research/cantera
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
3057 commits 2 branches 16 tags 37 MiB
Commit graph

3057 commits

This branch
This branch
All branches
Author SHA1 Message Date
Harry Moffat
08716fa3f1 Make sure that phaseIsStable flag is set correctly 2011-02-18 23:04:50 +00:00
Harry Moffat
6dd143640c Incremental progress on dogleg 2011-02-18 20:25:53 +00:00
Harry Moffat
3996c38b3c Changed the residual weights formula 2011-02-16 17:46:28 +00:00
Harry Moffat
80227d3314 More work on dogleg algorithm 2011-02-16 16:40:02 +00:00
Harry Moffat
96880bae26 Worked on trust region calculcations 2011-02-15 17:09:03 +00:00
Harry Moffat
6a902abd24 More work on the dogleg option. Figured out and verified the calculation 
of the steepest direction when there are scaled residuals and
scaled solution variables.
 2011-02-15 02:57:34 +00:00
Harry Moffat
ecc2d271c9 Worked on verifying the steepest descent calculation 2011-02-12 03:36:24 +00:00
Harry Moffat
0c178dbdd8 Prototyping a Dogleg method. 
Added in the calcualtion of the steepest descent and the Cauchy point
 2011-02-10 20:59:53 +00:00
Harry Moffat
3dbe5d658f Fixed an error in the rootsolver 2011-02-02 05:33:29 +00:00
Harry Moffat
8dc95dfbc1 Added accessor functions for phaseStability and phaseExistence() 2011-02-01 19:08:28 +00:00
Harry Moffat
60194831c1 Added a copy constructor and an assignment operator 2011-01-12 20:55:07 +00:00
Harry Moffat
bab66a8b89 Doxygen update 2011-01-12 00:52:01 +00:00
Harry Moffat
f493f8f39a Doxygen update 2011-01-11 17:34:31 +00:00
Harry Moffat
3a63c938ea Added a few more files 2011-01-11 15:27:59 +00:00
Harry Moffat
14bf2b1ee2 Started putting these files into the standard form. 
Started eliminating doxygen errors.
 2011-01-10 23:22:19 +00:00
Harry Moffat
f727cd44af Added in a more robust treatment of handling limits 2011-01-05 03:55:13 +00:00
Harry Moffat
a36864543a Got rid of a print statement. 2011-01-04 23:02:07 +00:00
Harry Moffat
b1750d579a Added return variables to a lot of Residual evaluation functions. 
These are now used by the nonlinear solver to bail if the residual
says to bail.
 2011-01-04 22:04:59 +00:00
Harry Moffat
5395e214b8 Fixed a doxygen error. 2010-12-31 23:54:29 +00:00
Harry Moffat
eb35f9e253 Some changes to reduce extra printing 
Some changes due to a bad algorithm - Known - need to fix this.
 2010-12-31 23:20:55 +00:00
Harry Moffat
8c4d47f59c Fixed an error in the phase stability calculation 2010-12-18 18:26:16 +00:00
Harry Moffat
fc7b7e8dc6 Initial attempt at exposing the phase stability routine to the 
outside without actually doing an equilibrium solution.
 2010-12-16 22:44:29 +00:00
Harry Moffat
32816b1fa7 Incremental commit of new ability to determine phase stability 
without actually solving the equilibrium equations.
 2010-12-15 20:20:13 +00:00
Harry Moffat
e10fa97af0 Added the phaseIsStable idea 2010-12-15 02:54:08 +00:00
Harry Moffat
73d8df7f4c Tightend up the tolerances on the nonlinear solvers. 2010-12-15 02:15:27 +00:00
Harry Moffat
ec8bbc5be2 Fixed the help section to be in line with what the code actually does. 2010-12-14 23:56:12 +00:00
Harry Moffat
d5f44a5e6b ResidJacEval: 
Added in support for returning a failure condition from calcDeltaSolnVariables()


NonLinearSolver:
   Added in support for returning a failure condition from bueler_jac().

   Added in a call to return an editable bounds constraint vector.

DenseMatrix:
   Added in comments about error handling from LAPACK routines.


mdp_allo:
   Added in a new routine, mdp_zero_int_1

thermo.h
electrolyteThermo.h
   Started to add in ifdef blocks.  These are not handled correctly currently.
 2010-12-07 18:05:25 +00:00
Harry Moffat
44e03fcf79 Basic upgrade to the DenseMatrix Class. 
We've added a capability to change the handling of singular matrix
    conditions.
     Options exist to either throw an exception or return an error code.
     Options exist to print out a warning or to not print out anything.
 2010-11-30 15:33:50 +00:00
Harry Moffat
2e50b8a929 Fixed an error. The m_speciesData XML tree wasn't being created. 
I created one in the constructor.
 2010-11-18 03:09:45 +00:00
Harry Moffat
1f53f7db72 Fixed minimal python build. 2010-11-15 22:34:21 +00:00
Harry Moffat
3e0d9cef21 Fixed an error that crept in from the recent merge 2010-11-12 22:06:23 +00:00
Harry Moffat
ff3ebb25c2 Merge of the main branch into the LiquidTransportDevelop branch. 2010-11-12 21:37:41 +00:00
Harry Moffat
c3405dad33 Added a file 2010-11-12 20:06:53 +00:00
Harry Moffat
b6e1b14216 Deleted a file 2010-11-12 20:00:56 +00:00
Harry Moffat
fa56309462 Modifications to the Makefiles to make the THREAD_SAFE_CANTERA ifdef 
block work again.
 2010-11-12 19:18:49 +00:00
Harry Moffat
4eb6e622be Fixed a compiler error on the THREAD_SAFE ifdef 2010-11-12 18:08:51 +00:00
Harry Moffat
56ad9ca9e8 Fixed a compile error in the THREAD_SAFE capability 2010-11-12 18:02:21 +00:00
Harry Moffat
58baed0f2e Fixed issue 17 
There was an obvious error in the THREAD_SAFE_CANTERA compile.
 2010-11-11 00:19:59 +00:00
Harry Moffat
4942cf30aa Fixed issue #17. 
There was an obvious error in the THREAD_SAFE_CANTERA ifdef block compile.
 2010-11-11 00:17:45 +00:00
Harry Moffat
54e1280e08 Added hooks in for filters for New step directions and for 
new solutions.
 2010-11-10 05:01:22 +00:00
Harry Moffat
813a98ddf1 Fixed the residual norm calculation. What was in there before 
was necessary but not sufficient. Since there is no good 
theory in tying the residual calculation to convergence short
of calculating the condition number, I've made its magnitude 
a slave to the solution update magnitude.
 2010-11-10 01:39:48 +00:00
Harry Moffat
12b6f572e6 Added in a new Residual evaluation type, Base_LaggedSolutionComponents, 
to properly handle Picard iterations within the damping algorithm.
 2010-11-08 22:22:13 +00:00
Harry Moffat
154b234da9 Added a printout of the jacobian, when an ifdef block is turned on. 2010-11-05 19:44:05 +00:00
Harry Moffat
c44b477285 Fixed an error in the specification of m_deltaBoundsMagnitudes 
Made the residual check in dampStep a lot harder to satisfy.
 2010-11-02 15:44:09 +00:00
Harry Moffat
7fc2841f85 Fixed an error in vcs_solve_TP. m_tPhaseMoles_new wasn't set 
before being used.
 2010-10-28 01:33:54 +00:00
Harry Moffat
c7ce446b32 Added a special constructor for setting the element potential directly. 2010-10-27 21:07:51 +00:00
Harry Moffat
9373fe02f7 These routines needed std:: added to the arguments and function bodies 
of std c++ library calls.
 2010-10-27 21:07:15 +00:00
Harry Moffat
a96f865403 Added FixedChemPotSSTP object. 
This is a utility object that will provide a bath species at a certain
chemical potential to the equilibrium solver.
 2010-10-27 17:00:46 +00:00
Harry Moffat
13f151a358 Modification to the structure of solve_TP() 
Took the special case of SS species deletion in RxnStep and moved
it into the main routine. It now counts as an interation step.
I'm trying to simplify the structure of the program.
 2010-10-25 22:12:55 +00:00
Harry Moffat
1bdee90289 Fixed an error that occured when components of multispecies 
phases were  zeroed because other phases were being zeroed.
Then the noncomponents of that same multispecies phase could 
comprise the entire phase and the Hessian diagonal would then be
zero. The fix to zero the species seems to work.
 2010-10-23 01:35:10 +00:00