Temperature dependence of dynamic Young's modulus and mechanical damping in SCS-6/Beta-21S titanium metal matrix composites

Abstract

Titanium-metal matrix composites (Ti-MMCs) are useful in applications that require high temperature performance, high specific strength, specific stiffness, creep resistance, and fracture strength. They are particularly useful in applications such as aerospace and engine components when conventional metals and ceramics are inadequate. The SCS-6/Beta-21S Ti-MMC gains its strength and stiffness from a network of silicon carbide fibers embedded in a titanium alloy matrix. This paper discusses an investigation into the temperature dependence of dynamic Young’s modulus and mechanical damping. The Piezoelectric Ultrasonic Composite Oscillator Technique (PUCOT) was used to determine the modulus and damping. The experimental data, together with findings from metallography, were applied to the Granato-Lucke theory of mechanical damping due to dislocations to give estimates of the dislocation density. The material was fabricated by hot-pressing alternating layers of matrix foils (Ti-15Mo-2.7Nb-3Al-0.2Si0.15O2) and continuous SiC (Textron SCS-6) fibers in a quasi-isotropic [0/±45/90]s orientation. The volume percent of the fibers was approximately 35% [1]. Strips 3 mm wide and approximately 30 mm in length were cut from the as-received sections with a low speed diamond saw. The PUCOT is a popular method of measuring the dynamic Young’s modulus, mechanical damping, and strain amplitude. Two α-quartz crystals (a drive crystal and a gauge crystal), a fused quartz spacer rod, and the specimen were glued together and excited by alternating voltages so that they experience resonant longitudinal stress waves. The gauge voltage is proportional to the strain amplitude. Values for the drive and gauge voltages, the resonant period of the system, and the density of the specimen were required for the PUCOT analysis, which returns the values of Young’s modulus, mechanical damping, and strain amplitude. The density was determined using Archimedes’ method. More information on the PUCOT theory is available elsewhere [2, 3]. For PUCOT tests at elevated temperatures, the specimen was glued onto the end of a fused quartz rod and inserted into a furnace. The spacer rod allowed the piezoelectric crystals to remain near room temperature while the specimen was brought to an elevated temperature within the furnace. A different length of rod was required for each test temperature to transmit longitudinal vibrations from the transducers to the specimen at the resonant frequency of the system. The damping that occurred within the rod is accounted for in the PUCOT analysis. PUCOT tests were conducted at test temperatures of 220, 350, 530, 680 and 850◦C. At each test temperature, the strain amplitude of the specimen was varied by modifying the gauge voltage, and hence the magnitude of the longitudinal vibrations, while holding the temperature constant. The strain amplitude was varied between a background level of approximately 3× 10−7 and a peak value of approximately 3× 10−4. Mechanical damping was measured at each strain amplitude. Micrographs were taken of two samples from the same specimen: a room temperature sample that did not experience elevated temperatures, and a sample that was held near 500◦C for 0.5 h, which was the duration of a typical PUCOT test. The samples were prepared by cutting the composite specimen with a diamond blade in a direction perpendicular to most of the fibers. The metallography facilitated investigation into the occurrence of oxidation and estimation of network length for application of the Granato-Lucke theory of dislocation damping. Table I summarizes the results of measurements of Young’s modulus, damping and density at room temperature for the quasi-isotropic [0/±45/90]s orientation. Figs 1 through 6 show micrographs of the etched specimens. The cross-sectional shapes of the fibers in Figs 1 and 2 reflect the fiber orientations. Two adjacent layers of fibers angled at 90◦ to the cut appear as perfect circles on the outside edge, followed by two layers at 45◦ that appear as ovals, and two layers angled at 0◦ that appear longitudinally in the center of the specimen. The micrographs in Figs 3 and 4 reveal the layered microstructure of the SiC fibers: a carbon core, a pyrolytic coating, the SiC region, a carbon-rich coating, and the reaction zone between the fiber and matrix [4]. Metallographic results showed little differences between the heated and unheated specimens. The