Introduction

Al–Zn–Mg–Cu alloys are widely used in the transportation and aerospace industries because of their ultra-high-strength, outstanding stress corrosion resistance, and high fatigue strength [1,2,3]. Weldability is essential for making these alloys widely used in production. Unfortunately, due to the notorious hot-crack susceptibility of these alloys, there is no special fusion welding wire available to connect them. Recent studies have found that introducing ceramic particles to the aluminum alloy welding pool can effectively modify the hot-crack susceptibility and enhance the mechanical properties of materials [4,5,6,7,8,9]. Huang et al. improved the weldability of spray-formed (SF) 7055 aluminum alloy by introducing TiB2 and ZrB2 ceramic particles [4,5,6]. Sokoluk et al. welded 7075 aluminum alloy with TiC/7xxx filler rods, which not only decreased the hot-crack sensitivity of 7075 aluminum alloy but also greatly improved the mechanical properties of welded joints [7].

However, those studies mentioned above only focused on the as-welded joints. As known, the properties of Al–Zn–Mg–Cu alloys, which are determined by the kind, size, and distribution of precipitates, could be improved by heat treatments [10,11,12]. Some studies have found that the introduction of ceramic particles can change the heat treatment behavior of aluminum alloys [13,14,15]. Some reported that ceramic particles accelerated the heat treatment behavior [16, 17], while others reported that ceramic particles decelerated the heat treatment behavior [18,19,20].

Xiao et al. found that TiB2/7050 aluminum matrix composites were more quench sensitive than 7055 alloys due to the dislocations caused by the coefficients of thermal expansion (CTE) mismatch and the high energy interfaces of TiB2 and α-Al [13]. Li et al. investigated the aging precipitation of TiN/7xxx aluminum matrix composites [16]. Compared with the alloys with only Ti or without any refiner, alloys with TiN promoted diffusely uniform precipitation and effectively improved the mechanical properties. Ma et al. found that TiB2 not only reduced the nucleation energy barrier of precipitation but also accelerated the diffusion of solute atoms and vacancies, thereby accelerating the aging kinetics [17]. However, Pal et al. found that SiC decelerated the aging kinetics of Al–Cu–Mg alloys due to lower vacancies, segregation of alloying elements, and inadequate dislocations [18]. Geng et al. found that TiB2 gathered along the grain boundaries (GBs) acted as shells to prevent the diffusion of alloying elements and inhabited the dissolution of the second phases, which decelerated the aging kinetics of alloys [19]. Hu et al. found that the formation of MgO in the B4C/Al interface depleted the Mg atoms and suppressed the precipitation during aging process [20]. Therefore, the effect of ceramic particles on the solution and aging behavior of aluminum alloy has not been determined, which depends on the kinds and content of ceramic particles, alloy composition and other factors.

Some literatures show that weld joints of Al–Zn–Mg–Cu alloys reinforced by ceramic particles preferred to fracture at the weld metal (WM) [4, 7, 21], which indicated that WM was the weakest position of joints. Therefore, appropriate heat treatment parameters can effectively improve the properties of ceramic particle reinforced weld joints. However, there is little research on the solution and aging behavior of 7xxx-WM reinforced by ceramic particles. The effect of TiC–TiB2 on the solution and aging behavior of SF 7055-WM was further investigated based on [21], which provided a foundation for the heat treatment process of ceramic particles reinforced weld joints.

Experimental

The preparation of TiC–TiB2/7xxx (BC) filler rods and the weld parameters for SF 7055 aluminum alloy can be found in [21].

The weld joints were solid solution treated at 475 °C for 1, 2.5, and 5 h and being quenching in water at 65 °C. Then these weld joints were artificially aged at 120 °C for 0–24 h, and taken out every 3 h. Metallographic analyses were observed by the optical microscope (Axioskop2-MAT). The phase composition of WM was observed by the X-ray powder diffraction (XRD, Shimadzu XRD-600, Cu Kα1), the scanning range of XRD is 30–80° and the scanning rate is 4°/min. Microstructures and fracture morphologies of WM were observed by the scanning electron microscope (SEM, Gemini SEM 300). The nano-strengthening precipitates of WM were observed by the transmission electron microscope (TEM, JEM-2010F). The microhardness of WM was obtained by the microhardness (MH-5D) tester at 100 g for the duration of 15 s. Based on GB/T 2651-2008, tensile tests were measured by the electronic universal testing machine (CMT 5205) at a drawing speed of 2 mm/min.

Results and discussion

Solution and aging treatment

Solution treatment

The SF 7055 aluminum alloy joints welded with 7055 and BC filler rods were treated with solution treatment. Figure 1 shows the micrographs of 7055-WM and BC-WM with solution treatment for 1, 2.5, and 5 h (S-0, S-1, S-2.5, and S-5). The microstructure of S-1 7055-WM is coarse dendrites with an average grain size of 115 μm (Fig. 1a). While the microstructure of S-1 BC-WM is fine equiaxed crystals with a uniform grain size distribution (30 μm), as shown in Fig. 1d. The grains of S-2.5 7055-WM coarsen into equiaxed crystals with an average grain size of 135 μm (Fig. 1b). However, the grain growth of S-2.5 BC-WM is not noticeable, and the average grain size is 35 μm (Fig. 1e). The grain size of S-5 7055-WM continues to grow, and the average grain size is 203 μm (Fig. 1c). The grain growth of S-5 BC-WM is still not noticeable, and the average grain size is 42 μm (Fig. 1f). Compared with the dissolution of alloying elements, TiC–TiB2 ceramic particles could not diffuse during solution treatment, which means ceramic particles can effectively pin the GBs and prevent the migration and growth of GBs, resulting in the grain refinement of WM [16, 22, 23].

Figure 1
figure 1

OM images of 7055-WM and BC-WM with different solution times: a–c 7055 S-1, S-2.5 and S-5; d–f BC S-1, S-2.5 and S-5

Full size image

Figure 2 shows the SEM and energy dispersive spectroscopy (EDS) images of 7055-WM and BC-WM with different solution times. Based on SEM images of WM, the second phases of WMs are measured by Image J, as shown in Fig. 3.

Figure 2
figure 2

SEM images and EDS of 7055-WM and BC-WM with different solution times: a–c 7055 S-1, S-2.5 and S-5; d–f BC S-1, S-2.5 and S-5

Full size image
Figure 3
figure 3

Area ratios of the second phases in 7055-WM and BC-WM with different solution times

Full size image

With the S-1 treatment, the are ratio of second phases in 7055-WM and BC-WM is 4.46 ± 0.16% and 2.61 ± 0.17%, respectively (Fig. 2a, d). With the S-2.5 treatment, the second phases in 7055-WM grains are almost completely dissolved, while the second phases in GBs still exist partially continuously (Fig. 2b), with an area ratio of 2.16 ± 0.23%. There are no large second phases in the grains and GBs of BC-WM, with area ratios of 1.40 ± 0.14% (Fig. 2e). According to EDS, the residual eutectic phases are rich in Al–Cu–Mg elements and Al–Cu–Fe elements.

With the S-5 treatment, the second phases of 7055-WM have been significantly dissolved, with an area ratio of 1.35 ± 0.17% (Fig. 2c), and the area ratio of the second phases in BC-WM is 1.19 ± 0.14% (Fig. 2f). Although the solution time has doubled, the effect of solution in BC-WM has no longer significantly increased. Therefore, the eutectic phases of BC-WM can be dissolved by S-2.5 treatment. In addition, the content of eutectic phases in S-5 7055-WM is the same as in S-2.5 BC-WM, indicating that ceramic particles can accelerate the solution of alloying elements in WM.

Figure 4 shows the XRD diffraction patterns of S-2.5 7055-WM and BC-WM. S-2.5 7055-WM is composed of α-Al, MgZn2, and Al2CuMg, and Al7Cu2Fe. S-2.5 BC-WM are mainly composed of α-Al, Al3Ti, TiC, Al2CuMg and Al7Cu2Fe. According to XRD (Fig. 4) and EDS (Fig. 3), MgZn2 dissolved into the Al lattice during solution treatment, and the two residual eutectic phases are Al2CuMg and Al7Cu2Fe. These WMs were further treated with S-5, but these residual phases were still residual and could not be completely dissolved (Fig. 2f). The results are consistent with [15, 24, 25].

Figure 4
figure 4

X-ray diffraction pattern of S-2.5 7055-WM and BC-WM

Full size image

The relationship between diffusion temperature and time of alloying elements during homogenization can be described as [26]:

$$ A\left( t \right) = A_{0} \,{\text{exp}}\left[ { - \frac{{4\pi^{2} D_{0} t}}{{L^{2} }}{\text{exp}}\left( { - \frac{Q}{{{\text{RT}}}}} \right)} \right] $$
(1)

where \(A_{0}\) is the initial amplitude of the composition segregation, \(D_{0}\) is the diffusion coefficient, \(Q\) is the diffusion activation energy, \(R\) is the gas constant, \(T\) is the temperature, \(t\) is the time of heat treatment, and \(L\) is the inter-dendritic spacing.

Equation (1) is used to calculate the theoretical solution time of BC-WM. The diffusion coefficient of Cu is much lower than Mg and Zn at the same temperature. Therefore, the solution process is suggested to be controlled by the diffusion of Cu. The solid solubility of Cu in the Al matrix is 5.6%. Assuming the alloying elements are almost completely soluted when the segregation amplitude of Cu is reduced to 5.6%, i.e.:

$$ \frac{A\left( t \right)}{{A_{0} }} = \frac{5.6}{{100}} $$
(2)

then

$$ A\left( t \right) = A_{0}\, {\text{exp}}\left[ { - \frac{{4\pi^{2} D_{0} t}}{{L^{2} }}{\text{exp}}\left( { - \frac{Q}{{{\text{RT}}}}} \right)} \right] $$
(3)

By taking natural logarithms of both sides, Eq. (3) can be rewritten as

$$ \frac{1}{T} = \frac{R}{Q}{\text{ln}}\left( {\frac{{4\pi^{2} D_{0} t}}{{2.88L^{2} }}} \right) $$
(4)

The average inter-dendritic spacing \(L\) of as-welded BC-WM is 26 μm [21], respectively. \(D_{0} \left( {{\text{Cu}}} \right)\) = 0.084 cm2/s, \(Q\left( {{\text{Cu}}} \right)\) = 136.8 kJ/mol, \(R\) = 8.314 J/(mol K), and the solution temperature \(T\) is 748 K. By substituting them into Eq. (4), the corresponding solution time is ca. 5 h (4.97 h). However, as shown in the SEM images (Fig. 2) and the area ratios of second phases (Fig. 3), there is no discernible change between the alloying elements of BC-WM with S-2.5 treatment and S-5 treatment, indicating that ceramic particles reduce the solution time of WMs. Both experimental results and theoretical calculations prove that ceramic particles can accelerate the solution of alloying elements.

Figure 5 shows the microhardness of 7055-WM and BC-WM with different solution times. It is found that at the same solution time, the microhardness of BC-WM is higher than 7055-WM. The microhardness of 7055-WM decreases with increasing solution time, from 111 HV0.1(S-1) to 106 HV0.1(S-5). However, the microhardness of BC-WM barely decreased and remained above 120 HV0.1. The superior microhardness of WM reinforced by ceramic particles was caused by the grain refinement strengthening (Fig. 1), ceramic particles with higher hardness, and the solution strengthening of more alloying elements [16].

Figure 5
figure 5

Microhardness of 7055-WM and BC-WM with different solution times

Full size image

Aging treatment

Figure 6 shows the age-hardening curves of 7055-WM and BC-WM (S-1, S-2.5, and S-5). WMs increase rapidly during the initial aging process and then exhibit peak-aging behaviors.

Figure 6
figure 6

The age-hardening curves of 7055-WM and BC-WM: a S-1; b S-2.5; c S-5

Full size image

With the S-1 treatment, BC-WM exhibit peak-aging at 18 h, and the microhardness is 203.48 ± 1.27 HV0.1. However, the microhardness of 7055-WM increases consistently during the whole aging process and is 195.24 ± 1.14 HV0.1 at 24 h. With the S-2.5 treatment, BC-WM exhibit peak-aging at 15 h, which is 206.86 ± 2.20 HV0.1, and the aging kinetics increase compared to S-1. However, the microhardness of 7055-WM still increases consistently, and is 198.08 ± 2.15 HV0.1 at 24 h. With the S-5 treatment, the aging kinetics of BC-WM no longer increase, and exhibit peak-aging at 15 h, which is 204.68 ± 2.62 HV0.1 and 205.60 ± 1.36 HV0.1, respectively. However, the aging kinetics of 7055-WM continues to increase, exhibiting peak-aging at 21 h, which is 198.18 ± 1.55 HV0.1.

It is found that compared with 7055-WM, the microhardness of BC-WM with ceramic particles has faster aging kinetics at the same solution time. The CTE dislocations and the particles/matrix interface could act as the pipe diffusion of alloying elements, accelerating the aging kinetic and shortening the time to peak-aging [17, 27]. By comparing the age-hardening curves of WM under different solution times, with each increasing solution time, the time of WM to peak-aging decreases. The optimal solution time of BC-WM is 2.5 h, in which eutectic phases are almost completely dissolved and possess the best aging kinetics. When the solution time is less than 2.5 h, the eutectic phases of BC-WM dissolve incompletely, and the aging kinetics is insufficient. When the solution time exceeds 2.5 h, the residual eutectic phases could not dissolve anymore, and the aging kinetics could no longer increase.

Tensile properties

Figure 7 shows the stress–strain curves of 7055 and BC weld samples with S-1A-15, S-2.5A-15, and S-5A-15 treatments. Since these specimens all fractured at WM, these values are all the strength of WM, which is consistent with the trend of the age-hardening curves (Fig. 6).

Figure 7
figure 7

The stress–strain curves of 7055 and BC weld samples with S-1A-15, S-2.5A-15, and S-5A-15 treatment

Full size image

At the same aging time (15 h), the tensile properties of 7055 joints improve with increasing solution time. The tensile strength and ductility of S-5A-15 7055 joints is 397 ± 15 MPa and 2.61 ± 0.24%, respectively. However, the tensile properties of BC joints exhibit the peak values with S-2.5A-15 treatment, which are 530 ± 15 MPa, 3.9 ± 0.29%, respectively (S-2.5A-15 7055: 358 ± 16 MPa, 2.38 ± 0.42%). The alloying elements in BC-WM have been nearly completely soluted with S-2.5 treatment. The grains of BC-WM could coarsen with the increasing solution time. Therefore, the tensile properties of BC-WM decreased with the S-5A-15 treatment.

Figure 8 shows the fracture morphologies of S-2.5A-15 7055 and BC joints. The grains of S-2.5A-15 7055-WM have continued to grow more than 150 μm in the aging process. It can be seen that some second phases exist in the fractures, which are residual eutectic phases from the solution process according to the EDS (Fig. 8a, b) [25]. The eutectic phases in the S-2.5A-15 7055-WM GBs are still partly continuously present, which are the initiations of cracks under load. However, the microstructure of BC-WM is fine equiaxed crystals less than 45 μm, whose growth is limited by ceramic particles during the aging process (Fig. 8c-e). In addition, the initiation and propagation paths of cracks have also decreased due to the reduction of eutectic phases. Therefore, the strength of BC joints is superior to the 7055 joints with S-2.5A-15 treatment.

Figure 8
figure 8

Fracture micrographs of 7055 (a, b) and BC (c–e) joints with S-2.5A-15 treatment

Full size image

Figure 8d, e shows that some dimples with some particles exist in the BC joints. According to the EDS and [25], these submicron particles are Al3Ti, TiC and TiB2, which could enhance the strength of joints by Zener pinning strengthening [28]. Therefore, the tensile strength of the BC joints is better than 7055 joints with S-2.5A-15 treatment.

Analysis of precipitates in aging process

Figure 9 shows the microstructure of ceramic particles in the BC-WM. The diameter of the ceramic particles is ca. 50–100 nm, as shown in Fig. 9a. Some CTE dislocations in the α-Al matrix, around ceramic particles, could accelerate the aging behavior of precipitates, as shown in Fig. 9b. To acknowledge the precipitation behaviors of WM without or with ceramic particles in the aging process more accurately, BC-WM and 7055-WM with S-2.5A-15 were further investigated by TEM analysis.

Figure 9
figure 9

The SEM image (a) and TEM image (b) of ceramic particles in the BC-WM

Full size image

Figure 10 shows the TEM images of S-2.5A-15 BC-WM. The moiré fringe exists in the BC-WM GBs. The precipitate free zone (PFZ) has a width of 37 nm. The selected area electron diffraction (SAED) pattern of the [112]Al projection is shown in Fig. 10c. As seen, the diffraction spots of α-Al have been identified. Meanwhile, there are also some weak spots in the SAED pattern, which have been indexed [29,30,31]. The weak spots located at the (311)Al/2 have been indexed as Guinier–Preston-II (GP-II) zones, as shown by the blue arrows in Fig. 10c [32]. The weak spots located at the 1/3(220)Al and 2/3(220)Al have been indexed as η' phases, as shown by the red arrows in Fig. 10c. The weak spots located at the (220)Al/2 have been indexed as η phases, as shown by the red arrows in Fig. 10c. Therefore, the precipitates of S-2.5A-15 BC-WM are GP-II zones, η', and η phases. Figure 10d shows the microstructure of precipitates in S-2.5A-15 BC-WM. The precipitates in BC-WM are evenly distributed in the matrix. The length of the η phase is ca. 15 nm and the thickness is ca. 3 nm (Fig. 10d).

Figure 10
figure 10

TEM images of S-2.5A-15 BC-WM: a morphology characterizations of GBs and the matrix; b the PFZ in GBs; c the SAED pattern of the matrix; d, e the microstructure of precipitates; f, g FFT and IFFT diffractograms of the Region 1 and 2

Full size image

Figure 10e-g shows the high-resolution transmission (HRTEM) images of precipitates in S-2.5A-15 BC-WM. According to the diffraction spots (Fig. 10f, g), the precipitates in the Region 1 and 2 (Fig. 10e) are indexed as GP-II zones and η' phases, respectively [33, 34]. The length of GP-II zones is ca. 4 nm, and the thickness is ca. 1 nm. The length of η' phases is ca. 5–8 nm, and the thickness is ca. 1–3 nm. It can also be found that the η' phases in the Region 2 (Fig. 10e) contact each other and present the trend to coarsen and grow into η phases, indicating that the over-aging behavior could occur if the aging time continues to increase.

There are two groups of lattice fringes at the GP-II/Al interface, which are 0.2360 nm and 0.2326 nm, corresponding to the (111)Al and GP-II zones, respectively (Fig. 10f). The well-bonding GP-II/Al interface indicates that the orientation relationship of GP-II zones and α-Al is coherent. As shown in Fig. 10g, there are also two groups of lattice fringes at the η'/Al interface, which are 0.2489 nm and 0.2421 nm, corresponding to (111)Al and (1003)η', respectively. A few dislocations exist at the η'/Al interface, indicating that the orientation relationship of the η' phases and α-Al is semi-coherent. The lattice fringes of (111)Al are larger than the standard value (d(111)Al = 0.2338 nm), indicating that the lattice of α-Al is distorted [33].

Figure 11 shows the TEM images of S-2.5A-15 7055-WM. The PFZ has not formed in the GBs, indicating S-2.5A-15 7055-WM is still in the early aging process and has not exhibited peak-aging (Fig. 11a) [35]. While the PFZ with a width of 87 nm, wider than that of BC-WM (Fig. 10b), forms in other GBs with the moiré fringe. (Fig. 11b). The SAED pattern of [110]Al projection is shown in Fig. 11c. There are no weak spots of other phases in Fig. 11c, indicating that few precipitates in S-2.5A-15 7055-WM, which is consistent with Fig. 11d. In addition, the large-size precipitates in S-2.5A-15 7055-WM are unevenly distributed (Fig. 11d). The corresponding FFT diffractograms of these precipitations are shown in Fig. 11e, f. The η' phases with a length of ca. 12–18 nm and a thickness of ca. 2–3 nm exist in the Region 1 of Fig. 11e and Region 3 of Fig. 11f. There are some disk-like precipitates in the Region 2 of Fig. 11e. However, there are no corresponding weak spots in the FFT pattern. This disk-like precipitates are GP-I zones, which is the early aging precipitates of the Al–Zn–Mg–Cu alloys [36]. There are also some GP-II zones with a length of ca. 7 nm and a thickness of ca. 1 nm in Fig. 11e. The number of precipitates in 7055-WM significantly decreases, the size increases, and the distribution is worse.

Figure 11
figure 11

TEM images of S-2.5A-15 7055-WM: a, b GBs and the PFZ; c the SAED pattern of the matrix; d the microstructure of precipitates; e, f HRTEM of precipitates

Full size image

The precipitation sequence of Al–Zn–Mg–Cu alloys can be described as: GP zones → η' → η [37, 38]. The η' (η) phases are the growth phases of GP zones. In the S-2.5A-15 state, the precipitates in BC-WM are GP-II zones and η' (η) phases (Fig. 10), and that in 7055-WM are GP-(I/II) zones and η' phases (Fig. 11). The appearance of η' (η) phases in BC-WM indicates that ceramic particles promote the aging behavior of precipitates. In addition, the finer and more dispersed precipitates in S-2.5A-15 BC-WM indicate that ceramic particles could modify the age-precipitation behaviors and enhance the mechanical properties.

Discussion

Evolution of precipitates and PFZs in WM during the aging process

Part 3.3 shows the differences in precipitates and PFZs in BC-WM and 7055-WM with S-2.5A-15 treatment, indicating ceramic particles modify the aging-precipitation behaviors. With the S-2.5 treatment, the alloying elements of ceramic particle-reinforced WMs (CRWMs) are almost completely dissolved, while those of 7055-WM are incompletely dissolved. In addition, the dislocations gathered by the CTE mismatch provide the diffusion paths of alloying elements. Therefore, the aging kinetics of CRWMs is faster than that of 7055-WM. Figure 12 shows the evolution of precipitates and PFZs in WM during the aging process.

Figure 12
figure 12

Evolution of precipitates and PFZs in WM without or with ceramic particles during the aging process: a, c, and e without ceramic particles; b, d, and f with ceramic particles

Full size image

With the S-2.5 treatment, the residual eutectic phases in 7055-WM GBs are more than CRWMs. Therefore, the alloying elements dissolved in the 7055-WM matrix are less than CRWMs. In addition, the CTE dislocations are formed in the CRWMs matrix (Fig. 12a, b).

In the early aging process, the precipitation behavior is driven by the Gibbs isotherm adsorption decoration, the repartition of solutes occurs between GBs and the matrix [39, 40]. Solutes tend to diffuse to the region with more vacancies. GBs, as vacancy sinks, can effectively capture the vacancies generated during solution treatment. Most solutes diffuse into the 7055-WM GBs. In addition, more alloying elements are in the 7055-WM GBs than in the matrix. Therefore, the larger rich-solute clusters form in the 7055-WM GBs, while the smaller rich-solute clusters form in the matrix (Fig. 12c). Although the same diffusion process of solutes occurs in the CRWMs, the CTE dislocations in the matrix, acting as vacancy sinks, can also capture the vacancies during solution treatment [41, 42]. Therefore, there is a competition for adsorbing solutes between GBs and the matrix in the CRWMs. The solutes diffused into CRWMs GBs decrease, while the concentration of solutes in the matrix remains high. Both larger rich-solute clusters form in GBs and the matrix of WMs with ceramic particles (Fig. 12d).

Subsequently, the precipitation behavior is driven by the spinodal decomposition [43,44,45]. The larger rich-solute clusters enrich the surrounding solutes and grow into precipitates. However, the smaller rich-solute clusters could not form precipitates and disappear due to the low content of solutes. Since most of the larger rich-solute clusters of 7055-WM exist at GBs, the solutes near GBs dissolve into GBs, which form the wider PFZ. In addition, due to a few larger rich-solute clusters and a relatively short aging time of 7055-WM, these larger clusters could not obtain sufficient solutes. Thus, the precipitates of 7055-WM are GP-(I/II) zones and η' phases (Fig.12e). In addition, the alloying elements solute in the matrix are few. Thus, the number of precipitates is small, and the size is large.

Compared with the 7055-WM, more large clusters exist in the matrix of CRWMs. The CTE dislocations could also capture solutes and accelerate the aging behavior (Fig. 9). Therefore, solutes near GBs are less likely to diffuse to GBs, resulting in a narrow PFZ. In addition, GP-II zones and η' (η) phases form in the CRWMs. Meanwhile, there is competition for adsorbing solutes among these larger clusters during the aging process. Thus, there are more fine precipitates with a dispersed distribution in the CRWMs (Fig.12f).

Strengthening mechanism

Compared with 7055-WM, CRWMs have superior mechanical properties with S-2.5A15 treatment, which are mainly strengthened through grain refinement strengthening, Zener pining strengthening, CTE mismatch strengthening, and precipitation strengthening.

  1. (1)

    Grain refinement strengthening.

After solution-aging treatment, the CRWMs grains are finer than the 7055-WM coarse grains [23]. The strength increment \(\Delta \sigma_{H}\) [46] caused by the grain refinement is as follows:

$$ \Delta \sigma_{H} = kd^{ - 1/2} $$
(5)

where \(k\) is the strengthening constant, and d is the grain size of α-Al. The severe coarsened grains in 7055-WM are more than 150 μm, while the grains in CRWMs are ca. 50 μm, indicating that the grain growth is inhibited by the ceramic particles. Therefore, the mechanical properties of CRWMs could be improved by grain refinement strengthening.

  1. (2)

    Zener pining strengthening.

When the tensile samples are stretched, the load is transferred from the soft phases α-Al to the hard phases submicron TiC, TiB2, and Al3Ti. These submicron ceramic particles could pin and hinder the migration of dislocations under load, thereby improving the mechanical properties of CRWMs. The strength increment \(\Delta \sigma_{{\text{z}}}\) [47] caused by the Zener pining is as follows:

$$ \Delta \sigma_{{\text{z}}} = \left( {3f\Gamma } \right)/\left( {2r} \right) $$
(6)

where \(\Gamma\) is the GBs tension, \(r\) is the radius of the pinning particles. The submicron ceramic particles in BC-WM are TiC, TiB2, and Al3Ti. As seen from Eq. (7), the improvement in BC-WM strength is significant as the number of submicron ceramic particles increases.

  1. (3)

    CTE mismatch strengthening.

When the temperature changes, thermal stress is generated at the particles/matrix interface by the CTE mismatch between particles and α-Al, resulting in distortions of lattices and increasing the resistance to dislocation migration. In addition, the CTE dislocations are also gathered at the particles/matrix interface. Therefore, the mechanical properties of CRWMs could be improved by CTE mismatch strengthening. The strength increment \({\Delta }\sigma_{{{\text{CTE}}}}\) [48] is as follows:

$$ {\Delta }\sigma_{{{\text{CTE}}}} = \eta Gb\sqrt {\left( {6{\Delta }T{\Delta }Cf} \right)/\left[ {bd\left( {1 - f} \right)} \right]} $$
(7)

where η is a constant equal 1 [48], G is the elastic shear modulus of the matrix \(b\) is the Burgers vector, \({\Delta }T\) is the temperature change from the heat treatment temperature to room temperature (RT), \({\Delta }C\) is the CTE difference between the matrix and ceramic particles. Therefore, the mechanical properties of CRWMs could be effectively improved when the weld joints cool from the aging temperature to RT.

  1. (4)

    Precipitation strengthening.

Precipitation strengthening [49], which includes the shearing and Orowan strengthening, is the most important strengthening mechanism in Al–Zn–Mg–Cu alloys. When the size of precipitates is small, dislocations could pass them by cutting, and the dominant strengthening mechanism is the shearing strengthening. While the size of precipitates is large, dislocations could not pass them, but bulge between and loop around, and the dominant strengthening mechanism is the Orowan strengthening [50].

The size of GP-II zones and η' phases is small, and the orientation relationships with the matrix are coherent and semi-coherent, respectively [51]. Therefore, dislocations could pass them by shearing. The strength increment \({\Delta }\sigma_{{\uptau }}\) [52] caused by shearing strengthening is as follows:

$$ {\Delta }\sigma_{{\uptau }} = {\text{c}}_{1} f^{m} r_{p}^{n} $$
(8)

where \({\text{c}}_{1}\),\( m\), and \(n\) are constants, \(r_{p} \) is the radius of shear precipitates. However, the orientation relationships between the η phases and the matrix is incoherent [35]. Therefore, dislocations bulge between η phases and bypass them. The strength increment \({\Delta }\sigma_{{{\text{or}}}}\) [52] caused by Orowan strengthening is as follows:

$$ {\Delta }\sigma_{{{\text{or}}}} = {\text{c}}_{2} f^{1/2} r^{ - 1} $$
(9)

where \({\text{c}}_{2}\) is constant. In this case, the Orowan strengthening decreases with increasing radius of precipitates [52].

For 7055-WM, the precipitates are mainly GP-(I/II) zones and η' phases, with only shearing strengthening [53]. However, the precipitates of BC-WM are GP-II zones, η', and η phases, indicating that the precipitation strengthening of BC-WM is the coupling of shearing strengthening and Orowan strengthening. In addition, the precipitates in 7055-WM are large and few, which contributes to little strength improvement. However, the precipitates in BC-WM are finer and more numerous, which contributes to a more significant strength improvement. Therefore, the strength of CRWMs is superior to 7055-WM with S-2.5A15 treatment.

Conclusions

  1. (1)

    During solution treatment, ceramic particles hinder the grains' coarsening, accelerate the solution behavior, and enhance the microhardness of CRWMs. The optimal solution time of CRWMs is 2.5 h. The eutectic phases are almost completely dissolved, and the aging kinetics of CRWMs is the best.

  2. (2)

    During the aging process, ceramic particles accelerate the precipitation behavior of CRWMs, decrease the width of PFZs, and modify the aging behavior of precipitates.

  3. (3)

    With the 475 °C/2.5 h + 120 °C/15 h treatment, the tensile strength (530 ± 15 MPa) and ductility (3.9 ± 0.29%) of the BC joints are high 7055 joints (358 ± 20 MPa, 2.38 ± 0.42%), respectively. The dominant strengthening mechanisms of CRWMs are Grain refinement, Zener pinning strengthening, CTE mismatch strengthening and precipitation strengthening.