Thermodynamic modelling of carbides in multicomponent systems : theoretical and experimental approach

Abstract: This thesis concerns thermodynamic modeling of carbides in multicomponent systems. Focus has been made on systems interesting for cemented carbide production but the results are also useful for many other application were the material consist of different carbides, for example tool steels/high speed steels. The Co-W-C system forms the basis in cemented carbide production. An accurate thermodynamic description of this system is therefore crucial for extrapolation into higher order systems. New experimental results on the liquid+fcc+graphite+WC and liquid+fcc+WC+M6C equilibrium temperatures, that has recently been published, shows that these equilibrium temperatures are higher than the values used in the available assessment of Co-W-C. Since an accurate description of these equilibrium temperatures are very important for production of cemented carbides and when extrapolating into higher order systems a reassessment of the Co-W-C system is presented. Cr is sometimes deliberately added to lower the melting point, reduce grain growth and/or increase corrosion resistance in the production of cemented carbides. When adding chromium there is a risk of forming an unwanted M7C3 carbide. It is therefore of great interest to know the stability of this carbide. New experimental results on the maximum solubility of Co in the M7C3 is presented as well as a new thermodynamic description of the Co-Cr-C system which accurately describes the solubility of Co in the M7C3 carbide in the temperature range 1373- 1723 K. The assessment of a system, and the determination of Gibbs energy functions, is straightforward when reliable and consistent thermochemical and phase equilibrium information is available. However, reliable experimental information is often lacking or does not give a unique set of model parameters, and therefore different strategies to estimate information have been developed. In the present work the excess energies for A1-xBxC mixed carbides (where A and B are metals) have been calculated using ab-initio calculations, for 14 systems. A thorough comparison has been made with experimentally assessed excess energies. The comparison shows that ab-initio calculations can be used to predict the sign, magnitude and symmetry of the excess energy for A1-xBxC mixed carbides. The calculated excess energies have also successfully been used to describe several AC-BC systems where the experimental information does not give a unique determination of the excess energy in traditional CALPHAD modeling. Experimental work has also been done on the C-Co-Ti-V-W-Zr system in order to determine the extension of the miscibility gaps in TiC-ZrC and VC-ZrC into the (TiC or VC)-ZrC-WC system. Thermodynamic calculations were used to design samples that will form a miscibility gap in equilibrium with liquid, WC and graphite. Samples were produced from powder and sintered for 1 week in controlled atmosphere at 1300, 1410 and 1500 °C. From the microstructure it could be concluded that the samples form a miscibility gap in equilibrium with liquid, WC and graphite at all temperatures. The composition of the MCx carbides was measured using an analytic SEM. The new experimental information was used to assess the thermodynamic description for the TiC-ZrC system.