Elsevier

Acta Materialia

Volume 188, 15 April 2020, Pages 720-732
Acta Materialia

Full length article
Mechanism of stress relaxation and phase transformation in additively manufactured Ti-6Al-4V via in situ high temperature XRD and TEM analyses

https://doi.org/10.1016/j.actamat.2020.02.056Get rights and content

Abstract

Additive manufacturing is being increasingly used in the fabrication of Ti-6Al-4V parts to combine excellent mechanical properties and biocompatibility with high precision. Unfortunately, due to the build-up of thermal residual stresses and the formation of martensitic structure across a wide range of typical processing conditions, it is generally necessary to use a post-thermal treatment to achieve superior mechanical performance. This investigation aims to obtain a deeper understanding of the micro/nanostructural evolution (α′ martensite phase decomposition), accounting for the kinetics of phase transformation during the heat treatment of 3D-printed Ti-6Al-4V alloy. As the mechanism of phase transformation and stress relaxation is still ambiguous, in this study the changes in crystal lattice, phase, composition and lattice strain were investigated up to 1000°C using both in situ high temperature X-ray diffraction (XRD) and transmission electron microscopy (TEM). Based on the result a mechanism of phase transformation is proposed, via the accommodation/substitution of Al, V and Ti atoms in the crystal lattice. The proposed mechanism is supported based on elemental concentration changes during heat treatment, in combination with changes in crystal structure observed using the high temperature XRD and TEM measurements. This study provides a deeper understanding on the mechanism of phase transformation through martensitic decomposition, as well as a deeper understanding of the influence of post-thermal treatment conditions on the alloy's crystal structure.

Introduction

Additive manufacturing (AM) has become increasingly important in the aeronautical, medical device and automotive sectors, due to its ability to produce consolidated parts for specific end-use applications [1]. Titanium and its alloy Ti-6Al-4V (Ti64) are the preferred materials in the AM production due to their corrosion resistance, biocompatibility, high specific strength, superior mechanical properties and fracture toughness [2]. Furthermore, the mechanical performance of AM processed Ti64 is equivalent if not better compared to Ti64 components produced with traditional routes [3]. In AM processes, distinctively laser-based systems, melted layers experience extremely high cooling rates (~103 − 108 K/s), which lead to the formation of metastable acicular martensitic phase with hexagonal close packed (HCP) crystal structure, whereas conventional processing (~0.5 K/s) yields typical α + β (equilibrium) microstructure exhibiting excellent fatigue performance [4], [5], [6]. In addition, thermal cycling occurs by successive layer deposition and melting during additive manufacturing which results in the formation of residual stresses. Liu Y et al. reported that during deposition of a new layer on the previously scanned layers, compressive stress increases within the underlying layers due to thermal cycling while tensile stress within the top layer is converted into compressive stress [7]. As a result, extensive research has been carried out in an attempt to optimise the as-fabricated parts to reduce and/or eliminate residual stresses through tailoring AM process parameters [8], [9], [10], [11], [12], [13]. Haider et al., for example, determined that residual stresses can be reduced by optimising parameters such as laser power, scan speed and scan pattern, whereby the cooling rate and temperature gradient is lowered [9].

A second common method used to address some of the fabrication limitations is through the post-processing treatment of parts. Hot isostatic pressing (HIP) and heat treatment (Annealing) are two methods used in the industry to enhance the performance of AM fabricated parts [14]. The improved properties are achieved by inducing microstructural changes during heat treatment, more specifically, grain growth. This growth can be controlled through the tailoring of the thermal arrest period and heating rates. For example, it was observed that by using a heating rate of 3.3°C/min and 10°C/min, resulted in a net grain growth of 0.49% and 0.32%, respectively, after annealing for 1 h at 1015°C [14]. HIP has become the preferred method for the closure of internal pores in both cast and AM produced parts. The process is conducted on chemically clean components, in a heated, argon-filled pressure vessel, usually at pressures of 69 to 103 MPa (10–15 ksi) between 900 − 955°C achieved using a heating rate of 5°C/min [14,15]. However, due to the high cost associated with HIP, optimised as-built parts with low porosity are commonly heat treated. The standard industrial aeronautical post-thermal treatment is based on DIN 17869, which states that stress relief heat treatment is recommended for component manufacturing using multi-layered weld seams. The process parameters are based on indications of DIN 65083 for thermal treatment of cast components made of titanium and titanium alloys for aerospace [15,16]. Uhlmann et al. for example investigated the effect of thermal treatment on the microstructure of printed parts under protective gas using a heating rate of 5°C/min, holding period of 60 min at 675°C followed by inert gas cooling based on these standards to [17].

As additive manufacturing has gained considerable industrial interest, the need to optimise the process has resulted in several studies focused on determining the optimal annealing strategy to relieve internal residual stresses and achieve tailored phase composition with application specific mechanical properties. Elmer et al. [18], for example, observed a lattice contraction of 0.57% after annealing for 2 h at 450°C, while Combres [19] observed complete relaxation after annealing at 730°C for 2 h. Vracken et al. reported that the fine martensitic structure was fully transformed to a mixture of α and β after heat treatment of 2 h at 780°C, with the α phase present as fine needles [20]. Vilaro et al. examined the effect of conventional and optimized heat treatments on the mechanical behaviour of parts. In their studies it was observed that the gradual decomposition of α′, achieved by annealing at 850°C for 2 h followed by air cooling, produced an increase in hardness in comparison to annealing at lower temperatures [21]. Coincidently, Zhang et al. reported an improvement in ductility after heat treatment at a temperature range of 850 − 900°C for 2 h followed by furnace cooling due to the full decomposition of α′ [13]. In contrast, Ter Haar and Becker [22] claim that superior tensile properties are achieved by duplex annealing (part held at 950°C for 4 h, followed by water quenching), resulting in bi-modal microstructure when compared to standard annealing strategies. Evidently, most investigations focus on understanding the effect of varying heat treatment cycles on the mechanical performance based on the final part microstructure. Therefore, a deeper understanding of the micro/nanostructural evolution (α′ martensite phase decomposition) accounting for the kinetics of phase transformation is critical, in order to optimise heat treatment procedures applied on 3D-Printed Ti-6Al-4V alloy. In 2002 Pederson used high temperature X-ray diffraction (HT-XRD) to study the phase transformation kinetics in cast Ti64 during heating up, isothermal hold and cooling down. In his studies, the thermal expansion behaviour was examined based on changes in both phase fraction and d-spacings [23]. Equivalently, in this study the processing conditions used for in situ characterisation are representative of that commonly used in industry for the additive manufacturing of Ti64 alloy.

Martensitic α′ is conveniently viewed as a distorted α-hcp structure with smaller lattice parameters (a = 2.931; c = 4.681 Å, with c/a = 1.597) [24]. A recent study revealed that after work hardening, equilibrium α and martensitic α′ phases can coexist as dual-phase α + α′ microstructure in a matrix of (α + β) lamellar [25]. It has been reported that during heat treatment, diffusion of aluminium (Al), vanadium (V) and titanium (Ti) occurs as the microstructure changes [26], [27], [28], [29]. Matsumoto et al. proposed that during solution treatment, V enrichment occurs as Al decreases in the α′ martensite [26]. Similarly, Xu et al. reported that during annealing, martensitic decomposition results in the segregation of Al in the α phase [27]. In contrast, Barriobero-Vila et al. argue that the overall V and Al concentration decreases in favour of Ti during the α′ → α + β transformation [28]. Furthermore, variation in phase composition has been reported to influence the lattice parameters due to differences in atomic radii [29].

Clearly, there is no overall agreement in literature on the mechanisms of martensite phase transformation, which occur during the thermal treatment of additively manufactured Ti64 alloy. The goal of this study is to investigate this mechanism by tracing the changes in crystal lattice, composition and lattice strain via in situ high temperature characterisation techniques. This should not only allow for the optimization of post-processing treatment of AM Ti64 alloy, but also provide a fundamental understanding to design new Ti-alloys for 3D-Printing.

Section snippets

Sample manufacturing and preparation

Test samples were additively manufactured using extra-low interstitial Ti64 (Grade 23, ELI-0406) powder, with particle size in the range of 10 – 45 µm. The print studies were carried out using the production scale powder bed system, Renishaw RenAM 500M equipment, which uses a 500 W laser with wavelength of 1070 nm. A detailed description of the additive manufacturing process can be found in the authors’ prior work [30]. Parts with dimensions 10 × 10 mm were built to facilitate in situ

Results

Fig. 1 exhibits a schematic representation of the additively manufactured sample and the corresponding orientations of the extracted lamellae used in in situ TEM analyses. Upon inspection, the samples exhibited structural features possessing high aspect ratios scattered across the surface, which resembles needles, a characteristic of martensitic phase (Fig. 1(a) and (d)) [36]. Examination of the side view reveals long, columnar prior β-grains, which grow epitaxially along the build direction

Discussion

In situ experiments up to 1000°C have demonstrated that there is a stepwise expansion of the Ti64 lattice, with an accompanied relief of internal residual stress and decomposition of α′ (martensite) → α + β (equilibrium). It is known that materials used in layer-by-layer deposition are subjected to cyclic heating and cooling which results in a primarily martensitic microstructure [4]. Comparatively, the cooling rate between selected laser melting (SLM) and traditional strategies such as water

Conclusion

In this study the mechanism of stress relaxation and phase transformation of Ti64 alloy fabricated using additive manufacturing, was observed by high temperature in situ XRD and TEM analyses. Based on experimental results it was observed that stress relaxation of the crystal lattice occurs as the alloy is heated from 25 to 400°C, without any evidence of phase transformation. Above 400°C, α′ (martensite) → α + β (equilibrium) phase transformation was determined to occur in a stepwise manner. The

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

This publication was supported in part by Science Foundation Ireland (SFI) through the I-Form Advanced Manufacturing Research Centre, (Grant Number 16/RC/3872). All TEM imaging and analysis including EDS was carried out at the Advanced Microscopy Laboratory (AML) at the AMBER centre, CRANN Institute, Trinity College Dublin, Ireland. The AML is an SFI supported imaging and analysis centre. VN would also wishes to acknowledge the European Research Council (CoG 3D2D Print) and Science Foundation

Reference (47)

  • W. Xu et al.

    In situ tailoring microstructure in additively manufactured Ti-6Al-4V for superior mechanical performance

    Acta Mater.

    (2017)
  • P. Barriobero-Vila et al.

    Role of element partitioning on the α–β phase transformation kinetics of a bi-modal Ti–6Al–6V–2Sn alloy during continuous heating

    J. Alloys Compd.

    (2015)
  • J.W. Elmer et al.

    In situ observations of lattice expansion and transformation rates of α and β phases in Ti–6Al–4V

    Mater. Sci. Eng. A

    (2005)
  • M. Canavan et al.

    Novel in-situ lamella fabrication technique for in-situ TEM

    Ultramicroscopy

    (2018)
  • A.R. Ghasemi et al.

    Measuring residual stresses in composite materials using the simulated hole-drilling method

    Residual Stresses Composite Mater.

    (2014)
  • T. Ungár

    Microstructural parameters from X-ray diffraction peak broadening

    Scr. Mater.

    (2004)
  • H. Ali et al.

    In-situ residual stress reduction, martensitic decomposition and mechanical properties enhancement through high temperature powder bed pre-heating of Selective Laser Melted Ti6Al4V

    Mater. Sci. Eng. A

    (2017)
  • J. Yang et al.

    “Formation and control of martensite in Ti-6Al-4V alloy produced by selective laser melting”

    Mater. Design

    (2016)
  • S.Q. Wu et al.

    Microstructural evolution and microhardness of a selective-laser-melted Ti–6Al–4V alloy after post heat treatments

    J. Alloys Compd.

    (2016)
  • R. Sivakami et al.

    “Estimation of lattice strain in nanocrystalline RuO 2 by Williamson–Hall and size–strain plot methods”

    Spectrochim. Acta Part A

    (2016)
  • W. Xu et al.

    Additive manufacturing of strong and ductile Ti–6Al–4V by selective laser melting via in situ martensite decomposition

    Acta Mater.

    (2015)
  • D. Thomas

    Costs, benefits, and adoption of additive manufacturing: a supply chain perspective

    Int. J. Adv. Manufactur. Technol.

    (2016)
  • W. Toh et al.

    Microstructure and wear properties of electron beam melted Ti-6Al-4V parts: A comparison study against as-cast form

    Metals

    (2016)
  • Cited by (132)

    View all citing articles on Scopus
    View full text