[mesa-users] helium main sequence binary

[Josiah Schwab]

Hello Ying,

On 12 September 2015 at 03:37 PDT, Ying Qin wrote:

Thank you for sending the additional files.  Now I can reproduce the
results you mention.

> I know WR star lost mass very fast, these results are not what I expect now.
> Why does WR star decrease mass from 3.9637246692135668E+000 ? I think the mass of WR should be decreasing from about 10.0.
> Why does binary_separation change from 9.6975446700026726E+000 ? Because I initialize this it with 0.2d0.
> For these results, I said there were random, Because they are not reasonable.

Before the main evolution begins, there was a relaxation stage to
satisfy the condition that the rotation of the star be sychronized with
the orbit (since you specified do_initial_orbit_sync_1 = .true.).

During this part --- which is printed to your terminal along with some
relax_to_omega messages, but is not included in the binary_history.data
file --- the WR wind you specified was active.  So by the time the
system gets synchronized and therefore ready for the main evolution, the
star has lost mass and the orbit has adjusted.

I'm surprised (as are you) that MESA isn't deactivating mass loss during
this relaxation phase.  Hopefully Pablo or one of the binary experts can
chime in.

I suggest the following work-around.  Make a tidally synchronized model
by running with the WR wind off.  Immediately save this model at
timestep 1.  Then load this saved model and run with the WR wind on.


> Also, I want star1 to evolve with zero age and zero model number, and
> I change these in inlist1:
>
>       set_initial_age = .true.
>       initial_age = 0 ! in years
>
>       set_initial_model_number = .true.
>       initial_model_number = 0
>
> But from the results, we can clearly find it did not  work.

These controls are mainly used when you are loading a saved model file,
which has an associated age and model number.  These controls are not
really relevant since your model is coming from one of the ZAMS files.

Still, why is it clear this "didn't work"?

When I look in binary_history.data, the first model number is 1 and the
first star_age is the duration of the first timestep.  That is what I
would expect to see using these flags (even if it is not really due to
these flags in this case).

Hope that helps,
Josiah