Observations of room-temperature creep recovery in titanium alloys

Conventional α(hcp) and α(hcp)/β(bcc) titanium alloys exhibit significant primary creep strains at room temperature and at stresses well below their macroscopic yield strength. It has been previously reported in various materials systems that repeated unloading during primary creep testing may either accelerate or retard the accumulation of creep strains. These effects have been demonstrated to depend on both microstructure and the applied stress. This article demonstrates that significant room-temperature recovery occurs in technologically relevant titanium alloys. These recovery mechanisms are manifested as a dramatic increase in creep rates (by several orders of magnitude) upon the introduction of individual unloading events, ranging from 1 minute to 365 days, during primary creep tests. Significant increases in both creep rate and the total accumulated creep strain were observed in polycrystalline single α-phase Ti-6Al, polycrystalline α/β Ti-6Al-2Sn-4Zr-2Mo-0.1Si, and individual α/β colonies of Ti-6242. Based on transmission electron microscopy (TEM) studies of the active deformation mechanisms, it is proposed that the presence of significant stress concentrations within the α phase of these materials, in the form of dislocation pileups, is a prerequisite for significant room-temperature recovery. M.F. SAVAGE, formerly with the Department of Materials Science and Engineering, The Ohio State University Columbus, OH. This article is based on a presentation made in the symposium entitled “Defect Properties and Mechanical Behavior of HCP Metals and Alloys” at the TMS Annual Meeting, February 11–15, 2001, in New Orleans, Louisiana, under the auspices of the following ASM committees: Materials Science Critical Technology Sector, Structural Materials Division, Electronic, Magnetic & Photonic Materials Division, Chemistry & Physics of Materials Committee, Joint Nuclear Materials Committee, and Titanium Committee.     

Observations of room-temperature creep recovery in titanium alloys

Conventional α(hcp) and α(hcp)/β(bcc) titanium alloys exhibit significant primary creep strains at room temperature and at stresses well below their macroscopic yield strength. It has been previously reported in various materials systems that repeated unloading during primary creep testing may either accelerate or retard the accumulation of creep strains. These effects have been demonstrated to depend on both microstructure and the applied stress. This article demonstrates that significant room-temperature recovery occurs in technologically relevant titanium alloys. These recovery mechanisms are manifested as a dramatic increase in creep rates (by several orders of magnitude) upon the introduction of individual unloading events, ranging from 1 minute to 365 days, during primary creep tests. Significant increases in both creep rate and the total accumulated creep strain were observed in polycrystalline single α-phase Ti-6Al, polycrystalline α/β Ti-6Al-2Sn-4Zr-2Mo-0.1Si, and individual α/β colonies of Ti-6242. Based on transmission electron microscopy (TEM) studies of the active deformation mechanisms, it is proposed that the presence of significant stress concentrations within the α phase of these materials, in the form of dislocation pileups, is a prerequisite for significant room-temperature recovery. This article is based on a presentation made in the symposium entitled “Defect Properties and Mechanical Behavior of HCP Metals and Alloys” at the TMS Annual Meeting, February 11–15, 2001, in New Orleans, Louisiana, under the auspices of the following ASM committees: Materials Science Critical Technology Sector, Structural Materials Division, Electronic, Magnetic & Photonic Materials Division, Chemistry & Physics of Materials Committee, Joint Nuclear Materials Committee, and Titanium Committee.     

Characterization and modeling for forging deformation of Ti-6Ai-2Sn-4Zr-2Mo-0.1 Si

The deformation behavior during upset forging has been determined for Ti-6242 in both the (α + β) and β starting microstructures. For (α + β), flow softening attributed to deformation heating was observed. Deformation heating accounted for only a fraction of the extensive flow softening of the β microstructure. The dependence of log σ on 1/T was linear for (α + β) and bilinear for the β microstructure, with an approximate transition temperature of 930 °C. The two temperature regimes for β corresponded to distinct deformed microstructures which were manifestations of different softening mechanisms, all promoting the flow-induced transformation of metastable β microstructure to the equilibrium (α + β) microstructure. Based on the experimental data, flow stress equations for both microstructures, and empirical equations describing the flow softening behavior of β have been developed. WithT and ġe as the only input variables, these equations can accurately predict the σ - ε relationships for process modeling of this alloy.     

Finite-Element Modeling of Titanium Powder Densification

A powder-level, finite-element model is created to describe densification, as a function of applied stress during uniaxial hot pressing, of CP-Ti and Ti-6Al-4V powders with spherical or spheroidal shapes for various packing geometries. Two cases are considered: (1) isothermal densification (in the α- or β-fields of CP-Ti and in the β-field of Ti-6Al-4V) where power-law creep dominates and (2) thermal cycling densification (across the α/β-phase transformation of Ti-6Al-4V) where transformation mismatch plasticity controls deformation at low stresses. Reasonable agreement is achieved between numerical results and previously published experimental measurements and continuum modeling predictions.     

A Study on Alpha-Case Depth in Ti-6Al-2Sn-4Zr-2Mo

Isothermal oxidation experiments in air were performed on Ti-6Al-2Sn-4Zr-2Mo (Ti-6242) with a bimodal microstructure in the temperature range 811 K to 922 K (538 °C to 649 °C) for up to 500 hours, and α-case depths were quantified using metallography. Alpha-case depth followed a parabolic variation with time. Alpha-case depths in excess of 10 μm formed above 811 K (538 °C) and 100-hour exposures. An activation energy of 244 kJ/mol was estimated for diffusion of oxygen in the α phase of Ti-6242.     

Crystallography of grain boundary α precipitates in a β titanium alloy

The crystallography of α(hcp) precipitates formed on the β(bcc) matrix grain boundaries has been studied with transmission electron microscopy (TEM) in a Ti-15V-3Cr-3Sn-3Al alloy. The α precipitates have a near-Burgers orientation relationship with respect to at least one of the adjacent β grains. Among the possible 12 variants in this orientation relationship, the variant that [11•20]_α is parallel to the 〈111〉_β closest to the grain boundary plane tends to be preferred by the α precipitates. Additionally, further variant selections are made so as to minimize the deviation of orientation relationship with respect to the “opposite“ β grain from the Burgers one. Such rules in variant selection often result in the formation of precipitates with a single variant at a planar grain boundary. Prior small deformation of β matrix changes the variant of α precipitates at the deformed portion of grain boundary. It is considered that the stress field of dislocations in the slip bands intersecting with the boundary strongly affects the variants of α precipitates. Discussion of these results is based upon a classical nucleation theory. Formerly Graduate Student, Department of Materials Science and Engineering, Kyoto University Formerly Graduate Student, Department of Materials Science and Engineering, Kyoto University This article is based on a presentation made during TMS/ASM Materials Week in the symposium entitled “Atomistic Mechanisms of Nucleation and Growth in Solids,“ organized in honor of H.I. Aaronson’s 70th Anniversary and given October 3–5, 1994, in Rosemont, Illinois.     

Microstructure and mechanical properties of Ti-40 Wt Pct Ta (Ti-15 At. Pct Ta)

Of the β-isomorphous Ti-X alloy systems, Ti-Ta is one of the least studied. In the current work, the microstructure and mechanical properties of Ti-40 wt pct Ta (Ti-15 at. pct Ta) are investigated. Annealing at 810 °C produces a two-phase microstructure consisting of Ti-richa idiomorphs in a continuous Ta-rich β matrix; this suggests the β-transus temperature is higher than indicated by the most recently published phase diagram. Water quenching from 810 °C causes the β phase to partially transform to orthorhombic martensite (α), while furnace cooling yields secondarya The primary α formed isothermally remains unchanged in both cases. Subsequent aging causes transformation of the martensite to type 1a plus residual β, with a corresponding increase in strength and decrease in ductility. The maximum ductility (20 pct elongation) occurs in the water-quenched condition in which metastable β is retained. Analysis of the true stresstrue strain data suggests that transformation-induced plasticity may contribute to the enhanced ductility of the water-quenched material.     

Microstructure and phase relations for Ti-Mo-Al alloys

The influence of aluminum additions to a Ti-7 at. pet Mo alloy on the phase equilibria was investigated. The microstructures of the alloys, Ti-7 pct Mo-7 pct Al and Ti-7 pct Mo-16 pct Al, were determined by light and electron microscopy. It was found that with increasing aluminum concentration the formation of the metastable w phase was suppressed. In the Ti-7 pct Mo-16 pct Al alloy the β phase decomposed upon quenching by precipitating coherent, ordered particles having a B2 type of crystal structure (β₂). At low temperatures the equilibrium phases for this alloy were β + α+ β ₂, whereas at high temperature (850° to 950°C) the Ti₃Al phase was in two-phase equilibrium with the β phase. The four-phase equilibrium which exists at a temperature of about 550°C involves the reaction β + Ti₃Al ⇌ α + β₂. G. LUETJERING, formerly Staff Member Materials Research Center, Allied Chemical Corp., Morristown, N. J.,     

Deformation of an alloy with a lamellar microstructure: experimental behavior of individual widmanstatten colonies of an α-β titanium alloy

The deformation behavior of individual Widmanstatten colonies comprised of aligned lamellae of ductile phases has been investigated. Based on the α β Ti alloy, Ti-8Al-lMo-lV, this study shows the existence of a large (>2X) variation in the critical resolved shear stress for yielding of individual colonies. Schmid’s Law is not obeyed except for prism slip parallel to the β lamellae. In addition, colonies with a high yield stress exhibit a high work hardening rate and fine, uniform slip. This behavior appears to be independent of the α-phase slip system (basal, prism, or pyramidal) and of the microstructure (α βvs α α′ (martensite)). The experimental behavior is correlated to several colony orientation parameters including the stress axis, the slip plane, the slip direction and the orientation of the α β interface. The yield stress of a colony is found to increase as the slip direction of the dominant macroscopic slip plane approaches normality to the α β interface. These results indicate that the macroscopic flow behavior of colonies comprised of ductile lamellae depends on the ability of a slip system, once activated in the softer phase, to shear through the harder phase. The data also indicate that the interaction stresses at the phase interfaces are not a principal factor controlling macroscopic yielding in α β Ti alloys. Finally, the alignment of a slip system in the α-phase with a potential slip system in the β-phase lamellae does not appear to affect the yield stress strongly. formerly Graduate Students, Department of Metallurgical Engineering, Michigan Technological University     

Deformation-Induced Grain Refinement and Amorphization in Ti-10V-2Fe-3Al Alloy

The microstructural evolution and grain refinement mechanisms of a Ti-10V-2Fe-3Al alloy, β-solution quenched and cold forged (CF) to strains of 0.1, 0.35, and 1.2 have been investigated using optical microscopy (OM), X-ray diffraction (XRD), and transmission electron microscopy (TEM). The results showed that the stress-induced martensitic transformation became a predominant deformation mode in the metastable Ti-10V-2Fe-3Al alloy during cold forging. These martensites α″ repeatedly divided the original β parent phase into a large number of micron-sized blocks when the forging strain was 0.1. Shear bands were observed to traverse α″/β lamellae and resulted in a significant grain refinement of the β phase, as the forging strain increased to 0.35. The degree of grain refinement inside shear bands was higher than the outside. Nanocrystalline and amorphous structures were produced in local areas of the original β phase, when the forging strain rose to 1.2. This dramatic grain refinement in the metastable Ti-10V-2Fe-3Al alloy could be attributed to the stress-induced martensitic transformation promoting the initiation and growth of shear bands across α″/β lamellae. More dislocations were produced and accumulated inside grains to accommodate plastic deformation. The crystal structure was collapsed and an amorphous structure was formed as soon as the dislocation density was accumulated to a critical value of 10¹⁴/cm². Moreover, some of the reverse martensitic phase transformation, α″→β, was observed to contribute to grain refinement of Ti-10V-2Fe-3Al alloy as well.     

A Self-Consistent Approach for Modeling the Flow Behavior of the Alpha and Beta Phases in Ti-6Al-4V

The flow behavior of the α and β phases in Ti-6Al-4V was interpreted in the context of a self-consistent modeling formalism. For this purpose, high-temperature compression tests were conducted at various temperatures for a single-phase α alloy (Ti-7Al-1.5V), a variety of near-β alloys, and the two-phase alloy Ti-6Al-4V, each with an equiaxed microstructure. The flow behavior of the α phase in Ti-6Al-4V was deduced from the experimental results of the single-phase α alloy. The flow behavior of the β phase, which was predicted by using the self-consistent approach and the measured flow behaviors of Ti-6Al-4V and Ti-7Al-1.5V, showed good agreement with direct measurements of the various near-β alloys. From these results, it was shown that the strength of the α phase is approximately three times higher than that of the β phase at temperatures between 1088 K and 1223 K (815 °C and 950 °C). It was also concluded that the relative strain rates in the two phases varies significantly with temperature. The usefulness of the approach was confirmed by comparing the predicted and measured flow stresses for other Ti-6Al-4V and near-α alloys.     

Observations on the faceted initiation site in the dwell-fatigue tested ti-6242 alloy: Crystallographic orientation and size effects

A novel use of the electron backscattered diffraction (EBSD) characterization technique for study of fracture has been demonstrated. This new approach has been employed for characterization and analysis that contribute to the understanding of crack initiation in Ti-6242 under dwell-fatigue loading conditions. A faceted crack initiation site is typically observed on the dwell-fatigue fracture surface of Ti-6242. The level of microtexture has a major influence on the dwell-fatigue failures in near-α titanium alloys, such as Ti-6242. In this study, serial sectioning and EBSD techniques were used to obtain the orientation images of almost the entire specimen cross section at different depths below the fracture surface. The orientation images are color coded on three different bases: the angle between the loading axis and basal plane normal, the Schmid factor for prism slip, and the Schmid factor for basal slip. The aim was to determine the important aspects of the crystallographic orientation and the size of the microtextured region that is associated with the faceted initiation site. The results of this study are used to explain the possible locations of crack initiation in a test specimen under dwell-fatigue loading condition. These results are also used to better understand the role of size of microtextured regions in determining which crack will outgrow the other cracks (for the case of multiple cracking typically observed in the alloy of current study under the dwell-fatigue loading conditions) to become the dominant crack that leads to eventual specimen failure. This understanding has important practical implications because the dominant crack effectively determines the specimen life.     

Determination of dihedral angle distributions in polycrystals: Application to {itα} + {itβ} brass

A method is described to determine the distribution of dihedral angles of a given type in a polycrystalline aggregate from the measured distribution of angles in random sections. Each of the two angular distributions (2D and 3D) is divided inton classes and the probabilityp _ik that a dihedral angle in the 3D classk originates an angle in the 2D classi is calculated. Ann ×n matrixP = (_p) is then introduced to relate the two distributions. The 9 × 9 matricesP andP ^{- 1} corresponding to intervals of 20 deg are given in the paper. The method is applied to the determination of the distributions of dihedral angles atααβ andαββ triple lines in aα + β brass, submitted to a long annealing treatment at 620 °C. Using the average angles and the standard deviations of the distributions, the relative values of the interfacial tensionsact, ββ, and αβ were calculated. It was found that γ_αα ≃γ _ββ ≃ 14γ_αβ.     

On the decomposition of β phase in some rapidly quenched titanium-eutectoid alloys

The decomposition of the β phase in rapidly quenched Ti-2.8 at. pct Co, Ti-5.4 at. pct Ni, Ti-4.5 at. pct, and 5.5 at. pct Cu alloys has been investigated by electron microscopy. During rapid quenching, two compctitive phase transformations, namely martensitic and eutectoid transformation, have occurred, and the region of eutectoid transformation is extended due to the high cooling rates involved. The β phase decomposed into nonlamellar eutectoid product (bainite) having a globular morphology in Ti-2.8 pct Co and Ti-4.5 pct Cu (hypoeutectoid) alloys. In the near-eutectoid Ti-5.5 pct Cu alloy, the decomposition occurred by a lamellar (pearlite) type, whereas in Ti-5.4 pct Ni (hypereutectoid), both morphologies were observed. The interfaces between the proeutectoid α and the intermetallic compound in the nonlamellar type as well as between the proeutectoid α and the pearlite were often found to be partially coherent. These findings are in agreement with the Lee and Aaronson model proposed recently for the evolution of bainite and pearlite structures during the solid-state transformations of some titanium-eutectoid alloys. The evolution of the Ti₂Cu phase during rapid quenching involved the formation of a metastable phase closely related to an “ω-type” phase before the equilibrium phase formed. Further, the lamellar intermetallic compound Ti₂Cu was found to evolve by a sympathetic nucleation process. Evidence is established for the sympathetic nucleation of the proeutectoid a crystals formed during rapid quenching.     

The microstructure of rapidly solidified Ti-Fe melt-spun ribbons

Ti-Fe binary alloys were rapidly solidified by the melt-spinning technique, and four compositions were examined: Ti-5 wt pct Fe, which is the critical composition for theβ to ω athermal transformation; Ti-10 wt pct Fe, which represents a hypoeutectoid composition; the eutectoid composition Ti-15 wt pct Fe; and Ti-20 wt pct Fe, as an example of a hypereutectoid alloy. The Ti-5 wt pct Fe rapidly solidified ribbons are composed of two different structures. The first consists of α′-martensite plates inβ matrix and the second, athermal ω particles inβ matrix. The Ti-10, 15, and 20 wt pct Fe alloys are also composed of two structures. These areβ grains and isothermal-like ω particles inβ matrix. A solidification model is suggested which explains the existence of two different microstructures at the same composition and the for-mation of two kinds of ω particles.     

Characterisation of solution annealed VT-14 titanium alloy using eddy current based electrical resistivity measurements

This paper discusses characterisation of microstructures in Ti-4.5Al-3Mo-1V (VT-14) alloy specimens solutionized in the temperature range of 923K–1373K, using room temperature electrical resistivity measurements. Using phase angle of impedance change in eddy current coil as a parameter, calibration curve between resistivity and phase angle has been established using three reference specimens of known resistivity. The changes in the electrical resistivity of the specimens have been correlated with optical microscopy investigations. The electrical resistivity of unstable β phase has been compared with α and α′ phases and possible reasons for higher resistivity of β phase and lower resistivity of α′ have been given based on the rule of mixtures and scattering and mobility of electrons. The electrical resistivity of α′ martensite phase of the chosen VT-14 alloy has been determined as 1.04 αΩ-m. The studies reveal that it may be possible to identify the β transus temperature of α+β titanium alloys from the electrical resistivity data.     

Comparison of the fatigue and fracture of α+β and β titanium alloys

The present study compares the fatigue and fracture properties of the high-strength β titanium alloy β-Cez with the conventional α+β titanium alloy Ti-6Al-4V, because of increasing interest in replacing α+β titanium alloys with β titanium alloys for highly stressed airframe and jet engine components. This comparison study includes the Ti-6Al-4V alloy in an α+ β-processed condition (for a typical turbine blade application) and the β-Cez alloy in two distinctly different α+β-processed and β-processed conditions (optimized for a combination of superior strength, ductility, and fracture toughness). The comparison principally showed a much lower yield stress for Ti-6Al-4V (915 MPa) than for both β-Cez conditions (1200 MPa). The Ti-6Al-4V material also showed the significantly lower high-cycle fatigue strength (resistance against crack initiation) of 375 MPa (R=−1) as compared to the β-Cez alloy (∼600 MPa, R=−1). Particularly in the presence of large cracks (>5 mm), the fatigue crack growth resistance and fracture toughness of the Ti-6Al-4V material is superior when compared to both β-Cez conditions. However, for small crack sizes, the conditions of both the alloys under study show equivalent resistance against fatigue crack growth. For the β-Cez material, where microstructures were optimized for high fracture toughness (conventional large crack sizes) by thermomechanical processing, maximum K _Ic-values of 68 MPa√m of the β-processed β-Cez condition (tested in the longitudinal direction) decreased by ∼50 pct in the presence of small cracks (1 mm). A similar decrease in fracture toughness was obtained by loading the β-processed β-Cez condition perpendicular to the flat surfaces of the pancake-shaped β grain structure (tested in the short transverse direction). These results were discussed in terms of the effectiveness of the crack front geometry in hindering crack propagation. Further, the results of this study were considered for alloy selection and optimized microstructures for fatigue and fracture critical applications. Finally, the advantage of the α+β-processed β-Cez condition in highly stressed engineering components is pointed out because of its overall superior combination of fatigue crack initiation and propagation resistance (especially against small fatigue cracks).     

Surface relief of α″ martensite in a Ti-Mo alloy

The surface relief of α″ martensite plates in a polycrystalline Ti-4.74 at. pct Mo alloy was studied by atomic force microscopy (AFM). The orientation of matrix grains was measured by electron backscatter diffraction (EBSD), and measured surface tilt angles were compared with calculation by the crystallographic theory of martensite transformation. The observed maximum tilt angle was close to the predicted value of 7.57 deg. The overall agreement between measured and calculated tilt angles was improved significantly by taking into account not only the inclination of habit plane to the specimen surface, but also the shear direction predicted from the theory. The tilt angle may vary with the moving direction of the interface unless the habit plane is perpendicular to the specimen surface. However, this effect was small in this transformation.     

The effect of deformation-induced transformation on the fracture toughness of commercial titanium alloys

The effect of deformation-induced transformation of metastableβ phase on the ductility and toughness of four commercial titanium alloys was investigated. Tensile tests, Charpy impact tests, and both static and dynamic fracture toughness tests were carried out at temperatures between 77 and 473 K on four titanium alloys containing metastableβ phase. Deformation-inducedα″ (orthorhombic martensite) was observed in an (α + β)-type Ti-6Al-2Sn-4Zr-6Mo alloy. The dynamic fracture toughness of this alloy increased considerably at 223 K compared to those at other temperatures. In another (α + β)-type Ti-6A1-4V alloy, the static fracture toughness at 123 K and the dynamic fracture toughness at 223 K were increased considerably by the presence of deformation-induced martensite compared to those at other temperatures. The strength increased as the temperature decreased in this alloy. An abnormal elongation of aβ-type alloy, Ti-15V-3Al-3Sn-3Cr, at 123 K was attributed to the mechanical twinning of theβ phase. However, the effect of deformation-induced transformation on the fracture toughness of Ti-3Al-8V-6Cr-4Mo-4Zr alloy was not observed. Formerly Visiting Associate Professor, Department of Metallurgical Engineering and Materials Science, Carnegie Mellon University, Pittsburgh, PA. Formerly with the Department of Production Systems Engineering, Toyohashi University of Technology.     

Application of a modified jogged-screw model for creep of TiAl and α-Ti alloys

Stress exponents for creep, in the range of 5, are typically associated with dislocation creep processes, normally associated with a strong tendency for subgrain formation. In this article, we will demonstrate that there are several important alloy systems that have similar stress dependence and, yet, lack this tendency for subgrain formation. Specifically, dislocations in the intermetallic compound γ-TiAl and the hexagonal close-packed (hcp) α phase of the commercial Ti alloy, Ti-6242, tend to be homogeneously distributed with a tendency for alignment along screw orientation. In both alloy systems, the screw dislocations exhibit a large density of pinning points, which detailed transmission electron microscopy (TEM) investigation indicate are locations of tall jogs. These observations suggest that the jogged-screw model for creep should be appropriate after suitable modification for the presence of these tall jogs. This modified jogged-screw (MJS) model is presented, together with a discussion of the assumptions made, and the results of this model are shown to compare favorably with experiment for both alloy systems. The possible criteria for the formation of tall jogs are also described, and the potential application of this modified model to other alloy systems is discussed. This article is based on a presentation made in the workshop entitled “Mechanisms of Elevated Temperature Plasticity and Fracture,” which was held June 27–29, 2001, in San Diego, CA, concurrent with the 2001 Joint Applied Mechanics and Materials Summer Conference. The workshop was sponsored by Basic Energy Sciences of the United States Department of Energy.