Effect of Pre-Heat Treatment on Microstructure and Properties of As-Extruded AZ91-CaO Alloy

Abstract

The effects of different pre-heat treatments on the microstructure, mechanical properties and corrosion resistance of extruded AZ91 and AZ91-CaO alloys were investigated. The results showed that the microstructure of AZ91 is clearly refined after adding CaO modifier. After solution heat treatment (T4), most of the second phase was dissolved into the matrix, while after aging heat treatment (T6), new fine and discontinuous Mg₁₇Al₁₂ were precipitated in both AZ91 and AZ91-CaO alloys. The average grain size of the extruded AZ91-CaO-T6 alloy was 0.9 μm, and the second phase has the smallest size and more uniform dispersion. Therefore, the mechanical properties of AZ91-CaO-T6-EX are optimal, and its ultimate tensile strength, tensile yield strength and elongation reach 367.6 MPa, 320.2 MPa, and 9.5%, respectively. Meanwhile, the electrochemical and salt spray corrosion experiments also showed that the AZ91-CaO-T6-EX alloy had the best corrosion resistance; the self-corrosion potential was −1.08 mV, the corrosion current density was 5.65 μA/cm², and the annual corrosion rate was 0.36 mm/y after 72 h salt spray treatment. This high corrosion performance was mainly attributed to the dispersed second phase with small size and fine grain size.

1. Introduction

In recent years, magnesium (Mg) alloys have been widely used in the aerospace and automotive industries as lightweight structures and components [1,2]; AZ91 is a typical representative alloy due to its excellent mechanical properties, absence of rare earth elements and wide solidification range [3]. However, the coarse second phase and large grain size in as-cast AZ91 alloy lead to poor mechanical properties and corrosion resistance, which limits its further application [4,5]. Based on this, many methods have been employed, such as hot deformation [6,7], heat treatment [8] and the addition of inhibitors [9] to refine the coarse second phase and the grain size of AZ91 alloy. In particular, the combined heat treatment and subsequent deformation became an effective way to improve the performance of AZ91 alloy.

Lee et al. [10] studied the effects of aging (T6), post-extrusion aging (T5) and solid solution (T4) on the mechanical properties of AZ80 alloy and the results showed that samples treated with T5 and T6 have higher tensile strengths and lower elongations than that of T4 sample, owing to the presence of more abundant Mg₁₇Al₁₂ precipitates in the former two materials. Jung et al. reported that the mechanical properties of AZ80 [11] and Mg-7Sn-1Al-1Zn [12] alloys were improved by aging prior to extrusion (APE) treatment and explained that the increase in strength was attributed to the decrease in grain size and the increase in the volume fraction of the precipitated phase. Zhu et al. [13] have studied the effects of APE and post-extrusion aging on the mechanical properties of AZ91 alloy and the results showed that both of the above methods could refine the grain size and improve the strength of AZ91 alloy, but the plasticity of the sample fabricated by aging after extrusion decreased due to the precipitation of a large amount of second phase.

Chelliah et al. [14] demonstrated that AZ91 alloy after T4 treatment had serious pitting corrosion in Ringer’s solution, which reduced its corrosion resistance. Ambat et al. [15] found that T4 treatment would destroy the barrier effect of the continuous network Mg₁₇Al₁₂ phase structure, but the Mg₁₇Al₁₂ phase precipitated by aging treatment could effectively reduce the degree of micro galvanic corrosion. Zhou et al. [16] found that the dissolution of Mg₁₇Al₁₂ phase in T4 microstructure decreased the cathode-to-anode area ratio leading to highly localized corrosion in the a-Mg matrix and intergranular corrosion and pitting were the main corrosion mechanisms in T6 microstructure.

However, the refinement degree of the second phase after T6 treatment was not enough to greatly improve the corrosion performance of the Mg alloys. In order to further improve the performance of Mg alloy, modifiers were added to modify the microstructure of Mg alloys, and the known modifiers are mainly rare earth elements (RE) [17], alkaline earth elements (Ca, Sr) [18,19], and ceramic particle (SiC, TiO₂) [20]. Adding Ca to Mg can effectively refine the grain and improve the performance [21], but Ca was mostly added to Mg melt through Mg-Ca master alloy [22,23,24]. Gu et al. [25] studied the effect of Ca and CaO on the grain size of the alloy and found that the same mass percentage of CaO had a more obvious refinement effect. Moreover, although the price of Ca is lower than that of REs and other alkaline earth elements, the extraction of Ca elements and the addition of Mg-Ca master alloy would also increase the production cost [26,27,28].

According to our previous research [29], after adding CaO to AZ91 alloy, a small number of Al₂Ca phase formed and consumed part of Al element in the matrix, which led to the reduction in the amount and refinement of the size of the Mg₁₇Al₁₂ phase, resulting in the improvement of the mechanical properties of as-cast AZ91 alloy. In this study, different heat treatment processes were performed to treat AZ91 and AZ91-CaO alloys, and then the alloys were extruded to study the effects of heat treatment on the microstructure and properties of the extruded alloys.

2. Experimental Methods

2.1. Material Preparation

The raw materials used were commercial AZ91 alloy and CaO ceramic particles with an average particle size of 500 nm. The commercial AZ91 alloy was melted in a resistance protected by a mixture of 99.9% N₂ and 0.1% SF₆ gas. The melting temperature was raised to 780 °C at a rate of 5 °C/min. After AZ91 block was melted, CaO powders with a mass percentage of 0.7 (0.7 wt.%) were added into the melt, and a hand-held electric drill (Jiangsu Dongcheng M&E Tools Co., Ltd., Qidong, China) with blades was used to disperse the particles initially for 10min. Subsequently, a high shear strong melt stirring device (Hangzhou Success Ultrasonic Equipment Co., Ltd., Hangzhou, China) and an ultrasonic stirring bar were used to stir the melt for 15 mins, respectively. The melt was cooled to 720 °C and kept for 20 min at this temperature and then it was poured into a pre-heated steel mold with a diameter of 60mm. In order to ensure the consistency of the experiment, the same process was used to remelt the commercial AZ91 alloy. Then, the remelted AZ91 ingot and AZ91-CaO ingot were solution-treated at 420 °C for 8 h and then quenched into warm water to obtain the T4 sample. After a solid solution was obtained, part of the T4 sample was subjected to aging treatment at 170 °C for 12 h, and then the T6 samples were obtained. Before extrusion, the T4 and T6 billets were preheated in muffle furnace at 300 °C for 1 h and then were extruded at this temperature with a speed of 0.5 mm/s and an extrusion ratio of 36:1.

2.2. Microstructural Characterization

The phase compositions of the alloys were measured by X-ray diffractometer (Rigaku, Wako, Japan) with Cu target Kα line (λ = 0.15418 nm), accelerating constant voltage of 40 kV and constant current of 100 Ma. The cast, heat-treated and extruded samples were ground by silicon carbide sandpaper of 80#, 320#, 800#, 1500#, 3000#, 5000#. Subsequently, the metallographic samples were prepared by physical polishing and chemical erosion with an abrasive polishing machine (Shenyang Kejing AUTO-INSTRUMENT Co., Ltd., Shenyang, China). The erosion solution consisted of 3.75 g picric acid, 5 mL glacial acetic acid, 5 mL deionized water and 45 mL absolute ethanol. The distribution of the second phase and the grain size of the samples were observed using a metallographic microscope (Olympus, Tokyo, Japan), and the grain size was measured by the mean linear intercept method. A field emission scanning electron microscope (Quanta FEG 250 (FEI Inc., Hillsboro, OR, USA)) was used to observe the microstructure characteristics, and EDS was used to determine the composition of the second phase.

2.3. Mechanical Performance Test

The sample was machined into tensile specimen according to GB/T 24176-2009 standard. The specimens were ground and polished to remove the processing coolant and oxide, and then were ultrasonic cleaned with absolute ethanol. The tensile properties of the specimens were measured by a universal testing machine (MTS, State of Minnesota, MN, USA) with a collet rate of 0.5 mm/min. Three parallel samples were set for each group of samples, and the average value was obtained.

2.4. Corrosion Resistance Test

Corrosion behavior was tested by an electrochemical workstation (CHI-7601 (Zahner, Kronach, Germany)). The sample was used as the working electrode, the saturated calomel electrode as the reference electrode, and the graphite electrode as the auxiliary electrode. The test solution was 3.5 wt.% NaCl solution, and the test area was 0.785 cm². At the beginning of the experiment, the samples ground to 3000# were subjected to the open circuit potential (OCP) test in solution for at least 30 mins. After the open-circuit potential was stabilized, electrochemical impedance spectroscopy (EIS) experiment was conducted with a frequency range of 10⁻² Hz–10⁵ Hz. Finally, a potentiodynamic polarization test was conducted with a scanning speed of 1 mV/s. In order to ensure the stability and credibility of the experiment, three parallel samples were set for each group of alloys.

The salt spray test was in 5% NaCl solution with pH between 6.5 and 7.2, and the salt spray temperature was 35 °C. The NaCl solution was sprayed through the spray device, and the sedimentation rate of salt spray was set to be between 1–2 mL/(80 cm²·h). The spray method was selected as interval spray, the interval time was 12 h, and the total experimental time was 72 h. Before the experiment, the samples were polished, cleaned with absolute ethanol, and dried with a hair dryer. The corrosion rate was calculated by weight loss method, and the samples before corrosion were weighed and recorded by electronic balance. All samples were placed in the salt spray test chamber 15–30° to the vertical plane of the test box, and ensure the samples were not in contact with each other and the salt solution did not drop on other samples. During the experiment, samples were taken at 12, 24, 48 and 72 h. At the end of the experiment, the sample was placed in chromic acid solution to remove corrosion products on the surface of the sample. After cold air drying, the surface corrosion morphology was observed, and the sample was weighed and recorded to obtain the weight loss. Three parallel samples were set for each group of samples.

3. Results and Discussion

3.1. Microstructure of the Heat-Treated Alloys

Figure 1 shows the OM microstructure of as-cast and heat-treated AZ91 and AZ91-CaO alloys. The microstructure of the as-cast alloys exhibits typical dendrite distribution in that the dendrite spacing of AZ91-CaO is much smaller than that of AZ91 alloy, indicating that CaO particles play a significant role in refining the microstructure of AZ91 alloy. This was mainly attributed to reaction between the added CaO and Mg melt, resulting in the formation of Al₂Ca and MgO nanoparticles. These particles promoted the heterogeneous nucleation of Mg, which not only refined the dendrite structure of Mg, but also significantly reduced the size of the second phase [29,30,31]. The volume fraction of second phase decreased from 56.25% in as-cast AZ91 (AZ91-Cast) to 45.47% in as-cast AZ91-CaO (AZ91-CaO-Cast) (Figure 1a,b). After T4 treatment, the second phase in AZ91-T4 alloy is basically dissolved into the Mg matrix, and only a small amount of them can be seen in the grain interior. The grain size distribution of the alloy is relatively uniform, and the average grain size is about 87.3 μm (Figure 1b). For AZ91-CaO-T4 alloy, the grains are equiaxed with a uniform size distribution, as shown in Figure 1e. The average grain size is about 48.4 μm, which is only half that of AZ91-T4. There are also some undissolved second phase particles at grain boundaries. After T6 treatment, the second phases were precipitated at the grain boundary for both alloys, as shown in Figure 1c and 1f. The precipitates of AZ91-T6 exist as clusters along the grain boundary (Figure 1c) and the magnification image inserted in the bottom left-hand corner in Figure 1c shows that the precipitates are in a semi-continuous distribution with a range of 10~100 μm. For AZ91-CaO-T6 alloy, the precipitates distribute in a short ribbon on the grain boundaries. Meanwhile, the area fraction of the precipitates in AZ91-T6 is apparently higher than that in AZ91-CaO-T6.

Figure 2 shows the SEM microstructure of AZ91 and AZ91-CaO alloys in T4 and T6 states. As can be seen from Figure 2a, the remaining second phases of AZ91-T4 alloy are mainly round particles or fine strip shapes.

Table 1. EDS result of AZ91-T4 and AZ91-CaO-T4 in Figure 2.

Analyzed Locations	Alloys	Elements (at. %)
		Mg	Al	Ca	Mn	Zn
A	AZ91-T4	5.96	68.08	0.00	25.95	0.01
B	AZ91-T4	0.75	67.59	0.00	31.66	0.00
C	AZ91-T4	62.43	36.57	0.00	0.00	1.00
D	AZ91-CaO-T4	1.03	62.91	0.04	36.01	0.00
E	AZ91-CaO-T4	59.87	37.97	1.09	0.04	1.03

shows the EDS data of the corresponding points in Figure 2. The EDS analysis shows these particles are identified as Al₈Mn₅ phase (point A and point B), which was formed during solidification. This suggests Al₈Mn₅ has a high thermal stability and can still be observed after solid solution. The fine strip phase is inferred to be Mg₁₇Al₁₂ phase, according to the EDS result of point C. Different from AZ91-T4, the AZ91-CaO-T4 alloy (Figure 2b) has a partial massive phase (point E) in addition to the undissolved Al₈Mn₅ phase (point D). EDS analysis at point E shows that the coarse phase consists of Mg, Al, Ca and Zn elements. It is speculated that some Ca and Zn elements were dissolved into Mg₁₇Al₁₂ phase [32,33,34].

After T6 treatment, a large number of irregular white bright patches were precipitated along grain boundaries in both alloys, and the area fraction of the light gray patches in AZ91-T6 alloy (Figure 2c) is more than that in AZ91-CaO-T6 (Figure 2d). Enlarged images show that these lumpy regions are composed of a large number of discontinuous needle-like phases. According to phase diagram analysis, Mg₁₇Al₁₂ phase is precipitated at 170 °C for both alloys. However, the precipitation location of the second phase of AZ91-CaO-T6 alloy is significantly less than that of AZ91-T6 sample, and the spiculate Mg₁₇Al₁₂ precipitated by aging is finer and more uniform. The reason is that under the same solution conditions, the grains of AZ91-CaO-T4 are fine, and more undissolved second phase remains on the grain boundary in AZ91-CaO-T4 alloy (Figure 2b). The supersaturation Al element in α-Mg solid solution of AZ91-CaO-T4 alloy is lower than that of AZ91-T4. During the aging process, the grain boundary and the second phase increase the precipitation sites of Mg₁₇Al₁₂ phase, inhibiting the tendency of the needle piece to grow, and resulting in the fine size (Figure 2f). In AZ91 alloy, most of the primary Mg₁₇Al₁₂ phases are dissolved into α-Mg during solid solution, which leads to high supersaturation of Al element in AZ91-T4. During aging, more Al elements are precipitated in the form of β phase, leading to a higher volume fraction of β phase in the AZ91-T6 alloy.

3.2. Microstructure of the Extruded Alloys with Pre-Heat Treatment

Figure 3 shows the OM micrographs of the longitudinal sections of extruded AZ91 and AZ91-CaO alloys after T4 and T6 treatment, respectively. Most of the microstructure of AZ91-T4-EX alloy is consisted of dynamic recrystallized grains, but there are still some non-recrystallization regions (Figure 3a). The grain size of AZ91-T4-EX is also the largest, and the grain size distribution is not uniform. For AZ91-CaO-T4-EX alloy, as shown in Figure 3c, the area fraction of dynamic recrystallized grains is increased, the microstructure is more uniform and the grain is finer than that of AZ91-T4-EX. After T6 treatment and extrusion, most of the Mg₁₇Al₁₂ spicules distributed along grain boundaries were broken into small particles during extrusion and arranged along the extrusion direction. Compared with T4 extruded alloy, the grain size of AZ91-T6-EX is further refined. The Mg₁₇Al₁₂ particles in AZ91-CaO-T6-EX alloy are finer and exhibit dispersion distribution.

The SEM microstructure and EDS analysis of the two alloys in the different conditions are shown in Figure 4. It can be seen from Figure 4a that in AZ91-T4-EX alloy, there are many coarse second phase and a small amount of fine punctate second phase, and the volume fraction of second phase is the lowest (19.31%) among the four extruded samples.

Table 2. EDS result of AZ91-EX and AZ91-CaO-EX in Figure 4.

Analyzed Locations	Alloys	Elements (at. %)
		Mg	Al	Ca	Mn	Zn
A	AZ91-T4-EX	62.36	36.51	-	0.06	1.07
B	AZ91-T4-EX	68.34	30.19	-	0.32	1.15
C	AZ91-T6-EX	75.01	21.47	-	3.19	0.34
D	AZ91-T6-EX	70.70	28.15	-	0.00	1.15
E	AZ91-CaO-T4-EX	65.30	32.19	1.69	0.03	0.79
F	AZ91-CaO-T4-EX	61.58	34.98	2.20	0.00	1.24
G	AZ91-CaO-T6-EX	57.95	37.34	3.50	0.00	1.21
H	AZ91-CaO-T6-EX	58.08	37.24	3.11	0.46	1.11

shows the EDS data of the corresponding points in Figure 4. The EDS result (point A, point B) revealed that the second phases are basically Mg₁₇Al₁₂. Compared with AZ91-T4-EX, the microstructure of AZ91-CaO-T4-EX alloy is more uniform and the volume fraction of the second phase increases significantly (35.74%). Meanwhile, the size of the second phase also decreases obviously but the type of the second phase is also mainly Mg₁₇Al₁₂ phase (point E, point F). For AZ91-T6-EX alloy, as shown in Figure 4b, the morphology of the second phase has changed. In addition to the relatively coarse second phase (point C), which is partially the same as that in AZ91-T4-EX, there are also some small, short needle-like second phase (point D). Both of these second phases are Mg₁₇Al₁₂ phase. As shown in Figure 4d, for AZ91-CaO-T6-EX alloy, the second phase (point G and point H) also presents the same two states as in AZ91-T6-EX, but the number and size of coarse second phase decreases. At the same time, the volume fraction of the short needle-like second phase is significantly reduced, and the distribution of the second phase is more uniform.

Based on the above experimental results, it is clear that the volume fraction of the second phase in extruded alloy after T4 treatment (Figure 4a,c) is much higher than that of the corresponding T4-treated alloy for both of the AZ91 and AZ91-CaO alloys (Figure 2a,b). This was mainly attributed to the precipitation of a large amount of Mg₁₇Al₁₂ phases from the matrix during preheating and subsequent extrusion. In addition, the reasons for the size reduction of the grain and the second phase particles, and the increase of the volume fraction and dispersion degree of Mg₁₇Al₁₂ phase in AZ91-CaO-T4-EX compared to those of AZ91-T4-EX are as follows: (1) The relatively small grain size in AZ91-CaO-T4 after solution treatment and the small Mg₁₇Al₁₂ phase remaining in the matrix. In the preheating process before extrusion, they provided more non-uniform nucleation sites for the precipitated phase, resulting in more Mg₁₇Al₁₂ phases precipitated and smaller size. These precipitates are broken into smaller size particles during extrusion. (2) The increase in deformation energy storage during extrusion will lead to the precipitation of part of the second phase from the saturated solid solution. Meanwhile, the large second phase particles (>1μm) trigger the particle-stimulated nucleation (PSN) mechanism during extrusion [35]. In this mechanism, high-density dislocations gather around the second phase during extrusion, and the deformation zone located near the second phase has a large amount of energy storage, so it can be used as the preferred location for recrystallized grain nucleation [36].

In addition, it can be seen from Figure 2e,f that after aging treatment, a large number of small discontinuous β phases with short needles or sheets precipitated at grain boundaries. In the process of extrusion, these tiny β-phases were broken into short needle-like shapes, which would pin the grain boundaries of the recrystallized grains and hinder the grain boundary growth, resulting in a finer grain size compared with that in extruded alloys after T4 treatment. Moreover, the β phase precipitated in AZ91-CaO-T6 sample was smaller than that of in AZ91-T6 alloy, and the grain boundary nailing effect was more obvious. Therefore, the smallest grain size was obtained in AZ91-CaO-T6-EX.

3.3. Mechanical Properties of the Pre-Heat Treated and Extruded Alloys

Figure 5 shows the tensile stress-strain curves of AZ91 and AZ91-CaO alloys in different states and

Table 3. Tensile test data.

Sample	UTS(MPa)	YS(MPa)	EL(%)
AZ91-T4-EX	322.8 ± 8.6	219.4 ± 14.8	10.5 ± 3.1
AZ91-T6-EX	355.9 ± 1.3	276.6 ± 37.2	9.5 ± 1.4
AZ91-CaO-T4-EX	362.5 ± 1.4	309.1 ± 13.7	9.7 ± 0.5
AZ91-CaO-T6-EX	367.6 ± 1.6	320.2 ± 8.6	9.5 ± 2.6

shows the corresponding tensile test data. Compared with the as-cast alloy, the mechanical properties of the extruded alloys are significantly improved. Among the extruded alloys, AZ91-T4-EX has the poorest ultimate tensile strength (UTS) and yield strength (YS) of 322.8 MPa and 219.4 MPa, but has a relatively higher fracture elongation (10.5 %) than that of other alloys. This may be due to the large grain size and the large size of the second phase in AZ91-T4-EX alloy. The UTS and YS of AZ91-CaO-T4-EX are 362.5 MPa and 309.1 MPa, which is increased by 12.3% and 40.8%, respectively, compared with AZ91-T4-EX. Meanwhile, the elongation is reduced slightly to 9.7 %. The increase in volume fraction and the reduction in the size of second phase help to improve the strength of the alloy but deteriorate its ductility according to the Orowan strengthening mechanism [37]. Compared with the T4-treated extrusion material, the tensile strength of the T6-treated extrusion materials is improved, and the UTS of AZ91-T6-EX reaches 355.9 MPa. However, due to the existence of the large number of brittle second phases in the T6-treated extrusion specimen, the elongation is decreased. The AZ91-CaO-T6-EX exhibits the best UTS and YS of 367.6 MPa and 320.2 MPa, respectively. Meanwhile, its elongation is essentially the same as that in AZ91-T6-EX alloy. The in-situ Al₂Ca and MgO result in the noticeably small grain size of AZ91-CaO-T6-EX alloy. At the same time, the nano-scaled and evenly distributed second phase particles further enhance the strength of the alloy. Both of the above factors contribute to the high strength of AZ91-CaO-T6-EX alloy.

Figure 6 gives the tensile fracture morphology of AZ91 and AZ91-CaO alloy in an extruded state. From the perspective of tensile fractures, the fracture forms of both materials are mainly ductile fractures. As shown in Figure 6a, the fracture of AZ91-T4-EX alloy is flatter, with less small mountain-like undulation, and there are also many dimple-like holes with large diameters. The reason for such morphology may be due to the small volume fraction and large size of the second phase in AZ91-T4-EX alloy. However, in AZ91-T6-EX alloy, many small dimple-like structures are generated in the fracture in addition to the large dimples, but the distribution of these small dimples is uneven. This improves the strength of AZ91-T6-EX alloy to some extent. In AZ91-CaO-T4-EX alloy, due to the small size and uniform distribution of the second phase, the fracture morphology is relatively flat. The small amount of large second phase particles reduces the number of large-diameter dimple-like structures. In AZ91-CaO-T6-EX alloy, there is almost no large-diameter dimple-like characteristic, and almost all of them are small dimple-like structures.

The change of mechanical properties is mainly attributed to the fine grain strengthening and the second phase strengthening. On the one hand, the microstructure of the extruded AZ91 and AZ91-CaO alloys is significantly different at the same state. AZ91-CaO alloy has a more complete degree of recrystallization, smaller grains, a more uniform distribution of the second phase and a larger number of second phases, leading to a higher mechanical property. On the other hand, in the same material, there are also obvious differences in the microstructure of samples in different states. The grain size of the T6-treated extrusion specimen is significantly smaller than that of the T4-treated extrusion specimen. Meanwhile, the reduction of the non-recrystallization area and the uniform structure are conducive to the improved performance of the T6-treated specimen.

3.4. Corrosion Resistance Analysis of the Pre-Heat Treated and Extruded Alloys

3.4.1. Electrochemical Performance

The material was placed in 3.5% NaCl solution for electrochemical experiments, and Figure 7 shows the electrochemical polarization curve and impedance diagram of the extruded AZ91 and AZ91-CaO alloys.

Table 4. Electrochemical results of AZ91-EX and AZ91-CaO-EX alloys.

Sample	Ecorr(V)	Icorr(µA/cm2)
AZ91-T4-EX	−1.243	21.56
AZ91-T6-EX	−1.221	5.34
AZ91-CaO-T4-EX	−1.096	17.79
AZ91-CaO-T6-EX	−1.086	5.65

gives the corresponding electrochemical data. As can be seen from Figure 7a and Table 4, although the self-corrosion potential of the alloys with the same composition has little difference, the self-corrosion potential of the T6-treated extruded (T6-EX) alloy is higher than that of the solid solution extruded (T4-EX) alloy. At the same time, the corrosion current density of the T6-EX alloy is significantly lower than that of the T4-EX alloy. The current density of AZ91-T6-EX is 75.2% is lower than that of AZ91-T4-EX and the current density of AZ91-CaO-T6-EX is 68.2% lower than that of AZ91-CaO-T4-EX. In addition, the self-corrosion potential of AZ91-CaO alloy is much higher than that of AZ91 alloy at the same state. However, the current density of AZ91-T6-EX is slightly lower than that of AZ91-CaO-T6-EX alloy. According to the AC impedance diagram of AZ91 and AZ91-CaO alloy in the extruded state, shown in Figure 7b, the arc resistance of AZ91-CaO-EX is greater than that of AZ91-EX in the same state. The order of capacitive-reactance arc of different alloys is AZ91-CaO-T6-EX > AZ91-T6-EX > AZ91-CaO-T4-EX > AZ91-T4-EX. Therefore, the corrosion resistance of AZ91-CaO alloy is better than that of AZ91 alloy, and AZ91-CaO-T6-EX has the best corrosion resistance, with a self-corrosion potential of −1.08 mV and corrosion current density of 5.65 μA/cm².

3.4.2. Corrosion Performance Analysis of Salt Spray

In order to further explore the corrosion performance of the material in the long-term test, the sample was placed in the salt spray testing machine for the salt spray test. Figure 8 shows the corrosion morphology of AZ91 and AZ91-CaO alloy after removing the corrosion products at different times. It can be seen from the figure that after 12 h of corrosion, strip corrosion traces appear on the surface of AZ91-T4-EX specimen, and slight corrosion appears on the surface and edge of AZ91-CaO-T4-EX specimen. Meanwhile, slight corrosion occurs on the edge of both of AZ91-T6-EX and AZ91-CaO-T6-EX alloys. At this time, all the samples present silver-white metallic luster. When the corrosion is carried out for 24 h, the corrosion is aggravated. There are obvious corrosion traces on the surface of all the extruded specimens, but there is no corrosion pit. The surface of AZ91-T4-EX and AZ91-CaO-T4-EX specimens are covered by a wide range of corrosion marks, while the surface of AZ91-T6-EX and AZ91-CaO-T6-EX are covered by only a smaller area of corrosion marks.

After 48 h of corrosion, the corrosion is further aggravated. The corrosion gradually spreads from the edge to the center of the AZ91-T4-EX sample. Most of its surface is corroded, while most of the surface of AZ91-T6-EX is still intact. The surface corrosion area of the AZ91-CaO-T4-EX is less than that of the AZ91-T4-EX specimen but is greater than that of the AZ91-T6-EX sample. The corrosion area of AZ91-CaO-T6-EX is not different from that at 24h, and obvious corrosion traces appear at the edge of the sample. At this point, the AZ91-T6-EX and AZ91-CaO-T6-EX specimens remain silvery-white, while the AZ91-T4-EX and AZ91-CaO-T4 specimens turn light gray.

After 72 h of corrosion, no corrosion pits appear on the surface of the extruded samples. The surface of AZ91-T4-EX is corroded severely, and the corrosion area accounts for 80.05% of the total sample area. The corrosion grade is determined as C0. The corrosion of AZ91-T6-EX is relatively mild and the surface of the sample exhibits a light gray color. The corrosion area reaches 65.95% of the total sample area; thus, the corrosion grade is determined to be C0. From the corrosion images at 48 h and 72 h, it can be seen that the corrosion rate of AZ91-T6-EX alloy is significantly accelerated. This may be due to the uneven size distribution of the second phase in AZ91-T6-EX alloy, which still contains larger Mg₁₇Al₁₂ phase, leading to the accelerated corrosion rate of AZ91-T6-EX alloy with the increase in salt spray time. After 72 h salt spray, the corrosion area accounts for 54.2% in AZ91-CaO-T4-EX alloy, and the corrosion grade is C1. The corrosion of AZ91-CaO-T6-EX is the least severe. Only a small range of corrosion occurs on the surface of the sample, and no corrosion pits appear. The corrosion area accounts for only 12.7%, and the corrosion grade is C2. For the AZ91-CaO extruded alloy, the corrosion area of the alloy does not change significantly after 72h of salt spray test, which may be attributed to the smaller grain size and uniform distribution of the second phase.

Figure 9 shows the weight loss and corrosion rate of the salt spray test at different times. It can be seen from the figure that the weight loss and corrosion rate of the extruded AZ91-CaO alloy changes more gently, which may be caused by the fine grain size and the uniform distribution of the fine second phases in extruded AZ91-CaO alloy. The order of weight loss and corrosion rate of extruded materials from high to low is AZ91-T4-EX > AZ91-T6-EX > AZ91-CaO-T4-EX > AZ91-CaO-T6-EX. Therefore, the results of weight loss and corrosion rate are consistent with the results of salt spray corrosion morphology. Among them, AZ91-CaO-T6-EX has the best corrosion resistance, and its corrosion rate was only 0.36 mm/y.

The electrochemical test and salt spray corrosion results show that the AZ91-CaO-T4-EX alloy exhibits better corrosion resistance due to the clear refinement of recrystallized grain size and the small and diffusely distributed second phase than that of AZ91-T4-EX alloy. After T6 treatment and subsequent extrusion, the recrystallization degree of the alloys is more complete and the grain size is finer. When corroding, the uniformly distributed second phase and the small grain size promote the formation of a compact passivation film, which could protect the matrix improve the corrosion resistance.

4. Conclusions

The effects of different heat treatment processes and subsequent extrusion on the microstructure, mechanical properties and corrosion resistance of AZ91 and AZ91-CaO alloy were studied. Conclusions were drawn as follows:

(1)

The microstructure of AZ91 alloy was clearly refined after adding CaO. The addition of CaO consumed Al element in the matrix and generated Al₂Ca phase. The distribution of coarse second phase changed from the continuous network structure in AZ91-Cast to the discontinuous state and size of them was clearly refined. After solution treatment, most of the Mg₁₇Al₁₂ were dissolved into the matrix, but there was still fishbone Mg₁₇Al₁₂ phase in AZ91-CaO-T4. After aging treatment, fine Mg₁₇Al₁₂ phase was precipitated at grain boundaries, and the Mg₁₇Al₁₂ in AZ91-CaO-T6 was even finer.

(2)

After extrusion, all the two alloys in different conditions underwent dynamic recrystallization, but the microstructures were different due to the different pretreatment processes. The size of grain and second phase of CaO-added alloys were smaller than that of extruded AZ91 alloy, while the T6-treated extruded alloy has a more fine and uniform microstructure compared with T4-treated extruded alloy.

(3)

The AZ91-CaO-T6-EX alloy has the best mechanical properties due to its small grain size, fine and dispersed second phase. The ultimate tensile strength, tensile yield strength and elongation reached 367.6 MPa, 320.2 MPa and 9.5%, respectively.

(4)

The corrosion resistance of the extruded material was significantly improved due to its small and diffusely distributed second phase and relatively uniform structure. The electrochemical test and long-term salt spray experiments revealed that the CaO-added alloy has a higher corrosion resistance than that of AZ91 alloy, while the corrosion performance of T6-treated alloy was better than that of T4-treated alloy. The AZ91-CaO-T6-EX alloy exhibited the optimal corrosion resistance among the extruded alloys.