Commits

Author Commit Message Labels Comments Date
Devin Silvia
minor modifications and changes to the plotting script to produce relative abundances.
Devin Silvia
merged.
Matthew Turk
Fixing plotting the legend
Daniel Reynolds
fixed bug in scaling
Daniel Reynolds
Updated BE_chem_solve.C to allow variable/equation renormalization
Matthew Turk
A few cleanups and messups of primordial, to pass off to Dan.
Devin Silvia
modifications to write_primordial_network and the plotting script
Devin Silvia
electrons are back in everywhere, the temperature does evolve, but I think we agree with the results. cooling is currently commented out.
Devin Silvia
merged.
Matthew Turk
Restricting oxygen to o1, o2, de and ge.
Devin Silvia
sympified 0's in chem network
Matthew Turk
Adding plotting script
Devin Silvia
added time and timestep attributes to the solution outputs.
Devin Silvia
Trying to make sure that 'ge' is being evolved in the right way, conversion factors added to convert from specific energy density to energy density.
Devin Silvia
added code to output intermediate and final solutions. also removed electrons from the number density calculations so that they wouldn't effect temperature. To put this back in, need to change the initial conditions generation and chemical_network.py (put 1/weight back in gamma_factor function). Can run O species to CIE in 1e9 years with n_O = 1. (128 cells)
Matthew Turk
Many fixes. * Removing most/all of the printout statements from inside loops. * Adding subcycling. * Number density everywhere. * Dropped maximum temperature in oxygen to 1e7, which now works. * No more ENZO_FAIL in BE_chem_solve
Matthew Turk
Merge
Matthew Turk
Primordial network now only outputs 8 cells, and just the ionization/recombination for HI. Seems to work to 1e-7. Had to make two changes: 1) Remove de from the weight calculation 2) Fix the Jacobian's row/column ordering.
Matthew Turk
We don't want to include electrons in the mass density, either.
Devin Silvia
Work through the creation of ICs and the computation of temperature and gas energy to make sure things were being done correctly. Things now appear to work correctly for the oxygen network. The IC temperatures are returned by the C code in a matching fashion. Additional, the values of the RHS for the oxygen species seem to be calculated correctly. The current issue is that the solver doesn't actually evolve the species and things are "solved" This appears to stem from issues with atol and rtol, which may related to the fa…
Matthew Turk
This crazy, bloated contraption is able to solve with a mere 10 iterations the Oxygen network. Several changes have been made, including dropping the rtol and atol very, very low.
Devin Silvia
added a write_oxygen_network.py to test the solver with oxygen. It seems to be suffering for the same issues as the primordial network. rhs_total is always equal to rhs[de]. Things explode into unrealistic values. This may still have something to do with how electrons are being handled.
Matthew Turk
Fixing up some issues with cooling terms and setting time to something reasonable.
Devin Silvia
multiplied by a factor of 'mh' in the equation to calculate temperature. this also changes the output of dT/dge. Still produces a non-convergent result.
Devin Silvia
printing logT info as well.
Matthew Turk
Fixing rtol/atol sizes
Devin Silvia
changed some print commands to get more information at runtime
Matthew Turk
This adds a few debugging outputs and fixes a few offsets and issues.
Matthew Turk
* Adding BE_chem_solve from Dan Reynolds' Enzo repository. Note that this requires a full set of Enzo definitions to compile. * Successfully create solver which can be executed * Typedef our functions. * Copy ICs into BE_chem_solve input style.
Devin Silvia
Re-wrote the header comment at in the C++ template code.
  1. Prev
  2. Next