Masanori Kajihara

Cellular precipitation involving both substitutional and interstitial solutes: cellular precipitation of Cr2N in CrNi austenitic steels

This paper describes a part of our series of studies on the cellular precipitation involving the diffusion of both substitutional and interstitial solutes. Following a brief review of a few previous studies on the cellular precipitation of alloy carbides and nitrides in steels and superalloys, the reason is given why the cellular precipitation of Cr2N in CrNi austenitic stainless steels with high nitrogen concentrations is best suited for the study on this type of cellular precipitation. The microstructural features of the present cellular precipitation of Cr2N, which have largely been ignored in the previous studies, are illustrated. Experimentally determined growth features are summarized in well-controlled specimens with four concentration levels of nitrogen. Several growth features were measured on the same specimens. They are the volume fraction of the cellular precipitate, the free surface area of the moving cell boundary, the average migration rate of the moving cell boundary, the interlamellar spacings and the lattice parameter of both the untransformed austenite matrix and the austenite within the cell. The concentration profiles of metallic elements as well as of nitrogen were also measured. These growth features are compared with those established in the “typical” cellular precipitation in binary substitutional systems, such as PbTi, FeZn and AlMg alloys. A qualitative explanation is offered for the observed non-steady state growth features of the present cellular precipitation. They are the transfer of the faster-diffusing element, nitrogen, from the untransformed matrix to the cell via long-range volume diffusion, the deceleration of the migration rate of the moving cell boundary with reaction time and the stoppage of the migration of the cell boundary at a stage when nitrogen supersaturation still remained in the untransformed matrix. The importance of the volume diffusion of the slower-diffusing element, chromium, ahead of the moving cell boundary is emphasized.

Kinetic Features of Reactive Diffusion in Binary Systems

The temperature dependence of the kinetics of the reactive diffusion was numerically analyzed for a hypothetical binary system composed of one compound phase (β) and two primary solid solution phases (α and γ). The growth rate of the β phase during the reactive diffusion between α and γ phases in a semi-infinite diffusion couple was expressed as a function of the interdiffusion coefficient Dθ and the solubility range of the θ phase ( θ = α, β, γ). For the reactive diffusion controlled by volume diffusion, the thickness l of the β phase is described as a function of the annealing time t by the parabolic relationship l2 = Kt. The equations K = K0 exp(−QK/RT) and Dθ = D0 θexp(−Qθ/RT) were adopted to express K and Dθ as functions of temperature T, respectively. The relationship between the temperature dependence of K and that of Dθ was evaluated according to the following assumptions: the molar volume, the solubility range and the value of D0 θ are constant and equivalent for all the phases. When Qα or Qγ is smaller than Qβ, QK is greater than Qβ. On the other hand, QK is close to Qβ, if both Qα and Qγ are greater than Qβ. In such a case, the temperature dependence of the kinetics represents that of interdiffusion in the growing compound.

Kinetic features of diffusion induced recrystallization in the Cu(Ni) system at 873 K

Abstract The kinetics of diffusion induced recrystallization (DIR) in the Cu(Ni) system was experimentally studied at 873 K. Ni/(CuNi)/Ni diffusion couples were prepared by a diffusion bonding technique using a pure Ni specimen and binary CuNi alloys with mol fractions of Ni of y0=0, 0.114 and 0.231, and then annealed at 873 K for various times between 1 and 72 h (3.6×103 and 2.59×105 s). After annealing, a fine grain region (DIR region) alloyed with Ni was observed to form at the interface owing to DIR and to grow mainly towards the Cu phase. The thickness l of the DIR region monotonically increases with increasing annealing time t according to the relationship l=k(t/t0)n, where t0 is unit time, 1 s. The exponent n takes a constant value of 0.45 independent of the composition y0. On the other hand, as the composition y0 increases from 0 to 0.231, the proportionality coefficient k decreases from 5.7×10−8 to 4.5×10−8 m. The experimental results were theoretically analyzed using a kinetic model proposed by the present authors. The growth behavior of the DIR region was explained well by the analysis. This means that the chemical driving force due to the compositional difference between the Cu phase and the DIR region is the most important driving force for DIR in the Cu(Ni) system under the present experimental conditions.

Observations on diffusion-induced recrystallization in binary Ni/Cu diffusion couples annealed at an intermediate temperature

Abstract Diffusion-induced recrystallization (DIR) was experimentally studied using diffusion couples consisting of a pure Cu single crystal and a binary Ni–Cu polycrystalline alloy. The (Ni–Cu)/Cu diffusion couples were annealed at 873 K for various times between 1 and 72 h (3.6×10 3 and 2.59×10 5 s). During annealing, a fine grain region (DIR region) was observed to form at an interface of each diffusion couple due to DIR and to grow mainly towards the Cu phase. The thickness l of the DIR region increases with increasing annealing time t according to the relationship l = k l ( t/t 0 ) n , where t 0 is unit time, 1 s. The proportionality coefficient and the exponent take constant values of k =2.18×10 −8 m and n =0.55, respectively, independent of the composition of the Ni phase within experimental uncertainty. The compositional differences across the moving boundaries of the DIR region on both the Cu phase and Ni phase sides decrease in inverse proportion to the almost fourth root of the annealing time. At each annealing time, the compositional difference Δ x c on the Ni phase side monotonically decreases with increasing Cu concentration x 0 in the Ni phase, whereas there is no clear relationship between the compositional difference Δ x i on the Cu phase side and the Cu concentration x 0 .

Formation and stability of a nitride with the structure of beta manganese in Ni-Cr-N ternary system

A nitride with the metal-atom arrangement of β manganese, which is designated as π phase, was confirmed to exist as an equilibrium phase in the ternary Cr-Ni-N system. A Cr-40 mass pct Ni binary alloy was nitrided under a pure nitrogen atmosphere at 1523 K to prepare a Cr-40Ni-5N alloy consisting of a nickel-rich face-centered cubic (fcc) γ phase and a dichromium nitride, Cr2N, designated as ε phase. Upon aging the alloy at 1273 K, the π phase was formed through a peritectoid reaction between the γ and ε phases. A volume fraction of the π phase reached 95 pct after 3.6 × 105 s, and no more change of the volume fraction was observed, even after 3.6 × 107 s aging. The chemical composition of the π phase was determined to be Cr13Ni7N4, whose lattice parameter wasa = 0.6323 nm. The π phase became unstable above 1473 K, which explains a previous unsuccessful attempt to produce the Cr-Ni-N ternary π phase by the replacement of molybdenum in Mo12Ni8N4 by chromium at 1473 K.

Microstructure formed by eutectic reaction in a binary Cu–12.3Zr alloy

Abstract A Cu–Zr alloy with a Zr concentration of 12.3 mass% was used to examine experimentally the morphology and crystallography of the microstructure formed by the eutectic reaction for the Cu phase and the most Cu-rich intermetallic compound in the binary Cu–Zr system. A typical lamellar microstructure with interlamellar spacings around 0.4 μm was developed during solidification after induction melting. According to the observations by transmission electron microscopy, the lamellar microstructure consists of the Cu and Cu9Zr2 phases. Here, the space group and Bravais lattice of the Cu9Zr2 phase are considered as F 4 3m and face-centered cubic, respectively. In the lamellar microstructure, the parallel orientation relationship with [100]Cu//[100]Cu9Zr2 and [010]Cu//[010]Cu9Zr2 exists between the Cu and Cu9Zr2 phases, and the Cu/Cu9Zr2 interface is almost parallel to the {100} planes of these phases. During annealing up to 60 h after solidification, the lamellar microstructure changes into a granular microstructure at 1123 and 1223 K due to discontinuous coarsening and spheroidization, whereas it is rather stable at 1023 K. For the specimen annealed at 1223 K for 60 h, the concentration of Zr in the Cu9Zr2 phase was determined to be 24.3±0.2 mass% by electron probe microanalysis. This value is close to the Zr concentration of 24.18 mass% for stoichiometric Cu9Zr2.

Boundary energies of Σ11 [110] asymmetric tilt boundaries in Cu determined from the shape of boundary silica particles

Abstract The boundary energies of Σ11 [110] asymmetric tilt boundaries with a misorientation angle of θ=50.5° in Cu have been experimentally determined by a silica particle observation method for various inclination angles between φ=0 and 90°. This method has no drawbacks that are inevitable in boundary grooving and tri-junction methods. In the boundary energy vs inclination angle diagram, three deep cusps exist at φ=10, 64.8 and 90° whereas a shallow cusp appears at φ=35.3°. Such inclination angle dependence of the boundary energy has been utilized to analyze the faceting of the boundary. The observations on the faceting reported for Σ11 [110] tilt boundaries in face-centered cubic (f.c.c.) metals have been quantitatively accounted for by the present analysis.

Activity of carbon in nickel-rich Ni-Mo and Ni-W alloys

Effects of molybdenum and tungsten on carbon activity in nickel have been experimentally determined at 1000, 1100, and 1200 °C. Seventeen nickel-molybdenum and thirteen nickel-tungsten binary alloys were carburized in a flow of purified methane and hydrogen mixed gas. A sealed capsule technique was also employed for carburization of a few series of nickel-molybdenum alloys. The carbon concentration was determined either by hot extraction techniques (LECO and Coulomatic) or by weight gains of these specimens. The carbon concentration at a constant carbon activity decreases with increasing either molybdenum or tungsten concentration in nickel. The effect of tungsten on the carbon solubility in nickel is slightly larger than that of molybdenum. The experimental data were analyzed in terms of the regular solution model with two sublattices due to Hillert and Staffansson. Temperature dependence of the interaction coefficients between carbon and molybdenum or tungsten was expressed as DGMo/RT = −4.45 + 11650/T andDGW/R = 1.21 + 9010/.

Optimizing POF/PCF based optical switch for indoor LAN

For indoor local area network (LAN) the Polymer optical fiber (POF) is mostly appropriate, because of its large core diameter and flexible material. A 1×2 optical switch for indoor LAN using POF and a shape memory alloy (SMA) coil actuator with magnetic latches was successfully fabricated and tested. To achieve switching by the movement of a POF, large displacement is necessary because the core diameter is large (e.g., 0.486mm). A SMA coil actuator is used for large displacement and a magnetic latching system is used for fixing the position of the shifted POF. The insertion loss is 0.40 to 0.50dB and crosstalk is more than 50dB without index-matching oil. Switching speed is less than 1s at a driving current of 80mA. A cycling test was performed 1.4 million times. Polymer clad fiber optical (PCF) switch also fabricated and tasted.

Growth rate of IMC in the binary sytems of Co/Sn and Cu/Sn

In this study we discuss superiority of Cobalt for using in 3D interconnection as alternative metal to Cu. Specimens composed of pure Sn and Co or Cu is aged under same aging condition varied time and temperature below melting point of Sn. Thickness of IMC (intermetallic compound) formed at the interface is then calculated for extraction of growth rate and kinetics like activation energy and power factor. Compared with each factors extracted from calculation, IMC formation and growth of Co/Sn has stronger time and temperature dependence than IMC of Cu/Sn. Furthermore, no voids was observed at any interface in Co/Sn bonding.