Materials Processing
Fabrication of Open-Channel Aluminum by Template-Extraction Technique
2023 Volume 64 Issue 11 Pages 2643-2647
Details
2023 Volume 64 Issue 11 Pages 2643-2647
Open-channel aluminum with directional holes is fabricated by extraction of stainless steel wires coated with release agent, boron nitride, which are embedded in solidified aluminum. In order to extract the wires from aluminum, tensile force is necessary. If a number of wires can be extracted simultaneously, the production of open-channel aluminum with many holes becomes feasible. Smaller extraction force is desirable to fabricate open-channel aluminum with holes as many as possible. It was found that the extraction force decreases remarkably by adding magnesium hydride to boron nitride because of forming porous layers between the wires and aluminum during melting of aluminum. Besides, the extraction force is also investigated as functions of diameter and length of stainless steel wires, which decreases with decreasing diameter and length of the template wires. The results show that small contact area of the wire to aluminum results in smaller extraction force, which is attributed to decrease in friction force of the wires to aluminum.
Porous metals with directional long pores1–3) have attracted extensive attention on applications to heat sinks,4) heat exchangers5) etc. For high performance of such devices, channel holes should possess small diameter and long size to obtain large contact area to the coolant. There are several perforation techniques for metals and alloys such as mechanical drilling, laser ablation and solidification techniques.
Channel holes can be mechanically formed by drilling and removing material. In the case of drilling to fabricate small holes, drill bits with thickness less than 1.0 mm in diameter are worn down through repeated mechanical drilling and are broken easily.6) Thus, creating perforation smaller than 1.0 mm in diameter is laborious, time-consuming and costly. Alternatively, Gillen and Moore adopted electron beam processing to make small holes. The melt zone irradiated by electron beam is evaporated so that the holes are formed.7) The operation should be done in a high vacuum atmosphere and so the technique is inefficient for practical application. A focused laser beam in air is available to melt small points.8) However, in particular, for low melting metals such as aluminum the holes were tapered; the size of holes spreads near surface. This technique is laborious and expensive. Haga and Fuse investigated the fabrication of open-channel aluminum and its alloys by pulling core-bar from a semisolid ingot.9) Metallic core-bars were immersed in aluminum melt, which were pulled out from the semisolid condition during solidification process. The semisolid condition is indispensable and the holes were easily deformed because of viscous semisolid. Thus, a process that does not require semisolid is desirable. On the other hand, Muto et al. developed rod-dipping process to fabricate open-channel aluminum alloy.10) Carbon rods were dipped into a molten aluminum alloy. After solidification carbon rods were removed by plastic deformation or evaporation. A method that does not use the plastic deformation nor long-time evaporation is required for low-cost performance. Therefore, none of previous perforation techniques satisfy the requirements of production of open-channel aluminum for high performance devices with low cost.
Recently the present author fabricated open-channel aluminum with long elongated channel holes by extraction of lubricated template wires from solidified aluminum.11,12) In order to perforate many holes in aluminum, a number of template wires should be extracted from aluminum by tensile force loaded to the template wires. Thus, smaller extraction force is desirable to fabricate open-channel aluminum with holes as many as possible. In the present work, first the extraction force of the stainless steel wires coated with boron nitride from solidified aluminum was measured and then the release agent was investigated to cause smaller extraction force, which should be smaller than that coated with boron nitride. Finally usefulness of porous layer formed in the release agent was found by adding magnesium hydride to boron nitride. This paper reports that addition of magnesium hydride to boron nitride reduced the extraction force of the stainless steel template wires. Furthermore, it is shown that small contact area of the wire to aluminum results in smaller extraction force, which is attributed to decrease in friction force of the wires to aluminum.
Open-channel aluminum was fabricated by extracting lubricated metallic wires embedded in a solidified aluminum. High purity aluminum (99.99% pure) was used as samples. Hard stainless steel wires, SUS304, were used as templates of open-channel holes. The thickness of the template wires was in the range from 0.28 mm and 4.0 mm in diameter. The length of the template wires was in the range from 20 mm to 100 mm. The inner dimensions of graphite crucibles were 50 mm × 20 mm × 150 mm. Boron nitride (grain size is 2 µm in diameter) or/and magnesium hydride was diluted with cyclopentane, which was coated onto the wires as a release agent. Both ends of lubricated template wires were mounted into holes of stainless steel plates perforated by high-speed precision micro-drilling machine, BDM-660, Nihon Seimitsu Kikai Kosaku Co., Ltd., Japan. The set of the templates and its supporting plates was inserted into a lower graphite crucible as shown in Fig. 1(a). Aluminum was melted at 953 K in atmospheric air in an upper crucible for 900 s. Then, the melt was dropped down by opening a stopper to be filled into the lower crucible (Fig. 1(b)). During the melting, a graphite piston was pressed onto the melt in order to fill the melt into the spaces among the template wires (Fig. 1(c)). Then, the crucible was directionally lowered at a speed of 30 mm min−1 to minimize shrinkage cavity of the ingot. The solidified aluminum ingot was taken out from the crucible. The embedded aligned wires were extracted by using a tensile test machine, and open-channel aluminum was fabricated (Fig. 1(d)). Tensile test machines, Shimadzu Co., Autograph AGX-V 5kN and AG-100kND were used to extract the template wires from the solidified aluminum and to measure the extraction force. Force necessary to extract the template wires from aluminum was defined as the extraction force. The extraction velocity was 3 mm min−1. The shape and the size of cross-sectional channel holes was observed with optical microscope. Three-dimensional images were measured by X-ray computed tomography (hereafter identified as X-ray CT.) with Micro-CT scanners TXS-32300FD, Toshiba IT & Control System Corporation. The voltage and electric current of the X-ray tube were adjusted at 160 kV and 120 µA, respectively. The target was tungsten.
Schematic drawing of the fabrication of open-channel aluminum. (a) The template wires coated with release agent are mounted in lower crucible in the furnace and aluminum is melted in upper crucible. (b) Aluminum melt drops down by opening the stopper. (c) The molten aluminum is filled up into the space among the wires by pressing a piston. (d) After solidification the template wires are extracted from aluminum.
An overview of the open-channel aluminum is shown in Fig. 2(a). The diameter and length of the holes are 1.1 mm and 34 mm, respectively. The number of holes and the porosity are 182 and 30%, respectively. The hard stainless steel wires whose diameter is 1.0 mm were used as the templates. The boron nitride was coated onto each template whose thickness was about 50 µm. X-ray CT scanning image of the open-channel aluminum are shown in Fig. 2(b). After the extraction of the template wires, the penetration ratio of the channel holes is 100%, and the diameter of the channel holes is in good agreement with that of extracted wires within the accuracy of ±5%.
(a) Overview of open-channel aluminum. Size 48 mm × 12 mm × 34 mm, hole size 1.1 mm ϕ × 34 mm. Porosity 30% and penetration ratio 100%. (b) X-ray CT scanning image of open-channel aluminum.
In order to investigate quantitatively whether coating of the release agent, boron nitride, is effective for extracting the template wires, first of all the template wires without coating of boron nitride were tried to be extracted. Figure 3(a) shows a plot of the extraction force versus extraction length of the uncoated template wire embedded in aluminum. The template wire was not able to be extracted and was fractured at the force of about 880 N, which is equivalent to the tensile strength of the stainless steel wire.13) The upper curve in Fig. 3(b) shows a plot of the extraction force versus extraction length of the template wire whose surface was coated with boron nitride of 50 µm in thickness. The extraction force was in the range of elastic tensile deformation. The template wires coated with release agent can be extracted by strength smaller than the yield strength of stainless steel wires. If the extraction strength increases to more than the yield strength, the template wires are fractured during the tensile elongation at a position which is not embedded in aluminum. In the upper curve, the template wire began to be extracted at the maximum force 43 N and the force decreased zigzag with increasing extraction length. It is surmised that ununiform friction through boron nitride located between the template wire and aluminum results in the zigzag shape.
(a) Extraction force-extraction length curve of template stainless steel wire of 1 mm in diameter embedded in aluminum, (b) extraction force-extraction length curve of template wire coated with BN, and extraction force-extraction length curve of template wire with the mixture of BN and MgH2.
When a template wire coated by the boron nitride is extracted from an aluminum ingot, the feasibility of extraction, that is, the extraction force, is thought to be attributed to the friction force mainly. The coefficient of friction μ is expressed by
| \begin{equation} \mu = F/W \end{equation} | (1) |
where F is the friction force and W is the load operated perpendicular to the friction plane. In the present case, since W is constant during the extraction process, the friction force is proportional to the coefficient of the friction. It is known that μ between iron and aluminum is 0.82 at room temperature14) and μ of boron nitride on metals is less than 0.2.15) Therefore, the friction force of stainless steel wire on aluminum is much higher than that of boron nitride on stainless steel or aluminum. The present result that the extraction force without boron nitride is much higher than that with boron nitride is reasonable, because the extraction is carried out due to the friction, although an alloying effect should be considered as another cause.
The lower curve in Fig. 3(b) shows a plot of the extraction force versus extraction length of the template wire whose surface was coated with a mixture of boron nitride and magnesium hydride of 50 µm in thickness. The extraction force of the template wire coated with the mixture of boron nitride and magnesium hydride is much smaller than that with boron nitride. Since the friction force became smaller in the case of the mixture, the template wire was extracted smoothly without showing zigzag behavior.
The extraction force is shown as a function of the weight ratio of WMgH2/WBN, where WMgH2 and WBN are the weight of MgH2 and BN, respectively, as depicted in Fig. 4. The thickness and length of template stainless steel wires are 1.0 mm in diameter and 40.0 mm, respectively. The extraction force decreases drastically in the range of WMgH2/WBN from 0 to 0.4 with increasing MgH2, and then, decreases slightly until about 1.5. This result indicated that even if more MgH2 was added, the extraction force kept almost constant. Since it is known that gas pores are formed by hydrogen decomposed with the reaction of MgH2 → Mg + H2 in the temperature range higher than 670 K,16,17) it is surmised that the porous layers may be formed to decrease the extraction force.
Dependence of extraction force on weight ratio of WMgH2/WBN in open-channel aluminum.
If such porous structure is formed around the interface between stainless steel wires and aluminum, dilation due to pores in the interface region is expected. In order to investigate the volume change due to the dilation, the cross-sectional views were observed. The boron nitride of 50 µm in thickness was coated to the stainless steel wires, while the mixture of 50 µm in thickness of boron nitride and magnesium hydride was coated to the stainless steel wires; the weight ratio of the mixture of BN and MgH2 was WMgH2/WBN = 1.5. Figure 5(a) shows the sectional view of stainless steel wire before extraction. The average diameter of the stainless steel wire was 1.08 mm, while the average diameters of channel holes of open-channel aluminum were 1.16 mm for coating with boron nitride and 1.45 mm for the mixture with boron nitride and magnesium hydride as shown in Fig. 5(b) and (c), respectively. Therefore, the increase in diameter is (1.16/1.18 = 0.98 ≅ 1.0) nearly zero before and after extraction. However, further increase of (1.45/1.18 = 1.23) 23% in diameter due to the addition of magnesium hydride was found, that is, the volume expansion of ((1.45/1.18)2 = 1.51) 51% was observed. Thus, it is considered that this expansion is attributed to formation of the porous structure as illustrated in Fig. 5(d) and (e).
(a) Sectional view of stainless steel wire embedded in aluminum, (b) sectional view of channel hole extracted by stainless steel wire coated with boron nitride, (c) sectional view of channel hole by stainless steel wire coated boron nitride and magnesium hydride, (d) schematic drawing of section of channel hole extracted by stainless steel wire coated with boron nitride, and (e) schematic drawing of section of channel hole extracted by stainless steel wire coated with boron nitride and magnesium hydride.
As shown in Fig. 4, the extraction force does not change and keep constant, even if the weight ratio of the mixture increases to more than 1.5. Accordingly it was observed that the diameter of channel holes do not increase in the weight ratio more than 1.5. Thus, additional MgH2 content does not influence the size of the channel holes. The reason is not solved at present. However, a few possibilities are considered.
The extraction force was measured as functions of the diameter and the length of template wires in open-channel aluminum. 50 µm in thickness of the mixture of MgH2 and BN whose weight ratio WMgH2/WBN = 1.5 was coated onto the template wires uniformly. Figure 6(a) shows a plot of the extraction force against the length of the template wires. The thickness of the template wires was 1.0 mm in diameter. Five data were taken to obtain an average value at each length. The result shows that the extraction force is directly proportional to the length of the wires. Figure 6(b) exhibits a plot of the extraction force versus the diameter of the template wires. The length of the template wires was constant to be 40.0 mm. The results were an average values of five measurement data. The extraction force increases linearly with increasing the diameter of the wires.
(a) Variation of extraction force with template length of open-channel aluminum. (b) Variation of extraction force with template diameter of open-channel aluminum.
When the template wires are embedded in the solidified aluminum, a contact area S of the template wire to surrounding aluminum is shown as
| \begin{equation} S = \pi d L, \end{equation} | (2) |
where d and L are the diameter and the length of the template wire embedded in aluminum, respectively. From the results of Fig. 6(a) and (b), the extraction force F is proportional to d and L, and thus, F ∝ dL.
Therefore, the followings are obtained
| \begin{equation} F \propto S. \end{equation} | (3) |
It is concluded that the extraction force F is directly proportional to the contact area. Therefore, the decrease in the contact area of the wires to aluminum reduces the extraction force of the wires.
The open-channel aluminum with directional holes is fabricated by extraction of template wires coated with boron nitride and magnesium hydride embedded in solidified aluminum. The results obtained are as follows.
The author wishes to express his appreciation to Director Kunihiko Koike, Mr. Jyun Miyazaki, Mr. Tetsuji Nakamura and Mr. Minoru Sato of Iwatani R&D Center, Iwatani Co. Ltd. for supporting this project and stimulating discussion.
