Effects of forming velocity on micro deep drawing performance with different blank thickness


 Micro deep drawing is a promising manufacturing method to produce the hollow, thin walled, cup or box like products at micro scale. Forming velocity can affect the products’ quality significantly due to the size effect, and this effect can be various with different thickness material. In this study, 30, 40, and 50 µm thickness stainless steels were annealed at 950 °C for 2 min under protection of argon gas ambient respectively. These different thickness steels were utilized in the micro deep drawing with different forming velocities. The experimental results show that, the profile accuracy and surface quality of the micro product are affected by changing the forming velocity with different thickness blanks. The micro cup has a less indentation area at the bottom and becomes rounder and more symmetrical with a thicker blank. Besides, the wrinkling phenomenon turns distinct with a thinner blank, and the earing becomes more significantly when increasing the drawing velocity or decreasing the blank thickness. When the drawing velocity or blank thickness increases, the surface of the micro cup becomes smooth and even. The experimental results are in good agreement with the simulation results, which confirms the developed finite element simulation model is applicable.


Introduction
Manufacturing at micro scale becomes important in the modern society due to the increasing demand of the micro products in auto, medical, electrical communication, and aerospace industries. Besides, the pollution, volume and cost of the products can be reduced significantly by minimising the forming scale [1,2]. The microforming technologies can be divided into two types, namely, microsyste m technologies (MST) and micro-engineering technologies (MET) [3]. Generally, MST contains micro wire EDM, lithographic technologies, and micromachining, which are widely used to produce micro components with high accuracy. However, these methods have high operation cost and can only be sued in a narrow range of workpiece materials. By contrast, MET are mainly related to mechanica l machining, and these technologies transfer the conventional manufacturing to the micro scale [3,4].
The MET can be utilized to manufacture high-precision mechanical components with the smooth surface. Furthermore, based on the type of machined materials, microforming can be classified as silicon-based or non-silicon-based manufacturing technologies [5,6]. Above all, plastic microforming is a preferred micro manufacturing technology due to the high productivity, low product cost, and less pollution. This microforming technology can be transferred from the conventional manufacturing process, which has well developed principle and theory at macro scale [7][8][9][10]. However, the principle and theory of conventional plastic forming cannot be directly transferred to the micro scale due to the size effect, which can cause the difference on the stress flow, formability, fracture behaviour, and friction of the material between the macro scale and micro scale [11][12][13][14]. To produce thin walled, hollow, box or cup like products at micro scale, micro deep drawing (MDD) is a fundamental process of the micro plastic forming, and more complex micro products can be produced based on this technology [2,15]. Therefore, it is essential to find a theory to be used in producing high-quality micro parts and explain the frictional behaviour during the MDD [16][17][18][19].
There are several researches have been conducted to improve the profile accuracy, surface quality, and forming ability of the micro products in MDD [16,[20][21][22][23]. Generally, friction and strain during the process are crucial elements to determine these characters of the micro cups [24]. Coating the diamond-like carbon (DLC) film on the tool surface is efficient to reduce the friction and improve the forming ratio in the MDD, and this method is not suitable for mass production [25]. Therefore, the nanoparticle lubricant can be used to improve the micro product's quality, namely, the lubricant can form an adhesive film to uniform the stress during the process [26][27][28][29][30][31]. Besides, the thickness of the drawing material and the drawing velocity can influence the friction and strain significantly, then affect the quality of the micro products [32][33][34].
This study presents an experimental study in forming different thickness austenite SUS301 micro cups with different forming velocities. The vertical view, side view and the surface quality of the micro products are observed by the 3-D laser microscope, and the efficiency of the MDD is expected to be enhanced via altering the drawing velocity. Besides, the products' quality can also be affected by the thickness of the material. Even though, the volume can be diminished distinctly with a thinner blank, the product quality could be defected more significantly. The finite element (FE) model of MDD process is established to simulate this study, the drawing velocity and experimental materia l thickness are altered to explore the relationship between these factors and micro product's quality, then compare the results between the experiment and simulation to confirm the developed FE model for MDD process [35][36][37].

Experimental material
The SUS301 sheet was used in this study, and the material was manufactured by cold rolling to obtain 30, 40, and 50 µm thickness metal pieces respectively, which are satisfied the specification of die sets in MDD. The chemical composition of SUS301 is listed in Table 1, and this material is annealed in 950 °C for 2 min in the KTL tube furnace under the protection of argon ambient, which is efficie nt to prevent the material oxidation.

Micro deep drawing
In order to manufacture the box-like product at micro scale, the MDD is carried out on a Desk-top press machine, which is shown in Fig. 1 (a). This press machine can provide up to 25 KN press force and the drawing velocity can be controlled by the control box as shown in Fig. 1 (b). The drawing process occurred in the die set, which contains the upper and lower dies, besides, the geometrica l parameters are shown in Table 2. During the MDD, the material was placed to the gap between the upper and lower dies, then the force was applied on the top of the upper die. As shown in Fig. 2, there are the force sensor, upper blank holder and punch in the upper die; the lower die contains spring, lower blank holder, and die cavity. This figure shows the principle of the MDD, and this process contains blanking stage and drawing stage. In the blanking stage, the lower die holder acts as the blanker punch, which moves upward and cut a round blank form the experimental material, then this blank is fixed by the upper and lower blank holders, and the punch moves downward to punch the round blank into die cavity. After that, a micro cup is manufactured.

FE simulation
To explore the effect of drawing velocity on the different thickness sample, FE modelling is an efficient method to determine this effect. In this study, the explicit dynamic analysis model was established as shown in Fig. 3, and 1/4 symmetrical model was set to simplify the simulatio n.
Furthermore, the blank holders and punch were set as rigid analytical bodies, and sheets were builded as deformed shell model with 0.03, 0.04, and 0.05 mm respectively. As the relevant process parameters, the lower blank holder and die cavity are immobile, and the punch moves along the vertical direction with 0.1, 0.2, and 0.3 mm/s velocity respectively. Besides, 20 MPa was applied on the upper blank holder before the blank leaved the blank holder, and the other tool parameter was set as same as Table 2.

Results and discussion
The influences of forming velocities on the micro part's quality were distinct at micro scale, since the quantity of formed grain is obvious lower compared with the macro scale, which means the effect of the forming rate on the single grain becomes more significant compared with macro forming. Besides, the volume of forming material can be changed significantly due to the variation of the blank thickness at micro scale, and the ratio of thickness reduction increases with a decrease of the forming scale.

Effect of blanking velocity on blank asperity with different thickness
The different forming velocities of the sample can alter the strain rate during the forming process, and this variation can affect the sample's forming ability with various thickness. Fig. 4 shows the profile and side view of the blank under different blanking velocities with different thickness (0.03, 0.04, and 0.05 mm). The edge profile and side view of the blank edge were captured by the laser scan microscope, and the sheet tends to bend into the die cavity during the blanking process. From Fig. 4 (a), the different colours on the blank edge show the bending area, and the blank edge is not smooth and even. It can be seen that 0.04 mm blank with 0.1 mm/s blanking velocity has the best edge quality comparing with other blanking velocity and thickness. Under this condition, the blank has the least bending area and jag phenomenon. Overall, the blank edge generates more jag phenomenon due to a rise of the blanking velocity with different sample thickness. Besides, the effect of blanking velocity becomes more obvious with a decrease of blanking thickness. The inhomogeneity of the material is more obvious with the thinner blank, because the less grain located on the formed path. Although, the forming ability of the blank is more homogeneous with the thicker blank, the blanking time can also increase with a rise of the thickness, then the possibility of bending and producing jag increases with more blanking time. Therefore, the bending and jag area of the 0.05 mm is more obvious compared with that of 0.04 mm sample due to the increase of blanking time.  force is the resistance of bending, and the other force is small relatively, then the drawing force increases slowly. After the blank was bended slightly, the blank was close to the die fillet and the drawing force decreased slowly. Since the blank is slipped in the die cavity, the main drawing force transfers to the forming force at the flange, the friction between the blank and blank holder, and the friction between the blank and the die cavity. The drawing force increases rapidly as the punch moves downward, then this force reaches the peak point and remains a short period. As the deformatio n continues, the drawing force decreases to the end of the drawing process, and the drawing force is not zero at the end of the forming process, because the drawn cup needs to release the stored strain energy. with a rise of the drawing velocity. Since the extending area is increased with a lower drawing velocity before the blank slips into the cavity, and the friction could increase with a larger contact area.
After the drawing force reached the peak point, the drawing force decreases significantly to the end of the drawing process, and the reduction of drawing force is more significant with a decrease of the drawing velocity. Furthermore, the drawing force is larger with a higher drawing velocity during this stage. Meanwhile, the energy can be stored in the drawn cup, the springback can release the stored energy gradually. The last drawing force can describe this phenomenon because the drawn cup squeezed the punch at the end of the forming process, and then the last stroke drawing force is generated. Furthermore, more force could compress the punch, when the more stored strain energy needs to be released. The last drawing force of all conditions are recorded and plotted in Fig. 6. It can be seen that the last drawing force increases with a higher drawing velocity, which means the micro cup can store more strain energy due to a higher forming velocity. The high drawing velocity can cause a large strain rate on the blank during the MDD, and the more force squeeze the punch at the end of the drawing stage. Since the blank thickness can alter the microstructure and formability of the material, the influence of the forming velocity on the last drawing force is also different with the various thickness samples. Generally, the last drawing force increased with a rise of the blank thickness under the same drawing velocity, and the difference of the last drawing force between 0.03 and 0.04 mm thickness is more obvious than that of this difference between 0.04 mm and 0.05 mm.
The effect of drawing velocity on the last drawing force is not significant with 0.05 mm blank, and this effect becomes obvious with a decrease of the blank thickness. The grain becomes more crucial in the smaller volume during the forming process. Therefore, the difference of grain growth under the heat treatments becomes more significant with the thinner blank due to the less blank volume, and the effect of forming velocity on the single grain is more obvious due to the lower quantity of grains on the forming path.

Micro conical-cylindrical cups
The profiles of the micro cups are almost same in the drawing die sets, and the experimental materia l thickness is a crucial element to affect the diameter and the height of the micro products. The profiles of different thickness micro cups are measured by the microscope, and the thickness of micro cup mouth can be measured by the inner and outer circle radius. Fig. 7 shows the micro cup mouth view with different thickness blanks, and the changes of the thickness are various. There is a slight increase on the micro cup mouth, when the blank thickness is 0.03 mm, and the thickness of cup mouth increased to 0.033 mm. However, the micro cup mouth became thinner when drawing the 0.04 or 0.05 mm blanks. The difference of micro cup mouth thickness variation among the 0.03, 0.04, and 0.05 mm blanks is caused by the distinction on the initial stage of the drawing process. Thus, the initial stage of drawing process contains the bending and compressing at the rim area of the blank and stretching on the wall area. When the blank thickness is 0.03 mm, the blank slips into die cavity easily, then the compression time and force diminish, and the bending occurs more on the interior of the blank due to the insufficient blank holder force. Besides, the stretching on the cup wall becomes significant at the initial stage of drawing process, then the micro cup mouth becomes thicker.
However, when the thickness is 0.04 or 0.05 mm, the blank can be fixed firmly at the initial stage, and then the rim area can be formed and stretched long. Thus, the micro cup mouth becomes thinner than that of the original blank when a 0.04 or 0.05 mm blank is drawn. Fig. 8 shows the stress states of different thickness blanks at the initial stage of drawing in the simulation, and these states are close to the real experimental results. In the simulation result, the stress focus on the lower rim and wall area, when the blank thickness is 0.03 mm. Besides, the focus area of stress moves to the rim of the blank with a rise of the thickness. Therefore, comparing with the blank original thickness, the micro cup mouth becomes thicker when drawing a 0.03 mm blank, and the cup mouth turns thinner when drawing a 0.04 or 0.05 mm blank. Furthermore, the profile accuracy of the micro cups could be improved due to the well formation on the blank rim. Fig. 9 shows the bottom view of the micro cups with different thickness, and the bottom of micro cup becomes rounder and more symmetrical with an increase of the blank thickness. When the thickness is 0.03 mm, the bottom of micro cup has a significant indentation, and this indention area decreases significantly with a thicker micro cup.    and 20 with 0.03, 0.04, and 0.05 mm blank thickness respectively. In the micro cup wall area, the stress increases with a rise of blank thickness, and this stress also be enhanced when decreasing the drawing velocity. Furthermore, the stress distribution is more uniform in the same height area with a thinner blank, and this distribution becomes more homogeneous under a higher drawing velocity.

Conclusion
This study presented an experimental research and numerical simulation on forming micro austenite SUS301 cups with different blank thickness under various blanking and forming velocities. The influence of forming velocity with different thickness blanks on drawn micro products' dimensio na l quality, drawing force, profile accuracy and surface quality were studied. The conclusions are as follows: 1. The drown product quality is affected with different blanking velocities and experime nta l blank thickness. 0.04 mm blank with 0.1 mm/s blanking velocity has the best edge quality compared with other blanking velocity and thickness. When the blank thickness increases, the effect of the blanking velocity on the side quality of drawn micro products becomes more significantly. More burr and uneven area can be generated with a higher blanking velocity, and the blank quality is crucial to determine the product's quality in the next forming stage.
2. The profile of micro cup can be modified by changing the blank thickness and forming velocity. It was found that, the micro cup has less indentation area, and becomes more round and symmetrical with a thicker blank. Furthermore, the wrinkling phenomenon becomes distinct with a decrease of blank thickness. The amount of wrinkling point increases and the wrinkling degree becomes more obvious with a rise of drawing velocity. The heights of micro cups are not even with different blank thickness under different drawing velocities, and the earing phenomenon becomes more significant with an increase of drawing velocity and a decrease of the blank thickness.
3. The surface quality could be improved by changing the drawing velocity and experime nta l material thickness. When increasing the drawing velocity or blank thickness, the surface becomes smooth and even. Furthermore, the micro cup surface could be rough and uneven with a rise of the blank thickness under the 0.1 or 0.2 mm/s drawing velocity, and the surface quality has an opposite trend when the drawing velocity is 0.3 mm/s.

Declarations
Ethical approval The authors declare that there are no conflicts with the ethical standards by Springer and the research conducted in context of this research paper.

Consent to participate
No research involving human subjects has been conducted in context of this publication and thus, a consent to participate is not needed.

Consent to publish
No research involving private companies has been conducted in context of this publication and thus, a consent to publish is not needed.

Competing interests
The authors declare no competing interests