The following results have been obtained through the study made on the subject. (1) Large non-metallic inclusions distribute mainly at the bottom part of an ingot. Whose compositions are (Fe, Mn) O, MnO-Al203 and Mn-Silicate. From the fact that larger inclusions have less value of MnO/FeO, coming closer to the composition of scum, it is estimated that large inclusionsstem partially from scum through rimming action. Since Ca is detected in the aluminate and silicate, they are considered to have origin in slag, refractory and deoxidation products. (2) With the “ingot parameter” Y-which is defined as ingot width/ingot thickness+0.02 (ingot height-width), relations between ingot dimensions and inclusion distribution and its composition can be clearly shown: when Y becomes bigger, total amount of inclusions and amount of large inclusions (more than 100μ) increase and the inclusions have less value of MnO/Fe0. The parameter Y is considered to have some relationship with thickness of viscous laver at the bottom part of ingot and with state of flow of rimming steel in the mold. Y therefore, is a good index to explain conditions of scum movement along rimming action, scum trap and precipitation of (Fe, Mn) O. (3) With shorter capping time, increase in total amount of inclusion is observed, but the amount of large inclusions decreases. This result is explained as follows: When capping time becomes shorter, more oxygen, which would be otherwise eliminated through rimming action as CO, remains to react with iron and manganese which results in higher amount of nclusion. Growth of inclusions is hindered as rimming action is terminated earlier, which results in less amount of large inclusions.
Some experiments were carried out by means of X-ray diffraction, electrolytic isolation and potentiostatic etching technique, in order to make clear the relation between precipitated carbides and σ-phase in 25Cr-20Ni heat-resistant cast steel. The results obtained were summarized as follows: (1) When electrolytic residues contain more than 5wt% of σ-phase, it can be identified by X-ray diffraction method. (2) For the electrolytic isolation of σ-phase, 45wt% FeCl3 aq. solution showed a favourable result as compared with 10vol% HCl-C2H5OH solution. (3) Not only M7C3 is distinguishable from M23C6 carbide but also σ-phase from M23C6 by means of potentiostatic etching in 10 n-NaOH solution. At 200mV vs. Hg-HgO reference electrode M23C6 carbide is etched more rapidly than M7C3 carbide or σ-phase, at 500mV M23C6 and M7C3 carbides, or M23C6 and σ-phase are simultaneously etched, and at 600mV, σ-phase etched most rapidly. (4) From these investigations it was found that σ-phase may be possible to precipitate even in the prolonged heat treated cast steel containing 0.35% carbon.
The temperture rise due to hot rolling was qualitatively evaluated by rolling plates with inserted thermocouples. In rolling at high reductions, thermocouples were shortcircuited by oxidation scale in the vicinity of the junction. However, this was found to exert no influence on temperature readings. The temperature rise ranged between 10°C and 45°C for the conditions studied (1150-800°C, 10-40% reduction), and showed a better correlation with mean flow stress rather than with amount of plastic work in rolling.