Re: [mesa-users] On "modus videndi with very thin shell burning"

[Falk H. <fh...@uv...>]

Kent,

as Bill pointed out some shell-shifting algorithms have been tried in  
MESA in the past, and they were ejected when it came to rigourous  
verification leading up to the MESA paper. These  showed in addition  
to the data that Bill sent on RGB stars, that AGB stars would not  
produce proper mixing behaviour, in particular the absence or weaker  
appearance of the third dredge-up event.

As to how long things can run to be practical of course depends on the  
goals you have in mind. MESA does in 'high-fidelity' mode whatever it  
does at least 10 if not 20 times faster than what people including  
myself have been using before (and it does it better). Which does not  
mean that we should not try to improve further in speed, but just to  
give you a users perspective.

Best, Falk.



>
> I haven't run a calculation with a very thin shell all the way up the
> AGB. I'm not that patient. My guess, based on my attempts, is: At
> *least* several days even with four threads running. Sometimes it  
> may be
> worth the brute calculation, but it seems to me that it would be
> worthwhile to experiment with some kind of front propagator method.
> Frank, you pointed me at some leads; I'll take a look at these but I
> can't promise I'll be able to put anything useful in the code anytime
> soon. (Or ever.)