An investigation of the thermoviscoplastic behavior of a metal matrix composite at elevated temperatures

This research investigates the response of a fiberless 13 layer hot isostatically pressed Ti-15-3 laminate to creep, constant strain rate, and cyclic constant strain rate loading at temperatures ranging from 482C to 649C. Creep stresses from 48 to 260 MPa and strain rates of .0001 to .01 m/m/sec were used. Material parameters for three unified constitutive models (Bodner-Partom, Miller, and Walker models) were determined for Ti-15-3 from the experimental data. Each of the three models was subsequently incorporated into a rule of mixtures and evaluated for accuracy and ease of use in predicting the thermoviscoplastic response of unidirectional metal matrix composite laminates (both 0 and 90). The laminates were comprised of a Ti-15-3 matrix with 29 volume percent SCS6 fibers. The predicted values were compared to experimentally determined creep and constant strain rate data. It was found that all three models predicted the viscoplastic response of the 0 specimens reasonably well, but seriously underestimated the viscoplastic response of the 90 specimens. It is believed that this discrepancy is due to compliant and/or weak fiber-matrix interphase. In general, it was found that of the three models studied, the Bodner-Partom model was easiest to implement, primarily because this model does not require the use of cyclic constant strain rate tests to determine the material parameters involved. However, the version of the Bodner-Partom model used in this study does not include back stress as an internal state variable, and hence may not be suitable for use with materials which exhibit a pronounced Baushinger effect. The back stress is accounted for in both the Walker and Miller models; determination of the material parameters associated with the Walker model was somewhat easier than in the Miller model.