Elsevier

Acta Materialia

Volume 58, Issue 9, May 2010, Pages 3303-3312
Acta Materialia

A study of the microstructural evolution during selective laser melting of Ti–6Al–4V

https://doi.org/10.1016/j.actamat.2010.02.004Get rights and content

Abstract

Selective laser melting (SLM) is an additive manufacturing technique in which functional, complex parts can be created directly by selectively melting layers of powder. This process is characterized by highly localized high heat inputs during very short interaction times and will therefore significantly affect the microstructure. In this research, the development of the microstructure of the Ti–6Al–4V alloy processed by SLM and the influence of the scanning parameters and scanning strategy on this microstructure are studied by light optical microscopy. The martensitic phase is present, and due to the occurrence of epitaxial growth, elongated grains emerge. The direction of these grains is directly related to the process parameters. At high heat inputs it was also found that the intermetallic phase Ti3Al is precipitated during the process.

Introduction

Selective laser melting (SLM) is one of the new additive manufacturing techniques that emerged in the late 1980s and 1990s [1]. During the SLM process, a product is formed by selectively melting successive layers of powder, e.g. Ti–6Al–4V powder, by the interaction of a laser beam. Upon irradiation, the powder material is heated and, if sufficient power is applied, melts and forms a liquid pool. Afterwards, the molten pool solidifies and cools down quickly, and the consolidated material starts to form the product. After the cross-section of a layer is scanned, the building platform is lowered by an amount equal to the layer thickness, in this research 30 μm, and a new layer of powder is deposited. This process is repeated until the product is completed. The non-irradiated material remains in the building cylinder and is used to support the subsequent layers. After the process, the unused powder is sieved and can be reused. To counteract curling of the material due to the build-up of thermal stresses during the SLM process, the part is built on a solid substrate. Due to the high reactivity of Ti alloys, the process needs to be conducted under an inert argon atmosphere. A typical SLM machine layout is presented in Fig. 1 [2].

Compared to conventional manufacturing techniques, SLM offers a wide range of advantages, namely a lower time-to-market, a near-net-shape production without the need of expensive moulds, a high material utilization rate, direct production based on a CAD model, and a high level of flexibility (e.g. products with different geometry can be produced in the same batch). Moreover, due to the additive and layer-wise production, the SLM process is capable of producing complex geometrical features that cannot be obtained using conventional production routes.

Unfortunately, this new production technique has to deal with some frequently observed problems. SLM is characterized by high temperature gradients, which results in the build-up of thermal stresses, and a rapid solidification, which gives rise to the occurrence of segregation phenomena and the presence of non-equilibrium phases. The stability, dimensions and behaviour of the melt pool will determine to a great extent the porosity and the surface roughness, other than the roughness created by the layer-wise building (i.e. the staircase effect). The limited availability of some particular materials in powder form and the extensive research that still is needed to optimize the process for a given material also restrict the material range that can currently be processed.

A better understanding of the process is necessary to deal with the problems mentioned above. However, SLM is a complicated process with a wide range of non-equilibrium phenomena taking place, depending on a large number of parameters. Not only the laser parameters, such as the laser power or the laser spot diameter, or the scanning parameters, such as the scanning velocity or scanning strategy determine the process, but also the material properties, such as the surface tension or the thermal conductivity and environmental conditions play a large role. An extensive treatment of the influence of the different process parameters on the SLM process is given in Ref. [2].

Previous research has concentrated entirely on the influence of the process parameters on the product properties such as the surface roughness and relative density [3], [4], [5], [6], or has investigated the obtained mechanical properties [7], [8], [9] and the feasibility of the SLM process for applications in, for example, the electronic, biomedical and aeronautical industries [10], [11], [12], [13], [14], [15].

Although the microstructure offers an understandable link between the process and the resulting mechanical, physical and even chemical properties, only limited research has included a microstructural characterization [2], [9]. It was found that the materials processed by SLM exhibit a very fine, non-equilibrium structure. In the case of Ti–6Al–4V, an acicular martensitic structure is found. However, neither the influence of the process parameters nor the influence of the scanning strategy on the development of the microstructure during the SLM process was taken into account in previous research. In this paper, a thorough description of the microstructure of Ti–6Al–4V formed during the SLM process is given and explained, while the influence of the process parameters is investigated.

Section snippets

Materials and methods

The material used for this investigation is the extra-low-interstitial grade of Ti–6Al–4V alloy. The powder, produced by a plasma atomization process, is spherical with a particle size between 5 and 50 μm. Half of the particle volume has a powder size smaller than 34.43 μm. The apparent density of the powder, measured following ASTM B212-48, is 59%.

All parts were made on the in-house developed LM machine of the PMA Division (Production Engineering, Machine Design and Automation) of KU Leuven.

Results

This research was aimed at obtaining an insight into the development of the microstructure of Ti–6Al–4V during SLM. Therefore, the microstructure of the reference sample will be discussed first. To determine the influence of the process parameters in the development of the specific microstructure, two different sets of samples will be studied afterwards. In the first set, the energy density applied to the samples will be varied by varying either the scanning velocity or the hatch spacing. In

Discussion

Due to the high temperature gradients during the SLM process, the microstructure of Ti–6Al–4V is martensitic [9], [17]. The martensitic phase present in Ti–6Al–4V is the acicular martensitic phase, or the α′ phase, which is hexagonally packed. This is confirmed by X-ray diffraction (XRD) measurements. The XRD diffraction patterns indicate the presence of a hexagonal phase with lattice parameters a = 0.293 nm and c = 0.467 nm. These values correspond well to the lattice parameter values given in

Conclusion

The specific process conditions, i.e. short interactions, high temperature gradients and the high localization of the SLM process, lead to a specific microstructure for Ti–6Al–4V.

  • The fast cooling gives rise to a martensitic phase.

  • Due to partial remelting of the previous layers, elongated grains of several hundred micrometers grow.

  • The direction of the elongated grains depends on the local heat transfer condition which is determined by the scanning strategy (and the part geometry).

  • Segregation of

Acknowledgements

The authors acknowledge KU Leuven for support through the Project GOA/2002-06.

References (21)

  • J.P. Kruth et al.

    Ann CIRP

    (2007)
  • I. Yadroitsev et al.

    Appl Surf Sci

    (2007)
  • L.E. Murr et al.

    J Mech Behav Biomed Mater

    (2009)
  • D.A. Hollander et al.

    Biomater

    (2006)
  • P. Rochus et al.

    Acta Astronaut

    (2007)
  • I. Yadroitsev et al.

    Appl Surf Sci

    (2007)
  • F. Verhaeghe et al.

    Acta Mater

    (2009)
  • Rombouts M. PhD thesis. KU Leuven;...
  • R. Morgan et al.

    J Mater Sci

    (2004)
  • R. Morgan et al.

    J Mater Sci

    (2002)
There are more references available in the full text version of this article.

Cited by (2365)

View all citing articles on Scopus
View full text