[Mesa-users] numerical noise in Brunt-Vaisala -- eos issue

[Josiah Schwab]

Hi Alina,

First, thank you for the excellent message. This is a great example of what
a useful problem report looks like!

Does anyone have a suggestion on how to get rid of this type of noise ?
>

As you suspected, I think you're hitting an eos discontinuity,
specifically, the Z-switch between HELM and PC.  By default, that occurs at
PC_min_Z = 0.999d0, which matches the location of the jump in your
eos_sample.

Evan was also displeased by these sudden switches between HELM and PC as a
function of composition in WD interiors and led a successful campaign to
eliminate PC_min_Z in future MESA versions.

For now, I'd suggest putting that switch at Z = 0, so that MESA always uses
PC.

Please try

    set_eos_PC_parameters = .true.
    PC_min_Z = 0d0
    mass_fraction_limit_for_PC = 0d0 ! default of 1d-3 can also cause small
discontinuities

in your star_job inlist and see if that helps.

Let us know how it goes.

Josiah
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20200715/73e2517f/attachment.htm>