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

[Patrick Neunteufel]

Hi Josiah,

Thanks for all the work!

I will try and have a look at this asap (probably tomorrow). 

> the coolest outer layers at Z ~= 0.04 don't go to PTEH, but flip to ELM/HELM.  That abrupt EOS switch seems to be why models want to stop when the surface metalicity nears 0.04.

This makes a lot of sense. I had a suspicion that this might be the problem. I am glad that I wasn’t completely wrong. 

In fact, I ran a few other models in the meantime that do not achieve the high metallicities seen in the contentious models and they did seem to converge without much of a problem (they’re also a little boring, unfortunately).

I’ll get back to you with the results of my own experiments.

I’d also be happy to assist in fixing this issue, if needed.

Have a pleasant weekend!

Best wishes

Patrick

> On 21. Apr 2019, at 00:57, Josiah Schwab <jwschwab at ucsc.edu> wrote:
> 
> Hi Patrick,
> 
> After some more experimentation, here are my current thoughts.
> 
> We're having some struggles with the high-Z (Z > 0.02) blend of OPAL & SCVH in the DT2 eos.  Probably best to fall back to the older DT version.  (The difference is the way the interpolation is done; they're the same OPAL / SCVH data underneath.)  So set
> 
>     use_eosDT2 = .false
> 
> in controls.  It is possible that you may no longer be able to satisfy the tight "gold" tolerances.  If not, you may have to set use_gold_tolerances = .false. .
> 
> When we reach Z near 0.04, we change between OPAL/SCVH (or DT2) and PTEH (controlled by Z_for_all_PTEH = 0.040d0 and Z_for_any_PTEH = 0.039d0 in star_job).  These EOSes have different limits in logQ (= logRho - 2 * logT + 12).  By default we have
> 
>       logQ_max_for_low_Z_PTEH_or_DT2 = 4d0
>       logQ_max_for_high_Z_PTEH_or_DT2 = -0.4d0
> 
> so the coolest outer layers at Z ~= 0.04 don't go to PTEH, but flip to ELM/HELM.  That abrupt EOS switch seems to be why models want to stop when the surface metalicity nears 0.04.
> 
> If you're willing to push PTEH a beyond its limits, you can make these cutoffs the same, i.e., set
> 
>       logQ_max_for_high_Z_PTEH_or_DT2 = 4d0
> 
> and prevent this transition.  We just don't have a good high Z EOS down at this Rho,T so it seems like picking whatever makes things smooth is reasonable.  Doing that, I got your model down to about M = 0.1 Msun.  Then it seemed to get into trouble again, which I don't have time to investigate at the moment.  Presumably also the EOS.
> 
> Josiah
> <grid1003100.png>