Molds with exothermic self-hardening sand bound by sodium silicate have a tendency to cause the movement of mold wall and induce casting defects such as external and internal shrinkage.
  In order to eliminate these defects, it is very inportant to control the return sand and the quantity of sodium-silicate, and to adopt a method for treating the return sand.
  Pouring of the molten iron into a heavy section casting is liable to cause the movement of mold wall because of a large quantity of heat transfer and the exertion of pressure on the surface of the mold.
  In this report the following problems are investigated
  (1) The quantity of sodium-silicate
  (2) The quantity of residual sodium-silicate round the return sand grains.
  (3) The degree of sand grain bonding.
  (4) The characteristics of sand against temperature.
  (5) The degree of ramming of mold.
The results obtained are as follows.
  (1) Adding a large quantity of sodium-silicate and using the sand over the bench life tends to cause the movement of mold-wall.
  (2) Mold with sand of smaller thermal expansion tends to cause more casting defects.
  (3) Mold with larger amount of sand with residual sodium-silicate tends to cause more casting defects.
  (4) When using sand with good flow, the degree of ramming does not affect so much the formation of casting defects.

View full abstract

  The first part of this article describes the prevention of inverse chill of spheroidal graphite cast iron melted by a 2 t coreless type low frequency furnace lined with SiO₂ and treated by magnesium-rare earth mixtured spheroidal reagent (KC-reagent) in foundry works. The inverse chill often appeared in the case of the spheroidized treatment of molten irons which were not held at all or held above 70 minutes at 1,500°C, but it was eliminated by holding the moten iron about 13 to 60 minutes at 1,500°C. It was found that inverse chill was generally associated with low graphite nodule number and imperfect spheroidal graphite forms, and the lower the graphite number the stronger was the chilling tendency.
  In the second part of this article, the effect of holding molten irons for applicable frequency of lined material and change of Sc was investigated. The change of chemical compositions and the behavior of gas elements (nitrogen and oxygen) were as follows. In either case carbon slightly decreased, while silicon slightly increased, but manganese, phosphorous, sulpher, chromium and titanium did not change. However when the applicable frequency was great (70 times), titanium slightly increased. In spite of Sc value, the Ns (hydrochloric acid soluble nitrogen) content changed little, but when Sc was low, the Ns content was higher than the value when Sc was high. This phenomenon corresponded with the thermodynamic equilibrium value of Ns. In either case the N_I (hydrochloric acid insoluble nitrogen) content slightly decreased, but only when the applicable frequency was great (70 times), the N_I content slightly increased to be as much as titanium value. In the case of high Sc, oxygen was increased by holding ever 50 minutes. This increase of oxygen may have promoted the tendency of chill effect and undercooled graphite formation. In the case of low Sc, chilling depth changed little and undercooled graphite did not appear. Though oxygen content showed a tendency to increase, it was essentially little different with high Sc melts.
  In the last part of this article, the contents of gases in molten irons melted by cupola and low frequency furnace were compared. The Ns content was little different in both melting methods, but the N_I content in cupola was a little higher than for the low frequency furnace. On the other hand, oxygend ontent in the case of cupola was almost twice as much as in the case of the low frequency furnace, but after spheroidizing treatment, it decreased remarkably to the oxygen content of spheroidized low frequency furnace melts.

View full abstract

  This study was carried out for the purpose of industrialization of oxygen enrichment which was not formally possible in spite of many advantages, because of economic problems but which now can be realized with the development of the special gaseous oxygen production equipment that enables the production of oxygen at a low price.
  For commercialization, first the metalurgical effects had to be grasped, and then an economical method had to be concidered. The acid lined, cold blast and water cooled cupola used for this study has a melting rate of 6 t/hr.
  As a result of this study, the expected objectives were accomplished as follows :
  (1) It was possible to decrease the coke ratio from 15% to 13% by oxygen enrichment.
  (2) The more the oxygen content in the blast air increased, the more the melting rate rose.
  (3) Oxygen enrichment has superb recovery from unexpected fall of tapping temperature.
  (4) As for the chemical composition of the metal, it was acknowledged that the carbon content increased and the sulpher content decreased by oxygen enrichment.
  (5) Economy was obtained with the decrease of coke ratio from 15% to 13%.

View full abstract

  In order to obtain a more detailed knowledge of the role of silicon in promoting graphite formation, the distribution behavior of silicon during the solidification of iron-carbon-silicon alloy with or without graphite-spheroidizing treatment was investigated. Carbon and silicon contents of alloys were varied from about 0% to 4% and 0.2% to 3.5%, respectively. The cooling rate employed was about 2°C/sec in the molten state.
  The distribution of silicon was detected by means of a potentiostatic etching technique. The following became clear after this study :
  (1) The distribution behavior of silicon is largely influenced by the carbon content of alloy, irrespective of the shape of the graphite phase deposited, indicating that the solidification process of spheroidal graphite cast iron is not much different from that of flake graphite iron. The segregation intensity of silicon, however, seems to be stronger in spheroidal graphite iron than in flake graphite cast iron.
  (2) The normal segregation of silicon is observed around the primary austenite, but is inverted at eutectic. So the primary austenite can be distinguished from the eutectic one by silicon content. Most of the graphite are found in the region rich in silicon and it is thought that the build-up of silicon as well as carbon in front of the growing austenite dendrite plays an important role in the formation of graphite in hypo-eutectic cast iron.
  (3) Also in hyper-eutectic iron, dendritic austenites can be observed, but they are enriched with silicon in contrast to the primary austenite occurring in hypo-eutectic iron and can be regarded as divorced eutectic austenite.
  (4) Thus, it is thought that the solidification process of both flake and spheroidal graphite cast iron can be explained according to the iron-carbon-silicon phase diagram.

View full abstract

  With the increase of the production speed of castings, it becomes important to cool the foundry sand forcedly. In this paper we deal with the cooling effect when the naturally falling sand is cooled by counter air flow.
  First the following assumptions have been made in working out the theoretical equations for the estimation of the cooling effect. (1) Sand falls separately and at a uniform speed. (2) Each grain of sand is regarded as a sphere which has the same projection area as the sphere itself has. (3) The temperature distribution in the grain of sand may be neglected. (4) Sand consists of dry sand and uniformly wetted sand. (5) The heat of sand is taken away by evaporation of water and heat transfer.
  Especially when dry sand is cooled, the temperature of sand at the outlet of the cooling tower can be calculated from equation (1)
      [Written in non-displayable characters.]····(1)
      [Written in non-displayable characters.], [Written in non-displayable characters.]
where c : specific heat (Kcal/Kg°C), d_s : mean diameter of the sand (m), t₀ : passing time of the sand from inlet to outlet of the cooler (hr), G : weight flow rate (kg/hr), α : coefficient of hea ttransfer (Kcal/m²hr°C), γ : specific weight (Kg/m³), η : cooling effect
subscripts 1, 2 : refers to inlet and outlet of the cooler, a : refers to air, s : refers to sand.
  If equation (2) which was obtained by Yuge for a sphere is used for fhe coefficient of heat transfer α, the theoretical curve almost coincides with the experimental results.
      [Written in non-displayable characters.]····(2)
where λa : coefficient of heat conduction of air (Kcal/m•hr•°C)
      Re : Reynolds number for mean diameter of sand
As we can see from equation (1), cooling efficiency η becomes greater with the decrease of φ'. When φ' is greater than 1.0, the maximum value of η is 1/φ', so it is useless to make the length of cooler too long.
  When sand is wetted, it can be cooled in much shorter time than with dry sand. The cooling process has been solved numerically and analytically from heat transfer equation, heat and mass balance equations and evaporation equation in which the Lewis' law is used. The numerical results considerably coincide with experimental results.
  Especially on the following conditions, the temperature of the wetted sand at the outlet can be calculated from equation (3). (1) Water content of sand is greater than 0.7% at the outlet and sand is uniformly wetted. (2) The temperature of sand is between 80°C and 40°C. (3) The mean absolute humidity X_m of cooling air in the cooler is less than 2%.
      [Written in non-displayable characters.]····(3)
Where a=0. 000433 b=0.0598
      [Written in non-displayable characters.]
      L_s : heat of vaporization (Kcal/kg)
  In this case the cooling efficiency η becomes greater with the increase of the temperature T_s1 of sand at the inlet, and the value of φ'=C_sG_s/ (C_aG_a) does not affect η very much when the humidity of cooling air is lower than the saturated humidity. As the critical water content of silica sandis about 0.4, 0.7%, it is desirable for the water content at the outlet to be greater than 0.7%.
  It is practically difficult to separate sand into each grain and get it uniformly wet, therefore η decreases.

View full abstract