2018 Volume 59 Issue 6 Pages 999-1004
An aluminum foam sandwich (AFS), which consists of an aluminum (Al) foam core and two dense metallic face sheets, is a lightweight component material with good energy and sound insulation properties. A fabrication method for AFSs that can simultaneously fabricate a foamable precursor and realize metallurgical bonding between the precursor and a dense metallic sheet was proposed using the friction stir welding (FSW) route. In this study, we produced a sandwich-type foamable precursor with two Al face sheets using the FSW route and, by foaming this precursor, an AFS was fabricated. Through X-ray computed tomography inspection of the AFS and observation by an electron probe microanalyzer, it was confirmed that no cracklike pores existed in the Al foam core and that only few pores generated by the foaming entered the Al face sheet. Moreover, static tensile tests of the AFS were carried out, and it was shown that the tensile strength of the Al foam core was not decreased by the existence of an oxide film and that the bonding strength of the interface between the Al foam core and the Al face sheet was higher than the tensile strength of the Al foam core.
Fig. 1 Schematic illustration of process for fabricating AFS by FSW.
An aluminum foam sandwich (AFS) consists of an aluminum (Al) foam core and two dense metallic face sheets. This AFS is a multifunctional component that can improve tensile and bending strength owing to the dense surface plates in addition to having the excellent lightweight, energy absorption and sound insulation properties of Al foam. Because of these superior properties, it is expected that AFSs can be applied as multifunctional components in various industrial fields such as automobiles, railway vehicles and machine tools.1,2) As fabrication methods of AFSs, the bonding of an Al foam core to dense metallic face sheets using an adhesive3–5) and the heating of a foamable precursor clad-bonded with two face sheets by extrusion or rolling6,7) have been proposed. However, the former makes the recycling of AFSs difficult and generates environmental problems because of the use of adhesive8) and the latter has the problem of low productivity because the production of a foamable precursor and the extrusion (or rolling) are time-consuming processes.
The authors previously proposed a fabrication method for an AFS that uses friction stir welding (FSW) to produce a foamable precursor.9–12) In this method, both the production of a foamable precursor by mixing blowing agent powder into Al sheets and metallurgical bonding between the foamable precursor and dense metallic face sheets can be conducted simultaneously; thus, high productivity of the foamable precursor can be realized. Moreover, using this fabrication method, we succeeded in fabricating an AFS comprising an A1050 Al foam core and SPCC low-carbon steel face sheets9,10) and an AFS comprising an ADC12 Al alloy foam core and dense Al face sheets.11,12) In the fabrication of the AFS, a foaming process, in which two foamable precursors with one face sheet placed face to face in a die are heated and the Al foam core is formed by bonding two foaming precursors, has been used. However, in the Al foam core fabricated using this foaming process, insufficiently bonded regions such as cracklike pores sometimes appeared owing to the existence of an oxide film.12) These insufficiently bonded regions in the Al foam core may lead to early failure and decrease the tensile strength of the AFS. Therefore, it is necessary to obtain an effective foamable precursor by a new FSW procedure to fabricate an AFS without insufficiently bonded regions in the Al foam core.
In this study, we produced a sandwich-type foamable precursor of ADC12 Al alloy with two A1050 Al face sheets using a new pass route of FSW and fabricated an AFS by foaming this foamable precursor. This AFS does not use materials other than aluminum and aluminum alloy. It can be expected that the AFS makes the recycling comparatively easy and reduce influences on environment. X-ray computed tomography inspection of the obtained AFS and microstructure observation by an electron probe microanalyzer were conducted to examine the pore structure and the bonding state of the interface between the Al foam core and the Al face sheet, i.e., whether no cracklike pores existed in the Al foam core and pores generated by foaming did not enter the Al face sheet. In addition, static tensile tests of the AFS were carried out to examine whether insufficiently bonded regions such as an oxide film in the Al foam core did not exist. From these results, it was shown that the AFS had an Al foam core with good pore structures, that the tensile strength of the Al foam core was not decreased by the existence of an oxide film and that the bonding strength of the interface between the Al foam core and the Al face sheet was higher than the tensile strength of the Al foam core.
Figure 1 shows a schematic illustration of the process for fabricating an AFS using FSW. Two as-cast ADC12 Al–Si–Cu Al alloy die casting sheets of 3 mm thickness were laminated with blowing agent powder (TiH2, <45 µm) and stabilization agent powder (α-Al2O3, ∼1 µm) distributed between them, and the laminated sheets were stacked on a commercial-purity A1050-H18 Al sheet of 3 mm thickness (cf. Fig. 1(a)). Arithmetic average roughness Ra for the A1050 and ADC12 sheets are 0.5 µm and 2.5 µm respectively. As shown in Fig. 1(b), the probe of a rotating tool was inserted into the laminated sheets, and multipass FSW13,14) of 4 lines × 4 passes was conducted to mix the powders into the laminated sheets and to join the laminated sheets.15,16) The tool used for FSW had a columnar shape with a screw probe. The diameter of the tool shoulder was 17 mm and the diameter and length of the screw probe was 6 mm and 4.8 mm, respectively. The tool material was SKH51 high-speed tool steel. The tool rotation speed and welding speed were 1000 rpm and 100 mm/min, respectively.9–12) A tilt angle of 3° was used with respect to the vertical axis of the sheet surface. The probe of the rotating tool was inserted into the surface of the A1050 Al sheet to a depth of 0.2 mm9–12) during 4th pass. This inserted depth of probe (0.2 mm) is much larger than Ra. It is expected that, using intense fluidity by FSW, the metallurgical bonding between the A1050 Al sheet and the laminated sheets can be conducted regardless the surface roughness and surface state. The amounts of TiH2 and Al2O3 relative to the mass of the ADC12 Al alloy die casting sheet with the dimensions of the area over which TiH2 and Al2O3 were distributed and the length of the screw probe were 1 mass% and 5 mass%, respectively. FSW was carried out using an FSW machine (SHH204-720, Hitachi Setsubi Engineering Co., Ltd.). Moreover, after the surface of the laminated sheets stirred by FSW was smoothed by milling, an A1050 Al sheet was stacked on the smooth surface. Then, as shown in Fig. 1(c), FSW of 4 lines × 1 pass was conducted to join the A1050 Al sheet to the laminated sheets. In this FSW, the same type of tool as that used for the laminated sheets but with a different length of the screw probe of 3 mm was used and the probe was inserted into the surface of the laminated sheets to a depth of 0.2 mm. The tool rotation speed and welding speed were 2200 rpm and 100 mm/min, respectively.9–12) After the surface of the A1050 Al sheet stirred by FSW was smoothed by milling, a sandwich-type foamable precursor consisting of the stirred ADC12 Al alloy core bonded with A1050 Al top and bottom sheets, as shown in Fig. 1(c), was cut from the region subjected to FSW by machining.
Schematic illustration of process for fabricating AFS by FSW.
The sandwich-type foamable precursor was turned over so that it faced downward and placed in a die made of SPCC steel as shown in Fig. 1(d). The distance between the A1050 Al top face sheet of the foamable precursor and the bottom plate of the die h was 8 mm (the distance between the bottom plate and top plate of the die was 18 mm) and the distance between the left/right side of the foamable precursor and the left/right block of the die b was 7.5 mm. Then, the sandwich-type foamable precursor set in the die was heated in a preheated electric furnace. The sandwich-type foamable precursor was foamed downward to bring the A1050 Al top face sheet facing downward into contact with the bottom plate of the die. The holding temperature (equal to the preheated temperature) was 948 K (675°C) and the holding time tH during the heating process was varied from 9 min to 14 min in steps of 1 min. After heating, the foamed precursor was taken out from the electric furnace and cooled to room temperature under ambient conditions. AFS specimens with an horizontal area of 20 mm × 20 mm were cut by electro-discharge machining.
2.2 Evaluation of pore structuresThe porosity p (%) of each AFS specimen was evaluated by the following equation:
| \begin{equation} p = \frac{\rho_{i} - \rho_{f}}{\rho_{i}}\times 100, \end{equation} | (1) |
To evaluate the pore structure of the Al foam core, the AFS specimens were scanned by a microfocus X-ray CT system (SMX-225CT, Shimadzu Corporation). The X-ray source was tungsten. A cone-type CT system, which can obtain a set of two-dimensional cross-sectional X-ray CT images for an entire specimen with a slice pitch equal to the length of one pixel in the X-ray CT image by a single rotation of the specimen, was employed. The X-ray tube voltage and current used in the inspection were 80 kV and 30 µA, respectively. The resolution of the X-ray CT image was 512 × 512 pixels and the length of one pixel was approximately 60 µm. The X-ray CT images displayed 16-bit gray-scale data. An appropriate threshold was set for all the two-dimensional cross-sectional X-ray CT images to distinguish the cell walls and the pores to construct binarized X-ray CT images. From the binarized X-ray CT images, the equivalent diameter d and circularity e19) of each pore were evaluated by the following equations:
| \begin{equation} d = 2\left(\frac{A}{\pi}\right)^{\frac{1}{2}}, \end{equation} | (2) |
| \begin{equation} e = \frac{4\pi A}{l^{2}}, \end{equation} | (3) |
The distribution of elemental Si, which exists in larger amounts in the ADC12 Al alloy than in the A1050 Al, on the surface of each AFS specimen was observed using an electron probe microanalyzer (EPMA-1620, Shimadzu Corporation). To avoid damaging the specimen, a cut surface obtained by electrodischarge machining of a separated part was used instead of the cut surface of the specimen. It was assumed that the elemental Si distribution on the surface of the separated part was almost the same as that of the specimen surface. The observed surface was ground using SiC paper (up to #2400) and then polished with alumina (1 µm) before the EPMA observation.
2.4 Tensile tests on AFS specimenA part for gripping during tensile tests was realized by bonding a jig to the A1050 Al face sheets of the AFS specimens using an adhesive. The tensile tests were carried out at room temperature using a universal testing machine with a load capacity of 98 kN (Autograph AG-100kNG, Shimadzu Corporation). The relative velocity between the cross head and the screw rod was set at 1.8 mm/min. During the tests, sequential deformation images of the specimens were visually observed by a video camera.
Figures 2(a), 2(b) and 2(c) show examples of AFS specimens with their values of porosity p fabricated with the holding times of 9 min, 11 min and 14 min, respectively. Figures 3(a), 3(b) and 3(c) show X-ray CT images at the center surfaces in the depth direction of the specimens in Figs. 2(a), 2(b) and 2(c), respectively. In Fig. 3, the upper and lower gray parts are the A1050 Al top and bottom face sheets, and the other gray and black parts between the upper and lower gray parts show the cell walls and the pores of the Al foam core, respectively. For the entire holding time, the A1050 Al top face sheets of the foamed precursors were in contact with the bottom plate of the die and, as shown in these figures, although the surface traversed by FSW can be observed on the A1050 Al top face sheets, the A1050 Al top face sheets of the AFS specimens were maintained parallel to the A1050 Al bottom face sheets. The heights of all the AFS specimens were 18 mm. By visual observation of the surfaces (cf. Fig. 2) and X-ray CT images (cf. Fig. 3) of the AFS specimens, we can judge that no cracklike pores existed in the Al foam core or at the interface between the Al foam core and the A1050 Al face sheets. As an example of the result of EPMA mapping analysis for elemental Si, Fig. 4 shows the result for the AFS specimen with the holding time of 11 min. The chain lines in this figure show the boundary between the Al foam core and the A1050 Al face sheets that the probe tip was traversed through. The gray regions indicate where elemental Si was detected. In this figure, more Si was detected in the Al foam core, whereas little Si was detected in the A1050 Al face sheet. Therefore, it can be considered that only few pores in the Al foam core generated by foaming entered the A1050 Al face sheet.
Fabricated AFS specimens with holding times of (a) 9 min, (b) 11 min and (c) 14 min.
X-ray CT images of fabricated AFS specimens with holding times of (a) 9 min, (b) 11 min and (c) 14 min.
Result of EPMA mapping analysis for elemental Si in AFS (holding time 11 min). (a) Overall side view and (b) detail in region A.
Figures 5, 6 and 7 show the changes in the porosity p, the average equivalent diameter dm and the average circularity em with the holding time tH, respectively. From these figures, it was found that, when tH was 9 min, owing to insufficient foaming, p was low (lower than 70%), dm was small (lower than 1 mm) and em was small (lower than 0.7). When tH was 10 min, the values of p, dm and em increased markedly to approximately 75% for p, 1.1 mm for dm and 0.71 for em owing to the intense decomposition and foaming of TiH2 caused by the temperature rise of the precursors. Then, they increased slightly with further increasing tH up to 12 min because the generation of gas by the decomposition of TiH2 was reduced. When tH was 13 min or larger, p and em decreased slightly, whereas dm rapidly increased to approximately 1.6 mm. The slight decrease in p was because, although the decomposition of TiH2 had almost finished, the gas in the pores generated by the decomposition of TiH2 was released from the softened precursor. The rapid increase in dm was because large pores were formed by the coalescence of pores. The slight decrease in em is mainly because distorted pores were formed by the coalescence of pores, and the pores were deflated and deformed by the release of the gas from the precursor.
Change in porosity p with holding time tH.
Change in average equivalent diameter dm with holding time tH.
Change in average circularity em with holding time tH.
From these results, it is considered that AFS specimens with good bonding interfaces and good pore structures of dm ≒ 1.2 mm and em ≒ 0.71 were fabricated when the holding time was 11–12 min for the distance between the foamable precursor and die used in this study (h = 8 mm and b = 7.5 mm, cf. Fig. 1(d)). That is, it was found that an AFS specimen with high porosity and good pore structures can be fabricated by optimizing the holding time for a given distance between the foamable precursor and the die.
3.3 Tensile test resultsFigure 8(a) shows a typical tensile stress-strain curve of an AFS specimen which was obtained from the specimen with a porosity of 75.6% and a holding time of 11 min. Figures 8(b)–8(e) show photographs of the specimens corresponding to positions (b)–(e) in Fig. 8(a). In this specimen, little local fracture occurred below the maximum tensile stress, and local fracture of the cell wall began to occur at the position in the Al foam core shown by the black arrow in Fig. 8(c) immediately after the tensile stress reached a maximum. After that, the tensile stress gradually decreased with increasing tensile strain owing to crack propagation due to the continuous plastic deformation and fracture of the cell wall (cf. Fig. 8(d)), then the final fracture occurred (cf. Fig. 8(e)). Figure 8(f) shows an example of the fracture surface, on which we observed no oxide film. From these results, it is concluded that there was no oxide film in the Al foam to decrease the tensile strength of the AFS and that the bonding strength of the interface between the Al foam core and the A1050 Al face sheet was higher than the tensile strength of the Al foam core of the AFS specimen.
Tensile test results for AFS specimen (holding time of 11 min). (a) Stress-strain curves, (b) initial state, (c) onset of crack growth, (d) final fracture, (e) after final fracture and (f) fracture surface.
Here, plastic deformation can be easily applied to the sandwich-type foamable precursor with two A1050 Al face sheets developed in this study. Although further studies are necessary, by foaming a plastically deformed sandwich-type foamable precursor with two A1050 Al face sheets, it is expected that AFSs having various shapes can be fabricated.
In this study, to fabricate an AFS without insufficiently bonded regions in an Al foam core, an AFS was fabricated using a new pass route of FSW. Moreover, static tensile tests were carried out on the fabricated AFS specimen to examine its tensile strength. The experimental results led to the following conclusions.
This work was partly financially supported by Grant-in-Aid for Scientific Research (C) (17K06842) and the Amada Foundation.
