Removed a number of variables relevant only to the gas-phase transport
coefficient models. These include the following:
- vector_fp m_cond;
- vector_fp m_polytempvec;
- vector<vector_fp> m_condcoeffs;
- std::vector<vector_fp> viscCoeffsVector_;
- vector_fp m_sqvisc;
- vector_fp viscSpecies_;
- DenseMatrix m_wratkj1;
- DenseMatrix m_phi;
- DenseMatrix m_wratjk;
Removed much code related to these variables.
Added a number of variables relevant to the liquid-phase transport
models including:
+ vector_fp m_condSpecies;
+ vector_fp m_viscSpecies;
+ vector_fp m_visc_A;
+ vector_fp m_visc_n;
+ vector_fp m_visc_Tact;
+ vector_fp m_thermCond_A;
+ vector_fp m_thermCond_n;
+ vector_fp m_thermCond_Tact;
Changed some of the relevant comments to
pertain to the liquid-phase models.
Added compatibility for sundials 2.4 along the 1.8.0 branch. I'll
add it to the main trunk as well shortly. This hasn't been fully tested
except for the cxx_examples problem.
gas-phase transport, we needed to pass the GasTransportParams rather
than TransportParams.
We made new methods to initialize liquid transport including
TransportFactory::setupLiquidTransport()
TransportFactory::initLiquidTransport()
TransportFactory::getLiquidTransportData()
Added new struct LiquidTransportData.
Added getArrhenius method copied from kinetics directory to parse
Arrhenius form XML.
of the same name holds core components including the thermo object,
XML nodes, species names and molecular weights.
A derived class GasTransportParams holds
kinetic-theory-of-gases-related parameters.
LiquidTransportParams in LiquidTransportParams.h now derives from
TransportParams class. Removed kinetic-theory-of-gases-related
parameters.
class LiquidTransportParams;
Changed the init(TransportParams& tr) method to
initLiquid( LiquidTransportParams& tr )
Removed a number of gas-related transport parameters from the
init/initLiquid() method including m_poly, m_diam, m_eps, m_alpha
In AqueousTransport.cpp there was a method to return a struct to a
GasTransportData. This now returns LiquidTransportData struct, but is
not yet working.
In LiquidTransport.cpp removed all references to m_polytempvec, a
vector of 5 entries of the form
- m_polytempvec[0] = 1.0;
- m_polytempvec[1] = m_logt;
- m_polytempvec[2] = m_logt*m_logt;
- m_polytempvec[3] = m_logt*m_logt*m_logt;
- m_polytempvec[4] = m_logt*m_logt*m_logt*m_logt;
We might want these back soon. This allowed up to comment out the
code in the methods that evaluate the temperature dependence of the
viscosity, thermal conductivity and diffusivity using polynomials in
these variables. --This line, and those below, will be ignored--
M AqueousTransport.h
M LiquidTransport.h
M AqueousTransport.cpp
M LiquidTransport.cpp
const doublereal* grad_T,
int ldx,
const doublereal* grad_X,
int ldf,
const doublereal* grad_Phi,
doublereal* fluxes)
after passing electrostatic potentialgradient as well as the rest of
the gradients.
Added initGas( GasTransportParams& tr ) and
initLiquid( LiquidTransportParams& tr ) because of the different
TranportParams subclasses. There might be a better way to do this.
Changed calcDensity to be a virtual function started at VPStandardStateTP().
It's a protected function as well.
At the VPStandardStateTP level, calcDensity() produces an error when called, because the EOS isn't specified. Derived classes must define this function, supplying the EOS.
WaterPropsIAPW: I added documentation to this routine, defining what functions changes the state of the object.
IdealMolalSoln: I had forgotten to update this object to employ the newer setState_TP() strategy that I had used in VPStandardStateTP().
system so that I can do kinetics with the ThermoPhase object.
For molten salts, we're going with a standard concentration
of 1.0, until there is an argument otherwise. This means
that the kinetics objects, homogeneous and heterogeneous,
will treat the molten salts as unitless.
This is a vector that is a baseline "safe" mole fraction
vector that can be applied anywhere. I was running into problems
with specifying this for molten salts, because just setting
the first component to a mole fraction of 1 causes
charge imbalance.
M thermo/ThermoPhase.h
M thermo/ThermoPhase.cpp
IonsFromNeutralVPSSTP : This is initially used for molten salts.
Molten salt thermo is treated via the neutral molecule
molten salt expressions. However, ions migrate by themselves.
therefore, you must treat ions within the transport equations.
This thermophase object treats the ions as species, and is
made to be consistent with the neutral molecules thermophase
object that is used to express the non-ideal thermodynamics
of the molten salt solution. Currently, MargulesVPSSTP
is used for this as the LiKCl system is parameterized using
this excess Gibbs free energy formulation.
MargulesVPSSTP
Various upgrades and fillers. It now reads and is initiated
from the input file.
setState_TP(), setTemperature(), and setPressure() within
VPStandardStateTP() formulations
There was some circular logic, previously. Now setTemperature() and setPressure()
are subserviant to setState_TP(). Basically, it doesn't make
sense to calculate the standard state without first getting
both the temperature and pressure (not just one and then the other)
before you calculate the standard state for VPStandardStateTP derivatives.
Think about the water EOS, which is a real fluid. You don't want
to be trying to calculate the standard state within the enveloppe where
neither gas nor liquid is stable.
PseudoBinaryVPSSTP
Deprecated.
not in the current directory. THe xml file is always written to the
current directory. And the subsequent reading of the xml file must
reflect that. The xml file was trying to be read from the directory
where the cti file was located. Now, the xml file will be read from
the current directory.
