Commit Graph

336 Commits

Author SHA1 Message Date
Ian Bell
b58e3cbe70 Fixed bug in https://github.com/CoolProp/CoolProp/issues/93
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-19 01:14:05 +02:00
Ian Bell
b506fb5d26 Tidied up derivative tests
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-19 00:36:08 +02:00
Ian Bell
881a6c4616 Added handling for first partial derivative as a string - added tests - they pass
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-19 00:34:26 +02:00
Ian Bell
569ebc43ac First derivatives are calculated by gridded tabular backend
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 22:00:25 +02:00
Ian Bell
e3fba36d81 Gridded table builds
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 19:28:54 +02:00
Ian Bell
ad15288c23 Byebye to Armadillo (maybe reconsider later on)
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 19:28:00 +02:00
Ian Bell
9c28d81fea Quit Brent if solved to within numerical precision
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 19:27:05 +02:00
Ian Bell
dd71b1ac22 Really this time un-specify the phase
More logical debug output for invalid phases

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 19:26:31 +02:00
Ian Bell
99da465b06 Unspecify the phase after the HP solver with Brent - test added
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 18:52:06 +02:00
Ian Bell
9293c6355c For posterity's sake, a version using armadillo - not an improvement... slower than Eigen
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-18 09:47:02 +02:00
Ian Bell
f7464fcaa2 Implemented Eigen solution for all() function, speed testing next
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-17 21:09:27 +02:00
Ian Bell
10f23992e7 Cleanup of all function - not any faster though
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 18:45:24 +02:00
Ian Bell
fc2e618fb7 Speed optimizations for saturation routines
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 17:50:04 +02:00
Ian Bell
4041e88a5e Fully implemented, but a lot slower, why?
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 16:13:03 +02:00
Ian Bell
9a1a0122e3 Renamed Derivatives to HelmholtzDerivatives; they all check fine
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 14:26:25 +02:00
Ian Bell
5cad463158 Parallel derivatives finished for NonAnalytic term
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 14:11:27 +02:00
Ian Bell
68d3725ffd Added tests for undefined parameters; documented undocumented parameters
Closes https://github.com/CoolProp/CoolProp/issues/97

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-16 12:40:35 +02:00
Ian Bell
6c4592fa64 Reorganized swig interface file
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 22:49:22 +02:00
Ian Bell
eee8139306 SAFT has transitioned to parallel evaluation of derivatives, though it isn't truly parallel since there is some duplication
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 22:34:24 +02:00
Ian Bell
29632c72e6 Transition complete to generalized exponential term
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 22:03:40 +02:00
Ian Bell
2270f9f076 Helmholtz derivatives are all calculated using parallel code - next step is to remove classes from Helmholtz
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 21:16:41 +02:00
Ian Bell
10b9060863 Fixed failing Catch errors due to humid air properties
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 20:28:15 +02:00
Ian Bell
61bc24e7b1 phases is now an enum, added (non-working) critical region VLE solver
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 17:13:00 +02:00
Ian Bell
3e0864e620 Ncomp is now a std::size_t
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 17:11:25 +02:00
Ian Bell
68f917c5f5 CMakeLists with profiling block 2014-08-15 16:40:46 +02:00
Ian Bell
0112436169 Updated main with speed testing for parallel derivatives of power term
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 16:05:15 +02:00
Ian Bell
d4e7a6015e Silenced a few warnings
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 15:57:33 +02:00
Ian Bell
2136f417b9 Updated codelite and added Tabular backend structure
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 15:44:55 +02:00
Ian Bell
c3508c3f55 Added parallel evaluation of residual helmholtz terms
Should in the end only slow down by 1.2x, but calculate at least up to second order derivatives all at once, so net speedup of 5 if calculating all first and second order partials !!

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-15 15:42:47 +02:00
Ian Bell
f8c8d104fc Added error message to PropSI output to close https://github.com/CoolProp/CoolProp/issues/95
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-13 15:09:20 +02:00
Ian Bell
a7ddd80d5a Much more progress on PY flash routines. Only a few edge cases that fail, otherwise all is good
Moved from bounded secant to Brent for stability

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-13 14:26:47 +02:00
Ian Bell
baf5a88a8e Added backup saturation_T routine
It actually has exactly the same form as saturation_p. Interesting.  Possibility for optimization?

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-13 12:16:14 +02:00
Ian Bell
a84edcdb57 Implemented saturation P 1D solver for low pressure for some fluids like n-Propane
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-13 11:24:27 +02:00
Ian Bell
f7c0cb0b06 Relaxed some pressure domain checks and pass state by reference
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 23:40:16 +02:00
Ian Bell
acd7f98b74 Fixed typo in PQ_flash
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 21:44:05 +02:00
Ian Bell
8f9c6ca438 Added hs_anchor test (passing) and function get_state to allow to get access to states stored in the backends
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 21:17:34 +02:00
Ian Bell
21f18bf104 Added aliases for molemass and p_triple
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 21:16:11 +02:00
Ian Bell
e4c8d3f061 Python wrapper is more or less back in action completely. Additional testing is required.
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 16:25:33 +02:00
Ian Bell
802495cad3 Last commit before rename of python module to CoolProp
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 15:07:57 +02:00
Ian Bell
5726d38975 Bump up max temps for melting lines to make them valid up to the maximum pressure of EOS
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 12:02:56 +02:00
Ian Bell
b7bd58d986 More work on flash routines, number of failures is rapidly decreasing - still have problems for low pressure sat_p calcs
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 11:21:12 +02:00
Ian Bell
79eca1d58e check both versions of REFPROP on windows always
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-12 08:50:14 +02:00
Ian Bell
ce5d8a919f D(T,P) falls back to Brent if can't get good solution with other solvers
A guaranteed solution if the solution domain made wide enough

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-10 21:39:09 +02:00
Ian Bell
8bd4227f38 Added T for phase determination p calculations 2014-08-10 21:02:15 +02:00
Ian Bell
86641af670 Fixed some errors with two-phase Trho inputs
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-10 20:38:59 +02:00
Ian Bell
36d601490f Lots more work on melting curves, can now make pretty phase diagram for water including the melting line 2014-08-10 19:28:04 +02:00
Ian Bell
7203107b67 Can use pseudo-pure fluids again with REFPROP backend
Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-10 16:15:43 +02:00
Ian Bell
5ee295f639 cleaned up whitespace 2014-08-09 15:12:31 +02:00
Ian Bell
c3cc696080 P,Y now works if P > pc
Add supercritical_gas and supercritical_liquid phase flags for a bit finer granularity

Signed-off-by: Ian Bell <ian.h.bell@gmail.com>
2014-08-09 15:10:09 +02:00
Ian Bell
1a6e516eb0 Merge branch 'master' of https://github.com/coolprop/coolprop 2014-08-09 13:59:48 +02:00