A Novel Process for Joining Ti Alloy and Al Alloy using Two-Stage Sintering Powder Metallurgy

Introduction

In the aerospace industry, lightweight structural material has become an excellent solution to reduce fuel consumption and improve load capacity and voyage. Titanium alloy and aluminum alloy show high specific strength, high specific stiffness and other excellent characteristics [1, 2], thus they are considered to be the key structural materials for advanced aircraft components. With the development of technology, the demand for various functions increased. Since it is difficult to satisfy the demand of multi-functions and lightweight in a single material, the joining of Ti and Al has been developed and applied in various fields.

A disadvantage of traditional joining techniques, such as fusion welding [3, 4] and friction welding [5, 6, 7], is the development of solidification cracking and high distortion stresses, these techniques significantly decrease the joints strength. Diffusion bonding (DB) does not form a molten zone at the joint interface and can thus avoid the above problems [8, 9, 10]. But high temperature and long bonding time are required for full diffusion, resulting in increased costs and long process cycles [11], also the grain coarsen and brittle intermetallic compounds (IMCs) that occur at the interface will deteriorate the properties of matrix and joint. These problems can be avoided by two-stage sintering powder metallurgy (P/M) process, which includes the pre-sintering process and powder metallurgy bonding. To differentiate with DB, P/M bonding is only a solid state that is applied to bond P/M bodies (pressed or pre-sintered or sintered). The new method is defined as the components being bonded with high temperature, and one in which of the components full densification and the greater strength of matrix is obtained. Due to the powder shows higher atoms activity and larger amount of diffusion channels, complete metallurgical bond with high strength of matrixes and joint can be obtained. To date, the DB of Ti/Al joint has been widely investigated [12, 13, 14], but a few work regarding the joining of Ti alloy and Al alloy by P/M bonding has been reported.

In the present work, the Ti-6Al-4V (wt.%) powder mixtures sintered body and pre-sintered 2A12 alloy were joined by P/M bonding, while TC4 alloy and 2A12 alloy were bonded by DB. A thorough interface characterization and joint performance of the Ti/Al joints boned by P/M were done and compared with that of the DB using the same and the optimized process parameters to evaluate the applicability of P/M to dissimilar metal joining.

Materials and experimental procedure

For P/M bonding experiment, a nominal composition of Ti-6Al-4V (wt.%) was synthesized adopting mixed element powder by high-energy ball-milling for 2 h under Ar gas atmosphere. The Ti-6Al-4V composites bulks and the 2A12 bulks with the thickness of 10 mm (Φ40 mm) were produced by pressing. Ti sintered bodies with good strength were prepared after the bulks were sintered at 1300 °C for 4 h in vacuum atmosphere. The 2A12 bulks can’t be applied pressure directly due to the poor mechanical properties, so they need to be pre-sintered at 550 °C for 2 h before joining. For DB experiment, the base metals used were TC4 alloy and 2A12 alloy bar, received as cylindrical shaped (Φ40 mm × 10 mm). To achieve effective bonding, both experiments were conducted in the FVPHP-R-10FRET-40 hot pressing furnace, using the heating temperature of 625 °C and 650 °C, under pressure of 4 MPa and 10 MPa for 2 h and 3 h in a vacuum environment. The cooling process was conducted in the vacuum chamber from bonding temperature to 100 °C.

Specimens with dimension of 10×10×10 mm, which were parallel to the load applied direction, were fabricated from the Ti/Al bonded joints by wire cut machine. Specimens for optical microscopy (OM) were prepared using standard metallographic techniques. The etchants were Kroll’s and Keller’s reagents for Ti and Al matrix, respectively [15].

The structure and morphology of fracture surface were examined by the Nova Nano 230 scanning electron microscope (SEM). Distribution of Ti and Al at the bonded joint was determined by JXA-8230 electron probe microanalysis (EPMA). The phase constitutions were identified using a RIGAKU RAPIDⅡR type selected area X-ray diffractometer (XRD). Vickers microhardness testing across the joint was conducted by a UNHT type nanoindentation testing (NIT) with the maximum load of 3 mN and 30 sec load time. The tensile strength of the joint was tested by an Instron 3369 type mechanical testing machine, with a loading speed of 1.0 mm/min, and the mean values were calculated. The samples for tensile strength testing were cut by wire-cutting, as shown in Figure 1. Shear properties of the samples were determined based on GB3252-82. Samples of 8 mm diameter and 20 mm length were subjected to tension using an Instron 3369 machine with a crosshead speed set at 5 mm/min.

Figure 1:

The samples for tensile strength testing.

Results and discussion

Interface characterization

The optimized process parameters were selected through our early investigation. Figure 2 shows the optical micrographs of Ti/Al bonds during different processes, and the SEM micrograph of the P/M joint is given in Figure 3. It can be seen from Figure 2(a) and 3, the joint obtained by P/M is well boned, Ti grains show the duplex microstructure with equiaxed-α and flake-α, which shows good ductility and fatigue properties. The reaction layers are too thin to be observed, so no visible IMC is detected in these Ti/Al P/M bonds, while a new layer formed at the interface in Figure 2(b). As shown in Figure 2(c), also no IMC can be found, the columnar Ti grains near the joint interface markedly grow at higher temperature, even after shorter holding time.

Figure 2:

Optical micrographs in interface region of Ti/Al bonds obtained by different methods (a) P/M; (b) and (c) DB.

Figure 3:

The SEM micrograph in interface region of Ti/Al bonds obtained by P/M.

In order to determine the diffusion behavior at the interface, EPMA line scans were obtained for the Ti/Al joints, as shown in Figure 4. Cu, Mg and V are also as the major component of the bonds, due to its low content the changes of these elements can be ignored at this stage. It can be seen from Figure 4(a) and 4(c) that Ti and Al content gradually changes across the interface from Ti matrix to Al matrix, while a platform which corresponds to intermetallic phase can be found in Figure 4(b). The diffusion of atoms can be promoted at higher temperature or longer holding time.

Figure 4:

The elemental distributions at the interface region of Ti/Al bonds by different methods (a) P/M; (b) and (c) DB.

Since the powder has higher distortion energy by surface pretreatment such as polishing in P/M process, the surface stress can be related to the concentration distribution and enhances diffusion by reducing the activation energy in the thermal part. Densification of the powder will consume energy, and also due to the loose contact between the particles, less diffusion paths occur, the growth rate of IMC in P/M is lower than that in DB with the same process parameters.

It is known that diffusion-induced stresses or chemical stresses have been referred as the stresses caused by diffusion [16]. Diffusion induced by applied stresses will be expressed by the exponential form of the Dorn equation, and then the stresses are [17]:

(1)σ−PGn=23kTZDGbdbf.divu‾

where P is the applied pressure, σ−P is the diffusion-induced stresses, G is shear modulus, k is the Boltzmann constant, T is absolute temperature, Z is the dimensionless constant, D is appropriate diffusion coefficient, b is the Burgers vector of dislocations, d is the grain size, f and n are undetermined coefficients, uˉ is the volume growth rate of phase. This equation shows that the volume growth of IMCs will inevitably lead to the generation of diffusion-induced stresses, which in turn promote atomic interdiffusion. The diffusion rate of atoms at the interface increases with the stress due to a faster growth rate of IMCs in the DB bonds, thus a greater diffusion distance can be obtained and the diffusion has promoted the growth of the IMCs.

To further study the phase composition of the selected areas as shown in Figure 2(b), selected area X-ray diffraction was taken, the results were shown in Figure 5. The diffraction pattern indicates that AlTi₃, AlTi and Al₃Ti IMCs occur in this process, these brittle phases might be responsible for the mechanical properties of bonds. The brittleness at room temperature of IMCs is mainly related to its crystal structure, represented by the hindrance of dislocation slip movement. Al shows a face-center cubic (fcc) structure that has small lattice dislocation slip resistance and activation energy, so it exhibits high plastic. However, these IMCs show complex crystal structure with poor ductility. Also, due to the differences in crystal structure from the matrixes, atoms combined defects occur at the interface which reduces the bonding strength.

Figure 5:

The selected area XRD result on the DB bonds.

Joint performance

Vickers microhardness

Vickers microhardness tests were performed across the interface of these joints, and the results were shown in Figure 6. It shows that the microhardness of the joints zone in all of the specimens is higher than the Ti and Al base metals, the major factors affecting the fluctuation in hardness from the matrixes to the joints zone is the solid solution strengthening effect. It can also be seen that the hardness of the joints zone is much higher than that of the matrixes at the joint obtained by DB, which can be attributed to the formation of various hard Ti-Al intermetallic phases at the joint interface. The distribution of Vickers microhardness across the joint is in agreement with the observed morphology of the Ti/Al joint.

Figure 6:

Vickers microhardness distribution of the bonded joints.

Tensile test

The tensile strength profile is shown in Figure 7. It can be indicated that the bonding method has a straight relationship with the mechanical properties, and the tensile strength of joints obtained by P/M is obviously higher than those of joints bonded by DB. Decrease in tensile strength of the joints obtained by DB can be mainly ascribed to the formation of continuous brittle IMCs. A bond made by P/M may provide less IMCs to produce a sound metallurgical bond, while bonds made by DB tend to form a thick IMCs layer at the interface. In general, these occurred due to the comprehensive action of interdiffusion rate, the growth of grain, solid solubility, and nucleation energy barrier and densification behaviors.

Figure 7:

Stress–strain curves of tensile strength.

To investigate the tensile behaviors, the fractographs of Ti/Al joints with the highest and lowest tensile strength are examined by SEM, as shown in Figures 8 and 9. It can be seen from the fracture morphology at low magnification in Figure 8, some multi-island-shaped structures can be observed in the fracture near the Ti-matrix, while the fracture near the Al-matrix shows a laminated structure. EDS analyses results obtained from the different regions of the fracture are shown in Table 1. The percentages of Al in region-Ⅰ, region-Ⅱ, region-Ⅲ, region-Ⅳ and region-Ⅴ are about 22.01 %, 47.15 %, 93.28 %, 76.89 % and 96.24 %, respectively. Taking into consideration these data, it can be concluded that the formation of various phases takes place, AlTi₃ in region-Ⅰ, AlTi in region-Ⅱ, Al₃Ti in region-Ⅳ and Al-based solid solution in region-Ⅲ and region-Ⅴ. Therefore, it is revealed that the tensile fracture occurs at the Al-IMCs layer.

Figure 8:

Fracture morphology of the Ti/Al diffusion bonded joint obtained at 625 °C.

Figure 9:

Tensile fractographs of the bonded joints.

Table 1:

EDS result of the mark positions in Figure 8.

Element	Region	Region	Region	Region	Region
(At.%)	-Ⅰ	-Ⅱ	-Ⅲ	-Ⅳ	-Ⅴ
Ti	77.99	52.85	6.72	23.11	3.76
Al	22.01	47.15	93.28	76.89	96.24

According to the SEM and EDS results, the multi-island-shaped structure suggests the presence of certain mechanical bonding at the interface of the DB bonds. These regions preserve the original matrix interfacial, with bond strength lower than that of the metallic bonding.

The fracture morphology at high magnification in Figure 9 shows that a brittle fracture occurs at the interface of both joints. There are some cleavage steps at the fracture of P/M bonds, and the crack propagation direction is consistent with the direction of cleavage steps. Due to the formation of the continuous brittle IMCs at the interface of DB bonds, IMCs directly bear the load until they break and form crack, crack grows along the grain boundaries, so granular substance is observed at the fracture surface. Thus, the formation of such compounds might be responsible for the mechanical deterioration of the DB bonds.

Shear test

The room temperature shear strength–testing curve of the joints diffusion bonded at different processing parameters is shown in Figure 10. It indicates that there is a big difference in the shear strength between the two diffusion bonds, the shear strength are 52 and 40 MPa respectively. The curve of the sample with lower shear strength approximately meet the Hooke’s law, the joint reaches the ultimate strength when fracture occurs, so it shows typical brittle fracture almost without plastic deformation occurring at the interface. While the other sample shows a significant plastic deformation behavior, the necking generates after exceeding the maximum load, and then the resistance-distortion ability of the joint decreased, the deformation increases until the joint fracture.

Figure 10:

The load–displacement curve of shear strength of Ti/Al diffusion bonded joints.

The shear fracture surface microstructures of the diffusion joints are shown in Figure 11. It shows that the fracture occurs in the IMCs layer due to the brittle Ti-Al intermetallic compound generated at the interface, so large granular substance is observed at the fracture surface, as shown in Figure 11(a) and 11(b). While in the other bonding process, since IMCs is invisible at the joint interface, the deformation of material takes place along the direction of strain, it increases with the increasing of load until the joint fracture. Therefore, the shear deformation shows significant tendency of direction, with a large number of zonal shear planes extending in the same direction.

Figure 11:

Shear fractographs of the Ti/Al diffusion joints obtained at different processing parameters.

Conclusions

Ti/Al was bonded successfully via diffusion bonding and two-stages sintering powder metallurgy respectively in this paper. The interface and joint performance of the Ti/Al joints were carefully characterized and compared. General conclusions of the current work can be summarized as following:

1. Optical micrographs in interface region of Ti/Al bonds show that no visible IMC is detected in the P/M bonds. The distribution of Ti and Al across joints was revealed by EPMA suggesting possible formation of reaction product layer at the interface of DB joint. The presence of AlTi₃, AlTi and Al₃Ti IMCs at the DB bonds obtained at the same process parameters with P/M was confirmed by XRD analyses.

2. Vickers microhardness was closely related to the concentration distribution and components, due to the formation of various IMCs, the joints zone is much harder than that of the matrixes at the joint obtained by DB. The tensile strength of Ti/Al joints is closely related to the formation of various IMCs. The highest tensile strength in the joint is 111 MPa, which is obtained at 625 °C with holding time of 180 min and pressure of 4 MPa.

3. Both fractures are found to be brittle failure, a few cleavage steps are observed at the tensile fracture of P/M bonds, while the fractographs of diffusion bonds show a few granular substances can also occur. Compared with DB, complete metallurgical bond with higher strength of joint can be obtained in P/M bonding. Therefore, P/M might be a new method for bonding dissimilar metals to prepare high-quality bonds.