Further:
* revert unit tests to previous species definitions (some case mis-matches)
* remove non-essential comments
* opt to maintain case-sensitive species maps with lowercase as fallback
* store species information with case sensitive names
* retain lookup for non-case sensitive species names, e.g. Phase::speciesIndex
* implement flag that enforces case sensitive species names as a member
variable of Phase
* add exception handling for species that are not uniquely defined unless case
sensitive (e.g. Cs and CS in nasa.cti if cs is specified and case sensitivity
is not enforced)
* deprecate Phase::species(std::string&)
Solving with auto=True and then solving with auto=False would leave the domain
width check in place, but without the logic for automatically increasing the
domain width, resulting in unexpected solver failures.
* The commit implements saving of data extracted from SolutionArrays
to HDF containers using pandas infrastructure.
* Two methods are introduced: `write_hdf` and `to_pandas`.
* Both methods only work if the pandas module can be imported; an
exception is raised only if the method is called without a working
pandas installation.
The "catching polymorphic type" warnings appear during compilation with GCC 8:
src/base/global.cpp: In function ‘void Cantera::setLogger(Cantera::Logger*)’:
src/base/global.cpp:28:19: warning: catching polymorphic type ‘class std::bad_alloc’ by value [-Wcatch-value=]
} catch (std::bad_alloc) {
^~~~~~~~~
src/equil/vcs_MultiPhaseEquil.cpp: In member function ‘int Cantera::vcs_MultiPhaseEquil::equilibrate_HP(doublereal, int, double, double, int, int, doublereal, int, int)’:
src/equil/vcs_MultiPhaseEquil.cpp:228:31: warning: catching polymorphic type ‘class Cantera::CanteraError’ by value [-Wcatch-value=]
} catch (CanteraError err) {
^~~
src/equil/vcs_MultiPhaseEquil.cpp: In member function ‘int Cantera::vcs_MultiPhaseEquil::equilibrate_SP(doublereal, double, double, int, int, doublereal, int, int)’:
src/equil/vcs_MultiPhaseEquil.cpp:354:31: warning: catching polymorphic type ‘class Cantera::CanteraError’ by value [-Wcatch-value=]
} catch (CanteraError err) {
^~~
This commit fix this warnings via caught by reference.
The changelog of Sundials 4.0.0 states:
"With the introduction of SUNNonlinSol modules, the input parameter iter
to CVodeCreate has been removed along with the function CVodeSetIterType
and the constants CV_NEWTON and CV_FUNCTIONAL.
Similarly, the ITMETH parameter has been removed from the Fortran interface
function FCVMALLOC. Instead of specifying the nonlinear iteration type
when creating the CVODE(S) memory structure, CVODE(S) uses
the SUNNONLINSOL_NEWTON module implementation of a Newton iteration by default."
so the appropreate conditional changes are added to control
the code execution via CT_SUNDIALS_VERSION preprocessor variable
to omit the parameters of Sundials solver that are no longer required.
In the atomic weights struct in Elements.cpp, ensure that there is no
space before the closing brace of an element and that there is one
space between the longest element name and a 3-digit weight.
Use data from the periodic table at
http://www.ciaaw.org/atomic-weights.htm and
https://iupac.org/wp-content/uploads/2018/12/IUPAC_Periodic_Table-01Dec18.pdf
Elements without any atomic weight in either table do not have a stable
isotope. These are deleted from elements.xml and have their atomic
weight set to -1.0 in Elements.cpp. Add elements after plutonium that
were not previously listed. None of these elements have stable
isotopes.
These elements are retained/added so their symbols, names, and atomic
numbers can still be retrieved and the mapping of atomic number to
index - 1 in the struct is maintained.
Modify the element weight lookup functions to throw errors when an
element with no weight is requested (i.e., the weight is -1.0 in the
struct).
* differentiated Valve::setValveCoeff from PressureController::setPressureCoeff
and introduced MassFlowController::setMassFlowCoeff for consistency.
* introduced FlowDevice::setTimeFunction and FlowDevice::setPressureFunction to
differentiate time-dependent and pressure-dependent functions.
* introduced arbitrary pressure dependence for PressureController
* deprecated FlowDevice::setFunction which is replaced by time and pressure
specific functions.
* introduced properties Valve.valve_coeff / PressureController.pressure_coeff /
MassFlowController.mass_flow_coeff in Cython interface and deprecated
Valve.set_pressure_coeff / PressureController.set_pressure_coeff
* deprecated corresponding function calls in clib interface
* deprecate FlowDevice.setParameters (which was only used by MATLAB interface)
Replace data/inputs/diamond.cti with test_problems version that has
more information. This results in a change in the default pressure and
mole fractions of the gas phase, which in turn changes the result of
one of the regression tests. This is fixed by setting the composition
and pressure of the gas phase in the test to their original values. The
default state from the CTI file matches from the paper.
In addition, there was a difference in the reversibility of reaction u
between the files. Since the thermo for C(d) specifies that the
reaction is irreversible, this is the sense of the reaction that is
chosen.
Include plotting in the diamond_cvd.py and use open properly.
The '${python_prefix}' substring for installation prefix path
was accepted as mapping key for '.format()' function resulting in
a 'KeyError' failure of 'cantera/interfaces/cython/SConscript' script
in case of `env[libdirname] == 'lib64'`.
Moreover the early applied pull request[1] didn't take into account
the additional setting of installation prefix path in the cases
when 'libdirname' takes values different from 'lib64'.
This patch resolves both those issues.
[1]: https://github.com/Cantera/cantera/pull/661
* add deprecation warning for int ReactorBase::type() (to be changed after Cantera 2.5)
* introduce temporary std::string ReactorBase::typeStr() (to be renamed after Cantera 2.5)
* deprecate all functions using the old call and introduce associated temporary functions
Update test_problems for rankine and pureFluid to avoid using the
deprecated PureFluid.h convenience wrapper classes. Update the
rankine.cpp test problem to match the rankine.cpp sample. Switch
both test_problems updated here to use writelog instead of
printf/cout.
Some distributions (e.g. Fedora/RHEL) use 'lib64' instead of 'lib'
on 64-bit systems or could use some other library directory name
instead of 'lib' depends on architecture and profile
(e.g. Gentoo 'libx32' on x32 profile).
If user didn't set 'libdirname' configuration variable then
set it to default value 'lib'
This commit is related to early closed issue:
https://github.com/Cantera/cantera/issues/318
The general intent here was to enable calculating reaction enthalpies in the
Matlab toolbox, as part of the li-ion battery simulations in PR #563.
This required several changes:
- Create getDeltaEnthalpies.m in Matlab toolbox/@Kinetics, as well as similar
methods for Gibbs free energy and entropy of reaction
- Add kin_getDelta to kineticsmethods.cpp.
- Add getPartialMolarEnthalpies to metalPhase class (it returns all zeros).
Note that similar methods are not enabled for the corresponding
'Standard State' methods, for the time being. Mainly because it is
difficult for me to envision a significant use case, but also because of
some lingering confusion between 'standard' and 'reference' states in
Cantera's codebase.
A user-defined mass flow rate function can modify the ThermoPhase object used by
a reactor, for example if it depends on calculating some property of a different
reactor. To make sure that the reactor governing equations are evaluated
correctly, the ThermoPhase state needs to be set after all user-defined
functions have been called.
