Hydrogen-induced internal-stress plasticity in titanium

Abstract

Internal-stress plasticity is a deformation mechanism resulting from the biasing of internal mismatch stresses by an external stress.[1] These internal stresses can be produced by a temperature change in materials with grains (or phases) exhibiting different thermal expansion coefficient[2] or different density during a phase transformation[3]; alternate sources of internal stresses include compressibility mismatch during pressure changes[4] and swelling mismatch during irradiation.[5] The most common type of mismatch generation is thermal, as it easily lends itself to cycling during which tensile strains well in excess of 100 pct can be accumulated; this phenomenon is known as internal-stress superplasticity.[1] The purpose of the present article is to report, for the first Fig. 1—Titanium-hydrogen phase diagram[14] showing chemical cycles time, internal-stress plasticity where internal stresses are used in the present study of chemically induced internal-stress plasticity induced by a reversible change of chemical composition (at at 805 or 860 8C ( pH2 5 0 } 4.1 kPa) as well as typical thermal cycles constant temperature and pressure) rather than a change used in the literature for thermally induced internal-stress plasticity in of temperature or pressure (at constant composition), as hydrogen-free Ti (T 5 860 } 900 8C). illustrated in Figure 1. We use the Ti-H system to demonstrate this effect, because hydrogen diffuses rapidly in and out of titanium where it produces mismatch strains both directly (by lattice swelling in the presence of concentration gradients) and indirectly (by inducing the hcp (a-Ti) to bcc (b-Ti) transformation), as demonstrated in a recent study of hydrogen-induced ratchetting.[6] We note that this chemically induced internal-stress plasticity is fundamentally different from thermochemical processing of titanium[7] where hydrogen alloying is used to produce a weaker, more easily workable b-Ti phase, and where internal stresses are unimportant. Titanium plates (99.7 pct pure, from Alpha Aesar, MA) were machined into dog-bone specimens with a gage 28 to 52 mm in length, 5.4 to 6.0 mm in width, and 2 mm in thickness. Specimens were tested in a low-stress creep apparatus described earlier,[8] modified to allow for a gas atmosphere flowing at a rate of 1.22 L/min, which could be changed from Ar (99.999 pct pure) to an Ar/4 vol pct H2 mixture. This 4.1 kPa hydrogen partial pressure corresponds to an equilibrium hydrogen concentration in titanium well within the b-Ti field (Figure 1). The specimen strain was continuously measured at the cold end of the load train and the temperature was maintained at a constant value of 805 8C 6 2 8C or 860 8C 6 2 8C by a thermocouple contacting the gage section or the pull head. Figure 2 shows the creep rate as a function of the applied Fig. 2—Isothermal creep rate at 805 8C and 860 8C for hydrogen-free astress for hydrogen-free a-Ti and for fully hydrogenated bTi and fully hydrogenated b-Ti. Ti at temperatures of 805 8C and 860 8C. The stress exponents