[Mesa-users] Problem with r12778. Bug?

[Ian Foley]

Hi Josiah, and Biil,

Applied your suggested mod and it works beautifully. I can now use gold
tolerances and get good performance
and better results too.

Thanks!
Ian


On Wed, 13 May 2020 at 10:20, Josiah Schwab <jwschwab at ucsc.edu> wrote:

> Hi Ian,
>
>
>> you have found a good work around for the problem until the next release
>> when this bug will be fixed.
>> as you noticed, the treatment in basic.net avoids the problem, so it is
>> only larger nets that will hit it when dealing with low temperatures.
>>
>
> To add a bit more, we noticed the rate chosen for li7 burning with a
> "soft" net (where your net file only specifies isotopes) isn't behaving
> correctly at low T.  (You might notice when you look closely the li7
> profile looks "funny" in your model.)  Nets like pp_and_cno_extras.net
> use a different rate, which is probably why this had gone unnoticed.
>
> As a workaround, you can swap out the rate for a better behaved version.
> Aaron Dotter suggested doing
>
> remove_reaction(r_h1_li7_to_he4_he4)
> add_reaction(r_li7_pa_he4)
>
> and when I added these lines to ian30r.net (after the
> add_isos_and_reactions part) I no longer saw the issue.
>
> I think the spikes in eps_nuc were associated with zones reaching logT > 7
> and then having their li7 suddenly burn.  (Physically it should have burned
> at lower T).
>
> Thanks for the bug report!
>
> Josiah
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200513/ef375a76/attachment.htm>