Papers by Author: Toshio Narita

Abstract: A Ni aluminide layer containing Zr or Hf was formed on a Ni specimen by the simultaneous electrodeposition of Al and Zr or Al and Hf using a molten-salt bath. When the simultaneous electrodeposition of Al and Zr was carried out using molten NaCl-KCl containing 3.5 mol%AlF₃ and 0.05 mol%ZrF₄, the electrodeposited layers were formed in the order of Ni₂Al₃, NiAl₃ and Al from the Ni substrate side. The ZrAl₃ particles were uniformly formed in the surface region of the NiAl₃ and Al layers. On the other hand, when the simultaneous electrodeposition of Al and Hf was carried out using molten NaCl-KCl containing 3.5 mol%AlF₃ and 0.05 mol%HfF₄, the electrodeposited layer consisted of Ni₂Al₃ as the inner layer and NiAl₃ as the outer layer were formed with HfAl₃ particles uniformly formed in the surface region of the NiAl₃ layer. For the sample treated with the simultaneous electrodeposition of Al and Zr, no significant change in the mass gain was observed during the cyclic-oxidation test at 1423 K, suggesting that the sample had a high cyclic-oxidation resistance. Similarly, the sample treated by the simultaneous electrodeposition of Al and Hf had a high cyclic-oxidation resistance. An adhesive scale, having localized inward penetrations consisting of Al₂O₃ containing ZrO₂ or HfO₂, was formed on the samples having the high cyclic-oxidation resistance.

Abstract: Cross-sectional structures of a Re-based diffusion barrier coating on Nb as ultra high temperature material were investigated in order to verify the crystalline structure and composition of the coated layer. Three types coating specimens were prepared by electroplating Re from an aqueous solution on an Nb substrate, followed by Cr-pack cementation in vacuum. The coating process produced three distinct layers; an outer Cr(Re) layer, an intermediate Re-Cr-Nb layer, and an inner Nb(Re) layer. A crystal structure of Cr(Re) and Nb(Re) layers possess similar single crystal bcc structure. The Re-Cr-Nb layer expected to act as a diffusion barrier between the substrate and the outer reservoir layer was comprised of cubic c phase and hexagonal Laves C14 phase. Moreover, several crystal defects such as dislocations and stacking faults as well as voids and cracks are observed in the coating specimen.

Abstract: High temperature cyclic oxidation behavior of γ'-base Ni-25Al-10Pt (in at.%) alloy was investigated at 1000°C in air with and without 30vol.%H₂O. The oxidation mass gain during the initial stage of oxidation was similar in both atmospheres, but the oxidation rate in air+H₂O was lower in the longer steady-state oxidation stage. Metastable Al₂O₃, which formed during the initial stage of oxidation, transformed completely after about 100hr of oxidation in dry air. The transformation to α-Al₂O₃ also occurred in air+H₂O, but complete transformation to α-Al₂O₃ was not observed during the oxidation time in the present study. θ-Al₂O₃ grains remained for longer on the α-Al₂O₃ layer in air+H₂O and became significantly coarser with oxidation time. The present results indicate that water vapor delays the metastable to α-Al₂O₃ phase transformation, and decreases the growth rate of α-Al₂O₃.

Abstract: The concept of the diffusion barrier coating system (DBC system) is summarized and the latest results are presented. The DBC system is comprised of alloy substrate/diffusion barrier/Al-reservoir/an external scale. Diffusion flux (JAl) of Al through the barrier layer will be given approximately by J_Al = D_Al x S_Al x (d C_Al/d x), where DAl and SAl are the diffusion coefficient and solubility limit of Al in the barrier layer, respectively as well as d CAl / d x is driving force given by the concentration difference across the barrier (d CAl) divided by the thickness of the barrier layer (d x). A slow diffusion flux can be obtained by using low values of DAl, SAl, or (d CAl /d x). Accordingly, a selection of a barrier layer with lower DAl and SAl is essential. A low driving force is also an important factor, and can be achieved by using lower C_Al with a constant barrier layer thickness dx. At higher temperatures, however, the barrier layer can react with the alloy substrate and Al-reservoir layer, resulting in gradual degradation of the barrier layer. This means that the thickness dx of the barrier layer tends to decrease and may finally disappear. With decreasing thickness of the diffusion barrier layer, the driving force (dCAl/dx) will increase, and the effectiveness of the barrier layer will be eliminated. Therefore, it is essential to maintain a constant thickness of the barrier layer for long exposure time. Several types of the DBC system are proposed, a single barrier layer and triple-layers with g + g’ and g’ inserted among these barrier layers.

Abstract: The oxidation behavior of Fe-20at.%Cr-10at.%Al alloys with a small amount of an additional element such as W, Cu, Mn, Nb, Mo, Re, Co or Ti was investigated at 900 °C for up to 625hr. The fourth element addition to the FeCrAl alloy could be classified into two groups; elements (Mn, Nb, Ti) that are contained in the Al2O3 scale, and elements (W, Mo, Re, Co) which are not present in the scale. In the latter case, the elements (W, Cu) caused scale spallation. The rumpling of alloys with Mn, Nb or Ti was smaller than that of the other alloys. The surface of the alloy with Ti was the smooth. Pt marker experiments suggested that the Al2O3 scale formed on the alloy with Ti grew by inward diffusion of O, whilst the Al2O3 scale formed on the FeCrAl alloy grew by both outward diffusion of Al and inward diffusion of O. This different growth behavior due to the elements incorporated in the Al2O3 scale could have an effect on the surface rumpling behavior.

Abstract: A duplex layer, outer Pt-modified γ’-Ni3Al + γ-Ni and inner multi-barrier σ- Re(Cr,Ni,W), coating system was formed on a Ni-based single crystal 4th generation superalloy. Oxidation behavior of the coated alloy was investigated under thermo-cycling conditions, and analyzed by EPMA and XRD. During cyclic oxidation 1hr at 1100°C and 20 min at room temperature, a slow growing α-Al2O3 formed for up to 400 cycles and its spallation was rare. The parabolic rate constant of mass change was 6.3x10-16 kg2m-4s-1. The Pt-modified γ’-Ni3Al + γ-Ni contained 19Al, 12Pt, 4Cr, and 3Co in at%, and their concentration profiles were almost flat across the outer layer. The multi-barrier, σ-Re(Cr,Ni,W) contained 40Re, 23Cr, 17Ni, 7Al, 4W, 3.5Mo, and 3Co in at%. Furthermore, the γ’-Ni3Al containing Pt was newly formed between the multibarrier and bulk alloy substrate. It was concluded that the σ-Re(Cr,Ni,W) is compatible with the Ptmodified γ’-Ni3Al in the multi-diffusion barrier coating on Ni-based single crystal, 4th generation superalloy at high temperatures.

Abstract: High temperature oxidation / creep deformation behavior of a diffusion barrier coated Hastelloy-X alloy, with large grain size ~500μm, was investigated at 970°C in air with external tensile stress of 22.5, 27.5, 32, and 40MPa. The diffusion barrier coating formed on Hastelloy-X consisted of a duplex structure with an inner diffusion barrier layer of Re-Cr-Ni alloy, and an outer oxidation resistant layer of β-NiAl. Un coated bare Hastelloy-X alloy with same grain size was also examined under the same conditions for comparison. The composition of the as-coated diffusion barrier coating was (15~21)Ni, (33~37)Cr, (30~33)Re, (11~15)Mo, and (9~14)Fe. This composition corresponds to σ-phase in the Ni-Cr-Re ternary system, which is known as a topologically close packed, TCP phase. The composition of this diffusion barrier layer did not change during the experiment. The oxide scales formed after creep testing on the coated and un-coated alloy surfaces were needle-like θ-Al2O3, and Cr2O3 with small amount of FeCr2O4, respectively. Grain boundary oxidation was also found in the subsurface region of the un-coated alloy. The Al2O3 scale exhibited severe spallation, and many cracks were formed perpendicular to the stress direction. However, no spallation or cracks were observed in the Cr2O3. The creep rupture times for the diffusion barrier coated alloy were about 1.5 times longer than those for bare alloy at all creep stress conditions. The fracture surface after rupture indicates that fracture occurred along alloy grain boundaries in both the coated and un-coated alloy substrate. Many cavities and cracks were observed within the diffusion barrier coated alloy substrate. These cavities and cracks tended to propagate from the substrate toward the diffusion barrier layer, and then stopped at the Re-Cr-Ni / β-NiAl interface. Cracks formed in the un-coated alloy initiated at the tip of grain boundary oxides, and propagated into alloy substrate. However no major cavities were observed inside the alloy substrate. The stress index, n, for both specimens was about 6, and this indicates that the deformation mechanism of both samples was dislocation creep. These results suggest that the Re-Cr-Ni diffusion barrier layer acts as a barrier against the movement of dislocations at the interface with the alloy surface.

Abstract: Single crystal superalloy TMS82+ and model alloys of Ni-12Al and Ni-6Cr-8Co-12Al were electroplated with 5-6μm of platinum and heat treated at 1000°C for up to 100hrs. In the model alloys the platinum concentration gradient in the interdiffusion region caused uphill diffusion of aluminium. The subsequent surface enrichment led to formation of aluminium-rich γ′ phase. In TMS82+ superalloy inward diffusion of platinum resulted in loss of the γ/γ′ microstructure and formation of new γ and γ′ grains. The initial dissolution of the γ′ cubes was due to the outward diffusion of aluminium. Again, the presence of a platinum gradient in the interdiffusion region resulted in uphill diffusion of aluminium and a net increase in aluminium content at the alloy surface, leading to a near-surface single-phase γ′ layer being formed, however topologically close-packed (TCP) phase formation was not observed.

Abstract: The effects of coatings on the creep and oxidation behavior of Ti-50Al alloy were investigated at 1173K in air at a constant loading of 30MPa. The coating was formed by a two-step Cr/Al diffusion treatment and consisted of an outermost TiAl2 layer, an outer Al-rich γ layer, an intermediate γ, Laves and β mixture layer, and a Cr diffusion zone. Creep tests were also carried out with sole Cr or Al coated TiAl and also of uncoated TiAl. The oxide scales formed on the uncoated TiAl and the sole Cr coated specimens were a mixture of TiO2 and Al2O3, which displayed several exfoliations. Both the two-step Cr/Al coated TiAl and the sole Al coated specimens formed a protective Al2O3 layer and little oxide exfoliation was observed here. Significant cracks were observed in the sole Al coated TiAl, while no cracks were observed in the sole Cr coated TiAl; the two-step Cr/Al coated TiAl showed a number of cracks in the coatings. Low creep rates in the two-step Cr/Al coated TiAl could be due to the Laves phase with a hexagonal C14 structure in the intermediate, γ, β and Laves phase mixture, and the high creep rates of the sole Cr coated TiAl may originate in the major β phase component with a B2 structure in the γ, β, and Laves phase mixture.

Abstract: The effects of the third element on the high temperature oxidation of γ'-Ni3Al with 5at%X (X=Ti, Ta, Nb, Cu, Co and Fe) alloys were investigated at 1173K in air, and oxidation behavior could be classified into three groups. The first group, comprised of alloys with Cu and Co, showed good oxidation performance with Al2O3 formation. A second group contains Ti, Ta, and Nb as alloying elements, and showed poor oxidation performance. With Fe or Mn addition the alloy oxidation performance was intermediate between the first and second group. The effects of these elements are discussed associate with partitioning factors for each element in the γ'-phase.