[Mesa-users] mix_type not set to rotation even if Eddington-Sweet diffusion coefficient is the largest

[Mathieu Renzo]

Dear all,

I have encountered a situation where looking naively at Kippenhahn plots 
might be misleading because rotational mixing is not shown even if 
dominant.

I have a stellar model where thermohaline mixing and rotational mixing 
are both active at the same time in some cells. Looking at the diffusion 
coefficients (e.g., from the Mixing plot in pgplots) I see that the 
dominant diffusion coefficient in these zones is the one for 
Eddington-Sweet circulations, so I would expect mix_type = 5 
(rotation_mixing as per $MESA_DIR/const/def/const_def.f90). However, 
since thermohaline mixing is also active (although subdominant), the 
mix_type is set to 4 (thermohaline).

This seems to be caused by the second `if` in the snippet of code below 
(from $MESA_DIR/star/private/mix_info.f90 - lines 342+ in r15140)

```

             do k = 2, nz
                if (s% D_mix(k) < 1d-10) s% D_mix(k) = 0d0
                s% cdc(k) = s% D_mix(k)*cdc_factor(k)
                if (s% D_mix(k) /= 0 .and. s% mixing_type(k) == 
no_mixing) then
                   s% mixing_type(k) = rotation_mixing
                end if
             end do
```

I have no reason to believe the calculation itself is wrong, but the 
choice of setting mix_type = 5 only if there is no other mixing active 
will effectively remove rotational mixing from the Kippenhahn plots 
(both in pgstar and with tools that use the mix_type to make the plot) 
even if it dominates.

I was wondering why this choice of setting mixing type to rotational 
mixing if and only if no other mixing is active although the rotational 
mixing diffusion coefficient might dominate over other diffusion 
coefficients.

Also, if someone has a workaround to show the rotational mixing in 
Kippenhahn diagrams without using mix_types, that would also be 
appreciated -- but I doubt this is easy to do without re-running the 
models and saving many profiles.

Bests,
Mathieu

-- 

Mathieu Renzo <https://users.flatironinstitute.org/~mrenzo/>
Flatiron Research Fellow
Center for Computational Astrophysics
Columbia University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210212/bb10e25b/attachment.htm>