[mesa-users] Scratchy surface metallicity profile with diffusion in 1.3Msun star
Warrick Ball
wball at astro.physik.uni-goettingen.de
Tue Nov 25 11:57:36 EST 2014
Okay, I'll get back to this in the morning, but it looks like
(a) T_mix_limit smooths out the behaviour in the expected fashion, and
(b) the jump of about 0.07 dex in [Fe/H] around 2.5 Gyr is a result of the
convection zone penetrating inwards past one of those peaks I found
before, caused by the diffusion velocity varying in ionization zones.
Watch this space!
W
On Tue, 25 Nov 2014, Bill Paxton wrote:
>
> On Nov 25, 2014, at 7:34 AM, Warrick Ball wrote:
>
> I'm mainly hoping to smooth out the outermost meshpoints so that the fluctuations are suppressed and
> the surface abundance matches the underlying layers.
>
>
> Hi Warrick,
>
> Your comment suggests that you want to have diffusion on for the effects deeper in the model and don't
> particularly care about the effects of diffusion near the surface. If so, you might just use some other trick to
> mix things near the surface. In some cases, you can use "T_mix_limit" for this, but it depends on having a
> near-surface convection zone that can simply be extended to the surface. That may not work for you. An
> alternative would be "radiation turbulence" -- here's the description from controls.defaults:
>
> !## radiation turbulence
>
> !### radiation_turbulence_coeff
>
> ! To counter depletion of h and metals in outer envelope of stars with M > 1.4 Msun.
> ! Morel, P., and Thevenin, F.,
> ! Atomic diffusion in stellar models of type earlier than G.,
> ! A&A, 390:611-620 (2002)
>
> ! D = radiation_turbulence-coef * 4*crad*T^4/(15*clight*opacity*rho^2)
>
> ! 1 is reasonable value for this coefficient
>
> radiation_turbulence_coeff = 0
>
> Give that a try --- if it isn't what you need, then we can go back to tweaking diffusion controls.
>
> Good luck,
> Bill
>
>
>
>
>
>
------------
Warrick Ball
Postdoc, Institut für Astrophysik Göttingen
wball at astro.physik.uni-goettingen.de
+49 (0) 551 39 5069
More information about the Mesa-users
mailing list