Microstructure and Micro-Hardness Properties of In-Situ LENS Fabricated Ti-Al-Si-xV Alloys

. This study presents laser in-situ alloying of Ti-Al-Si-xV alloys fabricated using the laser engineered net shaping (LENS) machine from elemental powders. The as-built samples were subjected to heat treatment at 1200 o C, 1300 o C, and 1400 o C for 1 hour and furnace cooled (FC) with subsequent homogenization heat treatment at 950 o C for 6 hours and FC. The microstructure was characterized by scanning electron microscopy (SEM) equipped with an electron dispersion spectroscopy (EDS). The micro-hardness properties were evaluated with the Vickers hardness tester. The results showed that alloying via in-situ powder deposition followed by heat treatment is practicable for the producing TiAl-based alloys with improved mechanical properties.


Introduction
The main advantage of additive manufacturing (AM) over conventional manufacturing routes is high precision of manufactured parts [1]. This class of technology includes directed energy deposition (DED) AM, also known as laser engineered net shaping (LENS). In the past couple of years, materials such as metals, composites, ceramics, and functionally graded materials (FGM) including titanium aluminide-based alloys (TiAl-based) have been successfully processed using LENS technique [1][2][3][4][5]. Despite the fact that several alloys have been used to fabricate superior engineering structures, the fabrication of superior structures with TiAl intermetallic alloys has proven to be problematic.
Owing to combined unique physical, chemical, and mechanical properties that make titanium alloys uniquely suitable for the aerospace, marine, and chemical industrial sectors. Titanium alloys have widely been used for a variety of applications with respect to aerospace, distributions of the graded TiAl-based intermetallic alloys were also investigated by the Vickers hardness tester.

Methodology
The Optomec 850R LENS machine with 1 kW IPG fibre laser was used to fabricate the samples of Ti-Al-Si-xV alloys with different atomic weight percent (at.%) of V. This machine uses an IPG fiber laser as energy source and argon gas to maintain the inert environment. It uses a deposition head with four copper feeding nozzles and a workstation with the computer software program. Two of the powders (Ti and Al) were delivered from the hoppers of the Optomec machine while the other powders (Si and V) were fed from a GTV powder hopper externally attached onto the other hoppers of the LENS machine. The SEM secondary electron images (SEIs) of the metal powders used are presented in Fig. 1. The specimens were fabricated based on a pre-programmed CAD file loaded with the workstation software of the Optomec application version 3.1.6 which automatically controls and operate the deposition head. Spherical shaped powders from TLS Technik GmbH & Co of 45-90 µm particle size were used in this study. The samples were built on Ti-6Al-4V alloy substrate plate without any pre-heating. Sample cubes of 15 mm by 15 mm by 10 mm was built with build parameters listed in Table 1. The process set-up adopted in this study is similar as in ref [5,23] with the optimized processing parameters as indicated in Table 1. The as-produced specimen cubes were heat treated at a rate of 500 o C/hour in an argon-rich environment at three different temperatures of 1200 o C, 1300 o C, and 1400 o C for 1 hour and furnace cooled (FC) at 100 o C/hour. This was followed by homogenization heat treatment at 950 o C for 6 hours and FC.  The cube samples of the LENS fabricated Ti-Al-Si-xV alloys were sectioned transversely along the builds. Prior to performing morphological characterization, surfaces of the samples were mechanically ground and polished using standard metallographic procedures. After etching with Kroll's reagent (2% HF, 6% HNO3 92% distilled water all in volume fractions) allows the microstructures to be visibly examined by using SEM equipped with EDS. Micro-hardness measurement was carried out using Zwick/Roell Indentec (ZHVµ) Vickers micro-hardness tester. Diamond indenter with pyramidal shape having an angle of 136 o between opposite faces was used. A load of 500 gf with dwell time of 10 s was applied for the indentation. The micro-hardness was measured at 15 different locations and the average value was calculated.

Microstructure of graded Ti-Al-Si-xV Alloys
The SEIs in Fig. 2 presents the as-built Ti-Al-Si-xV alloys processed with 0.18 g/min, 0.33 and 0.49 g/min V additions and the compositions are shown in Table 2. It has been reported by Appel, Paul and Oehring [24], that V moves the (α2+γ)/γ lamellae towards the Ti dominated area thereby improving the ductility of TiAl-based alloys and reducing the Al content in the γ phase. These quaternary alloys of Ti-Al-Si-xV demonstrate (α2+γ)/γ dominating the microstructure prompted by the V additions in the alloy. The SEM image in Fig. 2 corresponds to 5.11 at.%, 9.42 at.% and 10.78 at.% for the Ti-Al-Si-xV alloys with 0.18 g/min, 0.33 and 0. 49 g/min V additions, respectively. This is because of the low Al content that predominantly favours α2-Ti3Al phase in the microstructures of TiAl-based alloys. Although the alloys had (α2+γ)/γ lamellae, ζ-Ti5Si3 that are visible around the grain boundaries, it is still embedded within the α2-phases.  The heat treatment carried out on the Ti-Al-Si-xV alloys was done on all the as-built alloys. The SEIs and EDS images is shown in Fig. 3, Fig. 4 and Fig. 5; while the compositions of the Ti-Al-Si-xV alloys are presented in Table 3. The phases identified in the heat-treated Ti-Al-Si-xV alloy samples are α2+γ/γ, β and ζ-Ti5Si3 phases. These phases are similar to the as-built alloys, however, the α2+γ/γ lamellae was more obvious with smaller quantities of βphase formed after heat treatment. It was perceived that the strategy adopted for heat treating these alloys resulted in the over refinement of the microstructure. Thereby, further decreasing the amount of α2+γ/γ at higher temperatures owing to a reduction in Al content. This was easily detected for all samples heat-treated at 1400 o C/60 mins/FC/950 o C/6 h/FC as compared to those ones at 1200 o C/60 mins/FC/950 o C/6 h/FC that formed more α2+γ colonies. Since V is not a strong β-stabilizer like Mo, the ζ-Ti5Si3 was able to dissolve most of the β formed in the α 2 -matrix.
Generally, the heat-treated samples have lesser Al content when compared to their asbuilt alloys noticed in Table 3. The composition of Si and V stays almost the same like their as-built alloys notwithstanding the heat treatment conditions but the Al content reduces in all the heat-treated alloy samples. This promotes α2 phase formation that serves as nucleation sites for both β and ζ phases. Therefore, dissolution of the β phase was accomplished owing to the presence of ζ-phase and V being a weak β-stabilizer. Besides, all the heat-treated samples only showed a slight variation in Si content. Hence, promoting the ζ-Ti5Si3 phase formation in the α2 phase which results in the decrease of β phase formed.    The microstructure of the as-built Ti-Al-Si-xV alloy with 0.33 g/min of V shows a clearer representation of the morphological features of the fabricated alloy. The darker region depicts an Al-rich area, whereas the composition of the bright sections represents areas of lower Al content. The white precipitates detected in the alloys is composed primarily of α2-Ti3Al and ζ-Ti5Si3. The equiaxed α2/γ lamellae colonies particles comprise largely α2+γ/α2 structures due to limited amount of Al content. Consequently, the higher feed rate of V leads to increased quantities of β and ζ-Ti5Si3 contents of both as-built and heat-treated samples.
For the heat-treated samples with more than 2 at.% of Si, the hexagonal phase ζ-Ti5Si3 primarily solidifies from the matrix which consists of the eutectic α2+ζ-Ti5Si3. The quantity of primary precipitated ζ-Ti5Si3 in the matrix increases with increasing Si content. This is caused by the present of the β phase that is precipitated from the grain boundary interface that acts as nucleation sites during the solidification of TiAl phases leading to an equiaxed microstructure. Annealing at a temperature of 950 o C results in homogenization of the alloys and an increase in microstructural stability. However, the heat treatment strategy adopted does not reduce the amount of precipitate phases formed in comparison to the as-built alloys.
Generally, laser AM processing gives rise to rapid solidification which can cause the segregation of various elements. However, the fast solidification rate avoids the coarsening of the TiAl-based unlike traditional methods [1,[5][6]. Equiaxed α2 is present in some areas and absent in those areas where growth of γ phase occurs. This implies that presence of α2 dominate the solidification stage, hindering the formation of γ. This is visible in SEIs (Fig.  3, Fig. 4 and Fig. 5). According to Ti-Al phase diagram, the solidification takes place in the eutectic zone. During the solidification the expected phases to be in equilibrium are α2+γ matrix with inhomogeneous distribution of β+ζ precipitates. AMed samples experiences at least 3 solid-liquid cycles along with heat treatments consecutively at temperatures above β transformation coarsening the phase.
However, these observations are valid where β+ζ concentration was almost absent. This confirms that V distribution rather than the amount, volume or size is important during grain refinement of β+ζ. The α2-phase are known to have limited coarsening even at high temperatures [5]. This also prevents the coarsening of lamellar grains. In summary, the microstructure of the heat-treated Ti-Al-Si-xV alloys showed different morphological features from the as-built samples with refined equiaxed grains as the V content increases in the deposits. The formation of hard β and ζ-Ti5Si3 was more predominant at the grain boundaries. Fig. 6 shows the distribution of the microhardness for the as-built Ti-Al-Si-xV alloys. It was inferred that V additions leads to reduction the micro-hardness values of the Ti-Al-Si-xV alloys. The highest average microhardness value is approximately 675 HV0.5 corresponding to a YS of 2207 MPa (using Equation 1 in ref [5]). This is much lower than the values achieved in our previous works for Ti-Al-0.43 g/min Si [5] and Ti-Al-0.03 g/min Si-xMo [4] alloys. Thus, it depicts that V additions would improve the fracture toughness and ductility of typical TiAl-based alloys. All the same, the micro-hardness values obtained for further V additions above 0.49 g/min (not presented in this work) resulted in considerably high microhardness values with desirable microstructures and composition for consideration as TiAlbased alloys. Therefore, adding V can demonstrate both positive and negative behaviours on mechanical properties of TiAl-based alloys. This may necessitate including other alloying elements of small quantities to stabilize the microstructure and balance this drawback.  Fig. 8 and Fig. 9 presents the distribution of the microhardness of the heat-treated Ti-Al-Si-xV alloys fabricated at 0.18 g/min, 0.33 g/min and 0.49 g/min V additions, respectively and the average microhardness values are presented in Table 4. Generally, the microhardness values of all the heat-treated alloys increases with temperature and the 1400 o C heat-treated samples showing the highest microhardness values regardless of the amount of V deposited; while the hardness values of the 1200 o C and 1300 o C for 0.18 g/min and 0.49 g/min V additions were relatively close. This was credited to the Al content (<40 at.%) which supports formation of more α2 enabling high precipitation of ζ-Ti5Si3 phases. It is understood that the ζ-Ti5Si3 phase is the hardest phase present in the microstructure of the Ti-Al-Si-xV alloys resulting on increased micro-hardness values.

Micro-hardness of graded Ti-Al-Si-xV Alloys
Generally, the average microhardness values of the heat-treated Ti-Al-Si-xV alloys were all above 630 HV0.5 irrespective of V additions which tallies to a YS of approximately 2060 MPa. This establishes that even though V addition decreases the microhardness, these values were still quite high owing to the lower Al content and presence of Si that promoted ζ-Ti5Si3 phase formation in the presence of α2. Thus, it is necessary to decrease Si while retaining the amount of V added as alloying element to diminish the detrimental effects on the mechanical properties.
The high hardness of the heat-treated Ti-Al-Si-xV alloys is attributable to the formation of β and ζ-Ti5Si3 phase and excess amount of alloying elements in the solid solution phase. Conversely, the hardness of the as-built alloys were lesser due to relatively high amount of Al in the alloy compositions. According to the Archard's law, increased hardness generally improves the wear resistant properties of a material [25][26]. The microhardness values were observed to increase with increase in V feed rate. This was expected because of the increase in β and ζ-Ti5Si3 phase precipitating randomly and dispersedly distributed causing the strengthening of the alloys. From the work of Mathabathe, Modiba and Bolokang [14], it is understood that based on Hall-Petch relationship that hardness values are inversely proportional to the microstructure grain size. Thus, in this work the variation in microhardness can be attributed to the influence and distribution of the strengthening phases on the microstructures. The addition of V could increase the wear resistance and reduce the friction   Fig. 9. Micro-hardness of heat-treated Ti-Al-Si-0.49 g/min V alloys It is generally known that ζ-Ti5Si3 has some outstanding advantages. These include a high melting point, a low density, high hardness, excellent wear resistance, high thermal stability and excellent high temperature oxidation resistance [3,10,14,22]. This means that ζ-Ti5Si3 is an excellent reinforced phase for wear-resistance and oxidation-resistance. It can also be seen that the hardness of these alloys increases with an increase of V content. With the highest values in samples heat-treated at 1400 o C/60 mins/FC/950 o C/6 h/FC due to the slow cooling (100 o C/hour) rate that allows for complete transformation of β and ζ-Ti5Si3 phases. However, micro-hardness values of the as-built alloys were lesser because of the rapid heating and cooling effects in the LENS process, thereby, leaving little time for full transformation of the precipitating phases of β and ζ-Ti5Si3.