[Mesa-users] EOS/opacity (?) issues in proto-WDs

Josiah Schwab jwschwab at ucsc.edu
Fri Apr 19 23:29:01 EDT 2019

Hi Patrick,

I finally had a chance to look at this thread.  You're correct in
identifying this as an EOS issue.  Thank you for diving in and sending
models / inlists with helpful diagnostic plots prebuilt.

I think of the MESA EOS like a patchwork quilt, but in 4 dimensions (and
infested by hyperspatial moths).  When looking into EOS issues, it is handy
to know that Z = 0.04 is a "seam", because this is the highest Z provided
by OPAL.

Knowing this, the first thing I did was change your pgstar profiles to have
Z on the x-axis and there immediately pops up a discontinuity in the
entropy at Z = 0.04.  And so then I went hunting and found a bug the in
Z-blend in the PT eos.  That fixed the initial problem, but then the model
gets in similar trouble somewhat later once the Z-blend reaches the surface
(at mass 0.17 Msun, so still short of your goal).

A general piece of advice is that the default DT form of the EOS gets more
attention than the PT version, so if you can run these models without the
lnPgas form, I'd encourage that.  Long ago, there used to be a use-case for
lnPgas in getting smoother pressure profiles for asteroseismology, but I'm
not sure I know of a strong motivation to use it in current MESA versions.

The default DT form also has a Z blend, but in current versions it blends
into PTEH, which is usually much better match than HELM (b/c PTEH includes
H & He ionization).  See Appendix A of MESA V for some of these acronyms
and a little more detail.

I've been working on the EOS over the last few weeks, so I'd appreciate if
you could try your models with the development version

   svn co https://subversion.assembla.com/svn/mesa^mesa/trunk mesa-svn

Note this will require the latest release of the MESA SDK.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20190419/f1d81ee0/attachment.html>

More information about the Mesa-users mailing list