MATERIALS TRANSACTIONS
Online ISSN : 1347-5320
Print ISSN : 1345-9678
ISSN-L : 1345-9678
Engineering Materials and Their Applications
Effect of Si Concentration of a Brazing Precursor on the Bonding Strength of Aluminum Foam Bonded via Foaming Bonding
Ryosuke SuzukiYoshihiko HangaiYusuke AsakawaIkuo ShohjiHidetoshi FujiiMasaaki Matsubara
著者情報
ジャーナル フリー HTML

2021 年 62 巻 8 号 p. 1210-1215

詳細
Abstract

Pure Al foams were bonded via foaming bonding using Al–10.5Si brazing alloy. However, the pores around the bonded area were coarsened because Si diffused into the Al foam from the brazing foam, where it reduced the melting temperature, leading to melting of the cell walls of the Al foam. The bonding strength of the bonded foam varied greatly. An Al–Si brazing precursor with a low Si concentration was prepared and used for foaming bonding of two Al foam specimens. Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy were used to investigate the bonding state. Four-point bending tests were conducted to evaluate the bonding strength. The brazing foam was metallurgically bonded with the Al foam when brazing precursors with 9.4–10.5 mass% Si were used. The bonding interfaces were partially observed by SEM. The bonding strength of the Al foam bonded using 10.5 mass% Si brazing foam was lower than the minimum strength of the Al foam. The bonding strength of the specimen bonded using the 9.4 mass% Si brazing foam was lower than the minimum strength of pure Al foam and the bonding strength of the other bonded specimens. An Al–Si brazing precursor with an appropriate Si concentration prevents pore coarsening around the bonding interface. The bonding strength became low when the Si concentration in the Al–Si brazing precursor was excessively low.

1. Introduction

Al foams are composite metal materials that are lightweight because of their numerous pores.1,2) Al foams are expected to not only be used as lightweight structural materials but also as functional materials because of their various desirable characteristics, including thermal insulation,3) sound absorption,4) vibration absorption,5) and impact absorption.6) Large-sized Al foams are required for industrial applications because they especially expected to be used as lightweight functional structural materials in the transportation industry,7,8) and bonding is an effective method to manufacture large-sized components.9) However, joining Al foams using the bonding methods for other metals is difficult because Al foams have a cell structure. For example, screw threads cannot be cut into Al foam, and the cell walls of Al foam collapse by local compressive stress induced by bolt–nut fastening. In addition, adhesion is not an appropriate bonding method for Al foams because the effective bonding area is narrow. Welding is also difficult because the cell walls collapse by melting of the matrix. Friction stir welding10) likewise destroys the cell walls. However, a few studies have reported bonding processes for Al foams. Specifically, diffusion bonding11) using a superplastic Al brazing sheet and welding12) using a low-melting-point Al precursor as a filler material have been proposed.

In the diffusion bonding process, a superplastic Al sheet is placed between Al foam components, and they are pressed at an elevated temperature. The strength of the bonds created by diffusion bonding is greater than that of bonds created by adhesion at elevated temperatures. However, preventing the collapse of the cell walls during diffusion bonding is difficult because the process involves applying compressive stress. The bonding strength is only about 50% of the strength of bulk pure aluminum foam at room temperature. In addition, the aforementioned beneficial properties of Al foam are lost at the bonding area because the brazing material is dense.

In the welding process, a low-melting-point Al precursor is placed between Al foam components and the bonding area is heated using a welding torch. The strength of the bond between Al components joined by the welding method was found to be as high as that of pure bulk Al foam because the joined components broke before joint separation during mechanical testing.12) In addition, the beneficial properties of Al foams are preserved at the bonding area because the blazing material is also a foam. However, the cell walls collapse by melting as a consequence of the high heat input in the welding process.

Recently, we proposed a foaming bonding method as a brazing process for Al foam.13) This method can bond Al foam components using Al–10.5Si as an Al alloy brazing precursor with a melting point lower than the melting point of Al foam. The mean bonding strength resulting from foaming bonding is high—close to that of pure Al foam—and the industrially useful properties of Al foams are also preserved at the bonding area. However, the pores at the bonding area are coarsened by cell wall melting because the melting point of the pure Al used decreases as a result of the diffusion of Si from the Al–Si brazing foam into the Al matrix. This causes a large decrease in the bonding strength, leading to low reliability.

The pore coarsening effect can be prevented by decreasing the Si concentration in the brazing precursor. However, a low Si concentration leads to poor bonding. In the present study, to improve the reliability of the foaming bonding, we carried out foaming bonding using a brazing precursor with 9.4–10.5 mass% Si. To reduce the Si concentration in the brazing precursor, pure Al powder was added to a mixture of ADC12 (Al–11Si) powder, TiH2 powder, and Al2O3 powder to prepare a brazing precursor via powder metallurgy. The bonding strength was then evaluated using four-point bending tests.

2. Experimental Procedure

2.1 Materials

Pure Al powder with a particle diameter less than 180 µm (Koujundo Chemical Laboratory) was prepared as a matrix material for pure Al foam. The melting point of Al is 660°C. An Al–11Si alloy powder with a particle diameter smaller than 300 µm (Toyo Aluminium K.K.) was prepared as the matrix material for the brazing precursor. The Al–11Si alloy has a near-eutectic composition, and its solidus and liquidus temperatures were approximately 520 and 572°C, respectively.14) The literature includes numerous reports on Al–11Si alloy foams.15,16) Al–11Si alloy is easy to handle as a brazing precursor material. Titanium hydride (TiH2) powder with a particle diameter smaller than 45 µm and alumina (Al2O3) powder with a particle diameter of ∼1 µm (Koujundo Chemical Laboratory) were prepared as foaming and thickening agents, respectively. TiH2 and Al2O3 are widely used as foaming and thickening agents, respectively, for Al precursors.17,18)

2.2 Manufacture of pure Al foam

The process used to prepare the pure Al precursor is shown in Fig. 1. The pure Al powder was mixed with 1 mass% TiH2 powder and 3 mass% Al2O3 powder, and the resultant mixture was placed into a mold for powder metallurgy. The mold was made from SUS310 stainless steel and had a hole with a 25 mm diameter. The mold was placed on a uniaxial hot-press machine and heated to 500°C under 150 MPa compressive pressure. The uniaxial hot pressing was carried out for 180 min. A columnar precursor with a 25 mm diameter and ∼50 mm height was obtained. This was cut into cubes measuring 15 mm on each side. The density of the precursor, ρp, was measured using an electric hydrometer (EW-300SG, ALFAMiRAGE). The filling rate of the precursor, fo, is expressed as   

\begin{equation} f_{o} = \frac{\rho_{p}}{\rho_{o}} \end{equation} (1)
where ρo is the density of the pure Al.

Fig. 1

Manufacturing process for the bonding specimens.

A schematic of the foaming process is shown in Fig. 1. The cubic precursor was placed into the foaming mold made from a stainless steel rectangular tube. Both ends of the mold were plugged with stainless steel rectangular blocks. The internal dimensions of the foaming mold were 16 × 16 × 50 mm3. The mold was placed in an electric furnace (DSTR-11K, ISUZUSEISAKUSHO) heated at 710°C and maintained at this temperature for 7 min. The rectangular Al foam was approximately 16 × 16 × 50 mm3, and the ideal porosity of the foam was ∼74%. The density of the Al foam, ρf, was measured, and the actual porosity, po, was calculated using the equation   

\begin{equation} p_{o} = 1 - \frac{\rho_{f}}{\rho_{o}} \end{equation} (2)
The rectangular Al foam block was then cut into halves perpendicular to the longitudinal direction.

The second rectangular Al foam block cut into halves parallel to the longitudinal direction was also prepared for the investigation of cell morphology. A photograph of the cross-section was taken with a digital camera. The photograph was analyzed using the Image J (Ver. 1.45l) software. The mean pore diameter, dm, was calculated using the equivalent circle diameter:   

\begin{equation} d_{m} = \sqrt{\frac{4}{n\pi}\sum_{0}^{n} A_{i}}, \end{equation} (3)
where n is the number of the pores and Ai is the area of each pore on the cross-section. The average, maximum, and minimum pore diameters were obtained.

2.3 Manufacture of brazing precursor

The Al–11Si alloy powder was mixed with 1 mass% TiH2 powder, 3 mass% Al2O3 powder, and 0, 1, 2, 3, 4, 5, or 10 mass% pure Al powder. The pure Al powder was added to reduce the Si concentration in the brazing precursor, thereby preventing coarsening of the pores around the bonding area. The precursor mixtures with 0, 1, 2, 3, 4, 5, and 10 mass% Al added had corresponding Si concentrations of 10.5, 10.4, 10.3, 10.2, 10.1, 10.0, and 9.4 mass%, respectively. The mixtures were sintered using the previously described uniaxial hot-pressing method. The hot-pressing temperature, pressure, and time were 500°C, 150 MPa, and 180 min, respectively. The columnar brazing precursor was cut into sheets with dimensions of 16 × 16 × 2 mm3 for use in the foaming bonding experiments. The cross-sectional area of the brazing precursor was observed with an optical microscope, and the density of the precursor was measured using an electric hydrometer.

2.4 Bonding Al foams with a brazing precursor

The cross-sections of the bonding surfaces for pure Al foam were degreased with acetone. A flux (RZ-201, Shinfuji Burner) was applied to the cross-sections, and the samples were dried. The brazing precursor sheet was placed between the two pure Al foam blocks, and the specimens were then placed in the foaming mold (Fig. 1). Both ends of the foaming mold were plugged with stainless steel blocks. The foaming mold was placed in an electric furnace heated at 600°C and maintained at this temperature for 20 min, resulting in joined Al foam blocks. The joined Al foam was cleaned ultrasonically at 30°C for 5 min to wash away the flux.

The ideal porosity of the brazing foam, pb, is roughly estimated. The porosity is the volume fraction of pores in the brazing foam. The porosity of the brazing foam is expressed as next equation.   

\begin{equation} p_{b} = \frac{V_{\textit{bp}}}{V_{\textit{bf}}} = \frac{V_{\textit{bp}}}{V_{\textit{bm}} + V_{\textit{bp}}}. \end{equation} (4)
Here, Vbf, Vbp and Vbm are the volumes of the brazing foam, its pores and its cell wall material, respectively. Vbm is almost equal to the volume of the brazing precursor.   
\begin{equation} V_{\textit{bm}} = W^{2}t, \end{equation} (5)
where t is the thickness of the brazing precursor and W is the length of the one side of the square contact surface between the brazing precursor and the Al foam (Fig. 2). The brazing foam flows into the open pores of Al foam on the square contact surface to the maximum filling depth d/2. d is the depend on the cell morphology of the Al foam and is considered the mean pore diameter dm of the Al foam. When the all pores of the Al foam existing from the square contact surface between the brazing precursor and Al foam to the maximum filling depth are open and these open pores are completely filled with the brazing foam, Vbp is equal to the volume of these open pores of Al foam and expressed as next equation.   
\begin{equation} V_{\textit{bp}} = 2W^{2}\frac{d_{m}}{2}p_{o} = W^{2}d_{m}p_{o}. \end{equation} (6)
Here, po is the porosity of the Al foam. The ideal porosity, pb, of the brazing foam is obtained by substituting eq. (5) and eq. (6) to eq. (4).   
\begin{equation} p_{b} = \frac{p_{o}d_{m}}{t + p_{o}d_{m}}. \end{equation} (7)
Although this equation (eq. (7)) is derived with that the shape of contact surface is square, this equation is also satisfied for the other shape contact surface.

Fig. 2

Schematic showing the parameters used in the calculation of the porosity of the brazing foam.

2.5 Evaluation of bonding strength

The joined Al foam blocks were cut into halves parallel to the longitudinal direction, and the bonding interface was observed using scanning electron microscopy (JSM-5510, JEOL). The acceleration voltage was 15 kV, and the atmosphere was a high vacuum. Energy-dispersive X-ray spectroscopy (EDS) analysis was also carried out to investigate the Si concentration around the bonding interface.

The rectangular stainless steel tubes were adhered to both ends of the joined Al foam blocks with skin layer to prevent local collapse at the supports used for the four-point bending tests, which were carried out using a universal testing machine (UH-100KNA, Shimadzu). The cross-head speed was 0.5 mm/min. The outer and inner spans of the supports were 110 and 50 mm, respectively. The bending tests were carried out three times for each specimen. The bending strength, σb, was quantitatively evaluated on the basis of the maximum flexure stress:   

\begin{equation} \sigma_{b} = \frac{3F(L_{1} - L_{2})}{2wh^{2}} \end{equation} (8)
where F is the maximum applied load, L1 is the outer span of the supports, L2 is the inner span of the supports, w is the width of the specimen, and h is the height of the specimen.

3. Results and Discussions

3.1 Pure Al precursor and brazing precursor

The filling rates of the pure Al precursor and the brazing precursor were both greater than 99%. Both precursors would be foamable because their filling rates are greater than 94%.19)

3.2 Porosity and cell morphology of pure Al foam

The experimental porosity of the pure Al foam was ∼74%, which is equal to the ideal porosity. Cross-sections perpendicular and parallel to the longitudinal direction are shown in Fig. 3. The results of a cell morphology analysis of the pure Al foam are listed in Table 1. The mean pore diameter was ∼2 mm for both cross-sections. The ideal porosity of the brazing foam calculated from eq. (7) was 43%. Here, po was 74%, dm was 2 mm, and t was 2 mm. The minimum pore diameter was ∼0.8 mm, and the maximum pore diameter was 7 mm. This large variation of the pore diameter increases the porosity of the brazing foam and the bonding strength.

Fig. 3

Photographs of cross-sections of specimens cut (a) perpendicular and (b) parallel to the longitudinal direction of the Al foam.

Table 1 Results of cell morphology analyses.

3.3 Joined aluminum foam

Photographs of the cross-sections of Al foam blocks joined using 9.4 and 10.5 mass% Si brazing precursors are shown in Fig. 4(a) and Fig. 5(a), respectively. The brazing precursors formed a foam between the pure Al foam blocks. Pore coarsening due to a decrease in the Al melting point was observed in the bonded region for the 10.5 mass% Si brazing precursor but not for the 9.4 mass% Si brazing precursor.

Fig. 4

Results of cross-sectional observations of the specimen bonded with the brazing precursor with 9.4 mass% Si: (a) photograph of a cross section of the bonded specimen; (b) SEM micrograph of the position labeled A near the bonding interface; and (c) the Si concentration at the positions of the open circles around the bonding interface in (b).

Fig. 5

Results of cross-sectional observations of the specimen bonded using the brazing precursor with 10.5 mass% Si: (a) photograph of the cross section of the bonding specimen; (b) SEM micrograph of position B near the bonding interface in figure (a); (c) Si concentration at the open circle positions near the bonding interface shown in figure (b).

SEM micrographs at the marked positions (Fig. 4(a) and Fig. 5(a)) around the bonding interface for 9.4 and 10.5 mass% Si brazing precursors are shown in Fig. 4(b) and Fig. 5(b), respectively. It was clear that at least partial bonding had occurred.

The Si content around the bonding interface for 9.4 and 10.5 mass% Si brazing precursors is plotted as functions of position in Figs. 4(c) and 5(c), respectively. The Si content decreased with increasing distance from the brazing foam but was not zero in the Al foam near the bonding interface. These results show that metallurgical bonding between the brazing foam and the Al foam was achieved using both the 9.4 and the 10.5 mass% Si brazing precursors, although the bonding interfaces were only partially observed.

Photographs of the breakdown positions in the specimens after the four-point bending tests are shown in Fig. 6. All of the bonded specimens started to break from the tensile side because the tensile strength of the Al foam is lower than its compressive strength.20) These photographs show three typical fracture morphologies. The specimen is separated at the bonding interface in Fig. 6(a), broken in the Al foam in Fig. 6(b), and starts to break at the bonding interface followed by breakdown in the Al foam in Fig. 6(c).

Fig. 6

Photographs of the bending-test specimens. The specimens (a) were separated at the bonding interface, (b) were broken in the Al foam, and (c) started to break at the bonding interface, followed by breaking in the Al foam.

The bending load–displacement curves for the Al foam and three typical breakdown scenarios for the bonded specimens are shown in Fig. 7. The bonding strength was evaluated on the basis of the bending strength, which was calculated by substituting the maximum applied load into eq. (4). The bending strength is plotted against the Si content in the brazing precursor in Fig. 8. The open circles, closed circles, and gray triangles show interface separation, breakdown in the Al foam, and complex breakdown, respectively. The broken line shows the maximum and minimum bending strengths for the Al foam. The bonding strengths for the specimens with 10.0 to 10.5 mass% Si were higher than the minimum strength for the Al foam and were almost the same as each other. The bonding strength for the specimen with 9.4 mass% Si brazing foam was lower than that for the other specimens. One of the specimens with 10.5 mass% Si brazing foam was broken at a bending strength lower than the minimum strength of the Al foam. This result is attributed to pore coarsening. However, the bending strengths for all of the specimens with 10.0–10.4 mass% Si brazing foam were higher than the minimum strength of the Al foam. Interface separation occurred not only in the specimen bonded with 10.5 mass% Si brazing foam but also those bonded with 9.4–10.2 mass% Si brazing foam. A brazing precursor with an appropriate Si content would prevent pore coarsening around the bonding interface. However, the bonding state does not greatly improve when brazing foam with a low Si concentration is used because Si diffusion from the brazing foam into the Al foam becomes difficult.

Fig. 7

Load–stroke curves corresponding to the four-point bending tests of the specimens shown in Fig. 6.

Fig. 8

Maximum bending stress plotted against the Si content. The closed circles, open circles, and gray triangles correspond to Al foam destruction, interface separation, and interface/Al foam destruction. (a), (b), and (c) in Fig. 8 correspond the specimens shown in Fig. 6(a), (b), and (c), respectively.

4. Conclusion

To improve the bonding quality during foaming bonding, two pure Al foam blocks were bonded via using an Al–Si brazing precursor with different Si contents. Cross-sectional observations, Si concentration distribution analyses of the regions near the bonding interface, and four-point bending tests were carried out. The results led to the following conclusions:

  1. (1)    The Al foam blocks could be successfully bonded using the brazing precursors. Diffusion of Si from the brazing foam into the Al foam was confirmed by EDS analysis of the region around the interface between the brazing foam and the Al foam for all Si contents in the brazing precursors. The brazing foam and the Al foam were metallurgically bonded for a Si content ranging from 9.4 to 10.5 mass%, although the bonding interfaces were only partially observed.
  2. (2)    Pore coarsening was observed around the bonded region for 10.5 mass% Si, but not for 9.4 mass% Si. One of the specimens bonded using 10.5 mass% Si was broken at a bending strength lower than the minimum strength of the Al foam. This result was attributed to pore coarsening. The bending strengths for the specimens bonded using 10.0–10.4 mass% Si brazing foams were greater than the minimum bending strength of the Al foam. Thus, the use of an Al–Si brazing precursor with an appropriate Si content can prevent pore coarsening around the bonding interface.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Number JP20K05163. This work was supported by JST Mirai Program Grant Number JPMJMI19E5, Japan. We are grateful to Toyo Alulminium K.K. for providing ADC12 aluminum alloy powder.

REFERENCES
 
© 2021 The Japan Institute of Metals and Materials
feedback
Top