Elsevier

Materials Chemistry and Physics

Volume 183, 1 November 2016, Pages 238-246
Materials Chemistry and Physics

Orthorhombic martensite formation upon aging in a Ti-30Nb-4Sn alloy

https://doi.org/10.1016/j.matchemphys.2016.08.023Get rights and content

Highlights

  • A massive α″ martensite formation was observed after 24 h of heat treatment.

  • Martensite formation occurs in the vicinity of α phase laths.

  • Incorporation of Sn in the β phase reduces the strain needed to form α″ phase.

Abstract

The characteristics of orthorhombic martensite (α″) formed by step-quenching in a Ti-30Nb-4Sn (wt%) alloy have been investigated by transmission electron microscopy (TEM) and X-ray diffraction (XRD). According to literature, α″ lattice parameters depend mainly on the composition of the parent β phase. In this study, samples subjected to step quenching heat treatment presented α″ phase formation in the proximity of α phase laths, driven by two combined factors: solute rejection and lattice strain. Our results indicate that as the aging is prolonged, α″ becomes richer in solute content, which makes it more similar to the parent β phase. An average 2.55% lattice strain along [110]β directions was found to be necessary in order to obtain α″ from the β phase after 24 h of aging at 400 °C, followed by water-quenching. The initial lattice strain along the same direction was estimated at approximately 3.60% with zero aging time. The precipitation of the α phase does not inhibit a solute rich α″ phase formation.

Introduction

Titanium alloys have been extensively used in biomedical applications due to their biocompatibility, high specific strength, low elastic modulus and even superelasticity. These properties, however, depend a lot on microstructural features: the phases present, their morphology, volumetric fraction and behavior under strain [1], [2]. As an example, recent studies have shown the promising use of Ti-Nb-Sn alloys in functionally-graded biomaterials, with both the elastic modulus and the tensile strength being optimized for a femoral hip stem prosthesis application [3], [4].

Beta metastable titanium alloys display martensitic transformations from the β phase (bcc) to the α’ (hcp) or α″ (orthorhombic) phases at a specific solute content range. Among Ti-Nb alloys, the α″ phase is formed after water-quenching from the β phase field in alloys with an Nb content higher than 17.5 wt% [5] and less than 36.2 wt% [1]. The β phase can also be transformed into α″ at room temperature as a result of a stress induced martensitic (SIM) transformation [6], and thus the α″ phase can be reversed into a β phase after heating above the martensite start temperature (Ms), leading to an observable shape memory effect [7]. Furthermore, a few experiments with High Energy X-ray Diffraction (HEXRD) reported the formation of an α″-like phase during continuous heating or isothermal heat treatments, foregoing the initial stages of α (hcp) phase nucleation [8], [9]. This phenomenon was first observed by Duerig et al. [10] in Ti-10V-2Fe-3Al alloy, in which the isothermal α″ phase, presumably lean in Fe and V, has been formed at low temperatures (250–460 °C), favored by lower heating rates. At the time, the authors justified the lean α″ phase stabilization settled on the principle that the lattice strain needed to form α″ from the β phase is smaller than the strain needed to directly form the α phase, thus proposing the following phase transformation sequence during heating: β + α″ → β + lean α″ → β + lean α″ + α → β + α. The formation of α phase derived from the lean or isothermal α″ phase was recently elucidated by Barriobero-Vila et al. [11].

Several reports covered Ti-Nb based alloys in relation to the martensitic β/α″ transformation. According to Kim et al. [1], since there is a strict lattice correspondence between the β and α″ phases, the lattice transformation strain needed to form α″ from the β phase along a specific set of directions can be estimated using the lattice constants of each phase. The transformation strain needed is maximized when the loading axis is parallel to the [011]β directions. Furthermore, the transformation strain needed decreases with the increase of Nb, Sn and Zr contents as the martensite start (Ms) temperature declines [2]. Additionally, Bönisch et al. have demonstrated that the atomic rearrangement necessary to form α″ depends particularly on the parent β phase Nb content [12]. Liu et al. [13] identified clusters via HRTEM (high resolution transmission electron microscopy), which are rich in either Nb or Ti, proposing, in addition to the well-known α″ phase, the stress induced formation of another martensitic phase (δ martensite), also orthorhombic and apparently lean in Nb, in Ti-24Nb-4Zr-8Sn-0.1O (wt%) alloy. Most of these studies, however, were focused on the martensite formed from pure water quenching, or under stress/strain cycles [14], therefore none of them have properly characterized the α″-like phase formed upon isothermal heat treatments in Ti-Nb based alloys.

In this work, we aim to investigate the crystallographic characteristics of the α″ phase formed formerly and concurrently with the α phase upon an isothermal aging heat treatment in a Ti-30Nb-4Sn alloy. The samples were solution heat treated, step-quenched from 800 °C to 400 °C for 0.5 h, 8 h and 24 h, then subjected to water quenching (WQ). Samples were analyzed by scanning electron microscopy (SEM), transmission electron microscopy (TEM) and X-ray diffraction (XRD). Our results show crystallographic features of the α″ phase that are related to the Nb and Sn solute redistribution during aging.

Section snippets

Materials and methods

The Ti-30Nb-4Sn (wt%) ingot was prepared by melting high purity Ti (99.81%), Nb (99.99%) and Sn (99.99%) metals in an arc furnace under argon atmosphere. The ingot was homogenized at 1000 °C for 12 h followed by water quenching (WQ) and cold rolled to 2 mm thickness. The chemical composition of the experimental alloy (Table 1) was determined by X-ray Fluorescence Technique (Rigaku RIX 3100). O and N interstitial elements were evaluated by a LECO TC400 analyzer. Samples were cut and encapsulated

Previous data compiled from literature

All the electron and X-ray diffraction analyses have been performed based on the crystallographic data listed in Table 2, which shows some Inorganic Crystal Structure Database (ICSD) standards for the β, α and α″ phases. The α″ phase lattice parameters are displayed as a′, b′ and c′. Table 3 compiles the expected modification of the β and α″ lattice parameters with the addition of Nb and Sn, based on the results issued by Kim et al. [1] and Hao et al. [2]. Regarding the Sn addition, the β phase

Discussion

As already stated, there is a specific OR between the β and α″ phases, so the maximum lattice strain needed to form α″ from the β parent phase can be evenly estimated using the 3 × 3 lattice strain matrix T' presented by Kim et al. (2006) [1]:T'=[a'/a0000(b'+c')/22a0(b'c')/22a00(b'c')/22a0(b'+c')/22a0]

Taking a0=a(β)=3.316Å and using the XRD lattice parameters for the α″ phase obtained in each condition, the maximum lattice strain – i.e. the average along the [011] β crystallographic

Conclusions

In this work, the α″ and α phase formation upon isothermal step-quench heat treatment was successfully described. The following conclusions are highlighted:

  • 1.

    The longer the material is kept at 400 °C, the more the α″ orthorhombic phase resembles the parent β phase, in crystallographic terms. The lattice strain needed to form α″ from β along the [110]β axis drops from 3.60% (WQ) to 2.55% after 24 h of heat treatment.

  • 2.

    The α phase precipitation is detected only after 24 h of heat treatment. It occurs

Acknowledgements

The authors gratefully acknowledge the Brazilian research funding agencies FAPESP (#2012/10164-8 and #2013/50391-6) and CNPq (#484379/2012-7) for their financial support; Companhia Brasileira de Metalurgia e Mineração (CBMM) for the Niobium supplied and also the Brazilian Nanotechnology National Laboratory (LME/LNNano/CNPEM) and OSU (The Ohio State University) for the use of electron microscopy facilities and the technical support during the TEM work.

References (32)

Cited by (0)

View full text