Synthesis, microstructure, and mechanical properties of in situ TiB₂/Al-4.5Cu composites

3.1 Microstructure characterization

Figure 2 shows the X-ray diffraction (XRD) patterns of the in situ TiB₂/Al-4.5Cu composites with peaks corresponding to the Al, TiB₂, and CuAl₂. The intensity of TiB₂ peaks increases with the increase in the TiB₂ content in the composites.

Figure 2:

XRD diffraction patterns of the in situ TiB₂/Al–4.5Cu composites.

Figure 3A~B shows the SEM micrographs of the as-cast 1.7% TiB₂/Al-4.5Cu composite and 7.2% TiB₂/Al–4.5Cu composite, respectively. It can be observed that a mass of continuous CuAl₂ phases distribute along the grain boundaries, and the continuous CuAl₂ phases transformed into discontinuous phases gradually with the increasing TiB₂ content. From the higher magnification microstructure at the top right corner of Figure 3, it can be seen that most of the TiB₂ particles distribute along the grain boundaries and interweave with CuAl₂ phases together. Figure 3D~E shows the element area scanning by EDS of 7.2% TiB₂/Al-4.5Cu composite. Most of the Ti and Cu element are distributed in the grain boundaries, which can further confirm the distribution characteristic of the TiB₂ particles. This is different from the previous work that reported that the TiB₂ particles could uniformly distribute in α-Al [9], [18]. The primary reason for this is that the ratio of the thermal conductivity coefficient between the particles and matrix is <1. According to the theory of thermal conductivity, during solidification, the TiB₂ particles will be easily pushed onto the solidifying interface [19]. Agglomeration of TiB₂ particles has also been observed in the composites, which might affect the mechanical properties adversely.

Figure 3:

SEM micrographs of TiB₂/Al composites. (A) 1.7TiB₂/Al-4.5Cu; (B) 7.2TiB₂/Al-4.5 Cu, and surface distribution of the element of Al (C), Cu (D), and Ti (E) in 1.7TiB₂/Al-4.5Cu composite.

Figure 4 shows SEM at higher magnification and TEM micrograph of as-cast 7.2.vol.% TiB₂/Al-4.5Cu composites, which presents the typical morphology and size distribution of TiB₂ particles. The in situ formed TiB₂ particles show a mixture of different shapes such as spherical, hexagonal, and cubic, which is in accordance with the previous work in the literatures [19], [20]. It can also be seen in Figure 4B that the interface between the TiB₂ particles and the matrix is clean and clear. The well-bonded interface can be attributed to the in situ formation process within the Al melt. Besides, some θ′ phases around the TiB₂ particles can be seen. The θ′ phases may be precipitated during the casting cooling process because the distortion field around the TiB₂ particles could promote the precipitation of an unstable phase. During the reaction process, the in situ formed TiB₂ particles were protected from the oxidation or other contamination. Meanwhile, because of the thermodynamic stability of the TiB₂ particle, it avoided producing undesirable compounds between the particle and the matrix during the heat-preservation process [21], [22]. The contamination of reinforcements would lead to the poor wettability between the reinforcements and the matrix [7]. So, the in situ formed uncontaminated particles are good at preferable wettability and interfacial integrity between the aluminum matrix and the TiB₂ particle. Figure 4C reveals that the in situ formed TiB₂ particles in the composite are between 60 and 400 nm in size. The average particle size dp is 200 nm using the Image pro plus 6.0 software, much lower than that of the ex situ particles in discontinuously reinforced composites [6].

Figure 4:

The typical (A) SEM (BSE) micrographs of the TiB₂ particle; (B) TEM micrographs of the TiB₂ particle; (C) size distribution of TiB₂ particles in TiB₂/Al-4.5Cu composites at as-cast state.

Figure 5 shows the effect of TiB₂ content on grain structure and size of as-cast xTiB₂/Al-4.5Cu composites with a different particle content (x=0, 1.7, 3.6, and 7.2 vol.%). Figure 5A shows large equiaxed grains with coarsening dendritic substructure and great branch distance. Continuous phases were observed at the grain boundaries. The grain size of the casting base alloy was about 205.0 μm. Figure 5B~D illustrates that average grain size of the as-cast composite decreases with increasing particle content, which means that the TiB₂ particles have significant effect on refining grain size of as-cast composites. When the content of the TiB₂ particles reaches 7.2 vol.%, all the grains of the as-cast composite are almost fine and equiaxed. Figure 5E shows the relationship between the average grain size of the TiB₂/Al-4.5Cu composites and particle content. As the amount of TiB₂ particles increases, the grain size reduces. The grain size reaches as low as 40.8 μm with 7.2 vol.% TiB₂. Based on the published literature [23], [24], the high grain-refining effect is mainly due to the TiB₂ particles nucleating in the composites. However, it should be accepted that any nucleus acting as nucleating agents should be located at the center of the grain. In this work, most TiB₂ crystals are pushed into the grain boundaries and distribute in the intergranular region, indicating that they do not nucleate α-Al alone. The main reason for the refined grain size may be attributed to the influence of particles pinning on grain boundaries, which effectively limits grain growth [1], [16].

Figure 5:

Effect of TiB₂ content on grain structure and size of (A) Al-4.5Cu; (B) 1.7TiB₂/Al-4.5Cu; (C) 3.6TiB₂/Al-4.5Cu; (D) 7.2TiB₂/Al-4.5Cu composites; (E) Shows the average grain size of composites.

3.2 Mechanical properties and strengthening mechanisms

Table 1 and Figure 6 present the tensile properties [Young’s modulus, yield strength (R_0.2) at 0.2% strain, ultimate tensile strength (R_m), and El% (elongation)] of Al-4.5Cu base alloy and TiB₂/Al-4.5Cu composites at room temperature.

Figure 6:

Effect of TiB₂ content on the tensile mechanical properties.

In comparison to the base alloy, the in situ composites showed a significant improvement in the yield strength and ultimate tensile strength, while the elongation reduced with the increase in TiB₂ particle content. It is accepted that the increase in yield strength by incorporating TiB₂ is commonly attributed to the following three mechanisms: grain refinement strengthening [14], [23], Orowan strengthening [12], [25], and coefficient of thermal expansion (CTE) mismatch strengthening [13]. For grain refinement strengthening, the effect of grain size d on the mechanical properties of a composite can be commonly described by the Hall-Petch relationship [26], [27]:

where, the parameter σ₀ represents the yield strength of a single crystal in the absence of any strengthening mechanisms except the solid solution effect, and k_y is the magnitude by which crystal boundaries in a polycrystalline material resist slip. In this experiment, taking values of ky=0.04 MPam for Al and Al alloy [28]. In cases where the addition of particles reduces the size of the grains in the particulate-reinforced AMCs, the yield strength improvement due to grain refinement can be described by the following equation [29]:

where, d_MMC and d_m are the average grain diameters of polycrystalline matrix materials in the composite and unreinforced base alloy. This equation assumes that the Hall-Petch parameters k_y and σ_o remain unchanged in the composites during processing.

Figure 4 illustrates that the TiB₂ size is less than 1 μm; so, the Orowan strengthening should not be neglected in estimating the yield strength. The particulate-dislocation interaction by means of the Orowan bowing mechanism in the matrix can increase the strength. The yield strength contributed from the Orowan strengthening is given as in the following equations [13]:

where, G_m is the shear modulus of the matrix; b is the Burgers vector of the dislocations, b=0.286 nm; ν is the Poisson’s ratio, E is the elastic modulus of the matrix, E(Al-4.5Cu)=66.3 Gpa; λ is the interparticle spacing; d_p is the particle diameter, d_p=200 nm; V_p is the volume fraction of reinforced particles.

Because of the difference in the coefficient of thermal expansion (CTE) and the elastic modulus (Ec) mismatch between the matrix and reinforced-particle particles, enormous dislocations are produced as the result of residual plastic strain. CTE mismatch strengthening relates the contribution of dislocation density to the strength of the material. The effect of thermal mismatch on the strength of the composite is given [12], [13]:

where β is a constant and approximately equal to 1.25. ρ_CTE is the dislocation density induced by the CTE mismatch. Δα is the difference between the CTE of the matrix and reinforced particle, α_Al=24.6×10⁻⁶; αTiB2=7.8×10−6. ΔT is the difference between the processing and test temperatures, T_test=298 K; T_process=1123 K.

There are a lot of modeling methods to predict the yield strength of particle-reinforced aluminum composites. The arithmetic summation method proposed by relevant scholars [14] is one of the simplest and easily achieved prediction modeling. Obviously, it is assumed that different mechanisms do not influence each other, and they can independently contribute to the final yield strength of composites. According to the arithmetic summation method, the yield strength of the TiB₂/Al-Cu composite in this study could be predicted by the following equation [14]:

Figure 7 shows the experimental and predicted yield strength according to the arithmetic summation method. The parameters used in the predicted calculation of the yield strength are shown in Table 2. The predicted values using the arithmetic summation method are much higher than the experimental values at the as-cast state. The high discrepancies are mainly due to the micrometric agglomerate of TiB₂ particulates on the grain boundaries. As seen from the SEM microstructure in Figure 3, the TiB₂ particles are not in ideal uniform distribution among the composites. There must be less effective TiB₂ particles introducing the strengthening effect than the nominal TiB₂ particle content used in the theoretic calculation. Resultantly, the predicted values of yield strength should be overestimated in TiB₂/Al composites. In fact, Orowan strengthening, CTE mismatch strengthening, and Hall-Petch strengthening are closely linked to the distribution of the reinforced particles [13]. The Orowan mechanism and CTE mismatch strengthening just act on effective particulates in the grain; the micrometric agglomerate particulates on the grain boundaries should be excluded in the calculation [16]. So, the actual interparticle distance was larger than the theoretical distance λ calculated from equation (8). In particle pushing, reinforced particles are pushed ahead of the solidification front and can act to restrict grain growth. The particles pushed to grain boundaries only can contribute to the grain refinement [15]. So, the fraction in the grain and on the grain boundaries must have a great effect on the strength of the composites.

Figure 7:

Theoretically calculated yield strength using the arithmetic summation method.

In order to consider the effect of the particle distribution on the yield strength, it is assumed that the proportion of the particle content in grain and on grain boundaries is n/m. Then, the yield strength of the composites can be expressed as equation (12):

Figure 8A shows the comparison between the yield strength prediction using a different ratio of n:m with the experimental data in this study. As we can see, the yield strength prediction of the composites decreases with the decrease in the ratio of n to m experimental data. When all the particles distribute in the grain, the yield strength reaches a maximum value, the main strengthening mechanisms are Orowan strengthening and CTE mismatch strengthening. When n/m is zero, all the particles distribute on the grain boundaries, the strengthening effect is lowest, only the Hall-Petch effect contributes to the yield strength. The multistrengthening mechanisms work when the particles are distributed both in the grain and on the grain boundaries. The strengthening effect weakens as the proportion of particles inside the grain decreases. Compared to the experimental value in the study, it can be found that the predicted values agree well with the experimental value when n/m is 1/50. It means that the proportion of the particle content in grain and on grain boundaries is 1/50, which confirms the same conclusion in Figure 3 that most of the TiB₂ particles distribute along the grain boundaries.

Figure 8:

The comparison of yield strength prediction by the modified model with experimental data for TiB₂/Al alloy composites. (A) in this study; (B) deriving from reference.

Figure 8B shows a comparison between the predicted results using the appropriate ratio of n:m and the experiment results of yield strength for TiB₂/Al composites deriving from reference, which further verify the accuracy of the modified method. In these literatures, several types of in situ TiB₂/Al alloy composites are included, such as TiB₂/Al-7Si [30] composites, TiB₂/Al-4Cu [24] composites, TiB₂/A356 composites [8], and TiB₂/7055Al composites [31]. It is easy to see that the predicted results show good agreement with the experimental values by choosing the suitable value of the proportion of the particle content on grain boundaries and in grain. Generally, it is indicated that the phenomenological model can give a fairly precise estimation for the yield stress of the in situ TiB₂/Al alloy composites.