Re: [mesa-users] instability issue

[Robert F. <rjf...@as...>]

If you do want to limit the timestep then don't forget to look in

http://mesa.sourceforge.net/controls_defaults.html#timestep_controls

There are many options (plus always the option to set your own timestep
limiter in a run_star_extras file) including limits on the maximum timestep
taken.

Rob

On Wed, Jun 21, 2017 at 8:49 AM, Bill Paxton <pa...@ki...> wrote:

> To follow up on Josiah's excellent response,
>
> 1st) do you really mean 10^-11 seconds?  is that perhaps a typo?
> 2nd) when I'm doing core collapse supernovae explosions in mesa the
> timesteps get down to milliseconds at the start.
>         the steps grow in size as the timescale of the evolution increases.
>         the same is true for other types of evolution -- the timesteps
> adapt to the timescales of interest according to user set limits.
> 3rd) there are large differences between mesa's implicit Lagrangian
> approach and the explicit Eulerian method you seem to have in mind.
>         confusion is inevitable until you get clear on this
> 4th) mesa isn't the right tool for all problems.  you may well decide that
> your task is one that is better done using something else.
>
> Good luck,
> Bill
>
>
>
> On Jun 21, 2017, at 8:27 AM, Josiah Schwab wrote:
>
> > Hi Zoe,
> >
> >> We are going to study the behaviour of an external field during the
> >> evolution of star, therefore a second order equation of time and
> >> radius must be solved inside mesa code. The problem is that for the
> >> instability to be refused, " dt " must be about 10^(-11) seconds and
> >> the timesteps between mesa profiles are of order 10^(+15) seconds. Is
> >> it possible to study such a behaviour and apply the equation in mesa?
> >
> > It doesn't make much sense to evolve MESA models with a timestep of
> > 10^{-11} s.  Perhaps you could use MESA to generate a series of stellar
> > models, and then treat everything as "frozen-in" on the short timescale
> > of interest for your problem.
> >
> >> Another question is about mesa mesh adjustment and the CFL number. How
> is
> >> the CFL number defined in mesa and which quantities does it relate,
> (dt,dr)
> >> or (dt,dm) ?
> >
> > If you're using the hydrodynamics capabilities, then there are CFL-like
> > controls, e.g.,
> >
> >  http://mesa.sourceforge.net/controls_defaults.html#dt_div_
> min_dr_div_cs_limit
> >
> > but otherwise not.  Remember that MESA does not timestep explicitly.
> >
> > Best,
> > Josiah
> >
> > ------------------------------------------------------------
> ------------------
> > Check out the vibrant tech community on one of the world's most
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > _______________________________________________
> > mesa-users mailing list
> > mes...@li...
> > https://lists.sourceforge.net/lists/listinfo/mesa-users
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> mesa-users mailing list
> mes...@li...
> https://lists.sourceforge.net/lists/listinfo/mesa-users
>