This study presents a complete set of effective elastic-stiffness coefficients of a Ni 80 P 20 amorphous-alloy thin film deposited on an aluminum-alloy substrate by electroless plating. The film thickness was 12 m. The electromagnetic-acoustic-resonance method detected resonance frequencies of the triple-layered specimens ͑film/substrate/film͒, which enabled us to determine all five independent elastic-stiffness coefficients of the film using known substrate elastic properties. The resulting coefficients were those of a transverse isotropic material. There was strong anisotropy between the in-plane and normal directions; the in-plane Young's modulus is larger than the normal Young's modulus by 34%, for example. The anisotropic coefficients can be interpreted by considering a micromechanics model for local incomplete cohesion ͑thin ellipsoidal voids͒ aligned parallel to the film surface.

No abstract

IN1RODUCTIONWe studied electromagnetic acoustic resonance at elevated temperatures up to 1000 K. Using Lorentz-force transduction, we measured the temperature dependence of resonant frequencies and attenuation coefficients in SiC-fiber-reinforced Ti-alloy composites. Evolution of the resonance frequencies is consistent with the changes of elastic constants of the matrix, and attenuation evolution is explained by dislocation damping in the matrix.Ultrasonic measurement provides a powerful tool for studying the temperature dependence of elastic constants and damping of solid materials developed for elevated temperatures. For the elastic constants, resonant ultrasonic spectroscopy (RUS) [1-4] provides a good measurement method. The usual RUS configuration uses a rectangular-parallelepiped specimen, whose sides are typically a few millimeters, and two sandwiching transducers that contact the specimen comers. One of them transmits a cw oscillation and the other detects ultrasonic oscillation to acquire a series of resonance frequencies. Inverse calculation [l,2] makes it possible to determine all independent elastic constants from the resonance spectrum. When it is used at elevated temperatures, one needs to insert waveguides between the specimen and transducers because of the low Curie point of conventional transducers, around 520 K for PZT. Using such devices, it is possible to determine the resonance frequencies and then the elastic constants. However, it is difficult to isolate the specimen's ultrasonic-energy loss, that is attenuation, because ultrasonic waves propagate not only in the specimen but also in the waveguides and transducers. Kuokkala and Schwarz [5] used magnetostriction for noncontact ultrasonic transduction and measured internal friction up to 520 K. The temperature limit depends on the Curie point of magnetostriction, typically below 1000 K. This technique requires coating a magnetostrictive material onto specimen surfaces, which requires further interpretation for a double-layered

Using electromagnetic acoustic resonance (EMAR), we measured the complete set of elastic constants and internal friction of a silicon-carbide-fiber-reinforced Ti-6Al-4V-matrix cross-ply composite between room temperature and 1000 K. This material exhibits tetragonal symmetry and possesses six independent C ij . By changing the magnetic-field direction, we selected two vibration groups among the eight vibration groups of a rectangular parallelepiped. The elastic constants showed anomalous temperature dependence around 650 K and internal friction Q Ϫ1 increased rapidly above that temperature; neither the C ij nor Q Ϫ1 returned to the before-heating values. We attributed this to disbonding between the matrix alloy and SiC, and between SiC and the carbon core within the fiber caused by oxidation of carbon, which is supported by the microstructure observation. We present the temperature derivatives of the composite C ij and estimate those of the fiber C ij by comparing the matrix-C ij derivatives.

No abstract