[Mesa-users] More resolution at core boundary ?

[Warrick Ball]

Hi Gio,

Since r12778, there's an extra hook called `mesh_delta_coeff_factor`, which decreases the usual `mesh_delta_coeff` by the factor you set in the routine.  You can enhance the resolution in the regions you're interested by decreasing `mesh_delta_coeff`.  About a month ago, I experimented with enhancing the resolution around convective boundaries by using a Lorentzian in `gradr-grada`.  That worked reasonably well, though it required a bit of experimentation with the parameters of my function.

(I don't think there's any fundamental reason why one has to use a smooth function for `other_mesh_delta_coeff` but it seemed natural to me.)

If you don't want to upgrade to r12778, you can try the mesh function hooks.  Recall that the more mesh points are placed where the *gradient* of the mesh function is steeper, so in this case I tried a smooth step-like function (I think I used tanh), with the step centred at or near `gradr-grada == 0`.  Though I found something that worked for my one very specific case, I struggled to find coefficients for the custom mesh function that enhanced the resolution at convective boundaries by a similar amount in stars with different parameters (e.g. age or mass).

Let me know if you'd like more detail or if this is enough to steer you in a helpful direction!

Cheers,
Warrick



___________

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

On Mon, 1 Feb 2021, mesa-users at lists.mesastar.org wrote:

> 
> Dear all,
> 
> I am computing MESA models of upper main-sequence stars and I want to locate and describe the interface between the convective core and the radiative envelope precisely. What is the best way to increase the number of mesh points locally at the
> interface?
> 
> I have MESA 11701 and 12115 installed. I have read through the default files, but this thread here has got me confused: https://sourceforge.net/p/mesa/mailman/message/34628504/ .
> 
> Here is an example of my work: with MESA 11701, I have computed a 17Msun model with default mesh settings: my model has 1064 shells and the core-envelope interface is between the mesh points at m=6.46 and m=6.54. I would like to resolve this 0.08
> Msun shell much better than this, with about 100 mesh points there. I don’t need more resolution elsewhere though, and ideally I would like to simply restart from my model file, simply rederive my variables on a new mesh and save that in a profile
> file.
> 
> Thank you for any help you can provide,
> Best.
> 
> --
> Gio.
> 
> 
>