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.
This method returns the units of the concentration-like terms appearing
in rate expressions, and are needed in order to convert rate constants
from user-specified input units to Cantera's MKS+kmol system.
Only treat activation energies as a special case, rather than all molar
energies. Units of activation energy can be set either explicitly or by setting
units for energy and quantity. Only the case where activation energies are given
as temperatures needs to be specified explicitly.
Allow setting of default energy units, which allows calories to be used.
Also add dyn/cm^2 as an option for pressure units.
This will prevent floating point exceptions (sometimes enabled by third-party
codes) in case c[1] or c[2] are zero but will not change the current behaviour
if c[1] and c[2] are not zero.
When the nonreactant_orders option was enabled, specifying reactant orders for
species which were not present in the phase previously resulted in out-of-bounds
memory access.
The previous formulation will only consider a rection as electrochemical
if a beta value is supplied for that reaction *and* the reaction is an
'edge_reaction.' This is problematic for two reasons: (1) many/most
charge-transfer reactions of interest occur at two-phase boundaries (see,
for example, Li-ion batteries and PEM fuel cells), not the three-phase-
boundary-like edges (which are most relevant for SOFCs). (2) determining
whether a reaction is electrochemical or not should not rely at all upon
user input - the program itself should check to see whether charge is
transferred between phases, and the appropriate steps should be taken
during rate-of-progress calcuations.
This commit addresses the former issue. Currently, if a charge-transfer
reaction is written as a surface_reaction, the code does not apply the
voltage correction to the forward rate. By default, then, the entire
voltage correction is applied to the reverse reaction, which is the same
as setting beta = 0; not a good 'default' behavior (beta = 0.5 is a more
appropriate default). With this change, surface reactions can now be
supplied with a beta value in cti or xml formats, and will be recognized
as a charge transfer reaction.
Longer term, it would be better to change the constructor routines such
that charge transfer is automatically detected and handled, rather than
relying upon user-specified flags.
Fixes an error introduced in 37f71bd9 which caused these reactions to be
ignored. However, flux calculations for reactions such as H + HO2 -> 2 OH are
still incorrect.
