Materials Transactions, Vol. 49, No. 6 (2008) pp.

1365 to 1371 #2008 The Japan Society for Technology of Plasticity

Fabrication of Superplastic Microtubes Using Dieless Drawing Process*

Department of Mechanical Engineering, Tokyo Metropolitan University, Tokyo 192-0397, Japan Department of Mechanical Engineering, Takamatsu National College of Technology, Takamatsu 761-8058, Japan

Microtube is commonly used and required, particularly as micro components in micro-system technologies (MST) and micro electromechanical systems (MEMS). In this paper, fabrication process of microtube using superplastic dieless tube drawing was studied experimentally. Superplastic material used is Al-78Zn alloy tube with outer diameter of 2 mm and wall thickness of 0.5 mm. A high-frequency induction heating apparatus with air cooling nozzle was used for the dieless drawing. In the experiment of single pass dieless drawing, the eect of drawing conditions such as forming temperature, distance between heater and cooler and tensile speed on deformation prole, was claried. Furthermore, in three-pass dieless drawing, a microtube with outer and inner diameters of 343 mm and 161 mm respectively can be fabricated successfully. In addition, from the experimental results and fundamental principles, it is conrmed that the ratio of inner to outer tube diameters maintains a constant value during dieless drawing. In other words, the geometrical similarity with the minimization of dimension is satised in this process. Finally, it is found that the surface roughness of microtube maintains a constant value in the dieless drawing process. [doi:10.2320/matertrans.P-MRA2008810] (Received September 12, 2007; Accepted February 20, 2008; Published May 14, 2008) Keywords: dieless tube drawing, microtube, superplasticity, aluminum-78zinc alloy, drawing condition, drawing mechanism

1.

Introduction

Recently, the manufacture of ultrane and high-precision processed components has led to the enhancement of the functions of medical and biological equipment, and equipment used for communication, measurement, and information storage, and has contributed to developments in the elds of microsystem technologies (MST) and micro electromechanical systems (MEMS). The demand for precision components has been rapidly growing, and in particular, ultrane microtubes are expected to be applicable in various apparatus, such as mechanical/structural components, cooling micronozzles used in micromachines, electrode tubes used in electric discharge machining, contact probes, and painless needles.1) As a method of miniaturizing tubes, die drawing has been mainly used at present.2) However, in the fabrication of microtubes, high-precision microdies are required for conventional drawing, which makes it dicult to fabricate microtubes. Moreover, it is also dicult to insert a plug or a mandrel into a miniaturized tube. In addition, it is predicted that the increase in the friction resistance between the die and the tube upon miniaturization will result in a deterioration in the fabrication limit. In short, the use of microdies is a strong barrier to promoting the fabrication of microtubes. On the other hand, dieless drawing is a method in which a tube is miniaturized by locally heating a material, moving the heated part in the axial direction, and simultaneously drawing the tube.36) The characteristics of dieless drawing are that the miniaturization of tubes is realized by a high-exibility fabrication method without the use of dies and tools, and an reduction in area greater than that obtained using conventional drawing can be achieved. Although dieless drawing has long been proposed as a miniaturization process, its practical application has rarely been achieved because of the diculty in realizing high productivity and to
*This

date there have been no studies on dieless drawing being applied to micromachining. With this background, in this study, we aimed to fabricate ultrane microtubes by dieless drawing without the use of tools and ne dies, which are considered dicult to fabricate, by making use of the superplasticity underlying the fabrication principle of ne tubes used in glassworking. In particular, we experimentally examined the eectiveness of superplastic dieless drawing for fabricating microtubes. A prototype of a superplastic dieless drawing apparatus was constructed and the eects of the drawing conditions on the deformation behavior and shape of tubes were examined. Next, on the basis of the results obtained, microtubes with an outer diameter of approximately 340 mm were fabricated. Moreover, it was experimentally demonstrated that the ratio of the inner to outer diameters was maintained after deformation occurred during drawing, the underlying principle of which was additionally elucidated using elementary theory. Finally, the quality and surface properties of the fabricated microtubes were evaluated. 2. Superplastic Dieless Drawing

Superplastic dieless drawing is a fabricating method of a thin tube by the following procedure: x one end of a tubular material, as shown in Fig. 1, heat a part of the material then

V2 Heater A1

Cooler A2 V1 Z

Paper was Originally Published in Japanese in J. JSTP 47 (2006) 870874.

Fig. 1

Schematic illustration of superplastic dieless tube drawing process.

1366

T. Furushima, K. Manabe and T. Sakai

m value

r 1 A1 = A2 V1 = V1 V2

0.3 0.2 0.1 0 273 323 373 423 473 523 573 Temperature /K K value

300 200 100 0

Moreover, denoting the ow stress at the starting point of the deformation (the heated part) and that at the end point of the deformation (the cooled part) as 1 and 2 , respectively, the limiting reduction in area rc in the fabrication limit is obtained as7) rc 1 1 =2 2

From eq. (2), the fabrication limit during thinning depends on the ratio of the ow stress at the heated part to that at the cooled part. Namely, by applying superplastic materials, which have very low ow stress at the heated part, to dieless drawing, a signicant improvement of the fabrication limit is expected. 3. Experiment

Fig. 2

Material properties of Al-78Zn alloy at three dierent temperatures.

Universal testing machine High frequency induction heating device Cooling coil Air cooler Z-axis table Heating induction coil Al-78Zn alloy tube

3.1 Materials Al-78Zn superplastic alloy was used as a material. Because no tubular material of Al-78Zn superplastic alloy is available on the market, a commercial superplastic alloy sheet with a thickness of t 5 mm was machined to obtain a hollow billet with an outer diameter of D 4:4 mm and an inner diameter of d 1:1 mm, which was then extruded using a xed mandrel; thus, a tube with initial dimensions of D 2 mm and d 1 mm was prepared. The extrusion temperature was 523 K and a dry uoric lubricant was used. After extrusion, the tube was heated at 653 K for 2 h and then liqueed by rapid cooling in water containing ice. To determine the superplastic deformation characteristics of the tube, a jump tensile test was carried out. A tubular tensile specimen was used with a gage length of 16 mm. The test temperature was in the range from room temperature (R.T.) to 523 K, and the strain rate was in the range of 4:17 104 {102 s1 . The relationship between the ow _ during superplastic deformation stress  and the strain rate " _m , where K is the strength coecient is expressed as  K " and m is the strain rate sensitivity index. Figure 2 shows the values of m and K at each test temperature, which were obtained from the tensile test. The m increases with increasing test temperature and reaches approximately 0.5 at 523 K. In the case of superplastic materials, m is generally a high value of 0.3 or more; therefore, the results indicate that the specimen in this study exhibits sucient superplasticity. 3.2 Experimental procedure Figures 3 and 4 show a schematic and a photograph of the experimental superplastic dieless drawing apparatus used, respectively. In accordance with the necessity of achieving local heating and a temperature rise in a short time, a highfrequency induction heating device with a maximum output of 2 kW and a frequency of 2.2 MHz was used. A heating coil was xed at an output transformer and moved toward the

Fig. 3

Schematic illustration of superplastic dieless drawing apparatus.

Universal testing machine

Cooling coil

Tube

Heating coil High frequency heating apparatus Z-axis table

K value

draw the other end of the material at a speed V1 and move the heated deformation region at a speed V2 at the same time. The cross-sectional areas of the tube before and after deformation are denoted as A1 and A2 , respectively. Assuming that the deformation is in steady state, the desired reduction in area r is expressed as3)

0.6 0.5 0.4 m value

600 500 400

Fabrication of Superplastic Microtubes Using Dieless Drawing Process

1367

(a)
4 10

(b)
Insulation

1
6
Fig. 5 Schematic illustration of heating coils (a) Type 1 (b) Type 2.

sample using a z-axis table. Two types of heating coil were used. Figure 5(a) shows a coil made of two loops of copper pipe with an inner diameter of 10 mm and a width of 6.5 mm (Type 1). Figure 5(b) shows a special coil with an inner diameter of 6 mm and a width of 1 mm, which was made for the purpose of reducing the width of the heating region and the distance between the specimen and the coil (Type 2). The heated parts of the tube was rapidly cooled after deformation, leading to good progress of drawing without failure. Accordingly, a cooling coil connected to an air cooler using the Joule and Thomson eects was attached immediately above the heating coil. Superplastic dieless drawing experiments were carried out at a the speed of the z-axis table of V2 at 3.6 mm/min and at various tensile speeds, V1 2:5, 3.6, 5.0, 7.2, 8.4, 10.8, and 14.4 mm/min, Here, the maximum reduction in area for which the drawing was successfully completed without any breaks under the above speeds was dened as the limiting reduction in area. 4. Experimental Results of One-pass Superplastic Dieless Drawing

of a Al-78Zn tube shown in Fig. 2, it is found that the values of m are approximately 0.5 and 0.4 at 523 K and 473 K, respectively. When a part of the material starts to undergo local deformation during drawing at Th 523 K, the deformation resistance increases with increasing strain rate at that part, suppressing the growth of the local deformation. As a result, the unstable deformation was suppressed in the case of Th 523 K when m was greater, but was locally generated in the case of Th 473 K. During this drawing, most of the breakages occurred in the initial stage. Therefore, for successful fabrication, it is important to select an appropriate temperature corresponding to a suciently large value of m. Eects of distance between heater and cooler on limiting reduction in area To examine the eects of the shape of the heating coil on the optimal distance between the heater and the cooler, the temperature-time curve was measured using the thermocouple welded to the outer surface on the specimen under the condition of moving the z-axis table at V2 3:6 mm/min without applying tensile deformation to the specimen. Figure 7 shows the temperature distribution calculated by temperature-time curve of the specimen. When it is assumed that the region with a temperature of 333 K or lower corresponds to the cooled part in order to dene the distance between the heater and the cooler z, the temperature gradient between the heater and the cooler is greater in the Type 2 heating coil, than that obtained using Type 1. This indicates that the Type 2 heating coil has a shorter z: z ; 12 mm (z=D ; 6) for Type 1 and z ; 7 mm (z=D ; 3:5) for Type 2. Next, to examine the eect of z on the fabrication limit, we carried out dieless drawing for both heating coils under the conditions of V1 2:5 14:4 mm/min, V2 3:6 mm/min, and Th 523 K. Figure 8 shows a comparison between the limiting reduction in area for Type 1 (z=D ; 6) and Type 2 (z=D ; 3:5) heating coils. The limiting reduction in area for Type 1 was 58.1% and that for Type 2 was 80%. This reveals 4.2

Temperature /K

Eects of forming temperature on deformation prole We examined the eects of the forming temperature on the deformation prole. Photographs of longitudinal cross sections of the specimens deformed by drawing at 473 K and 523 K are shown in Fig. 6. Here, the drawing was carried out under the conditions of V1 7:2 mm/min and V2 3:6 mm/min (r 66:7%). Although drawing was possible at both temperatures, necking was generated at Th 473 K in the initial stage of the fabrication owing to the unstable deformation, whereas unstable deformation is suppressed in the case of Th 523 K. From the superplastic characteristics

4.1

6 30

573 523 473 423 373 323

Type1

(a)

Type2

(b)

Unstable deformation

1mm
-5

273 0 5 10 15 Distance from center of heating coil /mm

Fig. 6 Eect of drawing temperature on deformation prole (V1 7:2 mm/min, V2 3:6 mm/min, r 66:7%, Type 2) (a) Th 523 K, (b) Th 473 K.

Fig. 7 Comparison of temperature history in longitudinal direction of Types 1 and 2.

1368

T. Furushima, K. Manabe and T. Sakai

Limit of reduction in area r c /%

90
r c =80%

80 70 60 50 40 30 Type1 (z=12mm) Type2 (z=7mm) Distance between heater and cooler

ing with namely, an increase in the reduction in area. Under these conditions, the limiting reduction in area obtained during one-pass drawing was 80%. For a reduction in area of above 80%, fracture occurs at an unstably deformed part, resulting in failure of the tube. 5. Fabrication of Microtubes

Fig. 8 Eect of distance between heater and cooler on limiting reduction in area.

that the limiting reduction in area is improved by decreasing z. On the basis of these result, we hereafter adopted the Type 2 heating coil (z=D ; 3:5) in the superplastic dieless drawing experiments. 4.3 Eects of tensile speed on deformation prole Figure 9 shows the deformation proles of the tube obtained after superplastic dieless drawing under the conditions of a constant V2 of 3.6 mm/min and V1 in the range from 3.6 to 14.4 mm/min. With an increase in V1 , unstable deformation occurred in early stage of drawing accompany-

5.1 Multipass superplastic dieless drawing It is conducted from the results above mentioned that the following conditions are required to achieve a stable deformation prole of the tube during dieless drawing with large reduction in area in one-pass dieless drawing. (1) An appropriate forming temperature at which the strain rate sensitivity index m is high should be selected. (2) The distance between the heater and the cooler z should be decreased. (3) The tensile speed V1 should be greater than the z-axis table speed V2 . In addition, from the results of previous experiments, the conditions required for a ne tube prole without fracture during one-pass dieless drawing are as follows: forming temperature, Th 523 K; the speed conditions, V1 14:4 mm/min and V2 3:6 mm/min; a reduction in area, r 80%. Under these conditions, the outer and inner diameters of the tube obtained after one-pass drawing were D 894 mm and d 496 mm, respectively. On the basis of the above results, in order to fabricate ner microtubes, multipass superplastic dieless drawing was carried out starting from a tube. After one-pass dieless drawing with r 66:7% was carried out repeated three times, a microtube with D 343 mm and d 161 mm was fabricated successfully, as shown in Fig. 10 and Fig. 11(a). For comparison, the tube in the initial state before dieless drawing (D 2 mm, d 1 mm) is shown in Fig. 11(b).

(a)

Heating coil

(b)

(c)

(d)

Unstable deformation

5mm
Fig. 9 Tube proles drawn at various tensile speeds V1 by using heating coil type 2 (a) V1 3:6 mm/min, V2 3:6 mm/min, r 50% (b) V1 7:2 mm/min, V2 3:6 mm/min, r 66:7% (c) V1 10:8 mm/min, V2 3:6 mm/min, r 75% (d) V1 14:4 mm/min, V2 3:6 mm/min, r 80%.

Fabrication of Superplastic Microtubes Using Dieless Drawing Process

1369

500m
Fig. 10 SEM photograph of three-passed microtube (D 343 mm, d 161 mm).

343m

(b)
2mm 1mm

161m

(a)

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0

d Const. D

Diameter ratio, d /D

Heating coil Type 1 Type 2 500 1000 1500 Outer diameter, D / m 2000

Fig. 12 Variation in tube diameter ratio, d=D under various drawing conditions. Fig. 11 Comparison of vertical section in longitudinal direction of original tube and fabricated microtube under 1 pass-r 66:7%, 2 passesr 66:7%, 3passes-r 66:7% (a) 3 pass microtube (D 343 mm, d 161 mm) (b) Original tube (D 2 mm, d 1 mm).

6.

Principle Underlying Maintenance of Ratio of Inner to Outer Diameters

In conventional die drawing of non-superplastic materials, the reduction in area obtained by one-pass drawing was approximately 30%. Moreover, because the reduction in area per one-pass decreases with miniaturization of the tube, approximately 10 or more drawing passes are expected to be required for the fabrication of microtubes. Therefore, superplastic dieless drawing is an eective fabricating method of microtubes can be fabricated with a much smaller number of passes than that required in conventional die drawing. 5.2 Ratio of inner to outer diameters Figure 12 shows the relationship between the ratio of the inner to outer diameters d =D before and after deformation and the outer diameter D under each drawing condition. When the inner diameter becomes much smaller than the outer diameter i.e. d=D approaches 0, the inside tube is collapsed. The d=D of the initial tube was 0.5. It is seen in Fig. 12 that d=D is almost constant (approximately 0.5) even when the outer diameter decreases, namely, when the tube is miniaturized. During conventional die drawing, in general, the tube wall is thickened, and when the tube diameter is reduced, the inner diameter reaches 0; i.e. the tube becomes rod. In contrast the tube thickening is suppressed during dieless drawing, leading to fabricating microtubes with maintaining a constant diameter ratio. This means that dies such as mandrels and plugs is not necessary during superplastic dieless drawing. Miniaturization of tubes by using conventional die drawing must also be required for that of plugs and other dies, but it is very dicult to fabricate and insert such ne dies. Therefore, superplastic dieless drawing is a very useful method of fabricating microtubes.

As described in section 5.2, it was found that d=D was maintained after deformation by superplastic dieless drawing as shown in Fig. 12. In this section, we will ascertain the principle of this fact on the basis of elementary theory. First, we assume that the tube being deformed during superplastic dieless drawing is subject to uniaxial stress. Assuming that the tube is made of an isotropic material, the relation " "t holds for the circumferential strain " and the thickness strain "t from volume constancy. " and "t during dieless drawing are " lnD1 =D0 ; "t lnt1 =t0 ; 3

where t0 and t1 are the tube thicknesses before and after deformation, respectively. From Eq. (3), the following equation is obtained. D1 =D0 t1 =t0 4

Here, t0 and t1 can be expressed as eq. (5) using d0 and d1 . t0 D0 d0 =2; t1 D1 d1 =2 5

Using eq. (5), eq. (4) is expressed as d0 =D0 d1 =D1 const: 6

Namely, eq. (6) implies that the values of d =D before and after deformation are the same. The required condition for eq. (6) to hold is that the tube is subject to uniaxial stress during superplastic dieless drawing. Here, assuming that the length of the deformed part between the heater and the cooler is suciently greater than the tube thickness, it can be assumed that uniaxial stress condition is satised at the part. In addition, because superplastic materials, in general, have little anisotropy, the tube being subject to superplastic dieless drawing can be considered as an isotropic material; therefore, the above-mentioned principle of maintenance of d=D is satised. In short, this principle also shows the eectiveness

1370

T. Furushima, K. Manabe and T. Sakai

(a)

(b) Drawing direction

5m

(c)

(d)

Fig. 13 Microstructures of microtubes fabricated by superplastic dieless drawing process (a) 0 Pass (0.5 mm) (b) 1 Passes (0.7 mm) (c) 2 Passes (0.9 mm) (d) 3 Passes (1.1 mm).

of superplastic dieless drawing for the fabrication of microtubes.

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 1 2 Pass numbers 3 4 Grain size Surface roughness

2 1.8 1.6 1.2 1 0.8 0.6 0.4 0.2 0 Grain size /m 1.4

7. 7.1

Evaluation of Microtubes after Drawing

Eect of the number of drawing passes on microstructure The eect of the number of passes of superplastic dieless drawing on the microstructure of the microtube was examined. After mirror-polishing the surface of the microtube, we observed its microstructure using a scanning electron microscope. The grain size was calculated on the basis of the Hyde method. Figure 13 shows photographs of the microstructures of the microtubes before and after drawing with dierent numbers of passes. The average grain size of the initial tube was approximately 0.5 mm, whereas those of the tubes after one-pass, two-pass, and three-pass drawing were 0.7 mm, 0.9 mm, and 1.1 mm, respectively. The grain size slightly increased with increasing number of passes, but no marked grain growth was observed. Moreover, equi-axial grains were maintained, indicating the development of superplasticity during each pass of dieless drawing. 7.2 Eect of the number of drawing passes on tube surface property In superplastic dieless drawing, the smoothing of the surface using a die or a tool cannot be expected; thus, we must examine the changes in surface properties caused by the

Fig. 14 Eect of drawing pass numbers on surface roughness and grain size of microtubes.

deformation. To examine the surface properties of the specimen, we measured the roughness of the outer surface of the microtube using a non-contact surface roughness tester and evaluated the arithmetic mean deviation of the prole Ra . Figure 14 shows the eect of the number of drawing passes on Ra and grain size. Ra increases after one-pass drawing, but decreases after 2 or more passes, and the surface roughness after 3 passes was almost the same as that before deformation. From these results, the surface roughness of the

Fabrication of Superplastic Microtubes Using Dieless Drawing Process

1371

fabricated microtube depends on that of the initial tube, suggesting that the preparation of initial tubes with good surface properties is one method of ensuring the fabrication of high-quality microtubes. As described in the previous section, the grain size increases with increasing number of drawing passes. This means that increase in grain size does not request in increasing the surface roughness. 8. Conclusions

In this study, we examined the eectiveness of superplastic dieless drawing for the fabrication of microtubes. The study and its results are summarized in the following points. (1) The eects of test conditions during superplastic dieless drawing, such as the forming temperature Th , drawing speed V1 , and the distance between the heater and cooler z, on deformation behavior were experimentally examined. Moreover, the basic deformation characteristics of the drawing were determined and the index for fabricating microtubes has been obtained. (2) By carrying out the drawing with an reduction in area of r 66:7% three times, microtubes with an outer diameter of D 343 mm and an inner diameter of d 161 mm were successfully fabricated. (3) It was demonstrated that the ratio of the inner to the outer diameters of the miniaturized microtubes was maintained after deformation by superplastic dieless drawing, the underlying principle of which was claried on the basis of elementary theory. (4) It was demonstrated that the surface roughness of the microtubes does not increase when up to 3 drawing passes are carried out. The most signicant characteristic of superplastic dieless drawing is that dies and plugs, which are essential in conventional drawing, are not necessary. Therefore, the dieless drawing method outlined in this study enables the fabrication of components without exposure to friction, which would be

a serious problem due to the dimension eect. Furthermore, the reduction in area ratio can be controlled by arbitrarily controlling the fabrication speed without using any special dies or tools. Fields in which superplastic dieless drawing is applicable include the fabrication of micronozzles and stepped tubes. Moreover, as a benet of the dimension eect due to miniaturization, the improvement of fabrication eciency owing to the increase in heating and cooling speeds should also be noted. In the future, by realizing the accurate measurement and control of temperatures at a microscale, the use of dieless drawing as a highly exible microfabrication process is a strong possibility. Thus, superplastic dieless drawing is a method that shows signicant advantages for microfabrication compared with the fabrication of macroscale components. If hollow microcomponents with dierent shapes are fabricated by this process in the future, further development of microfabrication techniques can be expected. Acknowledgment This study was supported by a Grant-in-Aid for Scientic Research (Exploratory Research). REFERENCES
1) E. Nakamachi: Journal of the Japan Society of Mechanical Engineering 107 (2004) 45. 2) R. Sato: J. JSTP 29 (1988) 439. 3) H. Sekiguchi, K. Kobatake and K. Osakada: J. JSTP 17 (1976) 6771. 4) K. Kobatake, H. Sekiguchi, K. Osakada and K. Yoshikawa: J. JSTP 20 (1979) 814819. 5) K. Kobatake, H. Sekiguchi, K. Osakada and K. Yoshikawa: J. JSTP 21 (1980) 5258. 6) L. Yonggang, R. Q. Nathaniel and K. Aravinda: Mater. Sci. Eng. A 358 (2003) 5970. 7) H. Sekiguchi: Text from the 98th Symposium for the Technology of Plasticity (1985) 2135.