Influence of dimension, building position, and orientation on mechanical properties of EBM lattice Ti6Al4V trusses

A lattice structure is a space-filling unit cell that can be tessellated along any axis with no gaps between cells; hence, its architecture is formed by a collection of spatial periodic unit cells [1]. Such type of structures has been increasingly attracting the research community for the possibility to tailor their micro-architecture to match specific and multifunctional design constraints [2, 3], to design lightweight structures, as well as for their ability to provide high specific mechanical properties [4‐7]. The mechanical behavior of the lattice structure can be easily changed by varying the topology of the unit cell and the size of the struts [8]. Examples of lattice structure applications are lightweight structural components, energy absorbers, and heat exchangers [9‐11]. The use of lattice structures has been hindered in the past by the fact that the traditional subtractive technologies could not be used due to the several difficult and expensive cutting and welding processes mandatory for their manufacturing [12]. The advent of additive manufacturing (AM) technologies [13, 14] for metallic materials allowed designers to manufacture components/parts that were not conceivable at all until few years ago due to the previous manufacturing limits. Lattice structures are the type of structures that have benefited the most from this innovation.

The possibility of using AM technologies [13‐15] to produce complex metallic lattice structures at adequate costs increased the interest of the scientific community towards such type of structure. Several efforts have been spent to evaluate the properties of fully three-dimensional lattice structures made by AM processes [16, 17]. The advantages of the AM processes, with respect to the traditional subtractive manufacturing technologies, are numerous: increased design freedom, minimum human interaction requirement, reduced design cycle time, and limited environmental impact. Such advantages have opened a wide range of new opportunities as the possibility to topologically optimize a mechanical component for enhanced performances (e.g., weight saving) [18], the part count reduction of assemblies [19, 20], but also the possibility to manufacture components with a prescribed lattice structure micro-architecture.

To these aims, electron beam melting (EBM) is one of the most promising AM processes since it allows producing three dimensional metal parts, built layer-by-layer, starting from metal powders that are melted by a high intensity electron beam [21, 22]. During the EBM building process, a thin layer of powder is distributed, heated and selectively melted. Subsequently to the layer melting phase, according to the instructions contained in the CAD (computer-assisted design) model of the component/part to be manufactured, the building table is lowered and the sequence is repeated until the whole product is obtained. The process, firstly introduced in 1997 by the Swedish company ARCAM, is currently a promising technology for the manufacturing of lightweight, durable and dense end parts; it is mainly used in aerospace, medical and defense industries [23‐26]. Similar to electron beam welding (EBW), the EBM process takes place in a vacuum chamber, providing a clean and controlled environment which avoids porosity formation inherent in many structural alloys such as Al-alloy [27, 28] as well as an excellent thermal insulation. Another unique peculiarity of EBM consists in its ability in keeping the entire build at high temperatures, above 1000 °C if needed. For each layer, the electron beam heats the entire powder bed leading the powder to an optimal temperature, specific for the used material. This ensures a good microstructure of parts free from residual stresses. It also eliminates the need of post process heat treatments, which has a significant impact on the total production costs. The high temperature process and the resulting stress-relieved components provide improved material properties better than cast and comparable to wrought [29, 30].

The EBM process can be successfully used to manufacture both massive parts, that for components having a circular cross-section it means diameter larger than 2 mm, and lattice structures, consisting basically of several truss-like elements having diameters often lower than 2 mm. Depending on the type of the structure to build, different process themes (i.e., a set of process parameters which are optimized to obtain the best results) can be used in the EBM process: the melt processing theme can be used to melt massive components, whereas the net processing theme is often used to melt lattice structures.

Titanium alloys possess excellent properties such as low density and good mechanical properties up to 550–600 °C and thus they are widely used in aerospace applications [31‐33]. The development of the EBM technology in processing titanium alloys for the manufacturing of massive parts has brought the technology to an advanced state of maturity. Indeed, several studies [34‐36] have been focused on the mechanical characterization of the Ti6Al4V alloy which is today one of the most promising alloys that can be processed via EBM. Further investigations on the EBM process, in terms of surface roughness, dimensional accuracy and defects distribution have been deeply faced in literature [37‐41]. On the contrary, the existing literature on the EBM manufacturing of lattice structures in Ti6Al4V [42‐45] is quite lack and investigations are needed to better assess their mechanical behavior and their possible use for actual structural applications. According to the existing literature, the EBM process has shown encouraging results in the manufacturing of Ti6Al4V lattice structures. However, some manufacturing imperfections, such as deviation of the real geometry from the nominal geometry defined in the CAD model as input for the process can affect negatively the performances of such lattice structures with respect to the expected ones. In particular, some researchers [29] highlighted that for thin EBM components, such as for lattice structures, both surface roughness and dimensional accuracy, which depend on the process parameters, play a crucial rule for the structural strength. With the main goal to encourage the use of lattice structures in aerospace, further studies are needed to investigate on the repeatability and the reproducibility of the process, also in terms of deviation from the expected dimensions of the product.

The mechanical characterization of AM octet truss structures should start by a well assessed approach as the building block one used for advanced aerospace structures. Figure 1 shows a scheme of such a typical approach used to mechanically characterize this type of structures.

Specifically, the first level of the proposed building block approach (the lowest one of the pyramid in Fig. 1), consists in the selection and characterization of the powders to use in the additive process; the second level consists in the mechanical characterization of the structural behavior of the single octet truss struts; the third level consists in the characterization of the elementary cell, subsequently followed by the fourth level consisting in the characterization of simple specimens made of repeated elementary cells, and finally by the characterization of the full-scale components. It has to be highlighted that the proposed approach can be supported by numerical analyses, e.g., when the experimental tests involve the unit cells, complex to be tested especially under tensile loads. This paper particularly focused on the second level, since the first one is quite established in literature, depending on manufacturing parameters and on the 3D printing technology under consideration.

The aim of this paper is to evaluate the influence of dimension, building position, and orientation on mechanical behavior of Ti6Al4V single truss-like elements made by EBM, carrying out an extensive experimental campaign on single beams representative of single truss-like elements. The experimental campaign was characterized by the usage of different nominal diameters for trusses, different growth orientations and different positions with respect to both building plate and height inside the chamber. Several Ti6Al4V samples (i.e., trusses) for tensile tests were designed and manufactured by EBM to assess the effects of such aspects on the mechanical properties.

Ti6Al4V plasma atomized powder with spherical morphology with a nominal diameter in the range of 45 ÷ 100 μm obtained by atomization process was used to manufacture the specimens for this study. The spherical shape improves flowability and ensures high build rates and part accuracy [45]. The powder flow rate measured according to ASTM B213 [46] was found to be 24 s/50 g. The apparent density measured according to ASTM B212 [47] was 2.6 g/cm3. As regarding the particle size distribution, the percentage by mass of particle size in the range 45 ÷ 106 μm was found equal to 91%; The powder nominal chemical composition according to the ASTM F2924 [48] is reported in Table 1.

Figure 2 shows some details of a typical lattice structure based on an octet truss elementary cell. Each elementary cell can be considered the combination of two regular tetrahedra and a central body consisting of two further tetrahedra sharing the same square base. In this paper, the attention has been focused on the single trusses composing the cell, whose mechanical behavior was assessed experimentally.

To this aim, the tensile specimens adopted in this study (Fig. 3) were conceived to have the same geometry of the beam composing an octet truss structure. The gauge length of specimens presented a circular section. Three different truss diameters d (the nominal diameters defined as input for the process) were investigated. The length of the calibrated section was set up to 25 mm, although this parameter was not of particular interest.

The CAD software used for the drawing of the lattice structures was Solid Edge®. The geometry of the specimen was modeled in two parts and two separate files were generated, one representing the specimen heads and the other one representing the gauge length, in order to assign different process themes to heads and to the gauge length. Further details are provided in the followings.

According to Fig. 4, specimens were manufactured for each of the directions defined for the beams with reference to the octet truss cell of Fig. 2. This was needed so as to quantify the influence of the truss growth directions on their mechanical behavior. Octet truss cells are expected to provide the same mechanical behavior irrespective of their position within the lattice structure. Therefore, specimens were manufactured at various locations within the building chamber (sized 210 × 210 × 380 mm × mm × mm) so as to check whether the position and the height of samples with reference to the building plate played a significant role [49]. The specimens were manufactured considering the four parameters described below:

Considering all the possible combinations of the abovementioned 4 parameters, 36 different combinations were needed. Additionally, these combinations were replicated twice. Therefore, a total number of 72 specimens were manufactured.

This paper aims to investigate on the mechanical behavior of the single octet truss struts composing the lattice unit cell, as the octet-truss ones, according to the building block approach herein proposed to characterize such type of engineering structures.

This paper focused the attention on the mechanical characterization of single Ti-6Al-4 V struts composing an elementary octet truss cell, printed through EBM technology. For this purpose, several tensile tests were carried out by changing some relevant parameters, such as struts diameter, growth orientation, in-plane position on the building plate, and height in the building chamber. The main mechanical properties of each specimen group were derived from test data to arrange a discussion in terms of stress–strain curves, Young’s modulus, ultimate tensile strength, yield stress, and elongation at fracture.

According to these test results, it was observed that the horizontally manufactured specimens presented a very reduced mechanical resistance as well as a more brittle behavior. In all specimens manufactured at θ = 0°, unmelted powders were observed to affect the outer surfaces of all specimens and the whole cross-sections. Moreover, for such group of specimens, it was found that the cross-section shape did not correspond to the circular section defined as input for the AM process. Cross-section of d = 1 mm specimens resulted very small and characterized by irregular profiles. Such aspect became less evident as the diameter increased, nonetheless still present. In general, the larger the diameter was, the higher the resistance of specimens resulted to be. This was attributed to the fact that when specimens with small diameters were built, large discrepancies were obtained between the nominal and the actual specimen cross-section area. Finally, incomplete melting of powder was observed at the outer surface of the specimens for all the other growing directions.

With respect to the specimen dimensions, increasing performances were observed as the specimen diameter increased. Concerning the specimen location, specimens manufactured at the central region of the building chamber and closer to the building plate resulted to be stiffer than the others. Centrally manufactured specimens, with respect to the building plate, reported σ_UTS and σ_YS values higher than those ones manufactured at the building plate corners. Similarly, specimens manufactured closer to the building plate were slightly more performant in terms of σ_UTS and σ_YS, even though this tendency seemed to be less significant.

Concerning the specimen orientations and excluding the θ = 0° data (less reliable due to defects), θ = 45° specimens appeared stiffer than 90° ones, presenting performances slightly higher than those ones vertically manufactured. σ_UTS of 45° specimens was found to be higher than that one measured for θ = 90° specimens.