Sage Journals: Discover world-class research

Abstract

This paper presents a novel robot employed to manufacture space solar cell arrays. First of all including the mechanical configuration and control system, the architecture of the robot is described. Then the flow velocity field of adhesive in the dispensing needles is acquired based on hydrodynamics. The accurate section form model of adhesive dispensed on the solar cells is obtained, which is essential for the robot to control the uniformity of dispensing adhesive. Finally the experiment validates the feasibility and reliability of the robot system. The application of robots instead of manual work in manufacturing space solar cell arrays will enhance the development of space industry.

Keywords

Robot Space solar cell arrays Adhesive dispensing Laydown

1. Introduction

With the development of space industry, the technology and application are both required to be automatic. Because it is very complicated and meticulous, at present the manufacturing process of space solar cell arrays (SSCA) mounted on satellites is manual work absolutely. Supported by NSFC (National Science Foundation of China), our lab has been developing the special applications on manufacturing the space solar cell arrays and many research results have also been obtained. (Zhao Yanzheng & Fu Zhuang, 2005) presented a space solar cell bonding robot, including the system architecture and technology. It was the first application that can realize automatic bonding of the anti-irradiation cover-glass on the space solar cells. The methods of quality control and crack inspection of auto-bonding process were deeply studied respectively by (Zhao Y.Z. & Fu Z., 2004) and (Fu Z.; Zhao Y.Z. & Liu Y., 2004). The model of pipe flow velocity field in the process of adhesive dispensing was obtained by (Zhuang Fu & Yanzheng Zhao, 2005), which was necessary for the bonding robot to control dispensing adhesive on space solar cells. (Zhuang Fu; Yanzheng Zhao & Qinghua Yang, 2006) further studied the prediction model of the adhesive coating thickness on a space solar cell. All the above studies focus on the bonding robot system structure, the models of dispensing adhesive and inspecting the adhesive layer. This paper describes another latest robot developed for manufacturing solar cell arrays, which can realize the auto-laydown process from solar module to space solar cell arrays.

The process of manufacturing solar cell arrays is shown in Fig. 1. At present the manufacturing process from solar module to solar cell arrays is manual work absolutely. The process is as follows:

1)
First, many adhesive tapes with a lot of rectangle holes are mounted over the solar panel substrate.
2)
Then, the prepared adhesive is dumped over the substrate and struck off by a pattern strickle.
3)
Tear out the adhesive tapes and some rectangle adhesive would be left on the panel.
4)
Put the solar module on the adhesive and make sure each cell is correspondent with one rectangle adhesive field.

Fig. 1.
Manufacturing process of space solar cell arrays

Obviously, there are some severe deflects in manual work: 1)
First, the thickness of adhesive layer can't be controlled.
2)
Second, bubbles are easily brought about inside the adhesive which is vital to the solar cells in outer space.
3)
Furthermore, solar cells are easily stained by manual operation and it will result in very difficult clean operation and lower production rate.

The manual operation has not suited the development of new generation solar cells, which is described in (Martin A. Green, 2003). To avoid and reduce these problems, we developed a new robot system, as shown in Fig. 2, which can realize the auto-dispensing and auto-laydown operations.

Fig. 2.
The robot system

This paper is organized as follows. Section 2 describes the general robot system. Section 3 studies the dispensing process. And Section 4 presents the prediction model of adhesive shape dispensed on the solar cells. Experiment is implemented in Section 5. Finally conclusions are given in Section 6.
2. Architecture of the robot system

Our robot consists of the mechanism of adhesive dispensing and auto-laydown, a pneumatic system and a control system. The mechanisms of adhesive dispensing and auto-laydown are installed respectively on the Z-axis and X-axis of the XYZ three-DOF auto-moving mechanism. Since solar arrays are usually very large and at present the largest solar array is 4000 mm length and 2000 mm width, the XYZ three DOF auto-moving mechanism is employed as support foundation to make sure the high position precision. The main manufacturing process of solar arrays consists of two parts, dispensing adhesive on solar cells and auto-laydown solar module on solar panel. Controlled by robot, the adhesive dispensing mechanism is in charge of distributing adhesive on solar cells. It can be seen from Fig. 3 that ten syringes are fixed on the block of the dispening mechanism. Under the effect of mutual compressed air, syringes can coat adhesive on solar cells respectively.

Fig. 3.

Schematic diagram of dispensing mechanism

And the block can move along the lead screw by manual. Therefore, the dispensing height between the needles and solar cells can be adjusted at any moment. The laydown mechanism has two basic functions, gripping solar module and pressing it on the solar panel. And the mechanism diagram is shown in Fig. 4. Over 80 suction cups are mounted on the same air chamber. Powered by motor 4 and motor 5, both air chamber and suction cups can move vertically. Under the effect of motor 6, they can also turn around on the center axis.

Fig. 4.

Schematic diagram of laydown mechanism

The configuration of the robotic control system is illustrated in Fig. 5. The system consists of a 32-bit IPC computer with Intel Pentium 4 processor and a control box. It is responsible for the motion control of the robot and other I/O control. The Windows 2000 OS and control software are installed on the IPC. The computer can send signals to the electrical control box through the motion controller. At the same time, the encoders will feedback the position of the motor axes to the motion controller, which performs the precise position control.

Fig. 5.

The configuration of the robotic control system

3. Adhesive dispensing process

The adhesive used to be dispensed on the solar cells is a typical non-Newtonian fluid. Through studying the theory of dispensing adhesive, we can control the velocity and volume of adhesive flow. So the adhesive flow in the needles is looked on as the research object.

The coordinate is defined as shown in Fig. 6. The z coordinate coincides with the axis of the needle and the r coordinate is perpendicularly to it. In terms of dynamic balance, the following equation is obtained, $(P_{2} - P_{0}) π r^{2} - 2 π r l τ = 0$ (1)

Where, P₂ and P₀ are pressure on interface 2 and interface 1 respectively, τ is shear stress of adhesive. The constitutive equation of adhesive is described by $τ = η \dot{γ} = μ {\dot{γ}}^{n} = μ (\frac{d v}{d r})^{n}$ (2)

Where μ is viscosity coefficient, γ is shear rate, n is power exponent. So in terms of above two equations, we acquire the velocity field of adhesive flow in needles: $u_{z} = \frac{n}{n + 1} {(\frac{1}{2 μ} \frac{d p}{d z})}^{\frac{1}{n}} (R_{n} \frac{n + 1}{n} - r^{\frac{n + 1}{n}})$ (3)

Fig. 6.

Flow model of dispensing adhesive in the needle

The pressure on interface 2 is approximately equal to the one on interface 3. So pressure gradient can be written as: $\frac{d p}{d z} = \frac{p 2 - p 0}{L_{0}} \approx \frac{p 1 - p 0}{L_{0}}$ (4)

Where P₁ is the pressure of compressed air; P₀ is the standard atmosphere pressure. And the flow volume of adhesive in the needls is $Q = \int_{0}^{R} 2 π r μ d r = π {(\frac{1}{2 π} \frac{d p}{d z})}^{\frac{1}{n}} \frac{n}{1 + 3 n} R^{\frac{3 n + 1}{n}}$ (5)

4. Prediction model of adhesive dispensing

The dispensing tracks are determined and shown in Fig. 7. From the figure we can learn these parameters, the length h, the width m, the clearance d between tracks, and the rim a and k. All these parameters can be adjusted flexibly to meet requirements on the spot. All the solar cells on a solar array should be level approximately, which is mainly decided by laydown process on the solar panel. Furthermore, the key element affecting the laydown quality is the uniformity of dispensing adhesive. When it is dispensed on space solar cells along the tracks, the adhesive is in a certain section form. Determining the section parameters is necessary for robot to control the dispensing tracks and the thickness of adhesive layer. Based on this work, the weight of the whole solar array is also controlled effectively within a limited range which is helpful to reduce the payload of satellites. This paper proposes a prediction method and model of the adhesive section form on a space solar cell. The coordinate is defined as Fig. 8. The y axis coincides with the dispensing track center and the coordinate origin is on the needle center at the start point of dispensing. During dispensing adhesive, the syringe controlled by robot system moves ahead along y axis at the speed v₀. The circles, shown in Fig. 8, stand for the needle's movement track. Assume reference point p is any one of a solar cell. During dispensing process, the needle passes through the reference point along chord PP₁. Then, the distance between reference point and center of the needle changes all the time, namely that the dispensing radius r is variable. At the time of dispensing beginning, there is a relation r = op = R. And in terms of equation 1, the dispensing velocity of the reference point on the solar cell at this time is u_z = 0. Then, with movement of the needle, the dispensing radius of this point becomes small gradually and the dispensing velocity becomes great correspondingly. When point A reaches the reference one, the dispensing radius is the minimum and the dispensing velocity is the maximum: $u_{max} = \frac{n}{n + 1} {(\frac{1}{2 μ} \frac{d p}{d z})}^{\frac{1}{n}} (R^{\frac{n + 1}{n}} - c^{\frac{n + 1}{n}})$ (6)

Fig. 7.

Tracks of dispensing adhesive on a space solar cell

Fig. 8.

Movement sketch of a needle in the process of dispensing adhesive

Where, c = |OA|. And then the dispensing velocity of the reference point begins to become small accompanying the robot locomotion. When point P₁ reaches the reference one, the dispensing velocity returns back to zero. The change process of adhesive dispensing velocity of the reference point is presented in Fig. 9 and the shaded area indicates the dispensing velocity interval of some point on the solar cell. So the dispensing adhesive velocity of any point on the solar cell should be average and it can be described as following: $\bar{u} = \frac{1}{R - C} \int_{c}^{R} \frac{n}{n + 1} {(\frac{1}{2 μ} \frac{d p}{d z})}^{\frac{1}{n}} (R^{\frac{n + 1}{n}} - r^{\frac{n + 1}{n}}) d r$ (7)

Fig. 9.

The dispensing velocity interval of some point on the solar cell

Consequently, we have $\begin{aligned} \bar{u} \frac{1}{R - c} \frac{n}{n + 1} (\frac{1}{2 μ} \frac{d p}{d z})^{\frac{1}{n}} (\frac{n + 1}{2 n + 1} R^{\frac{2 n + 1}{n}} - R^{\frac{n + 1}{n}} c \\ + \frac{n}{2 n + 1} c^{\frac{2 n + 1}{n}}) \end{aligned}$ (8)

This is the definite expression of dispensing velocity field, where the space to the dispensing track center on the solar cell is c. Thus, we can obtain the thickness formula of adhesive layer dispensed on the solar cell before flowing around and it is described by $\begin{aligned} h (x) = \bar{u} \times Δ t = \frac{2}{v_{0}} \frac{n}{n + 1} (\frac{1}{2 μ} \frac{d p}{d z})^{\frac{1}{n}} (\frac{n + 1}{2 n + 1} R^{\frac{2 n + 1}{n}} \\ - R^{\frac{n + 1}{n}} | x | + \frac{n}{2 n + 1} | x | \frac{2 n + 1}{n}) \sqrt{\frac{R + | x |}{R - | x |}} \end{aligned}$ (9)

Fig. 10 shows the thickness distribution graph of adhesive of different distance to the dispensing track center at that time when adhesive is just dispensed on the solar cell. Here the needle's radius that we adopt is : R = 0.1mm.

Fig. 10.

The thickness distribution graph of adhesive

5. Experiment

If we adopt the manual way of dispensing adhesive, neither the volume of adhesive nor the area of dispensing can be controlled accurately. Then the phenomenon of outflow of adhesive takes place frequently, which will contaminate solar cells and deteriorate the quality of solar arrays. Therefore, the wiping operation is imperative in the manual operation and it will bring the danger of fragment solar cells. The auto-laydown robot system, shown in Fig. 11, is employed to implement the dispensing experiment and the test dispensing parameters are shown in Table 1. As shown in Fig. 12, it can be seen from the results that both adhesive shapes and dispensing areas are all controlled very well. Furthermore, no outflow occurs and all the solar cells have uniform adhesive dispensing areas and the same thickness layers.

Fig. 11.

Photo of the auto-laydown robot system

Fig. 12.

Results of dispensing adhesive controlled by robot

Table 1.

Experimental parameters of dispensing adhesive

No.	Test parameters	Values
1	Operating temperature (°C)	22
2	Setting track (mm)	a=4, k=5, d=1.6
3	Gas pressure (MPa)	0.15
4	Diameter of syringe needle (mm)	0.2
5	Speed for Robot's movement (mm/s)	1.5
6	Height for dispensing adhesive (mm)	1.5

6. Conclusion

We have described a novel robot, which can realize automatic manufacturing of space solar cell arrays mounted on satellites. The problems, resulting from manual work such as out of control of adhesive section form, outflow of adhesive, staining the solar cells and so on, can be avoided thoroughly. Based on studying both the robot movement and the dispensing process, the prediction model of adhesive shape on the solar cells is acquired and its validity is confirmed through experiment. The robot can be widely applied replacing manual work, so the quality of manufacturing solar arrays will be greatly enhanced, which is very helpful for the development of space industry.

Footnotes

7.

This work was partially supported by the National Natural Science Foundation of China under Grant No. 60675040.

References

Yanzheng

Zhao

& Zhuang

(2005). A space solar cell bonding robot. Journal of Shanghai Jiaotong university, Vol.39, No.6, (Aug., 2005), pp.888–891, ISSN 1006-2467

Zhao

Y.Z.

& Fu

(2004). Study on quality control in the bonding processing of space solar cell, Proceedings of 2004 International Conference on the Business of Electronic Product Reliability and Liability, Chan

Y.C.

(Ed.), pp.9–12, ISBN 0-7803-8361-3, Shanghai, Aprial 2004, the Institute of Electrical and Electronics Engineers, Inc., Hong Kong, China

Zhao

Y.Z.

& Liu

(2004). Solar cell crack inspection by image processing, Proceedings of 2004 International Conference on the Business of Electronic Product Reliability and Liability, Chan

Y. C.

(Ed.), pp.77–80, ISBN 0-7803-8361-3, Shanghai, Aprial 2004, the Institute of Electrical and Electronics Engineers, Inc., Hong Kong, China

Zhuang

& Zhao

Yanzheng

. (2005). Auto-bonding robot for space solar cells. Robotica, Vol.23, No.5, (9, 2005), pp.561–565, ISSN 0263-5747

Zhuang

Zhao

Yanzheng

& Yang

Qinghua

. (2006). A Prediction Model of the Adhesive Coating Thickness on a Space Solar Cell. Journal of Manufacturing Science and Engineering, Vol.128, No.2, (May 2006), pp.576–579, ISSN 1087-1357

Green

Martin A.

(2003). Third generation for photovoltaics: high conversion efficiency for low cost, Proceeding of 2003 Anniversary Solar Energy Conference of China Solar Energy Society, luguang

Yan

(Ed.), pp.2–8, ISBN 7-313-03492-X, Shanghai, Sept. 2003, Shanghai Jiaotong University, Shanghai, China