685 DO NOT DELETE THIS LINE
---------------------------------------------------------------------------
DO NOT USE TABS IN PERPLE_X DATA FILES, TAB CHARACTERS ARE NOT INTERPRETED AS
BLANK SPACES AND CAUSE FORMATTING ERRORS.
---------------------------------------------------------------------------
SUBDIVISION SCHEMES define the discreitization and range of compositions for a solution
during a calculation. he discreitization and range of compositions for a solution
during a calculation. In many cases, particularly for chemically complex solutions,
this range is restricted to reduce the time and memory required for calculations.
If such a restriction is encountered (**warning ver993**) during the exploratory
stage of a calculation the restriction may have consequences for the final
auto-refine stage result. For this reason it is important for users to correct
the scheme (unless the restriction is intentional). A resume of warnings written
at the end of the exploratory stage is written to the *_auto_refine.txt file.
This file should be examined before any results are accepted as final.
A brief description of how subdivision schemes follows (it reads worse than it is):
If the solution model has a simplicial composition space, then each scheme is
directly associated with an endmember (as identified in *_auto_refine.txt). For
a solution with N endmembers, there are thus N-1 subdivision schemes. In general
it is best to order the solution model endmembers so that the most important end-
member is listed last and therefore determined by difference.
If the solution model has a prismatic composition space, then a scheme is
associated with the S-1 independent compositions of the T simplexes that comprise
the prism (as identified in *_auto_refine.txt). Thus for prismatic composition spaces
there are T*(S-1) subdivision schemes for the compositional variables X(1..T,1..S-1).
X(1..T,S) is determined by difference and therefore then endmembers should be ordered
so that X(1..T,S) corresponds to the most important compositional variable.
The subdivision scheme for each independent compositional variable is specified by
four parameters, in order, XMIN, XMAX, XINC, IMOD:
IMOD - may be either 0 or 1:
IMOD = 0 - is the cartesian scheme, in which compositions are discretized with a
regular spacing.
IMOD = 1 - is a non-linear scheme, in which compositions become more closely spaced
toward zero.
XINC - in most cases the value of XINC is equated to the first (exploratory stage) or
second value of the initial_resolution keyword, which defaults to [1/16 1/48],
i.e., the value of XINC specified within the solution model text here is irrelevant.
The exceptions are if initial_resolution = [0 0] or if the solution model type = 20
(electrolytic_fluid). For these exceptions, the value of XINC is read from the model
text here. XINC controls the resolution of the compositional discretization, i.e.,
the number of compositions generated over the interval of interest is 1/XINC + 1 if
XINC < 0 and is XINC+1 if XINC > 0.
XMAX - is the maximum value of the compositional variable permitted by the scheme. This
value may be relaxed if the hard_limits option is TRUE (default).
XMIN - If IMOD = 0, XMIN is the minumum value of the compositional variable permitted by
the scheme. This value may be relaxed if the hard_limits option is TRUE (default).
If IMOD = 1 and XMIN > 0, then XMIN is the smallest non-zero value of the compositon
for statically generated compositions. Thus, the value of XMIN dictates the asymmetry
of the discretization. XMIN is not relaxed if this composition becomes limiting.
If IMOD = 1 and XMIN = 0, then XMIN has no significance and the asymmetry of the
discretization is specified by the stretch_factor option.
For convexhull minimization calculations (CONVEX) the resolution of a composition is dictated
entirely by the subdivision scheme and the initial_resolution option. For adaptive minimization
calculations (VERTEX) the resolution of statically generated compositions is controlled as in
convexhull calculations, but the compositional resolution for dynamically generated compositions
is limited by the final_resolution keyword. In particular for non-linear subdivision, this has
the consequence that the value of XMIN or stretch_factor should be viewed as the order of
magnitude the poorest acceptable resolution.
---------------------------------------------------------------------------
Solution model types are:
2 - simplicial composition space
6 - order-disorder, simplicial composition space
7 - prismatic composition space (reciprocal when correct)
8 - order-disorder, prismatic composition space
9 - order-disorder, simplicial composition space with a prismatic vertex
10 - simplicial composition space with a prismatic vertex
Special model types:
0 - internal (fluid) EoS
20 - Electrolyte (charge balance) model
26 - Haefner H2O-CO2-NaCl
29 - BCC Fe-Si alloy, Lacaze & Sundman 1990.
30-33 - FeSiC alloys (BCC/FCC/CBCC/HCP), Lacaze & Sundman 1990.
39 - generic hybrid fluid EoS
40 - Silicate fluid (MRK)
41 - COH fluid (hybrid MRK)
42 - FeS liquid, Saxena & Eriksson 2015 with ECRG corrections
---------------------------------------------------------------------------
Character data is format free.
---------------------------------------------------------------------------
comments can be placed between models provided, nothing is written in the first
10 columns. Comments may be placed after data if it is separated from the data
by a '|' marker. Comments may be placed within the data in some cases without the
'|' marker, but it is always safe to add a comment with the marker.
---------------------------------------------------------------------------
begin_model | solution models begin/end with the begin_model/end_model tags
model_name | the name (<11 chars) used to identify the solution model
abbreviation abb | an abbreviation used optionally for output (solution_names abb)
full_name type_2_model | a long name used to classify the model (e.g., liquid) and optionally for
| output (solution_names ful)
2 | model type: simplicial composition space (see list above)
N | number of endmembers
name_1 name_2
... name_N | endmember names, these must match the endmember names in thermodynamic data file
0 0 ... 0 | N endmember flags, if 0 the pure endmember is treated as part of the solution.
| subdivision schemes (see above) for the first N-1 endmembers
0 1 0.1 0 | scheme for endmember name_1
...
0 1 0.1 0 | scheme for endmember name_N-1
| if the solution model does not have an excess function, or the excess function
| is specified internally the excess function section may be replaced by the
| left-justified 'ideal' tag.
begin_excess_function | excess function data begin/end with the begin_excess_function/end_excess_function tags
| the terms of the excess function (J/mol, P and T in bar and K) are expressed
W(name_1 name_2 name_2) c_0 + c_1 * P + c_2
| where the order of the term with respect to the endmember fractions is indicated by
| the number of times the endmember name occurs within the (). e.g., the foregoing
| represents the term (c_0 + c_1 * P + c_2) * y_name1 * y_name2 * y_name2
| where, except in the case of van Laar models (see van_Laar_sizes below), y is the
| mole fraction of the endmember.
| The format used for the p-t dependence of the coefficient is flexible and described
| at www.perplex.ethz.ch/perplex/faq/Linear_P_T_function_input_format.txt
| for the Redlich-Kistler formulation a more compact format is allowed, and will be
| documented upon request. This formulation is indicated by a leading tag of the
| form:
Wk(name_1 name_2)
end_excess_function
| the number of identisites, site multiplicities, site fraction expressions needed to
| compute the configurational entropy of the solution follow the excess function
M | the number of sites (specify 0 for no model or a molecular entropy model)
| then for each of the M sites:
N q | N - number of species on site 1, q - multiplicity of site 1. if q = 0, then
| the site is a Temkin site
| if q > 0, then the N q line must be followed by N-1 linear expressions for
| the site fraction of a species in terms of the mole fractions of the endmembers
| (denoted by the endmember name), the stoichiometric coefficients c_i may be
| written as integers or real numbers or integer fractions (e.g., 11/31). An initial
| constant is optional.
| The tag to left-hand-side (e.g., z(1,1)) of the equals sign has no significance,
| it may be chosen for convenience or transparency. E.g.:
z(1) = [c0 +] c_1 name_1 + c_2 name_2 + ...
...
z(N-1) = [c0 +] c_1 name_2 + c_2 name_4 + ...
| if q = 0, then the N q line must be followed by N linear expressions for
| the molar amount of the N species in terms of the mole fractions of the endmembers
| (denoted by the endmember name), the stoichiometric coefficients c_i may be
| written as integers or real numbers or integer fractions (e.g., 11/31)
| it may be chosen for convenience or transparency. E.g.:
n(1,1) = [c0 +] c_1 name_1 + c_2 name_2 + ...
...
n(1,N) = [c0 +] c_1 name_2 + c_2 name_4 + ...
| the above sequence is repeated for each of the M sites.
| the configurational entropy data may be followed by several optional tags and/or sections:
reach_increment 0 | see www.perplex.ethz.ch/perplex_options_body.html#reach_increment
refine_endmembers | see www.perplex.ethz.ch/perplex_options_body.html#refine_endmembers_solution_model_file
| if the excess function is the van Laar formulation, then the following section specifies
| the size parameter (c_i) for each endmember:
begin_van_laar_sizes
alpha(name_1) c_1
...
alpha(name_N) c_N
end_van_laar_sizes
| if the model makes dqf corrections to an endmember specified in the thermodynamic data file,
| then the following section can be used to specify these corrections (it may be preferable to
| specify positive corrections as a make_definition in the header of the thermodynamic data file):
begin_dqf_corrections
| where c_0, c_1, c_2 specify the p-t dependence of the correction as described at
| at www.perplex.ethz.ch/perplex/faq/Linear_P_T_function_input_format.txt
dqf(name_2) = c_0 + c_1 * P + c_2 * T
end_dqf_corrections
end_of_model
--------------------------------------------------------
begin_model | solution models begin/end with the begin_model/end_model tags
model_name | the name (<11 chars) used to identify the solution model
abbreviation F | an abbreviation used optionally for output (solution_names abb)
full_name fluid | a long name used to classify the model (e.g., liquid) and optionally for
| output (solution_names ful)
39 | model type: model_type_39, see perplex.ethz.ch/Perple_X_generic_hyrbid_fluid_EoS.html
N | number of species
name_1 name_2
... name_N | species names, these must match the species names in thermodynamic data file
0 0 ... 0 | N species flags, if 0 the pure species is treated as part of the solution.
| subdivision schemes (see above) for the first N-1 species
0 1 0.1 0 | scheme for species name_1
...
0 1 0.1 0 | scheme for species name_N-1
| type 39 solution models use an internal EoS to compute excess properties, this is
| indicated by the ideal tag
ideal
| type 39 solution models use a molecular configurational entropy model, this is
| indicated by setting the the number of identisites to zero.
0 | the number of identisites (specify 0 for no model or a molecular entropy model)
| the configurational entropy data may be followed by several optional tags and/or sections:
reach_increment 0 | see www.perplex.ethz.ch/perplex_options_body.html#reach_increment
refine_endmembers | see www.perplex.ethz.ch/perplex_options_body.html#refine_endmembers_solution_model_file
| currently no provision is made for DQF corrections or van Laar excess funcitons in GFSM
end_of_model
--------------------------------------------------------
begin_model
Talc, ideal.
M1 M2 T2
______________________
Mutliplicity 2 1 2
______________________
1 en Mg Mg SiSi
Species: 2 fs Fe Fe SiSi
3 mgts Mg Al AlSi
______________________
reformulated from relict equipartion (model type 7) to simplicial
composition space (model type 2).
JADC 5/10/2018
T | solution name
abbreviation Tlc
full_name talc
2 | model type: simplicial
3 | number of independent endmembers
ta fta tats | endmember names
0 0 0 | endmember flags, indicate if the endmember is part of the solution.
0.0 1. 0.1 0 | range and resolution for ta, imod = 0 -> cartesian subdivision
0.0 1. 0.1 0 | range and resolution for fta, imod = 0 -> cartesian subdivision
ideal
3 | 3 site (M1, M2, T2) conigurational entropy model
2 2. | 2 species on M1, 2 sites per formula unit.
z(m1,mg) = 1 ta + 1 tats
2 2. | 2 species on T2, 2 sites per formula unit.
z(t2,al) = 1/2 tats
3 1. | 3 species on M2, 1 site per formula unit.
z(m2,mg) = 1 ta
z(fe,m2) = 1 fta
end_of_model
--------------------------------------------------------