Open Access
Add to BibSonomyAbstract
We introduce a type of LED light-gauge steel frame lamp with an adjustable illumination light field that does not require a diffusion plate. Base on the Monte Carlo ray tracing method, this lamp has a good glare rating (GR) of 17.5 at 3050 lm. Compared with the traditional LED light-gauge steel frame lamp (without diffusion plate), the new type has low GR. The adjustability of the illumination light field could improve the zebra effect caused by the inadequate illumination light field of the lamp. Meanwhile, we adopt the retinal image analysis to discuss the influence of GR on vision. High GR could reflect stray light on the retinal image, which will reduce vision clarity and hasten the feeling of eye fatigue.
© 2014 Optical Society of America
1. Introduction
To address the energy crisis and slow down global warming, energy conservation, environmental protection, and green energy have become vital global concerns. Thus, saving energy and developing alternative energy sources are critically important. Many types of lamp fixtures are currently available in the market. Light-emitting diodes (LED) have certain advantages over other lamps. These advantages include compact size, power conservation, and long lifespan. LED has been drawing considerable attention as a replacement for fluorescent and incandescent lamps as light source [1]. As a result, LED has been extensively applied to vehicles, buildings, traffic signs, and indoor light systems. Recent studies on LED lighting include those on special LED packaging [application-specific LED packaging (ASLP)], which is even smaller than the volume and size of traditional LEDs (merely an eighth of traditional LEDs). ASLP has higher system luminous efficiency (8.1%) and lower manufacturing cost [2]. After eight instances of feedback, the smooth and free lens design with rectangular luminance distribution exhibits uniformly increasing luminance from 18.75% to 81.08% [3]. Low-beam and high-beam lenses improve the illumination efficiency of automotive lighting with LED lamps [4]. An effective optical design is adopted to solve the problem of prescribed luminance caused by LED chip array packaging (LCAP), which could improve the uniform luminance and brightness of road lighting. Compared with the traditional optical system, the longitudinal and overall uniformity of LEDs are 0.7 and 0.5, respectively, and the optical efficiency also increases by 19.6% [5]. The lens design with free curved surface applied to LED collimation lighting validated that higher optical efficiency could be obtained with the same angle of view [6].
Aside from some challenges pertaining to research and development, improvements, and manufacturing technologies, lamps are also confronted with the glare problems. Recent studies on LED lighting glare include that of Boyce and Ngai on the glare problem of head-wear light source. Glare could be generated by the different angles of illumination. The researchers found that discomfort glare could occur when the horizontal angle of light exceeds 53° [7,8]. Lee et al. evaluated the study on LED lamps with discomfort glare and found that non-uniform lighting will cause eye discomfort [9]. Takahashi et al. demonstrated that the difference between uniform and non-uniform lighting will disappear at a larger perspective angle based on the light source position index matrix [10]. We found that substantial problems related to eye discomfort are attributed to glare when LED lamp is applied for lighting in these studies. These problems are worthy of further studies.
Glare factors, such as luminance on eyes, brightness, size, angle, and duration of glare source, affect visual performance. Hence, we divide glare into two categories, namely, disability glare and discomfort glare [11]. Disability glare refers to the decline in visual ability, that is, the scattered incident light seriously affects the visibility of the viewed target by developing a light curtain on the retina. Therefore, the viewed target has a diminishing contrast and visibility, which normally occur with a dark background. The high-intensity light source illumination on the eyes will temporarily blind a person and will trigger dangerous situations. Discomfort glare refers to the discomfort of viewers when the light source does not reduce the discernibility of the eye. This type often occurs indoors. Such discomfort will cause irritation, dazzling, or fatigue after extended exposure to glare from the light source illumination. Such discomfort also has a psychological influence.
Most anti-glare devices in LED lamps adopt diffusion plates as the main axis. However, these plates reduce illumination efficiency as well as increase the cost and discoloration (aging) of the diffusion plate. This study attempts to develop an anti-glare LED without diffusion plates through optical design. Major methods adopted emphasize the use of advanced systems analysis program (ASAP) optical software to discuss the optical and glare-related features of the light-gauge steel frame fluorescent lamps (LGSFFL), traditional light gauge steel framing LED lamps (LGSFLL), and self-designed new light gauge steel framing LED lamps (nLGSFLL). Light ray tracing is simulated by using ASAP optical simulation software. The software simulates the lamp luminance, luminance uniformity, distribution curve, uniform glare index (UGR), and lighting scenario. The software also analyzes the retinal image under different lamps and UGR [12].
2. Establishment of experimental steps and computation model
The optical calculation procedure is illustrated in Fig. 1. The first step is to design the lamp and light source models. First, the researcher measures the size of the purchased lamp. He then builds the lamp model with Solidworks drawing software in sequence. The establishment of a lamp model is particularly important to correct optical calculation. The primary work for optical analysis is correct, and the information related to lamp dimension are established. If the design has errors at the onset, subsequent optical calculation will be incorrect. The design of a group of indoor lamps might generally include two elements: lamp contour design and light source design. The lamp contour design is categorized into shape, material, function, and other perspectives. Different shape designs or choices of different materials will have different outputs of luminance. Light source design includes the choice of light source and its position against objects. These parameters all affect the lighting efficiency of lamps. The sizes of the three lamps will be described in detail in Chapter IV.
The second step is to prepare the model for ASAP optical calculation. ASAP has a smart translation procedure that interprets the lamp model established in Solidworks CAD. This convergence method could build a detailed and actual specification in the calculation procedure and solve the difficulty encountered by users when they build a complicated model in ASAP. ASAP optical analysis has four steps: a) building a system model, b) producing a light source (light beam), c) light ray tracing, and d) evaluating the optical calculation result. To examine the indoor glare problem further, the third step must first define the position of the indoor space, human eye, and E-chart. The indoor space is assumed to be a small office. The small office includes a ceiling, wall, and floor. The space dimension is 3.980x3.680x 2.980 m. The lamp at the center of the ceiling is 0.6 m long and 0.6 m wide. The small office with lamps and working planes is established in ASAP. The working plane is 1.985 m below the lamp with dimensions of 1.2x1.2 m. The human eye is 1.75 m away and 30° from the lamp. Light ray tracing is performed by ASAP to observe the path of the light ray being reflected inside the lamp or reflected upon the wall and reaching the working plane. Meanwhile, the path of light ray is classified and compared with the UGR obtained through actual measurement of lamps for further study. The fourth step is to input the data result of ASAP into DILux illumination software and MATLAB program to calculate the lamp glare and uniformity. The calculations will be described in the subsequent chapters.
Optical simulation and tracing through ASAP are possible when the spatial system is established. The simulation analysis is divided to two parts: (1) analysis of working plane illumination and (2) analysis of distribution curve. These parts are explained as follows:
(1). Analysis of working plane illumination
The working plane height is “850 mm above the ground” [13]. The actual measurement shows that the small office ceiling is 2835 mm above the ground. Hence, the working plane is 1985 mm beneath the lamp in this simulation. The working plane is 1200x1200 mm. After the working plane is established, the average illumination (lux), illumination uniformity and maximum value (lux), and total luminous flux of the lamp upon the working plane could be analyzed.
(2). Analysis of distribution curve
The distribution curve of the simulation lamp in ASAP has the same direction as the actual measurement by the apparatus. The distribution is divided into the vertical and parallel directions relative to the lamp axis, as shown in Fig. 2. The red line represents the distribution of spatial light intensity measured parallel to the light axis. The green line represents the distribution of spatial light intensity measured vertical to the light axis.
The final step evaluates whether the light ray causes discomfort to human eyes (UGR is < 19) [14]. If the light causes discomfort to human eyes, we will return to the first step to redesign the lamp model. Otherwise, the subsequent simulated tracing is incorrect. After the lamp design is completed, we will use the retinal image to observe the actual glare feeling to the human eyes.
3. UGR
The International Commission on Illumination proposed the UGR in 1995 [11] because the CIE glare index (CGI) glare revision formula cannot be used conveniently and is also difficult to employ for actual calculation. The advancement of UGR could solve these problems. To calculate UGR, we use the following formula:
where Lb: background luminance, unit is cd/m2 ; L: the luminance of light source toward the direction of the eye, the unit is cd/m2 ; ω: the solid angle formed by the light source and the eye of the observer; and P: Guth’s position index (the position index originally developed by Guth weighs a glare source based on its position in a view) [15], similar to the British Glare Index.
The comparison table of GR and human eye feeling is also used [16]. During general lamp design, the UGR must be less than 19 [14]. Hence, glare is more severe when the UGR is greater, which is exactly opposite the visual comfort probability value [17]. The UGR calculation procedure is divided into three parts. The first part involves the lamp parameters, such as size, quantity, distribution curve, luminous flux, and position of the lamp in the space. The second part involves the indoor space including the length, height, and width of indoor space, maintenance coefficient, and reflectivity of wall, ceiling, and floor. The third part involves the relevant position of lamps at the detection point. Before UGR calculation, the background brightness, light source brightness directed toward the eyes, solid angle formed by the light source and eye level of the observer, as well as Guth’s position index [15] are obtained. The calculation procedure is as follows:
(a) Calculation of the light source brightness directed toward the eye level of the observer L
Figure 3 shows that D is the projected horizontal distance from the lamp to the eyes, H is the height from the lamp to eye level, Q is the distance from the lamp to the eye level of the observer, O is the observer, S is the light source, Iθ is the luminosity of the distribution curve flex, AS is the area of lamp, and θ is the angle of elevation between the lamp and the eye level of the observer. The light source brightness directed toward the eye level of the observer L in Eq. (2) is calculated based on the result of Eq. (3).
(b)Calculation of the solid angle between the light source and the eye level of the observer
Figure 3 shows that the solid angle between the light source in Eq. (4) and the eye level of the observer is calculated based on the results of Eqs. (5) and (3).
(c)Calculation of background brightness
Background brightness does not include the wall brightness from the light source. The intensity of the light source is calculated based on the luminosity of the lamp distribution curve on the wall. However, the background brightness should include the usage rate and maintenance coefficient of the environment luminance in the office space. Hence, the reflection rate and maintenance coefficient should be considered and calculated using the following equations:
where DFw refers to the distribution factors of walls, FL is the total luminous flux of the lamp, MF is the maintenance coefficient, Aw is the total area of the wall, EW(D) is the direct luminance of the lamp on the wall, and EW is luminance of the lamp on the wall.
(d)Calculation of Guth’s position index
Position index is a spatial factor defined by International Committee on Illumination [18] that represents the coefficient factor defined by the ratio of length vs. the height and the ratio of width vs. length from the observer to the lamp. Finally, the UFR value is calculated by using Eq. (1).
4. Three lamp specifications and result of optical measurement
This chapter describes in detail the establishment of Solidworks CAD lamp model and inputs into ASAP with illustrations of the indoor light gauge steel frame fluorescent lamp, LED lamp, and self-designed LED lamp. Figure 4 shows the model incorporating the three lamp specifications into ASAP. The red arrow represents axis X, the blue arrow represents axis Z, and the green arrow represents axis Y. The actual specification for each object of the lamp and the model of each component in Solidworks CAD are included in the ASAP model.
4.1 Light-gauge steel frame fluorescent lamp
First, we measure the size of the light-gauge steel frame fluorescent lamp. The lamp has a length of 600 mm, width of 600 mm, and height of 65 mm, whereas the aluminum plate is 0.5 mm thick. The material is polished steel plate with baked varnish, mirror-surface aluminum plate (with the electronic stabilizer installed at the back), anti-glare grid, and T5 tube. The rated voltage is 220 V, the rated input current is 1.4 A, the rated frequency is 60 Hz, and total consumption is 96 W. The specification of T5 tube is Philips TL5 HE 28W/865. The optical flux of each tube is approximately 2400 lm. T5 is the code for a 16 mm (5/8 in.) diameter fluorescent lamp glass tube with length of 550 mm. Compared with the traditional T9 lamp tube (9/8 in diameter) or T8 lamp tube (1 in diameter), T5 fluorescent lamp has a small diameter, but its luminous efficiency is higher than that of T9 and T8 lamp tubes [19].
In the simulation, we first generate a drawing of the polished steel plate with baked varnish, mirror surface aluminum plate, and anti-glare grids by using the Solidworks drawing software. After encoding, we input the drawing into the ASAP optical software. The T5 lamp tube model is directly encoded in ASAP. The physical entity of the fluorescent lamp is completed.
4.2 LED lamp
We measure the actual LED lamp. The lamp has a length of 600 mm, width of 600 mm, height of 50 mm, and thickness of 0.5mm. The color temperature is daylight white 5000 K, luminous flux is 2300 lm, power consumption is 36 W, and the input voltage ranges from 90 to 264 (AC). The lamp includes a back plate with baked varnish, LED, and others. The LED light source is Philips Lumileds LXHL-LW3C. This LED has a maximum optical flux of 80 lm under 1000 mA current [20]. In the simulation, we first generate a drawing of the lamp model using Solidworks. Then, we assemble the model into ASAP. The LED model is directly drawn using the ASAP compiler program. This LEP lamp has four rows of LED, and each row has 11 LEDs for a total of 44 LEDs.
4.3 Self-designed LED
The lamp is designed with an outer shell made of mirror-surface aluminum plate with high reflectivity. LED strips are installed in four sides of the lamp. The light ray directly hits the working plane through reflection. As a result, the luminance on the working plane conforms to the luminance of LED bought in the market. The lamp measures 600 mm (L), 600 mm (W), and 100 mm (H). Similarly, the mirror-surface aluminum plate models are drawn using Solidworks and then assembled into ASAP. The LED specification is Philips Lumileds WR-WC090-20S and is incorporated into the lamp model. The lamp has four rows of LED, and each row has 10 LEDs or a total of 40 LEDs. LED lamp is generally used for direct illumination. However, this lighting design can cause direct and reflective glare to the eyes. Hence, the design of this model reduces the average LED luminance on the working plane through primary and secondary reflections. These reflections reduce the effect of glare and fulfill the goal of eye comfort.
The self-designed LED has an adjustable light field as shown in Fig. 5-(a). The LED includes a lamp shell, four side plates extending from four sides of the top plate, which could provide a beam emitted from this lamp strip, and four lamp skew plates. The lamp skew plate has an adjustable angle with the horizontal line. The skew plate structure is illustrated in Fig. 5-(b). The lamp shell plate structure is shaped through metal stamping and includes the parabolic mirror of the top plate and two side boards. The top plate is made of reflective materials and reflects the light beam emitted by two lamp strips. Adjustable skew plate angles could change the distribution curves and luminance on the working plane. Figures 6 (a) to (f) show the distribution curve when the skew plate has an angle of 5°, 15°, 30°, 45°, 60°, or 75°. The red line represents the luminosity curve C0–C180, where the green line represents the luminosity curve C90–C270 (the same direction as shown on Fig. 2) [21]. Figure 7-(a) shows the simulated lighting scenario of the traditional LED lamp. Excessive distance between lamps causes uneven lighting area and zebra effect. Figure 7(b) shows the simulated lighting scenario of the lamp designed for this study. Adjusting the skew plate angle obtains a uniform lighting environment to correct the zebra effect caused by excessive distance between lamps.
Fig. 6 Distribution curve at different angles of the skew plate (a) 5°, (b) 15°, (c) 30°, (d) 45°, (e) 60°, (f) 75°.
5. Measurement result of distribution curve flex and illumination of working plane
5.1 Analysis of distribution curve
Table 1 sets the total light ray number of the lamp as 40 million to ensure the high-precision of optical simulation. One fluorescent lamp set has four lamp tubes, and each fluorescent lamp tube emits 10 million rays. As for the distribution curve of light-gauge steel frame fluorescent lamp, the divergence angle of the distribution curve at A direction is between 40° and 140° or a total of 100°. The divergence angle at B direction is between 30° and 150° or a total of 120°. The intensity of distribution curve at A direction slightly declined from 60° to 40° and from 120° to 140° because the light-gauge steel frame fluorescent lamp is installed with a mirror-surface aluminum plate to increase the divergence angle. The angle of mirror-surface aluminum plate has certain relationship with the lamp divergence angle. As the angle of the mirror-surface aluminum plate becomes more inclined, the divergence angle of the lamp becomes larger. The light distribution becomes uneven. The distribution curve of the light-gauge steel frame fluorescent lamp has a maximum brightness of 1061 cd, minimum brightness of 0 cd, average brightness of 676 cd, total luminous flux of 2707 lm, and GR of 16.3. The traditional LED lamp has 40 million light rays. One group of LED lamps has a total of 44 LEDs, and each LED has 91000 light rays. As for the distribution curve of traditional LED lamp, the divergence angle of distribution curve at A direction is between 40° and 140° or a total of 100°. The divergence angle at B direction is from 40° to 140° or a total of 100°. The traditional LED lamp provides direct illumination. Hence, the lamp has uniform rays, and most rays are located at the center. The distribution curve of the traditional LED has a maximum brightness of 1206 cd, minimum brightness of 0 cd, average brightness of 699 cd, total luminous flux of 3050 lm, and GR of 19. The self-designed LED lamp has 40 million rays. One group of LED lamps has 40 LEDs, and each LED has one million light rays. As for the distribution curve of the self-designed LED lamp, the divergence angle of distribution curve at A direction is between 20° and 160° or a total of 140°. The divergence angle at B direction is from 20° to 160° or a total of 140°. Although the self-designed lamp provides indirect lighting, the center of distribution curve is quite low. Therefore, the self-designed lamp provides direct lighting, and its ray is not directly emitted but reflected by a large angle of emergence, which will cause a low light field at the center of the distribution curve. The distribution curve of the self-designed LED has a maximum brightness of 696 cd, minimum brightness of 101cd, average brightness of 477 cd, total luminous flux of 3055 lm, and GR of 17.5.
Table 1. Distribution curve, average brightness, total luminous flux, actual measurement of glare value, and calculated glare
5.2 Analysis on luminance of the working plane
Table 2 shows the luminance of the working plane of the light-gauge steel frame fluorescent lamp. The luminance has regular distribution. The maximum luminance is 277 lx, the minimum luminance is 155 lx, the average luminance is 222 lx, and the total luminous flux is 320 lm. The lamp has a general uniformity of 0.7. The average difference of uniformity is 75% [22]. General uniformity is also known as the distribution uniformity. General uniformity refers to the ratio between the minimum and average luminance of the space with a value between 0 and 1 [23,24], and 1 is the ideal value.
Table 2. Illumination distribution, illumination, total luminous flux, uniformity and average uniformity of three types of lamps on 1.2 m x 1.2 m working plane
Where E0: minimum luminance in the space, Ea: average luminance in the space.
The average difference can be computed by:
where the average difference indicates the degree of uniformity on the table plane. A lower percentage indicates better uniformity.The luminance of the working plane of the traditional LED lamp has a near-elliptical distribution. The LED light distribution field of an LED lamp with direct lighting is round in shape, and LED lamps have a regular arrangement. Hence, the accumulated light field has near-elliptical distribution. The maximum luminance is 310 lx, the minimum luminance is 215 lx, the average luminance is 271 lx, total luminous flux is 390 lm, and general uniformity is 0.79. If the average uniformity is introduced, its value is 65%. As to the luminance of the working plane of self-designed LED lamp, the luminance has a cross distribution because the light source structure of this lamp comprises four LED strips. The light source is reflected to the working plane through indirect luminance, which causes a non-uniform light ray. As uniformity is defined above 0.6, the luminance is distributed with in a tolerable scope. The maximum luminance is 365 lx, the minimum luminance is 165 lx, the average luminance is 256 lx, total luminous flux is 369 lm, and general uniformity is 0.64. The average uniformity is 85%. We calculate UGR based on the above description. UGR is calculated using the luminous flux and distribution curves of different lamps. Figure 8 shows the luminous flux of different lamps and relationship curve of UGR. The longitudinal axis is the UGR and the horizontal axis is the magnitude of luminous flux. The curves in Fig. 8 represent the light-gauge steel frame fluorescent lamp, traditional LED lamp, and self-designed LED lamp. Figure 8 also shows the relationship between the increase in luminous flux and UGR change. In an optical simulation system, we continue to increase the luminous flux of the three different lamps. We know from Eq. (8) that when the optical flux increases, UGR will also increase, and the square of the luminous flux is proportional to UGR. Hence, we obtain the result using Eq. (8). We conclude that UGR will rise linearly at a low luminous flux (below 5 x 103 lm). When luminous flux is greater than 1.0x104, UGR approaches saturation. When the luminous flux is fixed to a certain value, the traditional LED lamp has the largest UGR, whereas the fluorescent lamp has the smallest UGR. The self-designed LED lamp UGR has a values between the two. To confirm the luminous flux and UGR during actual lighting, we enlarge the portion in which the luminous flux is smaller than 5 x 103 lm. The three green dots are the actual luminous flux of the three types of lamps. The UGR of traditional LED is 19 (approaching the limit of qualification value). The self-designed LED lamp corrects the problem of the traditional LED with too large UGR, which is 17.5. The self-designed LED has a low UGR and is free of diffusion plate components. Although the light-gauge steel frame fluorescent lamp has low UGR, its luminous flux is also quite low, implying low illumination efficiency.
Fig. 8 Relationship chart on the luminous flux and UGR of three types of lamps. The illustration is the enlarged details.
6. Retinal image
To examine further the observation result of indoor glare on vision, we model the human eyes (ASAP biomedical module BI0 Toolkit) in space (as described in Chapter II), place the visual acuity test right below the lamp, and view straight into the lamp (as shown in Fig. 9). The visual acuity test is also placed horizontally below the lamp and view downward at an angle of 45。 (as shown in Fig. 10). Thus, we use the retinal image to observe the actual glare feeling of the human eyes. As for the up-right E-chart image of different lamps under varying luminous flux, the light-gauge steel frame fluorescent lamp has a luminous flux of 2707 lm. Based on the glare expression, the traditional LED lamp has a luminous flux of 3050 lm when UGR is 19.0. The self-designed LED lamp has a luminous flux of 3055 lm when UGR is 17.5. The retinal image has a UGR of 17.5, as shown in Fig. 11. From the light-gauge steel frame fluorescent lamp, the image has less stray light but the traditional fluorescent lamp has more stray light. The stray light of self-designed LED lamp has a value between the two; that is, UGR expression is based on the stray light.
Fig. 10 Lighting space of the lamp when human eyes observe E-chart at an angle of 45° and E-chart lies horizontally on the working plane
Fig. 11 Retinal image when human eyes look straight into the E-chart, (a) LGSFFL. (b) LGSFLL. (c) nLGSFLL.
However, applying the stray ray method for fuzzy representation is impossible in a plane-type retinal image because the light ray reflected by E in the Vision Acuity Table to human eyes in the plane-type retinal image is close to the increased quantity of light ray reflected to human eyes from other areas, which is different from the up-right glare image. The up-right image has strong light near the lamp, and the light is sufficiently strong to affect the fonts of the E-chart. Hence, in plane-type retinal images, only the distribution of stray light is seen in the E-chart, but cannot make the fuzzy feeling, as shown in Fig. 12.
Fig. 12 Retinal image when human eyes observe the E-chart at an angle of 45°, (a) LGSFFL. (b) LGSFLL. (c) nLGSFLL.
7. Conclusion
The goal of this study is twofold. The first goal is to develop an anti-glare evaluation method with a retinal imaging system. The second goal is to produce a group of anti-glare LED lamps through optical design, as well as to simulate the traditional light-gauge steel frame fluorescent lamp and LED lamp available in the market, analyze their luminance, uniformity, distribution curve, retinal image, and UGR, design a group of LED lamps, and cause the light emitted from the LED to hit the working plane through primary or secondary reflection with the skew plate and mirror-surface aluminum board. We analyze the luminance, uniformity, distribution curve, retinal image, and UGR, and then observe the retinal image of each lamp. When the light-gauge steel frame fluorescent lamp has a luminous flux of 2707 lm, the glare is 16.3. When the luminous flux of traditional LED is 3050 lm, the glare value is 19.0. When the self-designed LED lamp has a luminous flux of 3055 lm, the UGR is 17.5. We found that the stray light in a retinal image has a definite relationship with UGR. As for the set of anti-glare lamps, the retinal image has less stray light than traditional LEDs. As for the lamp illumination efficiency, the two values of optical flux of 3055 lm and glare 17.5 are better than those of the light-gauge steel frame fluorescent with a luminous flux of 2707 lm and UGR of 16.3 and is significantly lower than those of the LED lamps available in the market with a luminous flux of 3050 lm and UGR of 19.0. The main reason for this finding is that the light ray is reflected out of the lamp after repeated reflection, and the lamp unavoidably absorbs some energy.
Acknowledgment
This research was supported by the National Science Council, The Republic of China, under the Grants NSC 100-2221-E-194-043, 101-2221-E-194-049, 102-2221-E-194-045, and 102-2622-E-194-004-CC3.
References and links
1. W. D. van Driel and X. J. Fan, Solid State Lighting Reliability: Components to Systems (Springer, 2012).
2. K. Wang, F. Chen, Z. Liu, X. Luo, and S. Liu, “Design of compact freeform lens for application specific Light-Emitting Diode packaging,” Opt. Express 18(2), 413–425 (2010). [CrossRef] [PubMed]
3. Y. Luo, Z. Feng, Y. Han, and H. Li, “Design of compact and smooth free-form optical system with uniform illuminance for LED source,” Opt. Express 18(9), 9055–9063 (2010). [CrossRef] [PubMed]
4. F. Chen, K. Wang, Z. Qin, D. Wu, X. Luo, and S. Liu, “Design method of high-efficient LED headlamp lens,” Opt. Express 18(20), 20926–20938 (2010). [CrossRef] [PubMed]
5. S. Wang, K. Wang, F. Chen, and S. Liu, “Design of primary optics for LED chip array in road lighting application,” Opt. Express 19(S4Suppl 4), A716–A724 (2011). [CrossRef] [PubMed]
6. J.-J. Chen, T.-Y. Wang, K.-L. Huang, T.-S. Liu, M.-D. Tsai, and C.-T. Lin, “Freeform lens design for LED collimating illumination,” Opt. Express 20(10), 10984–10995 (2012). [CrossRef] [PubMed]
7. P. Ngai and P. R. Boyce, “The effect of overhead glare on visual discomfort,” J. Illuminating Eng. Soc. 29(2), 29–38 (2000). [CrossRef]
8. P. R. Boyce, C. M. Hunter, and C. B. Carter, “Overhead glare and visual discomfort,” J. Illuminating Eng. Soc. 73–87 (2003).
9. C. M. Lee, H. Kim, and D. S. Choi, “A study on the estimation of discomfort glare for LED luminaires,” Conf. Proc. 26th Session of the CIE, Beijing, D3–33–D3–36 (2007).
10. H. Takahashi, Y. Kobayashi, S. Onda, and T. Irikura, “Position Index for the matrix light source,” Light & Vis. Env. 31(3), 128–133 (2007). [CrossRef]
11. CIE 117–1995 Discomfort Glare in Interior Lighting. CIE. (1995).
12. H.-Y. Sun, H.-C. Wang, and S.-F. Yang, “Vision correction via multi-layer pattern corneal surgery,” Opt. Commun. 300, 293–298 (2013). [CrossRef]
13. M. G. Helander and B. A. Rupp, “An overview of standards and guidelines for visual display terminals,” Appl. Ergon. 15(3), 185–195 (1984). [CrossRef] [PubMed]
14. The Illuminating Engineering Institute of Japan, Lighting Handbook (2006).
15. M. Luckiesh and S. K. Guth, “Brightnesses in visual field at borderline between comfort and discomfort,” Illum. Eng. 44(11), 650–670 (1949). [PubMed]
16. H. Heuer, G. Hollendiek, H. Kröger, and T. Romer, “Dieruhelage der augen and ihreinflu β auf beobachtungsabatand and visuelle ermudung bei bildschirmarbeit (Rest position of the eyes and its effect on viewing distance and visual fatigue in computer display work),” Zeitschrift Fur Experimentelle und Angewandte Psyologie No.36, 538–566 (1989) (Journal of Experimental and Applied Psychology).
17. J. A. Clarke, M. Janak, and P. Ruyssevelt, “Assessing the overall performance of advanced glazing systems,” Sol. Energy 63(4), 231–241 (1998). [CrossRef]
18. J. Lin, “A light diet for a giant appetite: An assessment of China's proposed fluorescent lamp standard,” Conference: Right Light 5, Nice (FR) (2002).
19. National Research Council, Video Displays, Work, and Vision (National Academy Press, 1983).
20. http://www.lenalighting.pl/en/knowledge/light-distribution/
21. Dialux 4.9, Light building software, DIAL GmbH, www.dial.de/
22. C.-H. Tsuei, J.-W. Pen, and W.-S. Sun, “Simulating the illuminance and the efficiency of the LED and fluorescent lights used in indoor lighting design,” Opt. Express 16(23), 18692–18701 (2008). [CrossRef] [PubMed]
23. M. Y. Hsieh and C. Y. Chen, “The Effects of interior illuminance Distribution on Spatial impression,” Journal of Architecture of Taiwan 75, 81–98 (2011).
24. D. Ramane and A. Shaligram, “Optimization of multi-element LED source for uniform illumination of plane surface,” Opt. Express 19(S4Suppl 4), A639–A648 (2011). [CrossRef] [PubMed]
