Calphad, Vol. 26, No. 2, pp.

143-154, 2002
Pergamon 0 2002 Published by Elsevier Science Ltd
0364-5916/02/$ - see front matter

PII: SO364-5916(02)00031-7

“EH”ERkf’: The Interactive Support Package of Experimental Thermodynamics.

Josef Tomiska*
Institute of Physical Chemistry, University of Vienna
Waehringerstr. 42, A- 1090 Vienna, Austria
*: Corresponding author’s e-mail: joseEtomiska@univie.ac.at
Homepage: http://www.univie.ac.at/tomiska

(Received October 5,200l)

Abstract. The presented program package “ExTHERM” consists of three parts: (i) cETD, a simple interac-
tive open data bank module covering all types of metal alloys (over 150 binary phases and some ternaries),
(ii) cA4od, a module for algebraic conversion among different representations of excess data, and (iii)
cM3 (KA&/ EkfFlHTC), three interactive evaluation units suitable for overall best fits of experimental
investigations expandable to mass spectrometric (KMS), EMF, and calorimetric investigations (HTC). The
concept is based upon the multi-component TAPS (thermodynamically Adapted Power Series) concept of
excess quantities covering all types of solutions, includingn liquid alloys showing short range ordering. The
data bank is based upon the Hultgren data of binary systems and on new experimental data. Individual
supplements may be performed easily. The conversion module employs a modular concept procedure which
makes the conversions independent of the actual system data. For all software modules contact the author.
0 2002 Published by Elsevier Science Ltd.

1. Introduction

The topics of the presented interactive support package for experimental thermodynamics ExTHERA4
are not focused on phase diagram computations, but covers the surrounding topics of it: (i) Conversion of
data represented by means of different approximation formulas, (ii) Evaluation of experimental data, and (iii)
A convenient interactive handbook of mixing behavior solution systems as an improved and enlarged
“electronic Hultgren”. Individual supplements may be performed easily.
The support package ExTHERM is configured as an interactive supporting tool even for those who are
not familiar with sophisticated thermodynamic models, and data treatment of applied thermodynamics. Em-
phasis is placed on simplicity, efficiency, general applicability, and convenient running. Therefore:
ExTHERM is simple in use even by occasional user, and it is split into simple algorithms. The program
package has been developed stepwise for many years, and is written in PASCAL 7 employing individual gra-
phical support. Multi-tasking working is possible, and all diagrams and tables may be copied easily into arbi-
trary graphic programs such as CorelDraw or WinWord by means of the intermediate copy technique.
Simple color codes will help to prevent errors in mind. Both, color as well as black-and-white plots are
possible. Automatic and individual scaling and changing the rounding limits are possible. Full cursor support
is integrated, and all diagrams may be zoomed either by mouse click or by keyboard. Optional plot of a com-
parability graph possible for all diagrams except the ternary triangle plots.

2. The fundamental algorithm

Emphasis is placed on (i) dealing all options by means of one and the same algorithm, (ii) most
possible convertibility into other thermodynamic approximation concepts, and (iii) employing the simplest
possible mathematical formalism which will yield not only consistent data but guarantees strict accordance

143
144 J. TOMISKA

with the laws of physical chemistry as well, Common characterization of the molar Rmctions Z of K-compo-
nent systems by means of the contributions made by the alloy constituents k which are called partial molar
fhCtiOnS zk, are used:

Z(x,T) = Z xk &(Xk,T), (k = l,..., K) (1)

where Z = Gibbs energy G, enthalpy (heat) H or entropy S. In the phase a the molar partial properties
Zka(T,xa) (Z = G, H, S) are split by common modeling as following:

Zka(Xa,T) = Zkgo(T) + Zkqid(Qa,T) + zksE(Xka,T) + zkqmag(Xka,T), (2)

where Zk’, and Zkid denote the contributions of the pure species, and of the ideal mixing, respectively.
Concerning the free Gibbs energy this is:

G91d(aa,T) = RT h (xka), (3)

In Eq. (2) the terms ZkE and Zkmagare assumed commonly to be due to the intermolecular and magnetic
forces, respectively. As chem&y cannot annul any mathematical law, all arbitrary splitting of sums should
therefore be prohibited. Consequently, employing of a separate magnetic term Zkmagas performed in Eq.(2) is
problematically, because experimental investigations will only yield the deviation between the “ideal”
solution property (Zk”+Zkid) and the molar property Z. This difference will be determined experimentally, and
it is nothing else than the total excess term (ZkE+Zkmag)[OlTom]. Nevertheless, widely known databanks in
literature as offered by the ThermoCalc group are based upon the use of magnetic terms Zkmagas expressed
by the model of Sundman and Agren [81Sun]. ExTHERM also makes also the conversion of binary Zmag-
parameters among various representations possible.
ExTHERM is based upon the multi-component TAPS (thermodynamically Adapted Power Series)
concept of excess properties ZE as developed earlier [86Toml, 98Tom], and summarized in the appendix. It
satisfies all these requirements: (i) Algebraic representing of the activity & as well as of all mixing properties
of all types of alloys, including metallic melts showing short-range ordering. (ii) Using only one common
parameter set for traightforward descriptions of combined temperature and concentration dependences. (iii)
Extrapolation from lower component alloys, (iv) Suitable best fit formulas putting together ail experimental
measurements of an alloy system for performing one algebraic overall best fit, and (v) A very convenient,
simple modular procedure which makes the conversion coefficients independent of actual system data.

Table 1. Options of Temperature Dependencies of Excess Properties Z in ExTHERM

GTZ: TRP_Uers: G = Ho - T.S”

TCA_Uers: G = HD + T.P 6lY = -SO)

GT3: TAP_Uers: G = Ho - T.SD + 01/T

H= Ho + 2.01/T

s= S” t Dl/TS

TCh_Uers: G = Ho t T.P t 01. T. 1 nlT3

H= Ho - D1.T
S= ~-lF3 - 01. cl+1 n<TS>
HT3: TfW_Uers: H = Ho t 2. Dl/T + 3.02/T’

TCh_Uers: H = Ho - O1.T + 3.D2.,TP

 ExTHERM: THE INTERACTIVE SUPPORT PACKAGE 145

Following the TAPS concept, temperature dependencies are described by substituting the TAP-parame-
ters C,,’ (appendix, Eqs. (Al)-(A3)) by the suitable temperature dependent TAP parameters (appendix, Eqs.
(A4)--(A6)), only. In EXTHERM this is performed easily by means of only two switches: (i) Select between
TAT-version and TCA_version as the ThermoCalc formalism is called in the presented tool. (ii) Select the
appropriate temperature dependence from the options as shown in Table 1.

Table 2. Performance Chart of the Conversion Module cMOD.

Step 1: Switch selecting of the options.

1) Input: (i) Keyboard, (ii) ExTHERM data bank, or (iii) individual data file.
2) Data Type: (i) Excess data or (ii) Magnetic data.
3) In case of ErlHERM data bank input: Phase and System.

Step 2: Switch selecting of the desired conversion.

1 TAP from TPl 3 TAP from RKx 5 TAP from RKl 7 TAP from TCA
2 TPl from TAP 4 RKx from TAP 6 RKl from TAP 8 TCA from TAP

9 TAP from WLx 11 TAP from SAN 13 TAP from MRG 15 TAP from SPS
10 WLx from TAP 12 SAN from TAP 14 MRG from TAP 16 SPS from TAP

TAP Thermodynamic adapted power series, TPl TAP expanded in mole fraction xl,
RKx Redlich-Kister expansion in x2, RKl Redlich-Kister expansion in xl,
ICA ThermoCalc Expansion,
WLx Wohl equation in x2, SAN Sanchez approximation,
MRG Margules equation, SPS Simple power series.

3tep3: The TCA parameter set as converted from TAP parameters by means of module 8, and
ltting outs of the control tables of the corresponding excess properties.
system: Fe-Ni /Phase : LI Q
Trst-Page: 1 of 2

New Par: TCf’a Old Par: TCIP

C-l : -iaooo c-1 : - 10500

c-2 : 9000 c-2 : - i 2000
c-3 : -1500 c-3 : - 6000
x-2 Z_ex <TCC)> x-2 Z_ex <TCIP>
0.0 0 0.0 0
0.1 -1058 0.1 -1058
0.2 -2102 0.2 -2102
0.3 -3074 0.3 -3074
Trgt-Page: 2 of 2

New Par: TCfi Old Par: TM’

u-1 : 4.287 C-l .+ -1.850

C-2 : -3.700 c-2 : -2.350
c-3 : 1.262 c-3 : -5 .oso

x-2 Z_ex (TCCI) x-2 Z_ex CTAP)

0.0 0 0.0 0
0.1 -[I. 4922 0.i -I). 1922
0.2 -0.4035 0.2 -0.4035
n? rc.-sc. n _) C?_
146 J. TOMISKA

The codes “GT2” and “GT3” indicate full Gibbs energy descriptions with 2 and 3 temperature terms,
respectively: “GT2” switch on temperature independence of both the heat of mixing II, and the entropy S,
and Gibbs energy shows the simple temperature dependence G = H-TS. In case of arising temperature
dependencies of H and/or S, use of the code “GT3” switches on the corresponding, additional term. Tempera-
ture dependences which require two temperature dependent terms have been observed by calorimetric in-
vestigations only, and therefore it is programmed for the heat of mixing H at the present stage of develop-
ment of ExTherm (code “HT3”), only.
The TAPS concept is designed to employ SGTE data for pure elements as presented by Dinsdale
[91Din] in all application in which the single element contributions of Eq. (2) are required, especially for
computational phase diagram assessments.

3. The conversion module cMOD

The stand alone module cMOD is of interest for everybody. As carried out previously [89Tom2] a
simple modular concept is possible for the algebraic conversion among arbitrary polynomial approximations
of the molar binary excess properties showing essential features: the conversions are independent of the actu-
al values of the system parameters, and the modular design diminishes considerably the number of necessary
conversion formulas. The program flow of the conversion tool cA4OD is performed in three steps as shown in
Table 2. Additional conversion modules are in progress.

4. The Evaluation modules cM3_ @MS, EMF, HTC)

As carried out also elsewhere, power&l evaluation techniques are obtained by putting together all expe-
rimental measurements to one algebraic overall best-fit employing the TAPS concept [81Tom, 82Tom,
95Tom, OlTom]. E.g., if all (K-1)-component boundary systems are well-known, then it is only necessary to
determine the best-fit values of the K-component interaction terms KCTIZ.This tool of modern experimental
thermodynamics makes possible a combined, simple and consistent description of both the temperature, and
the concentration dependence of the excess mixing behavior of multi-component solutions systems by means
of one set of parameters. The concept allows the representation of all types of solutions, and includes even
liquid alloys showing short range ordering [98Tom].
According to the multifarious experimental efforts splitting of the evaluation and best fit tools into
three stand alone modules makes sense: The three modules cA43 are designed specially for experimental
data from (i) mass spectrometric investigations (Kn/rs>, (ii) measurements on the electromotive force (EMF),
and (iii) calorimetric investigations (HTC). Evaluation of all types of experimental data by means of the
modules cM3_ is performed by means of the same few steps as following:

4.1 Preparing the Input Data.

The evaluation modules cM3 makes possible an overall best fit of up to 450 measuring points, split
into up to 28 runs containing on a maximum 70 measuring points may be taken together to one overall best
fit. However, mass spectrometric investigations by means of the ratio techniques are restricted to a maximum
of 15 ratios per run, because in this case the best fit computations require far more space than the other
evaluation techniques.
The best fit tools cM3_ are designed for a variety of input formats as summarized in Table 3. Input
data may be written by means of any text editors, and each data tile is headed by the system name, the phase
code, and the input format code as given in Table 3. In a second line all run specifica have to be summarized
as run identification, temperature, and investigated components. In case of ternary data, also the names of the
parameter files of the three binary boundary systems must be given, Use of the data bank ETD is marked by
the simple code “/ETD”.
 ExTHERM: THE INTERACTIVE SUPPORT PACKAGE 147

Table 3. Table of cM3- Input Formats.

1Code 1Type of experimental data 1Component

lR2 Ln(*(x) /aj(x)) = Q +di/T (j,k = 1,2,3)

IL1 Ln(&(x)) = do +di/T (k= 1,2,3)
EMF GM(x) = -nF.EMFk= do +di/T, transformed into IL1 (k= 1,2,3)
ICD Ln(T.Ik) = do +di/T of all components, transformed (k= 1,2,3)
into IR2

ITC 1 Data points: Ln(&(x,T)) or G”(x,T) or &(x,T) 1 ( k = 1,2,3)

HTC 1 Data points: H’(x,T) 1 (k = 1,2,3)
1GTC 1 Data points: G’(x,T) 1 ( k = 1,2,3) ]

4.2 Interactive Specification of the Actual Best Fit.

The modules cM3_ analyze automatically the input and all possible best fit options are listed on the
screen for interactive specification of the desired best fit performance. Table 4 gives the interactive evalua-
tion screen of calorimetric measurements on ternary In-Sb-Sn alloys, showing short range ordering, and Tab-
le 5 that of mass spectrometric investigations on ternary Fe-Ni-Cr alloys using the intensity ratio technique.
The evaluation window of EMF measurements may be a mixture of the presented two facilities.
The key 7” exchanges the best fit version from TAP to TCA, and vice versa, and the key “F7” makes
possible to restrict the best fit from the total data to a selected number of runs. In Table 4 the key “F6”
switches on an enlarged best fit, containing not only the ternary parameters, but also those of one of the three
binary boundary system (BbS), whereas in case of mass spectrometric investigations these option is not pro-
grammed yet.
The calorimetric data of In-Sb-Sn alloys best fit employ up to three temperature terms (see code I-IT3 in
Table l), and consequently in Table 4 the corresponding selection keys “Fl” to “F4” are shown, but in Table
5 these buttons are suppressed, as vapor pressure lines require strictly T-dependence as coded with GT2 in
Table 1. Then the number of adjustable parameters are selected. As vapor pressure lines will not specify eva-
luation temperatures, the mean best fit temperature is required in Table 5 also.

Table 4. Interactive Evaluation Screen of Module cM3 HTC for Best Fit
of Calorimetric Measurements on Ternary In-Sb-Sn Alloys, Sho&rg Short Range Ordering.

System : In-Sb-Sn /Phase : LI Q /HT3-Paranrter /HTC-Input

Default: Overall Best Fit of all 3 Runs

Total of Experinental Points: 34

TCIP_Uers ion

Best Fit Options:

F6 = Include a BbS Fit
F7 = Run Select ion 4nax. 3)
U = TCCI_Uers

Tenperature Dependence of Paraneters:

Fl = C*
F2 = C* + 1 T-Tern
F3 = C* + 2 T-Terns
F4 = Co + Indiu. T-Terns

HT3: TFIP_Uers: H = HO + P.Dl/T

Select No. x-Paraneters <O to 4; df It= 1): 2
I
148 J. TOMISKA

Table 5. Interactive Evaluation Screen of Module cM3_KMS for Best Fit of Mass Spectrometric
Investigations on Ternary fee Solid Fe-Ni-Cr Alloys by Intensity Ratio Technique.

srsten : Fe-Ni-Cr /Phase: FCC /GT2-Paraneter /IR2-1nrrut

IC-Rat io: Fe&k-

Default : Overall Best Fit of all 2 Runs

Total of Experinental Points: 16

TfiP_Uersion

For IR2 No BbS Fit srost-anned

F7 = Run Select ion <max. 2)
u = TCCI_Uers i
GT2: TCIP_Uers: G = HO - T.SP
Select No. x-Paraneters CO to 4; df It= 1): I

Systen Temperature CKI Cdflt: 1650) = 1650

I

4.3 Control of Best Fit, and of the Mixing Behavior.

All best fit computations are performed by means of a really quick and sophisticated algorithm which
makes possible regression of both extreme large as well as small values yielding the same high quality of
accuracy: All best fit terms will be transformed to a value of about unit by multiplying with automatically
scaled zooming factors, and the resultant best fit parameters will be re-transformed finally by means of the
same zooming factors to correspond to the original input data. For the proper best fit algorithm see [UTom].

FIG 1.
Best Fit Control Diagram of Calorimetric Measurements on Ternary In-Sb-Sn Alloys, Showing Short Range
Ordering.

Step 5:
Best fit control.

k = In, Sb or Sn.

1- Explicit value; or
2- Ratio of a selec-

Actual curser values:

Point identification:
 ExTHERM: THE INTERACTIVE SUPPORT PACKAGE 149

Fig. 1 illustrates the graphical best fit control of liquid In-Sb-Sn alloys. The diagrams are plotted in co-
lor, of course, to differentiate easily among the experimental points of the various runs. In the center of the
right column there is the reading of best fit codes (number of C,-parameters (in Fig. 1: CN=2), and number of
temperature terms (in Fig. 1: CT=2)), and of the plotted runs. The cursor gives the mole fraction xk of the
right hand component as well as the y-value of the plot (below the diagram). In case of a ternary system the
mole fractions of all three species are shown additionally (right column, bottom). A point identification key
makes possible additional reading of the input file name, the identification codes of the selected measuring
point, and the exact (x,y)-input values of it (below and right of the cursor reading)

Table 6. The Parameters of Molar Excess Properties

of Ternary In-Sb-Sn Alloys Showing Short Range Ordering.

systen : In-Sb-Sn /Phase: LIQ /HT3-Paraneter /HTC-Input

Parameter values fron a total of 34 points. /Tenaerature: 944 K

" C_HXS C-H" C-H-I

J/W01 J/no1

I I5090 68760 -25.33 E+6

2 -39520 -253700 101.1 E+6

In-Sb /GT3 C_Hxs C-H' C-S' C_Di

1 -I4960 I6650 0 -14.92 E+6
2 1432 -40750 0 19.91 E+6
3 4771 37970 0 -15.67 E+6

Sb-Sn /HT3 C_HXS C_HQ C-H-1

i -4317 360.6 -2.208 E+6
2 890.7 -470.1 642300
3 -22300 -76600 25.63 E+6
4 37790 165 100 -60.09 E+6
5 -16540 -94290 36.70 E+6

Sn-In /HT3 C_HXS C-H’

i -703. a -703 .a
2 -190.2 -140.2

Table 7. The Parameter Set of Molar Excess Properties of Ternary FCCSolid Fe-Ni-Cr Alloys.

Parameter values fron a total of I6 points. 1Tenperature: I650 K

" C_GXS C-H' c_so

J/MO1 J/no1 .JhrPlK

0 25320 -10400 -21.65

1 6109 39320 20.13

Fe-Ni /GT2 C_Gxs C-H* c-s 0

0 0 0 0
i -4990 -6300 -o.aOoo
2 -10220 -I2200 -1.200
3 3720 -7500 -6.800

Ni-Cr /GT2 C_Gxs C-H' c-so

0 0 0 0
1 -14490 -2I550 -4.280
2 5719 59790 32.77

Cr-Fe /GT2 C_Gxs C-Ho C-S'

0 0 0 0
I I810 I2700 6.600
2 1275 -4500 -3.500
150 J. TOMISKA

For ternary systems there is an option of interactive comparing arbitrary cuts of constant mole fractions
xi&~_, with the experimental data: Enter either an explicit ratio value or select an arbitrary measuring point
from the input control list for this purpose. The control diagram will then give a plot the resultant excess
property curve together with all experimental points not scattering within a pre-selected deviation limit from
this cut. On the top of the right column the number of these measuring points together with the deviation
limit is shown. Fig. 1 shows that the no measuring point scatters more than 2% from the overall best fit
curves, and ternary short range ordering may be described excellently by means of the TAP concept.
All information as plotted in the diagram may be printed as best fit control lists as well. Tables 6 and 7
gives the resultant table of parameters. In Table 6 we notice that the parameters of the integral and partial
heat of mixing as well as the heat capacities of alloy systems showing ternary short range ordering (H” and
IT1 parameter) is really simple. Data of all three binary boundary systems are from the data bank ETD: In-Sb
and Sb-Sn are showing short range ordering as well, whereas binary Sn-In alloys show normal liquid
behavior. Table 7 shows the description of total mixing behavior of ternary fee solid Fe-Ni-Cr alloys has
been obtained (Z, Zk, Z = G, H, S; k = Fe, Ni, Cr; a& Again, the data of all three binary boundary systems
are from the data bank ETD. The parameters COare the calibration constants necessary for best fit of the mass
spectrometric intensity ratio data.
Controlling of the mixing behavior may be performed by means of the menu of data tables and
diagrams as described with the data bank module cETD. As mentioned already in the introduction, compa-
ring with graphs produced by means of an imported, second set of parameters is possible in all types of
diagrams except the ternary triangle plots.

5. The Data Bank Module cETD.

The parameter sets of the data bank ETD are based on (i) the binary metal alloy data as compiled by
Hultgren et al. [73Hul], and (ii) literature data of experimental investigations after Hultgren’s supplement.
Consequently the data bank module makes it possible to differentiate Hultgren’s compilation data and newer
experimental results as can be seen from Table 8. The key “5” yields the data sets reproducing the values as
comiled in [73Hul]. The TAP parameters of this table have been converted mainly from the parameters as
presented by Williams [91Wil], only in case of a minor number of binary systems best fit improvements have
seemed to be necessary.

Table 8. The Research Options of the Data Bank Module cETD.

Code Type of Table

1 = G-Bank All binaries with G’-TAP-data
2 = H-Bank All binaries with H”-TAP-data
3 = G-New-Bank All binaries with experimental G”- TAR-data newer than Hultgren’s supplement
4 = H-New-Bank All binaries with experimental H’- TAR-data newer than Hultgren’s supplement
5 = Hult-Bank Binary data from Hultgren’s supplement represented by TAP concept.
6 = G_Mult-Bank All binary G’-TAP-data listed in ETD (multiple parameter sets)
7 = HMult-Bank All binary H’-TAR-data listed in ETD (multiple parameter sets)

Fl = G-Bank All temaries with G”-TAP-data

F2 = H-Bank All ternaries with HO-TAP-data
 ExTHERM: THE INTERACTIVE SUPPORT PACKAGE 151

Table 9. Number of Systems Stored in the Data Bank Module cETD

(In Brackets: Nuniber of Systems Showing Short Range Ordering).

Table 10. Main Menu of the Data Bank Module cETD.

Default: Molar Excess zxs T.1.R.

Term : 1650 K cut : Fe/C+ = 2.28 x: x_Ni

F2 = Sys_Uieu Fl = F~mula~/Hel~

F = Fct_Plot 3 = Tri_Plot E = End of PRG

Z = Z_Partial P = paraneter D = Data Quest
X = Nixing I = NO T.I.CI. Cl = Quit Menu
c = cut Select T = Temperature S = Save Data 2001-lo-01/15:42

Table 11. Molar Excess Properties as Tabulated by the Data Bank Module cETD
(FCC Solid Ternary Fe-Ni-Cr Alloys at 1650 K Along The Cut of x&cr = 2.28).

Srst~n: Fe-N i -Cr /Phase: LIQ /GT2-Parameter /ETD_E-Inout

Parameter ualues /Tenmsrature: 1650 K

n C_Gxs C-Ho c-s*

J/n01 J/n01 J/no 1K

0 0 0 0
1 10400 43700 20.18

Fe/Cr = 2.28
X-Fe x_N i GXS HXS SXS a_&- a_Fe a_N i
J/MO1 J/no1 J/no 1K

0.70 0.0 546 -1427 -1.2 0.324 0.717 0

0.63 0.1 -240 -1168 -0.5 0.293 0.642 0.0592
0.56 0.2 -983 -1253 -0.1 0.261 0.565 0.124
0.49 0.3 - 1630 -1580 0.0 0.225 0.482 0.198
0.42 0.4 -2136 -2037 0.0 0.186 0.395 0.285
0.35 0.5 -2455 -2503 -0.0 0.145 0.307 0.387
0.28 0.6 -2549 -2646 -0.1 0.105 0.222 0.503
0.21 0.7 -2380 -2923 -0.3 0.0691 0.146 0.630
0.14 0.8 -1917 -2565 -0.4 0.0387 0 .OG26 0.761
0.07 0.9 -1132 -1667 -0.3 0.0155 0.0337 0.888
0.00 1.0 0 0 0 0 0 1.00
152 J. TOMISKA

All other facilities of Table 8 are split into key codes for systems with parameter sets describing till
mixing behavior (G, H, S, s), and those where heat of mixing information available. The keys “1” and “2”
in Table 8 yield the upgraded data banks, whereas the keys “3” and “4” give the corresponding lists of the
upgrading systems, that means of all experimental data after Hultgren’s supplement, and already included
into the data bank ETD. The keys “6” and “7” shows all sets of binary alloy parameters as listed in the data
bank ETD as a combination of the keys “S’, “3” and “4”. The tunction keys “Fl” and “F2” yield the cor-
responding lists of stored ternary parameters.
The newer experimental excess Gibbs energy data are from literature with most of them due to the
mass spectrometric investigation of the author, whereas the majority of in-cooperated calorimetric data have
been reported by several authors, most of them from members of the research group of B. Predel (MPI Stutt-
gart; see [98Hej]).
The main menu of the data bank module cETD is given in Table 10. The key “C” selects an individual
ternary cut of constant mole fractions xk/xk_i, and the key ‘7” for temperatures. The ternary interaction
(T.I.A.) may be exclude by the switch “I” to illustrate its contribution. The key “X” allows exchange between
dealing with the molar excess properties ZE, and the molar mixing properties Z”, whereas the key ‘77’
switches between molar integral, and molar partial properties.
The key “F” yields the menu for squared diagrams, enabling plots both of a single temperatures as well
as of temperature ranges. Comparable systems can be imported by means of the menu switched on by the
function key “F2”. The key “3” switches on the corresponding menu for ternary triangle plots, which are
restricted to single temperatures, and no comparable graphs are possible.
The parameters as presented in Tables 5 and 6 are shown by means of the key “P”, and additional
pressing the “return’‘-key yields then the lists of the molar mixing behavior as presented in the upper part of
Table 10 for fee solid ternary Fe-Ni-Cr alloys at 1650 K along a cut of constant mole fraction xre/xc, = 2.28.
Fig. 2 gives, as an example, the black-and-white version of the colored triangle plot of the heat of mixing of
fee solid in the Fe-Ni-Cr ternary alloys. Zooming as well as cursor reading are also possible. The selected cut
ofXFek!r = 2.28 is plotted by means of the dashed line. Minimum and maxl’mum are indicated.
FIG. 2
Example of Triangle Diagrams of Data Bank Module cETD.
The Heat of Mixing of Fee Solid Fe-Ni-Cr Ternary Alloys.

EXThrrn-GrlJOh:
m “ax 4328 Fe-Ni-Cr
0 nin -3960
cnntnur Llncs:

Hxs /16xl Y
-1947
.J/"Ol
CN=I; CT=2

B.BB 0.28 8.48 8.68

Fl = HELP X = x_Ni = 0.5171
 ExTHERM: THE INTERACTIVE SUPPORT PACKAGE 153

5. Discussion

It has been demonstrated that conversion among various model parameters of the molar mixing
properties may be performed conveniently. The presented interactive evaluation tools cM3 may increase the
quailty of best fit, especially, for ternary experimental data, and will diminish considerably both, experimen-
tal effort as well as evaluation time. This work demonstrated also the concept of the data bank ETD really
describes all types of mixing behavior of metal alloys - even short range ordering by means of the same
formalism. This simplifies considerably the handling of thermodynamic mixing data. As can be seen from
Table 9, the data bank ETD contains free Gibbs energy data of over 150 binary phases and of some ternaries,
and heat of mixing data of additional 25 liquid binaries. The presented support package ExTherm may
encourage more experimental investigations as well as more use of thermodynamic data for solving problems
arising in every day research and industry work.
Two additional modules not presented in this work exist: One module, cM_REG, is configured for
simple best fitting of arbitrary (xj,yj)-data sources. Data input by means of the keyboard as well as file-input
are possible. Various input conversion modes exist concerning vapor pressure data. The second module,
cPD Bin, makes interactive computations of binary phase diagrams applying the generalized Newton method
to s&e the equations of thermochemical equilibria with arbitrary high accuracy. Starting points of the itera-
tion algorithm may be determined by means of interactive plots of the tangents of the Gibbs energy curves of
phases. For all software modules contact the author.

Acknowledgments

The author would like to thank em. Prof. Dr. B. Predel, Dr. R. Luck, and Dr. I. Arpshofen, all with the
MPI Stuttgart, for so much help, interest, encouragement, cooperation and discussions throughout the many
years of development. Grate&l acknowledgment is made to Mag. J. Theiner (Vienna) for essential parts of
the graphic unit and other usetil help, Dr. K. Kopecky (former Vienna) for programming the ternary triangle
plots as well as Dr. H. Wang (former MPI and Vienna), Mag. J. Havrankova (Masaryk Univ. Brno), and
Mag. D.M. Hejze (Vienna) for preparing the input data of the best fits of emf and calorimetric investigations.

References

73Hul R. Hultgren, P. D. Desai, D. T. Hawkins, M. Gleiser, and K. K. Kelley: Selected V&es ofthe
Thermodynamic Properties of Binary Alloys, Am. Sot. for Metals, Metals Park, OH, 1973.
81Sun B. Sundman and J. Agren, J.Phys.Chem.Solids, 42 (1981) 297-301.
81Tom J. Tomiska, CALPHAD, 5 (1981) 81-92.
82Tom J. Tomiska, J. High Temp.-High Pressures, 14 (1982) 417-422.
85Tom J. Tomiska, CALPHAD, 9 (1985) 15-28.
86Tom 1 J. Tomiska, CALPHAD, 10 (1986) 91-100.
86Tom2 J. Tomiska, CALPHAD, 10 (1986) 239-252.
89Tom J. Tomiska, nermochim. Acta, 151 (1989) 145-158; 159-170.
91Din A.T. Dinsdale, CALPHAD, 15 (1991) 317-425.
91Wil R.O. Williams, CALPHAD, 15 (1991) l-10.
95Tom J. Tomiska and H. Wang, Ber. Bunsenges. Phys. Chemie, 99 (1995) 633-640.
98Hej D.M. Hejze, Diploma Theses, Univ. of Vienna, Vienna, May 1998.
98Tom J. Tomiska, Ihermochim. Acta, 314 (1998) 145-153.
OlTom J. Tomiska, J. of Mining andMet., (2001) 9-28.
154 J. TOMISKA

Appendix: The Multi-Component TAP Concept of Excess Quantities ZE.

Al. Concentration dependence ZE = ZE(x)

Al.l. Binary systems (i,k: constituents; N: number of adjustable parameters j,kC,z; j*Ci’: regular solu-
tionjkCzz: sub-regular solution):

LkzE(x) = xj c LkC,Za”, (n=l,...,N) (Ala)

“kzEk(xk)= Xj’ c J’kC,Zxk”-‘n, (n=l,...,N) (Alb)

j’“z”j(Xj) = Xk’ c j,kCnz xk”*( I -n+r&). (n=l,...,N) (Ale)

A1.2. K-component systems (K = 3,4 ,...)

A1.2-1 Extrapolation formulas:

(K-l)BsZE(~)= C [all ZE-terms with (K-l) components] (A2)

Ternary systems: 2mBsF(x) = ‘f2f(x) + 2S3f(x) + 3~‘~(x).

A1.2-2 K-component interaction included (“Cj”: adjustable parameters of the K-component interaction):

ZE(X) = “ZE(X) + (K-l)ssZyX), (ma)

KZE(X)=x I... XKIKCIZ+ (KC2zXI+ . ..+ KCZK+lXK)+ . ..I. (A3b)

Ternary systems: f(x) = !f(x) + 2-Bs~(x);

‘f(x) = x1x2x3(cl’+ (+2,” XI+ ‘c3’ x2+ ‘c,” x3) + . ..I.

A2. Temperature dependence ZE = ZE(T)

A2.1 Heat of mixing HE(T) = Ho + 2.Dr.T-’ + 3.Dz.T’ +... (CHon,Di,“: adjustable parameters):

C”, = CHo,,+2.D,,“.T-’ + 3.D2,n.T.’+ (A4)

A2.2. Excess Gibbs energy GE(T) = H(T) - T S(T):

HE and SE temperature independent (C”,, C’,: adjustable parameters):

C’,(T) = CH, - T C’,. (As)

HE and SE temperature dependent (Di,“: adjustable parameters):

C’,(T) = CHo,II’T- CSon + D1,nT-’ + D2,”T-2... (W

A3. Combined temperature and concentration dependence ZE = Z’(x,T)

For all types of solution systems: Substitute in Eqs. (Al) - (A3) the TAP-parameters by the correspon-
ding temperature dependent TAP parameters (Eqs. (A4) - (A6)), only. Applicable not only for all of metal
alloys including metallic melts showing short range ordering, but also for all other types of multi-component
systems. Consistency between the excess properties is guaranteed. Interchange of the components may be
performed by mean of the conversion module cA4OD (TPl from TAP and TAP from TPl).