The macrostructural examination of the Mg-5Nb-SiC_n as-cast billets and extruded rods indicated that the surface of the materials were smooth and did not show any presence of macropores, blowholes and/or shrinkage cavities. The absence of macro-defects in the samples indicated the suitability of the processing parameters that result in sound, defect-free castings.

Table 1 shows the results of the density and porosity measurements made on the extruded sections. From the table, it is seen that the density values increase with increasing content of SiC_n, which is due to the higher density of SiC (3.21 g/cc) when compared to that of pure Mg. It can also be observed that the experimentally measured density values of the Mg-5Nb-SiC_n composites are comparable to those of their theoretically estimated values. This indicates very low levels of porosity (<0.1 %) and is in agreement with macroscopic observations.

Table 2 and Figure 1 show the results of the microstructural investigations conducted on the extruded materials. It can be observed that when compared to pure Mg, fine and near-equiaxed grains are observed in all the materials, with the average grain size <10 μm (Table 1, Figure 1a–d). The observed refined microstructure is due to the presence of reinforcement particles (both Nb and SiC_n), which act as heterogeneous sites for grain nucleation [15]. Further, as the hot extrusion was carried out at 350 °C (>0.5 T_m), it results in recrystallization, thereby leading to the formation of fine and nearly-equiaxed grains [9]. It can also be seen from Table 1 that when compared to Mg-5Nb, the additions of SiC_n up to 0.27 V_f did not contribute to grain size reduction. On the contrary, for V_f > 0.27% grain refinement can be observed. This indicates that at these SiC_n contents, the nanoparticles act as an obstacle for grain growth [9], thus resulting in a fine-grained structure.

Figure 2 shows the distribution of Nb and SiC_n particulates in the Mg-matrix. Microstructural observations indicated that the Nb particles were distributed uniformly throughout the matrix, while higher magnification images showed clustering of Nb particles that were present in all the composites. Representative images showing these features are given in Figure 2a, b. It should be noted that there are several factors that can contribute to particle clustering (often referred to as “particle local inhomogenity” [16], that include: (i) processing parameters such as insufficient melt stirring time, low melt temperature or low solidification rate; (ii) poor wettability between the particle/matrix; (iii) irregular shape of the particle and (iv) chemical incompatibility between the particle/matrix [16,17,18].

In the current work, standardized process parameters were employed. Further, the DMD technique has a relatively high solidification rate when compared to conventional die casting processes. Regarding the particle shape, the Nb-particles were dominantly cuboidal in shape of varying sizes and in the as-received condition, the relatively smaller-sized particles are found to be clustered (arrow, Figure 2c). While the wettability value of Mg/Nb is not known, the high magnification image shown in Figure 2d shows good interface bonding (arrow), indicating good wettability. However, it should be noted that based on the Mg-Nb phase diagram [20], Nb and Mg has no solubility (mutually insoluble) and does not form any intermetallic phase. Hence, considering the relatively high density of Nb (8.57 g/cc) and the lack of chemical compatibility (i.e., absence of chemical bonding due to no solubility) between Nb-particles and Mg-matrix, the Nb particles will tend to agglomerate during processing, thereby forming Nb-clusters. Earlier works on Mg-Ti and AZ31-Cr also showed similar agglomerations containing Ti and Cr [19,21]. In these works, Ti showed good wettability with the Mg-matrix, while the wettability of Cr with Mg was not known. However, not only was there no solubility between the Ti/Cr particles and Mg, but also that the particles were irregular in shape, such that it resulted in particle segregation/clustering [19,21].

For the distribution of SiC_n particles, EDX analysis was conducted on the polished samples. From Figure 3 that shows the x-ray mapping of Si, C, Nb and Mg, it can be seen that while the SiC_n particles were uniformly distributed in the matrix, they were also seen to be preferentially located on the Nb-clusters (arrows). These observations suggest that the distribution of the micron-particles in the Mg-matrix strongly control the dispersion of the nano-reinforcements.

Figure 4 shows the x-ray diffraction patterns of the developed Mg-5Nb-SiC_n composites. It can be seen that in all the composites, Mg and Nb peaks were prominent. As mentioned above, no secondary phase formation occurs between Mg and Nb due to the lack of mutual solubility. While SiC peaks could not be identified due to its very low volume fraction, low intensity peak corresponding to Mg₂Si has been observed. It is well-known that the interaction of Mg and SiC results in the instantaneous formation of Mg₂Si and has been reported in Mg-based micron SiC reinforced composites [15].

The mechanical properties of the developed Mg-5Nb-SiC_n composites with varying SiC_n volume fractions are shown in Table 3 and Table 4. The properties are also compared with those of pure Mg and Mg-5Nb metal–metal composite. From Table 3, it can be seen that the addition of SiC_n increases the hardness significantly. The hardness increases with increasing V_f of SiC_n such that at 1.10% V_f, the value is almost 2.5 times more than that of pure Mg and Mg-5Nb.

From Table 3 it can be seen that unlike the hardness, the strength properties do not increase with increase in SiC_n content. Rather, for V_f ≤ 0.27%, there is a decrease in strength properties. For V_f > 0.27%, there is a significant improvement in both the yield strength and tensile strength, however at the expense of ductility. The Mg-5Nb-1.0 SiC_n (0.55% V_f) composite shows the highest values of strength amongst all the materials, with ~50% increase in yield strength and ~33% increase in tensile strength. Comparatively, the increment in strength in Mg-5Nb-2.0 SiC_n (1.1% V_f) composite is slightly lower. However, among these two composites, 1.1% V_f shows improved ductility.

Table 4 shows the results of the compression tests. It can be seen from the table that, similar to that observed under tension, no improvement in yield strength occurs until ~0.27% V_f and the values are similar or lower than that of pure Mg and Mg-5Nb. In comparison, for composites with 0.55% V_f and 1.1% V_f of SiC_n, an increase of ~25% and ~30% in compressive yield strength can be observed. Further, for SiC_n V_f ≥ 0.27%, nominal increase in the ultimate compressive strength values are also seen, with little variation in the compressive ductility.

Figure 5 shows the representative engineering stress-strain curves of pure Mg, Mg-5Nb and Mg-5Nb-SiC_n composites under tensile and compressive loading conditions.

The fracture surface analyses of the tensile and compressive tested samples were conducted and representative images are shown in Figure 6a–h. Under tensile loading, pure Mg exhibits dominant cleavage fracture due to its h.c.p. structure [19], whereas dimple-like ductile features were reported in Mg-5Nb [8]. In the Mg-5Nb-SiC_n composites, dominant cleavage fracture with brittle morphology is seen (Figure 6a, c). Regions showing dimple-like ductile features were also observed, particularly in Mg-5Nb-2.0 SiC_n composite (Figure 6c). The brittle mode of fracture is largely due to the presence of hard SiC particles, Nb/SiC_n clusters and the formation of Mg₂Si, which is a brittle intermetallic. Further, the good interfacial bonding between Mg/(Nb/SiC_n) contributes only to mechanical bonding, but has low shear strength across the interface due to the absence of chemical bonding (no intermetallic phase formation) [21,22]. This would result in particle debonding due to increasing applied loads (Figure 6d). While a crack extending into the matrix near the Nb/SiC_n cluster is seen in Figure 6e, f shows an Nb-particle that is perfectly embedded in the matrix. However, voids coalescing around the particle (arrow) can also be seen, which would eventually result in debonding and/or matrix cracking. These consequently lead to low fracture strains (Table 3). Under compressive loads, fracture occurs by dominant shear fracture in all the materials (Figure 6g, h).

Based on the experimental results, it is observed that the addition of nano-sized SiC particles have significantly altered the mechanical properties of Mg-5Nb. The effect of SiC_n volume fraction, processing, distribution of metallic and ceramic reinforcements and their inherent properties in influencing the mechanical behavior of the composites are discussed in the following sections.

Mg-5Nb (Mg-turnings, Acros Organics, Singapore; Nb-powder of average particle size <10 μm, Prochem, Rockford Illinois) were used as the base materials to produce Mg-5Nb metal–metal composite. The Mg-5Nb composite was reinforced with 0.25 to 2 wt.% nano-SiC particles (average particle size ~50 nm, Nanostructured and Amorphous, US). The disintegrated melt deposition (DMD) technique which employs bottom pouring of the melt was used to produce the Mg-nanocomposites. The DMD technique adopts the simultaneous vortex stirring of the melt and its disintegration, with argon (Ar) gas for inert atmosphere and for melt disintegration, resulting in a high yield process which offers finer grain size [9]. The Ar gas was purged into the Mg-5Nb-nSiC molten metal at ~3 L/min at a super-heat temperature of ~750 °C, with intermittent stirring for 5 min. The melt was released through an orifice of diameter ~10 mm, situated at the bottom of the crucible and was disintegrated with two Ar gas jets maintained at ~25 L/min (oriented in the direction normal to the melt stream) before its deposition onto the steel mould.

Cylindrical billets of 36 mm diameter cut from the deposited ingots and soaked at ~400 °C for 1 h, were extruded at ~350 °C using a 150 ton hydraulic press at an extrusion ratio of 20.25:1 to produce rods of 8 mm diameter that were used for further characterizations.

Density measurements were performed on the extruded samples using the Archimedean Principle [2]. Microstructural characterization of finely polished and etched samples taken from the extruded rods were conducted using optical microscope (Olympus) and field emission scanning electron microscope (Hitachi FESEM-S4300) coupled with energy dispersion analysis (EDS). The grain size of the developed materials was determined using the Scion image analysis software. X-ray diffraction analyses were conducted on the samples using an automated Shimadzu LAB-XRD-6000 x-ray diffractometer (Cu, Kα; λ = 1.54056 Ǻ), with a scanning speed of ~2 min⁻¹. Phase identification was carried out by matching the Bragg angle and the intensity of the peaks with the standard peaks of Mg, Al and other related phases.

Microhardness measurements were performed on polished flat specimens using a Shimadzu HMV automatic digital microhardness tester with a Vickers indenter (load: ~25 gf and dwell time: ~15 sec). More than ten measurements were performed for each specimen. Uniaxial tensile tests and compression tests were performed at room temperature using a Materials Test System (MTS 810). Tensile tests were conducted on smooth bar tensile specimens (ASTM Standard E8M-08) with diameter ~5 mm and gauge length ~25 mm, at a constant strain rate of ~1.69 × 10⁻⁴ s⁻¹. A strain gauge (MTS extensometer) was used to measure the strain during the test. Compression tests were conducted on cylindrical samples of length ~8 mm and diameter ~8 mm (aspect ratio, l/d = 1) in accordance with ASTM E9-89a guidelines, at a constant strain rate of ~8.3 × 10⁻⁴ s⁻¹. A minimum of 5 tests were conducted for each composition to obtain repeatable values. Fractographic analysis was performed using Hitachi FESEM-S4300.