Re: [mesa-users] noisy Brunt near convective boundary

[Bill P. <pa...@ki...>]

Hi Santiago,

Thanks for sending your inlists!  

On Mar 3, 2015, at 5:38 AM, Santiago Andres Triana wrote:

> Hello mesa-users,
> 
> I'm trying to get the proper softening + mesh resolution settings to
> smooth out things near the bottom convective boundary in a
> shell-burning RGB star of about 1.2 M_sun. A symptom of the problem is
> reflected on the noisy Brunt-Vaisala frequency (figure attached). This
> not good for calculating pulsation frequencies through e.g. gyre,
> there is a bit too much variation from time step to time step.
> 
> I'm including softening and additional resolution near the convective
> boundary by settting

Are you sure you only want extra resolution for non-burning convective zones?  That's what the "nb" means.
For hydrogen burning convective zones you need to use "hb" instead (or in addition to).
> 
>      xtra_coef_czb_full_on = 1
>      xtra_coef_czb_full_off = 1
> 
>      xtra_dist_a_l_nb_czb = 2
>      xtra_coef_a_l_nb_czb = 1d-1
>      xtra_dist_b_l_nb_czb = 2
>      xtra_coef_b_l_nb_czb = 1d-1

recall that the"softening" isn't about resolution -- it is modifying diffusion coefficients.
> 
>      conv_bdy_mix_softening_f  = 0.001
>      conv_bdy_mix_softening_f0 = 0.003
>      conv_bdy_mix_softening_min_D_mix = 0

Here's something new to consider instead of trying to clean up the brunt by smoothing.
Turn off the smoothing
	num_cells_for_smooth_gradL_composition_term = 0 
	num_cells_for_smooth_brunt_B = 0 
and turn on a small background level of mixing such as might be the result of even small rotation
	set_min_D_mix = .true.
	min_D_mix = 1d3
that should remove the tiny cell-to-cell variations in composition that make the brunt noisy.
just for comparison, recall that D in convective regions will be 10 orders of magnitude larger than this.
but you should still experiment with this --- don't just use it!  try smaller values of min_D_mix.
check to see what else it might be changing.   these are of course standard warnings!

Another trick-of-the-trade I've used with some success is to have the code remove tiny gaps in convective regions.
You can try this by setting these
      min_convective_gap = 0.01
      min_semiconvection_gap = 0.01

As you can tell, this is an area where we as a community are still learning!  Glad to have your help. ;D

Cheers,
Bill


> 
> I attach figures showing the mixing type, composition, gradr_sub_grada
> all as function of radius around the zone in question. I'm using the
> astero module, inlists also attached.
> 
> I have tried already a number of things like small overshooting,
> semiconvection and diffusion but the results were not much better.
> 
> Any advice? I have the feeling that I'm not using the settings
> properly... Many thanks in advance for any help!
> 
> Santiago
> <brunt_N2.png><mixing_type.png><composition.png><gradr_sub_grada.png><inlist_astero><inlist_astero_search_controls>------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the 
> conversation now. http://goparallel.sourceforge.net/_______________________________________________
> mesa-users mailing list
> mes...@li...
> https://lists.sourceforge.net/lists/listinfo/mesa-users