Engineering Materials and Their Applications
Effect of Laves Phases on Creep Strength for a Mg–Al–Ca Alloy
2024 Volume 65 Issue 5 Pages 583-586
Details
2024 Volume 65 Issue 5 Pages 583-586
The effect of Laves phases, C36–(Mg,Al)2Ca and C15–Al2Ca, on high-temperature creep strength was quantitatively investigated for the Mg–5.0Al–1.5Ca alloy produced by die-casting. The homogenization treatment at 750 K for 1 h was carried out to divorce the interconnected skeleton of C36 phase, and the aging treatment at 523 K for 1–1000 h was performed to precipitate the C15 phase within the α-Mg grains. The creep tests to evaluate the creep strength were conducted at 447 K and 70 MPa. When the C36 skeleton is divorced, the minimum creep rate dramatically increases by a factor of 330. The coarsening of the C15 phase within the α-Mg grains increases the creep rate by a factor of 2.6. It was identified that the creep strength of the Mg–5.0Al–1.5Ca alloy is predominantly ascribed to the interconnected skeleton of C36 phase rather than the precipitation strengthening of C15 phase.
This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 88 (2024) 27–30.
Fig. 2 Creep rate vs. time in a log–log diagram at 448 K under a stress of 70 MPa for the Mg–5Al–1.5Ca alloy: as die-cast (open) and after homogenization (solid).
Magnesium alloys are emerging as promising lightweight structural materials, increasingly used as alternatives to steels and aluminum alloys in the transportation industry, encompassing automobiles, aerospace, and railways.1–4) Mg–Al–Ca alloys, noted for their cost-effectiveness and excellent non-flammability, are poised for application in automotive powertrain components operating at high temperatures ranging between 423 and 473 K.5) The creep properties of Mg–Al–Ca alloys at high temperatures have been extensively studied across various production methods,6–9) including die-casting10–16) and squeeze-casting,17) leading to evaluations of their creep strength and mechanisms.18)
The microstructure of the die-cast AX52 alloy, a standard composition in the Mg–Al–Ca alloy family, comprises a primary α-Mg phase with a grain diameter of approximately 5 µm, in which Al and Ca are supersaturated, and an interconnected skeleton of C36–(Mg,Al)2Ca phase18–20) covering the primary α-Mg phase.21) While this C36 interconnected skeleton remains morphologically stable up to 473 K, it divorces over time at temperatures above 473 K.22) Notably, aging treatment of the AX52 die-cast alloy at temperatures between 473 and 623 K results in the precipitation of the C15–Al2Ca phase within the primary α-Mg grains.23) Specifically, after aging at 523 K for 30 h (an over-aged condition), the precipitated C15 phase within the primary α-Mg grains is extremely fine, measuring about 20 nm in length and 1.5 nm in thickness.24) These fine C15 phase precipitates coherently at a high density (precipitation density of 6.1 × 1014 m−2) on the (0001)α plane of the α-Mg matrix.25) The creep strength of Mg–Al–Ca die-cast alloys is predominantly influenced by these two types of Laves phases: C36–(Mg,Al)2Ca and C15–Al2Ca.
The objective of this study is to quantitatively assess the effect of (i) the C36 interconnected skeleton and (ii) the fine C15 precipitates on the high-temperature creep strength of the die-cast Mg–Al–Ca alloy. In our approach, die-cast alloy specimens were initially subjected to homogenization treatment just below the eutectic temperature26–28) of the eutectic reaction (L → α-Mg + C3629)), to divorce the C36 interconnected skeleton covering the primary α-Mg phase. After homogenization, the samples underwent aging treatment at 523 K to induce precipitation of the fine C15 phase within the α-Mg grains. Creep tests were conducted at 448 K on specimens that had undergone both homogenization and aging treatments, to evaluate their high-temperature creep strength. It is important to note that the microstructure of the C36 and C15 phases remains stable morphologically at the creep test temperature of 448 K.
In this study, the AX52 alloy, comprising Mg–4.98Al–1.46Ca–0.34Mn (mass%), henceforth referred to as the Mg–5.0Al–1.5Ca alloy, was utilized. Alloy plates, measuring 50 × 70 × 3 mm3, were produced using a cold chamber die-casting machine under a 1 vol% SF6–99 vol% CO2 atmosphere. The casting and die temperatures were set at 993 K and 473 K, respectively. The die-cast samples underwent homogenization treatment in air at 750 K for 3.6 × 103 s (1 h), followed by water quenching (WQ). Subsequently, the homogenized specimens, referred to as AH (as-homogenized) specimens, were aged at 523 K for durations ranging between 3.6 × 103 and 3.6 × 106 s (1–1000 h). The AH specimens were then mechanically polished using emery paper and colloidal alumina slurry. The microstructure of these specimens was examined using field-emission scanning electron microscopy (FE-SEM).
Hardness measurements were carried out with a micro-Vickers hardness tester. The load was set at 2.94 N with a consistent holding time of 10 s. Seven measurements were taken for both the AH and aged specimens, and the hardness value was determined by averaging five measurements, excluding the maximum and minimum values. Flat specimens for creep tests, with dimensions of 6 × 3 × 28 mm2, were prepared from the as-die-cast, AH, and aged specimens.30,31) The creep tests, conducted in tension, were performed in air at 448 K under a stress of 70 MPa.
Figure 1 presents a FE-SEM image of the AH specimen of the Mg–5.0Al–1.5Ca alloy. The C36 phase, initially forming an interconnected skeleton around the primary α-Mg grains in the as-die-cast specimen, is observed to have transformed into a granular shape following the homogenization treatment (750 K/1 h/WQ). This indicates the dissociation of the interconnected skeleton. Figure 2 illustrates the creep rate over time in a double logarithmic plot for both the as-die-cast and AH specimens, with the creep tests conducted at 448 K and a stress of 70 MPa. For the as-die-cast specimen, the initial creep rate is 2.0 × 10−5 s−1, recorded 15 seconds after stress application. This rate continuously decreases over time, dropping by more than four orders of magnitude in the transient stage, eventually reaching a minimum creep rate ($\dot{\varepsilon }_{\text{min}}$) of 1.5 × 10−9 s−1. The creep test for the as-die-cast specimen was terminated upon the onset of the accelerating creep stage after 3.6 × 106 s (1000 h).
FE-SEM image of the Mg–5Al–1.5Ca alloy after homogenization.
Creep rate vs. time in a log–log diagram at 448 K under a stress of 70 MPa for the Mg–5Al–1.5Ca alloy: as die-cast (open) and after homogenization (solid).
In contrast, the AH specimen exhibited a creep rate of approximately 2.0 × 10−5 s−1 at 30 seconds after stress application, similar to that of the as-die-cast specimen. However, the rate of decrease during the transition stage is less pronounced, reducing by less than two orders of magnitude, with $\dot{\varepsilon }_{\text{min}}$ recorded at 5.0 × 10−7 s−1. This observation, as depicted in Fig. 2, indicates a dramatic increase in $\dot{\varepsilon }_{\text{min}}$ by a factor of 330 when the C36 interconnected skeleton is divorced. Notably, the initial creep rate in the transient creep region remains relatively unchanged.
3.2 Effect of C15–Al2Ca precipitates on creep strengthTo understand the age-hardening behavior of the AH specimen, its hardness was evaluated using a micro-Vickers hardness tester. Figure 3 illustrates the age-hardening behavior of the AH specimen at 523 K. Initially, the hardness of the AH specimen is measured at HV55.3, which is 14% lower than that of the as-die-cast specimen (HV64.032)). This reduction in hardness can be attributed to the dissociation of the C36 interconnected skeleton. At 523 K, hardness begins to increase before reaching 1.0 × 103 s, attaining a peak value of HV59.2 at 3.6 × 104 s (10 h). Subsequently, the hardness gradually declines with extended aging time, registering HV57.0 at 1000 h. In this study, creep tests were conducted on the AH specimens aged at 523 K for periods between 10 and 1000 h. These durations correspond to the peak- and over-aged conditions. The objective was to elucidate the effect of C15 precipitates on creep strength. It should be noted that, under peak- and over-aged conditions, the supersaturation of solute elements (Al, Ca) within the α-Mg matrix grains is depleted.
Plots of Vickers hardness vs. aging time at 523 K for the Mg–5Al–1.5Ca alloy after homogenization. The data for the as-homogenized (AH) specimen is included.
Figure 4 depicts the relationship between creep rate and time at 448 K and 70 MPa for the AH specimens subjected to aging treatment at 523 K for 10, 30, and 1000 h. The creep rate–time curve of each specimen demonstrates a downward curvature from the moment of stress application until creep rupture. Initially, normal transient creep is observed after the application of stress. This is followed by a minimum creep rate, and then a gradual increase in the creep rate occurs in the accelerating region, ultimately leading to the creep rupture of the specimen. For the specimen aged for 10 h, corresponding to the peak-aged condition, $\dot{\varepsilon }_{\text{min}}$ of 6.2 × 10−7 s−1 is recorded at 1.0 × 104 s (2.8 h), and creep rupture occurs at 7.0 × 104 s (19.2 h). The creep rate–time curves for the specimens aged for 30 and 1000 h, representing the over-aged conditions, are similar to that of the 10 h-aged specimen. However, $\dot{\varepsilon }_{\text{min}}$ increases, and the rupture life (trup) decreases with longer aging times at 523 K.
Creep rate vs. time in a log–log diagram at 448 K under a stress of 70 MPa for the Mg–5Al–1.5Ca alloy after homogenization, followed by the aging treatment at 523 K for 10, 30, and 1000 h.
Figure 5 plots $\dot{\varepsilon }_{\text{min}}$ against aging time at 523 K for specimens that underwent creep at 448 K and 70 MPa, including data for the AH specimen. The AH specimen does not contain C15 precipitates within the α-Mg grains, but exhibits maximized solid-solution strengthening due to the supersaturation of solute elements (Al, Ca). For the specimen peak-aged at 523 K for 10 h, the fine C15 precipitates are densely present within the α-Mg grains, maximizing precipitation strengthening. Conversely, the supersaturation of solute elements is depleted, minimizing solid-solution strengthening. In the over-aged condition, the C15 precipitates coarsen over time, reducing the total interfacial energy in the minimized solid-solution strengthening. At a later stage of the over-aged condition (523 K/1000 h), both solid-solution strengthening and precipitation strengthening are minimized due to the coarsening of the C15 precipitates.25) $\dot{\varepsilon }_{\text{min}}$ of the specimens aged at 523 K increases by a factor of 1.5, from 6.2 × 10−7 s−1 to 9.6 × 10−7 s−1, as aging time extends from 10 h to 30 h. Beyond 30 h, the increase in $\dot{\varepsilon }_{\text{min}}$ becomes less pronounced, approaching 1.6 × 10−6 s−1 after 1000 h, when the C15 precipitates are sufficiently coarsened and both solid-solution strengthening and precipitation strengthening are minimized. This variation in $\dot{\varepsilon }_{\text{min}}$ for the specimens aged between 10 and 1000 h at 523 K corresponds to the precipitation strengthening effect of the C15 precipitates. It is inferred that for the Mg–5.0Al–1.5Ca alloy, the coarsening of the C15 precipitates within the α-Mg grains results in an approximate 2.6-fold increase in the creep rate.
Plots of minimum creep rate vs. aging time at 523 K for the Mg–5Al–1.5Ca alloy after homogenization, where the creep tests were carried out at 448 K under a stress of 70 MPa. The data for the as-homogenized (AH) specimen is included.
This study involved a Mg–5.0Al–1.5Ca die-cast alloy, which was subjected to homogenization (750 K/1 h) and aging treatments (523 K/1–1000 h) to manipulate the microstructure of two types of Laves phases (C36–(Mg,Al)2Ca, C15–Al2Ca). Creep tests conducted at 448 K under a stress of 70 MPa aimed to quantitatively assess the effect of these Laves phases on the high-temperature creep strength of the alloy. The key findings are as follows:
This research was supported by the JSPS KAKENHI Grant JP22K04735, Japan. One of the authors (Y. Terada) greatly appreciates the support received from the Light Metal Education Foundation, Japan. The authors would like to thank Prof. Susumu Onaka of Tokyo Institute of Technology for the kind assistance in microstructure observation using electron microscopy.
