COST 507 - Repositório Aberto da Universidade do Porto
COST 507 - Repositório Aberto da Universidade do Porto
COST 507 - Repositório Aberto da Universidade do Porto
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
approximate agreement with de Boer, would be needed in order to describe the<br />
L+Al 3 Ti+Al 3 C 3 +TiC equilibrium in the Al-Ti-C system [96cJon]. To get a heat of<br />
formation of TiC they considered reasonable they found it necessary to use five<br />
independent parameters for the description of the liquid phase. Jonsson and<br />
Albertsen et al. used only two. In a recent assessment Dumitrescu accepted Jonsson<br />
description of C P and S 298 but for A f °ίί she chose a value as high as possible (i.e.<br />
less negative) with respect to measurements, having the same arguments as Seifert.<br />
4.2 Binary system Ti-C<br />
So far we have only considered stoichiometric TiC. We may also like to present the<br />
whole of the binary Ti-C system. Fig.8 shows a comparison between three of the<br />
existing descriptions of the Ti-C system [96bJon], [97Dum] and [96Sei].<br />
The solution properties of δ-TiC..,. according to Jonsson, Dumitrescu and Seifert et<br />
al. are fairly similar but Albertsen et al. have a different description because they<br />
wanted to fit their own <strong>da</strong>ta in preference to several other studies. For the liquid phase<br />
Seifert et al. used five parameters in order to fit information primarily concerning<br />
other phases. That may be a <strong>da</strong>ngerous strategy. For example, their liquidus line<br />
shows two inflexion points close to the Ti side. Jonsson and Dumitrescu used only<br />
two parameters for the liquid. Albertsen et al. and Seifert et al. also differ from<br />
Jonsson and Dumitrescu on the solubility of C in the β phase. Admittedly, the<br />
information on the solubility of C in α and β may not be very well established<br />
experimentally but it <strong>do</strong>es not seem as a natural choice to trust the information from<br />
Ca<strong>do</strong>ff et al.[53Cad] on the α phase, Fig.9, but not on the β phase, Fg.10.<br />
All the assessments show a very strong temperature dependence of the solubility of C<br />
in the α phase, in agreement with information from Ca<strong>do</strong>ff et al. By using only one<br />
adjustable parameter Jonsson and Dumitrescu reproduced a similar temperature<br />
dependence for the β phase, in agreement with information from Ca<strong>do</strong>ff et al.<br />
Albertsen et al. used two independent parameters and calculated a solubility almost<br />
independent of temperature. Seifert et al. did the same but then they had to use a very<br />
large third Redlich-Kister parameter in the liquid in order to adjust the eutectic<br />
composition. Dumitrescu has now reassessed the Ti-C system using her higher value<br />
of A f "H m (29S) for TiC and got a reasonable result showing a variation of the<br />
solubility in α with temperature.<br />
It is recommended that one should use the assessment by Dumitrescu, which is very<br />
close to Jonsson's, except for the choice of a different value of A f °// m (298).<br />
5. Application of the descriptions of TiC and Ti-C to higher order<br />
systems<br />
G m of TiC and the description of the Ti-C system should then be tested, as it was<br />
<strong>do</strong>ne for TiN, against information on higher-order systems of direct interest.<br />
5.1 Ti-C-N<br />
- 180 -