The Effect of Interrupted Homogenization on β-Al₅FeSi → α-Alx (Fe and Mn) Si Transformation in the A6063 Aluminum Alloy

Abstract

The aluminum alloys corresponding to the 6000 series are mainly manufactured by mechanical forming processes. Their properties are enhanced by the homogeneous distribution of intermetallic phases such as β-Al₅FeSi or α-Al_x (Fe, Mn) Si. By thermal homogenization treatment, the intermetallic compound β-Al₅FeSi changes its morphology from a needle type with a monoclinic structure to an acicular form known as α-Al₁₂(Fe, Mn)₃Si with an fcc structure. In the present study, samples of the 6063 alloy were subjected to different temperatures of homogenization (798, 823, and 848 K) and treatment times from 0 to 660 min (in intervals of 30 min) to evaluate their effects on the microstructures and morphologies of the intermetallic phases. For the kinetic study, the microstructures of the β and α intermetallic phases were quantified using the Image-Pro software. The results indicate that as the temperature and homogenization time increase, the percentage of phase α also increments. The results of the kinetic analysis revealed that the β → α transformation is controlled by two stages; the first corresponds to the diffusion of Mn atoms from the matrix to the interface of reaction for the formation of the intermetallic phases, while the second corresponds to the nucleation and growth of the iron- and manganese-rich intermetallic phases.

1. Introduction

The aluminum alloy A6063 is composed mainly of magnesium and silicon as alloying elements. These elements confer improved conditions for enhancing mechanical properties during forming processes. Moreover, artificial aging can greatly harden such alloys, as they are widely used in architectural, marine, and sometimes aerospace parts [1]. The dendrites of the A6063 alloy in the as-cast condition contain particles of the intermetallic compound β-Al_xFeSi, which has a needle-like morphology with a hexagonal crystallographic structure (hcp). This fact makes it brittle during plastic deformation [2,3]. Such deformation means that the alloy should be used in a heat-treated condition. Homogenization heat treatment consists of heating up the metal to produce supersaturating conditions, maintaining a constant temperature to ensure the dissolution of most second-phase precipitates, and cooling the metal down at a certain speed. The holding time and the cooling rate have a great influence on the microstructure of the thermally treated pieces [4,5].

On the other hand, the β-α transformation is important, as β particles are considered responsible for several surface defects in subsequent mechanical forming processes, such as hot cracking [6,7]. Al-Si-Mg alloys form several iron-rich intermetallic compounds such as β-Al_xFeSi and α-Al_x(Fe, Mn)_ySi. The compound Mg₂Si is also precipitated during aging thermal treatment [8,9,10]. During the homogenization stage, a phase transformation occurs in the A6063 alloys. The intermetallic phase β-Al_xFeSi is transformed into intermetallic particles of α-Al_x(Fe, Mn)_ySi. This phase transformation improves the formability of these aluminum alloys considerably. The β particles can lead to the initiation of cracks and cause surface defects in the extruded material. During the first stage of homogenization, α particles have been found to nucleate at the edges of β particles [11,12]. Some authors have reported that the β phase not only changes the alloy’s shape, but also changes its crystalline structure from hexagonal to a face-centered cubic crystal lattice [13]. The dissolution of precipitates and the transformation of intermetallics during homogenization are controlled by atomic diffusion.

The mass transport of each alloying element is accomplished by a concentration gradient between the aluminum matrix and its interface with the β intermetallic compound. Because there is an interfacial difference in terms of concentration between the aluminum matrix and the β intermetallic phase, here the chemical elements are provided by the β intermetallic phase; these are diffused through the aluminum matrix towards the α particles, and with these attributes, the α phase grows [14].

During the homogenization heat treatment of aluminum alloys, several microstructural changes occur, such as the dissolution of precipitates and the transformation of intermetallics. Therefore, microstructural changes are controlled by diffusion of the alloying elements. The mass transport of the alloy elements can be explained by its diffusion coefficients (K).

In the kinetic study of chemical reactions, the velocity equation that describes the degree of conversion of reactants or formation of products at a certain time is determined. Experimental data are compared with theoretical kinetic expressions to determine which one of the velocity equations describes the experimental measurements in a more precise way.

Kinetic analysis determines mainly the influence of temperature on the reaction rate. The main basis of kinetic analysis is a study of the transformed fraction (θ) as a function of time and temperature. The mathematical models that have been proposed for solid-state reactions can be divided into four groups: diffusion models, nucleation and growth models, contraction models, and reaction-order models [15,16,17]. When a model is selected, the next step is to determine the kinetic parameters, which are determined by analytical methods, i.e., mathematical models that describe solid-state reaction kinetics [18].

The transformed fraction (θ) changes as the reaction takes place, until the reagents are transformed into products. During each chemical change, the stoichiometry and kinetics of each stage can be analyzed separately. In a reaction, the value of (θ) can be calculated from the reduction of the area of the compound intermetallic β to time (t), related to the total decrease of the area, which corresponds to the end of the reaction. The transformed fraction (θ) is determined using Equation (1):

$$\left(\mathsf{\theta }\right)=\frac{{\mathsf{\theta }}_1}{\left({\mathsf{\theta }}_0+{\mathsf{\beta }}_0\right)}\ast 100$$

(1)

where θ₀ is the initial transformed fraction of phase α, θ₁ is the percentage of phase α transformed at time t, and β₀ is the percentage of phase β. The term activation energy is considered as the energy barrier to be overcome to allow phase transformation to take place.

The rest of the paper is organized as follows. In Section 2, the materials and methods used in this paper are presented. Section 3 provides the results and discussion regarding the microstructure of the A6063 alloy after heat treatment homogenization, the quantification of the α and β phases, and kinetic analysis. Finally, in Section 4, we draw conclusions from the experimentation carried out.

2. Materials and Methods

To prepare the A6063 alloy, recycled aluminum (burr, profile, pipe, etc.) was melted at a temperature of 1023 K, to obtain a homogeneous chemical composition. For the preparation of the A6063 aluminum pre-forms, an induction furnace with a capacity of 14 kg was used. Argon (UAP) was injected through a graphite lance for 10 min for degassing. At the end of the degassing time, the injection lance was withdrawn, and the slag was removed to fill the mold with liquid metal. The mold was preheated to 400°C to ensure its filling with liquid metal, so as to obtain the pre-forms.

The pre-forms have a rectangular shape and dimensions of 15″ × 4″ × 0.5″. The pouring temperature used for the experiments was 1023 K. Samples with dimensions of 7.5″ × 2″ × 0.5″ were obtained from the pre-forms, which were subjected to homogenized heat treatment.

The chemical composition of the aluminum alloy A6063 used in this work is presented in

Table 1. Chemical composition of aluminum alloy A6063.

Element	Si	Fe	Cu	Mn	Mg	Cr	Al
wt. %	0.439	0.238	0.0218	0.0501	0.471	0.0275	98.77

.

To perform the kinetic analysis of the β → α transformation, the quenching method was used, which consists of subjecting the material to homogenization treatment (solubilization) at temperatures of 798, 823, and 848 K, varying holding times from 0 to 660 min (at intervals of 30 min), and subsequently quenching the material in water at room temperature to retain the phases β or α obtained in each quenching condition.

On the other hand, to evaluate the effect of these variables on the transformation of β → α in the A6063 alloy, the samples were prepared for metallographic analysis, subjecting them to grinding with abrasive discs (60, 80, 120, 320, 500, 800, 1200, and 2400); they were then polished with diamond paste of 3 µm and 1 µm and finished with mirror-like finishing. An Olympus™ optical microscope was used to obtain photomicrographs of the intermetallic compounds precipitated during the thermal treatment of interrupted homogenization.

Samples were chemically attacked with Murakami reagent, which was heated to 348 K to reveal both α-phases of β, which have different colors, after chemical attack with the reagent. For the quantification of phases β and α for each treatment condition, the Image-Pro software was used, determining the percent of each of the phases obtained as a function of time for every experimental condition imposed.

3. Results and Discussion

3.1. Microstructure of an A6063 Alloy in the As-Cast Condition

To carry out the quantification of the phases α and β, it is necessary to analyze the microstructures obtained. Figure 1 shows photomicrographs of the alloy A6063 in the as-cast condition; images were taken at 500× and 1000×. Elongated precipitates in the form of needles or lamellas are observed, with sharp edges.

This type of morphology is typical of the intermetallic compound β-Al₅FeSi, according to the quantification performed using the scanning electron microscope, as shown in the attached table in Figure 2.

3.2. Microstructure of Samples with Homogenization Heat Treatment

The samples treated after the application of the afore mentioned conditions were subjected to chemical attack with the Murakami reagent at 348 K. This reagent has the particularity of revealing different shades or contrasts of the present phases. The intermetallic compounds α-Al₁₂(Fe, Mn)₃Si and β-Al₅FeSi are shown in Figure 3, as well as the patterns obtained from the semi-quantification of the phase α-Al₁₂(Fe, Mn)₃Si by EDS analysis in the SEM.

Figure 4 shows micrographs of samples subjected to different temperatures and treatment times. It is observed that at 798 K, after a period of 30 min, only the intermetallic compound β is present, so the transformation of β → α has not started (Figure 4a). As time increases and after 300 min of treatment, the first α-Al₁₂(Fe, Mn)₃Si precipitates are observed at the edges of the β-Al₅FeSi particles (Figure 4b). As the holding time increases to 480 min, a greater amount of precipitate of the α-Al₁₂(Fe, Mn)₃Si phase is observed. However, the transformation is not 100% completed (Figure 4c).

It is observed that at 823 K, after a period of 30 min, only the intermetallic compound β is present, so the transformation of β → α has not started (Figure 4d). In the range of 0 to 300 min, only the presence of the intermetallic compound β is observed (Figure 4e). At 480 min, it is found that the compound α-Al₁₂(Fe, Mn)₃Si occurs in a greater quantity than β-Al₅FeSi, close to a conversion of 100% (Figure 4f). At 848 K, after a period of 30 min, only the intermetallic compound β is present, so the transformation of β → α has not started (Figure 4g) from 0 to 300 min the same phenomenon occurs as at the two previous temperatures, that is, only the intermetallic compound β is present. After 300 min, the first precipitates of α-Al₁₂(Fe, Mn)₃Si are observed (Figure 4h). As time increases, β-Al₅FeSi decreases, and the increment of the phase α is evident (Figure 4i). However, 100% transformation is not achieved, as some particles of β-Al₅FeSi still can be observed.

3.3. Quantification of β-Al₅FeSi and α-Al₁₂(FeMn)₃Si Phases

It is established that the area of the β phase (as cast condition) is of the order of 2.5%, which will be considered as the initial amount of this phase, from which the percentage of transformation to α phase is determined. For this purpose, 9 fields of the sample are evaluated, to have a more precise result of each percentage of the β and α intermetallic compounds. In Figure 5, photomicrographs of heat-treated samples used in the quantification of β and α phases are presented. The different tonality in the colors is observed, which is obtained by attacking the samples with Murakami’s reagent, which allows quantifying each of the phases at the experimental conditions used, by means of the Image-Pro Software.

Table 2. Results obtained for the percentages of phases β and α, determined by the Image-Pro Plus software at different temperatures of homogenization treatment.

Time (min)	Temperature (K)
	798 K		823 K		848 K
As Cast	β (%)	α (%)	β (%)	α (%)	β (%)	α (%)
0	2.5	0	2.5	0	2.5	0
30	2.5	0	2.5	0	2.5	0
60	2.5	0	2.5	0	2.5	0
90	2.5	0	2.5	0	2.5	0
120	2.5	0	2.5	0	2.5	0
150	2.5	0	2.5	0	2.5	0
180	2.5	0	2.5	0	2.5	0
210	2.5	0	2.5	0	2.5	0
240	2.5	0	2.5	0	2.3	0.2
270	2.5	0	2.36	0.14	2.2	0.3
300	2.4	0.1	2.2	0.3	2.0	0.5
330	2.4	0.1	2.1	0.4	1.8	0.7
360	2.3	0.2	2.07	0.43	1.74	0.76
390	2.2	0.3	2	0.5	1.60	0.9
420	2.2	0.3	1.9	0.6	1.2	1.3
450	2.1	0.36	1.8	0.7	0.6	1.90
480	2.07	0.43	1.72	0.78	0.2	2.30
510	2.02	0.48	1.66	0.84	0.19	2.31
540	2.0	0.50	1.62	0.88	0.18	2.32
570	1.99	0.51	1.59	0.91	0.17	2.33
600	1.95	0.55	1.55	0.95	0.15	2.35
630	1.90	0.60	1.50	1	0.13	2.37
660	1.87	0.63	1.45	1.05	0.10	2.40
TOTAL (%)	74.8	25.2	58	42	4	96

shows the results obtained for all the samples, including the percentages of each intermetallic particle of both β and α at the indicated conditions.

Table 2 shows the percentage of the transformed fraction (θ) of [α-Al₁₂(Fe, Mn)₃Si], which increases with time and temperature. The results are used to construct plots of percentages of phases β and α for each time and temperature, as shown in Figure 6.

Additionally, the Brinell hardness test was performed on each of the samples. The results are shown in

Table 3. Values obtained of Brinell hardness (HB) of the studied alloy at different times and temperatures of homogenization treatment.

Treatment Temperature 798 K		Treatment Temperature 823 K		Treatment Temperature 848 K
Time (min)	Hardness (HB)	Time (min)	Hardness(HB)	Time (min)	Hardness (HB)
As-Cast	64.6	As-Cast	64.6	As-Cast	64.6
30	64	30	64.4	30	64.2
60	64	60	64.2	60	64
90	60	90	61.3	90	64
120	60	120	56.8	120	63.8
150	59	150	5.4	150	62.5
180	58	180	53.1	180	62.1
210	58	210	50.9	210	59.9
240	58	240	50	240	54.8
270	58	270	46.7	270	50.1
300	55	300	44.4	300	44.4
330	50	330	40.9	330	40
360	47	360	39.8	360	38.1
390	47	390	37.3	390	36.2
420	42	420	34.4	420	35.5
450	42	450	34.2	450	31.4
480	42	480	30.9	480	29.9
510	41.5	510	28.5	510	29
540	40.6	540	28.1	540	28.6
570	39.5	570	27.9	570	28
600	38.1	600	27.5	600	27.1
630	37	630	27.2	630	25
660	35	660	27	660	24

. It is observed that as the temperature and treatment time increase, the hardness values decrease. This is because the percentage of the alpha phase increases, while the percentage of the beta phase decreases. The alpha phase has a face-centered cubic crystal structure, while the beta phase has a close-packed hexagonal crystal structure, which makes the alpha phase more ductile, and thus the hardness values decrease.

It can be seen from the results reported in this table that the holding time and temperature have an important effect on the hardness values, as more ductile alloys are obtained at longer holding times and higher temperatures. This effect is mainly due to the β → α phase transformation, which later will be explained in detail.

3.4. Kinetic Analysis

To determine the speed of the transformation reaction and to establish the kinetic mechanism controlling the process, mathematical models for solid-state reactions were evaluated. To perform the kinetic analysis, the transformed fraction values were replaced in the equations of the mathematical models reported in the literature for homogeneous reactions. The values of α are replaced in the g(θ) functions of these models to determine which function best matches the experimental results obtained. It is determined that two mathematical models fit the behavior of the experimental data, which correspond to the diffusion model (the Jander equation, Equation (2)) and nucleation and growth through the PN energy law (Equation (3)).

$$\mathrm{g}\left(\mathsf{\theta }\right)={[1-{\left(1+\mathsf{\theta }\right)}^{1/3}]}^2$$

(2)

$$\mathrm{g}\left(\mathsf{\theta }\right)={ \mathsf{\theta }}^{1/\mathrm{n}}$$

(3)

The value of the constant (n) in the kinetic models varied in the range from 1 to 3, resulting in the value of 1; this fits the curves better, as shown in Figure 7a. The diffusion model describes the first part of the system from 0 to 330 min, while the nucleation and growth model describe the second part from 330 min to 660 min, as presented in Figure 7b. During the first stage (0–330 min), diffusion of manganese atoms from the bulk towards the edges of the intermetallic compound β-Al₅FeSi occurs. The thermodynamic conditions for the formation of groups of Mn atoms achieving the critical radius are then attained. Therefore, the nucleation and growth of the intermetallic compound α-Al₁₂(Fe, Mn)₃Si are presented.

The values of the rate constants of the transformation reaction at different temperatures are calculated, as is the value of activation energy of the system. To determine the rate constant of the transformation reaction, the value of k is determined from the function g(θ) = kt of the selected mathematical models.

According to the kinetic analysis of the transformation process from the beta phase into the alpha phase, there are two stages, described as follows:

• The process is controlled by the diffusion of Mn atoms from the bulk to the intermetallic beta interface or boundary layer.
• The process is controlled by the nucleation and growth of the alpha phase at the matrix–intermetallic beta interface.

With the values of the reaction constants for each stage of the process, it is possible to calculate the activation energy value. Because the value of the reaction constant is known for each temperature, it is possible to obtain the value of its natural logarithm and calculate the inverse of the corresponding temperature. The data are tabulated and plotted for each stage. To determine the activation energy of the two stages of the process, the Arrhenius equation is used [19].

Table 4. Reaction constants calculated at each homogenization treatment temperature, with its respective values of Ln K and 1/T.

Diffusion
T (K)	K	ln K	1/T
798	3.22 × 10−4	−8.04095	1.253 × 10−3
823	2.07 × 10−3	−6.18020	1.215 × 10−3
848	3.25 × 10−3	−5.7291	1.179 × 10−3
Nucleation and growth
798	1.32 × 10−4	−8.9360	1.253 × 10−3
823	2.78 × 10−4	−8.2091	1.215 × 10−3
848	4.81 × 10−4	−7.3822	1.179 × 10−3

shows the calculated values of reaction constants with their respective values of ln K versus 1/T.

The transformation values of the reaction constant at different temperatures allowed the construction of Arrhenius graphs, which are depicted in Figure 8. From these graphs, the values of the activation energy for each mathematical model describing the behavior of the experimental data were obtained. The activation energy value was 94.13 kJ/mol for the nucleation and growth model, while a value of 255.92 kJ/mol was obtained for the diffusion model.

Figure 9 shows the two mathematical models that fit the experimental data. The first model corresponds to the three-direction diffusion model (the Jander equation), in which the Mn atoms diffuse from the bulk to the beta phase interface. This phenomenon occurs in the time interval from 0 to 300 min. The second model corresponds to the nucleation and growth of the alpha phase, which begins at the matrix–beta phase interface, presenting growth towards the interior; this phenomenon occurs from 300 to 660 min. It should be mentioned that as the time increases, the percentage of the alpha phase also increases, reaching up to 98% at the end of the treatment.

4. Conclusions

The homogenization treatment transforms the as-cast structure, providing better mechanical properties to the material. The intermetallic particles of β-AlFe₅FeSi are transformed into α-Al₁₂(Fe, Mn)₃Si particles, and this is favored by the permanence time. The α-Al₁₂(Fe, Mn)₃Si particles improve the formability of the A6063 alloy.

The transformation from β → α of intermetallic compounds occurs at the edges of β intermetallic particles. With the increase of the temperature, there is a shorter time of permanence in the mentioned transformation.

The kinetic study revealed that the transformation of the intermetallic compound from β → α is governed or controlled by two stages. The first stage corresponds to the diffusion model (D); this stage is essential for such a transformation to occur. The second stage corresponds to the nucleation and growth model.