Effects of microstructure on the short fatigue crack initiation and propagation characteristics of biomedical α/β titanium alloys

This article presents the results of a study of the effects of microstructure on the fatigue strength and the short fatigue crack initiation and propagation characteristics of a biomedical α/β titanium alloy, Ti-6Al-7Nb. The results are compared to those obtained from a Ti-6Al-4V extra-low interstitial (ELI) alloy. Fatigue crack initiation occurs mainly at primary α grain boundaries in an equiaxed α structure, whereas, in a Widmanst?tten α structure, initiation occurs within the α colonies and prior β grains, where α plates are inclined at around 45 deg to the stress-axis direction. In an equiaxed α structure, the short fatigue crack initiation and propagation life, where the length of the crack (a) is in a microstructurally short fatigue-crack regime (2a < 50 μm), occupies around 50 pct of the total fatigue life. On the other hand, the fatigue crack in a Widmanst?tten α structure initiates at very early stages of fatigue, and, therefore, the fatigue crack-initiation life occupies a few percentages of the total fatigue life in an α structure. Then, the short fatigue crack propagates rapidly and is arrested at the grain boundaries of α colonies or prior β grains for a relatively long period, until the short crack passes through the boundaries to specimen failure. Therefore, the short fatigue crack-arrest life occupies more than 90 pct of the total fatigue life in a Widmanst?tten α structure. These trends are similar between the Ti-6Al-7Nb and Ti-6Al-4V ELI alloys and biomedical α/β titanium alloys. The total fatigue life for the Ti-6Al-7Nb alloy with an equiaxed α structure is changed by the volume fraction of primary α phase and the cooling rate after solution treatment. By increasing the volume fraction of the primary α phase from 0 to 70 pct, the fatigue limit of the Ti-6Al-7Nb alloy is raised. Changing the cooling rate after solution treatment by switching from air cooling to water quenching improves the fatigue limit of the Ti-6Al-7Nb alloy significantly.     

Improvement in mechanical properties of dental cast Ti-6Al-7Nb by thermochemical processing

Hydrogenation and dehydrogenation, that is, thermochemical processing (THP) and its variation with a post-heat treatment (THPH), are investigated in order to improve the balance of strength, elongation, and fatigue strength of cast Ti-6Al-7Nb and Ti-6Al-4V for dental applications. Microstructures of both cast alloys change from coarse Widmanst?tten α structure to super fine α structure with an average diameter of 3 μm by conducting THP or THPH. Tensile strength and fatigue limit of cast Ti-6Al-7Nb and Ti-6Al-4V increase by around 10 and 40 pct, respectively, as compared with those of both as-cast alloys. The balance of strength and ductility of cast Ti-6Al-7Nb is improved by conducting THPH as compared with the case where THP is conducted. This improvement is due to the plastic deformability of unstable β phase because the lattice constant of β phase in each alloy conducted with THPH is much greater than that of each as-cast alloy.     

Young’s modulus and mechanical properties of Ti-29Nb-13Ta-4.6Zr in relation to α″ martensite

An investigation on the formation of α″ martensite and its influence on Young’s modulus and mechanical properties of forged Ti-29Nb-13Ta-4.6Zr (wt pct) alloy is reported in this article. For ice-water-quenched specimens after solution treatment at 1023, 1123, and 1223 K in the single β-phase field for 1.8, 3.6, 14.4, and 28.8 ks, X-ray diffraction and internal friction measurements showed that the volume fraction of the α″ martensite changes with both solution temperature and time. This effect has been attributed mainly to the influence of grain size of the β-parent phase on the stability of the β phase and, consequently, on the martensitic start (M _s) temperature. A critical grain size of 40 μm was identified for the β phase, below which the martensitic transformation is largely suppressed because of low M _S temperature. With the β grain size increasing above this critical value, the volume fraction of the α″ martensite increases significantly at first and then decreases gradually with further grain growth. The α″ martensite was shown to possess good ductility and, compared to the β phase, lower strength and hardness but nearly identical Young’s modulus in the studied alloy.     

The effects of alloy microstructure refinement on the short-term thermal oxidation of NiCoCrAlY alloys

Three γ + β NiCoCrAlY alloys (a cast alloy, a laser-surface-melted (LSM) alloy, and a coating as deposited by electron beam-physical vapor deposition (EB-PVD)) with similar average composition (Ni-20Co-19Cr-24Al-0.2Y in at. pct), but with different microstructures prior to oxidation, were oxidized for 0.5 and 1 hours at 1373 K in an Ar 20 vol pct O₂ atmosphere (i.e., at a partial oxygen pressure of 20 kPa). It was found that on the alloy with β precipitates larger than 20 μm, the oxide layer was nonuniform in thickness, and had a laterally inhomogeneous composition and phase constitution. In this case, the oxide layer developed on top of the γ phase was thicker than that formed on top of the β phase and consisted of a NiCr₂O₄/Cr₂O₃ outer and an α-Al₂O₃ inner layer. For the thinner oxide formed on top of the β phase, the outer layer was constituted of a Cr and Co containing NiAl₂O₄ spinel and the inner layer also consisted of α-Al₂O₃. For the alloys with β precipitates smaller than 3 μm, a uniform and laterally homogeneous oxide formed, consisting of a Cr and Co containing NiAl₂O₄ outer layer on top of an α-Al₂O₃ inner layer. After oxidation, Y was distributed as numerous, small precipitates within the oxide layer for a homogeneous Y distribution prior to oxidation, or as a few, very large pegs along the γ/β phase boundaries of the alloy for an inhomogeneous Y distribution prior to oxidation. The performance of the alloys upon thermal cycling was improved for smaller β precipitates and for a more homogeneous Y distribution in the alloy prior to oxidation.     

Quantitative analysis on boundary sliding and its accommodation mode during superplastic deformation of two-phase Ti-6Al-4V alloy

A study has been made to investigate boundary sliding and its accommodation mode with respect to the variation of grain size and α/β volume fraction during superplastic deformation of a two-phase Ti-6Al-4V alloy. A load relaxation test has been performed at 600 °C and 800 °C to obtain the flow stress curves and to analyze the deformation characteristics by the theory of inelastic deformation. The results show that grain matrix deformation (GMD) is found to be dominant at 600 °C and is well described by the plastic state equation. Whereas, at 800 °C, phase/grain boundary sliding (P/GBS) becomes dominant and is fitted well with the viscous flow equation. The accommodation mode for fine-grained microstructures (3 μm) well agrees with the isostress model, while that for large-grained structures (11 μm) is a mixed mode of the isostress and isostrain-rate models. The sliding resistance analyzed for the different boundaries is lowest in the α/β boundary, and increases on the order of α/β≪α/α≈β/β, which plays an important role in controlling the superplasticity of the alloys with various α/β phase ratios.     

The effects of stress level and grain size on the ambient temperature creep deformation behavior of an alpha Ti-1.6 wt pct V alloy

Ambient (room) temperature studies have been carried out on an α-Ti-1.6 wt pct V alloy to determine the effects of stress level and grain size on ambient temperature creep behavior. Creep tests were performed at five different stress levels ranging from 75 to 95 pct of the yield stress value on specimens with an average grain size of 226 μm. It has been found that the alloy exhibits appreciable creep at stress levels far below the yield stress, with creep occurring at values as low as 75 pct of the yield stress. The extent of creep strain was found to decrease with a decrease in stress level. Creep tests were also performed on this alloy with different grain sizes ranging from 38 to 226 μm at a stress level of 90 pct of the yield stress. It was seen that the extent of creep strain decreased with a decrease in grain size. Fine slip and time-dependent twinning were found to be the creep deformation mechanisms. Based on the results of this investigation and earlier studies, it is suggested that time-dependent twinning is a major creep deformation mechanism in α-titanium alloys that contain small amounts of alloying elements. The time-dependent twinning phenomenon has been attributed to the diffusion of oxygen away from the twin-matrix interface, permitting the growth of twins.     

The effect of grain size on the shock-loading response of 304-type stainless steel

The residual microstructure and mechanical response of shock-loaded stainless steel (AISI-304) of four different grain sizes—23, 55, 85 and 187 μm-was investigated. In addition to mechanical twinning and planar dislocation arrays, transformation to both ɛ and α martensite occurred in all shock-loaded specimens but became more extensive with decreasing grain size. In comparison to the Hall-Petch behavior of yield and early flow stress observed for the material after 5.2 pet cold rolling, the strengthening efficiency of shock loading decreased with increasing grain size. Shock loading enhanced the strain-induced transformation to α martensite during subsequent tensile deformation.     

The effect of grain size on the ambient temperature creep deformation behavior of a beta Ti-14.8 V alloy

The effect of grain size in the range of 18 to 350 μm on the ambient-temperature creep behavior of a metastable β Ti-14.8 pct V alloy was investigated at a stress level of 95 pct yield stress. The main modes of deformation were found to be stress-induced plate (SIP) formation and slip. In the 350-μm grain-size material, the plates were found to nucleate and grow with time, thereby contributing to the ambient-temperature creep strain. The extent of SIP formation and growth were found to decrease with decreasing grain size, i.e., creep strain was found to decrease with decreasing grain size. The SIPs observed during creep were found to be similar to those observed in an earlier investigation during tensile testing, which were reported to be {332}〈113〉 twins.     

Microstructural evolution of a nanocrystalline Ti-47Al-3Cr alloy during annealing in the α + γ-phase field

Prealloyed, gas-atomized (GA) Ti-47Al-3Cr alloy powder, containing about 70 pct of the α ₂ (Ti₃Al) phase and 30 pct of the γ (TiAl) phase, was fully amorphized by mechanical alloying. The amorphous phase was stable during heating to 600 °C, but decomposed at higher temperatures, with an exothermic reaction peak at 624 °C as the material transformed to a mixture of α ₂ and γ and then to a fully γ structure at 722 °C. A nanocrystalline compact with a mean grain size of 42 nm was obtained by hot isostatic pressing (HIP’ing) of the amorphous powder at 725 °C. Isothermal annealing experiments were conducted in the two-phase α+γ field, at 1200 °C, using holding times of 5, 10, 25, and 35 hours, followed by air cooling. The X-ray diffractometry and analytical transmission electron microscopy investigations carried out on annealed and air-cooled specimens revealed only the presence of the γ grains, which coarsened on annealing. Initially, the grains grew, followed by a saturation stage after annealing for 25 hours, with a saturation grain size of about 1 μm. This grain growth and saturation behavior can be described with a normal grain growth mechanism in which a permanent pinning force is taken into account. Twins formed in the γ grains as a result of annealing and air cooling and exhibited a common twinning plane of (111) with the matrix phase. The minimum γ grain size in which twinning occurred in the annealed specimens was determined to be 0.25 μm, which suggests that twinning is energetically unfavorable in the nanometer-sized grains.     

Grain size and grain growth in an equiaxed alpha-beta titanium alloy

Methods of revealing grain size in a two-phase α-β titanium alloy have been examined and observations on beta grain growth in the presence of alpha have been carried out. The technique proposed by Greenfield and Margolin¹ for revealing β matrix grain sizes has been shown not to produce grain growth. However, for grain sizes of about 10 μm the G.M. technique does not reveal all the grains because of the similarity in orientation in neighboring grains. These clusters of similarly oriented grains are shown to persist as grain growth takes place but the misorientation between grains within a cluster decreases. Both the beta grain growth and alpha particle coarsening follow the same time dependency from which it is shown that a linear relationship exists between α particle size and β grain size. It is proposed that α particles must dissolve from theβ grain edges for β grain growth to occur. The linear dependency between beta grain size,D _β, and alpha particle size,d _α, can be rationalized either on the basis of geometrical or surface tension considerations. Formerly with New York University. Formerly Graduate Student with New York University.     

Superplastic behavior of two ultrahigh boron steels

The high-temperature deformation behavior of two ultrahigh boron steels containing 2.2 pct and 4.9 pct B was investigated. Both alloys were processedvia powder metallurgy involving gas atomization and hot isostatic pressing (hipping) at various temperatures. After hipping at 700 °C, the Fe-2.2 pct B alloy showed a fine microstructure consisting of l-μm grains and small elongated borides (less than 1μm) . At 1100 °C, a coarser microstructure with rounded borides was formed. This alloy was superplastic at 850 °C with stress exponents of about two and tensile elongations as high as 435 pct. The microstructure of the Fe-4.9 pct B alloy was similar to that of the Fe-2.2 pct B alloy showing, in addition, coarse borides. This alloy also showed low stress exponent values but lacked high tensile elongation (less than 65 pct), which was attributed to the presence of stress accumulation at the interface between the matrix and the large borides. A change in the activation energy value at theα-γ transformation temperature was seen in the Fe-2.2 pct B alloy. The plastic flow data were in agreement with grain boundary sliding and slip creep models. J.A. JIMéNEZ, Postdoctoral Fellow, formerly with Centro Nacional de Investigaciones Metalurgicas, C.S.I.C.     

An evaluation of the flow behavior during high strain rate superplasticity in an Al-Mg-Sc alloy

An Al-3 pct Mg-0.2 pct Sc alloy was fabricated by casting and subjected to equal-channel angular pressing to reduce the grain size to ∼0.2 μm. Very high tensile elongations were achieved in this alloy at temperatures over the range from 573 to 723 K, with elongations up to >2000 pct at temperatures of 673 and 723 K and strain rates at and above 10⁻² s⁻¹. By contrast, samples of the same alloy subjected to cold rolling (CR) yielded elongations to failure of <400 pct at 673 K. An analysis of the experimental data for the equal-channel angular (ECA)-pressed samples shows consistency with conventional superplasticity including an activation energy for superplastic flow which is within the range anticipated for grain boundary diffusion in pure Al and interdiffusion in Al-Mg solid solution alloys.     

The effect of metallic elements on the crystallization behavior of amorphous Fe-Si-B alloys

The crystallization behavior of amorphous Fe_84-X Si₆B₁₀M_X (M=Nb, Zr, V, or Cu) alloys was examined using differential scanning calorimetry (DSC) and transmission electron microscopy (TEM) with the aim of clarifying the effect of additional M elements. The compositional dependence of the first crystallization temperatureT _x1 increased in the order of Zr > Nb > V; however, the addition of 1 at. pct Cu caused a decrease inT _x1. Such an effect of the M elements on the thermal stability of an amorphous phase was interpreted in terms of the difference in the atomic size. These alloys were composed of a mixed structure ofα-Fe and amorphous phases after aging for 3.6 ks in the first exothermic temperature range. The addition of more than 3 at. pct Nb or Zr significantly affected the morphology and grain size of theα-Fe phase. However, their particles possessed dendritic morphology with a grain size of 0.1 to 0.3 μm, when the Nb or Zr content was less than 2 at. pct. Further addition of these elements brought about the formation of sphericalα-Fe particles. The average grain size, for instance, was as small as 20 nm in the aged alloy containing 6 at. pct Nb, which shows that a remarkable grain refinement occurs with increasing Nb content.     

Rapid solidification effects in martensitic Cu-Zn-AI Alloys

The effects of rapid solidification on martensitic transformations were studied in Cu-Zn-AI samples prepared by the method of melt-spinning, with an estimated cooling rate of about 10⁶ K per second near the freezing point. A diffusionless solidification reaction L → β occurs, and a very fine-grained β structure is obtained, with highly structured grain boundaries. The average β grain diameter (∼5 μm) is about two orders of magnitude smaller than that obtained by conventional solid state solution and quench treatment. The β:β grain boundaries contain extraordinary features such as large steps, and the matrix dislocation density is abnormally high. The M^s temperature is depressed significantly in as-melt-spun ribbon material, but as the martensitic transformation is cycled, it shifts upward in temperature and obtains a more narrow hysteresis loop. The martensite has the usual 9R structure (ABCBCACAB stacking) found in bulk alloys, and while the morphology is similar to that in bulk alloys, it is finer in scale. It is suggested that the β → 9R transformation is affected through the combined influence of rapid solidification on parent β grain size, disorder, β:β grain boundary structure, internal stresses, and dislocation substructure. Shape memory behavior is qualitatively similar in the rapidly solidified alloys.     

Texture evolution and mechanical anisotropy in dual-phase Ti<Subscript>3</Subscript>Al-based alloy loaded at 700 °C to 1000 °C

The super α ₂ Ti₃Al-based alloy with a fine grain size of ∼2.2 μm exhibits superplastic elongations over 1000 pct at 920 °C to 1000 °C, 600 pct at 900 °C, 330 pct at 850 °C, and 140 pct at 750 °C. Mechanical anisotropy is observed in this alloy, and relatively lower flow stresses and higher tensile elongations are obtained in the 45 deg specimen loaded at 25 °C to 960 °C. The texture characteristics appear to impose significant influence on the mechanical anisotropy at temperatures below 900 °C (under the dislocation creep condition), and the {111}〈2 〉 and {0001} basal textures evolve in the β and α ₂ phases after tensile straining. At loading temperatures higher than 900 °C (under the superplastic flow condition), the anisotropy effect is less pronounced and the grain orientation distribution becomes basically random in nature. Rationalizations for the mechanical anisotropy in terms of the Schmid factor calculations for the major and minor texture components in the β and α ₂ phases provide consistent explanations for the deformation behavior at lower temperatures as well as the initial straining stage at higher temperatures.     

A thousandfold creep strengthening by Ca addition in die-cast AM50 magnesium alloy

The effect of calcium addition on the microstructure and creep strength of the die-cast AM50 magnesium alloy was investigated. The α-Mg grains with the diameter of 4.9 μm are surrounded by the eutectic phases for the AM50-1.72 mass pct Ca alloy, while the β(Mg₁₇Al₁₂) particles are located mainly on the grain boundaries of the α grains for the AM50 alloy. The minimum creep rates of the AM50-1.72 mass pct Ca alloy are three orders of magnitude lower than those of the AM50 alloy at 423 K typically below 120 MPa. The thousandfold creep strengthening by the Ca addition is ascribed to the thermally stable eutectic phases appearing in the AM50-1.72 mass pct Ca alloy, which is expected to yield effective grain boundary strengthening or to resist the plastic flow of the α-Mg grains.     

Effects of volume fraction of tempered martensite on dynamic deformation properties of a Ti-6Al-4V alloy having a bimodal microstructure

The effects of the volume fraction of tempered martensite on the tensile and dynamic deformation properties of a Ti-6Al-4V alloy having a bimodal microstructure were investigated in this study. Five microstructures having various tempered-martensite volume fractions were obtained by varying heat-treatment conditions. Dynamic torsional tests were conducted on them using a torsional Kolsky bar. The test data were analyzed in relation to microstructures, tensile properties, and adiabatic shear-band formation. Under a dynamic loading condition, the maximum shear stress increased with increasing tempered-martensite volume fraction, whereas the fracture shear strain decreased. Observation of the deformed area after the dynamic torsional test indicated that a number of voids initiated mainly at α-phase/tempered-martensite interfaces, and that the number of voids increased with increasing martensite volume fraction. Adiabatic shear bands of 6 to 10 μm in width were formed in the specimens having lower martensite volume fractions, while they were not formed in those having higher martensite volume fractions. The possibility of adiabatic shear-band formation was explained by concepts of absorbed deformation energy and void initiation.     

Microstructure and mechanical properties of Fe-Ni-Cr-Al steel wires produced by in-rotating-water spinning method

Nonequilibrium austenite, γ, or duplex austenite + lath martensite,γ+ α′ _L, phase wires with high strengths and large elongation have been produced in Fe-Ni-Cr-Al-C alloy system by the in-rotating-water spinning method in which a melt stream is ejected into a rotating water layer. These wires have a circular cross section and a white luster, and the wire diameter is in the range of 80 to 180 μm. The γ phase has a grain size as small as about 1 to 4 μm. The yield strength, Σ_y, tensile fracture strength, ay, and elongation, ε_p, are about 340 to 655 MPa, 440 to 975 MPa, and 12 to 22 pct for the γ single phase wires and about 465 to 865 MPa, 640 to 1350 MPa, and 2 to 18 pct for the α′_L+γ duplex phase wires. A cold drawing causes significant increases in Σ_y and Σ_f, and the attained values are about 3200 MPa and 4030 MPa for Fe-8Ni-12.5Cr-2.5Al-3C wire drawn to about 95 pct reduction in area owing to the formation of a strain-induced α′_L phase and a remarkable work-hardening ability of γ and α′_L phases. On the subsequent low-temperature annealing around 673 K, the Σ_y and Σ_f increase further to 4000 MPa and 4240 MPa, respectively, probably because of the enhancement of the interaction between dislocations and interstitial carbon atoms. Around the temperature (≃800 K) where the γ phase decomposes into a stable mixed structure of α + ordered bec compound + M₇C₃ on annealing, the ε_p decreases drastically and the fracture surface morphology changes from a dimple pattern to a cleavage pattern. It has been therefore inferred that the high strengths and good ductility of the melt-quenched y and γ + α′_L wires are due to the suppression of the phase transformation of y to a mixed structure of γ + ordered bec compound + M₇C₃ carbide by the melt-quenching technique.     

Microstructure and mechanical properties of Fe-Ni-Cr-Al steel wires produced by in-rotating-water spinning method

Nonequilibrium austenite, γ, or duplex austenite + lath martensite,γ+ α′ _L, phase wires with high strengths and large elongation have been produced in Fe-Ni-Cr-Al-C alloy system by the in-rotating-water spinning method in which a melt stream is ejected into a rotating water layer. These wires have a circular cross section and a white luster, and the wire diameter is in the range of 80 to 180 μm. The γ phase has a grain size as small as about 1 to 4 μm. The yield strength, Σ_y, tensile fracture strength, ay, and elongation, ε_p, are about 340 to 655 MPa, 440 to 975 MPa, and 12 to 22 pct for the γ single phase wires and about 465 to 865 MPa, 640 to 1350 MPa, and 2 to 18 pct for the α′_L+γ duplex phase wires. A cold drawing causes significant increases in Σ_y and Σ_f, and the attained values are about 3200 MPa and 4030 MPa for Fe-8Ni-12.5Cr-2.5Al-3C wire drawn to about 95 pct reduction in area owing to the formation of a strain-induced α′_L phase and a remarkable work-hardening ability of γ and α′_L phases. On the subsequent low-temperature annealing around 673 K, the Σ_y and Σ_f increase further to 4000 MPa and 4240 MPa, respectively, probably because of the enhancement of the interaction between dislocations and interstitial carbon atoms. Around the temperature (≃800 K) where the γ phase decomposes into a stable mixed structure of α + ordered bec compound + M₇C₃ on annealing, the ε_p decreases drastically and the fracture surface morphology changes from a dimple pattern to a cleavage pattern. It has been therefore inferred that the high strengths and good ductility of the melt-quenched y and γ + α′_L wires are due to the suppression of the phase transformation of y to a mixed structure of γ + ordered bec compound + M₇C₃ carbide by the melt-quenching technique.     

Surface hardening of Ti alloys by gas-phase nitridation: Kinetic control of the nitrogen surface activity

A new concept for surface hardening of titanium alloys has been developed and successfully applied to Ti-6Al-4V alloys: gas-phase nitridation under kinetic control of the nitrogen activity. This method avoids the formation of detrimental second-phase nitrides by nitriding under a very low nitrogen activity, combined with rapid diffusion of nitrogen into the specimen. The surface hardness of the Ti-6Al-4V alloy was increased by a factor of 2 to ≈ 12 GPa, with only modest attenuation of ductility. We have realized conditions for generating case depths of ≈ 25 μm in reasonable nitridation times. The nitrogen activity in the gas phase is generated by heating a powder pack of Cr and Cr₂N. A closed two-zone system allows the powder pack and the specimen to be at different temperatures, optimizing both the nitrogen partial pressure and the nitrogen diffusion into the specimen. This low-cost, conformal nitridation process generates a smoothly graded nitrogen concentration profile and can be applied to finished Ti alloy components.