[Mesa-users] Question to an unusual behaviour at the end of hydrogen burning in massive star

[Francis Timmes]

> Interestingly, if I increase the f-parameter from the exponential decaying diffusion model for the convective boundary mixing and approximately double the spatial resolution, the strange behaviour disappears.

so you've found a solution to the sudden mixing event. super.
the question then is why increasing f and/or mass resolution alleviates it.

i notice the inlist has f = f0.
carefully read the paragraph above 
http://mesa.sourceforge.net/controls_defaults.html#overshoot_alpha
that begins "Overshooting works by taking ...".
what do you conclude about how much overshoot mixing occurs when f = f0 and when f > f0?

fxt






> On Nov 30, 2018, at 4:24 AM, Etienne Kaiser via Mesa-users <mesa-users at lists.mesastar.org> wrote:
> 
> Dear MESA community
> 
> I'm trying to calculate a 25 Msun star at pre-solar metalicity using MESA revision 10108 and I stumbled over a rather unusual behaviour of the model at the end of hydrogen burning.
> 
> At a point at the end of hydrogen burning, i.e. when the central hydrogen mass fraction is <1d-4, there is a sudden increase of central hydrogen to about 0.1 % in mass fraction within one step and the burning & convection revive. Moreover, the code drastically decreases the timesteps. I attached a Kippenhahn diagram (reddish are the burning regions and blueish the convective regions), a Herzsprung-Russel (HRD) and a Tc-Rhoc plot, showing the unusual region. After this unusual behaviour, the model calculates fine.
> 
> I tried different things to avoid this unusual behaviour, such as various increased spatial and temporal resolution, including the Type2 opacity tables, MLT++ or use a different reagion network (approx21 -> mesa206) but nothing of these seems to help. I tried to use the hydro debugging tool as well but didn't get a good clue.
> Interestingly, if I increase the f-parameter from the exponential decaying diffusion model for the convective boundary mixing and approximately double the spatial resolution, the strange behaviour disappears. It also disappears when going to smaller masses. I attached the inlist as well.
> 
> I wonder if anybody came across a similar problem and has found a solution. I unsuccessfully tried to find something in the mailing archives.
> 
> Thanks in advance.
> Cheers,
> Etienne
> <Kippenhahn_25M.png><TcRhoc_25M.png><HRD_25M.png><inlist_project>_______________________________________________
> mesa-users at lists.mesastar.org
> https://lists.mesastar.org/mailman/listinfo/mesa-users
>