The activation volume V^∗ and the dislocation velocity-stress exponent m^∗ were obtained by stress relaxation tests at low temperatures in sintered pure molybdenum. The effect of grain size (30–1500 μm) on these parameters was investigated over the temperature range from 300 to 500 K. The values of V^∗ and m^∗ remained constant independent of grain size at temperatures below 400 K within the experimental error. It is concluded, therefore, that the effective stress σ^∗(=2k_BTm^∗⁄V^∗) calculated using the parameters does not vary with grain size.
Subsequently, the grain size dependence of the yield and flow stresses was discussed in relation to the internal stress σ_i and well explained by the change in σ_i due to the grain size at low plastic strains. Namely, the Hall-Petch relationship which holds up to about 1% plastic strain is expressed by the Ashby type equation, σ_i∝ε^1⁄2d^−1⁄2, which does not hold above that strain.

View full abstract

The structural changes and surface morphology concerning phase transformations in indium-rich solid solutions, such as In–Tl, In–Cd, In–Pb, In–Sn and In–Hg, have been studied by means of crystallographic and metallographic methods. Four types of phase transformation are observed in indium-rich solid solutions as follows: (1) fcc\ ightleftarrowsfct (c⁄a>1), (2) fcc\ ightleftarrowsfct (c⁄a<1), (3) fct (c⁄a<1)\ ightleftarrowsfco\ ightleftarrowsfct (c⁄a>1) and (4) fct (c⁄a<1)\ ightleftarrowsfct (c⁄a>1). All the transformations, except the fco\ ightleftarrowsfct (c⁄a>1) one, are of martensitic type, and the alloy surface of low-temperature phase shows a surface relief associated with the phase transformation, which is ascribed to a banded structure due to {110} transformation twinning. Under an assumption that these transformations take place in the form of a kind of the Bain lattice deformation, the lattice correspondence and the orientation relation between high- and low-temperature phases are well explained with respect to each phase transformation. The crystallographic orientation of the twin plane is explained on the basis of the Sapriel theory available for ferroelastic substances.

View full abstract

The Ni–20Cr alloy was oxidized at 1373 K at various oxygen partial pressures, ranging from 10⁻⁷ to 10⁴ Pa, which were controlled by Ar–O₂ mixtures. A large difference was observed between the oxidation behavior in a high P_O2 (≥10² Pa) range and that in a low P_O2 (≤1 Pa) range. Detailed experiments were carried out in P_O2 of 10⁴ and 1 Pa, the former being typical in high P_O2 and the latter in low P_O2. The mass gain-time curves showed that the growth rate of scale in the low P_O2 lie between the parabolic and linear law, while that in the high P_O2 obeys the parabolic law, which indicates that the scale formed in the low P_O2 is less protective than that in the high P_O2. It was found, on the other hand, that the oxide scales formed in the low P_O2 were more adherent to the alloy than those formed in the high P_O2. In the high P_O2 a large amount of spalling was observed on cooling even in short time oxidation. The oxide scales formed in the low P_O2 were uniform in thickness and rather porous, while those formed in the high P_O2 were dense and non-uniform in thickness. In the low P_O2 a number of small voids were found at the oxide-alloy interface. In the high P_O2, however, large voids were found; this suggests the occurrence of plastic deformation of the oxide and the alloy by the stress induced in the growing scale. The above difference in oxidation behavior can be ascribed to the difference between growth mechanisms of the oxide scale in the high and low P_O2.

View full abstract

Distribution behaviour of minor elements has important implications for copper making processes. The arsenic distribution among the three phases of gas, slag and matte was evaluated thermodynamically and is described in this article.
The arsenic distribution is basically dependent on the process factors such as the arsenic content of charge, the amount of waste gas, the degree of vapour saturation with arsenic or temperature. The thermodynamic quantities such as the activity coefficient of arsenic in matte, the activities of iron and copper, SO₂ or O₂ potentials and the ratio of arsenic distribution between slag and matte phases L_As^s/m are decisive factors for the arsenic distribution, and the behaviours in various smelters could be explained systematically and reasonably by considering changes in these quantities during smelting.
In oxidizing smelting at 10.1 or 101.1 kPa (0.1 or 1 atm) SO₂, the proportion of arsenic in gas phase is predominant in both the smelting and the converting stages, the degree of vapour saturation with arsenic seriously affecting the distribution behaviour. The arsenic content of final matte does not change too much even if the content of charge or of starting matte increases substantially. Furthermore, the proportion of arsenic in matte does not decrease appreciably even if L_As^s/m gets extremely large.
In reducing smelting at 10⁻¹¹ of p_O2, the proportions in gas and slag phases decrease remarkably, owing to a small activity coefficient of arsenic in matte and also to extremely small L_As^s/m.

View full abstract

The reaction efficiency of hot metal desulfurization by lime powder injection with nitrogen gas is improved by prior addition of aluminum. That is, the rate of desulfurization is higher in a Fe–C_sat.–Si–Al–S melt than in a Fe–C_sat.–Si–S melt. The reaction layers of lime lumps immersed in the melts were examined by electron probe microanalysis and X-ray diffraction. The investigation revealed the mechanism by which the addition of aluminum improves the reaction efficiency of the subsequent lime injection.
In the Fe–C_sat.–Si–S melt, solid 2CaO·SiO₂ and 3CaO·SiO₂ layers are formed on the lime surface, and thereby retards the transfer of sulfur to its internal part. On the other hand, a fusible CaO–Al₂O₃–FeO layer is formed on the lime surface in the Fe–C_sat.–Si–Al–S melt, promoting the inward transfer of sulfur. Furthermore, the layer dissolves as much sulfur as up to approximately 50%.
The aluminum concentration needed to obtain this improvement of desulfurization in a 250 ton torpedo ladle was the aluminum content of (0.005+ΔAl)%, where ΔAl is the aluminum lost during the desulfurization process.

View full abstract

The controlling step of dissolution of ferrous alloys into molten aluminium was studied. Commercially pure iron, Fe–Si, Fe–Ni, Fe–Cu, Fe–Mn and Fe–C alloys were dipped into molten aluminium (99.8%Al) at 973, 1023 and 1073 K for various times. The rate of dissolution of each alloy into molten aluminium was examined theoretically from the viewpoint of natural convection mass transfer. For the dissolution of the Fe–Cr, Fe–Cu and Fe–Ni alloys at 973 K, the experimental value of the mass transfer coefficient k_ob is nearly equal to the theoretical value k_c. Therefore, the dissolution of these alloys is controlled by the diffusion of Fe in molten aluminium. At 1023 and 1073 K, k_ob decreases with increasing dipping time. This is due to the increase of the resistance to the chemical reaction or mass transfer in the compoud layer. For the Fe–Si alloy, k_ob is a little larger than k_c. The resistance to chemical reaction or mass transfer in the compoud layer is negligible. For the commercially pure iron, k_ob is smaller than k_c by about 30–50%. The resistance to the chemical reaction or mass transfer in the compound layer is not negligible. For the Fe–Mn alloy, k_ob is smaller than k_c by about 40–70%. The resistance to the chemical reaction or mass transfer in the alloy layer is not negligible. But, further study may be necessary for this alloy. For the Fe–C alloy, k_ob is much smaller than k_c. The dissolution of this alloy is controlled by the chemical reaction or mass transfer in the compound layer.

View full abstract