
> For my 1d-1 mesh_delta_coeff model the average radial shift of a cell in
> the modified model compared to its nearest cell in the base model is
> 5.361718481731894e-05 rsun, while the relative radial shift is
> 2.2747013522397857e-04 rsun.
>
>
>
> I am using MacOS Catalina 0.15.7 (19H524) , MESA Star version 12115, and
> Mesa SDK x86_64-osx-macos-19.11.2.
>
>
>
> Cheers,
>
> Josh
>
>
>
> ____________________________________________________________________
>
> Joshua Clarke
>
> PhD Student
>
> Department of Physics & Astronomy, University of Exeter,
>
> Stocker Road, Exeter, EX4 4QL, UK
>
> ____________________________________________________________________
>
>
>
>
>
>
>
>
> _______________________________________________
> 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/20220112/4aeb1f5d/attachment.htm>