High strength and high creep resistant ZrB₂/Al nanocomposites fabricated by ultrasonic-chemical in-situ reaction

1. Introduction

With strong demand for energy saving and emission reduction, aluminum alloys have been widely regarded as a substitute for steel and/or cast iron in transport systems, due to their low density, good formability and acceptable strength [[1], [2], [3]]. However, the application of aluminum alloys is greatly limited at high temperatures, because of their low creep resistance arising from the coarsening of precipitations and the easy climb of dislocations at elevated temperatures [[4], [5], [6]]. Many strategies were employed to enhance the creep resistance by introducing thermostable phases, especially the nano-dispersed phases, inside and/or on the boundaries of the aluminum grains to hinder the dislocation motion and the grain coarsening at elevated temperature, such as generating thermostable nano-precipitated phases via microalloying (as Sc, Er, La) [[7], [8], [9]], introducing oxide dispersions by powder metallurgy [10] and introducing reinforcements to form aluminum matrix composites [[11], [12], [13], [14], [15]]. Nevertheless, the strategies of microalloying and powder metallurgy are always plunged into the dilemma of high performance vs. high cost and complex processes, due to the expensive rare-earth elements and difficulty in the preparation of complex parts through powder metallurgy route [4,5,[16], [17], [18]]. Thus, the aluminum matrix composites with high creep resistance, have been thought to be a promising light materials used at elevated temperature [11,19], especially the in-situ aluminum composites, which have the advantages of small reinforcements, good thermodynamic stability and strong particle-matrix bonding compared with the ex-situ ones, i.e., even better performance at high temperature [[20], [21], [22], [23], [24]].

It is well known that the uniform distribution, small size and good interface bonding to matrix of the reinforcements are the eternal pursuit to fabricate composites with high performance [14,20,25,26]. In addition, our previous studies on in-situ aluminum composites have shown that the ZrB₂/Al nanocomposites can be synthesized from the industrialized-adaptive Al-K₂ZrF₆-KBF₄ reaction system [20,27]. Thus, in this study, the Al-K₂ZrF₆-KBF₄ in-situ reaction, assisted by melt ultrasonic vibration and spiral electromagnetic stirring, was employed to prepare ZrB₂/Al nanocomposites with fine size and uniform distribution of the in-situ nanoparticles. The microstructure and room temperature tensile property of the nanocomposites were invested and analyzed, firstly. Then, the high-temperature creep behavior of the aluminum matrix and the fabricated nanocomposites containing 1 vol.%, 2 vol.%, 3 vol.% and 5 vol.%, ZrB₂ were emphatically studied. Finally, the creep mechanism of the nanocomposites was further discussed.

2. Experimental

The raw materials were commercial pure Al ingots (≥ 99.5%) and powders of inorganic salts K₂ZrF₆ (≥ 99%) and KBF₄ (≥ 99%). And the reaction process of Al-K₂ZrF₆-KBF₄ system was drawn by the following chemical equations [20,28]:

2KBF₄ + 3Al = AlB₂ + 2KAlF₄ (1)

3K₂ZrF₆ + 13Al = 3Al₃Zr + K₃AlF₆ + 3KAlF₄ (2)

AlB₂ + Al₃Zr = ZrB₂ + 4Al (3)

The overall reaction can be expressed as:

K₂ZrF₆ (l) + 6KBF₄ (l) + 10Al (l) = 3ZrB₂ (s) + 9KAlF₄ (l) + K₃AlF₆ (l) (4)

To avoid the explosion hazard induced by the adsorptive water vapor of the air on the powder surface, the K₂ZrF₆ and KBF₄ reactant powders were dehydrated at 300 °C for 3 h firstly, and then mixed with mass ratio of 112:120 (112:(100 + 20)), about 20% excess of the KBF₄ powder calculated by overall reaction formula to ensure the reaction completed, due to the easy burning loss of B element in molten aluminum [20,27]. Then, the mixed salts were added to the molten Al at 1143 K, and the Zr and B elements could be absorbed and synthesize ZrB₂ reinforcements via spontaneous in-situ reaction. During the reaction, the high-intensity ultrasonic vibration with power and frequency of 2 kW and 20 kHz was introduced into the composite melt and a spiral electromagnetic stirrer (DJMH-250) was also utilized to increase the mass transfer with the reaction time of about 30 min.. After the reaction and slag removal, the composite melt was refined by the plunging of C₂Cl₆ at 1023 K, poured at 973 K into a copper mold (mold cavity size: 30 mm × 200 mm × 300 mm) and cooled to room temperature to obtain the ZrB₂/Al composite ingots. Through the above procedure, the nanocomposites with in-situ ZrB₂ volume fraction of 1 vol.%, 2 vol.%, 3 vol.% and 5 vol.% were prepared, where the conversion of mass fraction to volume fraction is realized through density of the ZrB₂ reinforcement of 5.8 g/cm³, and the content of ZrB₂ was further confirmed via the alkali dissolution of aluminum matrix alloy. In addition, the commercial pure Al with out in-situ nanoparticles was also melted and casted via the same route for comparison.

The compositions and microstructures of the fabricated materials were characterized by direct-reading spectrometer (SPECTRO MAXx mm06), X-ray diffraction (XRD, Rigaku Dmax/2500), optical microscopy (OM, Axio Observer Z1M), scanning electron microscopy (SEM, FEI Nova Nano450) and transmission electron microscopy (TEM, JEM-2100 F). The room temperature mechanical properties of the flat specimens (machined from the corresponding casted ingots with thickness of 4 mm, gauge length of 25 mm and gauge width of 8 mm) were tested on Shimadzu AG-X plus 10 kN universal testing machine at a strain rate of 10^-3 s^-1. In addition, the constant stress tensile creep tests of the specimens (machined from the corresponding casted ingots with a gauge length of 25 mm and a gauge diameter of 5 mm) were performed using a CTM504-B1 electronic creep testing machine in the air at 498 K, 523 K and 548 K with the applied stresses of 20 MPa, 25 MPa and 30 MPa, respectively.

3. Results and discussion

3.1. Microstructure and room temperature properties

The chemical compositions of the fabricated pure Al and ZrB₂/Al nanocomposites are shown in Table 1. It can be seen that, except the elements of Zr and B, the fabricated pure Al and nanocomposites exhibit the similar element content. Fig. 1 shows the XRD patterns of the fabricated 2 vol.% ZrB₂/Al nanocomposites treated without and with ultrasonic vibration. We can see that the diffraction peaks of Al and ZrB₂ phases could be clearly observed in the both composites. However, three are still peaks of Al₃Zr phase observed in the composite prepared without ultrasonic vibration treatment. Thus, the ultrasonic vibration could promote the synthesis of ZrB₂, which exhibits smaller size and higher performance than that of Al₃Zr [20,28].

We know that the high-energy ultrasonic could generate cavitation, acoustic streaming and radiation pressure in the melt, and results in promoting the chemical reaction, breaking clusters and uniform dispersion of the particles [27,29]. Fig. 2 gives the typical microstructure of the fabricated 2 vol.% ZrB₂/Al nanocomposites without and with ultrasonic vibration treatment. It can be seen that with the introduction of ultrasonic vibration in the composite melt, the in-situ particle clusters could be effectively broken into small ones in the final composites, shown as the OM images of Fig. 2 a and b, as well as the SEM images of Fig. 2(c and d). Meanwhile, we also found that the ultrasonic vibration can promote the synthesis and prevent the fast growth of the in-situ ZrB₂ ceramic particles, due to the fact that there are lath-shaped residual intermediate reaction product of Al₃Zr and some larger ZrB₂ particles (larger than 100 nm) observed inside the composites without ultrasonic treatment, while there are few observed in the ultrasonic-treated composites, shown in Fig. 2(c-f). In addition, according to Fig. 2(e and f), it can be seen that the synthesized particles are mainly in the size range of smaller than 100 nm, and the corresponding selected area electron diffraction (SAED) patterns further confirmed the synthesized particles to be ZrB₂. Thus, the ultrasonic vibration treatment in the melt is the key factor to the synthesis and uniform distribution of the nano-sized ZrB₂ ceramic particles. And the following microstructure characterization and property analysis are all performed on the composites fabricated by ultrasonic-chemical reaction.

Fig. 3 shows TEM images of the in-situ ZrB₂ nanoparticles in the fabricated composites. It can be clearly seen that the average diameter of ZrB₂ particles is about 10-50 nm with nearly equiaxed shape. The interfaces between the matrix and particles are net and no interfacial outgrowth is observed. Furthermore, we can also see, from the HRTEM image of Fig. 3(b), that the in-situ ZrB₂ nanoparticles inside the same grain exhibit different crystal orientations, i.e., various interface lattice mismatch, which has been proved to be benefit to hinder the movement of dislocations and improve the strength and creep resistance of the composites [4,8].

Fig. 4(a) gives the room temperature tensile curves of the prepared pure Al and ZrB₂/Al nanocomposites. It can be seen that with the volume fraction increasing of the in-situ ZrB₂ nanoparticles, the composites exhibited a dramatic increase in strength, and a maximum elongation at the ZrB₂ content of 2 vol.% in this study. Usually, the strain hardening ability of metals, dominated by the dislocation propagation and storage, is the key factor of strength and ductility for the unique matrix [30]. And the work hardening rates Θ of the prepared pure Al and ZrB₂/Al nanocomposites are derived from the tensile true stress-true strain data (Fig. 4(b and c)).

where σ is the true stress, ε is the true strain. It can be seen that all the samples exhibit a steep decrease in work hardening rate at stress below the yield stress, due to the elastic-plastic transition. The work hardening rates of the samples decrease with the strain after the elastic-plastic, with an obvious shoulder at strain of about 5%, which is respond to the dislocation propagating and storing process. After the shoulder, the Θ value decreases linearly with stress, which can be analyzed using the Kocks-Mecking model [31]:

where Θ₀ is the initial work hardening rate and K reflects the rate of dynamic recovery of existing dislocations during plastic deformation. And the values of Θ₀ and K for the samples were further established by fitting the Θ-σ curves, shown in Table 2.

We can see that with the volume fraction increasing of the in-situ nanoparticles, the values of Θ₀ increase from 1881.8 ± 66.5 MPa to 2990.5 ± 165.7 MPa and the values of K decrease from 26.7 ± 0.5-13.3 ± 0.1. Hence, one may deduce that the in-situ nanoparticles induced higher work hardening rate and lower dynamic recovery rate of dislocation during plastic deformation. Thus, the introduction of in-situ nanoparticles to the soft aluminum matrix could generate significant enhancement in strength via dislocation pinning and propagating caused by the nanoparticles, i.e., Orowan strengthening. For example, the prepared 1 vol.% ZrB₂/Al composite exhibits yield strength of 75 ± 2 MPa and ultimate strength of 166 ± 5 MPa, about 1.27 times and 1.46 times of the pure Al matrix.

Meanwhile, we can also find that except the strength improving with the content increase of the in-situ nanoparticles, the elongations of the composites increase when the content of nanoparticles is below 2 vol.%. This phenomenon can be also interpreted by the higher work hardening rate of the nanocomposites. It is well known that the higher the work hardening rate of the metals is, the larger the elongations are [30,32]. According to the Conside’re deformation criterion, the maximum of uniform tensile strain will be achieved, when [30]:

And the enhanced uniform elongation of the composites was caused by an improved normalized strain hardening rate defined Θ_N by:

The inset in Fig. 4(b) gives the variation of the normalized work hardening rate with true strain. It is clear that the composites with in-situ nanoparticles lower than 2 vol.% exhibit improved values of Θ_N. As a result, the composites achieve higher strength and larger elongation. However, when the volume fraction of the in-situ nanoparticles is higher than 2 vol.%, the values of Θ_N decrease dramatically after the uniform deformation (for example: about 13% of the 5 vol.% ZrB₂/Al nanocomposite) and result in low elongation of the composites, which is related to the in-situ nanoparticle agglomeration of the composites and the generation of stress concentration during plastic deformation [27,28].

3.2. High-temperature creep behavior

Fig. 5 shows the tensile creep curves of the casted pure Al and ZrB₂/Al nanocomposites under the temperature of 523 K and constant stress of 25 MPa. We can clearly see that with the content increasing of the nanoparticles, the creep fracture time increases, dramatically. Especially, it can be also found that the creep curves are all divided into three typical stages of Ⅰ: primary creep, Ⅱ: steady-state creep, and Ⅲ: creep fracture, where the steady-state creep (Ⅱ) enlarges and the creep rate decreases with content increase of the nanoparticles. Thus, it can be easily concluded that, the creep resistance of the nanocomposites is higher than the aluminum matrix and improves with the volume fraction increasing of the nanoparticles. The steady-state creep rates of the fabricated nanocomposites were calculated to be 2.12 × 10^-6 s^-1, 1.26 × 10^-6 s^-1, 5.56 × 10^-7 s^-1 and 2.71 × 10^-7 s^-1 for the composites with ZrB₂ nanoparticle volume fraction of 1 vol.%, 2 vol.%, 3 vol.% and 5 vol.% respectively, almost 2-18 times lower than that of the aluminum matrix (4.84 × 10^-6 s^-1) under the same creep condition.

Fig. 6 shows the creep behavior of the 2 vol.% ZrB₂/Al nanocomposite under different stress and temperature. We can see that the applied stress has a remarkable influence on the creep fracture time, as shown in Fig. 6(a). In order to investigate the effects of applied stress on the creep rate clearly, the variation of instantaneous creep rate vs. normalized creep time, t/t_f, where the t_f is the time from the applying of stress to the final fracture of the composites, was calculated by the differential of the creep curves of Fig. 6(a). As shown in Fig. 6(b). the steady-state creep rates, i.e., the minimum creep rate, were established to be 2.95 × 10^-7 s^-1, 1.26 × 10^-6 s^-1 and 5.32 × 10^-6 s^-1 for the applied stress of 20 MPa, 25 MPa and 30 MPa, respectively. Via the same calculation method, the steady-state creep rates of the 2 vol.% ZrB₂/Al nanocomposite at temperatures of 498 K and 548 K can also be established and shown in Fig. 6(c). We can find that the steady-state creep rate is rather sensitive to the creep temperature, compared with the applied stress. For example, the steady-state creep rate value was 3.43 × 10^-4 s^-1 at 548 K/25 MPa, about 272 times of that at 523 K/25 MPa (1.26 × 10^-6 s^-1), while only about 4-5 times of the prior one every increasing of 5 MPa.

It is well known that the steady-state creep rate, ε˙, can be represented as a power law of the applied stress and deformation temperature as [19,33]:

where A is the constant, σ is the applied stress, T is the absolute temperature, R is the gas constant, n is the apparent stress exponent and Q is the apparent activation energy. And a better high-temperature creep resistance can be expressed by a higher apparent stress exponent and higher apparent activation energy [6,19,34].

By taking the logarithm and natural logarithm of the power-law Eq. 9, the apparent stress exponent, n, and apparent activation energy, Q, can be obtained from slopes of the linear fitting of logε˙ vs. logσ and lnε˙ vs. 1/T, as indicated by the Eqs. 10 and 11. Fig. 7 gives the variation of logε˙ vs. logσ and lnε˙ vs. 1000/T based on the creep data of the composites. And the calculated values of the apparent stress exponent and apparent activation energy are established and listed in Table 3. It can be seen that the fabricated ZrB₂/Al nanocomposites exhibit the apparent stress exponents of 7.78-11.77, higher than that of pure aluminum and aluminum alloy (n = 3-5) [10] and the apparent stress exponent values of the nanocomposites increase with the increasing of in-situ nanoparticle content and decreasing of test temperature. Meanwhile, we can also find that the apparent activation energy was established to be 365.05-426.76 kJ/mol and rather higher that of the lattice diffusion of aluminum (142 kJ/mol) [35]. These creep results are the same with other reports for the nano-phase dispersed aluminum alloy and discontinuous particle reinforced metal matrix composites [[36], [37], [38]].

In order to interpret the mechanism of the improved creep resistance of the nanocomposites compared with the aluminum matrix, a threshold stress, σ0, which was always used to evaluate the creep mechanism of the dispersion-strengthened alloys and discontinuous particle reinforced metal matrix composites, was introduced here. In fact, the driving stress of creep should be an effective stress, σ-σ0, rather than the applied stress [39]. Therefore, the power-law creep equation can be modified as:

where A′ is a constant, σ0 is the threshold stress, n is the true stress exponent, and Q is the true activation energy. Generally, the values of n are taken as 3, 5 and 8 for the creep processes of dislocation slide, high-temperature dislocation climb controlled by lattice self-diffusion and a constant structure model controlled by lattice self-diffusion, respectively [10,17]. Then, applied stress independent σ0 can be estimated to be equal to σ, at ε˙1n=0 by linear fitting of ε˙1n vs. σ. And in this study, the true stress exponent value of 5 gives the best linearity of the creep data with correlation coefficients larger than 0.98 as shown in Fig. 8, which indicates that the creep mechanism of the composite is controlled by dislocation glide and climb as conventional aluminum [5,40]. Table 4 lists the estimated threshold stress values of the nanocomposites. As the apparent stress exponent and apparent activation energy, the fabricated nanocomposites exhibit an improved threshold stress with the volume fraction increasing of in-situ nanoparticles.

Fig. 9 shows the TEM images of the 2 vol.% ZrB₂/Al nanocomposite before and after creep test. It can be seen that the dislocations inside the matrix gains decrease significantly after the creep test, which is relate to the creep mechanism of dislocation climb of the composites. However, there are no evidence of grain coarsen (with sub-grain size of 0.5-$\widetilde{2}$ μm), which should be attributed to the grain boundary pinning effects of the in-situ stable ZrB₂ nanoparticles on or near the grain boundaries.

4. Conclusions

In this study, the ultrasonic-chemical in-situ reaction of the Al-K₂ZrF₆-KBF₄ system was employed to fabricate ZrB₂/Al nanocomposites. Through the microstructure, tensile property and creep behavior analysis of the fabricated nanocomposites, the following conclusions can be drawn:

(1)The in-situ ZrB₂/Al nanocomposites, with ZrB₂ nanoparticles size of 10-50 nm, were successfully fabricated from Al-K₂ZrF₆-KBF₄ system via ultrasonic-chemical in-situ reaction, where the ultrasonic vibration is benefit to preventing the fast growth as well as improving the uniformity of the in-situ synthesized nanoparticles. The in-situ nanoparticles are located on the boundary of and/or inside the aluminum matrix grains and exhibit various lattice-mismatch interfaces in the final composites.

(2)The fabricated nanocomposites exhibited an enhancement in both strength and ductility, due to the elevated work hardening ability, i.e., improved dislocation propagating ability and decreased dynamic recovery of the existing dislocations induced by the in-situ nanoparticles.

(3)The nanocomposites exhibit excellent creep resistance ability, about 2-18 times higher than the aluminum matrix. The stress exponent of the fabricated nanocomposites was identified to be 5, indicated that the creep mechanism of the nanocomposites is related to dislocation climb. And the enhanced creep resistance of the nanocomposites was attributed to the Orowan strengthening and grain boundary strengthening induced by the ZrB₂ nanoparticles.

Thus, the introducing nano-size ZrB₂ ceramic particles to the aluminum matrix by the ultrasonic-chemical reaction of Al-K₂ZrF₆-KBF₄ system, is a promising strategy to fabricate high strength and high creep resistance aluminum alloys with low cost and industrial scale.

Acknowledgments

This work was financially supported by the Natural Science Foundation of China (Nos. U1664254, 51701085, 51801074), the Natural Science Foundation for Young of Jiangsu Province, China (Nos. BK20160516 and BK201705433), the Six Talents Peak Project of Jiangsu Province (No. 2018-XCL-202), the Jiangsu Province Key Laboratory of High-end Structural Materials (No. HSM1803), the Research Foundation for Advanced Talents of Jiangsu University, China (No. 14JDG125), the Postdoctoral Science Foundation of Jiangsu Province, China (No. 1501029B), and the Postdoctoral Science Foundation of China (No. 2016M591780).

The authors have declared that no competing interests exist.

---