Effect of Er on the Hot Deformation Behavior of the Crossover Al3Zn3Mg3Cu0.2Zr Alloy

Abstract

The effect of an erbium alloying on the hot deformation behavior of the crossover Al3Zn3Mg3Cu0.2Zr alloy was investigated in detail. First of all, Er increases the solidus temperature of the alloy. This allows hot deformation at a higher temperature. The precipitates resulting from the Er alloying of the Al3Zn3Mg3Cu0.2Zr alloy were analyzed using transmission electron microscopy. Erbium addition to the alloy produces the formation of more stable and fine L1₂-(Al₃(Zr, Er)) precipitates with a size of 20–60 nm. True stress tends to increase with a decline in the temperature and an increase in the deformation rate. The addition of Er leads to decreases in true stress at the strain rates of 0.01–1 s⁻¹ due to particle-stimulated nucleation softening mechanisms. The effective activation energy of the alloy with the Er addition has a lower value, enabling an easier hot deformation process in the alloy with an elevated volume fraction of the intermetallic particles. The addition of Er increases the strain rate sensitivity, which makes the failure during deformation less probable. The investigated alloys have a significant difference in the dependence of the activation volume on the temperature. The flow instability criterion allows better deformability of Er-doped alloys and enables the alloys to be formed more easily. The evenly distributed particles prevent the formation of shear bands with elevated storage energy and decrease the probability of crack initiation during the initial stages of hot deformation when only one softening mechanism (dynamic recovery) is working. The microstructure analysis proves that dynamic recovery is the main softening mechanism.

1. Introduction

Wrought Al-Zn-Mg-Cu alloys are widely used in the aviation and aerospace industries due to their high strength [1]. The highest content of alloying elements provides a great strengthening effect after aging treatment. The Zn/Mg ratio is more than 1 in the commercial Al-Zn-Mg-Cu alloys, ensuring a high yield strength (YS), but the casting properties, heat resistance, and corrosion resistance are low [1,2,3,4,5,6]. The Al-Zn-Mg-Cu alloys with a Zn/Mg ratio of 1 have medium strength and good castability, corrosion resistance, and heat resistance [1,3,4,5,6,7,8]. Another way to improve the casting properties is alloying using eutectic forming elements [7,8,9,10,11,12,13]. Scandium and zirconium are well-known for improving the mechanical properties of the Al-Zn-Mg-Cu alloysat room and elevated temperature, due to the nucleation of nano-sized precipitates of the L1₂-Al₃(Sc,Zr) phase during solution treatment [14,15,16,17,18,19,20,21]. There is a similar effect on the microstructure and properties of Al that may render the combination of Er and Zr unnecessary [22,23,24,25,26]. However, Er is significantly cheaper than Sc. At the same time, Er is not only a precipitate-forming element but also a eutectic-forming element [27,28,29,30,31,32,33]. Er-rich eutectic origin phases have an improved resistance to growth with high-temperature solution treatment, which contributes to achieving a high strength of the alloy [27,28,29,30,31,32,33]. The novel cast and wrought Al-3Zn-3Mg-3Cu-Zr-Y(Er) alloy with improved heat resistance was developed based on these principles [34]. Developed alloys fit into the basic concept of the crossover alloys based on the use of aluminum alloy scraps [35]. The crossover alloys cover the structure and properties of combined alloys of the different groups [36,37,38,39,40,41,42,43,44]. The developed Al-Zn-Mg-Cu-Zr-Y(Er) alloys combined the best properties from the cast A-Zn-Mg alloys, high-strength Al-Zn-Mg-Cu alloys, and heat-resistant Al-Cu alloys [34,44]. The presented features of the new alloys make them very promising for deep investigations.

The main way to improve the mechanical properties of the Al-Zn-Mg-Cu alloys is to control the grain structure under thermo-mechanical processing [14,45,46,47,48,49,50,51,52,53]. Xu et al. showed that dynamic recovery is the main dynamic softening mechanism in the Al-5.6Zn-1.9Mg-0.3Cu-0.09Sc-0.09Zr alloy; at the same time, three dynamic recrystallization mechanisms (discontinuous, continuous, and geometric) lead to grain refinement [54]. A lower value of the effective activation energy was obtained in the as-forged alloy [55] compared to the as-cast alloy [50]. Tang et. al. investigated the influence of Zn content on the hot deformation behavior of Al-xZn-2Mg-2Cu alloys: the increase in Zn concentration from 6.3 to 10.1% decreases the rate of dynamic recovery and dynamic recrystallization and decelerates dynamic softening [56]. However, the increase in Zn content does not have a significant influence on the static softening after hot deformation [57]. Zirconium content has a more significant influence on the static softening [58]. An elevated volume fraction of Al₃Zr dispersoids pinned dislocation and inhibited the dynamic recovery and/or recrystallization, resulting in a higher level of stored deformation energy in the hot deformed alloys. Consequently, more driving force for static softening is presented with increasing Zr additions. A similar influence of the hot deformation conditions (strain rate and temperature) on the post-deformation behavior was observed by Long et. al. [48]. The volume fraction of statically recrystallized grains is influenced by the stored deformation energy: at low deformation temperatures and high strain rates, the driving force for static recrystallization is higher. Hot deformation may also influence the phase composition of the alloys. Zhang et. al. has shown that hot rolling changes the morphology of Al₃Zr particles in 7055 aluminum alloys and leads to an increase in the dislocation density in subgrains and, consequently, results in an increased number and variety of quenching-induced phases [59].

Previously, a power tool for the investigation and modeling of hot deformation behavior called “processing mapping” was developed. This approach lets us determine the optimal deformation conditions using minimal experimental data. This method was successfully applied to the 5A06 [60], 2195-O [61], and ECO-7175 [62] aluminum alloys and the TiC [63], CNT [64], and ZrB₂ [65] aluminum-based composites.

The present investigation aims to determine the effect of Er alloying on parameters of the hot deformation of the Al-3Zn-3Mg-3Cu-0.2Zr alloy using the modeling of the 3D processing maps.

2. Materials and Methods

2.1. Alloy Melting

The Al-2.8Zn-2.7Mg-3Cu-0.2Zr-0.1Ti-0.15Si-0.15Fe (Al3Zn3Mg3Cu) and Al-2.9Zn-2.8Mg-3Cu-1.4Er-0.2Zr-0.1Ti-0.15Si-0.15Fe (Al3Zn3Mg3CuEr) alloys were melted in the graphite–chamotte crucibles in the resistance furnace. The melting and pouring processes are described in [34]. The samples for microstructure investigation and compression tests were cut out from the ingots that measured 20 mm in thickness, 40 mm in width, and 120 mm in height. The ingots were obtained in the water-cooling copper mold.

2.2. Sample Preparation and Structure Investigation

Evaluations of grain structure, microstructure, phase composition, and precipitates were conducted using a Neophot 21 optical microscope (OM) (Carl Zeiss AG, Oberkochen, Germany), a Tescan Vega 3 LMH scanning electron microscope (SEM) (Tescan, Brno, Kohoutovice, Czech Republic) with an XMax-80 electron diffraction X-ray (EDX) detector (Oxford Instruments Advanced AZtecEnergy, High Wycombe, UK), and a JEM2100 transmission electron microscope (TEM) (Jeol Ltd., Tokyo, Japan). The samples for microstructure investigation were grinded and polished using equipment from Struers Labopol (Struers APS, Ballerup, Denmark). The oxidation was applied to etch the grain structure. The oxidation proceeded in the Barker’s reagent [34]. The samples for TEM were electrochemically polished in the A2 electrolyte using equipment from Struers Tenupol (Struers APS, Ballerup, Denmark). The heat treatment was carried out in the Nabertherm furnace based on the differential calorimetry data obtained in [34]. The calculation of the volume fraction of the precipitates was performced in the Thermo-Calc software (https://thermocalc.com/, TCW5, Thermo-Calc Software AB, Stockholm, Sweden).

2.3. Hot Compression Testing

The hot compression was carried out using a hydraulic-type Gleeble 3800 (Dynamic Systems Inc., Poestenkill, NY, USA) thermomechanical simulator. The cylindrical samples with a diameter of 10 mm and a height of 15 mm were compressed at temperatures of 350 –500 °C and strain rates of 0.01, 0.1, 1, and 10 s⁻¹. Graphite foil between the sample’s edges and dies was used to minimize the influence of friction. Control of the temperature was carried out using a chromel/alumel thermocouple welded to the center of the sample. The error in temperature determination was about 1 °C. The heating rate was 5 °C/s. The sample was held for 30 s at the compression temperature for unification of the temperature distribution before deformation. The overall strain was about 1. The primary curves that were obtained were corrected for friction and adiabatic heating [66,67] using the OriginLab Software (Version 9.1, Northampton, MA, USA).

3. Results and Discussion

3.1. Microstructure Analysis of As-Cast and Solution-Treated Alloys

The grain structure and phase composition (size and volume fraction of solidification origin phases and precipitates) affected the hot deformation behavior, as was noted earlier. The solution treatment was applied to the ingots of the as-cast alloys before hot deformation. The initial grain microstructure and as-solution-treated phase composition were investigated before the compression tests. The detailed investigation of the microstructure and phase composition of the as-cast and solution-treated Al3Zn3Mg3Cu and Al3Zn3Mg3CuEr alloys is presented in [34]. An average grain size in the Er-free alloy is 100 ± 15 µm (Figure 1a). A part of erbium atoms may substitute the titanium and/or zirconium atoms in the primary Al₃(Zr,Ti) clusters. As a result, the fraction of the clusters may be increased. These clusters play the role of the substrates for the primary aluminum solid solution during solidification due to a similar lattice parameter. The erbium addition to the Al3Zn3Mg3CuEr alloy provides significant refining of the grain structure (Figure 1b). The average grain size of the Er-rich alloy is 45 ± 10 µm. The effectiveness of the erbium as a grain refiner has been demonstrated by many researchers [68,69,70,71,72,73].

The solidus temperature, which was determined by differential scanning calorimetry in [34], of the Al3Zn3Mg3Cu and Al3Zn3Mg3CuEr alloys is 493 and 477 °C, respectively. The homogenization annealing during solution treatment at 480 °C for 3 h provides for the dissolving of the non-equilibrium low-temperature melting phases and increases the solidus temperature in the Al3Zn3Mg3CuEr alloy only. The SEM microstructure of the alloys after solution treatment at 480 °C for 3 h is presented in Figure 2a,b. The solidus temperature of the Er-rich alloy increased to 533 °C after the first stage of homogenization annealing [34]. The second stage of homogenization annealing at 520 °C was performed for both alloys. Homogenization treatment at 520 °C is the burnout for the Er-free alloy. The low-temperature melting phases on the dendritic cell and grain boundaries were melted and solidified in the near-spherical particles (Figure 2c). The solidification origin intermetallic phases of the Al3Zn3Mg3CuEr alloy change the morphology to a more globular structure (Figure 2d). As a result, the final microstructure of the Al3Zn3Mg3Cu alloy after solution treatment at 480 °C for 3 h and before hot deformation was characterized by 10–11% of the solidification origin phases with a size of 0.5–3 µm (Figure 2a). The Er-rich alloy contains 11–12% of the solidification origin phases with a size of 0.5–4 µm (Figure 2d) after two-step solution treatment (480 °C for 3 h + 520 °C for 6 h).

An accompanying process during homogenization is heterogenization. Heterogenization is the decomposition of the supersaturated zirconium or zirconium and erbium aluminum solid solution. The decomposition of the aluminum solid solution under annealing of the as-cast alloy following with L1₂-Al₃M (M = Zr,Er) precipitates nucleation. The measured EDX SEM content of the zirconium in the aluminum solid solution is 0.2–0.3%. The additional 0.3% fraction of erbium was determined in the aluminum solid solution in the as-cast Al3Zn3Mg3CuEr alloy. The higher content of the precipitates that form elements in the Al3Zn3Mg3CuEr alloy must provide the higher volume fraction of the L1₂ precipitates. Erbium substitutes the zirconium atoms in the lattice of the L1₂-structured Al₃(Zr) precipitates [21,22,23,24,25,26,30,31]. The TEM microstructures of the alloys after solution treatment are presented in Figure 3. The Al3Zn3Mg3Cu alloy was solution-treated at 480 °C for 3 h, and the Al3Zn3Mg3CuEr alloy was solution-treated at (480 °C for 3 h + 520 °C for 6 h). The L1₂-(Al₃Zr) precipitates with sizes of 40–100 nm were nucleated in the Er-free alloy (Figure 3a) after 3 h of solution treatment at 480 °C. The fast Fourier transformation (FFT) confirms the L1₂ structure of the precipitates (insert in the right image in Figure 3a). Erbium addition in the alloy provides the formation of more stable and fine L1₂-(Al₃(Zr,Er)) precipitates after two-stage solution treatment with the higher temperature of the second stage. The L1₂-(Al₃(Zr,Er)) precipitates have a size of 20–60 nm.

The characteristics of the microstructure components are summarized in

Table 1. Parameters of the solution-treated microstructure.

Alloy	Grain Size, µm	Intermetallic Particles Size, µm	Intermetallic Particle Volume Fraction, %	Precipitate Size, nm	Precipitate Volume Fraction, %
Al3Zn3Mg3Cu	100	0.5–3	10–11	40–100	0.26
Al3Zn3Mg3CuEr	45	0.5–4	11–12	20–60	0.35

. The grain size, volume fraction, and parameters of the intermetallic particles of the solidification origin, and precipitates were measured via the secant method using images of Figure 1, Figure 2 and Figure 3. The volume fraction of the precipitates was calculated using a multicomponent phase diagram of the Al-Zn-Mg-Cu-Zr system and binary Al-Er system. The fraction of the Al₃(Zr,Er) precipitates was calculated as the sum of the volume fraction of the Al₃Zr phase in the Al-3Zn-3Mg-3Cu-0.2Zr alloy and the Al₃Er phase in the Al-0.3Er alloy. The microstructure after solution treatment is the initial one for hot deformation. The differences in the microstructure characteristics may affect the hot deformation behavior. The main difference between the two alloys is the solidus temperature. The erbium addition provides an increase in the alloy solidus and increases the possibility of increasing the temperature of hot deformation.

3.2. Hot Deformation Behavior

The investigated alloys show typical hot deformation behavior (Figure 4 and Figure 5). True stress tends to increase with a decline in the temperature and an increase in the deformation rate. The diffusion process mainly determines the non-conservative movement of the dislocations, which is the main mechanism of deformation at elevated temperatures. As a result, the increase in temperature accelerates diffusion and makes the deformation easier. At the same time, at low strain rates, the dislocations have more time for sliding and creeping to other planes, which leads to decreases in the stress. All curves show similar shapes with a peak in stress. Three main processes of the microstructure evaluation may proceed under hot deformation: strain hardening, dynamic recovery, and dynamic recrystallization. The strain hardening and dynamic recovery took place before peak stress was reached. Dynamic recrystallization is the main process that underlies the decrease in the stress. At the last stage of the deformation, all processes are mutually compensated, and the steady state has a constant stress value. The position of the peak depends on the temperature and strain rate. Following the general trend, the strain for peak stress is shifted to the larger values, with increases in the strain rate and decreases in the temperature. The addition of Er leads to decreases in true stress at strain rates of 0.01–1 s⁻¹ due to particle-stimulated nucleation-softening mechanisms. At a high strain rate, the process of non-conservative dislocation movement is inhibited for both alloys.

The hot deformation behavior of the alloys may be compared in terms of the effective activation energy (Q). The Q-value is usually determined by the functional dependence of the stress (S) at a steady state and the Zener–Hollomon parameter (Z) [74]:

$$Z=\dot{\epsilon }{e}^{{\displaystyle \frac{Q}{RT}}}$$

(1)

where $\dot{\epsilon }$ and T are strain rate (s⁻¹) and temperature (K). The hyperbolic sine is the universal law that describes the dependence of the Z-parameter at all stress values:

$$Z=A_3{\left[\mathit{sinh}\left(\alpha S\right)\right]}^{n_2}$$

(2)

where A₃, n₂, and α are the material’s constants. However, the approximate determination of the α constant needs the construction of particular dependencies between stress and the Z-parameter, such as the exponential law for high stresses:

$$Z=A_2{e}^{\beta S}$$

(3)

and the power law for a low level of stress:

$$Z =A_1{S}^{n_1}$$

(4)

where the material’s constants A₁, n₁, A₂, and β should be determined using the values of the flow stress.

The value of α may be found using the following formula:

$$\alpha \approx {\displaystyle \frac{\beta }{n_P}}$$

(5)

The value of the Q-value was determined by the least squares method after the logarithmization of Equations (2)–(4). The values of the stress during the steady state were used for the determination of the material constants. Comparisons of the experimental and predicted true stress values using Equation (3), as well as constants of the constitutive equation, are presented in Figure 6. The value of Pearson’s coefficient for both collations is close to 1, indicating that the constructed model is accurate. The absolute average relative error has a value of less than 1%. The effective activation energy for the alloy with the Er addition has a lower value, suggesting that the hot deformation process is easier with the alloy with an elevated volume fraction of intermetallic particles.

The addition of Er to the alloys also influenced another hot deformation characteristic—the activation volume [75]:

$$V_a=3\sqrt{3}kT({\displaystyle \frac{\partial \left(ln\dot{\epsilon }\right)}{\partial S}})$$

(6)

where k is the Boltzmann constant (k = 1.38·10⁻²³ J/K).

This parameter quantifies the minimal volume that is necessary for the process of elemental deformation. The dependence of the activation volume on the temperature for both alloys at a steady state is shown in Figure 7. The activation volume increases with an increase in the temperature, which can be explained by a decrease in the dislocation density as the temperature increases. At low temperatures, V_a values for the Er-free alloy are higher than for the Al3Zn3Mg3CuEr alloy, which can be explained by the additional accumulation of the dislocations at the intermetallic particles. However, at temperatures higher than 450 °C, the activation volume of the alloy with Er becomes significantly higher due to a decrease in the dislocation density caused by particle-stimulated nucleation dynamic recrystallization.

One more important parameter for hot deformation behavior analysis is the strain rate sensitivity coefficient:

$$\mathrm{m} ={\displaystyle \frac{1}{n_1}},$$

(7)

where n₁ is the coefficient from Equation (4). The value of this coefficient is 0.174 ± 0.002 for the Al3Zn3Mg3Cu alloy and 0.181 ± 0.002 for the alloy with the Er addition. A larger value of the coefficient denotes a larger sensitivity to the strain rate increases and the prevention of deformation localization. The absence of the local large dislocation density (a larger strain) advances the deformation process and prevents premature failure. It shows an elevated technology adaptivity of the alloy with the Er addition.

3.3. Hot Processing Maps

For comparison of the hot formability of the alloys, the processing maps approach from [76] was also applied in this study. The approach involves the calculation of two criteria: the efficiency of power dissipation (η) and flow instability (ξ). The first criterion may be calculated as

$$\eta ={\displaystyle \frac{2m}{m+1}}.$$

(8)

This criterion shows the part of the energy that dissipates in the material due to microstructural softening, such as dynamic recovery and dynamic recrystallization. The usual values of these parameters for aluminum alloys are in the range of 10–40% and are dependent on deformation temperature, strain rate, and strain.

The second criterion shows the instability of flow during the deformation and also may be calculated using the strain rate sensitivity coefficient value:

$$\xi \left(\dot{\epsilon }\right)=\left[{\displaystyle \frac{\partial \ln \left(\frac{m}{m+1}\right)}{\partial \dot{\ln \epsilon }}}\right]+m.$$

(9)

Its positive value shows that the deformation proceeds evenly without significant material flow localization and may provide potential places for local crack formation, stress concentration, and, consequently, material fracturing.

Comparisons of processing maps of different deformation parameters for both alloys is shown in Figure 8. Both alloys have a wide region of good energy dissipation (more than 30%) at high temperatures and low strain rates. However, the flow stability significantly differs. Compression is one of the softest deformation processes. The deformation of the Er-free alloy at strain rates higher than 0.1 s⁻¹ may lead to flow localization. The flow localization may produce a fracture in deformation conditions that are harder than compression. At the same time, the $\xi \left(\dot{\epsilon }\right)$-value for the Er-doped alloy is lower than 0 only at high strain rates and temperatures below 420 °C, indicating that the alloy is better able to form under these conditions. The evenly distributed particles prevent the formation of shear bands with an elevated level of storage energy and decrease the probability of crack initiation during hot deformation during the initial stages when only one softening mechanism (dynamic recovery) is working.

Evaluations of the grain structure of the Al3Zn3Mg3Cu and Al3Zn3Mg3CuEr alloys under compression at different temperatures and strain rates are shown in Figure 9 and Figure 10. The polarized light in the OM makes the different grains appear as different colors. In general, the deformed non-recrystallized structures are visible in all conditions. The microstructure proves that dynamic recovery is the main softening mechanism. Figure 11 illustrates a comparison of the grain structure of the Al3Zn3Mg3Cu and Al3Zn3Mg3CuEr alloys after deformation at 450 °C and a strain rate of 0.01 s⁻¹. The thin boundaries are seen as one deformed grain. The observation that the areas divided by these boundaries are the same color may indicate that these areas have very similar crystallographic orientations. In other words, the subgrain boundaries are clearly seen in the one deformed grain of both alloys. The slightly finer grain and substructure were formed in the Er-rich alloy due to finer precipitates with higher volume fractions (Figure 3 and Table 1). The L1₂-(Al₃(Zr,Er) precipitates effectively suppress the recrystallization. By comparison, the static recrystallization in the cold rolled alloys proceeds at 350 °C after 1 h of annealing, and erbium provides a finer grain structure [34].

To summarize, we can provide three positive points of the Er alloying effect on the deformation behavior of the Al3Zn3Mg3Cu alloy:

-

Increasing the solidus temperature increases the hot deformation temperature;

-

Increasing the volume fraction of the solidification origin particles decreases the effective activation energy;

-

Increasing the volume fraction of the finer L1₂ precipitates suppresses recrystallization and provides stable hot deformation.

4. Conclusions

The effect of erbium as a eutectic and a precipitate-forming element on the hot deformation behavior of the novel crossover Al3Zn3Mg3Cu0.2Zr alloy was investigated in detail using optical, scanning, and transmission electron microscopy. The initial grain structure and as-solution-treated phase composition were investigated, and 3D processing maps were constructed after compression tests. The most important conclusions are presented below.

• The formation of precipitates during the Er alloying of the Al3Zn3Mg3Cu0.2Zr alloy was investigated. Erbium addition to the alloy enabled the formation of more stable and fine L1₂-(Al₃(Zr,Er)) precipitates, with sizes of 20–60 nm after two-stage solution treatment and higher temperatures during the second stage.
• The investigated alloys demonstrated typical hot deformation behavior. True stress tended to increase with a decline in the temperature and an increase in the deformation rate. The addition of Er led to decreases in true stress at strain rates of 0.01–1 s⁻¹ due to particle-stimulated nucleation-softening mechanisms.
• The effective activation energy for the alloy with the Er addition had a lower value, allowing the hot deformation process to proceed more easily in the alloy with an elevated volume fraction of intermetallic particles.
• The strain rate sensitivity coefficient was higher in the alloy with the Er addition due to elevated particle-stimulated softening.
• The activation volume for deformation of the Er-free alloy was higher at low temperatures and lower at elevated temperatures, which can be explained by different rates of hardening and softening due to the presence of more intermetallic particles.
• The investigated alloys had a wide region of good energy dissipation at high temperatures and low strain rates. The flow instability criterion for the Er-doped alloy had a negative value only at high strain rates and temperatures below 420 °C, indicating that the alloy was better able to form under these conditions. The evenly distributed particles prevented the formation of shear bands with elevated storage energy and decreased the probability of crack initiation during hot deformation during the initial stages when only one softening mechanism (dynamic recovery) was working. The microstructure proved that the main softening mechanism is dynamic recovery.