Creep behavior of near-alpha titanium alloys.

Abstract

The effect of microstructure, silicon content and thermal exposure on the creep behavior of Ti-6Al-2Sn-4Zr-2Mo has been studied at 650$\\sp\\circ$C and 760$\\sp\\circ$C. At these high temperatures, $\\beta$-solution treated microstructures exhibited generally superior creep behavior compared to $\\alpha$/$\\beta$-solution treated microstructure. Moreover, a pronounced influence of cooling rate from the $\\beta$-solution treated temperature on primary and steady state creep has been observed. Analysis of the stress dependence for steady state creep showed that at 760$\\sp\\circ$C, the creep mechanism changed from diffusional creep at low stress levels to dislocation creep at high stress levels. A similar transition in creep mechanism was not found at 650$\\sp\\circ$C, where creep was controlled by dislocation motion at all stress levels. Similarly, the observed dislocation structure depended on the creep test regime. The dislocation structure developed during creep at 650$\\sp\\circ$C and 760$\\sp\\circ$C/high stress levels consisted of homogeneous arrangements of a dislocations and dislocations with a c component Burgers vector. However, at 760$\\sp\\circ$C and low stress levels, the dislocation structure was characterized by a much lower density of a dislocations and dislocations with a c component Burgers vector, and an increasing occurrence of heterogeneous dislocation arrangements. At 650$\\sp\\circ$C and 760$\\sp\\circ$C/high stress, increasing silicon content decreased the primary creep strain and the steady state creep rate. At 760$\\sp\\circ$C/low stress, where diffusional creep dominated, addition of silicon reduced the primary creep strain but did not affect the steady state rate. Microstructural changes resulting from thermal exposure have been analyzed after 100-600 hours at 650$\\sp\\circ$C and 760$\\sp\\circ$C, in terms of silicide precipitation and $\\beta$ phase decomposition. The steady state creep rate was not significantly affected by these microstructural changes. A phenomenological model was developed to rationalize the influence of microstructure on creep behavior, in terms of the $\\alpha$ plate width and the amount of retained $\\beta$ phase. The existence of an optimum cooling rate from the $\\beta$ phase field can be explained in terms of the trade-off between increasing retained $\\beta$ phase (detrimental to creep resistance) and decreasing $\\alpha$ plate width (beneficial to creep resistance) which occur as cooling rate increases.