MATERIALS TRANSACTIONS
Online ISSN : 1347-5320
Print ISSN : 1345-9678
ISSN-L : 1345-9678
Foaming Process and Properties of 6063 Aluminum Foams by Melt Foaming Method
Tong ShiXiang ChenYing ChengYanxiang Li
Author information
JOURNAL FREE ACCESS FULL-TEXT HTML

2017 Volume 58 Issue 2 Pages 243-248

Details
Abstract

Aluminum foams made by 6063 aluminum alloy have been prepared by melt foaming method. The uniformity of foam pores is much concerned for that homogeneous pores attribute to excellent properties. It is mainly up to the stirring intensity when adding the blowing agent into the melt, and the cooling rate of the liquid foams. In this paper, 6063 aluminum foams have been fabricated by different stirring times of TiH2 at different cooling conditions. Microstructure, cell size uniformity and compressive property of the foams have been studied. The melt which mechanically stirred for 6 minutes before foaming and the foam cooled in the air show homogenous pore structure and good mechanical property. The use of 6063 aluminum alloy offers a new raw material to prepare foams with a uniform cell size distribution as well as good mechanical property.

1. Instruction

Aluminum foam, which was proposed by De Meller for the first time in the year of 1925, is a kind of cellular material combined with aluminum or its alloys and pores.1,2) The unique structure of aluminum foams brings about a series of superior properties, such as lightweight, high energy absorption capacity, high sound absorption capacity, fire resistant property, good electromagnetic shielding property and so on.36) Aluminum foams therefore have great application potential in aviation, aerospace, building industry and automobile industry.79) Among numerous manufacturing technologies, melt foaming method, also named as Alporas route, is widely adopted at the advantages of its simple fabrication process and industrialized mass production.10,11)

Various raw materials of aluminum foams have been focused on in recent years, such as pure aluminum, Al-Si alloy, Al-Mg alloy and so on.1218) However, reports on 6063 aluminum foams fabricated by melt foaming method are few. Liu, etc. have examined the impact behavior, deformation mode and energy absorption ability during axial crushing of 6063 aluminum foams, which showed that the dynamic impact contained two steps: initial compression and gradual crushing, and specimens deformed in “V”-shape or inverted “V”-shape mode among different density values.19) B.Y. Hur pointed out that porosity and pore size were uniformly distributed with the decreasing in foaming temperature of 6063 aluminum foam.20) There are some advantages for using 6063 aluminum alloy as raw materials. Firstly, reducing the surface tension of the melt by adding Mg can reduce the interfacial energy between pores and melt, so that pores are stabilized not to combine together and pore diameter is decreased.21) But too low the surface tension will reduce the melt's stability and incur collapse.22) The amount of Mg (0.45–0.9 mass%) element in 6063 alloy is just appropriate to reduce the surface tension a little. Secondly, liquidus temperature of raw materials applied in melt foaming method should meet the rapid decomposition temperature of the blowing agent. On one hand, the decomposition of blowing agent is speeding up with the raising temperature, so that a higher foaming temperature leads to excess decomposition of blowing agent; on the other hand, a lower foaming temperature helps enhance the stability of aluminum foams for that the molten viscosity increases with the decreasing temperature.23) According to Al-Si phase diagram, 0.2–0.6 mass% Si in 6063 aluminum alloy can help decrease the liquidus temperature of the melt to adapt the condition well. Meanwhile, the strengthening phase Mg2Si in 6063 alloy helps increase the compressive strength, so that the energy absorption is superior to that of pure aluminum foam.19) Taking these factors into consideration, it is appreciable for choosing 6063 aluminum alloy as a kind of raw material for foaming.

The uniformity of foam pores is much concerned for that homogeneous pores attribute to excellent properties.23) It is mainly up to the stirring intensity when adding the blowing agent into the melt.24) Besides, cooling rate is another important factor for which it influences not only the pore size but also the grain size in the cell walls of the foams.25) Moreover, the cooling rate determines the solidification time so that the macrostructure of the foams can be affected.26)

In the present work, aluminum foams made by 6063 alloy were fabricated by melt foaming method. Foams with different stirring times of blowing agent and at different cooling conditions were studied. Aluminum foams with homogenous pore structure as well as superior mechanical properties were prepared. The compressive property of the 6063 aluminum foams is better than that of pure Al in other's work.27)

2. Experimental Procedure

The raw materials used were the section bars of commercial 6063 alloy, an Al-Mg-Si alloy (Al, Mg: 0.45–0.9 mass%, Si: 0.2–0.6 mass%) widely applied in construction. Calcium granules (average size 3–4 mm) stored in a vacuum drying oven were used as the thickening agent. Titanium hydride powder (TiH2 < 48 μm, purity 99.4%) was added as the blowing agent. TiH2 was pre-treated at 773 K for 120 minutes in air to delay its decomposition in foaming process.25)

Aluminum foams were prepared by melt foaming method. The alloy was first melted in a crucible heated by a resistance furnace. 2 mass% of calcium was then added to the molten alloy at 953 K and been stirring for 10 minutes. The approximate melting temperature of the 6063 aluminum alloy with 2 mass% calcium is 923 K, according to the differential scanning calorimetry (DSC) spectrum shown in Fig. 1. After thickening, 0.6 mass% of TiH2 was added into the melt at 923 K and then dispersed by mechanical stirring for 3, 4, 5, 6 minutes, respectively. After foaming, aluminum foams were taken out at different cooling conditions shown in Fig. 2: cooling in the air, cooling in the heat preservation materials and cooling in the furnace, respectively. The stirring time and the cooling condition directly influence the uniformity of cell size distribution. For every experimental condition, the foams produced were named as A, B, C, D, E and F, showing in Table 1.

Fig. 1

DSC spectrum of 6063 aluminum alloy with 2 mass% calcium.

Fig. 2

Cooling conditions of (a) cooling in the air, (b) cooling in the heat preservation materials, and (c) cooling in the furnace.

Table 1 Experimental conditions for all the foams produced.
  A B C D E F
TiH2 stirring time, t/min 3 4 5 6 6 6
cooling condition in the air in the air in the air in the air in the heat preservation materials in the furnace

After cooling, the samples were taken out from the foam ingots. After grinding, polishing and eroding, the microstructure of foam samples was investigated on a Zeiss MERLIN-VP-COMPACT scanning electron microscope (SEM) using an accelerating voltage of 15 kV and the elemental analysis was done using an energy dispersive spectrometer (EDS). X-ray diffraction for phase analysis was performed using a D8-ADVANCE X-ray diffractometer. The specimens were scanned using Cu radiation at 40 kV and 100 mA. The scanning speed (2θ) was 1°·min−1.

Cuboid aluminum foam samples cut by an electrical discharge machining (EDM) with the size of 50 mm × 150 mm × 30 mm were prepared for macrostructure analysis. Foam density, porosity, pore diameter, cell size uniformity were calculated. Among these, foam density and foam porosity was acquired by physics method. Pore diameter was calculated on image analysis software Image-Pro Plus 6.0. And cell size uniformity was characterized by cell size distribution and its accumulative fraction. The surfaces of the samples were painted in black to have a better contrast. After binarization, all the pore areas were analysed and then converted to equivalent diameter by regarding the irregular pore shape and the cross-section as a sphere and a circle, respectively. Then the pore diameter can be acquired.

Quasi-static compression tests were carried out at a loading rate of 5 mm·min−1 on a WDW computer-controlled electronic universal testing machine (Model WDW-50). Cylinder specimens with the size of Ø50 mm × 100 mm were cut from the foams using the electrical discharge machining. The upper rigid plate was pressed down on the cylindrical sample while the lower one was stationary. The loading force and the displacement of the upper rigid plate were recorded by a computer. Energy absorption and energy absorption efficiency can be achieved from the stress-strain curve. To precisely quantify the macrostructure character and the compression behavior of the pores, 3 individual foams for every condition were prepared for averaging.

3. Results and Discussions

3.1 Microstructure

SEM microstructure and X-ray diffraction patterns of re-melting 6063 alloy, re-melting 6063 alloy with 2 mass% calcium and cell wall of the foam are shown in Fig. 3. The re-melting 6063 alloy (Fig. 3(a)) consists of Al dendrites and a continuous network of Al-Mg-Si eutectic, which is the strengthen phase. XRD profile in Fig. 3(d) demonstrates the main phases of the re-melting 6063 alloy are Al and Al9Si phase. As illustrated in Fig. 3(b), the re-melting 6063 alloy with 2 mass% calcium consists of an interdendritic network of Al-Si-Ca phase in a matrix of aluminum solid solution. XRD spectrum (Fig. 3(e)) shows that Al4Ca and CaAl2Si2 phases enhance the viscosity remarkably. Al-Mg-Si-Ca phase is present in the cell wall of the foam (Fig. 3(c)). Moreover, large Al-Ti-Mg-Ca phase is randomly distributed within the cell wall. It is presumably a kind of residue from the foaming agent reacted with the melting alloy. XRD analysis in Fig. 3(f) indicates that this reaction product is Al20CaTi2.

Fig. 3

Microstructure and XRD patterns of (a)(d) melting 6063 alloy, (b)(e) melting 6063 alloy with 2 mass% calcium, and (c)(f) cell wall of the foam.

3.2 Macrostructure of pores

Figure 4 shows the cross-sections of foams at different TiH2 stirring times and different cooling conditions. A part of cell walls are apparently squeezed when TiH2 stirring time is short in Fig. 4(a), while this phenomenon is not obvious when TiH2 stirring time is long in Fig. 4(d). With the decreasing of cooling rate, as shown in Fig. 4(d)(e)(f), pores tend to get bigger. Foam density ($\rho^*$) was measured by mass ($M$) and volume ($V$) of the foams, which can be given as:   

\[ \rho^* = M/V \](1)
Foam porosity ($\theta$) was expressed by relative density according to the following equation:   
\[ \begin{split} \theta & = (1 - \rho_r) \times 100\% = (1 - \rho^*/\rho_s) \times 100\% \\ & = [1 - M/(V\rho_s)] \times 100\% \end{split} \](2)
where $\rho_r$ is the relative density, $\rho^*$ is the foam density calculated by eq. (1), and $\rho_s$ is the density of corresponding dense material, which is the density of pure aluminum, 2.7 g·cm−3.
Fig. 4

Foam cross-sections of samples (a) A, (b) B, (c) C, (d) D, (e) E, and (f) F.

The calculation results are represented in Fig. 5(a). The density of the foams increases with the increasing of TiH2 stirring time and decreases with the decreasing of cooling rate. And the porosity of the foams presents an opposite trend. Apparently, sample D has the highest density of 0.6 g·cm−3, and the lowest porosity of 79%.

Fig. 5

Density (a) and porosity (b) of aluminum foam samples A, B, C, D, E, F.

Pore diameter is another parameter to characterize the foam structure. The equivalent coefficient 0.785 between the equivalent diameter and the real pore diameter is derived as eqs. (3)(4).28) The pore is regarded as a sphere called equivalent sphere. According to the Pythagorean Theorem, the diameter $d_1$ of the pore cross-section can be calculated by the expected value:   

\[ d_1 = E (d_1) = \int_{0}^{a} 2\sqrt{a^{2} - z^{2}} dz/a = a\pi /2 \](3)
where $E(d_1)$ is the expectation of the cross-section circle diameter, $a$ is the radius of the equivalent sphere, $z$ is the distance from the great circle of the equivalent sphere to the pore cross-section.

Then the relationship between the pore cross-section diameter $d_1$, i.e. the equivalent diameter, and the equivalent sphere diameter $d$, i.e. the real pore diameter, is   

\[ d_1 = a \pi/2 = d\pi/4 = 0.785d \](4)

Pore diameters of the six samples are varying from 4 mm to 7 mm, shown in Fig. 5(b). The diameter decreases with the increasing of TiH2 stirring time and increases with the decreasing of cooling rate. Among the foams, sample D has the smallest pore size of 4 mm.

A scattered column set illustrates a scattered cell size distribution, which means foam pores are heterogeneously existed. On the contrary, a gathered column set illustrates a gathered cell size distribution, which means foam pores are homogeneously existed. According to Fig. 6, pores in foam D with the narrowest column set are more uniformly distributed than those in other foams. A comparison of six aluminum foams' accumulative fractions are shown in Fig. 7. The curve tendency can reflect the pore diameter distribution greatly. Long platforms ahead and after, and a steep rising region demonstrate a gathered cell size distribution. This also well explains that foam D has the most uniform pores. It is worth mentioned that pores with less than 1.5 mm diameter are not involved in the statistics, because these pores are not caused by the blowing agent.

Fig. 6

Cell size distribution of aluminum foam samples (a) A, (b) B, (c) C, (d) D, (e) E, (f) F.

Fig. 7

Accumulative fractions of 6 aluminum foams.

The blowing agent must be mixed into the melt as soon as possible on account of its rapid decomposition. TiH2 stirring time directly influences the uniform dispersion of TiH2 and the amount of effective TiH2. Long stirring time contributes to homogeneous dispersion of the blowing agent in the melt, thus homogenous pore structure obtained correspondingly. Cross-sections of the foams in Fig. 4 indicate that prolonging TiH2 stirring time promotes the sufficient dispersion of blowing agent before foaming, so that the pores are uniformly distributed after cooling down. Furthermore, extending stirring time makes redundant H2 escape out of the melt, which reduces gas volume left in the melt. After solidification, the foam porosity is reduced and the density is increased. In addition, stirring time represents stirring intensity to some extent. An increase in stirring intensity not only breaks up the bubbles which intend to combine together, but also contributes to the well dispersing of the foam agent. These all, in hence, make the pores smaller and more uniform.

Cooling condition is related to the cooling rate. Cooling in the air has the rapidest cooling rate among these three cooling conditions, while cooling in the furnace has the slowest cooling rate. And the rate of cooling in the heat preservation materials is between the two. Due to the low thermal conductivity coefficient of aluminum foams, the solidification process is of vital importance. A slow cooling rate provides the pores a longer and more adequate time to coalesce into a larger one. As a consequence, the density is smaller with more gas, the porosity is bigger, and the pore diameter is larger. In addition, the temperature difference between outer surface and inner foam is the main factor affecting pore growth. The lower outer temperature makes the outer cell wall melt solidify at first, and then comes the inner cell wall melt. Compared with the outer cell wall melt, the inner cell wall melt has more time to grow and coalesce. This, finally, leads to a non-uniform distribution of the cell sizes.

Based on the reasons mentioned above, the most homogenous pore structure can be prepared by the foaming process of sample D, i.e. stirring the blowing agent for 6 minutes and cooling the foam in the air. In summary, a longer stirring time and larger cooling rate are necessary for preparing aluminum foams with better morphology.

3.3 Compression behavior

During the compression process, aluminum foams squashed layer by layer are shown in Fig. 8. The stress-strain curves and energy absorption efficiency obtained from aluminum foams of different stirring times and cooling conditions are shown in Fig. 9. At first, a very small linear elastic deformation displays and the stress increases rapidly when the pressure head acts on the surface of the aluminum foam where partially reversible cell walls bending occur. Then the plastic deformation of aluminum foam occurs after reaching the yield strength and an extended plateau region appears when the cell walls collapse layer by layer shown in Figs. 8(a) and 9. Energy is absorbed as the foam cell walls yielding. When all the cell walls collapse and become pressed together, shown in Fig. 8(b), densification begins and the stress increases rapidly.

Fig. 8

Compression process of an aluminum foam with (a) plateau period and (b) densification period.

Fig. 9

Stress-strain curves and energy absorption efficiency of aluminum foams with (a)different TiH2 stirring times and (b)different cooling conditions.

The energy absorption property can be represented by the following equation, which is the area under the stress-strain curve up to densification:29)   

\[ W(\varepsilon_d) = \int_{0}^{\varepsilon_d} \sigma (\varepsilon)d \varepsilon \](5)
where $W$ is energy absorption capacity of aluminum foam, $\varepsilon_d$ is the densification strain, $\sigma (\varepsilon)$ is the plateau stress at the strain of $\varepsilon$. Energy absorption efficiency $\eta$ is a parameter to describe the ability of the real porous material to absorb energy, which can be defined as:30)   
\[ \eta (\varepsilon_d) = \left[ \int_{0}^{\varepsilon} \sigma (\varepsilon)d \varepsilon \right] / \left[ \sigma_{max}(\varepsilon)\varepsilon \right] \](6)
where $\sigma_{max}(\varepsilon)$ is the maximum stress with the increasing of the strain. In the experiments of different stirring times, yield strength $\sigma_b$ and energy absorption $W$ increase with the time and have the highest values for foam D (7.6 MPa, 5.8 MJ·m−3). In the experiments of different cooling conditions, these two values decrease with the cooling rate slowing down and have the highest values for foam D. Table 2 summarizes all the foams investigated.
Table 2 Density, porosity, pore diameter, yield strength and energy absorption of six foam samples.
  A B C D E F
Density, $\rho^*$/g·cm−3 0.24 0.29 0.31 0.57 0.43 0.34
Porosity, $\theta$/% 91 89 88 79 84 87
Pore diameter, $d$/mm 7.0 5.6 5.6 4.0 4.5 5.3
Yield strength, $\sigma$/MPa 1.5 2.5 2.7 7.6 4.3 2.8
Energy absorption, $W$/MJ·m−3 1.1 1.8 2.0 5.8 2.7 1.9

The mechanical properties of foams are mainly determined by the density and the yield strength of cell wall material.3133) Aluminum foams with small pores tend to be denser. While a foam with large pore diameter has more breakup called solidification defects on the cell walls, the deformation usually initiates at larger pores.34) Therefore, the smaller pores a foam has, the higher strength it gets. In other words, the denser, the stronger. Besides, defect regions in foams, such as holes, cracks, squashed pores, can also lead to low strength. What is more, heterogeneous cell size distribution results in the unevenly stressing. As a whole, defects and heterogeneity are caused by low stirring intensity and slow cooling rate. Thus, foam D with long stirring time and high cooling rate has high yield strength and high energy absorption.

According to M. F. Ashby and L. J. Gibson, the yield strength of cell wall material has a linear influence on the plateau stress of aluminum foams.35) Foams can be made from many kinds of materials including metals, plastics, and even ceramics. The properties depend on two parameters: the first is the intrinsic properties of the cell wall material, and the second is the special distribution of the cell wall material. If the second kind of parameter is somehow unalterable, the modification and refinement of the cell wall material is a feasible way to improve the foam performance. Microstructure of foam cell walls in Fig. 10 apparently shows that the dendritic crystal is getting bigger with the decrease of the cooling rate. As everyone knows, the small and dispersed precipitation phases help strengthen the intensity of the cell wall. High strength cell walls make up the high strength foam. Thus, the compressive behaviour can be increased. In the experiments of different cooling rates, cooling in the air decreases the time for dendritic structure to grow, which makes the dendrite small and dispersed. This is another reason why foam D with small dendritic structure in its cell wall has high yield strength and high energy absorption.

Fig. 10

Microstructure of aluminum foams in different cooling conditions of samples (a)D, (b)E, (c)F.

4. Conclusion

(1) Aluminum alloys made from 6063 aluminum alloy have been successfully foamed with a uniform cell size distribution as well as good mechanical property.

(2) Appropriately extending TiH2 stirring time contributes to homogeneous dispersion of the blowing agent. A fast cooling rate leaves gas pores no time growing up and coalescing together. These finally help make a uniform cell size distribution of the foam.

(3) For the compressive property, the smaller the porosity of the foam, the greater the density, the greater load the foam can bear, the higher plateau stress for the aluminum foam.

REFERENCES
 
© 2017 The Japan Institute of Metals and Materials
feedback
Top