[Mesa-users] Fwd: other_momentum subroutine

[Francis Timmes]

one standard technique to investigate if an effect is numerical or perhaps physical
is to conduct a convergence study with respect to mass resolution and time resolution.
if the behavior changes significantly with improved mass and/or time resolution,
then its probably numerical.

you don't include your run_star_extras, so no one knows how you are coding other_momentum,
but you should carefully study how mesa sets up its momentum equation, paying  attention
to what quantities are defined at cell faces and which are defined at cell center.

another suggestion, since your changing the momentum equation, is to explore the 
velocity and radial coordinate of your models to see if they make sense relative 
to the other assumption of your model (e.g., mass loss, etc).

good luck!

fxt




> On Oct 20, 2017, at 4:07 AM, sheyda najafi via Mesa-users <mesa-users at lists.mesastar.org> wrote:
> 
> 
> ---------- Forwarded message ----------
> From: sheyda najafi <sh.najafi8590 at gmail.com>
> Date: Fri, Oct 20, 2017 at 1:57 PM
> Subject: other_momentum subroutine
> To: mesa-users at lists.sourceforge.net
> 
> 
> Hi MESA users
> 
> I am trying to find the evolution of the RGB star in the presence of an extra force ((-Beta) * dx/dm), which is added to the other_momentum subroutine as follows
> 
> dp/dm = - G m/4 pi r^4 - (Beta) * dx/dm
> 
> where Beta is an arbitrary constant.
> 
> I have run mesa for different Betas and as you can see in the attached plot (Tip of the RGB.png), the temperature of the Tip of the Red Giant Branch increases for (0<Beta<300) but for ( 300<Beta<400) there are some instabilities in the evolutionary track of the star and for (450<Beta<500) the instabilities increase in such a way that  the star leaves the standard track which is shown in the plot (HR.png) for Beta=450, and the code terminates therefore the evolution stops after RGB phase.
>  Also for (Beta>500) the code does not converge even in zams phase.
> 
> My question is whether we can have physical interpretation for the instabilities in the evolutionary track of the star or it is just the instability in the numerical code?
> 
> In other words, is it admissible that the extra force becomes so large for beta>300, and makes the star unstable so the star can not live any longer and these models are not physical, or it is just a large quantity that makes the numerical code unstable?
> 
> Thank you in advance
> 
> 
> 
> Sheyda Najafi
> 
> 
> <HR.png><Tip of the RGB.png>