Enhancement of Tensile Properties of AZ91–Ca–Sb Magnesium Alloy with SiC Nanoparticles Additions

Abstract

The present work investigates the effect of SiC nanoparticles (SiC_np) additions on the tensile properties of the squeeze-cast AZ91 + 2.0Ca + 0.3Sb (wt%) alloy at ambient, 423 K and 473 K temperatures. All the AZ91 + 2.0Ca + 0.3Sb + xSiC_np [x = 0.5 wt% (0.29 vol%), 1.0 wt% (0.58 vol%), and 2.0 wt% (1.16 vol%)] nanocomposites illustrate greater yield strength and ultimate tensile strength in contrast to the AZ91 + 2.0Ca + 0.3Sb alloy at all the temperatures employed. The yield strength, ultimate tensile strength and elastic modulus of both the alloy and the nanocomposites decline as the temperature increases, whereas the work to fracture increases with a rise in temperature. Among the nanocomposites, the one comprising of 2.0SiC_np demonstrates the best tensile properties. All the nanocomposites display superior strain hardening response than the alloy, and the maximum strain hardening is perceived in the AZ91 + 2.0Ca + 0.3Sb + 2.0SiC_np nanocomposite. The improved tensile properties of the nanocomposites are ascribed to the reduced grain size, the increase in dislocation density owing to Coefficient of Thermal Expansion (CTE) mismatch between the alloy and the SiC nanoparticles, the Orowan strengthening as well as the presence of a relatively higher amount of Al₂Ca phase in the nanocomposites. The contribution to the improvement of strength of the nanocomposites in decreasing order of their influence are the strengthening due to CTE mismatch, the Hall–Petch strengthening, and Orowan strengthening. The fracture surfaces of the tensile specimens tested at 298 K confirmed the presence of transgranular cleavage fracture which remained unchanged at 473 K as well.

Graphic Abstract

1 Introduction

The automobile industries have shown a rising interest in the magnesium (Mg) alloys because of their admirable blend of high specific strength, exceptional castability, and superior response to damping [1]. Among the conventionally engaged Mg–Al based alloys, the most used AZ91 alloy exhibits impressive mechanical and corrosion features at ambient temperature. Nevertheless, the use of AZ91 alloy is limited in the powertrain owing to a considerable reduction in its tensile and creep properties beyond 403 K [2, 3]. To overcome these difficulties, researchers have made plentiful efforts during the past two decades to ameliorate the mechanical behavior of the AZ91 alloy by introducing calcium (Ca), antimony (Sb), bismuth (Bi), strontium (Sr), silicon (Si) and other alloying elements, in individual or combined form [4, 5]. These elements chemically combine with Mg and Al form intermetallic possessing superior thermal stability. The intermetallic phases are precipitated at the grain boundaries and they block the dislocations as well as prevent the sliding of grain boundaries at elevated temperature. Even though the alloying additions to the AZ91 alloy improved its mechanical properties, however, the modified AZ91-based alloys showed an accelerated decrease in yield strength (YS) with an increase in temperature, which limits their application beyond 473 K [6]. Therefore, developing novel magnesium matrix nanocomposites (MMNCs) by adding nano-size reinforcements have been considered a promising technique to improve the Young’s modulus, strength and elongation of Mg alloys without significantly increasing their density [7, 8].

The addition of nanoparticles SiC, Al₂O₃, ZnO, MgO, TiC, TiB₂, B₄C, and so on enhanced the mechanical behavior of Mg alloys compared to their unreinforced counterpart [9, 10]. Goh et al. [11] concluded that there was an enhancement in YS as well as ultimate tensile strength (UTS) of Mg containing Y₂O₃ nanoparticles [0.5, 1.0 and 2.0 (vol%)], while the elongation remained intact. The improvement in the mechanical response of the AZ31/AZ91 hybrid alloy reinforced with 1.5Al₂O₃ (vol%) nanoparticles was examined by Paramsothy et al. [12]. Jia et al. [13] concluded that the AZ91 alloy with SiC nanoparticles remarkably enhanced both the YS and UTS, and the improvement in YS was 117% in their study. Nie et al. [14] reported that reinforcing 1.0SiC (vol%) nanoparticles into the AZ91 alloy escalated the YS and UTS, whereas, the same decreased for the nanocomposite containing 2.0SiC (vol%). Meenashisundaram et al. [15] stated that by incorporating TiO₂ nanoparticles into pure Mg, the tensile and compression properties were improved than the pure Mg. A substantial enhancement in the tensile and compression behavior of Mg owing to the existence of ZnO nanoparticles was reported by Sankaranarayanan et al. [16]. Khandelwal et al. [17] inspected the mechanical behavior of the AZ31/Al₂O₃ nanocomposites produced by ultrasonic treatment and concluded the superior mechanical properties of the nanocomposites than the unreinforced alloy. He et al. [18] examined the mechanical behavior of the hierarchical Mg–SiC nanocomposites containing 10 and 15 vol% of SiC nanoparticles and observed ductile rupture along with improved malleability. The number of literature investigating the effect of nanoparticles on the elevated temperature tensile response of Mg nanocomposites is limited. Hassan et al. [19] investigated the addition of Al₂O₃ nanoparticles into pure Mg and reported a noteworthy increase in the YS, UTS and ductility at ambient temperature, whereas the YS and UTS decreased as temperature increased to 200 °C. Mallick et al. [20] studied the elevated temperature deformation behavior of the Mg–Y₂O₃ nanocomposites and concluded that the Y₂O₃ nanoparticle addition substantially enhanced the YS and UTS values within the temperature range of 25–250 °C compared to the pure Mg. Cao et al. [21] concluded that addition of 1.0 AlN (wt%) nanoparticles improved the YS and UTS of the AZ91 alloy at high temperature.

Recently, the creep performance of the AZ91 + 2.0Ca + 0.3Sb (wt%) alloy comprising of different mass percentage of SiC nanoparticles was investigated, and it was inferred that the insertion of SiC nanoparticles in the alloy considerably delayed the creep deformation [22]. However, the ambient and elevated temperature tensile behavior of these freshly developed nanocomposites intended for the automotive powertrain applications is not yet reported, which is examined in the present manuscript. Thus, the present work explores the ambient and elevated temperature tensile performance of the AZ91 + 2.0Ca + 0.3Sb alloy with different mass fractions of SiC nanoparticles developed by squeeze-casting. In addition, a detailed analysis of the strain hardening response along with the strengthening mechanisms operated in the nanocomposites is presented.

2 Experimental Methods

The detailed fabrication process of the AZ91 + 2.0Ca + 0.3Sb alloy and AZ91 + 2.0Ca + 0.3Sb + xSiC_np nanocomposites has been discussed in our previous work [22] and are briefly highlighted here. The AZ91 + 2.0Ca + 0.3Sb (AZXY9120) alloy and AZ91 + 2.0Ca + 0.3Sb + xSiC_np [x = 0.5 wt% (0.29 vol%), 1.0 wt% (0.58 vol%), and 2.0 wt% (1.16 vol%)] nanocomposites with nominal compositions of AZ91 + 2.0Ca + 0.3Sb + 0.5SiC_np (NC1), AZ91 + 2.0Ca + 0.3Sb + 1.0SiC_np (NC2), and AZ91 + 2.0Ca + 0.3Sb + 2.0SiC_np (NC3) (wt%) were fabricated using AZ91 ingots (commercially pure), Ca, Sb, and SiC nanoparticles (average diameter 50 nm) (purity for all above 99.9%) in association with a squeeze-casting setup. The AZ91 alloy was melted in a resistance furnace at 1023 K under a non-reactive gaseous environment of Ar and SF₆. After melting, the warm Ca, Sb, and SiC nanoparticles in requisite quantities were integrated to the melt and agitated for 5 min to assure homogenization of the mixture. Finally, the melt was transferred into a steel mold preheated to 523 K, and concurrently a 200 MPa squeeze-pressure was applied. The melt solidified in about 20 s and a squeeze-cast block of 50 mm diameter and 220 mm length was produced.

The AZ91 + 2.0Ca + 0.3Sb alloy and the AZ91 + 2.0Ca + 0.3Sb + xSiC_np nanocomposites were tested under uniaxial tension at ambient and elevated temperatures of 298 K, 423 K, and 473 K. The cylindrical tensile specimens were made from the squeeze-cast alloy and nanocomposites having 20 mm gauge length and 4 mm gauge diameter, according to the ASTM B557M standard. All the tensile tests were executed on a 50 kN UTM (Model: BISS ELECTRA 50). The tests were conducted at a strain rate of 8.334 × 10^–5 s⁻¹ corresponding to a crosshead velocity of 0.1 mm min⁻¹. An extensometer was utilized to evaluate the actual elongation took place at the gauge length, and the elastic modulus was estimated from the linear region of the stress–strain plot. The work to fracture was determined from the entire area within the stress–strain diagram. Three tensile specimens for each composition were examined both at ambient and elevated temperatures, and the average values are reported here as tensile properties. After obtaining the tensile plots, the tensile properties were determined from each plot. Then average of each property was calculated. The tensile curves having the properties close to the average value was plotted as representative tensile curves for the alloy and nanocomposites. The error bars were calculated from the standard deviations from the average values of the tensile properties.

The fractured surfaces of the alloy and the nanocomposites following ambient and elevated temperature tensile tests were successfully characterized employing a scanning electron microscopy (SEM) (Model: JEOL JSM 6480LV) with energy-dispersive X-ray spectroscopy (EDS) (Model: OXFORD INCA X-ACT) attached to it. A transmission electron microscopy (TEM) (Model: JEOL JEM 2100) with EDS (Model: OXFORD INCA X-SIGHT) attached to it was used to understand the dislocation interaction in the nanocomposite specimens following tensile tests. The specimens for TEM analysis were prepared by punching small discs of 3 mm diameter, followed by mechanical thinning to 1 µm thickness. Finally, ion milling (Model: Gatan Duo Mill 600) was employed on the foils.

3 Results and Discussion

3.1 Microstructure of the Alloy and Nanocomposites in Squeeze-Cast Condition

The detailed microstructural analysis of the as-fabricated alloy and the nanocomposites was described in our earlier work [22] and are briefly summarized here. Figure 1a–d portrays the SEM micrographs of the AZXY9120 alloy and its three AZXY9120 + xSiC_np nanocomposites. The alloy constitutes of α-Mg (primary Mg), β-Mg₁₇Al₁₂, reticular Al₂Ca phases and all these phases were also seen in all the nanocomposites. The SiC nanoparticles addition to the alloy decreased the grain size, and the values were 21.42 ± 3.52, 19.08 ± 2.79, and 16.22 ± 1.97 μm for the NC1, NC2 and NC3 nanocomposites, respectively against the 24.52 ± 2.52 μm grain size of the AZXY9120 alloy. Thus, the grain size was the finest in the NC3 nanocomposite. Further, the volume fraction of the β-Mg₁₇Al₁₂ phase calculated using the ImageJ software was 9.45 ± 0.32% in the alloy, and it reduced to 8.18 ± 0.21, 6.78 ± 0.27, 5.43 ± 0.18% in the NC1, NC2 and NC3, respectively following the SiC nanoparticles additions. The corresponding volume fraction of the Al₂Ca phase was 4.49 ± 0.19% in the alloy, and it increased to 5.34 ± 0.24, 6.38 ± 0.27, 7.81 ± 0.1% in the NC1, NC2 and NC3, respectively. Thus, the addition of SiC_np in the alloy reduced the volume fraction of the β-Mg₁₇Al₁₂ phase and increased it for the Al₂Ca phase. The incorporation of SiC nanoparticles produced a large numbers of heterogeneous nucleation sites for the α-Mg phase. In the present investigation, the α-Mg phase is a solid solution of Al, Ca and other elements in Mg. As the number of nucleation sites of the α-Mg phase increased following the incorporation of SiC nanoparticles, a greater amount of Al went into the solid solution. The electronegativity difference between Al and Ca is 0.6 and that between Al and Mg is 0.3. Accordingly, Al has more affinity towards Ca rather than Mg and thus, the remaining free Al readily combined with Ca to form Al₂Ca phase. Subsequently, the amount of Al available for Mg to form the β-Mg₁₇Al₁₂ phase is further reduced resulting in the overall decrease of the volume fraction of the β-Mg₁₇Al₁₂ phase in all the nanocomposites. Nie et al. [14] in their investigation too reported that with the addition of SiC nanoparticles into the AZ91 alloy, the volume fraction of the β-Mg₁₇Al₁₂ phase reduced considerably.

Fig. 1

SEM micrographs of the a AZXY9120 alloy; b NC1, c NC2, and d NC3 nanocomposites [Reprinted with permission from Ref. 22]

3.2 Ambient and Elevated Temperature Tensile Behavior

3.2.1 Yield Strength (YS), Ultimate Tensile Strength (UTS) and Elongation

Figure 2a–c displays the true stress-true strain plots of the AZXY9120 alloy and the AZXY9120 + xSiC_np nanocomposites tested in tension at a temperature of 298, 423, and 473 K. All the curves exhibited an initial elastic region followed by a plastic region. The plastic flow of the alloy and the nanocomposites commenced without a noticeable yield point, as expected. The variation of yield strength (YS) and ultimate tensile strength (UTS) as a function of the SiC nanoparticles content (wt%) at different temperature is portrayed in Fig. 3a, b. It is obvious that all the nanocomposites exhibited the superior YS and UTS at ambient as well as elevated temperature. Further, the YS and UTS increased with the increase in nanoparticles content with the NC3 demonstrating the highest YS and UTS among the nanocomposites. The values of YS and UTS of both the alloy and the nanocomposites declined with a rise in temperature, as expected. The enhancement in YS of the NC3 nanocomposite compared to the alloy was 30% at 298 K, 29% at 423 K, and 22% at 473 K; and that of the UTS was 34% at 298 K, 33% at 423 K and 30% at 473 K. The enhancement in YS and UTS owing to the nanoparticle addition to the Mg matrix has been reported earlier as well [14,15,16,17,18]. Further, all the nanocomposites exhibited the superior elongation than the alloy at all temperatures, as shown in Fig. 3c. Among the nanocomposites, the improvement in elongation of the NC3 was the most pronounced at all the temperatures employed, and it was increased to 45% at 298 K, and 69% at 473 K compared to the alloy. The tensile properties of Mg-based nanocomposites produced by different procedures have been compared with that of the NC3 nanocomposite in Table 1. There is a variation in composition of the matrix alloy, the type of reinforcement employed and its size. However, the squeeze-cast NC3 nanocomposite in the present investigation exhibited the superior YS and UTS compared to the nanocomposites produced by disintegrated melt deposition (DMD) or ultrasonically stir cast (USC) techniques.

Fig. 2

True stress-true strain diagrams of the AZXY9120 alloy and the nanocomposites tested at a 298 K, b 423 K, and c 473 K temperature

Fig. 3

Tensile properties of the AZXY9120 alloy and the nanocomposites at 298 K, 423 K, and 473 K temperature: a yield strength vs. SiC nanoparticle content, b ultimate tensile strength vs. SiC nanoparticle content, c elongation vs. SiC nanoparticle content, d Elastic modulus vs. SiC nanoparticle content, and e work to fracture vs. SiC nanoparticle content

Table 1 Comparison of ambient tensile properties of the AZ91-2.0Ca-0.3Sb-2.0SiC nanocomposite with the reported values for other nanocomposites

The dispersion of the SiC nanoparticles in the alloy resulted in the improved strength of the nanocomposites following several factors. The first factor was the significant reduction in grain size in the nanocomposites following the additions of the nanoparticles in the alloy. The increase in strength due to the reduction of grain size was attributed to the Hall–Petch strengthening. The second factor was the increase in dislocation density because of the Coefficient of Thermal Expansion (CTE) mismatch between the AZXY9120 alloy matrix and SiC nanoparticles. The third factor behind the improvement in strength of the nanocomposite was the Orowan strengthening. The dispersion of the SiC nanoparticles obstructed the moving dislocations. With increased stress, the dislocations crossed the barrier, however, they created loops around the nanoparticles, resulting in reduced interparticle distance. Thus, the stress necessary to move the subsequent dislocations between two nanoparticles increased, which escalated the UTS of the nanocomposites. The detailed contribution to strengthening of the nanocomposites following the grain size reduction, increased dislocation density due to CTE mismatch and Orowan strengthening have been analyzed further and discussed in Sect. 3.2.4. The fourth factor behind the improved YS and UTS of the nanocomposites was the increased amount of the Al₂Ca phase and reduced amount of the β-Mg₁₇Al₁₂ phase. The inferior YS and UTS of the AZXY9120 alloy at ambient and elevated temperatures compared to the nanocomposites were accredited to the relatively coarse grain size and the absence of the strengthening resulting in from the presence of SiC nanoparticles. At elevated temperature, the YS and UTS decreased owing to the easy movement of the dislocations overcoming the obstacles due to the thermal activation. The amount of β-Mg₁₇Al₁₂ phase present was more in the alloy and accordingly, the amount of the Al₂Ca phase was less. The β-Mg₁₇Al₁₂ phase provided lesser resistance to the dislocation motion compared to the Al₂Ca phase at elevated temperature as the former (M. P. 710 K) became soft, and its ability to impede the subsequent dislocations reduced drastically, which lowered the YS of the alloy. On the contrary, at elevated temperature, the nanocomposites displayed greater YS and UTS in contrast to the alloy owing to the higher thermal stability of both the SiC nanoparticles and Al₂Ca phase in the nanocomposites. The Al₂Ca phase (M. P. 1352 K) has better high temperature stability than the β-Mg₁₇Al₁₂ phase and therefore, the former significantly inhibited plastic deformation in contrast to the later.

3.2.2 Elastic Modulus and Work to Fracture

Figure 3d, e displays the variation of elastic modulus and the work to fracture with the SiC nanoparticles content at all temperatures. As the temperature increased, the elastic modulus decreased, whereas the work to fracture increased for both the alloy and nanocomposites, as anticipated. Nevertheless, the nanocomposites illustrated the higher elastic modulus and higher work to fracture in comparison to the alloy at all temperatures. Among the nanocomposites, the highest elastic modulus and the highest work to fracture were revealed by the NC3. The improvement in work to fracture of the NC3 nanocomposite was 95% at 298 K, 205% at 423 K, and 156% at 473 K compared to the alloy. The enhanced elastic modulus of all the nanocomposites was attributed to the presence of SiC nanoparticles having higher elastic modulus (i.e., 450 GPa) compared to that of the alloy (i.e., 42 GPa). Nevertheless, at elevated temperature, the atomic vibration escalated which gave rise to a reduction in elastic modulus. As the temperature was raised, the strain to fracture of both the alloy and nanocomposites increased with a reduction of UTS. The work to fracture ($({\frac{1}{2}}\sigma\epsilon)$) increased with an increase in temperature, suggesting that it was more dependent on the strain than the stress. The trend witnessed in the improvement of elastic modulus, as well as the work to fracture, matches the trend reported in the literature [15, 19, 23].

3.2.3 Strain Hardening Rate and Strain Hardening Exponent

Strain hardening refers to the phenomenon where the flow stress of a material intensifies with an increase in strain, and it is primarily dependent on the dislocation interactions. The strain hardening rate is defined as $\frac{{{\text{d}\upsigma }}}{d\varepsilon }$ where σ and ε stand for the true stress and the true plastic strain. Figure 4a–c portrays the strain hardening rate as a function of effective stress (σ − σ_YS) after yielding in tension at different temperatures, where σ_YS and σ are the yield stress and flow stress from yield stress to UTS. For both the alloy and nanocomposites, the strain hardening rate decreased linearly with an increase in effective stress at all temperatures. Further, the rate of strain hardening of the alloy and nanocomposites was higher at 298 K and decreased with a rise in temperature, as expected. All the nanocomposites exemplified higher strain hardening rate in contrast to the AZXY9120 alloy. Amid the nanocomposites, the NC3 having 2.0SiC_np exhibited the maximum strain hardening rate.

Fig. 4

Typical plots showing the variation of strain hardening rate (dσ/dε) with effective stress (σ–σ_YS) for the AZXY9120 alloy and the nanocomposites tensile tested at a 298 K, b 423 K, and c 473 K

The strain hardening exponent is an estimation of the capability of a material to undergo strain hardening, and it is calculated from the Hollomon’s relationship as follows [24]:

$${\upsigma }_{{{\text{true}}}} = {\text{K}\upvarepsilon}_{{{\text{true}}}}^{{\text{n}}}$$

(1)

where $\sigma_{true}$ and $\varepsilon_{true}$ denote the true stress and true plastic strain, n and K represent the strain hardening exponent and strength coefficient. Taking logarithms of Eq. (1) gives:

$$\log {\upsigma }_{{{\text{true}}}} = \log {\text{K}} + {\text{n}}\log {\upvarepsilon }_{{{\text{true}}}}$$

(2)

Figure 5a–c shows a representative log ($\sigma_{true}$) vs log ($\varepsilon_{true}$) plot for the specimens tested at all the temperatures. The values of n were estimated from the slope of the curves and the values were 0.35 at 298 K, 0.31 at 423 K, and 0.63 at 473 K for the alloy. The values of n vary from 0.36 to 0.46 at 298 K, 0.32 to 0.68 at 423 K, and 0.63 to 0.68 at 473 K for the nanocomposites. Thus, all the nanocomposites showed higher n values contrasting to the unreinforced alloy. The n values of the NC1 and the NC2 nanocomposites lie in the midst of the alloy and the NC3 nanocomposite. Amongst the nanocomposites, the NC3 exhibited the highest n values at all the temperatures.

Fig. 5

Hollomon plots showing the variation of log (True stress) vs. log (True strain) for the AZXY9120 alloy and the nanocomposites tensile tested at a 298 K, b 423 K, and c 473 K

3.2.4 Assessment of Strengthening Mechanisms

The significant improvement in the ambient and elevated temperature YS and UTS of the AZXY9120 + xSiC_np nanocomposites was accredited to the collective contribution of Hall–Petch strengthening, Orowan strengthening, strengthening due to CTE mismatch, and the effect from load-partitioning. The SiC nanoparticles additions to the AZXY9120 alloy substantially reduced the grain size. The increase in strength (${\Delta \upsigma }_{{{\text{Hall}} - {\text{Petch}}}}$) due to grain size refinement as proposed by Hall–Petch is given by [25, 26]:

$${\Delta \upsigma }_{{{\text{Hall}} - {\text{Petch}}}} = {\text{k}}_{{\text{y}}} {\text{d}}_{{\text{m}}}^{{ - \frac{1}{2}}}$$

(3)

where k_y is a material constant and d_m signifies the grain size. It is evident from Eq. 3 that finer is the grain size, higher is the strength of a material. The Orowan strengthening took place as a result of the dislocation loop formation around the reinforced particles, which provided resistance to the subsequent dislocations and increased the rate of strain hardening. The increase in stress due to Orowan strengthening (${\Delta \upsigma }_{{{\text{Orowan}}}}$) is described by the following expression [27, 28]:

$${\Delta \upsigma }_{{{\text{Orowan}}}} = {\text{M}}\frac{{0.4{\text{G}}_{{\text{m}}} {\text{b}}}}{{{\uppi }\overline{{{\lambda }}} \left( {\sqrt {1 - \vartheta_{{\text{m}}} } } \right)}}{\ln}\left( {\frac{{{\overline{\text{d}}}}}{{\text{b}}}} \right)$$

(4)

and $\bar{d} = \sqrt {\frac{2}{3} }d$ where the average orientation factor is M and a value of 3 for M is taken for Mg [29], G_m represents the modulus of shear of the matrix, $\vartheta_{{\text{m}}}$ represents the Poisson’s ratio of the AZ91 alloy, and $\overline{\lambda }$ is the spacing between the particles and is given by ${\overline{\lambda }} = {\overline{\text{d}}}\left( {\sqrt {\frac{{\uppi }}{{4{\text{v}}_{{\text{p}}} }}} - 1} \right).$ The v_p represents the volume fraction of the nanoparticles. There is a noteworthy dissimilarity in the CTE values between the alloy and SiC nanoparticles. The dislocation density on account of CTE mismatch between the AZXY9120 alloy and SiC nanoparticles is theoretically predicted from the following expressions [27, 28]:

$${\uprho }_{{{\text{CTE}}}} = \frac{{12{\text{v}}_{{\text{p}}}\Delta {\upalpha } \cdot \Delta {\text{T}}}}{{{\text{bd}}\left( {1 - {\text{v}}_{{\text{p}}} } \right)}}$$

(5)

where ${\uprho }_{{{\text{CTE}}}}$ represents the dislocation density due to CTE mismatch, Δα signifies the difference in the CTE values, ΔT denotes temperature difference from the processing to test temperature. Figure 6 illustrates the dislocation densities calculated from Eq. (5) for all the nanocomposites at 298 K, 423 K, and 473 K. It is evident that the dislocation densities due to CTE mismatch increased with an increase in nanoparticle additions. Among the nanocomposites, the NC3 displayed the highest dislocation density at all the temperatures. The contribution to the improved strength (Δσ_CTE) of the nanocomposites following CTE mismatch is expressed as [10, 28]:

$$\Delta {\upsigma }_{{{\text{CTE}}}} = \sqrt 3 {\beta G}_{{\text{m}}} {\text{b}}\sqrt {{\uprho }_{{{\text{CTE}}}} } { }$$

(6)

where β is a constant and takes a value of 1.25 for Mg [30, 31]. Considering the nanoparticles were equiaxed, the contribution to the improvement of YS from the load partition between the alloy matrix and the SiC reinforcements is as follows [11]:

$$\Delta {\upsigma }_{{\text{l}}} = 0.5{\text{v}}_{{\text{p}}} {\upsigma }_{{\text{m}}}$$

(7)

Fig. 6

Dislocation densities due to CTE mismatch calculated for the alloy and the nanocomposites at 298 K, 423 K, and 473 K

To predict the YS of the nanocomposites, the contributions from the individual strengthening mechanism such as Orowan strengthening, Hall–Petch strengthening and strengthening due to CTE mismatch were theoretically calculated using Eqs. (3) to (7) and the data are provided in Table 2. Figure 7a depicts the contribution of each mechanism to the improvement of YS for each of the nanocomposite tested at ambient temperature. The contribution to the improvement of YS of the nanocomposites in decreasing order of their influence are the strengthening due to CTE mismatch, Hall–Petch strengthening, Orowan strengthening, and load-partitioning. However, the contribution to YS from the load-partitioning was negligible. The contribution was the most prominent in the NC3 nanocomposite. The contributions from strengthening due to CTE mismatch and the Orowan strengthening were 101% and 114% more in the NC3 compared to that of the NC1 nanocomposite. Further, the relative contribution $\left( {i.e., \, \frac{Contribution\, by \,a \,single\, mechanism}{{Contribution \,by \,all\, the \,mechanisms}} \times 100\% } \right)$ to the improvement of YS of the nanocomposites was calculated. In this regard, the contribution from the load partitioning was excluded owing to its negligible value. The relative contributions from the strengthening due to CTE mismatch, Orowan strengthening, and Hall–Petch strengthening for the NC1 were 42, 19, and 39%; for the NC2 were 46, 22, and 32%; and for the NC3 were 50, 24, and 26%, respectively. Thus, the relative contribution to YS of the nanocomposites from the strengthening due to CTE mismatch as well as the Orowan strengthening increased, whereas the same from the Hall–Petch strengthening reduced compared to the other two mechanisms with an increase in nanoparticles content in the nanocomposites.

Table 2 Parameters used to estimate the contribution of various strengthening mechanisms

Fig. 7

a Contribution of different strengthening mechanisms to the tensile yield strength (TYS) of the nanocomposites tested at ambient temperature, and b comparison of the calculated and the experimentally determined values of TYS of the nanocomposites at ambient temperature

The theoretical YS values of the nanocomposites were further calculated using different strengthening models, and the values were compared with the corresponding experimentally determined one as shown in Fig. 7b. The contribution to strengthening from the ‘Summation model’ ($\Delta \sigma_{Summation}$) is given by [28]:

$$\Delta \sigma_{Summation} = \sigma_{m} +\Delta \sigma_{Hall - Petch} +\Delta \sigma_{Orowan} +\Delta \sigma_{CTE}$$

(8)

where, $\sigma_{m}$ stands for the theoretical YS of the AZXY9120 alloy matrix. It is apparent from Fig. 7b that the values of the YS estimated from the ‘Summation model’ are greater compared to the experimentally determined one. In the ‘Summation model’, the contributions from the various strengthening mechanisms are simply added together to determine the overall strength of the nanocomposite. However, the method is not applicable in most cases, especially when the numbers of factors influencing the strength are more as it neglects the effect of different mechanisms on each other and assumes each factor behaves independently. The method is not applied to many cases because such linear summation predicts higher yield strength compared to the experimentally determined one. Thus, the summation method does not provide accurate prediction when compared to the experimental data. Therefore, the ‘Modified Clyne model’ [32] was adopted as follows in order to obtain the overall strengthening in the nanocomposites.

$${\upsigma }_{{{\text{modified }}\,{\text{Clyne }}}} = {\upsigma }_{{\text{m}}} +\Delta {\upsigma }$$

(9)

$$\Delta {\upsigma } = \sqrt {\left\{ {\left( {\Delta {\upsigma }_{{\text{L}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{Hall}} - {\text{Petch}}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{Orowan}}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{CTE}}}} } \right)^{2} } \right\}}$$

(10)

The modified Clyne method combines the root of the sum of the squares of the individual strengthening effects as shown in Eq. (10). The method was originally suggested for microcomposite and not for nanocomposite. The modified Clyne method assumed that the strengthening mechanisms do not behave independently; instead, the influence of each mechanism was considered together with the other mechanisms. The modified Clyne model i.e., Eq. (10) too overestimated the YS for the nanocomposites. Therefore, Eq. (10) was further modified as follows to suit the experimental results obtained in the present investigation.

$$\Delta {\upsigma } = { }\left\{ {\left( {\Delta {\upsigma }_{{\text{L}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{Hall}} - {\text{Petch}}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{Orowan}}}} } \right)^{2} + \left( {\Delta {\upsigma }_{{{\text{CTE}}}} } \right)^{2} } \right\}^{0.4}$$

(11)

The calculated values following Eq. (11) were closer to the experimental values. It is evident that the improvement in YS of the nanocomposites is precisely equal to the sum of the squares of individual strengthening mechanisms to the power raised to 0.4 instead of 0.5 and its precision to predict the theoretical values of YS is clearly observed in Fig. 7b. The YS values of the nanocomposites was also calculated using ‘Zhang and Chen model’ as follows [27, 28]:

$${\upsigma }_{{\text{c}}} = \left( {1 + 0.5{\text{v}}_{{\text{p}}} } \right)\left( {{\upsigma }_{{\text{m}}} + {\text{A}} + {\text{B}} + \frac{{{\text{AB}}}}{{{\upsigma }_{{\text{m}}} }}} \right)$$

(12)

where,

$${\text{A}} = 1.25{\text{G}}_{{\text{m}}} {\text{b}}\sqrt {\frac{{12\left( {{\text{T}}_{{{\text{Process}}}} - {\text{T}}_{{{\text{Test}}}} } \right)\left( {{\upalpha }_{{\text{m}}} - {\upalpha }_{{\text{p}}} } \right){\text{v}}_{{\text{p}}} }}{{{\text{bd}}_{{\text{p}}} \left( {1 - {\text{v}}_{{\text{p}}} } \right)}}}$$

(13)

$${\text{B}} = { }\frac{{0.13{\text{G}}_{{\text{m}}} {\text{b}}}}{{{\text{d}}\left[ {\left( {\frac{1}{{2{\text{v}}_{{\text{p}}} }}} \right) - 1} \right]}}{\text{ ln}}\frac{{\text{d}}}{{2{\text{b}}}}$$

(14)

The ‘Zhang and Chen model’ considered the effect of load-partitioning, Orowan strengthening, and strengthening due to CTE mismatch only. In this method, they assumed that the strengthening effects were not independent but simultaneously influenced each other and were synergistic in nature [33]. In the present investigation, with an increase in SiC nanoparticles content, the relative contribution to YS of the nanocomposites from the strengthening due to CTE mismatch as well as the Orowan strengthening increased, whereas the contribution from load bearing was negligible. Therefore, the ‘Zhang and Chen model’ estimated the YS of the nanocomposites satisfactorily as seen in Fig. 7b.

3.3 Fractography of Tensile Tested Specimen

Figure 8a, b displays the characteristic fractographs of the AZXY9120 alloy and AZXY9120 + 2.0SiC_np (NC3) nanocomposite tensile tested at 298 K. The fractographs of both the alloy and nanocomposite showed cleavage fracture with secondary cracks present along with the facets suggesting the mode of cleavage failure was transgranular (shown by A). An abundant number of cleavage planes (B) conglomerated to form cleavage steps (C) that in turn coalesced together to form river patterns (D). The features marked by A, B, C, and D are the distinguishing features of transgranular cleavage fracture [34]. The EDS spectra taken from the two cracked regions (shown by A1 and A2) of the fractograph of the AZXY9120 alloy in Fig. 9a constitute of Mg–13.1Al–3.3Ca–0.6Zn and Mg–22.1Al–3.9Ca–0.2Zn (at%) confirming that the cracks initiated at the β-Mg₁₇Al₁₂ phase. Similarly, the EDS spectra taken from the two cracked regions (shown by B1 and B2) of the fractograph of the NC3 nanocomposite in Fig. 9b constitute of Mg–22.1Al–4.9Ca–0.7Zn and Mg–21.8Al–3.8Ca–0.9Zn confirming that the crack initiated at the β-Mg₁₇Al₁₂ phase. The presence of the β-Mg₁₇Al₁₂ phase made the alloy as well as nanocomposites vulnerable to brittle fracture. Under the applied tensile stress, the micro-cracks originated at the β-Mg₁₇Al₁₂ phase and it propagated further with the increase in stress leading to the brittle failure of the alloy and nanocomposites. Patel et al. [35] too concluded from the fracture surface of the AZ91 alloy that the secondary cracks formed along the β-Mg₁₇Al₁₂ phase. The Al₂Ca phase was not present in the cracked regions. Therefore, the β-Mg₁₇Al₁₂ phase largely governed the fracture characteristics of both the alloy and nanocomposites. The appearance of fracture surfaces of the alloy and nanocomposites remain the same at 473 K as depicted in Fig. 8c–d. The dislocation activities play an important role in determining the tensile as well as the fracture behavior of the alloy and nanocomposites. The influence of dislocations on the tensile as well as the fracture behavior is discussed in Sect. 3.4.

Fig. 8

SEM fractographs of the AZXY9120 alloy (a, c) and the NC3 nanocomposite (b, d) tensile tested at (a, b) 298 K and (c, d) 473 K

Fig. 9

SEM micrographs of the fractured surfaces of the a AZXY9120 alloy, and the b NC3 nanocomposite with the results of EDS analysis corresponding to the cracked area shown

3.4 Observation of Dislocations in the Tensile Tested Specimen

Figure 10a depicts the representative bright field (BF) TEM image of the NC3 nanocomposite prior to the tensile test. The micrograph demonstrates the dislocation-free secondary precipitates and SiC nanoparticles throughout the matrix. Figure 10b–d presents the representative TEM micrographs of the NC3 tensile tested at 298 K. A plentiful of dislocation pile-ups were observed in the matrix as represented in Fig. 10b. The presence of the SiC nanoparticles and the Al₂Ca phase restricted the moving dislocations causing dislocations to pile-up surrounding them. Figure 10c illustrates the interaction of the dislocations with the precipitates. It is evident that few dislocations formed loops, some of them were piled-up at the secondary phases, while some of them were bent around the precipitates. These interactions resulted in the reduction of dislocation mobility which as a consequence, increased the dislocation density further in and around the SiC nanoparticles and secondary phases. Figure 10d shows a considerable dislocation pile-ups on one side of the Al₂Ca phase, whereas on the other side of the phase, few dislocation pile-ups are observed. Wang et al. [36, 37] testified the presence of similar numerous dislocation tangles and dislocation pile-ups in the tensile tested specimen of the AZ91 and AM50 alloys. It is well-established that when dislocations pile-up near the secondary phases, the strength of the alloy is increased, however, a high amount of dislocation pile-ups near the secondary precipitates causes an increase in the back stress and aids in the nucleation of micro-crack. With further increase in flow stress, the micro-cracks propagate readily and results in the brittle failure of the materials [34]. The dislocation pile-up present at the vicinity of the β-Mg₁₇Al₁₂ and Al₂Ca phases as well as the SiC nanoparticles as observed in Fig. 10b–d, resulted in micro-crack initiation leading to the brittle (cleavage) failure. Figure 10e–f shows the representative TEM images taken from the NC3 nanocomposite tested at 473 K. The interaction of dislocations with the β-Mg₁₇Al₁₂ phase can be noted in Fig. 10e. The dislocations bypassed as well as cut the β-Mg₁₇Al₁₂ phase at elevated temperature. Figure 10f illustrates several dislocations pile-ups in the neighborhood of the SiC nanoparticles. Goh et al. [11] also reported that the dislocations were trapped by the Y₂O₃ nanoparticles in pure Mg, which obstructed their motion and considerably improved the UTS of the Mg-based nanocomposites.

Fig. 10

Bright field TEM micrographs of the NC3 nanocomposite: a in as-cast condition; tensile tested at 298 K showing the b dislocation pile-ups, c dislocation interactions with intermetallics and d dislocations nearby the Al₂Ca phase; tensile tested at 473 K showing the e dislocation interaction with precipitates and f dislocation pile-ups around the SiC nanoparticles

4 Conclusions

The ambient and elevated temperature tensile performance of the AZ91 + 2.0Ca + 0.3Sb alloy with different mass fractions of SiC nanoparticles developed by squeeze-casting were investigated. A detailed analysis of the strain hardening response along with the strengthening mechanisms operated in the nanocomposites were carried out. The conclusions inferred from the present work are listed below.

• 1. The incorporation of the SiC nanoparticles in the AZ91 + 2.0Ca + 0.3Sb (AZXY9120) alloy improved its elastic modulus, YS, UTS, and the work to fracture. The highest improvement in tensile properties at all the temperatures employed was in the nanocomposite containing 2.0SiC nanoparticles.

• 2. All the nanocomposites exhibited the superior strain hardening response than the AZXY9120 alloy and it was the maximum in the nanocomposite consisting of 2.0SiC nanoparticles.

• 3. The improved strength of the nanocomposites was attributed to the reduction in grain size, the increase in dislocation density owing to CTE mismatch between the alloy and the SiC nanoparticles, the Orowan strengthening as well as the presence of a relatively higher amount of the Al₂Ca phase in the nanocomposites.

• 4. The contribution to the improvement of strength of the nanocomposites in decreasing order of their influence were the strengthening due to CTE mismatch, the Hall–Petch strengthening, and the Orowan strengthening.

• 5. The fractured surfaces of the broken tensile specimens at 298 K confirmed the presence of transgranular cleavage fracture which remained the same at 473 K.