[Mesa-users] custom rates

[Josiah Schwab]

Hi Amber,

> ​For weak rates, I see that according to file in the method is similar
> to the combination of rate_list.txt and rate files, but the weak rate
> files are expected to be in hdf5(hdfs?) rather than .txt format?

When you're in the regime that the weak rates are non-trivially density
and temperature dependent (see Zach's email from earlier in the week),
you need to provide several 1D & 2D tables per reaction.  To save the
user the hassle of formatting a text file, we use hdf5 for this.  The
custom_rates test suite case has an example python script that shows how
to populate an appropriate hdf5 file.

> ​Also for weak rates, I see that there is also the file ​weak_info.list
> which contains slightly different info, the half life and the quantity Qneu
> (the neutrino energy?). Can someone talk about the difference between that
> file (other than the obvious) and the weak rates tables (similar in format
> to weakreactions.tables)?

weak_info.list is primarily used to set the rate at low temperature (see
rates_initialize.f90, and also Zach's email).

If you are doing work that requires low-T weak rates, it is worth being
a little careful here and double checking that the values for your
reactions of interest are what you think they should be.  The transition
between the tables and low-T (REACLIB/terrestrial) rates is something
that could use some restructuring; it is often confusing.  Physically,
this is a bit of a tough spot because we're going from the tables (which
assume complete ionization) and moving to rates for neutral-ish atoms.

Most problems can safely ignore this sort of thing, but for an example
of a case where one needs to be a little more careful, see

  Shen, K. J. and Schwab, J. (2017)
  Wait for It: Post-supernova Winds Driven by Delayed Radioactive Decays
  http://adsabs.harvard.edu/abs/2017ApJ...834..180S

where Section 3 explains some of the rates gymnastics we undertook.

> Also, the info in weakreactions.tables appears to conform to the same
> data set as described in weak_rate_list.txt, but is not, to my
> knowledge, in hdf5 format.

Yes, when I added the ability to specify custom weak rates I left the
existing master weakreactions.tables in its plain text format.  There
was no reason to change something that was working.

New reaction rates (i.e. the Suzuki 2016 rates) that have been added
since are in the hdf5 format.

Hope that helps,
Josiah