1. Introduction
Due to suitable high-temperature properties, such as high specific high-temperature strength (100–150 MPa cm
3/g), high specific rigidity (35–42 GPa cm
3/g), high oxidation, and creep resistance, intermetallic γ titanium aluminide alloys with a density of ρ = 3.9–4.1 g/cm
3 are considered to be substitutes for nickel-based superalloys with approximately twice the density (ρ = 7.9–8.5 g/cm
3) [
1,
2]. The limited ductility and fracture toughness of the material [
3,
4], as well as the high reactivity of titanium, especially with oxygen, water, and nitrogen, are challenging both for the use and the production of titanium aluminide components. Applications are mainly in the aerospace industry and, in some cases, in the automotive and energy sectors [
1,
3]. One of the main applications is the use of low-pressure turbine blades at an operating temperature of approx. 700 °C. The best-known alloys currently in service are GE4822 (Ti-48Al-2Nb-2Cr, at.%), a two-pha TiAl alloy used by GE Aviation/Avio Aero in the GE90, CFM International LEAP (formerly LEAP X) and GEnX engines, and the TNM-B1 (B1 = 0.1 wt.-% B) [
5]. TNM
TM is a group of forging alloys fully solidified through the β phase and protected by MTU Aero Engines AG. Leistritz Turbinentechnik GmbH uses it to manufacture low-pressure turbine (LPT) blades [
5], which are installed in the geared turbofan PW1110G. The target element distribution of the main alloy elements is adjusted according to the desired distribution of the three phases—α2, γ, and β/β0(B2) [
1]—and is in the range Ti(42-44)Al-(3-5)Nb-(0.1-2)Mo-(0.1-0.2)B (in at.%).
The alloy investigated in this paper is designated TNM-B1. The nominal composition may differ, and the one used here is Ti-43.5Al-4Nb-1Mo-0.1B (in at.%) [
6]. The density is 4.16 g/cm³. The alloy solidifies via the β phase by means of β phase stabilizing elements, such as Mo and Nb [
7]. Diffusion processes in the α2 and γ phase are crucial for heat treatment results and slowed down by Nb. An increase in the activation energy of diffusion in these phases is achieved by the addition of Mo [
8]. Both serve to improve the creep properties. Boron is added in the range of 0.1–1 at% due to its grain refining and texture-reducing effect [
2,
4,
7]. The addition of boron leads to the formation of stable Ti-boride precipitates, which also inhibit grain growth during heat treatment.
Casting and forging processes are established for production, which are very complex due to the low ductility and fracture toughness as well as the high oxygen affinity of the intermetallic material. The industry’s interest in the tool-free additive production of titanium aluminide components is, therefore, great. Functional tests, e.g., at GE Avio Aero with blades manufactured by additive electron beam melting (EBM), are already being carried out [
9]. Other approaches in research include powder metallurgy (PM) via hot isostatic pressing (HIP) [
10], metal injection molding (MIM) [
11], or spark plasma sintering (SPS) [
12]. With the increasing use of γ-TiAl in aircraft turbines, the need for repair or modification solutions will increase, irrespective of the original production route of the components concerned. Currently, the research is performed on transient liquid phase (TLP) bonding for the repair of γ-TiAl, using Fe or Ni as melting point depressing elements (MPD) [
13]. Furthermore, in the recently completed LuFo-project (FKZ 20T1526A), the diffusion welding of TiAl was investigated with, again, the aim of the repair.
With direct energy deposition (DED), components can not only be built up additively but also be repaired near net-shape. The suitability of DED as an additive manufacturing process for coating and repair tasks has already been proven for numerous material groups (nickel-based alloys, steel materials, etc.). Especially the ability to build on 3D surfaces enables DED as a potential method to individualize parts manufactured via a different route.
Crack-free processing of titanium aluminides with DED is possible by preheating the substrate during the process [
14,
15,
16,
17,
18]. The temperature of the substrate or the already built-up volume required to produce crack-free traces must be in the brittle-ductile transition range (550–800 °C) for Ti-based alloys [
19]. The actual preheating temperature required depends on the alloy, the geometry, and the process parameters. To protect the melt from oxidation, the process zone is shielded locally with inert gas [
15] or a global inert gas atmosphere [
17,
20,
21,
22,
23,
24] with < 100 ppm [
20] or even smaller amounts of O [
25]. The oxygen content after DED has been measured in [
26,
27] at approx. 500–1200 ppm with complete shielding. The influence of process conditions and process parameters on the oxygen content in the material produced has hardly been investigated so far [
27], and the contents of <1000 ppm O in DED GE4822 and TNM-B1 have been documented in some publications [
15,
25,
28].
According to [
29], phase transformations can be achieved for TNM at temperatures between 1230 °C and 1300 °C, depending on the composition and thus, for example, the aluminum content. In [
30], results are published on the influence of heat treatment temperature, duration, and cooling rate. A typical composition for a TNM type alloy (MTU/Pratt Whitney) is approximately 79% γ, 25% α2, and 5% β [
31]. However, the heat treatment methods applied in the literature vary widely. Thus, a direct comparison of the microstructure and properties is hardly possible but crucial to determine whether materials from the same alloy are compatible, as the final part can be treated only as a whole.
Following this approach, samples in this work produced with different manufacturing methods, including conventional as well as additive manufacturing, were identically heat treated. The material was examined regarding microstructure and properties, taking into account the respective chemical composition. It was experimentally demonstrated that DED combined with cast and laser powder bed fusion (LPBF) resulted in dense, crack-free samples, which were analyzed as well. The properties of γ-TiAl produced with DED in the investigations carried out corresponded overall to those achieved with processes used to produce components made of γ-TiAl. DED was, therefore, suitable as a process for processing γ-TiAl. However, the oxygen content was still above the critical value of 1000 ppm, which reduced the ductility, especially at room temperature, and thus made reworking and assembly of components even more difficult.
4. Discussion
The high-temperature compressive strength (700 °C) of DED-manufactured material is approx. 1200 MPa, which is also comparable to that of the identically heat-treated cast, forged, and LPBF material. The biggest differences are found in the almost non-existent β0 phase and an approx. 15% higher γ phase content compared to cast, forged, and LPBF material, which can be attributed to an approx. 0.4% higher aluminum content of the DED samples. Since the oxygen content of all tested materials is similarly high at approx. 2000 ppm, this can be ruled out as the main cause of the differences between the materials. The overall low strength values for TNM-B1 can be explained both by the possibly less favorable heat treatment and the high oxygen content. Thus, the properties of γ-TiAl produced with DED correspond, in the context of the investigations carried out, overall to those achieved with processes used to produce components made of γ-TiAl.
The oxygen content of all samples is above the limit value of 1200 ppm. This allows for a more specific analysis of other influences in this study but is known to be too high for application purposes [
46]. The ductility, especially at room temperature, is highly influenced by O, which is detrimental for at least any post-processing. For cast and thus forged material, O contents are regularly lower, usually well below 800 ppm [
47]. Samples with higher content are deliberately selected in order to make the composition comparable and to be able to observe the influence of the different temperature-time curves of the production routes. For AM (Additive Manufacturing), the processing steps of powder production and the additive manufacturing itself will always lead to an unavoidable increase in oxygen. However, EBM (Electron Beam Melting) already allows us to minimize the uptake during AM close to zero due to the vacuum, which is why it is the most frequently selected AM method for processing γ-TiAl. At the same time, aluminum loss is comparatively higher [
48]. The resulting EBM microstructures [
49], are quite like the ones achieved with LPBF in this work, which suggests a transferability of the results shown.
The hardness of hybrid DED and cast samples varies measurably depending on the test area.
Contrary to other measurements, the hardness of the cast area is higher than that of the DED area in this test [
30,
32]. Due to the small grain sizes of the three-phase DED material, the positioning and size of the measuring points are particularly important. The transition zone shows the high lamellar microstructure of the cast with partly smaller grain sizes of DED so that there is a tendency to slightly higher hardness.
Due to the similarity of the material properties and the flawless transition zones, DED is generally suitable to be combined with AM, casting, or forging. The ability to build on 3D structures allows for repair and modification or individualization of series production parts likewise.