1. Introduction

Nowadays, the solar energy market size is still growing rapidly, driven by the rapid climate change caused by an increase of greenhouse gas emissions in recent years. However, the construction cost of solar energy conversion facilities is still expensive in terms of cost per unit electrical energy. Concentrated solar power technology focuses a large area of sunlight or solar thermal energy onto a much smaller receiver or exit using, which is able to produce energy more economically and efficiently than conventional solar systems by increasing the energy conversion efficiency and reducing the facility cost.

A linear Fresnel lens can focus collimated light to a strip in one dimension, and realize light concentration with much thinner lens materials compared to conventional lens. It has a wide variety of applications due to its light weight, small size and excellent optical performance, especially in concentrated solar power systems [1, 2]. However, mass production of large-area linear Fresnel lenses using traditional plastic injection molding is costly, because of the expensive inserts and the low-throughput manufacturing process.

Fresnel lenses are usually fabricated by ultra-precision diamond turning piece by piece or by plastic injection molding batch by batch. Roll-to-roll(R2R) embossing is an advanced continuous manufacturing method which is able to produce large-volume high-quality micro/nano surface structures on flexible film substrate at significantly lower cost and higher throughput compared to traditional manufacturing processes [3, 4]. It has already been widely used for mass production of brightness enhancement film as an important unit in the backlight modules of liquid crystal displays (LCDs) [5]. Unfortunately, very few studies can be found on manufacturing of optical linear Fresnel lens film using R2R embossing until now [6]. Hence, in this study, a complete manufacturing cycle of optical linear Fresnel lens polymer film are first studied in the aspects of ultra-precision diamond machining of metal roller mold, roll-to-roll embossing, and measurement test on film profile and functionality.

2. Machining of metal roller mold

2.1. Linear Fresnel lens design

A Fresnel lens is able to significantly reduce the volume of lens material compared to a conventional spherical lens by dividing the lens into a set of concentric annular prism rings. Through a Fresnel lens, a large area of sunlight is collected and directed onto a significantly smaller area by refracting the light rays, where a high-efficiency solar cell is located.

There are two basic design principles used in deriving Fresnel lenses from conventional spherical lenses: (1) equal ring height, and (2) equal ring width. In this study, the R2R UV embossing process will coat a thin layer of UV-curable liquid resin on the film substrate using a slot die coater, which is later embossed using a metal roller mould patterned with the desired functional microstructures. Due to such characteristic nature of R2R embossing, a constant height of the Fresnel lens is preferred in order to avoid large-volume transverse flowing of the liquid resin during the embossing process. Hence, the first design principle is applied to realize an even distribution of the resin along the transverse direction and precisely replicated microstructure pattern. For each ring, the light incidence angle, ${\alpha }_i$, and the refraction angle, ${\beta }_i$, should follow the refraction law:

The linear Fresnel lens has been designed to have 166 (2 × 83) prism rings with a constant ring height of 100µm and a width W of 55.6mm. The designed focal distance is 82.5mm, which is approximately 1.5 times of the lens width. It also has a constant draft angle θ of 2þ, and its designed facet angle φ_i ranges from 4.76þ to 29.49þ from the lens center to the edge. Figure 1 shows a schematic section view of the designed linear Fresnel lens.

 

Fig. 1 Construction features of the designed linear Fresnel lens.

Download Full Size | PDF

2.2. Ultraprecision machining of roller mold

In the R2R embossing system, the high-precision roller mold acts as the critical component and is very difficult to be fabricated. As the functional microstructures on the optical films are directly replicated from the roller, a high-precision roller mold is required with strict profile accuracy control and optical surface quality which is difficult to be fabricated in terms of tool trajectory optimization and tool/workpiece setting. The roller mold is made of brass with a diameter of Ø170mm and a length of 235mm. It was machined using a single crystal natural diamond tool with a 12.7µm nose radius and a 40þ included angle. In the design of linear Fresnel lens, all the corner edges of the prism rings are assumed to be perfectly sharp for simplicity. However, in generating trajectory of the diamond cutting tool, the peak corner edge of each prism ring has to be rounded with a small-radius curve in order to avoid micro cracking of sharp edges, and the valley corner must be larger or equal to the tool nose radius, limited by the inherent characteristic of such machining process. Hence, the peak and valley corner edge radii are determined to be 5µm and 12.7µm respectively in this study. Compared to a sharp-nose diamond tool, the theoretical diffractive efficiency for the designed Fresnel lens machined using a 12.7µm tool is calculated to be 5.1% lower. A total of three identical linear Fresnel lenses with the above design were machined on the brass roller mold.

The brass roller mold was held on a 5-axis ultra-precision machine system (Moore Nanotech 350FG) together with a diamond tool, as shown in Fig. 2(a). Based on the lens design above, a tool path for ultra-precision machining of the metal roller mould patterned with rotatory-symmetrical Fresnel lens is generated. In machining of the linear Fresnel lenses on roller mold, the integral lens profile cannot be generated by one-off setting of the diamond tool. Hence, the lens profile was divided into two sectional profiles (left and right sections), and two settings with different tool swivel angles were applied by rotating the machine B-axis to generate the whole Fresnel lens profile on the roller, as shown in Fig. 2(b). An optical inspection system attached to the ultraprecision machine was used to measure and record the location of the diamond tool to the B-axis at the two different swivel angles. The dividing position for the left and right lens profiles is selected at the peak corner of the left first prism ring, in order to avoid the generation of a visible parking line if the separation position is located at the centre.

 

Fig. 2 (a) Machined brass roller mold held on an ultra-precision machining system. (b) Schematic illustration showing the tool path and tool-workpiece setting for the machining process (Not drawn to scale).

Download Full Size | PDF

2.3. Profile measurement of the roller mold

A Taylor-Hobson stylus profilometer was used to measure the machined linear Fresnel lenses on the metal roller mold, as shown in Fig. 3(a).The stylus tip used to measure the lens profile is made of tungsten carbide, with a tip radius of 20µm and an included angle of 45þ. This large tip radius will not affect the measured profile accuracy, because its radius value has already been considered and compensated in processing the measurement data. The roller mold was located on a linear & rotary stage, which can realize fine adjustment of the roller position using micrometre. Pre-alignment was conducted to ensure that the roller axis is parallel with the horizontal movement of the stylus head, and the stylus was located at the highest position of the roller. Figure 3(b) shows the measured profile of one linear Fresnel lens on the roller mold. It can be seen from the enlarged figure areas that the width of the prism ring gradually decreases from the centre to the edge, which is caused by the increasing facet angle with the radial distance from the centre.

 

Fig. 3 (a) Close view of the measurement setup. (b) Measured surface profile of one linear Fresnel lens pattern. (c) Measurement process considering the shape of stylus tip. (d) Deviation of measured facet angles on the roller mold from the designed ones.

Download Full Size | PDF

It is interesting to observe that the measured prism ring height is also decreasing in the meantime, but the designed prism ring should have a uniform height. This mismatch is caused by the inherent measurement mechanism using the stylus profilometer. Due to the steep profile caused by the small draft angles (2þ in this study), the finite shape of the stylus cannot penetrate into the valley corners, hence not giving the true measurement data around the valley corners. When the stylus is measuring the (i + 1)^th prism ring, the stylus tip will contact the facet surface at a position higher than the i^th ring, because of the increasing facet angle (${\phi }_{i+1}>{\phi }_i$) from the center to the edge, as shown in Fig. 3(c).

Facet angles play a critical role in the profile of a Fresnel lens, and its angular accuracy for the machined roller mold will directly determine the quality of R2R embossed linear Fresnel lens film, and accordingly the capability of light concentration and optical efficiency. By filtering the incorrect profile data caused when measuring the draft faces, the deviation of measured facet angles from the designed ones for each prism ring can be calculated, as shown in Fig. 3(d). It can be observed that 93% of the deviation errors fall in the limit of ± 0.05þ, and the maximum deviation is −0.082þ. It can also be observed that the deviation error becomes relatively larger when the index of prism ring increases. Such increased deviation should be caused by the more significant arcing motion that occurs when the profilometer stylus vertically swings more sharply during scanning a steeper profile, and the decreased ring width which leads to less data points calculated.

3. Roll-to-roll embossing

R2R embossing applies the continuous replication process in order to drastically increase the patterning speed and thus the throughput. In this study, the fabricated high-precision roller mold was installed on a self-developed R2R ultraviolet (UV) embossing system. The system is constituted of two main modules: a coating and embossing module, and a web handling module. The coating and embossing module is located at the centre of the machine and contains all the components required for the coating, embossing, and curing of photosensitive resin on the flexible film substrate through UV light exposure. The key components include a slot die coater, an embossing roller, and a UV lamp. The coating width of the slot die is 200mm. By using different flexible moulds, various micro structures can be fabricated by R2R UV embossing.

As shown in Fig. 4, the web handling module is located at two ends of the embossing system, to drive the flexible film substrate, control the web-speed, and provide tension to the film substrate in the coating and embossing processes. A raw polymer film roll is installed on a motorized roller at the de-reeler station, which fed the film substrate into the coating and embossing module in a web speed of 2 m/min. Then the film substrate is evenly coated with a certain thickness of uncured resin using a slot die coater. The liquid uncured resin has a viscosity ranging from 70 to 100cP, and the coating thickness can be varied by changing the pneumatic pressure applied to the liquid resin reservoir. Four different pneumatic pressures (0.35bar, 0.5bar, 0.6bar and 0.65bar) were applied in this test to study the optimized pressure. The embossing process is realized by replicating the microstructure pattern from the roller mold onto the coated film, which is cured by UV light meanwhile. The polymer resin has a shrinkage ratio of 1% in volume after the curing process, which should pose limited effects on the profile accuracy of embossed Fresnel lenses. After the coating, embossing and curing processes, the film substrate consisting of embossed micro features is received by another motorized roller at the rewinder station.

 

Fig. 4 Schematic and physical views of the self-developed roll-to-roll UV embossing system.

Download Full Size | PDF

4. Measurement test

4.1. Profile measurement of the embossed lens film

After fabricating the optical linear Fresnel lens film, it is necessary to evaluate this manufacturing technique based on measurement results of the embossed films. However, it is difficult to obtain accurate surface profile data of the embossed optical Fresnel lens films, because they are flexible, transparent, and of high surface finish quality. Traditional optical measurement methods (e.g. 3D microscope, confocal laser scanning microscope, white light interferometer, etc.) will not work because very few light rays from the lens profiles could pass through the lens module and be captured by the light detection sensor.

Stylus profilometer is generally utilized to measure various surface profiles with high accuracy, but it is challenging to maintain levelness of the flexible film in a large area without warping and deflection. Hence, in this study, a high-flatness porous vacuum chuck was used to hold the film evenly, as shown in Fig. 5(a).2D and 3D measurement was conducted using the stylus profilometer with the assistance of an additional positioning stage, which is perpendicular to the traverse direction of the profilometer. It can be seen from Fig. 5(b) that the linear Fresnel lens pattern were well replicated from the embossing roller mold onto the film substrate, and no pores or defects were observed. The deviation of measured facet angles from the designed ones for each prism ring is shown in Fig. 5(c). It can be observed that 93% of the deviation errors fall in the limit of ± 0.5þ, and the maximum deviation was found to be −0.957þ. Such 10x higher angular deviations than the mold are expected, because R2R embossing will introduce larger mounting and replication errors compared to conventional plastic injection molding due to the large size of roller mold and its unique continuous replication and curing processes.

 

Fig. 5 (a) Measurement of R2R embossed linear Fresnel lens film using a stylus profilometer. (b) Measured 3D profile of one linear Fresnel lens pattern. (c) Deviation of measured facet angles on the Fresnel lens film from the designed ones.

Download Full Size | PDF

4.2. Functionality test

Finally, the capability of light concentration of the R2R embossed Fresnel lens film was evaluated by measuring the relative light irradiance distribution at the focal plane. Figure 6(a) shows the measurement setup for the light concentration test. The measurement was carried out using a reference solar cell with an aperture (4mm × 4mm), a computer controlled linear stage, and a Wacom super solar simulator. The known short-circuit current of the reference solar cell at 123.9 mA was used to calibrate the super solar simulator with a beam divergence angle of <2.5þ to obtain a solar irradiance of 1000 W/m². The linear Fresnel lens film was pasted on a transparent plastic plate made of PC, which was then mounted vertically and aligned to the direction of measurement. The Fresnel lens film was irradiated normally by the super solar simulator, and the reference solar cell was vertically mounted on the linear stage and positioned at the focal distance.

 

Fig. 6 (a) Measurement setup for light concentration test of linear Fresnel lens film (three identical lens patterns). (b) Measured concentration ratio with respect to the position of the referenced solar cell.

Download Full Size | PDF

The reference solar cell was scanned along horizontal axis with interval of 0.5mm and the readings of voltage V_f were taken at each point and recorded in the computer. Then the linear Fresnel lens film was removed and the voltage reading was recorded as V₀ and the voltage reading with the solar simulator turned off was recorded as V_d. Hence, the light concentration ratio, R, can be calculated as follows:

Figure 6(b) shows the measured relative irradiance distribution on the focal plane of the Fresnel lens film. The concentration ratio was measured to be 6.33 for the receiver area of 4 × 4mm². As R2R embossing usually leads to larger lens profile errors compared to the lenses manufactured by injection molding, the embossed Fresnel lenses may have slightly smaller concentration ratio with the same lens design.

5. Summary

In summary, we have firstly experimentally investigated the complete manufacturing cycle of linear Fresnel lens polymer film for solar concentration in terms of ultra-precision diamond machining of metal roller mold, roll-to-roll embossing, and measurement test on film profile and functionality. The roller mold is machined on a 5-axis ultra-precision machining system with 93% of measured facet angle deviations less than 0.05þ against designed ones. Linear Fresnel lens film is embossed on a self-developed R2R UV embossing system using the machined high-precision roller. Profile measurement of the film is conducted on a stylus profilometer assisted by a high-flatness porous vacuum chuck with 93% of the measured facet angle deviations less than 0.5þ. Functionality test is conducted on a solar simulation system with a reference solar cell, and results show that strong light concentration is realized on the designed focal plane of the linear Fresnel lens film. This study will help researchers and engineers to better understand and practically apply R2R manufacturing of functional optical films for solar concentration. Further studies will be conducted on R2R manufacturing of radial Fresnel lens films, and constantly improving the profile accuracy of manufactured lenses and their concentration efficiency.