Influence de la vitesse de refroidissement sur l'évolution de la macroségrégation dans les lingots de grande taille

Abstract

Large-size ingots (40-ton) are exposed to chemical heterogeneities after solidification called macrosegregation. Such heterogeneities have to be considered and monitored in order to better predict the evolution of mechanical properties and insure the required high quality of the products. The experimental identification of macrosegregation is cost and time-consuming, fostering the use of numerical tools such as FEM software for the prediction of macrosegregation patterns in large-size ingots. However, a proper prediction requires accurate material properties as input parameters. As part of a partnership with Finkl Steel (Sorel), a research project was carried out to improve macrosegregation prediction in large-size ingots. Such improvements are based on considering the influence of cooling rate on transition temperature and secondary dendritic arm spacing (SDAS). Hence, two experimental methodologies were developed to measure the behaviour of such parameters with cooling rates during solidification. A thermal analysis with a Netsch 404 and a microstructural analysis with a bifocal microscope were carried out. Two numerical FEM models were developed and implemented in the FEM software, TherCast. The first one simulates unidirectional solidification and the second one predicts macrosegregation in a 40- ton ingot. Both were used to analyse the influence of adaptive parameters as a function of the cooling rate on macrosegregation patterns. Experimental results were used to develop a relation describing SDAS evolution as a function of cooling rate during solidification. Then, the behaviour of the transition temperatures was measured emphasizing an insignificant variation with the order of magnitude of the range of cooling rates. Concerning the numerical analyses, the unidirectional solidification model showed a negligible impact of adaptive SDAS (between 50 and 350 μm) on macrosegregation patterns. Finally, some recommendations were developed to simulate macrosegregation in the large-size ingots.

Date	24 Aug 2018
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Mohammad Jahazi (Supervisor)