1 Introduction

For the purpose of energy savings and emission reductions, developing high-performance materials in terms of thermo-mechanical properties is needed for air-transport industry [1]. TiAl-based alloys are considered to be really attractive for high-temperature applications owing to their low density, high specific modulus, and mechanical strength [2], acceptable creep behavior and good oxidation resistance [3]. From the perspective of weight reduction, TiAl-based alloys are superior to Ni-based superalloys [4]. However, one challenge is that TiAl-based alloys are subjected to the poor weldability and machinability, which is not beneficial to practical application [5].

Appropriate joining technologies for TiAl-based alloys are urgently required to expand their applications, including the joining of different TiAl components and the repair of TiAl castings. Brazing technique has been proved effective as a unique joining process for releasing the residual stress and avoiding cracks to a certain extent [6]. Recently, progress in brazing process such as friction stir vibration brazing (FSVB) provides a new insight to enhance the metallurgy reaction between the brazing seam and the base metal [7]. In general, a sound brazing joint could be achieved by design of the filler alloy with suitable composition [8], optimizing brazing parameter [9] as well as controlling brazing process [10].

Although Ag-based brazing filler alloy could braze TiAl alloy successfully, the joint suffers from insufficient bonding strength. For instance, within TiAl intermetallic joints brazed with BAg-8 [11], the formation and excessive growth of brittle Al-Cu-Ti reaction layer would deteriorate the joint properties, especially at high temperatures. Compared with Ag-based brazing alloy, Ti-based brazing alloy used to join TiAl-based alloy could represent high bonding strength. For example, the joint brazed by Ti-Zr-Cu-Ni-Co [12] filler showed a relatively high tensile strength at room temperature, which was up to 316 MPa. But brittle phases at the joint interface suppressed the improvement of joint strength.

For solving the joining challenges of the advanced TiAl material, it is of great importance to search for new joining filler alloys [13]. In our previous studies, brazing filler alloys of Ti-Zr-Fe [14, 15], Ti-Ni-Nb-Zr [16, 17], and Ti-Zr-Cu-Ni [18] were proposed for TiAl brazing. However, the concentration of Zr element within the brazing fillers should be strictly limited to 8 ~ 11 wt.% [19] for sufficient diffusion. Differently, Fe element was demonstrated not only possessing excellent diffusion rate but also accelerating self-diffusion rate in α-Ti [20]. More importantly, certain amounts of Fe [21] and Mn [22] were added in TiAl alloy for solid solution strengthening, based on the Ti-Mn-Fe ternary system [23]. To decrease the melting point of the filler alloy, another Ni-Zr binary alloy with a certain percentage was added based on Ni-Zr [24] binary eutectic compositions. With the design concept of multi-principal element alloys (MPEAs) [25, 26], in this paper, a 5-element medium-entropy brazing filler alloy with Ti-Mn-Fe-Ni-Zr composition system was thus designed for TiAl joining.

The joining of TiAl alloy was attempted by using the newly developed 5-element medium-entropy brazing filler alloy, and the aim of this study was to verify the feasibility of using the new system filler alloy to achieve high joint strength. The detailed compositions and phases across the brazed joint were analyzed. Additionally, interfacial microstructures and mechanical properties of the brazed joint were investigated. Room-temperature tensile test was performed to evaluate the joint performance, and the joining mechanism was discussed.

2 Experimental procedures

The nominal composition of TiAl intermetallic was Ti-46Al-(3 ~ 4)Nb-(2 ~ 3)(Cr, Ta, B) (at.%), which was fabricated by vacuum arc remelting and thermo-mechanical treatment. The Ti-(18 ~ 24)Mn-(12 ~ 18)Fe-(3 ~ 8)Ni-(3 ~ 8)Zr (wt.%) filler alloy ingot was fabricated by arc melting technique in high purity argon gas atmosphere. The cast ingot was broken down into pieces and subsequently ground into powder. Then, the particles were preset into the 90° V-shaped groove butt joint with a narrow assembly gap of 50 μm, as shown in Fig. 1a.

Fig. 1
figure 1

Schematic of assembly specimens (a) and geometric dimensions (b)

Full size image

The TiAl alloy was cut into specimens with the size of 9 mm × 11 mm × 18 mm. Prior to brazing experiment, the sample surfaces to be joined were polished by SiC grit paper and then ultrasonically cleaned for 15 min in ethyl alcohol solutions. The brazing experiments were carried out in a vacuum brazing furnace, with a high vacuum of 5 × 10−3 Pa to 9 × 10−3 Pa. The brazing temperature varied from 1140 to 1210 °C, which was higher than the liquidus temperature of filler alloy. On this basis, not only the weak brazing parameters of 1140 °C/20 min but also the strong brazing parameters such as 1180 °C/75 min or 1210 °C/45 min were chosen. Meanwhile, three different dwell times of 20 min, 45 min, and 75 min were selected. Based on the geometric dimensions for plates in Fig. 1b, joint specimens with a joining area of (1.0 ~ 1.5) mm × 2.5 mm were prepared for tensile test.

Differential scanning calorimetry (DSC) with a heat rate of 10 °C/min was performed to determine the thermal behavior of Ti-Mn-Fe-Ni-Zr system brazing alloy. The average contact angle on the TiAl alloy after heating at 1110 °C for 10 min in vacuum was calculated from the four measured value of the sample cross-section in vertical direction. Scanning electron microscope (SEM) equipped with an energy-dispersive X-ray spectrometer (EDS) attachment was used to observe joint microstructure. By means of phase extraction software, area percentage of γ-TiAl phases within the joint was measured. The average value was calculated by at least three different zones from the joint cross-section. Joint tensile strength was measured by a universal testing machine with a loading speed of 0.5 mm/min at room temperature, and the reported tensile strength was the average value of at least three measurements for the same brazing condition.

3 Results and discussion

3.1 Characteristics of Ti-Mn-Fe-Ni-Zr medium-entropy filler alloy

Figure 2a shows that the cast filler ingot and Fig. 2b shows the backscattered electron image of the 5-element medium-entropy filler alloy, and three main phases could be observed. Based on the EDS analysis results in Table 1, the filler was mainly composed of Ti-(Fe, Mn) compound dissolved with 2 ~ 4 at.% Ni and Zr (microzone “1”), and Ti-(Fe, Mn) compound and Ti-rich phases dissolved with 2 ~ 8 at.% Ni and Zr (microzone “2” and “3”), as well as Ti-rich phases dissolved with Mn, Fe, Ni, and Zr (microzone “4” and “5”) (Table 2).

Fig. 2
figure 2

Cast ingot (a), backscattered electron image (a) and DSC results (b) of Ti-Mn-Fe-Ni-Zr medium-entropy filler alloy

Full size image
Table 1 EDS analysis results for microzones marked in Fig. 2
Full size table
Table 2 Calculation results of empirical parameters for the Ti-Mn-Fe-Ni-Zr braze alloy
Full size table

The Ti-Mn-Fe alloy system exhibits good compatibility within the three elements and there is a eutectic composition of Ti-(19 ~ 25)Mn-(14 ~ 20)Fe (wt.%), with the liquidus temperature of about 1139°C [27] in this ternary alloy system. In the meantime, Ni-Zr binary eutectic composition of Ni-46.9Zr (wt.%) was also added as an effective melting point depressant with the liquidus temperature of 1061°C [24]. On this basis, it can be deduced that there might exist a 5-element eutectic composition of Ti-(18 ~ 24)Mn-(12 ~ 18)Fe-(3 ~ 8)Ni-(3 ~ 8)Zr (wt.%) with lower melting point.

According to DSC analysis results in Fig. 2c, the melting range of the filler alloy was 1005.3 ~ 1060.1 °C. Due to the fact that the liquidus temperature is lower than some reported Ti-based brazing filler alloy, such as Ti-Ni-Nb [28, 29] eutectic braze alloy and Ti-Fe-Mn [27] eutectic braze alloy, the brazing experiment could be performed at a lower brazing temperature, which might be beneficial to control the interface reaction [30].

Figure 3 presents the wettability experiment results of the novel filler on TiAl alloy at 1110 °C for 10 min. As shown in Fig. 3, the filler melted and reacted with the base metal. The contact angle on the TiAl alloy was measured as 30°. The maximum reaction layer thickness of the Ti-Mn-Fe-Ni-Zr braze alloy in the TiAl alloy was about 95 ~ 105 μm in Fig. 3b-c, which was close to that of the Fe-Ni-Co-Cr-Si-B [31] filler alloy with 110 μm at 1180 °C for 10 min. This illustrated that a stronger reaction and sufficient spreading behavior of filler still occurred with the TiAl substrate.

Fig. 3
figure 3

Wettability of Ti-Mn-Fe-Ni-Zr filler alloy on TiAl at 1110 °C for 10 min: wetting morphology (a), contact angle from the cross-section of A-A (b) and B-B (c)

Full size image

From the thermodynamic point of view, the empirical parameters, including four main parameters, were calculated for the Ti-Mn-Fe-Ni-Zr braze alloy to predict the phase formation in high-entropy alloys (HEAs) [32,33,34], as shown in Table 2. Among the four parameters, the mixing entropy (\({\Delta S}_{{\text{mix}}}\)) of 9.42 J·K−1·mol−1, the mixing enthalpy (\({\Delta H}_{{\text{mix}}}\)) of − 17.55 kJ·mol−1, the parameter (\(\Omega\)) of 0.98, and the atomic size difference (\(\delta\)) of 6.72% were presented. Then, due to the \({\Delta S}_{{\text{mix}}}\) being between 1.0 \(R\) and 1.5 \(R\) (8.314 ~ 12.471 J·K−1·mol−1), the Ti-Mn-Fe-Ni-Zr filler metal should be classified as medium-entropy alloy [35].

3.2 Microstructural analysis of the brazed joints

Several choices of brazing parameters were made to join TiAl alloy at 1140 °C/20 min, 1180 °C/45 min, 1180 °C/75 min, and 1210 °C/45 min, respectively. The backscattered electron images of the TiAl joints brazed by the four different brazing parameters are shown in Fig. 4. The EDS analysis results for the typical microzones marked in Fig. 4 are displayed in Table 3.

Fig. 4
figure 4

Backscattered electron images of the TiAl joints brazed at 1140 °C/20 min (a), 1180 °C/45 min (b), 1180 °C/75 min (c), and 1210 °C/45 min (d), respectively

Full size image
Table 3 EDS analysis results for the microzones marked in Fig. 4
Full size table

As the typical interfacial microstructure of TiAl joint at 1140 °C/20 min, the thickness of the brazed joint reached 171 μm in Fig. 4a. The white phase (microzone “3”) in the center of the brazing seam exhibited as almost continuous layer, with a thickness of 29 μm. Obviously, the quantities of gray phase (microzone “2”) were far more than that of the white phase in the brazing seam. Since the composition of white phase in the brazed joint was close to that of the braze alloy, it could be identified as the residual brazing filler reaction phase. Based on the Ti-Al binary alloy phase diagram [36,37,38], the gray phase in the brazing seam showing a similar Ti/Al ratio to that of the parent metal might be regarded as the γ-TiAl + α2-Ti3Al phase [39, 40].

From the EDS analysis results shown in Table 3, except for the residual brazing filler reaction phase, generally Ti element concentration within the brazing seam was comparable to that of the parent material. Al element significantly diffused from the TiAl base metal to the brazing seam, with the concentration high up to 32.40 ~ 36.30 at.%. For microzones “3”, “6,” and “9” in the central part of the joint, their compositions were characterized by 29.65 ~ 31.58 at.% Ti, 34.30 ~ 40.93 at.% Al, 7.13 ~ 10.40 at.% Fe, 6.55 ~ 9.55 at.% Mn, 4.58 ~ 5.65 at.% Ni, and 5.07 ~ 6.02 at.% Zr, and thus they should be regarded as residual brazing filler reaction phase, and this was agreement with the area distribution map of elements in Fig. 5. Therefore, for eliminating the residual brazing filler reaction phase within joint, higher brazing temperature or longer dwell time is needed.

Fig. 5
figure 5

Backscattered electron images of magnified morphology for the joint brazed at 1140 °C/20 min (a), and area distribution map of elements Ti (b), Al (c), Fe (d), Mn (e), Ni (f), Zr (g)

Full size image

In comparison, the joint thickness was significantly increased to 258 μm and the thickness of the residual brazing filler reaction phase (microzone “6” in Fig. 4b) was sharply decreased to 14 μm by the brazing parameter of 1180 °C/45 min, as shown in Fig. 4b. But the residual brazing filler reaction phase still remained continuous. Interestingly, in this case, the dark lath-like phase (microzone “5”) appeared in the brazing seam. According to the Ti-Al binary alloy phase diagram [36,37,38], it is reasonable to deduce the dark phase to be γ-TiAl phase [39, 40].

The Al element concentration in the brazing seam was increased to 33.45 ~ 44.12 at.%, indicating the enhanced diffusion of Al element by the brazing parameter of 1180 °C/45 min. However, the expected sufficient diffusion of elements Fe, Mn, Ni, and Zr had not been accomplished due to the presence of the residual brazing filler reaction phase. For eliminating the residual brazing filler reaction phase, it is necessary to further prolong the dwell time or increase the brazing temperature.

On the one hand, with prolonging the dwell time to 75 min at the brazing temperature of 1180°C, the joint thickness was slightly increased to 278 μm and the thickness of the residual brazing filler reaction phase (microzone “9”) was decreased to only 6 μm, as shown in Fig. 4c. More importantly, the residual brazing filler reaction phase became discontinuous. It appeared that the lath-like γ-TiAl partially began to aggregate in the joint. Compared with the joint brazed at 1180 °C/45 min, the concentration of Al element was between 33.60 ~ 45.20 at.%. Although amount of the residual brazing filler reaction phase disappeared by prolonging the dwell time, it was still necessary to further increase the brazing temperature for completely eliminating the residual brazing filler reaction phase.

On the other hand, with increasing the brazing temperature to 1210 °C, the joint thickness was evidently increased to 308 μm and the size of the lath-like γ-TiAl phase gradually became larger (Fig. 4d). Significantly, this brazing temperature increase almost eliminated the residual brazing filler reaction phase, which might bring potential benefits to the joint strength.

Under the brazing condition of 1210 °C/45 min, the content of the main elements Ti and Al in the brazing seam (microzones “12” and “13” in Fig. 4d) was close to that of the base metal (microzones “7” and “8” in Fig. 4c). As shown in Fig. 6, the elements Ti, Fe, Mn, Ni, and Zr exhibited homogeneous distribution. It seemed that sufficient diffusion within the joint occurred and the overall composition of the joint was characterized by 47.98 ~ 52.45 at.% Ti, 34.00 ~ 45.34 at.% Al, 0.86 ~ 3.24 at.% Fe, 1.45 ~ 3.91 at.% Mn, 0.27 ~ 0.74 at.% Ni, and 0.36 ~ 0.40 at.% Zr. In other words, a sound brazing joint can be obtained with the composition close to that of the base metal.

Fig. 6
figure 6

Backscattered electron images of magnified morphology for the joint brazed at 1210 °C/45 min (a), and area distribution map of elements Ti (b), Al (c), Fe (d), Mn (e), Ni (f), Zr (g)

Full size image

The effect of brazing parameter on the area percentage of γ-TiAl phase within the brazed joint, the thickness of residual brazing filler reaction phase, and the joint thickness is shown together in Fig. 7. On the whole, when the brazing parameter becomes stronger, the joint thickness and the area percentage of γ-TiAl phase wthin the joint increased, and the thickness of residual brazing filler reaction phase decreased. Individually, the new formation of γ-TiAl phase was observed at 1180 °C/45 min and the residual brazing filler reaction phase within the joint almost disappeared at 1210 °C/45 min. Figure 7 also signified that the brazing condition of 1210 °C/45 min caused the maximum joint thickness of 308 μm and the maximum area percentage of γ-TiAl phase of 33.77%, as well as the almost elimination of the residual brazing filler reaction phase.

Fig. 7
figure 7

Effect of brazing parameter on the area percentage of γ-TiAl, the thickness of residual brazing filler reaction phase, and the joint thickness

Full size image

From the perspective of diffusion mechanism, the diffusivity of alloying elements Fe, Mn, and Ni followed the interstitial diffusion mechanism in α-Ti, which belonged to the faster diffusion elements. On the contrary, Zr element exhibited slow diffusion rate due to its vacancy mechanism [41]. For the brazing parameter of 1140 °C/20 min or 1180 °C/45 min, the residual brazing filler reaction phase remained in the central part of the joint due to its too weak high-temperature diffusion. On the brazing condition of 1210 °C/45 min, it was believed that the sufficient diffusion within the joint occurred. Moreover, the Zr element concentration of 3 ~ 8 wt.% in the brazing alloy was quite low. As a result, the joint composition of 0.86 ~ 3.24 at.% Fe, 1.45 ~ 3.91 at.% Mn, 0.27 ~ 0.74 at.% Ni, and 0.36 ~ 0.40 at.% Zr exhibited homogeneous distribution and residual brazing filler reaction phase was almost eliminated after the brazing, and this undoubtedly improved the joint strength.

3.3 Mechanical properties of the brazed joints

The effect of brazing parameter on the tensile strength of the brazed joint at room temperature is shown in Fig. 8. With enhancing the brazing parameter, the tensile strength first dramatically increased to 367 MPa at 1180 °C/45 min from 194 MPa at 1140 °C/20 min. Obviously, the diffusion was not sufficient at 1140 °C/20 min, resulting in the thick residual brazing filler reaction phase and the weak joint. The improvement of the tensile strength might be attributed to the strong diffusion behavior of elements and the partial dissolving of the residual brazing filler reaction phase, as well as the formation of γ-TiAl within the brazed joint. However, the decrease of the tensile strength to 228 MPa under the brazing condition of 1180 °C/75 min was also noticeable, and this might be associated with the microstructure of γ-TiAl phases within the brazing seam and the interface between the brazing seam and the base metal, as shown in Fig. 4c. Finally, the maximum tensile strength of 418 MPa was achieved with the joining parameter of 1210 °C/45 min, which was caused by the almost complete elimination of the residual brazing filler reaction phase and the homogeneous distribution of the main element in the brazing alloy throughout the whole joint, as shown in Fig. 6. Compared with the tensile strength 590 MPa of the parent material, the brazing condition of 1210 °C/45 min offered the highest strength coefficient of 70.85%.

Fig. 8
figure 8

Effect of brazing parameter on the tensile strength and the joint strength coefficient at room temperature

Full size image

Due to the sufficient diffusion between the brazing seam and the base metal, the Ni concentration within the joint was decreased to the low value of 0.27 ~ 0.74 at.%, and this was favorable to suppress the formation of brittle Ti-Ni intermetallic compounds. Moreover, the dissolved elements of 0.86 ~ 3.24 at.% Fe, 1.45 ~ 3.91 at.% Mn, and 0.36 ~ 0.40 at.% Zr should play an important role of solid solution strengthening, and thus had a beneficial effect on the joint strength [14, 21, 22]. However, it seems that the composition of the brazing filler alloy still needs to be optimized in future for further improvement of the joint strength [16, 19, 28, 42].

The fracture location and the fracture path are shown in Fig. 9. The cracks initiated and propagated at the center of the joint brazed at 1140 °C/20 min and 1180 °C/45 min, indicating that the continuous residual brazing filler reaction phase in the center of the joint caused the formation of cracks. In other words, the residual brazing filler reaction phase was the main factor to the lower strength of the brazed joint. On the contrary, for the joint brazed at 1180 °C/75 min and 1210 °C/45 min, the cracks initiated and propagated at the interface between the brazing seam and the base metal. Due to the fact that the discontinuous residual brazing filler reaction phase in the joint brazed at 1180 °C/75 min was thin and remained in small amount (shown in Fig. 9c), the tensile test specimen fractured at the interface. In this case, the detrimental effect of the residual brazing filler reaction phase on joint strength might be weakened.

Fig. 9
figure 9

Fracture paths of the joints brazed at 1140 °C/20 min (a), 1180 °C/45 min (b), 1180 °C/75 min (c), 1210 °C/45 min (d), and the magnified morphology of the red dash rectangle zone (e)

Full size image

Different from the joint brazed at 1180 °C/75 min, the fracture interface obtained by the brazing parameter of 1210 °C/45 min displayed far less smoothness and contained a small quantity of base metal, as shown in Fig. 9d. The roughened interface might be attributed to the stronger dissolution of TiAl and the more sufficient interdiffusion between the TiAl base metal and the brazing seam. Indeed, the adhered TiAl base metal at the fracture interface inferred that a strong metallurgical bonding between the brazing seam and the base metal had been formed and the strength of the brazing seam was comparable to that of base metal. The mixed fracture path passing through the base alloy and the brazing seam is shown in Fig. 9e. The fracture behavior exhibited a mixed-mode of intergranular and transgranular fracture. This fracture mode should be favorable to the improvement of the joint strength.

4 Conclusions

The main conclusions of the present study can be summarized as follows:

  1. 1

    The Ti-(18 ~ 24)Mn-(12 ~ 18)Fe-(3 ~ 8)Ni-(3 ~ 8)Zr (wt.%) medium-entropy filler alloy was proposed for TiAl joining, and the joints brazed at 1210 °C/45 min exhibited the maximum room-temperature tensile strength of 418 MPa, the highest joint strength coefficient of 70.85%.

  2. 2

    The brazed joint was composed of γ-TiAl, α2-Ti3Al, and residual brazing filler reaction phase. Under the brazing condition of 1210 °C/45 min, a sound joint was achieved with a desirable composition characterized by 0.86 ~ 3.24 at.% Fe, 1.45 ~ 3.91 at.% Mn, 0.27 ~ 0.74 at.% Ni, and 0.36 ~ 0.40 at.% Zr.

  3. 3

    The joint brazed at 1210 °C/45 min offered the maximum joint thickness of 308 μm and the maximum area percentage of γ-TiAl phase of 33.77%, as well as the almost complete elimination of residual brazing filler reaction phase.

  4. 4

    Under the brazing condition of 1140 °C/20 min and 1180 °C/45 min, the joint fractured along the residual brazing filler reaction phase in the center of the brazed joint. Differently, the joint brazed at 1180 °C/75 min and 1210 °C/45 min fractured at the interface between the brazing seam and the base metal, and the joint fracture showed a mixed-mode of intergranular and transgranular fracture.