Prediction of volume fraction of primary austenite at solidification of lamellar graphite cast iron using thermal analyses

• Published in: Journal of thermal analysis and calorimetry 2016-04, Vol.124 (1), p.215-225
• Main Authors: Diószegi, Attila, Diaconu, Vasile-Lucian, Fourlakidis, Vasilios
• Format: Article
• Language: English
• Subjects: Algorithms; Analysis; Analytical Chemistry; Austenite; Carbon equivalent; Cast iron; Chemistry; Chemistry and Materials Science; Coarsening; Cooling; Cooling conditions; Correlation; Forecasting; Good correlations; Graphite; Inorganic Chemistry; Iron; Iron compounds; Lamellar graphite cast iron; Measurement Science and Instrumentation; Microstructure formation; Numerical algorithms; Phase diagrams; Physical Chemistry; Polymer Sciences; Primary austenite; Solidification; Technical applications; Technology application; Thermal analyses; Thermal analysis; Thermal properties; Thermoanalysis; Thermocouples; Volume fraction
• ISSN: 1388-6150; 1588-2926; 1588-2926; 1572-8943
• DOI: 10.1007/s10973-015-5158-z

Lamellar graphite cast iron was investigated with carbon equivalents varied between CE = 3.4 and 4.26, cast at various cooling rates between 0.195 and 3.5 °C s −1 covering the limits used for technical applications in the production of complex-shaped lamellar graphite cast iron. Registered cooling curves displaced in two positions in the casting were used to predict the solidification and microstructure formation mechanisms. The predicted volume fraction of primary austenite was compared with the fraction of primary austenite measured on colour micrographs with the help of image analyses. A good correlation has been obtained for medium and slow cooling conditions, while a less good correlation at fast cooling condition was attributed to the used protective environment to preserve thermocouples. The observed fraction and the predicted fraction of primary austenite were in good correlation and followed a consequent variation dependent on the carbon equivalent. Furthermore, the quality of the prediction was dependent on the used numerical algorithm involving cooling information from either one or two thermocouples.