The Effects of Pre-Consolidation Heat Treatment on the Tensile and Fracture Toughness Behavior of the Rapidly Solidified Mg–Zn–Y–Al Alloys

Abstract

This study is aimed to optimize the tensile properties and fracture toughness of rapidly solidified (RS) ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al (at%) alloys by changing the pre-consolidation heat-treatment temperature. The alloys prepared from RS ribbons heat-treated below 673 K consisting of bimodal α-Mg grains; coarse-worked grains (∼2.8 µm) with high Kernel average misorientation (KAM) values (∼1.8°), and ultrafine dynamically recrystallized (DRX) grains (∼0.68 µm) with intermediate KAM values (∼1.1°). The DRX grains involve cluster arranged layers (CALs) and thin plate-shaped long-period stacking ordered (LPSO) phase precipitation. The fine grain structure strengthens the alloy, significantly while they have little beneficial effects in retarding the crack propagation, thus resulting in low fracture toughness. The alloys that were heat-treated above 723 K prior to consolidation possessed a three kind of α-Mg grain structure; it consisted of fine DRX grains (∼2.3 µm) with low KAM values (∼0.5°) in addition to the coarse-worked grains and ultrafine DRX grains. The fine DRX grains improved strain hardening and ductility, resulting in fracture toughness increase. Furthermore, high-temperature pre-consolidation heat treatment produces block-shaped LPSO phase grains associated with α-Mg. This LPSO phase appears to work as an effective feature to toughen the materials, due to the frequent formation of microcracks in the phase and promotion of crack deflection.

1. Introduction

Mg alloys are attractive structural materials for use in transportation vehicles, including for automotive and aerospace applications, owing to their low densities.¹⁾ Numerous types of Mg alloys have been developed for casting and/or die casting product forms.^2,3) Recently, casting and wrought alloys have been actively developed.^4,5) In 2001, high-strength Mg–1Zn–2Y (at%) alloys featuring a novel long-period stacking ordered (LPSO) structure were developed by rapidly solidified powder metallurgy (RS P/M) processing.⁶⁾ The excellent properties of the RS P/M Mg–1Zn–2Y alloy are attributable to the grain refinement of the α-Mg matrix grains and the high dispersion of the LPSO phases in the matrix grains. More recently, the RS ribbon consolidation method, which is significantly safer than the RS powder-consolidation method, has been developed.^7–11) Additional to the improvement in RS ribbon consolidation processing, the optimization of alloy compositions suitable for the process has also been performed. Mg–Zn–Y–Al has been reported to exhibit excellent corrosion resistance, high mechanical properties, and high fracture toughness.^12–14) Therefore, the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al (at%) alloys could be excellent material choice for aerostructures such as stringers and brackets.¹⁵⁾ In structural design, plane-strain fracture toughness is an important property requirement. In our previous study, we demonstrated that a pre-consolidation heat treatment of the RS ribbons was effective in improving the ductility and fracture toughness of the Mg–0.85Zn–2.05Y–0.35Al alloy.^13,14) However, the influence of the pre-consolidation heat treatment on the multimodal microstructure evolution and toughening mechanisms has not yet been elucidated, particularly in terms of the heat-treatment temperature dependence. Therefore, this study attempts to optimize the pre-consolidation heat-treatment temperature to enhance fracture toughness and to understand the mechanism. The changes in the tensile and fracture toughness properties of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys are expressed as a function of the pre-consolidation heat-treatment temperature, and the relationship between the multimodal microstructure evolution and fracture toughness is discussed.

2. Experimental Procedure

A master alloy ingot with a nominal composition of Mg–0.85Zn–2.05Y–0.35Al (at%) was prepared via high-frequency induction heating in an argon atmosphere. The RS ribbons were prepared using a single-roller melt-spinning method with a roll-circumferential velocity of 42 m·s⁻¹. The cooling rate at this roll-circumferential velocity was estimated at 1.4 × 10⁵ K·s⁻¹.⁸⁾ The RS ribbons were compacted into copper billets and degassed for 15 min at 523 K. Before extrusion for consolidation, the billets were heat-treated at 473 K, 623 K, 673 K, 723 K, and 738 K for 24 h. The non-heat-treated and head-treated billets were extruded at an extrusion ratio of 10 at 623 K and an extrusion ram speed of 2.5 mm·s⁻¹. Hereafter, the samples prepared from non-heat-treated ribbons are designated as NHT, and the samples prepared from the ribbons heat-treated at 473 K, 623 K, 673 K, 723 K, and 738 K are designated as HT473K, HT623K, HT673K, HT723K, and HT738K, respectively. The details of the RS ribbon consolidation process have been described in previous reports.^8,9)

The mechanical properties of the RS ribbon-consolidated alloys were evaluated using tensile and fracture toughness tests. Tensile tests were performed using an Instron testing machine (Instron Model 5584) at room temperature at an initial strain rate of 5 × 10⁻⁴ s⁻¹. The gauge sections of the tensile specimens were 2.5 mm and 15 mm in diameter and length, respectively. The tensile axis was placed along the extrusion direction. The 0.2% proof strength was used as the yield strength. Mode-I fracture toughness tests were conducted according to the ASTM E399-12^ε3 standard¹⁶⁾ using a servo-hydraulic mechanical testing machine (Shimadzu EHF-EM). Compact tension (CT) specimens with thickness B = 4.2 mm and width W = 8.4 mm were used in this study. The V-notch was normal to the extrusion direction. Before the plane-strain fracture-toughness test, fatigue cracking tests were performed under load control in tension using a maximum stress intensity of 2.3–3.2 MPa·m^1/2 at a frequency of 10 Hz with a load ratio of R = 0.1, to introduce a fatigue pre-crack in the specimens. At this time, the pre-crack length, a, has to be satisfied 0.45 ≦ a/W ≦ 0.55, as required by the standard for valid K_IC. Using pre-cracked specimens, the plane-strain fracture toughness test was performed at a constant displacement rate of 0.012 mm·s⁻¹ (∼0.55 MPa·m^1/2·s⁻¹) at room temperature. The fracture toughness test method has been described previously.¹⁴⁾

The microstructure was characterized by scanning electron microscopy (SEM; JEOL JSM-4601F) and transmission electron microscopy (TEM; JEOL JEM-2100F). The SEM samples were polished using a cross-sectional polishing (JEOL SM-09010). The TEM samples were prepared by ion milling (GATAN PIPS M-691) and a focused-ion beam (FIB) system (FEI Versa 3D). The crystallographic orientations of the alloys were analyzed using electron backscatter diffraction (EBSD) analysis using orientation imaging microscopy (OIM; TSL Solutions K.K.). The EBSD patterns were acquired with a step size of 0.1 µm. TSL OIM analysis software was used for EBSD data analysis: evaluation of the grain size and internal stain distribution was performed via Kernel average misorientation (KAM) mapping. In this study, the KAM values were calculated as the average misorientation between the data and the third-neighbor points. Fractography was conducted using a confocal optical microscopy (OM; Lasertec C-130), SEM, and TEM.

3. Results

3.1 Tensile and fracture toughness properties

Figure 1 shows the tensile properties of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloy. The typical tensile true stress–true strain curves and strain hardening rate (dσ_t/dε_t) curves are shown in Fig. 1(a). The true stress, σ_t, and true strain, ε_t, were calculated from the engineering stress, σ₀, and engineering strain, ε₀, using the following equations:   

\begin{equation} \sigma_{\text{t}} = \sigma_{0} (1 + \varepsilon_{0}), \end{equation} (1)

  

\begin{equation} \varepsilon_{\text{t}} = \ln (1 + \varepsilon_{0}). \end{equation} (2)

The NHT and HT623K specimens exhibit the highest yield strength of ∼500 MPa and a moderately small elongation of 5%. The yield strength of the HT673K specimen slightly decreases to 480 MPa. The HT723K and HT738K specimens show lower yield strength levels of 400–430 MPa, but large elongations of over 10%. The intersection between the true stress–true strain curve and strain hardening rate curve is the Consideré criterion, or the plastic instability, as shown in the following equations:¹⁷⁾   

\begin{equation} \sigma_{t} \geq \frac{d\sigma_{t}}{d\varepsilon_{t}}. \end{equation} (3)

For the NHT, HT623K, and HT673K specimens, plastic instability was observed nearly immediately after yielding. This result indicates that these three alloys have highly strain-hardened characteristics, where no further strain hardening occurred. The HT723K specimen also exhibited plastic instability immediately after yielding, but the work-hardening rates first decreased to ∼280 MPa, then increased to ∼360 MPa, and then gradually decreased once again. In contrast, the HT738K specimen showed high work-hardening rates, consequently resulting in the plastic instability to have only occurred way passed the yielding points. Figure 1(b) shows the changes in the tensile yield strength and elongation of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys as a function of the pre-consolidation heat-treatment temperature. The alloys heat-treated below 623 K maintains high yield strength values of ∼500 MPa, but those of the alloys heat-treated above 673 K decrease with increasing heat-treatment temperature. However, starting at 673 K, the tensile elongation increased with increasing pre-consolidation heat-treatment temperature.

Fig. 1

Tensile properties of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys. (a) True stress–true strain curves and the strain hardening curves in tension at room temperature. (b) Change in the tensile yield strength and elongation as a function of pre-consolidation heat-treatment temperature.

Figure 2 shows the change in the fracture toughness as a function of the pre-consolidation heat-treatment temperature. All K_Q values measured in this study satisfied the requirements of the ASTM E399-12^ε3 standard. Thus, they were regarded as the plane-strain fracture toughness values, K_IC. Unlike the variations in strength and ductility, the variations in the K_IC values exhibit a concave curve. The K_IC values, which are ∼10 MPa·m^1/2 for the specimen heat-treated at less than 473 K start decreasing at 473 K, then reach a minimum at ∼5 MPa·m^1/2 at 623 K, and then increase up to 15 MPa·m^1/2 at 723 K with increasing heat-treatment temperature. The results indicate that the pre-consolidation heat treatment above 723 K is effective in improving the fracture toughness. The tensile yield strength σ_0.2, ultimate tensile strength σ_uts, elongation δ, and K_IC values of all the specimens used in this study are listed in Table 1. The values of P_max/P_Q, W-a, 2.5(K_Q/σ_0.2)² are also displayed, where P_Q is critical load, P_max is maximum load, and K_Q is critical stress intensity factor in plane-strain fracture toughness tests.

Fig. 2

Change in the plane-strain fracture toughness, K_IC, of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys as a function of pre-consolidation heat-treatment temperature.

Table 1 Mechanical properties of the Mg–0.85Zn–2.05Y–0.35Al alloys fabricated via consolidation of as-quenched and heat-treated RS ribbons.

3.2 Microstructure characterization

3.2.1 NHT alloy prepared from as-quenched RS ribbons

The solute distribution of the Mg–0.85Zn–2.05Y–0.35Al alloy prepared from the as-quenched RS ribbons was investigated using SEM and TEM. Figure 3(a) shows a backscatter electron (BSE) image of the longitudinal section of the NHT specimen. The BSE signals have an atomic number contrast (Z-contrast), but the NHT alloy exhibits no distinct contrast in the SEM-BSE image because of the fine-grained microstructure. TEM work was needed and therefore conducted. Figure 4 shows the HAADF-STEM image and selected area electron diffraction (SAED) pattern obtained from the fine DRX grain region of the NHT specimen. In the HAADF-STEM image, profuse sparse line-shaped Z-contrasts and bundles are observed. The SAED pattern exhibits diffraction spots originating from α-Mg with strong streaks along the c^*-axis direction, and extra spots originating from LPSO phases are negligibly detected. Therefore, the line-shaped Z-contrasts and their bundles are referred to as cluster arranged layers (CALs).¹⁴⁾

Fig. 3

(a) SEM image, (b) IPF map, (c) KAM map, and (d) grain size distribution of the Mg–0.85Zn–2.05Y–0.35Al alloy prepared from the as-quenched RS ribbons (NHT). The color codes in each bin in the grain size distribution histogram (d) indicate the KAM angles.

Fig. 4

(a) HAADF-STEM image and (b) SAED pattern of the Mg–0.85Zn–2.05Y–0.35Al alloy prepared by consolidation of the as-quenched RS ribbons (NHT).

The crystallographic orientation and grain-size distribution of the NHT alloy were investigated using EBSD. Figure 3(b), (c), (d) shows the inverse pole figure (IPF) map, KAM distribution map, and grain size distribution histogram of the longitudinal section of the NHT specimen. EBSD data with confidence index (CI) values higher than 0.1 were used in this study. Therefore, the non-indexed black areas correspond to the regions where CI values are less than 0.1, i.e., the OIM system could not properly analyze the Kikuchi diffraction patterns as hcp-Mg phase. These black areas may correspond to the LPSO phase of 18R structure. The KAM value indicates the magnitude of the internal strain or geometrically necessary dislocation (GND) density in the individual grains.¹⁸⁾ The IPF map revealed that the NHT alloy consists of two matrices: sub-micrometer-sized dynamically recrystallized grains (ultrafine DRX) and coarse hot-worked grains elongated along the extrusion direction. The KAM map indicated that the worked grains tended to have higher KAM angles than the ultrafine DRX grains. The grain size distribution indicates that the NHT alloy possesses an obvious bimodal distribution in grains size, with an average grain size of ∼0.68 ± 0.21 µm of the DRX grains. The color codes in each bin in the grain-size distribution histogram indicate the KAM angles. The grains with high KAM angles appear to be distributed with grain sizes of over 1 µm.

3.2.2 HT alloys prepared from heat-treated RS ribbons

Figure 5 shows typical SEM-BSE images of the longitudinal sections of the Mg–0.85Zn–2.05Y–0.35Al alloys prepared by consolidation of the RS ribbons that were heat-treated at 623 K, 673 K, 723 K, and 738 K for 24 h. While the HT623K alloy showed a uniform gray contrast similar to that of the NHT alloy, the other three alloys (HT673K, HT723K, and HT738K) possessed three regions that could be characterized by bright, gray, and dark Z-contrast. The bright region corresponds to a moderately coarse LPSO phase enriched in both Zn and Y. The dark region corresponds to the α-Mg matrix without the LPSO phase. It was challenging to clarify the microstructure in the gray region using only the SEM observations. Figure 6 shows the HAADF-STEM images and SAED patterns obtained from the DRX grain regions of the HT623K, HT673K, HT723K, and HT738K alloys. The solute distributions of the HT623K and HT673K alloys are similar to that of the NHT alloy in that they showed profuse line-shaped Z-contrasts in the STEM image and strong streaks along the c^*-axis in the SAED pattern. This indicates that profuse CALs and plate-shaped LPSO phases were formed in the α-Mg grains. However, there were apparent differences in the mechanical properties. The HT623K alloy showed a lower K_IC than the NHT and HT673K alloys. Besides, the HT673K alloy showed lower strengths and higher elongation to failure than those of other two alloys. The difference in their mechanical properties may originate from slight difference in the CALs precipitates morphology. The CALs in the HT623K alloy precipitate to cover the entire grain in comparison with the NHT alloy. On the other hand, the CALs in the HT673K alloy tend to grow up along basal plane stacking direction and to reduce the CALs dispersion. Conversely, the HT723K and HT738K alloys prepared from the RS ribbons that were heat-treated at temperatures above 723 K exhibit significantly different solute distributions; the CALs disappear and a block-shaped LPSO phase is formed around the α-Mg grain boundaries. The thickness of the CALs or LPSO phase toward the c-axis tends to increase with increasing heat-treatment temperature. The SAED patterns indicate that the LPSO phases in the HT723K and HT738K alloys exhibit intergrowth of the 14H and 18R structures developing in the 2H structure of the matrix.

Fig. 5

SEM images of the Mg–0.85Zn–2.05Y–0.35Al alloy prepared by consolidation of heat-treated RS ribbons: (a) HT623K, (b) HT673K, (c) HT723K, and (d) HT738K.

Fig. 6

(a)–(d) HAADF-STEM images and (e)–(j) SAED patterns of the ultra-fine DRX grain region in the Mg–0.85Zn–2.05Y–0.35Al alloys prepared by consolidation of RS ribbons: (a), (e) HT623K, (b), (f) HT673K, (c), (g), (h) HT723K, and (d), (i), (j) HT738K. The SAED patterns are taken with the incident beam parallel to the $[11\bar{2}0]$.

Figure 7 shows the IPF maps, KAM distribution maps, and grain size distribution histograms of the HT623K, HT673K, HT723K, and HT738K alloys. The color codes in each bin in the grain-size distribution histogram correspond to the KAM angles. The bimodal grain-size distribution of the HT623K and HT673K alloys are similar to that of the NHT alloy. Specifically, the HT623K and HT673K alloys have bimodal microstructure consisting of coarse-worked grains (2.2∼2.8 µm) with high KAM angles (1.6∼1.9°) and ultrafine DRX grains (0.64∼0.8 µm) with intermediate KAM angles (0.9∼1.2°). In contrast, the HT723K and HT738K alloys exhibit completely different grain structures compared with the HT623K and HT673K alloys. Specifically, the microstructures of the HT723K and HT738K alloys possess micrometer-sized DRX grains (fine DRX grains) (2.3∼2.4 µm) with low KAM angles (0.5∼0.6°), in addition to coarse-worked grains and ultrafine DRX grains.

Fig. 7

(a)–(d) IPF maps, (e)–(h) KAM maps, and (i)–(l) grain size distribution histograms of the Mg–0.85Zn–2.05Y–0.35Al alloys prepared by consolidation of the heat-treated RS ribbons: (a), (e), (i) HT623K, (b), (f), (j) HT673K, (c), (g), (k) HT723K, and (d), (h), (l) HT738K. The color codes in each bin in the grain size distribution histogram (i)–(l) indicate the KAM angles.

3.3 Fractography

To elucidate the fracture characteristics of the alloys in detail, fracture topology was investigated by using the confocal optical microscopy, SEM, and TEM. Figure 8 shows the SEM images of the fracture surfaces of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys after the plane-strain fracture toughness tests. In the macroscopic SEM image of the NHT specimen, Fig. 8(a), periodic protrusions are visible; they are along the crack propagation direction. The spacing between two neighboring of protrusions was ∼17 µm, which was the exact thickness of the RS ribbons before consolidation. Therefore, these protrusions may be caused by the delamination of the bonding interface between the RS ribbons. The fracture surface that formed beyond the ridge zone is moderately smooth. In the other alloys prepared from heat-treated RS ribbons, Fig. 8(b)–(d), the thickness of the ridge zone became narrow, although the roughness of the fracture surface tends to increase with increasing heat-treatment temperature. The SEM images of the top view of the fracture surface of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys, Fig. 8(e)–(h), reveal that all the alloys had a mixed-mode fracture, which consisted of fine transgranular tearing and intergranular facet features. The size of the intergranular facet is in a good agreement with the grain size of each alloy. The SEM images of the cross-sectional fracture surfaces of the NHT and HT623K specimens, Fig. 8(i), (j), reveal a very fine microstructure; however, the fracture surface of the HT623K specimen is flatter than that of the NHT specimen. In the HT673K specimen (Fig. 8(k)), the cracks appear to propagate with little deflection in the gray contrast region that corresponds to ultrafine DRX grains. But the cracks often deflected in the coarse LPSO-dispersed region. In the HT723K specimen (Fig. 8(l)), crack deflections were often observed even at the ultrafine DRX grain region. To quantify the surface roughness of the fracture surfaces of the alloys, the average surface roughness, R_a, was calculated using the following equation:   

\begin{equation} R_{a} = \frac{1}{l}\int_{0}^{l}|H(x)|dx, \end{equation} (4)

where l is the evaluation length in the confocal optical microscope observation, and H(x) is the profile height function. Figure 9(a) shows the pre-consolidation heat-treatment temperature dependence of R_a for the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys. The variation in the R_a values for the alloys prepared from heat-treated RS ribbons exhibit a concave curve, similar to the variation in the K_IC value. Specifically, the R_a values, which are approximately 2 µm at temperatures below 473 K, start decreasing at 473 K, then reach the lowest value of ∼1 µm at 623 K, and then increase to ∼3.5 µm with increasing heat-treatment temperature. Figure 9(b) shows the change in K_IC as a function of R_a of the fracture surface. Notably, K_IC and R_a of the fracture surface are well correlated.

Fig. 8

SEM images of the Mg–0.85Zn–2.05Y–0.35Al alloys prepared via RS ribbon consolidation. (a)–(d) Macroscopic images of fracture surface and pre-crack regions. (e)–(h) top-views and (i)–(l) cross-section views of fracture surface regions. (a), (e), (i) NHT, (b), (f), (j) HT623K, (c), (g), (k) HT673K, and (d), (h), (l) HT723K.

Fig. 9

(a) Pre-consolidation heat-treatment temperature dependence of R_a for the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys. (b) K_IC as a function of the R_a of the fracture surface for the alloys.

Figure 10 shows the HAADF-STEM images of the internal structure underneath the fracture surface of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys. In the NHT specimen, voids are often observed only in the vicinity of the fracture surface. However, in the HT623K specimen, which exhibited the lowest K_IC value in this study, voids are negligibly observed. Microcracks and voids are often observed in the HT673K and HT723K alloys. In particular, these microcracks tend to form in the block-shaped LPSO phase with a thickness of over 50 nm toward the c-axis as indicated by blue arrows in Fig. 10(d), even at a distance from the main crack. The effect of the blocky LPSO phase formation on fracture resistance of this Mg alloy microstructure is discussed in the next section.

Fig. 10

HAADF-STEM images and schematic of cross-sections of fracture surface of the Mg–0.85Zn–2.05Y–0.35Al alloys prepared by consolidation of RS ribbons: (a), (e) NHT, (b), (f) HT623K, (c), (g) HT673K, and (d), (h) HT723K.

4. Discussion

4.1 Effect of multimodally grained microstructure evolution

The OIM analysis via EBSD measurements indicated that bi- or tri-modal grained Mg–0.85Zn–2.05Y–0.35Al alloys were prepared from the as-quenched and heat-treated RS ribbons. To elucidate the grain structure of the alloys, a bivariate data of the grain size (d) and the mean KAM angle (V_KAM) of the NHT, HT623K, HT673K, HT723K, and HT738K alloys are plotted in Fig. 11. The d-V_KAM bivariate data show that the α-Mg grains in the NHT, HT623K, and HT673K alloys form two groups consisting of coarse-worked grains (not recrystallized) and ultrafine DRX grains. The coarse-worked grains (Group 1) are characterized by large grain sizes (0.5–6.5 µm) and large KAM angles (>1.2°) with large amounts of scatter. The DRX grains (Group 2) are characterized by submicron-sized ultrafine DRX grains (0.3–1.5 µm), intermediate KAM angles (0.6–1.2°), and a large negative relationship between the grain size and KAM angle (ΔV_KAM/Δd is negative). In contrast, the α-Mg grains in the HT723K and HT738K alloys form a trimodally grained microstructure. Specifically, in the HT723K and HT738K alloys, the third microstructural feature (Group 3) is characterized by a micrometer-sized grain size (1.5–5.0 µm), low KAM angles (<0.6°), and a small negative relationship. The formation of the third group changes the profiles of the stress–strain curves of the alloys, as shown in Fig. 1(a). Specifically, the alloys possessing only Groups 1 and 2 exhibit higher tensile yield strength accompanied by work-softening after yielding, whereas the alloys containing three groups show moderately large elongations with strain hardening. In the latter case, plastic deformation may start with the plasticity of the grains of Group 3, because the mean Schmid factor for the basal slip in Group 3 was higher than those of the other groups. The average Schmid factors for the basal slip in groups 1, 2, and 3 of the HT738K alloy were 0.13, 0.26, and 0.32, respectively. That is, the share of Group 3 with large grain size and low KAM angles (low dislocation density) may become dominant in the strain accommodation, resulting in strain hardening, although the shares of Groups 1 and 2 strengthen the alloys. Therefore, developing a multimodal microstructure with a tailored distribution of dislocation densities can increase the yield strength and elongation of Mg–Zn–Y–Al alloys prepared by the consolidation of RS ribbons. Moreover, the higher level of work hardenability of the alloys can improve their fracture toughness.

Fig. 11

Bivariate data of the grain size and mean KAM value for the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al alloys prepared at different pre-consolidation heat-treatment conditions.

4.2 Effect of LPSO phase precipitation on fracture behavior

The pre-consolidation heat treatment influenced the grain structure of the α-Mg phase and the morphology of the LPSO phase. As shown in Fig. 9(b), fracture toughness and fracture surface roughness are well correlated; a greater fracture surface roughness corresponds to a higher fracture toughness. The cross-sectional SEM images of the fracture surface shown in Fig. 8 indicate that in this case the roughness of the fracture surface is related to the crack deflection process. Furthermore, the crack deflection tends to increase with increasing volume fraction and size of the LPSO phase. HAADF-STEM observations of the alloys prepared from the RS ribbons heat-treated at temperature above 673 K, as shown in Fig. 10(c) and (g) for the HT673K and Fig. 10(d) and (h) for the HT723K, reveal that microcracks are often formed inside and around the LPSO phase. Mine et al. conducted micro-fracture testing of α-Mg + LPSO two-phase and LPSO single-phase micrometer-sized specimens and reported that the K_IC values of the former and the latter alloys are estimated to be ∼21 and ∼1.5 MPa m^1/2, respectively.¹⁹⁾ That is, the single-phase LPSO is brittle and fracture resistance is small. But in the α-Mg + LPSO two-phase alloys, the blocky LPSO phase has a beneficial effect in improving the fracture toughness by deflection the crack path and advance.

In general, it is claimed that toughening by crack deflection is the result of a reduction in the local stress intensity factor at the crack tip when a crack deviates from its original straight path.^20,21) The crack deflection causes a change in the mode of the stress intensity factor, K, from near mode I to the mix of modes I and II, even in the macroscopic mode I loading state, and leads to a decrease in the effective crack tip stress intensity, consequently resulting in an improvement in K_IC. Figure 12 shows the effect of the thickness of CALs or LPSO on fracture surface roughness R_a and fracture toughness K_IC. Relationship between them are well correlated, and the R_a and K_IC increases with increasing thickness. The pre-consolidation heat treatments affect the thickness of the CALs or LPSO phase (Fig. 6). When the CALs or LPSO phase thickness are less than 30 nm toward the c-axis direction, as for the NHT, HT623K, and HT673K alloys, they precipitate coherently with the α-Mg matrix, and the alloys behave like single-phase alloys. Meanwhile, if the thickness of the LPSO phase exceeds 50 nm, as for the HT723K and HT738K alloys, the block-shaped LPSO phase is formed along the grain boundaries as well as in the grain, and the alloys assume the aspect of the two-phase microstructure. Figure 13 shows schematics of the crack propagation behavior for single-phase and two-phase alloys. In single-phase alloys, a less tortuous crack tends to propagate along the grain boundaries. In this case, the crack propagation also causes some intragranular deformation of the grains near the main crack; evidence of tearing fracture is visible in Fig. 8, and must be distinguished from the brittle intergranular fracture. In contrast, cracks in two-phase alloys tend to advance in various directions. This is probably because the inside of the block-shaped LPSO phase and the interface boundary between the LPSO phase and α-Mg matrix are the nucleation sites for voids and microcracks as shown in Fig. 10(d). At the crack tip, which is subjected to a strong strain field, such voids and microcracks are generated extensively prior to crack propagation, and the crack is largely deflected through a close-by void or microcrack.^22,23) In addition to the crack deflection, the formation of the secondary crack (as reported previously^13,14)) induced by the block-shaped LPSO phase can contribute to enhancing the fracture toughness. Therefore, the morphological change from the thin CALs or LPSO phase to the block-shaped LPSO phase is effective in improving the fracture toughness of the RS ribbon-consolidated Mg–Zn–Y–Al alloys.

Fig. 12

The effect of the thickness of CALs or LPSO on R_a and K_IC. The colored area indicates the blocky LPSO size of 50–100 nm.

Fig. 13

Schematics of the crack propagation behavior. (a) pseudo-single-phase type alloy (the thickness of the CALs or LPSO < 30 nm). (b) Two-phase type alloy (the thickness of the LPSO > 50 nm).

In the α₂ + β two-phase Ti₃Al-based alloy, Chan²⁰⁾ estimated toughening by crack deflection by introducing the toughening ratio, λ_d, which is defined as the ratio of the applied stress intensity factor, K_∞, to the stress intensity factor in the matrix, K_m, or in the equivalent (effective) crack tip stress intensity factor, K_eq. λ_d can be expressed using the deflection angle ϕ as follows.   

\begin{equation} \lambda_{\text{d}} = \frac{K_{\infty}}{K_{\text{m}}} = \frac{K_{\infty}}{K_{\text{eq}}} = \frac{1}{\cos^{2}(\phi/2)} \end{equation} (5)

This equation indicates that crack deflection toughening increases with increasing crack deflection angle. For LPSO-phase-containing Mg–Zn–Y–Al alloys the thickness of the LPSO grains as well as their intervals may be important parameters for controlling the crack deflection angle.

Moreover, Chan and Shih measured the fracture resistance of two phase γ-TiAl materials to further examine the contribution of microstructural and other variables.²⁴⁾ In future investigation, more detailed K-resistance studies that also examine the effect of loading condition - e.g., monotonic vs. cyclic - will be conducted. The aim is to clearly define at what level of intrinsic plasticity in the Mg–Zn–Y–Al alloy system in which the singularity at crack-tip evolves from an HRR (Hutchinson-Rice-Rosengren) type stationary crack^25,26) to an advancing crack.²⁷⁾

5. Conclusions

The multimodal microstructure evolution as well as tensile and fracture toughness behavior of the RS ribbon-consolidated Mg–0.85Zn–2.05Y–0.35Al (at%) alloys have been investigated to optimize the pre-consolidation heat-treatment temperature for fracture toughness improvement. The results are summarized as follows.

1. (1)    Pre-consolidation heat treatment above 723 K for 24 h was effective in improving fracture toughness. The Mg–0.85Zn–2.05Y–0.35Al alloys heat-treated below 673 K showed high strength (480–500 MPa), but low fracture toughness values (5–10 MPa·m^1/2). Conversely, alloys subjected to pre-consolidation heat treatment above 723 K showed moderate strength (400–430 MPa), large ductility (10–12% elongation), and high fracture toughness (13–15 MPa·m^1/2).
2. (2)    The microstructure of the Mg–0.85Zn–2.05Y–0.35Al alloys with pre-consolidation heat treatment below 673 K, consisted of bimodal α-Mg grains: coarse-worked grains (2.2∼2.9 µm) with high KAM values (1.6∼2.0°) and the ultrafine DRX grains (0.64∼0.8 µm) with intermediate KAM values (0.9∼1.2°). CALs and plate-shaped LPSO phases (thickness < 30 nm) formed along the basal planes of the α-Mg matrix. These grains mainly contribute to strengthening the alloy but scarcely hinder crack propagation in the fracture toughness test.
3. (3)    The Mg–0.85Zn–2.05Y–0.35Al alloys heat-treated above 723 K formed trimodal α-Mg grains consisting of coarse-worked grains (1.5∼1.8 µm), ultrafine DRX grains (0.91∼0.93 µm), and fine DRX grains (2.3∼2.4 µm) with low KAM values (0.5∼0.6°). The appearance of fine DRX grains with low KAM values improves the work hardenability and ductility, resulting in an enhancement of the fracture toughness.
4. (4)    Pre-consolidation heat treatment above 723 K had a significant effect on the morphology of the LPSO phase. The formation of the block-shaped LPSO phase (thickness > 50 nm) is important in enhancing fracture toughness in the Mg–0.85Zn–2.05Y–0.35Al alloys. Voids and microcracks frequently formed inside the LPSO phase and the LPSO/α-Mg interface. They promote crack deflection, and further enhancement of fracture toughness can be achieved.

Acknowledgements

The authors are indebted to and thank Prof. Donald S. Shih (Magnesium Research Center, Kumamoto University) for fruitful discussions. This study was supported by the JSPS KAKENHI for Scientific Research on Innovative Areas “MFS Materials Science” (JP18H05476) and for Scientific Researches (JP20H00312 and JP21H01673). S.N. is also thankful for support from the Research Fellowships of the JSPS for Young Scientists (21J13431).

REFERENCES

• 1)   T.M.  Pollock: Science 328 (2010) 986–987. doi:10.1126/science.1182848
• 2)   A.A.  Luo: J. Magnes. Alloy. 1 (2013) 2–22. doi:10.1016/j.jma.2013.02.002
• 3)   J.P.  Weiler: J. Magnes. Alloy. 7 (2019) 297–304. doi:10.1016/j.jma.2019.02.005
• 4)   S.  You,  Y.  Huang,  K.U.  Kainer and  N.  Hort: J. Magnes. Alloy. 5 (2017) 239–253. doi:10.1016/j.jma.2017.09.001
• 5)  M.O. Pekguleryuz: Advances in Wrought Magnesium Alloys, ed. by C. Bettles and M. Barnett, (Woodhead Publishing, Cambridge, 2012) pp. 3–62.
• 6)   Y.  Kawamura,  K.  Hayashi,  A.  Inoue and  T.  Masumoto: Mater. Trans. 42 (2001) 1172–1176. doi:10.2320/matertrans.42.1172
• 7)   S.  Izumi,  M.  Yamasaki and  Y.  Kawamura: ECS Trans. 16(32) (2009) 65–72. doi:10.1149/1.3091905
• 8)   S.  Izumi,  M.  Yamasaki and  Y.  Kawamura: Corros. Sci. 51 (2009) 395–402. doi:10.1016/j.corsci.2008.11.003
• 9)   M.  Yamasaki,  S.  Izumi and  Y.  Kawamura: ECS Trans. 16(32) (2009) 81–88. doi:10.1149/1.3091907
• 10)   K.  Suzawa,  S.  Inoue,  S.  Nishimoto,  S.  Fuchigami,  M.  Yamasaki,  Y.  Kawamura,  K.  Yoshida and  N.  Kawabe: Mater. Sci. Eng. A 764 (2019) 138179. doi:10.1016/j.msea.2019.138179
• 11)   D.  Drozdenko,  M.  Yamasaki,  K.  Máthis,  P.  Dobron,  P.  Lukác,  N.  Kizu,  S.  Inoue and  Y.  Kawamura: Mater. Des. 181 (2019) 107984. doi:10.1016/j.matdes.2019.107984
• 12)   M.  Yamasaki,  S.  Izumi,  Y.  Kawamura and  H.  Habazaki: Appl. Surf. Sci. 257 (2011) 8258–8267. doi:10.1016/j.apsusc.2011.01.046
• 13)  S. Nishimoto, M. Yamasaki, S. Inoue and Y. Kawamura: Proceedings of the 12th International Conference on Magnesium Alloys and Their Applications, ed. by A. Luo, M. Pekguleryuz, S. Agnew, J. Allison, K. Kainer, E. Nyberg, W. Poole, K. Sadayappan, B. Williams and S. Yue, (Springer, Cham, 2021) pp. 71–77. doi:10.1007/978-3-030-72432-0_8.
• 14)   S.  Nishimoto,  Y.  Koguchi,  M.  Yamasaki and  Y.  Kawamura: Mater. Sci. Eng. A 832 (2022) 142348. doi:10.1016/j.msea.2021.142348
• 15)   H.  Okouchi,  Y.  Seki,  T.  Sekigawa,  H.  Hira and  Y.  Kawamura: Mater. Sci. Forum 638–642 (2010) 1476–1481. doi:10.4028/www.scientific.net/MSF.638-642.1476
• 16)  ASTM E399-12^ε3, Standard Test Method for Plane–Strain Fracture Toughness of Metallic Materials, (ASTM, West Conshohocken, PA, 2012).
• 17)   I.S.  Yasnikov,  A.  Vinogradov and  Y.  Estrin: Scr. Mater. 76 (2014) 37–40. doi:10.1016/j.scriptamat.2013.12.009
• 18)   M.  Calcagnotto,  D.  Ponge,  E.  Demir and  D.  Raabe: Mater. Sci. Eng. A 527 (2010) 2738–2746. doi:10.1016/j.msea.2010.01.004
• 19)   Y.  Mine,  H.  Yoshimura,  M.  Matsuda,  K.  Takashima and  Y.  Kawamura: Mater. Sci. Eng. A 570 (2013) 63–69. doi:10.1016/j.msea.2013.01.069
• 20)   K.S.  Chan: Metall. Trans. A 23 (1992) 183–199. doi:10.1007/BF02660864
• 21)   S.  Suresh: Metall. Trans. A 16 (1985) 249–260. doi:10.1007/BF02816051
• 22)   F.M.  Heim,  Y.  Zhang and  X.  Li: Adv. Eng. Mater. 20 (2017) 1700605. doi:10.1002/adem.201700605
• 23)   S.  Osovski,  A.  Srivastava,  L.  Ponson,  E.  Bouchaud,  V.  Tvergaard,  K.  Ravi-Chandar and  A.  Needleman: J. Mechan. Phys. Solids 76 (2015) 20–46. doi:10.1016/j.jmps.2014.11.007
• 24)   K.S.  Chan and  D.S.  Shih: Metall. Mater. Trans. A 28 (1997) 79–90. doi:10.1007/s11661-997-0084-8
• 25)   J.W.  Hutchinson: J. Mechan. Phys. Solids 16 (1968) 13–31. doi:10.1016/0022-5096(68)90014-8
• 26)   J.R.  Rice and  G.R.  Rosengren: J. Mechan. Phys. Solids 16 (1968) 1–12. doi:10.1016/0022-5096(68)90013-6
• 27)  J.R. Rice, W.J. Drugan and T.L. Sham: Fracture Mechanics: 12th Conference, ASTM STP700, (ASTM, Philadelphia, PA, 1980) pp. 189–221.