Abstract
We used micro-double-staged laminated object manufacturing process (micro-DLOM) to fabricate 3D micromold. Moreover, the error of the micro-DLOM was also studied. Firstly, we got the principle error of the micro-DLOM. Based on the mathematical expression, it can be deduced that the smaller the opening angle α and the steel foil thickness h are, the smaller the principle error δ is. Secondly, we studied the error of femtosecond laser cutting. Through the experimental results, we know that the error of femtosecond laser cutting is 0.5 μm under 110 mW femtosecond laser power, 100 μm/s cutting speed, and 0.75 μm dimension compensation. Finally, we researched the error of microelectric resistance slip welding. Based on the research results, we can know that the minimum error of microcavity mold in the height direction is only 0.22 μm when welding voltage is 0.21 V and the number of slip welding discharge is 160.
1. Introduction
In recent years, with the development of microelectromechanical systems (MEMS), the microparts have been widely applied in industry areas. By using the conventional method, it is difficult to fabricate micro-parts due to their small size and high dimensional accuracy. With research in this field, it has appeared a series of modern manufacturing technologies for fabricating micro-parts [1–4].
By using UV-LIGA technology, Yan et al. fabricated hollow microneedle structure and they replicated the metal micro-needle structure via electroplating technology [5]. Using multidirectional ultraviolet (UV) lithography, Yoon et al. fabricated the specific 3D microstructure with inclined wall structure [6]. Based on the laminated manufacturing principle, Pfeiffer et al. used femtosecond laser to ablate 100 μm depth 3D micro-structures in cemented carbide and stainless steel plate [7]. Using a combined machining process based on microelectrical-discharge machining (EDM) and electrochemical polishing, Richter et al. fabricated microchannel with surface roughness of R a = 80 nm and star probe with sphere diameters of 40–200 μm [8]. Through the surface micromachining technologies and deep reactive-ion etching (DRIE) process, Zhang and Dong fabricated the active microprobe on a silicon-on-insulator (SOI) substrate and they applied it in the cellular force sensing and materials characterization [9].
The above researchers have done a lot of explorations and experiments on the fabrication of 3D micro-structure, which could be a good reference for the fabrication of 3D micro-structure. Based on the studies above, we proposed a novel forming process based on femtosecond laser cutting and microelectric resistance slip welding which can fabricate 3D micromold [10, 11]. This process is based on laminated object manufacturing (LOM) and it is composed of femtosecond laser cutting station and microelectric resistance slip welding station. Based on this, we named it as micro-double-staged laminated object manufacturing process (micro-DLOM). In the femtosecond laser cutting station, 10 μm thick 0Cr18Ni9 stainless steel foil is ablated to obtain 2D micro-structures. In the micro electric resistance slip welding station, multilayer of 2D micro-structures is welded and connected for approximately fitting the 3D micro-mold.
The machining error of process is an important parameter for assessing the quality of the process. The error of the micro-DLOM includes principle error of micro-DLOM, error of femtosecond laser cutting, and error of microresistor sliding welding. In this paper, we analyzed the error of the micro-DLOM and put forward some methods to reduce the error of the micro-DLOM.
2. The Micro-DLOM Process
The technological process of micro-DLOM is shown in Figure 1 and the main process is described as follows. (1) The desired 3D micro-mold which was constructed by the 3D modeling software was sliced into a stack of layers by a special converter. After doing that, the profile data of 2D micro-structure and the path data of resistance slip welding were obtained (Figure 1 (a)). (2) The mold base which was installed on the workbench fixture is moved to the micro electric resistance welding station and some welding spots are made to fix stainless steel foil on the mold base (Figure 1 (b)). Then, the mold base moves to the femtosecond laser cutting station. (3) In the femtosecond laser cutting station, the stainless steel foil is cut by the focused femtosecond laser beam based on the profile data of 2D micro-structure. After that, single layer 2D metal micro-structure would be obtained (Figure 1 (c)). (4) The waste generated by the femtosecond laser cutting is removed and the mold base should be moved down by a step distance (Figure 1 (d)). (5) The mold base is moved to the micro electric resistance welding station and the above process is repeated (Figures 1 (f) and 1 (g)). Finally, through the multilayer of 2D micro-structures superposition fitting, a 3D micro-mold is obtained (Figure 1 (h)). (6) Each layer 2D micro-structure of the 3D micro-mold is only connected through some welding spots and these welding spots could not achieve the full connection between each layer 2D micro-structure. So, the gap inevitably exists between each layer of 2D micro-structure. In order to eliminate the gap, the 3D micro-mold should be moved to the micro electric resistance welding station again and welded by the slip welding with multiple discharging (Figure 1 (i)).
Technological process of micro-DLOM.
Using the micro-DLOM, we fabricated micro-molds in five shapes: the square micro-cavity mold (Figure 2 (a)), the circular micro-cavity mold (Figure 2 (b)), the half-moon micro-cavity mold (Figure 2 (c)), the square micro-cavity mold with micro-channel (Figure 2 (d)), and the circular micro-cavity mold with cross keyway (Figure 2 (e)). The circular micro-cavity mold with cross keyway is a straight wall, and the other micro-cavity molds are uniformly extended based on the geometric center in the height direction.
SEM photograph of micro-cavity mold: (a) Square micro-cavity mold; (b) circular micro-cavity mold. (c) half-moon micro-cavity mold; (d) square micro-cavity mold with micro-channel; (e) circular micro-cavity mold with cross keyway.
3. Error Analysis of Micro-DLOM
3.1. Principle Error of Micro-DLOM
Micro-DLOM is based on laminated object manufacturing (LOM) and fabricates the 3D micro-molds through the superposition of multilayer of 2D micro-structures. So the micro-DLOM has a principle error, that is, the stepped effect. As the micro-DLOM is usually used to prepare micro-cavity molds, we used the cross section of the micro-cavity molds for calculating the principle error of micro-DLOM. As shown in Figure 3, d is dimension difference between adjacent steel foils in the horizontal direction, h is the thickness of the steel foil, α refers to the opening angle of micro-cavity along the vertical direction, and δ is the principle error of the micro-DLOM. According to trigonometric function, δ can be expressed as
Schematic diagram of principle error to micro-DLOM.
Based on the expression above, we calculated the principle error of micro-cavity molds which are shown in Figures 2 (a) and 2 (b): the maximum and minimum principle errors of square micro-cavity are 8.74 μm (h = 10 μm, d = 18 μm) and 1.96 μm (h = 10 μm, d = 2 μm), respectively; in the same way, the maximum and minimum principle errors of circular micro-cavity are 2.14 μm (h = 10 μm, d = 2.19 μm) and 0.99 μm (h = 10 μm, d = 1 μm).
By analyzing (1), we can know that within the range of 0∼90°, the smaller the α, the smaller the δ (principle error of the micro-DLOM) and vice versa. Therefore, for the calculation results above, we believe that the principle error of square micro-cavity is bigger than that of the circular micro-cavity because the opening angle α of square micro-cavity is bigger than that of the circular micro-cavity. Based on (1), it can be also inferred that for reducing the principle error of micro-DLOM, we can use the thinner stainless steel foils. For example, by replacing 10 μm thick stainless steel foils with 5 μm thick stainless steel foils to fabricate square micro-cavity mold (Figure 2 (a)), the maximum principle error of square micro-cavity reduced from 8.74 μm to 4.8 μm and the minimum principle error reduced from 1.96 μm to 1.8 μm. But when the thinner stainless steel foils are used in the micro-DLOM, it would sacrifice processing efficiency and foils cracking would easily happen.
3.2. Error of Femtosecond Laser Cutting
In the micro-DLOM, single layer of 2D micro-structure is mainly obtained through the interpolation of mechanical platform and femtosecond laser ablation. Therefore, the mechanical platform motion error and femtosecond laser cutting error have an important influence on the accuracy of 2D micro-structure. As the mechanical platform motion error is 0.2 μm, in this paper, we will mainly consider the influence of femtosecond laser cutting on the accuracy of single layer micro-structure. The factors which could influence femtosecond laser cutting accuracy mainly include femtosecond laser power and cutting speed.
In order to investigate the influence which femtosecond laser cutting speed has on the accuracy of 2D micro-structure, we adopted 10 μm thick 0Cr18Ni9 stainless steel foils for microcutting experiment. Steel foils were cut into 300 μm × 300 μm square hole under the same femtosecond laser power (200 mW) and different cutting speeds (1000 μm/s, 800 μm/s, 500 μm/s, 300 μm/s, 100 μm/s, and 50 μm/s). Through optical microscope, we observed the cutting edge morphology and measured the size of square holes. The measurement results are shown in Figure 4. From the results, we can know that, when the cutting speed increases from 50 μm/s to 1000 μm/s, the size square hole changes a little. Moreover, zigzag structure would appear in the cutting edge when the cutting speed is 1000 μm/s, 800 μm/s, 500 μm/s, and 300 μm/s (as shown in Figure 5). For this phenomenon, we believe that, when the cutting speed is more than 300 μm/s, the vibration of mechanical platform will occur and it would cause the zigzag structure of cutting edge. Therefore, for the considerations of cutting quality and machining efficiency, the femtosecond laser cutting speed was set to 100 μm/s in the paper.
Relation between cutting accuracy and cutting speed.
The cutting edge under the different cutting speeds (a) v = 800 μm/s; (b) v = 500 μm/s; (c) v = 300 μm/s; (d) v = 100 μm/s.
In order to investigate the influence which the laser power has on the accuracy of 2D micro-structure, we used the same stainless steel foils for the cutting experiments. Steel foils were cut into 300 μm × 300 μm square hole under the same cutting speed (100 μm/s) and different femtosecond laser powers (500 mW, 400 mW, 300 mW, 200 mW, and 110 mW). By using optical microscope, we observed the cutting edge morphology and measured the size of square holes. The measurement results are shown in Figure 6. From the results, we can know that, when the laser power increases from 110 mW to 500 mW, the size of the square hole gradually increases from 301 μm to 305.2 μm. Moreover, when the laser power is 110 mW, the cutting edge morphology is the best (as shown in Figure 7). From the results, we believe that, with laser power increasing, the ablation of steel foils gradually aggravates and this would affect the cutting accuracy. In addition, when the laser power is higher, the heat affected zone would appear near the cutting edge. The heat affected zone could also affect the cutting accuracy and cutting edge morphology. So, for the consideration of cutting accuracy and cutting quality, we set the femtosecond laser power as 110 mW.
Relation between cutting accuracy and laser power.
The square holes under the different laser powers. (a) P = 500 mW; (b) P = 400 mW; (c) v = 300 μm/s; (d) P = 200 mW; (e) P = 110 mW.
In order to investigate stability of the femtosecond laser cutting parameters above, we cut ten 300 μm × 300 μm square holes under 110 mW femtosecond laser power and 100 μm/s cutting speed. The size of each square hole was measured by the optical microscope and the measurement results are shown in Table 1. From the measurement results, we found that the maximum size of square hole is 303.1 μm and the minimum size of square hole is 300.3 μm. Based on the measurement results, we can calculate the dimension error of square hole as follows: the maximum error is 3.1 μm and the minimum error is 0.3 μm; the size of square holes which were obtained by the femtosecond laser ablation is basically larger than that of the designed square holes. The obtained size of 2D micro-structures could be larger than the designed size because the laser spot has a certain size. According to the measurement results, the average error of square hole is approximately 1.5 μm and it can be inferred that the laser spot radius is 0.75 μm.
Cutting accuracy with/without dimension compensation.
In order to obtain higher cutting accuracy, dimension compensation was used during the laser cutting and the value of dimension compensation was laser spot radius, which was similar to cutter radius compensation in NC machining. By using optical microscope, we measured the size of square holes and the measurement results are shown in Table 1. Based on the measurement results, we found that, before dimension compensation, the average values of side length 1 and side length 2 are 301.16 μm and 301.65 μm; after dimension compensation, the average values of side length 1 and side length 2 are 300.52 μm and 300.56 μm. So, the dimension compensation can improve the femtosecond laser cutting accuracy.
3.3. Error of Microelectric Resistance Slip Welding
In accordance with the micro-DLOM, a group of 3D micro-cavity molds were fabricated and their height was measured. From the measurement results, we know that the maximum height of micro-cavity mold is 93.33 μm and the minimum height is 65.79 μm, which are greater than the designed height (50 μm). Before femtosecond laser cutting, we only made some welding spots between the steel foils and these welding spots could not achieve full connection between steel foils. Therefore, a gap between stainless steel foils inevitably exists (as shown in Figure 8) and this gap could affect the dimensional accuracy of micro-cavity in height direction. In order to eliminate the gap between stainless steel foils and ensure dimensional accuracy of micro-cavity in height direction, micro-cavity mold should be moved to the micro electric resistance slip welding station for slip welding.
The gap between stainless steel foils.
The slip welding technology which was adopted in the paper is similar to the seam welding. With microbar electrode slipping on the steel foils, the steel foils gradually weld together and the gap between the steel foils would gradually eliminate. During the slip welding, the weld nugget will be formed between steel foils when the welding voltage is too high. As shown in Figure 9, serious deformation would appear around the weld nugget and this deformation will greatly influence dimensional accuracy of micro-cavity. Therefore, in order to ensure the accuracy of the micro-cavity mold, we should avoid forming weld nugget. Through the above analysis, the welding voltage should be as small as possible and we set it as 0.21 V (the smallest welding voltage of electric resistance welding machine is 0.2 V). With the micro-bar electrode on the steel foils and repeated discharge of the electric resistance welding machine, the heat is gradually accumulated, which could achieve the full connection between steel foils in the style of thermal diffusion.
Weld nugget forming during the micro electric resistance slip welding.
Based on the literature in [10], the micro-cavity mold will have the optimal dimension (50.41 μm) in height direction when the number of slip welding discharge reaches 160. In order to investigate the stability of the welding parameters above, we welded 14 micro-cavity molds under 0.21 V welding voltage and 160 time's slip welding discharge. After slip welding, the height of micro-cavity mold was measured and the results are shown in Figure 10. The experimental results show that, except number 1 and number 7, the rest of the micro-cavity molds obtain ideal height size (49.33 μm∼52.22 μm). With the experimental results analyzed, average error of micro-cavity mold in the height direction is 0.76 μm.
The height of micro-cavity mold after 160 time's slip welding discharge.
4. Conclusions
This paper used micro-DLOM process to fabricate 3D micro-cavity molds and the error of micro-DLOM is also analyzed. The details of the analysis are described as follows.
Micro-DLOM has a principle error. By analyzing the mathematical expression of principle error, it can be deduced that the smaller the opening angle α and the thickness of the steel foil h, the smaller the δ.
The femtosecond laser cutting was studied in detail and the results show that the minimum error of the femtosecond laser cutting is only 0.3 μm when femtosecond laser power is 110 mW, the cutting speed is 100 μm/s, and the value of dimension compensation is 0.75 μm.
Micro electric resistance slip welding is used to eliminate the gap between the steel foils and the average error of micro-cavity mold in the height direction is only 0.76 μm under 0.21 V welding voltage and 160 time's slip welding discharge.
Footnotes
Acknowledgments
This work was financially supported by the National Natural Science Foundation of China (no. 51175348) and the National Youth Science Foundation of China (no. 51205258). The authors are also grateful to the colleagues for their essential contribution to the work.