[mesa-users] Mass Transfer + Rotation Problems

[Bill Paxton]

Hi Pablo,

I notice that the model you sent is already in trouble even before the convergence fails.
There is a pressure inversion in the outermost 2 grid points.   The 1st couple of steps
converge in spite of this, but then remesh adds a bunch of points at the surface and
that seems to be enough to push things over the edge.  The many retries and small
timesteps are an attempt to deal with the pressure inversion (I think).  

It may well be a problem in the mass accretion that has led to the P inversion, but
we'll need to go back to an early model to see it happen.

Cole, as a work around, turn off velocities.  That seems to be all it takes.
Let me know it that works for you.

Cheers,
Bill






On Mar 3, 2013, at 8:05 AM, Pablo Marchant wrote:

> Ok, saved a model of the accretor for the binary run, just before mass transfer, to experiment with it with single star mesa.
> 
> Attached is the model star with the inlist I'm using to test it. For mass_change = 1d-5 the model has problems converging and ends up with very small timesteps right away. Slightly smaller mass_change can cause the model to behave at the first timesteps, but end up in the same situation.
> 
> Just to note, this also seems to happen with mass loss, though things start getting messy at around mass_change = -1d-4.
> 
> The value 1d-5 is actually around the same value that the implicit scheme attempts before ending up with tiny timesteps, and it's not very high, I would actually expect it to rise significantly more than that.
> 
> I'm certainly interested in digging into the mass adjustment routines and how it deals with angular momentum, as I'm starting to check how to implement the missing features regarding rotation into binary MESA.
> 
> Cheers!
> 
> 
> On Fri, Mar 1, 2013 at 6:38 PM, Bill Paxton <paxton at kitp.ucsb.edu> wrote:
> Hi Pablo,
> 
> On Mar 1, 2013, at 5:53 AM, Pablo Marchant wrote:
> 
> > Will try to see if I can get something similar experimenting with single rotating stars to check if this is effectively not a problem with the rlo routines.
> 
> Thanks for looking into this.
> If it turns out not to depend on the rlo routines, feel free to punt it back to me.
> But if you'd like to dig into the details of the interactions of rotation and
> mass transfer, I'd be delighted.  ;-)
> 
> It is quite possible that this is the first case of accreting onto a rotating model.
> The case of mass loss with rotation has been run alot, but not mass gain.
> 
> Cheers,
> Bill
> 
> 
> 
> 
> -- 
> Pablo Marchant Campos
> M.Sc on Astrophysics, Universidad Católica de Chile
> PhD student, Argelander-Institut für Astronomie
> <accretor_rotating.dat><inlist>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20130303/466527f8/attachment.html>