search

MESAHub / mesa

Modules for Experiments in Stellar Astrophysics
http://docs.mesastar.org
GNU Lesser General Public License v2.1
138 stars 38 forks source link

Fix eps_nuc_categories and approx21 #59

Closed rjfarmer closed 4 years ago

rjfarmer commented 4 years ago

Seems to be broken

rjfarmer commented 4 years ago

Turning off rates and comparing (eps_total + eps_neu) with the (eps_XX) it seems we are not accounting properly for iroag, ircag, irpen, and irnep.

rjfarmer commented 4 years ago

For irpen and irnep Qneu_rpen and Qneu_rnep are both zero so their contribution to eps_neu is 0 but we still have an energy change in eps_total due to the change in rest mass from y(ineut) * rate(irnep)

iroag, ircag are just straight up missing from approx21_eps_info

rjfarmer commented 4 years ago 
I’m looking forward to hearing how we can be getting such big differences in lg_Lnuc for
approx21 vs big net - the jump to lg_Lnuc ~24 for approx21 vs ~17 for the big net is 
beyond strange.    It means we have big differences in eps_nuc itself, not just in the 
categories.   The differences were present back in the last public release.
rjfarmer commented 4 years ago

Rates that still need fixing:

irpen/irnep/ni56_ec/co56_ec -> Their rates, Q's and Qneu's are set in approx21_weak_rates which means the information in reaction_Qs and reaction_Qneus are bogus, thus we use bogus information in approx21_eps_info.

We also need to worry about fe56ec_fake_factor not being available in approx21_eps_info

irnag and ir1616 show differences bigger than i'd expect. For instance the reaction_q for irnag is 13.542599999999995 but summing the mass excess's you get 13.542721824999996 (1.5*2.424915649999999 + 2.863417040000 + 7.0419313099999). Given that the mass excess and reaction_Qs all come from get_mass_excess() i have no idea how they have gotten slightly out of sync.

There may also be issues with the compound rates but i have not fully checked those.