A. Preparation for titanium aluminide Al₃Ti alloy

Titanium aluminide Al₃Ti alloy was prepared via a diffusion reaction between Al and Ti alloy foils at a temperature of 685 °C in a vacuum hot pressing furnace. The foils used in this study were commercial pure Al 1100 and commercial Ti–6Al–4V alloy (TC4), with the thickness of 1.8 and 0.5 mm, respectively, they were cut into square of 100 × 100 mm. The oxides and contaminations on both surfaces of pure Al and TC4 alloy foils were removed using 800Cw grit silicon carbide paper, and then cleaned in alcohol bath using ultrasonic cleaning machine. After drying, the foils were stacked in alternating layers of titanium alloy and aluminum with titanium alloy covers on the top and bottom. In the preparation process, the vacuum hot press furnace with stacked foils was gradually heated to the maximum temperature ∼685 °C and held for 7 h at this temperature. The pressure applied on the plate during fabrication was 2–3 MPa when the temperature was less than 500 °C, while it was 1–1.5 MPa when heating temperature is in a range of 500–685 °C. The pressure was free during the initial 5 h of the heat preservation. The temperatures and pressures used at different processing stages were shown in Fig. 1. The optimum parameters, for instance, thickness ratio between TC4 alloy and pure Al, heating rate and pressure were confirmed through many trials. Such that the reaction between the TC4 alloy and pure Al foils was exactly done, i.e., TC4 alloy and pure Al foils were completely converted to titanium aluminide Al₃Ti except for the residual TC4 alloy layers as thin top and bottom covers. At the last period of synthesis, when the TC4 alloy layers were completely consumed with the Al layers, the applied pressure was increased to 2–3 MPa again to ensure that the residual Al was squeezed out by a higher pressure. The dense bulk Al₃Ti alloy was eventually prepared using these optimized processing parameters, including right thickness ratio of the TC4 alloy and pure Al foils, and the variable pressures and temperatures used in different processing periods. It should be pointed out that the right thickness of the TC4 alloy and pure Al foils must be selected to ensure that the TC4 alloy and pure Al foils were exactly consumed without any residual TC4 alloy or pure Al left in the bulk Al₃Ti alloy after sintering. The last procedure to fabricate pure titanium aluminide Al₃Ti alloy is to remove the residual thin TC4 alloy layers on the top and bottom surface since the diffusion reaction only occurred on one side of covered TC4 alloy layers, while there was no reaction in the other side of covered TC4 alloy foil that is against the high temperature stainless steel plate.

FIG. 1.

The elevated temperature curve.

B. Measurement method of physical properties

The density of the fabricated titanium aluminide Al₃Ti alloy ρ was measured using the buoyancy method, the equation to calculate the density has the following form ^{Reference Goldstein27} :

(1) $$\rho = {{{g_{\rm{a}}}} \over {{g_{\rm{a}}} - {g_{\rm{w}}}}}{\rho _{\rm{w}}}$$

where the g _a, and g _w are the masses of the sample measured in air and water, respectively, ρ_w is the density of water. Here, a rectangular sample of Al₃Ti with 20 × 20 × 8 mm (height) was used to weight the mass using a balance (JA2003) with a minimum scale of 0.001 g.

The Possion's ratio and Young's modulus of the Al₃Ti alloy were measured according to ASTM E494-10 standard. Firstly, an ultrasonic pulse-receiver (Olympus 5072PR, Olympus Co. Ltd., Beijing, China) was used to measure the propagation velocities of longitudinal and transverse waves in the sample, respectively. The fundamental material properties of Al₃Ti alloy, including Young's modulus, shear modulus, and Poisson's ratio were determined using measured sound velocities and material density ²⁸ as:

(2) $${\rm{Velocity}}\left( V \right) = {{{\rm{thickness}}} \over {{{{\rm{round\ trip\ transit\ time}}} \mathord{\left/ {\vphantom {{{\rm{round\ trip\ transit\ time}}} {\rm{2}}}} \right. \kern-\nulldelimiterspace} {\rm{2}}}}}$$

(3) $${\rm{Poisson,s\ ratio}}\left( \nu \right) = {{1 - 2{{\left( {{{{V_{\rm{T}}}} \mathord{\left/ {\vphantom {{{V_{\rm{T}}}} {{V_{\rm{L}}}}}} \right. \kern-\nulldelimiterspace} {{V_{\rm{L}}}}}} \right)}^2}} \over {2 - 2{{\left( {{{{V_{\rm{T}}}} \mathord{\left/ {\vphantom {{{V_{\rm{T}}}} {{V_{\rm{L}}}}}} \right. \kern-\nulldelimiterspace} {{V_{\rm{L}}}}}} \right)}^2}}}$$

(4) $${\rm{Young,s\ modulus}}\left( E \right) = {{\mathop {{V_{\rm{L}}}}\nolimits^2 \rho \left( {1 + \nu } \right)\left( {1 - 2\nu } \right)} \over {1 - \nu }}$$

(5) $${\rm{Shear\ modulus}}\left( G \right) = \mathop {{V_{\rm{T}}}}\nolimits^2 \rho$$

where, V _T, V _L are the transverse and longitudinal velocities, respectively.

C. Microstructure characterization

The microstructure and phase type of the Al₃Ti alloy were investigated using the SEM, energy dispersive x-ray spectrometer (EDS), and x-ray diffractometer (XRD), respectively. The specimens used for SEM and EDS analysis were well-polished and etched for 15 s in the mixed solution of HF acid (10 mL), HNO₃ acid (30 mL), and water (70 mL), while the specimens prepared for XRD scan were just ground by 320Cw and 1000Cw silicon carbide papers, respectively. The x-ray diffractometer with a target of Cu, the working voltage of 40 kV, working current of 40 mA, the scanning range (2θ) of 20–90°, and the step size of 0.033°/step were used in the present XRD test.

D. Thermal analysis

The thermal analysis of the Al₃Ti alloy was performed by DSC/TG method. The powder samples used for thermal analysis were prepared using the fine fragments collected from the tested compressive specimens. The thermal analysis tests were carried out on NETZSCH STA 449F3 (NETZSCH Scientific Instruments Trading Ltd., Shanghai, China) in argon atmosphere. Testing temperature range was 35–1470 °C, and the heating rate was 20.0 K/min.

E. Mechanical testing

The compressive specimens were rectangular with the dimension of 6 mm (length) × 6 mm (width) × 8 mm (height), which were cut from the same Al₃Ti alloy bulk plate using electrical discharge machine. The compressive tests were performed using an Instron 5500R load frame (Instron Co. Ltd., Beijing, China) at a strain rate of 0.001/s. The load and displacement histories that were measured by the load cell and deflectometer, respectively, were used to obtain the stress–strain curves, strain to failure and compressive strengths of Al₃Ti alloy. The fracture surface of the tested specimen was observed by SEM to investigate the failure mechanisms of the synthesized titanium aluminide Al₃Ti alloy.

A. Physical properties

The physical properties of the synthesized Al₃Ti alloy, including density, Poisson' ratio, and Young's modulus are shown in Table I. As can be seen from Table I, the average density is ∼3.40 g/cm³, which is slightly higher than the theoretical value ∼3.36 g/cm³. The reason is that the titanium alloy foil used to synthesize the Al₃Ti alloy is TC4 alloy rather than pure titanium, i.e., the chemical element vanadium from original TC4 alloy foils is therefore introduced into the final synthesized Al₃Ti alloy. The later analysis results by XRD in this work demonstrate that the element vanadium exists in Al₃Ti alloy as the Al₃Ti_0.8V_0.2 compound (see Figs. 2 and 3 for details). To further confirm this point, Al₃Ti alloy was synthesized again using same sintering process by the foils of commercial pure titanium and pure aluminum, and the density of Al₃Ti alloy synthesized using pure titanium is determined to be ∼3.07 g/cm³, which is lower than the theoretical value of Al₃Ti alloy, ∼3.36 g/cm³. Therefore, the higher density of Al₃Ti alloy can be considered as the contribution of Al₃Ti_0.8V_0.2 compound due to the existence of element vanadium. In addition, as can be seen in Table I, the average Young's modulus is ∼209.07 GPa, also lower than the theoretical value of 216 GPa. This difference may result from the unreacted residual aluminum, since the later analysis results in this work reveal that there is a trace of aluminum phase in Al₃Ti alloy detected by DSC although the SEM observation and XRD can not find it due to a very small mount of residual aluminum. It should be recognized that the residual aluminum (even a small amount) may play an important role on the density, Young's modulus and mechanical properties of the Al₃Ti alloy. The difference between the current experimental results and theoretical values is related to chemical element vanadium, a trace of the residual aluminum and the micro-pores. Further investigation is therefore being performed to study the variation mechanisms of the physical and mechanical properties of the Al₃Ti alloys synthesized by different titanium (commercial pure titanium and TC4 alloy) foils, of which results will be reported in a separated paper.

TABLE I.

Physical properties of Al₃Ti alloy.

FIG. 2.

The EDS results scanned from the area as marked in Fig. 4(a).

FIG. 3.

XRD pattern of titanium aluminide Al₃Ti alloy.

B. Microstructure

The SEM images of the Al₃Ti alloy with different magnifications are shown in Fig. 4. It is noted that the grain size of the Al₃Ti is nonuniform and varies in a range of 3–20 μm, there is no other phase visible in boundary and inside of the grain, Al₃Ti is the only phase as shown in Fig. 4(b). This experimental result is confirmed by the later results from EDS and XRD, as shown in Figs. 2 and 3, respectively. It is noted from Fig. 2 that the elements of the reacted products are titanium, aluminum, and vanadium, and the atomic percentages of Al and Ti are ∼73.65 and ∼25.41%, respectively. Thus there is no doubt that the newly synthesized alloy mainly consists of the intermetallic Al₃Ti phase, which is identical with the result obtained later in this work.

FIG. 4.

SEM micrographs of titanium aluminide Al₃Ti alloy, (a) low magnification and (b) high magnification, here the sample was etched using the solution of HF acid (10 mL), HNO₃ acid (30 mL) and water (70 mL).

Figure 3 shows the XRD patterns of the Al₃Ti alloy. It is obvious that the most of the newly synthesized products are DO₂₂ Al₃Ti phase (PDF card no. 03-065-4202), a trace of Al₃Ti_0.8V_0.2 (PDF card no. 00-043-1154) is also identified by XRD analysis, Al₃Ti_0.8V_0.2 has a similar phase structure as the Al₃Ti intermetallic compound (see Fig. 3 for details). The experimental fact that the compound of Al₃Ti_0.8V_0.2 is identified by XRD is in coincide with the EDS result as shown in Fig. 2, in which the chemical element vanadium is detected by EDS technique. Here the chemical element vanadium is believed to come from original TC4 alloy foils, as aforementioned. Both the SEM observation and XRD analysis indicated that the starting titanium alloy and pure aluminum foils are completely reacted and transformed to monolithic intermetallic Al₃Ti alloy and a trace of Al₃Ti_0.8V_0.2, no any remaining unreacted titanium or aluminum is detected by the conventional SEM and XRD characterization techniques. As a matter of fact, a trace of residual aluminum is also found to exist in the final products, which just cannot be detected by conventional SEM, XRD experimental techniques due to a relatively small amount.

It should be pointed out that there should have some other intermetallic compounds formed during the reaction between the titanium and aluminum, such as TiAl₂, Ti₂Al₅, Ti₃Al, and TiAl apart from the Al₃Ti phase, in terms of the Ti–Al binary phase diagram. Why both the SEM and XRD analysis techniques show only Al₃Ti phase available in a newly reacted intermetallic compound? It has been demonstrated that the free energy of Al₃Ti formation is lower than that of Ti₃Al and TiAl compounds in the temperature range of 0–1200 °C, ^{Reference Kattner, Lin and Chang29} which is the reason for Al₃Ti alloy firstly generated at the current reacting temperature of 685 °C. On the other hand, the formation of TiAl₂ and Ti₂Al₅ need a series of liquid–solid and/or solid–solid state reactions under high temperature based on the TiAl phase formation that acts on one of the starting phases. ^{Reference Murray30} Therefore without the formation of the TiAl phase, it is hard to transform into other two compounds from the initial TiAl phase. That is to say, the formation of TiAl₂ and Ti₂Al₅ must be on the base of TiAl phase, while according to the thermodynamics and reaction kinetics conditions for forming the Al₃Ti and TiAl phases, the later one can hardly form in the current low temperature condition. Accordingly, there are no TiAl₂ and Ti₂Al₅ formation and the Al₃Ti phase is the only newly formed product at 685 °C.

It is no doubt that the formation of the Al₃Ti is a result of diffusion of the titanium and aluminum atoms. When the heating temperature is higher than the melting point of Al, there are liquid phases and small intermetallic nodules formed between TC4 and Al foils. ^{Reference Harach and Vecchio13,Reference Mackowiak and Shreir31–Reference Meyers, Yu and Vecchio35} The existence of the liquid phase is beneficial for the reaction process, which enables the reaction to be faster and the newly formed intermetallic phase to be dense. Once the complete Al₃Ti layer forms, two interfaces of Al–Al₃Ti and Al₃Ti–Ti also form, and the Al and Al₃Ti are in the two phase (L + Al₃Ti) region of the Ti–Al phase diagram. Therefore the titanium atoms need to diffuse across the previously formed Al₃Ti layer (solid phase) for forming new Al₃Ti phase with aluminum atoms coming from liquid phase. With the migration of the solid Al₃Ti/liquid Al interface toward to the liquid aluminum, the solid Al₃Ti phase grows up and the liquid aluminum phase gradually decreases, when the solid Al₃Ti phase meets another solid Al₃Ti layer moving from another side, the reaction ends at the mid-thickness of the each initial Al foil, and the entire dense solid Al₃Ti plate is synthesized. Clearly, the uniformity of the chemical composition, grain size, and microstructure is strongly dependent upon the reaction time and the initial foil thickness.

C. Thermal analysis (DSC/TG) results

The results of thermal analysis (DSC/TG) are shown in Fig. 5. It is noted that the mass of the sample is almost constant (∼−0.1%) during the whole heating in a temperature range up to 1470 °C, indicating the intermetallic Al₃Ti alloy has a superior thermal stability. As can be seen from Fig. 5, there are two endothermic peaks visible in the DSC curve during the heating process. The lowest peak associated with the temperature of 661.1 °C corresponds to the melting point of aluminum (660.4 °C). This lowest temperature peak detected by DSC reflects the existence of the residual aluminum, although the SEM and XRD experimental techniques do not find the residual aluminum due to the small amount, as mentioned previously in this work. In our another study, it is found that the amount of the residual aluminum increases with decreasing applied pressure that is used to squeeze the superfluous aluminum in last processing period. Our further study also reveals the effect of the different content of residual aluminum on the mechanical properties of Al₃Ti alloy. While in the current work, the amount of the residual aluminum is too less to be detected by the XRD analysis technique and found by SEM observation. For second peak corresponding to temperature of 1078.1 °C, it is probably related to the compound phase of Al₃Ti_0.8V_0.2 that is identified by XRD, as shown in Fig. 3. However, this conclusion needs further study for figuring out the physical nature of this peak. The temperature corresponding to the intersection point of the two tangent lines is 1365.8 °C, which exactly represents Al₃Ti melting point of 1360 °C.

FIG. 5.

The DSC/TG curves of titanium aluminide Al₃Ti alloy.

D. Mechanical properties

To investigate the mechanical behavior of the intermetallic Al₃Ti alloy, the quasistatic compressive tests were performed under a strain rate ∼0.001/s. Figure 6 depicts six typical compressive stress–strain curves. Clearly, all stress–strain curves show linear elastic behavior without any plastic deformation evidence, indicating a brittle fracture occurred in the Al₃Ti alloy. Such a stress–strain behavior reflects the inherent nature of the brittle intermetallic Al₃Ti alloy. On the other hand, relatively large differences are found in the maximum compressive strengths and failure strains of the six samples. For example, the maximum (sample #5) and minimum (sample #3) compressive strengths are 901.96 MPa and 594.76 MPa, respectively. Similar large scatters are also found in the failure strain for six compressive samples, which also reflects the inherent brittle nature of the Al₃Ti alloy.

FIG. 6.

Typical compressive stress–strain curves of Al₃Ti alloy tested under a strain rate of 0.001/s.

A brief summary of the compressive test results is presented in Table II, the average compressive strength is 765.20 MPa and the average failure strain is 0.74%. Both the average compressive strength and failure strain of the Al₃Ti alloy synthesized using metal foil metallurgical techniques in this work are higher than the previously reported elsewhere ^{Reference Vecchio12} in which the Al₃Ti alloy was synthesized by two-step processing (ultrasonic consolidation plus sintering technique). In their work ^{Reference Vecchio12} commercial pure titanium and aluminum foils were used to prepare Al₃Ti alloy. While in the current work, TC4 alloy and pure aluminum foils are used to fabricate Al₃Ti alloy by sintering technique in a vacuum pressure. The comparison with the mechanical properties of the Al₃Ti alloys prepared by other processing techniques indicates that the present technique proposed in this work is more competitive.

TABLE II.

The compressive strengths and failure strains of Al₃Ti alloy tested under a strain rate of ∼0.001/s.

The typical compressive failure characteristics of the Al₃Ti alloy were observed by SEM, as shown in Fig. 7. It can be clearly seen that the intermetallic Al₃Ti alloy shows a brittle failure features that are similar with that observed in the Al₃Ti matrix of the Ti/Al₃Ti laminate composites under three-point bend loading conditions. ^{Reference Adharapurapu, Vecchio, Jiang and Rohatgi26} As marked in Fig. 7, a mixed failure mode of the transgranular and intergranular fracture features is visible, which is a typical failure characteristic of the brittle Al₃Ti alloy no matter what the stress state applied on it.

FIG. 7.

The SEM image of fracture surface of Al₃Ti alloy, showing the mixture failure mode of the transgranular and intergranular fracture.