[mesa-users] mesa release 3790 + a cool movie

[Bill Paxton]

Hi,

Release notes will be found below, but I believe in eating dessert first,
so, here's something neat from Josiah Schwab <jwschwab at berkeley.edu>

He used mesa/star plus tioga to make a movie from pdf's giving very high quality text and graphics.
Thank you, Josiah, for creating this and sharing it with us.  ;-)
I wish I'd had something like it when I was first learning about stellar evolution!!!!

-------------- next part --------------
A non-text attachment was scrubbed...
Name: lifetime_720p.mp4
Type: video/mp4
Size: 2574228 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20111212/62be057a/attachment.mp4>
-------------- next part --------------




Now onto the information about mesa version 3790.

There have been the usual small changes and bug fixes. 
For example, we now add a point at the true center, r=0,
for FGONG and OSC files.   However, most of my time has 
been spent working on producing a more realistic abundance profile
in newly nonconvective regions left behind by retreating convective boundaries.

For example, in a massive star with convective center burning, there will be periods when
the top of the central convection zone is moving out and other times when it is moving inward.
And at the same time that the boundary is moving, the composition of the zone is
changing because of the strong nuclear burning.  Both the position of the boundary and
the composition in the zone are changing smoothly in time.   But in the code, we take
finite jumps in time and leave small irregularities in abundances behind a retreating
convection boundary.  For most applications, these composition "wiggles" are unimportant,
but for asteroseismology, they are an issue since the brunt frequency is sensitive
to the slope of the abundances, and little wiggles in abundances will produce
BIG wiggles in the derivatives of the abundances.

One way to deal with this, perhaps the best way, is to turn on overshooting
and let it clean up the trailing mess.  (Ledoux + semiconvection can also help
in some cases.)  But if you'd like to study the effect of overshooting by varying
its strength, you'd like to be able to include the case where the strength is set
all the way down to zero -- which means you'd like to have reasonable results
for the case of plain-vanilla-Schwarzschild without overshooting.   And to do that,
you need to account for the sub-step changes in position and composition at
the retreating edge of a convection zone.   

The current release of mesa includes my current best attempt to deal with this.
At the start of each step, before rezoning, it notices regions that were convective
previously but will be nonconvective now.  It flags these so that they will be given
higher than normal resolution during the rezoning.   Then, after the normal step
has generated a new structure and composition in the old way, complete with
wiggles, the code revisits the newly nonconvective regions and smooths the 
abundances to reflect that fact that intermediate positions ceased to be convective
at intermediate times when the abundance was at intermediate values.

If all this sounds complicated and difficult to implement, you are right!  
Thanks goes to Anne Thoul and Pieter Degroote for encouraging me to keep
trying on this problem in spite of repeated failures.  Thanks also to 
Aldo Serenelli and Andrea Miglio who, along with Anne and Pieter, had
to put up with my irrational tantrums as I struggled with this frustrating task.

Here's an example that shows the change in the Brunt we get by doing
the new scheme.

9M, Z=0.02, stop when center H reaches 0.4
"pure" Schwarzschild, no overshooting.

relevant inlist controls:
      r_div_R_smooth_brunt_dlnY_dlnP = 0.001
         ! width in fraction of stellar radius of top-hat smoothing function
      nsmooth_brunt_dlnY_dlnP = 1 
         ! the number of times to do r_div_R_smooth_brunt_dlnY_dlnP


OLD (version 3709)   --- small wiggles in he4 profile produce ragged brunt frequencies.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: brunt_frequency.pdf
Type: application/pdf
Size: 27926 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20111212/62be057a/attachment.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: he4.pdf
Type: application/pdf
Size: 26777 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20111212/62be057a/attachment-0001.pdf>
-------------- next part --------------


NEW (version 3790)
relevant inlist controls:
	same as above, plus 
              smooth_convective_bdy = .true.                  
              convective_bdy_weight = 1
-------------- next part --------------
A non-text attachment was scrubbed...
Name: brunt_frequency.pdf
Type: application/pdf
Size: 28223 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20111212/62be057a/attachment-0002.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: he4.pdf
Type: application/pdf
Size: 27225 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20111212/62be057a/attachment-0003.pdf>
-------------- next part --------------



Finally, I continue to use Rich Townsend's mesasdk and encourage you to do the same unless you have
a special need to do things the hard way.    To reduce the size of an already too huge mesa download, the mesasdk is
not included -- get it directly from http://www.astro.wisc.edu/~townsend/static.php?ref=mesasdk#toc-Downloads
The latest version is dated Nov 11, 2011.

Good luck!

-Bill