High-temperature damping capacity of fly ash cenosphere/AZ91D Mg alloy composites

2.1 Materials

AZ91D Mg alloy ingots and FAC particles were selected as the raw materials for preparing FAC/AZ91D composites. The chemical composition of AZ91D Mg alloy ingots (wt.%) is given as follows: Al, 9.07; Zn, 0.62; Mn, 0.21; Si, 0.034; Cu, 0.003; Fe, 0.0022; Ni, 0.00033; and Mg, balance. The contents of various phases in FAC particles (wt.%) are given as follows: SiO₂, 62.79; Al₂O₃, 24.24; Fe₂O₃, 3.86; CaO, 1.78; MgO, 1.28; K₂O, 2.29; Na₂O, 1.30; SO₃, 0.21; and others, balance.

2.2 Sample preparation

Before the experiment, the FAC particles whose median particle size is 145.7 μm were washed with 5 wt.% NaOH solution to remove the impurities adsorbed on the surface of FAC. The experimental setup for preparing the FAC/AZ91D composites is shown in Figure 1. The technology for preparing the composites comprises the following steps: first, AZ91D Mg alloy ingot was melted at 720°C at an electrical resistance furnace under a (3 vol.% SF₆+CO₂) protection gas. Then the melt was moved into another little furnace under the protection gas and stirred at 590°C as semisolid fluid. Simultaneously, 2, 6, and 10 wt.% FAC particles were added to the melt, respectively, and the slurry was stirred at a rotation speed of 675 rpm for 3 min with a stirring paddle to ensure the uniform distribution of FAC particles. After that, the slurry was rapidly reheated to 720°C and stirred for 1 min, and then the slurry was poured into a graphite mold preheated to 120°C. To study the evolution of the microstructure of the composites at elevated temperatures, 6 wt.% FAC/AZ91D composites were cut into the size of ϕ28 mm×20 mm, heated to 100°C, 150°C, and 200°C, respectively, and held for 30 min, then taken out for water quenching (approximately 20°C) rapidly to preserve the microstructure characteristics at elevated temperatures.

Figure 1:

The experimental setup for preparing the FAC/AZ91D composites.

2.3 Measurements and instruments

The damping capacity was measured by a dynamic mechanical analysis meter (DMA-Q800) in the double cantilever beam mode. The specimens for damping test were cut into the size of 60×6×1 mm with electric spark cutting. The test conditions were as follows: the strain amplitude (ε) was 1×10⁻⁵; the vibration frequencies (f) were 0.5, 1, and 5 Hz, respectively; the temperature (T) range was from room temperature to 320°C; and the heating rate (dT/dt) was 5°C/min. The damping capacity was evaluated as loss tangent (tan ϕ).

The microstructure was observed with an optical microscopy (OM, Nikon Epiphot-300), and the etching solution was 5 wt.% oxalic acid solution. The crystalline phases were determined by X-ray diffraction (D/MAX-2000 PC) with Cu K_α radiation.

3.1 Microstructures at elevated temperatures

Figure 2 shows the microstructures of AZ91D Mg alloy and FAC/AZ91D composites at room temperature. It can be found that the discontinuous second phase (Figure 2A) was distributed along the grain boundaries in AZ91D Mg alloy. After FACs were added, the interfacial reaction occurred between FAC and AZ91D Mg alloy, and the reaction products increased with the increase in FAC content (Figure 2B–D). It can also be found that the grain sizes of the matrix in FAC/AZ91D composites are finer than that of AZ91D Mg alloy, indicating that the addition of FAC particles has an obvious effect on the grain refinement, which accords with the report in the literature [20]. The discrete second phase became continuous and fine after the addition of FAC. Figure 3 shows the XRD patterns of AZ91D Mg alloy and 6 wt.% FAC/AZ91D composites. It can be found that the phases in the matrix are mainly α-Mg and β (Mg₁₇Al₁₂), and the reaction between FAC and AZ91D Mg alloy matrix forms two new phases of Mg₂Si and MgO. According to Figure 2, there are fewer reaction products (Mg₂Si) when the content of FAC is 2 wt.%. When the content of FAC is up to 6 wt.%, some sharp polygonal Mg₂Si particles distribute in the matrix or near the wall of FAC. When the content of FAC is 10 wt.%, massive cloddy Mg₂Si particles are attached on the wall of FAC. The more the content of FAC, the more intense the reaction between AZ91D Mg alloy and FAC. Some Mg₂Si particles dispersed along the grain boundaries can hinder the boundary motion and limit the growth of the matrix grain. Hence, the reaction between the wall of the FAC and the molten Mg matrix can lead to the refinement of microstructure of the composites. It can also be found that some tiny pores exist at the walls of FAC in FAC/AZ91D composites.

Figure 2:

Microstructures of AZ91D Mg alloy and FAC/AZ91D composites with different FAC contents at room temperature. (A) AZ91D Mg alloy matrix, (B) 2 wt.% FAC/AZ91D composites, (C) 6 wt.% FAC/AZ91D composites, and (D) 10 wt.% FAC/AZ91D composites.

Figure 3:

XRD patterns of (A) AZ91D Mg alloy and (B) 6 wt.% FAC/AZ91D composites.

The microstructures of 6 wt.% FAC/AZ91D composites after being heat treated at 100°C, 150°C, and 200°C, respectively, are shown in Figure 4. It can be found that a certain amount of the β phase is dissolved in the matrix after heat treatment and becomes smaller (compared with Figure 2C). The β phase is the smallest and more continuous after heat treatment at 200°C, which means that the β phase is gradually dissolved in the matrix as the temperature increases, and then the β phase shows a continuous reticular structure. It can also be found that the element diffusion capacity of Si is enhanced with increasing temperature. Some Mg₂Si particles are distributed in the vicinity of FAC as a small granular at first and then move to the matrix and grow at elevated temperatures, and a partial agglomeration situation appears at 200°C.

Figure 4:

Microstructures of 6 wt.% FAC/AZ91D composites after heat treatment at different temperatures: (A) 100°C, (B) 150°C, and (C) 200°C.

3.2 High-temperature damping capacity

Figure 5 shows the temperature dependence of the damping capacities of AZ91D Mg alloy and FAC/AZ91D composites at ε=1×10⁻⁵ and f=1 Hz. It can be noted that the damping capacities of both the composites and the AZ91D Mg alloy enhance with the increase in temperature, and the damping capacities of the composites are higher than that of AZ91D Mg alloy under different temperatures. As reinforcements, FACs significantly enhance the damping capacity of AZ91D Mg alloy, but FACs themselves are very low intrinsic damping materials, and the high damping capacity of FAC/AZ91D composites is derived from the interaction between FAC and AZ91D Mg alloy matrix. The damping capacities of the matrix and the composites improve slowly at temperatures lower than approximately 125°C, and 10 wt.% FAC/AZ91D composites show the highest damping capacity. The damping capacities of the matrix and composites increase quickly at temperatures higher than 125°C, and 2 wt.% FAC/AZ91D composites possess the highest damping capacity. When the temperature approaches 225°C, the damping capacity shows a sharp increasing trend. However, the damping capacity of AZ91D Mg alloy is always the lowest compared with the composites.

Figure 5:

High-temperature damping capacities of FAC/AZ91D composites and AZ91D Mg alloy.

Figure 6 shows the high-temperature damping capacity of 6 wt.% FAC/AZ91D composites under different frequencies. It shows that the damping capacities of AZ91D Mg alloy and the composites are almost frequency independent at low temperatures; however, they become much sensitive to frequency at elevated temperatures. The damping capacity of the composites increases with deceasing frequency, which shows a typical anelastic relaxation characteristic. After subtracting a background damping consisting of a superposition of an exponential term and fitting damping peaks according to the Gaussian function, bare damping peaks of 6 wt.% FAC/AZ91D composites at three testing frequencies can be obtained, as shown in Figure 7. The background damping Qb−1 can be expressed by the following equation [21]:

Figure 6:

Temperature dependence of damping capacity under different frequencies of 6 wt.% FAC/AZ91D composites.

Figure 7:

Damping peak of fitting for 6 wt.% FAC/AZ91D composites.

where A is a material constant, k is the Boltzmann constant, T is the absolute temperature, and B and C are the empirical parameters determined by the intercept and slope of a plot of ln Qb−1−1/T. The high-temperature background damping is actually caused by the crystal defects in the grain, which results from several mechanisms [22]. It can be found that the damping peaks shift to higher temperatures with the increasing frequencies, which shows that the peaks have the characteristic of the thermal activation relaxation process [18]. The relaxation process proceeds by atom diffusion; thus, the relaxation time and the activation energy correspond with the Arrhenius equation [18]:

where τ₀ and k are the exponent factor and the constant and H is the activation energy. At the peak position, ωτ=1 is met. Thus, the activation energy (H) can be calculated by the slope of ln (2πf)−1000/T_p, where T_p is the peak temperature. In light of the data in Figure 7, the Arrhenius plot of 6 wt.% FAC/AZ91D composites can be obtained and is shown in Figure 8. The data (T_p, H) of the matrix and the composites obtained by the same process are listed in Table 1.

Figure 8:

Arrhenius relation between the circle frequency and the peak temperature of 6wt.% FAC/AZ91D composites.

4.1 The influence of FAC on damping capacity

As demonstrated in Figure 5, the damping capacities of the composites are better than that of AZ91D Mg alloy because of the addition of FAC. The addition of FAC can modify the microstructures of the AZ91D Mg alloy matrix, including the grain size, the thermal mismatch inducing dislocations, and the interfaces between FAC and matrix.

The thermal mismatch due to the different coefficients of thermal expansion between AZ91D Mg alloy and FAC results in the residual strain, which can cause high dislocation density in the matrix alloy. The dislocation density in FAC/AZ91D composites is higher than that of AZ91D Mg alloy, which can lead to higher dislocation damping. Thus, the dislocation damping is one reason for the higher damping capacity of the composites when compared with AZ91D Mg alloy at the testing temperature range. In polycrystalline materials, the grain boundaries tend to exhibit anelastic characteristic, which can play an important role in improving the damping capacity. Grain boundary damping is very sensitive to temperature, and the viscous flow at grain boundaries converts the strain energy into other energy [23]; thus, the damping capacities substantially increase at high temperatures. The refinement of grains will lead to massive grain boundaries. Thus, the material with fine grains shows high damping capacity at high temperatures. According to Figure 2, the addition of FAC has an obvious influence on the grain refinement of AZ91D Mg alloy matrix. The grain size of FAC/AZ91D composites is finer than that of AZ91D Mg alloy, and they have higher grain boundary damping. Thus, the grain boundary damping is another reason for the high damping capacity of the composites as compared with AZ91D Mg alloy at the testing temperature range. The interfaces between FAC and AZ91D Mg alloy matrix are easy to slip at high temperatures; thus, interfaces have important influence on the damping properties of the composites. The interface damping is mainly affected by the bonding strength, interfacial reaction products, and porosities at the interface. After FACs were added into AZ91D Mg alloy, the elements of the FAC wall diffused to the matrix and formed the reaction products Mg₂Si and MgO, and then some tiny holes appeared in the walls of FACs. In addition, the differences between the crystal structure of FAC and AZ91D Mg alloy are large; thus, the interface between FAC and matrix may be incoherent, so the interface bonding is weak. In the case of a weak bonding interface, Lederman [24] has derived an upper bound of the damping component (η) owing to the initiation of the interfacial slip at the interface:

where C is the correction factor, μ is the coefficient of friction between the particulate and the matrix, k is the coefficient of the radial stress concentration at the interface, and V_P is the volume fraction of the particulates. According to Equation (3), interface damping is proportional to the volume fraction of FAC particulates. The existence of the interface damping is one reason for the better damping capacity of the composites at high temperatures. Therefore, the damping capacity of FAC/AZ91D composites is superior to that of AZ91D Mg alloy.

The difference in the damping capacity between AZ91D Mg alloy and composites can be explained by a dislocation damping mechanism at temperatures lower than approximately 125°C. According to the Granato–Lücke theory [25], [26], the damping mechanism of alloys is determined by the dislocation movement. The dislocation loops bow out between weak pinning points, which contribute to strain-independent damping (δ₀). The value of the strain-independent damping (δ₀) can be expressed as

when the breakaway stress is reached, the dislocation loops break away from the weak pinning points but are still pinning by strong points, which causes strain-dependent damping (δ_H). δ_H can be expressed as

where ρ and L_d are the dislocation density and the average distance between weak pinning points, F_B is the binding force between dislocations and weak pinning points, L_N is the average distance between strong pinning points, E is the elastic modulus, b is the magnitude of Burger vector, K is a constant, and η is the size ratio of the solvent atoms to solute atoms. The more the content of FAC, the more the products (Mg₂Si), which can reduce the L_N and enlarge the dislocation density ρ. At temperatures lower than approximately 125°C, the δ₀ component may play a leading role due to a low strain amplitude (ε=1×10⁻⁵); thus, 10 wt.% FAC/AZ91D composites possess the highest damping capacity because of their highest dislocation density ρ. At temperatures higher than 125°C, the grain boundary damping and the interface damping become more important. The existence of Mg₂Si phase around grain boundary can hinder the slide of the grain boundary, making the slide difficult, which will reduce the grain boundary damping. In addition, a lot of Mg₂Si phase gather in the wall of FAC in 10 wt.% FAC/AZ91D composites according to Figure 2, which could make the interface slip between FAC and matrix become difficult. Thus, 2 wt.% FAC/AZ91D composites show the highest damping capacity at temperatures higher than 125°C because of fewer Mg₂Si particles. The damping capacities of both composites and AZ91D Mg alloy show a sharp increasing trend due to the grain boundary damping after the temperature approaches 225°C.

4.2 Analysis of damping peak

As mentioned at Table 1, a damping peak around 150°C is found in the damping-temperature curves of AZ91D Mg alloy and FAC/AZ91D composites. Actually, the damping peak around 150°C in the damping-temperature curves of Mg alloy and its composites has been investigated in previous reports. Gu et al. [27] found the damping peak around 150°C in pure Mg and (SiC+Al₂O₃·SiO₂)/Mg composites, and the values of the calculated activation energy of the damping peaks were 1.22 and 1.44 eV, respectively. The damping peak is believed to be attributed to the dislocation movement. Deng et al. [28] also found the damping peak around 150°C in SiC_p/AZ91 and AZ91 Mg alloy in the solution state, and the values of the calculated activation energy were 1.76 and 1.30 eV, respectively. In Gr/AZ91 Mg alloy composites prepared by Wu et al. [29], the calculated H value of damping peak around 150°C was 1.28 eV. The peaks at approximately 150°C were also found in TiC/AZ91D magnesium alloy composites heat treated with T4 and T6 by Zhang et al. [30] and the values of the calculated activation energy of these composites were 2.18 eV for foundry composites, 1.52 eV for T4-treated composites, and 1.96 eV for T6-treated composites. The literature has shown that the grain boundary diffusion energy was 0.85–1.09 eV and the self-diffusion activation energy was 1.39 eV [28]. In this study, the calculated H values of AZ91D Mg alloy, 2 wt.%, 6 wt.%, and 10 wt.% FAC/AZ91D composites are 1.90, 1.98, 1.94, and 2.49 eV, respectively. The result shows that the addition of FAC can increase the activation energy of the composites, which are much larger than the grain boundary diffusion energy and the self-diffusion energy of the pure magnesium. Combined with previous studies, the damping peaks at approximately 150°C of AZ91D Mg alloy and its composites should be associated with the motion of dislocation. With the increase in temperature, the β phase gradually dissolves into the matrix, and the Mg₂Si phase slowly moves into the matrix and grows up in light of Figure 4. According to the G-L mechanism, the dislocation is pinned by the weak pinning such as solute atom and the strong pinning such as reinforcement particulates (Mg₂Si) and precipitation (β phase). The β phase dissolved in the matrix can result in the decrease in the average distance between the weak pinning points (L_d), and the decrease in the β phase, agglomeration, and growth of Mg₂Si phase can increase the average distance between the strong pinning points (L_N). Thus, the damping capacity can increase continuously at low temperatures. When temperature reaches a certain value (approximately 150°C), the dislocation would move quickly and then break away in the avalanche-like mode from the weak pinning points. It would only be pinned by the strong pinning points, and then the dislocation damping would not increase; consequently, the damping peak is generated. After that, the damping can increase again owing to the contribution of the grain boundary damping and interface damping.