[mesa-users] approx13

[Bill Paxton]

Good news!  Thanks for being persistent.  Maybe someday we can get rid of the T_limits stuff.  Frank and I are working on it.

-B


On May 15, 2014, at 9:50 PM, Ken Shen wrote:

> On Thu, May 15, 2014 at 6:33 PM, Bill Paxton <paxton at kitp.ucsb.edu> wrote:
> >
> > i wonder if T dependent rate limits are somehow getting triggered.    (my guess is you've never heard of those, right?)
> >
> > are you running 1 zone burn?  if so, try adding these to your inlist:
> >
> >       turn_on_T_limits = .false.
> >       turn_off_T_limits = .true.
> 
> 
> BINGO!  Setting turn_off_T_limits = .true. fixed everything.  Neglecting or keeping neutrons and protons in the isotopes list keeps things the same when that flag is .true. (when they don't exist in the actual reactions).  The rates gets changed in net/private/net_derivs.f.  If there are neutrons, rates are changed by factor_neut; otherwise, they are changed by factor_prot.
> 
> factor_neut and factor_prot are set in net_eval.f, with parameters from net_def.f.  The normalization for the neutron and proton factors there is different by 1000, which is why including neutrons mattered a lot more than including protons in the isotopes list.
> 
> In summary, if you are using MESA networks at high temperatures >2e9 K, note that there are flags that work to increase the timestep if you are near detailed balance but at the cost of changing the burning results if you are not in chemical equilibrium.  The rates are reduced by many orders of magnitude, even if neutrons and protons don't exist in the reactions.
> 
> Phew, I thought I was going crazy!!  Thanks so much for all your help!
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140515/6cd8d361/attachment.html>