Solidification of Hypereutectic Thin Wall Ductile Cast Iron

Abstract

Abstract. Hypereutectic ductile iron was cast in green sand moulds with four plates with thickness of 1.5, 2, 3 and 4 mm in each mould. Temperatures were measured in the 3 and 4 mm plate. The temperature curves showed that eutectic solidification was divided into two stages: primary and secondary eutectic solidification. The first stage, which was relatively short, had none or very little recalescence. Further under cooling, followed by reheating during recalescence, was necessary to initiate the second part of the eutectic solidification. Both the secondary under cooling and recalescence was larger in the 3 mm plates than in the 4 mm plates. All 1.5 mm plates contained carbides but the other plates solidified without carbides. Metallographic examination showed two populations of graphite nodules. A small group of nodules was larger than rest of the nodules. Colour etching revealing the segregation of Si showed a higher Si content in the ferrite around the larger nodules compared to the ferrite around the rest of the nodules. This indicates that solidification took place along the following path: The solidification starts with nucleation and growth of primary graphite nodules. This probably starts during the filling of the mould. The primary nodules act as nuclei for austenite. As austenite easily nucleates on graphite the temperature will be the same for the 3 and 4 mm plate for the first part of the eutectic solidification. This first part of the solidification ends when concentration of carbon around austenite dendrites is too large and new nodules have to nucleate and grow. The larger under cooling for the 3 mm plates compared to the 4 mm indicates that the nucleation of new nodules is governed by kinetics even in very well inoculated melts.

Experimental validation of error in temperature measurements in thin walled ductile iron castings

“…They have also made simulations of how the TC influences the temperature locally in thin walled ductile iron castings. 8,9 The simulations showed that the shape of the measured cooling curve is correct in comparison with the cooling curve in an unaffected casting. There was however a 15 to 20 uC parallel shift towards lower temperatures on the measured cooling curve in plate thicknesses of 2 to 4 mm.…”

“…This corresponds very well with what was shown by the numerical simulations. 8,9 Discussion Effect of segregation During solidification there can be some segregation of Si to the austenite. This would lower the Si content of the liquid and by that increase the T EW .…”

“…It is intended that the TC is placed in the hot spot of the casting but the simulation showed that the measured temperatures were 10-15uC lower than the temperature in the hot spot. Pedersen and Tiedje 7,8 have examined the microstructure around a TC placed in thin walled cast iron. It was not possible to detect any difference in microstructure around the TC compared with the rest of the casting.…”

“…This error in the measured temperature depends especially on the thermal conductivity of the TC wire while the heat capacity of the TC wire and the thermal properties of the ceramic tube only had minor influence on the measurement error. 8,9 In ductile iron it can be very difficult to experimentally confirm the results of the numerical simulations. This is due to that the undercooling of ductile cast iron can be very high during the solidification and it will not be possible to detect if the measured cooling curve is a result of undercooling during solidification or error in the temperature measurement.…”

Undercooling and nodule count in thin walled ductile iron castings

“…The sample rate for temperature measurement was 500 Hz and every 100 values were averaged to reduce noise giving a time increment of 0?2 s. The measured temperatures were corrected for the measurement error caused by the response time of the thermocouple. [3][4][5] The temperature was measured in all plates except for the 1?5 mm plates.…”