Our workmen believe that cast iron melted in an electric furnace cools more rapidly than cupola iron. The belief is of course hardly comprehensive from the scientific stand point. The present research was undertaken to see if it is true or not by making thermal analysis of two differently melted irons in a large scale, using a hand ladle containing 75 kgs of molten iron. Various precautions were made to get the same cooling condition in each measurement, i. e. the size of the ladle, its thickness of lining, the drying of the ladle, the temperature of it just before the reception of molten iron, the teeming temperature of iron into the ladle, and the position of the hot junction of a platinum and platinum-rhodium thermo-couple having a length of two meters. The mean cooling velocity of that part of cooling curves between 1300°C and 1250°C of molten iron was determined. The rate of cooling above 1300°C was excluded as the thermal equilibrium between iron and the ladle, the ladle and the air was not settled, and the rate below 1250°C was also excluded owing to the heat of evolution of primary and eutectic crystallization. The measurements were made with respect to 4 heats of electric iron and 12 heats of cupola iron.
  Comparing the rate of cooling we found out that electric iron had higher cooling velocity than cupola iron contrary to our expectation. The fact can only be explained by the evolution of more gases in electric iron which received higher degree of superheating than cupola iron. The gases evolved agitate molten iron by the invigoration of convection current, which accelerates the cooling of iron in the ladle. In addition to this it may be suspected that the liability of oxide film formation on the surface of molten electric iron has made our workmen to have an impression that it cools more rapidly than cupola iron, because the fluidity of iron is thereby decreased irrespectively of its true temperature. It may be concluded that if the electric furnace melting of cast iron was more skillfully carried out than the present practice, we shall have surely the same cooling rate as that of cupola iron.

View full abstract

  The investigation is intended to find how the molten metal fills up the vertical sand mould. Fine wire of high electric resistance is inserted in a mould space where the metal is to be fiiled, and the change of wire resistance caused by rising up the metal is recorded using Braun tube oscillograph. The experiment has been carried out as to aluminum cast at 750°, and the following results are obtained.
  (1) With same gate, runner and pouring basin, the amount of metal passing through the ingate varies with the size of casting attached to it.
  (2) When the casting is comparatively large enough to the ingate, the metal will flow steadily. But when it is not so large, the metal will flow less steadily.
  (3) The metal rushed into the mould vacancy from ingate runs straight to the end of it and rises up there, but proper design of the mould can make the metal flow steadily.

View full abstract

  Iron castings in Hokkaido have more defects than those in Honshu, we supposed that a cause of such unsoundness came from properties of moulding sands used in Hokkaido, so at first we studied properties of new sands.
  The results are as follows ;
 (1) On the rational analysis, quantities of clay substance and feldspar in sands are large, and silica is less.
 (2) Refractoriness and sintering points are low.
 (3) Shapes of sand grains are comparatively angular.
 (4) Moisture content which sands in Hokkaido obtain the maximum permeability or compression strength is generally larger than that in Honshu.

View full abstract

  By comparing the shrinkage curve of nodular cast iron produced by magnesium and cerium addition with that of ordinary cast iron, the following results were obtained :—
 1. The primary expansion of nodular iron is larger than that of ordinary iron. This is a reason of large amount of gas dissolved into the molten metal. Therefore magnesium or cerium dissolved into the molten metal in the state of gas and this gas prevents the free growth of graphite and induces the graphite to nodulize.
 2. The growth of graphite is slower in nodular iron than in ordinary iron, but the final extent of graphitization is the same.
  And ordinary cast iron which has a large primary expansion and a small total construction, i. e. has much dissolved gasses and difficult graphitization is generally difficult to nodulize.

View full abstract