[Mesa-users] Hydrogen burning with no hydrogen?

[Nicholas Rui]

Hi everyone,

I've been experimenting a bit with ways to artificially construct stellar
merger remnants in MESA. However, I have noticed a strange issue in some
models where a significant amount of hydrogen burning can happen even if
there is no hydrogen (down to ~machine precision?).

I have attached two zip files with two illustrative runs. I am using
version r24.08.1.

>> run 1: pure helium envelope (directory: "1_helium_envelope")

I initialize a pre-main-sequence model made entirely of helium (no H). I
then relax the initial mass and radius to some nonzero values (in this
case, to "match" an oxygen-neon WD model). To get the model into thermal
equilibrium, I have used dxdt_nuc_factor = 0 (preventing abundances from
changing due to burning) while evolving for a long time. In addition to the
inlist, I have also attached a pgstar plot.

Bizarrely, even though the model should be 100% helium by construction, the
envelope is actually supported by hydrogen burning (there is no helium
burning). Because dxdt_nuc_factor=0, I don't think this can be because
protons are being produced by some nuclear reaction somehow.

I also suspected that maybe the machine-precision levels of hydrogen (X ~
1e-99) are able to burn significantly at high temperatures due to the
strong temperature dependence, but that sounds difficult to believe and I
don't know how to check it for sure. I couldn't find any option to disable
energy release due to burning at sufficiently low mass fractions either, so
I'm not sure how to practically deal with this.

>> run 2: RG+HeWD "merger remnant" (directory: "2_rg_hewd_merger")

I am attempting to create a model qualitatively similar to the "Model C2"
red giant--He WD merger in Zhang & Jeffery 2013. In this model, a cooled He
WD "subducts" to the center of the "warm" helium core of a red giant, and
the resulting thing quickly ignites helium.

However, I had some trouble accreting helium onto the WD model using
mass_change, as was done in the original paper. I also wasn't sure if
relax_mass would do the desired thing of adding warm degenerate helium onto
the surface of the WD without modifying the pre-existing entropy
profile---I could only find a description in the documentation for what
this does when relax_mass is used to decrease the mass of the star
(although if it does, I think this would be acceptable for my purposes).
The difficulty with using mass_change in newer versions for applications
has been discussed on this list in the past:
https://lists.mesastar.org/pipermail/mesa-users/2015-November/005915.html

Instead, I did a very roundabout procedure of generating a "warm"
degenerate non-burning shell of helium by creating a pre-MS helium star,
disabling burning, and relaxing M_center and R_center. I then manually in
python "stitched" the He WD model to this helium shell. I added a hydrogen
envelope in the same way. The resulting star is "rung down" in the
hydrodynamical mode without burning until the velocities are small.

I then try to turn hydro mode off and burning back on. The MESA run somehow
loads the initial model and starts, although the timestep becomes very
small and eventually crashes (even with convergence_ignore_equL_residuals
enabled). "Full physics" is on in this case (I don't set
dxdt_nuc_factor=0). Given the pathological way I made this model, I'm not
really surprised it doesn't run for long.

However, again, the pgstar snapshot shows that, in addition to the expected
amount of hydrogen shell burning, there is also a significant amount of
both pp- and cno-burning in the helium core itself, with reaction rates
exceeding even that of triple alpha. The core has solar metallicity but
negligible hydrogen, which I confirmed by plotting the hydrogen fraction
(plot included). I don't know what's going on.

So... does anyone have any idea what's going on?

Best wishes,
Nicholas Rui
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20250725/cd60d62d/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1_helium_envelope.zip
Type: application/zip
Size: 11213 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20250725/cd60d62d/attachment.zip>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 2_rg_hewd_merger.zip
Type: application/zip
Size: 169248 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20250725/cd60d62d/attachment-0001.zip>