using micro structures to light into thin light guides462939/fulltext01.pdf1.1 project background...

Using micro‐structures to couple light into thin light‐guides

Yun Chen

Master of Science Thesis

TRITA-ICT-EX-2011:112

2

This work was carried out at Visual Experience & Lighting at Philips Research Europe, Eindhoven, The Netherlands. Approved by Examiner: Assoc. Prof. Sergei Popo Supervisor: Dr. M.P.C. Krijn;

Dr. G.E. Onac Assoc. Prof. Sergei Popo

3

Contents Abstract ........................................................................................................................................... 5 Acknowledgement .......................................................................................................................... 7 Chapter 1 Introduction .................................................................................................................... 9

1.1 Project background ........................................................................................................... 9 1.2 Newly proposed configuration ......................................................................................... 9 1.3 In this report ................................................................................................................... 11 1.4 Outline ............................................................................................................................ 11

Chapter 2 Theoretical background ................................................................................................ 12 2.1 Radiometry and photometry ........................................................................................... 12 2.2 Total internal reflection (TIR) ........................................................................................ 13 2.3 Light emitting diodes (LEDs) ........................................................................................ 14

2.3.1 Physics of LEDs ...................................................................................................... 14 2.3.2 The I-V characteristics of p-n junction ................................................................... 14 2.3.3 L-I characteristics .................................................................................................... 15 2.3.4 Optical property ...................................................................................................... 16

2.4 Lambertian scattering ..................................................................................................... 16 2.5 ETENDUE ..................................................................................................................... 17 2.6 Polymethyl methacrylate (PMMA)and its properties .................................................... 17 2.7 Retroreflection film ........................................................................................................ 18

Chapter 3 Simulation methods ...................................................................................................... 19 3.1 Light guide introduction ................................................................................................. 19 3.2 Descriptions of the applied micro-structures ................................................................. 19 3.3 Solutions ......................................................................................................................... 21

Chapter 4 Initial results and model selection ................................................................................ 33 4.1 Components in Light-Tools ........................................................................................... 33

4.1.1 Modeling of the LEDs ............................................................................................ 33 4.1.2 Modeling of the light-guide .................................................................................... 34 4.1.3 Modeling of micro-structure ................................................................................... 34 4.1.4 Modeling of detectors ............................................................................................. 34 4.1.5 Modeling of Lambertian reflectors, mirrors ........................................................... 35

4.2 Simulation result and discussion:Sphere model ............................................................. 35 4.2.1 Sphere model 1 ....................................................................................................... 37 4.2.2 Sphere model 2 ....................................................................................................... 39

4

4.2.3 Sphere model 3 ....................................................................................................... 41 4.2.4 Conclusion .............................................................................................................. 46

4.3 Simulation result and discussion: Light guide model .................................................... 47 4.3.1 Method A1 (Prism) &A2 (Pyramid) ....................................................................... 47 4.3.2 Method B1 (Prism) & B2 (Pyramid) ...................................................................... 48 4.3.3 Method C1 (Prism) & C2 (Pyramid) ...................................................................... 49 4.3.4 Method G1 (Prism) & G2 (Pyramid) ...................................................................... 49 4.3.5 Method H1 (Prism) & H2 (Pyramid) ...................................................................... 50 4.3.6 Method I1 (Prism) & I2 (Pyramid) ......................................................................... 51 4.3.7 Method K1 (Prism) &K2 (Pyramid) ....................................................................... 51 4.3.8 Discussion ............................................................................................................... 52

4.4 Models selection ............................................................................................................. 55 Chapter 5 Measurements in lab ..................................................................................................... 56

5.1 Test-module introduction ............................................................................................... 56 5.2 Integrating sphere introduction ...................................................................................... 58 5.3 Test results ...................................................................................................................... 59

5.3.1 Linear Response of LEDs ....................................................................................... 59 5.3.2 Test-prototype experimental result ......................................................................... 61

5.4 Conclusion ...................................................................................................................... 63 Chapter 6 Results and conclusion ................................................................................................. 64

6.1 Conclusions .................................................................................................................... 64 6.2 Future Work ................................................................................................................... 65

Appendix 1 .................................................................................................................................... 66 1.1 Interface Elements .......................................................................................................... 66 1.2 Immersion manager ........................................................................................................ 67 1.3 Optimization ................................................................................................................... 68 1.4 Finding help.................................................................................................................... 69

Appendix 2 .................................................................................................................................... 70 References ..................................................................................................................................... 71

5

Abstract The task of this project is to investigate the possibilities of using micro-structuring on the surface of thin light-guides to efficiently couple light from top-emitting LEDs into such light-guides. Areas of application are backlighting for LCD TV and flexible light emitting layers for lighting purposes. The micro-structures considered are prisms and pyramids. The micro-structures can be on the same side of light-guide as the LEDs or on the opposite side. When located on the opposite side, the micro-structures are coated with a specular reflecting layer or with a diffuse reflecting layer. The LEDs can either be in optical contact with the light-guide or not in optical contact. Optical ray-tracing software package Light-Tools is used for all ray-tracing simulations of these geometries. In simulations, a two-step approach is taken: Firstly, we build in Light-Tools a simple model of an LED in proximity to a micro-structured light-guide in order to ascertain which geometry is most likely to show a high in-coupling efficiency. In the next step, we made a more elaborate and more accurate model of the most promising geometries. The results of these simulations show that, for micro-structures that are located on the light-guide on the side opposite to the LEDs, a diffuse reflecting layer (i.e. a Lambertian scatterer) is more useful than a specular reflecting layer (i.e. a mirror). Also, in general, the prisms structures perform better than the pyramid structures. The highest efficiency reached in the simulations is 60% for Model I1 in which the light source is not in optical contact with the light guide and the mirror is on the opposite side of the light guide. Compared to the result reported in a previous paper, the in-coupling efficiency improvement is 8%. We also looked into using Retro-reflection Films for improving the in-coupling efficiency. Several prototypes with a Retro-reflection Film were made and tested in the laboratory. Measurements were performed for three thicknesses of the light guide: 0.25 mm, 0.5 mm and 1 mm. The best in-coupling efficiencies measured are 37% for a 0.25 mm thick light guide, 53.1% for a 0.5 mm thick light guide and 57.1% for a 1 mm thick light guide. Compared to the samples without Retro-reflection Film, the best Retro-reflection Film results in a 7% increase for the 0.25 mm thick light guide, 6.2% increase for the 0.5 mm case and 2.8% increase for the 1 mm case, respectively.

7

Acknowledgement My master project has started in Visual Experience, Philips Research, Netherlands and the Royal Institute of Technology (KTH), Sweden since 1st, August, 2010. During this period, I learnt a lot from my supervisors and colleagues and was supported and encouraged by them. Firstly, I would like to thank one of my supervisors in Philips, Dr. Marcel P.C.M. Krijn. Throughout my project, he helped me not only on theoretical knowledge, but also on communication skills with other colleagues. There is no doubt that what he has done for me is more than a supervisor. His attitude toward the research will influence me in my future research. I also would like to give my thanks to my other supervisor in Philips, Dr. Eugen Onac, for his kindly modifying my models in Light-Tools software. During the time I made my simulation models, he offered me lots of precious suggestions, which really improved my models a lot. I am quite thankful to Dr. Sergei Popov, the supervisor in KTH. His timely concern of my project and useful suggestions are very important to me. Without his help, my project will not be accomplished successfully. Many thanks to Dr. Hugo Cornelissen, who introduced how to use the integrating sphere to me in the laboratory, and Ir. Daniel Santos Canelles and Ir. Dominika Switlik, my dear colleagues, who shared their experince in Light-Tools with me. Wihout their assistance, I can not finish my project so smoothly. Also I received helpful support from Cong Mu, Etienne Geurts and other friends in our office. Then my thanks go to Prof. Urban Westergren, my programme director of Photonics and Microwave Engineering, KTH. Your suggestions about how to choose courses and master thesis project left on me a deep impression. Finally, I want to give my most precious appreciation to my beloved and respectable parents. For the last more than 20 years, your thoughtful support accompanied me wherever I was in Shanghai, Sweden or Nertherlands. During the last 2 years in Europe, I missed you all every minute. I also owe my thanks to my boyfriend. 最后，要把我最最珍贵的感谢送给我最亲爱的父母。在过去的 20 多年里，无论我在上海瑞典还是荷兰，感谢你们都无微不至地支持我，照顾我。在过去两年里的 700 多个日日夜夜，我一直在遥远的欧洲思念着你们。谨把这篇论文献给我至爱的父母。 Thank you all, Yun Chen

9

Chapter 1 Introduction

1.1 Project background According to Philips’ new concept of display, which is “The TV's aim is to tap into people's desires to hang their TV on walls just as they would do with a painting”, the Philips research group is focusing on creating thinner backlight panel which should be reduced from the normal size (25 mm) to ultra-thin size. With the great improvement in the brightness and lumen efficiency of light-emitting diodes (LEDs), they have become the most popular light source for liquid crystal display (LCD) backlights now. The advantages for using LEDs as light source are:

• Small size • High efficiency • Very fast response, long lifetime, • Low voltage driver • Local control of luminance resulting in better contrast and lower energy use • Saturated colors

Due to the small die size ( 11× mm2) and high efficiency of LEDs, the ultra-thin TV can be achieved by coupling LEDs light into thin backlight-guide by means of total internal reflection (TIR). Recently, in the IFA exhibition, Philips has demonstrated the world’s thinnest LCD TV in which a very thin backlight-guide was used. However, as discussed in the previous study [1], with the reduction in the thickness of the light guide, the light in-coupling efficiency decreases. Low light in-coupling efficiency results in two problems:

• Large light loss. More light escapes without being waveguided, giving rise to hot spots

• Inefficient and inhomogeneous illumination of the TV screen. While brightness and uniformity of TV screens need to be kept the same as normal size TV, these problems limit the thickness reducing of the backlight panel. Now the challenge we face is to get high light-coupling efficiency while the backlight panel becomes thinner and thinner.

1.2 Newly proposed configuration Several optical solutions are proposed to improve the light-coupling efficiency, based on wave-optics and geometrical optics. Firstly, the wave-optics way has been studied. In this case, a dielectric multilayer is used as an angular filter to keep more light inside the light-guide by selectively transmitting the light fulfilling the TIR condition. (See Fig. 1.1). By using designed thin filter, 45% of LEDs emitting light is kept inside a 0.3 mm thin light guide. However, there are still some problems left: non-uniform irradiance pattern for planar light source, more light leakage at small angles, wavelength dependence and so on [2].

The obindepen

This gesoftwarpresentemm thinmeasurethan 0.5do to im

Fig. 1.

bvious mendent. Two 1) Reflecto

the ligh

Fig. 1

eometrical ore, which wed in an ean light guided in-coupl5 mm mainmprove it. 2) Micro-s

guide (

.1 Optical confi

erit of geodifferent mors are used

ht-guide (Se

1.2 Optical con

optics way, ill be expla

arly study [de, 70% foring efficien

nly due to al

structures ar(See Fig. 1.3

iguration for us

metrical opmethods in ge

d on the tope Fig. 1.2).

figuration for u

using reflecained in App1]. The bes

r 0.5 mm thncy is less thlignment is

re used to m3).

10

sing angular filt

ptics, compeometrical o

p of the ligh

using a reflecto

ctors, is despendix I. Rest theoretica

hin light guihan the resusues [1]. Th

make the lig

ter to in‐couple

pared to woptics way

ht-guide to r

or to in‐couple

signed by uesults of theal in-coupliide, and 74%ult from simherefore the

ght bent bef

e light into ligh

wave opticare studied reflect the e

light into light‐

sing Light-Te simulationing efficien% for 1 mm

mulation where will be

fore entering

t‐guide [2].

cs, is the now:

escaped ligh

‐guide [1].

Tools, the lns and expencies are 60m thin lighthen the thick

lots of furth

g or escapin

wavelength

ht back in to

light tracingeriments are0% for 0.25t guide. Thekness is lessher work to

ng the light-

h

o

g e 5 e s o

-

11

Fig. 1.3 Optical configuration for using micro‐structures to in‐couple light into light‐guide.

1.3 In this report The aim of this report is to find if the second method of proposed geometrical optics way works for achieving high in-coupling efficiency with thin (0.2 mm) light-guide. We will model the second method of proposed geometrical optics way in Light-Tools, and measure the coupling efficiency of some similar modules in the laboratory. Two micro-structures will be used in this case, prisms and pyramids. The models can be arranged in two groups: light source in optical contact with light-guide and not in optical contact. Further in each group, there will be two cases: micro-structure on the same side or opposite side as the light source. These will be explicitly discussed in Chapter 3.

1.4 Outline In this master thesis, 6 chapters are included: Chapter 1: introduces the information of my master thesis; Chapter 2: reviews theoretical background, e.g., total internal reflection (TIR), and the property of PMMA material; Chapter 3: describes the methods used for simulation; Chapter 4: discusses the initial results, followed by model selection based on these results; Chapter 5: describes the measurements in lab and summarizes the results Chapter 6: analyzes both simulation and measurement results leading to a conclusion.

Light source:

waveguide

Light in‐couple

Micro‐structures

Mirror or Lambertian scatterer

12

Chapter 2 Theoretical background In this chapter, several definitions will be introduced that are useful for understanding the targets and results of this project.

2.1 Radiometry and photometry Radiometry [3], characterizing the physical parameters of radiation, describes radiant quantities. It applies to the whole electromagnetic spectrum, while photometry deals with only part of the electromagnetic spectrum that can be seen by a standard observer, for example the human eye. The eye’s sensitivity of light is not a constant in the whole light spectrum, and changes with light wavelength. The vision in normal lighting conditions during the day is defined as photopic vision. The vision under dim lighting conditions is called scotopic vision. The vision at intermediate conditions is called mesopic vision. When the source emits light in the space with a range of different wavelengths, the power distribution is a function of wavelength. The strength of the related total visual sensation is:

∫Φ=Φ λλκ λυ dVm )( (2.1)

Where mκ =683 lm/W is a constant (the number of lumens in one Watt of energy at 555 nm),

υΦ is the total luminous flux measured in lumens (lm), λΦ is radiant flux in watts per unit wavelength, and )(λV is normalized luminosity function (also called eye sensitivity curve). Some quantities in radiometry are defined as the following:

Radiant fluxΦ : Radiant flux in W is the quantity of energy flowing through a surface per unit time.

tQ

dd

=Φ (2.2)

Where Qis the energy and t is the time slot. Radiant intensity I: Radiant intensity in W/sr is defined as the radiant flux per unit solid angle Ω:

I ΩΦ

=dd (2.3)

From formula (2.3), it is obvious that radiant intensity is independent of distance between source and receiver. Irradiance E: Irradiance (W/ 2m ) is radiant flux received per unit area A:

dAdE Φ

= (2.4)

Where dA is the infinitesimal area that receives radiation. Radiance L: Radiance in ( )2msrW ⋅ is the radiation flux per unit projected area and per unit solid angle, given by:

13

dSd

dLΩΦ

= (2.5)

Where θcosdAdS = , and θ is the angle between normal n to area dA and direction of the solid angle Ωd . The formula (2.1) is the relationship between radiometry and photometry, from which the corresponding parameters (see Table 2.1) in photometry can be easily got.

Two sources will be discussed: • Point source: The radiant intensity is constant,

( )π

θ4Φ

== OII (2.6)

• Lambertian source: Luminance of this source is a constant over its surface and the

maximum radiant intensity is at o0 :

( ) ( ) ( )θπ

θθ coscos ⎟⎠⎞

⎜⎝⎛ Φ== OII (2.7)

2.2 Total internal reflection (TIR) When a ray of light strikes the interface of two media with different refractive index 1n and 2n(see Fig. 2.1), it will change its direction according to Snell’s law, partially be reflected back and partially be refracted.

When 1n > 2n , it might occur that ( ) 1sin 12

1 ≥θnn , which means ( ) 1sin 2 ≥θ . In these cases, total

internal reflection [4] happens that all the rays are reflected back to the incident medium (n1)

without energy loss. The incident angle in one particular case, in which ( ) 1sin 12

1 =θnn

,is called

critical angle, above which all the rays are kept in the incident medium.

Table 2.1 Corresponding parameters in Radiometry and Photometry.

14

Fig. 2.1 Rays travel from media 1 to media 2: 1θ is the angle formed by the ray of light in the first medium, and 2θ is the

angle formed by the refracted ray of light in the second medium.

When a light guide is made of a medium with high refractive index ( airnn > ), the rays of light, striking the boundary under TIR condition, will be trapped inside the light guide and propagate along the light guide. This is called trapped light phenomenon.

2.3 Light emitting diodes (LEDs) In this part, only the aspects of LEDs relevant to this project will be explained.

2.3.1 Physics of LEDs A light emitting diode (LED) is composed of a semiconductor p-n junction which is created by doping impurities into semiconductor materials. When a voltage is applied, the current flows from p-type (hole) material to n-type (electron) material and electrons and holes are recombined one by one. During this recombination, spontaneous light is emitted simultaneously because of electron falling into a lower energy level (hole) [5]. The spectrum of the light is dependent on the energy band gap of the material.

2.3.2 The I‐V characteristics of p‐n junction The Shockley equation should be introduced at first:

)1( −= TnVv

s eII (2.8)

Where I the current of LEDs, sI is the saturation current, V is the voltage drop on the LEDs, n is the quality factor, and TV is the thermal voltage. When a voltage is applied to a p-n junction, two parts of the response process will be included (see Fig. 2.2). In the first part, when the applied voltage is below the threshold voltage, the

n1 n2

θ1

θ2

θc

15

response of the p-n junction is fairly slow. However, once the applied voltage exceeds the threshold voltage the current will increase rapidly. This threshold voltage depends on the type of semiconductor material itself and the temperature [6].

Fig. 2.2 Current‐voltage characteristics of a p‐n junction [7].

2.3.3 L‐I characteristics The other important curve describing the behavior of LEDs is the L-I curve which shows the relation between luminous flux and current. In the Fig. 2.3 this relation is graphically shown and it can be seen that the L-I relation is almost a straight line. The light output power is a linear function of current. The light output power versus current starts to deviate from a straight at high currents (also called ‘droop’), mainly due to thermal effects.

Fig. 2.3 LEDs L‐I curve.

16

2.3.4 Optical property In ideal LEDs, all photons emitted by the semiconductor material should escape from the LEDs die and into free space. However, not all the photon will escape successfully in reality. Some of them are reabsorbed in the LEDs substrate which is absorbing at the emitting wavelength. Some of them are absorbed by a metallic contact surface. Others will be trapped inside the LEDs by TIR at the semiconductor-air interface. As a result, the external efficiency will be reduced, compared with the internal quantum efficiency [6]. For TIR cases, Lambertian emission pattern should be introduced. Think about a source that emits in all direction inside a semiconductor material with a high index of refraction. The rays of light are refracted at the material-air surface with an angle θ according to Snell’s law. Then the light intensity in air does resemble a Lambertian emission pattern, given by:

θcos2

2

s

airS n

nII = (2.9)

Where sI is the light intensity in the semiconductor, and sn is the refractive index of semiconductor. From formula (2.9), we can see that the light intensity of Lambertian emission in air varies with angleθ . The relation between them is illustrated in Fig. 2.4. At angles larger than o70 , the light intensity is very low and only 1/10 of its maximum value which is at o0=θ . When the angle is

o80 , nearly no light is emitted into air.

Fig. 2.4 Lambertian emission distribution integrated in (θ, π) or (-π, θ).

2.4 Lambertian scattering If a surface is a Lambertian scatterer, the incident light onto it will be scattered in all the directions, which is called diffuse reflection (red lines in Fig. 2.5). The intensity of reflected light depends on the angle from the normal to the reflected ray rather than the angle from the normal to observer, which means the intensity of the reflected rays are a constant to an observer. When

00,10,20,30,40,50,60,70,80,91

0 20 40 60 80Emission

intensity

. Normalized

refracted angle θ (degree)

Lambertian emission. Intensity within the angular intervals (θ, π) or (‐π, ‐θ)

17

light hits the Lambertian scatterer, it will be scattered with the same property of Lambertian emitter which is defined by: θcosOII = (2.10)

Fig. 2.5 When light hits the smooth surface, the reflected ray is traced by the blue line showed in this figure, the specular reflection. When the light hits the Lambertian scatterer, the reflected rays are red ones. The intensity of reflected ray

depends on the cosθ , θ is the angle from the normal to the reflected rays [8].

2.5 ETENDUE Etendue [3] is a very important concept in designing light in-coupling methods, which calculates how much radiation will flow through an area dA into a direction at an angle θ normal to the area and within a solid angle Ωd . The etendue is quantified as: Ω= ddAndU θcos2

(2.11) In an ideal optical system, etendue is always conserved, which is used in the transfer of radiation from a light source (LED) to a receiver in nonimaging optics. However, in a real optical system, the etendue is not conserved anymore and may be lost or increased.

2.6 Polymethyl methacrylate (PMMA)and its properties Polymethyl methacrylate (PMMA) is a crystal clear amorphous plastic, which is one of the polymers, acrylates [9]. The density of PMMA is 1.17–1.20 g/cm3. PMMA is widely used as a material that can efficiently guide light (for example, in light-guides as used in most backlights for LCD TV). Its main properties are summarized below:

• Weak absorption in visible spectrum, which is down to 0.1% per inch [10] [11].

Transmittance is almost 92% of visible light when the thickness is 3 mm, and reflection is about 4% at each polymer-air surface with its refractive index being 1.4914 at 587.6 nm.

• More transparent. • Hard and rigid, but light and easily to shape. • Better resistance. • Melting point is 125 . • The refractive index of PMMA slightly changes with the wavelength of light.(see

Fig. 2.6)

Fig. 2.6 Re

2.7 RBasicaland retrreflectiodirectiominimudirectio

Fig. 2.7

efractive index

Retrorefly, there arro-reflectionon, we use

on as the incum but is inon normal to

7 The types of

1,4651,47

1,4751,48

1,4851,49

1,4951,5

1,5051,51

1,5151,52

Inde

x of

refr

actio

n

x of PMMA in d

flection fe three diffn [12] (see a retro-refl

cident light n practice no the retro-r

Reflection: red

different light wthe UV light

film ferent main Fig. 2.7). I

flective thinis coming f

not always areflective fil

d lines represen

18

wavelength. Tht at wavelength

types of reIn this proje

n film. Suchfrom. Scatteavoidable, elm [12].

nts the incidentreflected ray

Wavele

PMM

e range of wavhs below 300n

eflection: diect we consh films do rering of lighespecially a

t rays and the bys.

ength [nm]

MA

velength is 350nm.

iffuse reflecsider all thrreflect lightht into otherat large ang

blue lines give t

nm‐1000nm.PM

ction, mirroree of themt back alonr directions

gles with res

the informatio

MMA will filter

or reflectionm. For retro-ng the same is kept at aspect to the

n about the

r

n -e a e

19

Chapter 3 Simulation methods

3.1 Light guide introduction In the nature world, waves spread in all directions, hence the energy propagates in all the directions as well. Waveguide, the device that guides waves in one or two dimensions, keeps waves inside it by means of total internal reflection (see Chapter 2). According to their different geometrical shape, the energy can be confined in either one dimension such as energy in fiber waveguide or two dimensions such as energy in slab waveguide. Additionally, waves of different frequencies are guided by different kinds of waveguides. Light guide, one of the waveguides that guides light wave, maintains light wave propagating inside due to total internal reflection. In this project, the planar waveguide is used, which is made by polymethyl methacrylate (PMMA). When light hits the surface 2 in Fig. 3.1 with the angle larger than critical angle, which means the TIR condition (see Chapter 2) is fulfilled, the light will be restricted inside the light guide. (Fig. 3.1)

Fig. 3.1. Trajectory of light travelling from PMMA to air. When the incident angle Ө in the surface 2 is larger than cθ ,

which is the critical angle, PMMA

airc n

narcsin=θ the light will be confined inside the PMMA. When the incident angle

θ is smaller than cθ , the light will escape.

3.2 Descriptions of the applied micro-structures In order to confine more light inside the light guide, which means more light should fulfill the TIR condition, the incident angle of light incident on the surface 2 (see Fig. 3.1) ought to be changed larger than critical angle. According to the aim of this report, which is to increase the light in-coupling efficiency by micro-structure, the micro-structures on the surface of light guide should bend light that will hit the surface 2 (see Fig. 3.1 ) with the angle larger than critical angle.

Light source: LED

Waveguide(PMMA)

Light in‐coupled

Escaping light

Ө

Surface 2

Surface 1

In opticprisms of incidlight gubend that angleThree cthe lighlight go

Fig. 3.2 (bfrom left

Fig. 3.2 (cis from ri

In the Fguide (nguide isbended

cs, the prismand pyrami

dent angle ouide using She light rayses larger thacases (normht through aoing through

Fig. 3.2 (a) T

b) The light trside.

c) The light traight side.

Fig. 3.2 (a), n=1.49) is s smaller thcompared w

m is often usids are choson the light Snell’s law s in such a wan critical anal incidence

a light guideh a light gui

The light trace

ace of slanted

ace of abnorm

light impinlarger than han incidenwith dashed

sed as the esen to be coguide-air inon the air-p

way that mongle. e and two se without mide with mic

of normal inci

incidence with

mal incidence w

nges the lighair (n=1),

nt angle. Frod line inside

20

element thatombined onnterface canprism interfost of them

lanted incidmicro-structucro-structur

idence without

hout prism (da

without prism (

ht guide perpdue to Sneom the Fige the light g

t can bend tnto the surfan be determifaces. In ge

m will hit the

dences) are ure is markere is traced b

t prism (dash l

ash line) and w

(dash line) and

rpendicularlell’s law theg. 3.2 (a), itguide, and w

the light. In ace of the liined by traceneral, the me parallel to

demonstrated with the by the solid

line) and with

with prism (sol

d with prism (s

ly. As the re refractivet is easy towill hit the

this projecight guide. cing a light micro-structop and botto

ted in Fig. 3dashed line

d line.

prism (solid lin

lid line). The in

solid line). The

refractive ine angle insido find that s

light guide

t, triangularThe changethrough thetures wouldom surfaces

3.2 in whiche, while the

ne).

ncident light is

e incident light

ndex of lightde the lightsolid line istop surface

r e e d s

h e

s

t

t t s e

21

with larger angle (θ ) than dash line. Thus the chance for TIR condition to be fulfilled is increased. The Fig. 3.2 (b) and (c) are two cases that show the deviation of light going through the light guide from left and right side of the normal at light guide bottom surface respectively for the slanted incidence case. One result (in Fig. 3.2 (b)) is quite similar to the normal incidence case, in which the angle θ becomes larger. In Fig. 3.2 (c), the result is in other way round, in which the angle θ is smaller. However, this will not have effect on in-coupling efficiency, since according to TIR theory, all the light passing through the parallel light-guide without prism could not be kept inside the light guide. Therefore, in theory, the prism and pyramid structures are helpful for bending the light.

3.3 Solutions As analyzed in Chapter 3.2, without structures, light always escapes; with structure, some light is captured. So now the challenge is: which geometry for the micro-structure captures most of the light. Some solutions will be discussed in this chapter. Considering the position of the light source with respect to the light-guide, in-coupling methods are classified into direct-lit and side-lit methods. In direct-lit methods, the light source is positioned parallel to the light-guide top and bottom facets. In side-lit methods, the source is positioned at the edge of the light-guide [1]. In my report, only the direct-lit method is discussed. In this case light sources can be distributed over the entire surface of the light guide. The methods in this chapter will be explained in two groups: light source using optical contact with light-guide or non-optical contact. Further in each group, there will be two cases: micro-structure on the same light guide side or opposite light guide side as light source. The Light-Tools software is used to simulate the designed module before implementation.

1: Prism Micro-structure

Group 1: light source in optical contact with the light-guide

Method A1: light source is in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source.

The light source is on the center of the light guide. The micro structure is aligned between the light source and the light guide. This is showed in Fig. 3.3.

Fig. 3.3 Mguide as l

Methodsame silight gu

The lighsource. mirror rA1, andin Fig. 3

Method A1: liglight source.

d B1: light side of light guide.

ht source isThe mirrorreflected sod gives them3.4.

ght source in o

source is inguide as lig

s on the cenr also contacome light wm one more

optical contact

n optical conght source. A

nter of the licts the other

which will echance to p

22

Side view

Top view

t with the light

ntact with tA mirror in

ight guide. Tr side of ligscape the li

propagate in

w

w

t-guide. The m

the light-gun optical con

The micro sght guide anight guide tnside the lig

micro-structure

ide. The mintact is on t

structure is d is alignedthrough theght guide. T

e is on the sam

icro-structuthe opposite

aligned undd with light bottom in his behavio

me side of light

ure is on thee side of the

der the lightsource. Thethe method

or is showed

t

e e

t e d d


Methodsame siopposit

The lighsource. light soescape the ligh

Method B1: liglight source. A

d C1: light side of lightte side of the

ht source isThe Lamb

ource. The Lthe light gu

ht guide. Thi

ght source in o mirror in opti

source is int guide as e light guide

s on the cenertian scatteLambertian uide throughis is explain

To

optical contactical contact is

n optical conlight source.

nter of the lierer also coscatterer (s

h the bottomned in the to

23

Side view

op view

t with the lighton the opposit

ntact with te. A Lambe

ight guide. Tontacts the see Chapterm, and giveop part of Fi

Side view

w

t-guide. The mte side of the lig

the light-guertian scatt

The micro sother side o

r 2) recycleses them oneig. 3.5.

w

micro-structureght guide.

ide. The miterer in opt

structure isof light guis some lighe more chan

e is on the sam

icro-structutical contac

aligned undide and is aht which othnce to propa

me side of light

ure is on thect is on the

der the lightaligned withherwise willagate inside

t

e e

t h l e


Methodopposit

The lighthe cent


Method C1: liglight source. A

d D1: light ste side of lig

ht source ister of the op

Method D1: lighlight source.

ght source in o Lambertian s

source is inght guide as

located on pposite side

ht source in op

optical contactcatterer in opt

n optical cons light sourc

the central e of light gu

ptical contact w

24

Top view

t with the lighttical contact is

ntact with tce.

face of the ide. This de

Side view

Top vie

with the light-g

w

t-guide. The mon the opposit

the light-gu

light guideesign is dem

w

ew

guide. The mic

micro-structurete side of the li

ide. The mi

e. The micromonstrated in

cro-structure is

e is on the samight guide.

icro-structu

o structure in Fig. 3.6.

s on the opposi

me side of light

ure is on the

is aligned in

ite side of light

t

e

n

t

Methodoppositsurface

The lighthe oppreflects them onof Fig. 3


Methodoppositpropert

The lighthe oppscattersone mo3.8.

d E1: light ste side of li.

ht source isposite side o

some lightne more cha3.7.

Method E1: lighlight source.

d F1: light ste side of lties.

ht source isposite side os the light thore chance t

source is inight guide a

on the centof light guidt which othance to prop

ht source in op

source is inlight guide

on the centof light guidhat otherwito propagate

n optical conas light sou

ter of the ligde. The topherwise wilpagate insid

ptical contact w

n optical conas light so

ter of the ligde. The surfise will escae inside the

25

ntact with turce. The t

ght guide. Tp surface of l escape th

de the light g

Side view

Top view

with the light-g

ntact with tource. The

ght guide. Tface of prismape the ligh light guide

the light-guop surface

The micro sf prism is a e light guidguide. This

w

w

guide. The mic

the light-gusurface of

The micro sm has Lambht guide thre. This beha

ide. The miof prism is

structure is aspecular re

de through behavior is

ro-structure is

ide. The mif prism has

structure is abertian scatrough the bavior is show

icro-structus a specula

aligned in theflective surthe bottom showed in

s on the opposi

icro-structu Lambertia

aligned in thtterer properottom, and wed in top

ure is on thear reflective

he center ofrface which

m, and givesthe top part

ite side of light

ure is on thean scatterer

he center ofrties, whichgives thempart of Fig

e e

f h s t

t

e r

f h

m .


Group 2

In this gGroup are illus

Methodthe sam

Method F1: lighlight source.

2: light sour

group, all th1, except thstrated in th

d G1: light sme side of lig

ht source in op

rce not in op

he moduleshat the lighthe following

source is noght guide as

ptical contact w

ptical conta

s, method Gt source is ng figures.

ot in opticas light sourc

26

Side view

Top view

with the light-g

act with the

G~L, are schnot in optic

l contact wce.

w

w

guide. The mic

light-guide

hemed respecal contact w

ith the light

ro-structure is

ectively simwith the lig

t-guide. The

s on the opposi

milar to methht guide. M

e micro-stru

ite side of light

hod A~F inMethod G~L

ucture is on

t

n L

n

Fig. 3.9 Mlight guid

Methodthe samthe ligh

Method G1: lighde as light sour

d H1: light sme side of light guide.

ht source not irce.

source is noght guide as

in optical conta

ot in opticas light sour

27

Side view

Top view

act with the lig

l contact wrce. A mirro

Side view

w

w

ght-guide. The

ith the lightor in optical

w

micro-structu

t-guide. Thel contact is

ure is on the sa

e micro-struon the oppo

me side of

ucture is onosite side of

n f

Fig. 3.10 light guid

Methodthe samopposit


Method H1: lide as light sour

d I1: light some side of lite side of the

Method I1: ligde as light sour

ight source notrce. A mirror in

ource is noight guide ae light guide

ght source not rce. A Lambert

Top

t in optical conn optical conta

ot in opticalas light soue.

T

in optical conttian scatterer i

28

p view

ntact with the lact is on the op

l contact wirce. A Lam

Side view

op view

tact with the liin optical conta

light-guide. Thpposite side of t

ith the lightmbertian sca

ight-guide. Theact is on the op

he micro-structthe light guide

t-guide. Theatterer in op

e micro-structupposite side of

ture is on the se.

e micro-struptical conta

ure is on the sathe light guide

same side of

ucture is onact is on the

ame side of e.

n e

Methodthe opp


Methodthe oppsurface

d J1: light sosite side of

Method J1: ligde as light sour

d K1: light sposite side .

source is nof light guide

ght source notrce.

source is noof light gui

ot in opticale as light so

t in optical con

ot in opticaide as light

29

l contact wiource.

Side view

Top view

tact with the li

l contact wt source. T

Side view

ith the light

w

w

ight-guide. Th

ith the lightThe surface

w

t-guide. The

e micro-struct

t-guide. Theof prism is

e micro-stru

ture is on the o

e micro-strus a specula

ucture is on

opposite side of

ucture is onar reflective

n

f

n e

Fig. 3.13 of light gu

Methodthe oppscattere

Fig. 3.14 of light gu

Method K1: liuide as light so

d L1: light sposite side er property.

Method L1: liuide as light so

ight source notource.

source is noof light gu

ight source notource.

t in optical con

ot in opticaluide as ligh

t in optical con

30

Top view

ntact with the l

l contact wiht source.

Side view

Top view

ntact with the l

w

light-guide. Th

ith the lightThe surfac

w

w

light-guide. Th

he micro-struct

t-guide. Thece of prism

he micro-struct

ture is on the o

e micro-strum A has a L

ture is on the o

opposite side

ucture is onLambertian

opposite side

n n

31

2: Pyramid micro-structure

For pyramid structures, configurations are quite similar to the ones for the prism structures, but with pyramid replacing prism. We list the configurations below:

Group 3: light source in optical contact with the light-guide

Method A2: light source is in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. The configuration is same as Fig. 3.3, but with pyramid.

Method B2: light source is in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. The configuration is same as Fig. 3.4, but with pyramid.

Method C2: light source is in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. A Lambertian scatterer in optical contact is on the opposite side of the light guide. The configuration is same as Fig. 3.5, but with pyramid.

Method D2: light source is in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source.The configuration is same as Fig. 3.6, but with pyramid.

Method E2: light source is in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. The top surface of prism is a specular reflective surface. The configuration is same as Fig. 3.7, but with pyramid.

Method F2: source is in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. The surface of prism has Lambertian scatterer properties. The configuration is same as Fig. 3.8, but with pyramid.

Group 4: light source not in optical contact with the light-guide

Method G2: light source is not in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. The configuration is same as Fig. 3.9, but with pyramid.

Method H2: light source is not in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. A mirror in optical contact is on the opposite side of the light guide. The configuration is same as Fig. 3.10, but with pyramid.

Method I2: light source is not in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. A Lambertian scatterer in optical contact is on the opposite side of the light guide. The configuration is same as Fig. 3.11, but with pyramid.

32

Method J2: light source is not in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. The configuration is same as Fig. 3.12, but with pyramid.

Method K2: light source is not in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. The surface of prism is a specular reflective surface. The configuration is same as Fig. 3.13, but with pyramid.

Method L2: light source is not in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. The surface of prism A has a Lambertian scatterer property. The configuration is same as Fig. 3.14, but with pyramid.

33

Chapter 4 Initial results and model selection Nowadays, computer simulation before manufacturing is more popular because it is economical and time-saving. Simulation helps scientists get rid of the obviously inefficient model, analyze the details of different methods, and get the optimized parameters of a model before implementation. However, there are still some differences between a model and real devices, therefore the aim of the simulation is to study the efficiency of several models and identify the most efficient and stable ones. The optical simulation software, Light-Tools will be used to perform the computer simulations of the models described in Chapter 3, which will be introduced in the Appendix A.

4.1 Components in Light‐Tools 4.1.1 Modeling of the LEDs 1. Normal cube light source:

As mentioned in Chapter 1, LED sources now become the most popular light source. In some models, LED sources were modeled as cuboids with size: 1.011 ×× mm2, made of a homogenous optically transparent material with refractive index 1.8 and no absorption. 100 lumen photometric fluxes were emitted outwards from the top surface with uniform Lambertian style emission, while the bottom surface was defined as a simple mirror with 100% reflectivity in the sphere models and with 60% reflectivity and 40% absorption in the light guide models. The model of the light source is illustrated in Fig. 4.1. The emission spectrum was defined as having Gaussian shape: A full width at half maximum of 25 nm and a central wavelength at 550 nm. Because methods of light in-coupling are hardly dependent on wavelength, it does not matter what color of LED source is used. All the parameters setup here refer to the Philips Rebel Lumiled [13], which will be used in the implementation of any prototype.

Fig. 4.1 Model of light source. LEDs was approximated as only consisting of semiconductor die.

34

2. Ray data source: Instead of generating light rays from a model of an LED source, it is also possible to generate rays by reading them from a file containing these rays. In chapter 4.2.3 of this report, we use a simple model of a light source that is not in optical contact with the light-guide. In this model, called “sphere model”, actually sphere model 3 as shown in Fig. 4.4 (b), it is more convenient to use ray data from a file instead of generating them from a model of the LED for the following reason: when using ray data from a file there is no LEDs model that can interfere with reflected rays. The ray data file is obtained by collecting rays by means of a receiver located very close to the emitting surface from a model of the LEDs source. The collected rays are exported to a file.

4.1.2 Modeling of the light‐guide The light guide was modeled as a cubic piece of material PMMA or Polycarbonte. The size is

2.01515 ×× mm2, and the refractive index is defined for two different cases: 1.49 as the value for PMMA in the non-optical contact case and 1.59 for Polycarbonate in the optical contact case, because in the latter case a glue with a refractive index of 1.43 was used to achieve optical contact and the refractive index difference between the two contacted materials (glue and light guide) was still needed to get refraction from the micro-structure and to fulfill the TIR condition. The four side-surfaces of the light guide should be defined to absorb all the rays which would hit them. Otherwise they would be counted more than once by the surface receiver when they hit the surface again, note that TIR also happens at the side-surfaces thereby potentially trapping light rays inside the light guide. This phenomenon is called billiard ball effect [1].

Fig. 4.2 Light guide in the Light‐Tools. Four side‐faces absorb all the light hitting them.

4.1.3 Modeling of micro‐structure In this project, two micro-structures were used, prisms and pyramids. In Light-Tools, a 3-D texture can be specified on the surface, the geometry and optical property of which can be modified according to the user’s requirement. Prisms and pyramids, two default geometry element shapes in Light-Tools, can be easily applied with adaptable parameters, variables and functions.

4.1.4 Modeling of detectors In Light-Tools, a surface can be defined as a detector. In order to know what percentage of light rays will be in-coupled into the light guide, the four side-surfaces were configured as detectors. The efficiency of the in-coupling light of light is defined as:

35

In-coupling efficiency ≡ η ≡ source thefrom emittedLight

j surface sideon light Incoming4

1∑=j

(4.1)

Where side surface j (j=1, 2, 3, 4) refers to one of the four side-surfaces of the light guide.

4.1.5 Modeling of Lambertian reflectors, mirrors In some models, a property zone, a Lambertian scatterer or a normal mirror, was added opposite to the LEDs to give light rays that formerly would escape a second chance to be coupled into the light guide. The location, size and optical property of the property zone can be adjusted according to the customer‘s demand in the Light-Tools software.

4.2 Simulation result and discussion:Sphere model In this section, it will be verified by a simple simulation experiment that the models with micro-structures, prisms and pyramids, described in Chapter 3.2 are helpful for in-coupling light rays.

Fig. 4.3 (a) The model used to make sure the micro‐structure is useful. All the rays hitting the surface of the sphere are nearly normal incidence which means all the rays will escape from the sphere without any refraction or reflection. (we did not consider Fresnel reflections.)

Fig. 4.3 (b) Angles from 0V to 180V in latitude‐longitude coordinates.

This moredirectLight-Tplane. T(size: 2make sdrawn b180V (geometrhappen related Three sFig. 4.4

Fig. 4.4 (athe light 4(pyramid

Fig. 4.4 (blight guid3(pyramid

odel is a sited by the m

Tools is shoThe light so

02.022 ×× mure each raby the blac(see Fig. 4rics, the anto be the s

light guide simple spher4) and light

a) Sphere modguide model ds).

b) Sphere modde model wheds) .

impler modmicro-structwn in Fig. ource (size:mm3) whichay from theck -lines) w4.3 (b)) in ngles, from same as themodel illusre models aguide mode

el 1, light sourwhere the mi

del 2, light sourre the micro‐s

del. From thtures into an4.3(a). It is 1.011 ×× mh is four time source hit

was used to latitude-lo0V to 180

e angles witstrated on thare used andels (in chapt

ce outside of tcro‐structure i

rce inside of thstructure is on

36

his model, wngles that fus a half sphemm3) is locmes the areats the micrrecord the

ngitude co0V (see Figth which thhe left side od the relatioter 3, on the

he sphere and is on the sam

e sphere and mn the opposite

we can bettulfill TIR coere with thecated abovea of the lighro-structure.e emitting pordinates (

g. 4.3 (b)), he light raysof Fig. 4.4.on between e left of Fig

micro‐structure side as the

micro‐structuree side as the l

ter understaondition. The micro-strue the center ht source. T. A far fielpower for e(see Appenmarked on

s hit the gu

the sphere m. 4.4) is sho

res on the top light source i

es on the top slight source in

and how mhe simulatioucture on th

of the micThis is done,ld receiver each angle fndix 2). Acn the far fieuide-air inte

models (on own below (

surface , is usen group 2(pris

urface, is used n group 1(prism

much light ison model inhe top circlero-structure, in order to(the spherefrom 0V toccording toeld receiverrface in the

the right of(Fig. 4.4)

ed to representsm) and group

to present thems) and group

s n e e o e o o r e

f

t p

e p

Fig. 4.4 (top surfaclight guid4(pyramid

The sim 4.2.1 Light soWith thin Lightin Fig. 4

Fig. 4.5 T

( o 20,10

0

10

20

30

40

50

60

70

80

90

100

percen

t pow

er( %

)

c) Sphere modce, and one wade model wheds).

mulation par

Sphere moource outsidhe models, pt-Tools, 1004.5 and disc

The curves of tooo 5,40,30,0

0

0

0

0

0

0

0

0

0

0

0

3 9

dels 3, includingave guide where the micro‐s

rameter of th

odel 1: de of the sphprisms or py0,000 rays acussed in th

he percentageooo 70,60,50

15 21 27

T

g one sphere wre the light soustructure is on

hese model

here and miyramids on are traced. T

he following

s of power varooo 89,80, , n

7 33 39

reflect

Texture ty

37

with a ray dataurce is not in on the opposite

s is the bott

icro-structuthe same si

The results fg.

rying with angl

o texture).

45 51 5

tive angle(deg

ype: Pris

a source insideoptical contact e side as the l

tom angle o

ures on the toide of the lifrom simula

les from 0V to

57 63 69

gree)

sm

e of the spherewith the light light source in

of the prisms

op surface.ight source ations for th

180V for 10 d

75 81

e and micro‐strguide, is used

n group 2(prism

s or pyrami

respectivelhe prisms ar

different prism

87

b89b80b70b60b50b40b30b20b10b1 no

ructures on theto present thems) and group

ds.

ly, availablere presented

bottom angles

bottom angle: 9 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: 0 degreebottom angle: degreeo texture

e e p

e d

s

38

Fig. 4.5 shows the information from the far field receiver, in which the horizontal axis represents the latitude angle on the far field receiver and the vertical axis gives the data about how much total power which is the integrated power from o0 is detected by far field receiver. In the right side of the figure, curves with different colours present the results for different values of the prism bottom angles while the black one gives the information of the model without any texture. We can see from the Fig. 4.5 that, when the latitude angle is o42 , the critical angle of PMMA, all the other curves with different colours are below the black one, which means the model with the prisms will have a higher in-coupling efficiency than the model without any texture and the lowest point is on the light blue curve where the bottom angle is o40 . In addition, it can be easily found that in the Fig. 4.5, the orange line where the bottom angle is o50 is very close to the light blue line, and same phenomenon can be seen in other 4 couple lines, purple line where the bottom angle is o30 and dark blue line where the bottom angle is o60 , green line where the bottom angle is o20 and yellow line where the bottom angle is o70 , brown line where the bottom angle is o10 and light yellow line where the bottom angle is o80 , black line which represents without any texture and red line where the bottom angle is o89 . Hence, Fig. 4.6 is made to give a detailed view about how much light power will be kept inside the related light guide in the right of Fig. 4.4 (a), where the x-axis presents the prism bottom angle and the y-axis gives the information of the percentages of in-coupling power. Obviously, the highest one is, at the point that the prism bottom angle is around o40 , while without prism, the percentage is almost zero.

Fig. 4.6 The curve for how much light power will be kept inside the light guide, when prism is used.

Next, the micro-structure prisms are replaced by pyramids, and the simulations are repeated again under the same conditions. Fig. 4.7 summarizes the results; similar as for the prisms, the values of all the curves at x-axis value of o42 (the critical angle for TIR) are below the black one (the model without texture). These values represent the in-coupling efficiency, i.e. the relative amount of power present within the angular range from o0 to o42 . The highest in-coupling efficiency, nearly 47%, is achieved for the case in which the bottom angle is about o50 , which is illustrated in Fig. 4.8.

0102030405060708090100

0,00 20,00 40,00 60,00 80,00

Efficiency ( %

)

Prism bottom angle(degree)

Efficiency vs.Bottom angle

39

To sum up, by means of Light-Tools simulations, the in-coupling efficiency does benefit from texture by prisms and pyramids under the condition of no -optical contact and the micro-structure on the same side as the light source.

Fig. 4.7 The curves of the percentages of power varying with angles from 0V to 180V for 10 different pyramid bottom angles

( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

Fig. 4.8 Curve showing how much light power will be kept inside the light guide, when pyramids are used.

4.2.2 Sphere model 2: Light source inside the sphere and micro-structures on the top surface.

0

10

20

30

40

50

60

70

80

90

100

3 9 15 21 27 33 39 45 51 57 63 69 75 81 87

prcent pow

er(%

)

reflective angle(degree)

Texture: Pyramid

bottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

0102030405060708090100

0 20 40 60 80

Efficiency ( %

)

Pyramid bottom angle(degree)

Efficiency vs. Bottom angle

40

Compared to sphere model 1 in Light-Tools, this model shown in Fig. 4.4 (b) differs: the light source is turned o180 and located inside of sphere. A mirror and a Lambertian scatterer are added separately on the top of micro-structure. The other conditions are the same as in the previous case. With the 6 different models of prisms and pyramids in Light-Tools, simulations are carried out with 100,000 traced rays. Results for the prisms case are shown in Fig. 4.10 and discussed in the following: For these models, it can be easily found from Fig. 4.9 that the value of the black line (the model without texture) is the smallest in all the lines at points where the x-axis value is o42 . Almost all the curves, except for the mirror one and the Lambertian scatterer one, overlap after o90 , which means the model without any texture holds more power than the others. However, the mirror curve is very similar to the Lambertian scatterer one. They both are well above the black one (corresponding to the case without texture). This indicates that the mirror and the Lambertian scatterer models cannot save more power than the model without any texture.

Fig. 4.9 The curves of the percentages of power varying with angles from 0V to 180V for 10 different prism bottom angles

( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

Similar results can also be found for the pyramid case, as shown in the Fig. 4.10. The black line (the model without any texture )is on the bottom when the value of horizontal axis equals o42 ; when the value is larger than o90 , nearly all the lines have the same value but for mirror one and Lambertian scatterer one.

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

percen

t pow

er( %

)


Texture: Prism

bottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degree

41

Fig. 4.10 The curves of the percentages of power varying with angles from 0V to 180V for 10 different pyramid bottom

angles ( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

In a word, the optical contact models with prisms or pyramids located on the side of the light guide opposite to the LEDs cannot conserve more power as we have predicted before. 4.2.3 Sphere model 3: Including one sphere with ray data source inside of the sphere and micro-structures on the top surface, and one wave guide where light source is not in optical contact with light guide. Sphere model 3 in Fig. 4.4 (b) which is same as sphere model 2, except for the light source whose ray data are now imported from a ray data file (for details please see 4.1.1), is made of micro-structures, prisms or pyramids on which 2 different reflectors, a mirror or a Lambertian scatterer, are added respectively. After 100,000 rays are traced in Light-Tools, the results of prism are showed in Fig. 4.11: Although the black line is the lowest curve in Fig. 4.11 (a), it is almost zero until the reflective angle is larger than o90 , which means the model without the micro-structure has the smallest power, nearly zero. The same phenomenon also can be found in light purple line where the bottom angle is o30 and dark blue line where the bottom angle is o1 , while other color curves are all non-zero during the angle from o0 to o90 . Obviously, prisms here are helpful to increase the in-coupling efficiency. The mirror (R=100%) and the Lambertian scatterer (R=100%) are added to the top of the micro-structure respectively in order to find out if it is useful, and the simulation results are showed in Fig. 4.11 (b) and Fig. 4.11 (c). For the mirror case, it is really good for increasing the in-coupling efficiency, since all the curves are above the black one, while in the

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

percen

t pow

er(%

)


Texture : Pyramidbottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

mirror

42

figure with results for the Lambertian scatterer all the curves almost overlap. However, it is difficult to say from the figure directly with which prism bottom angle the in-coupling efficiency is the highest. Based on the formula,

=efficiencyη percent power in o90 –percent power in o42 (4.2) all the in-coupling efficiencies are calculated and drawn in Fig. 4.12.

Fig. 4.11 (a) The curves of the percentages of power varying with angles from 0V to 180V for 10 different prism bottom


0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

Percen

t pow

er(%

)

Reflective angle(degree)

Texture: Prismbottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

43

Fig. 4.11 (b) The curves of the percentages of power varying with angles from 0V to 180V for 10 different prism bottom

angles and a mirror on top of the micro structure ( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

Fig. 4.11 (c) The curves of the percentages of power varying with angles from 0V to 180V for 10 different prism bottom

angles and a Lambertian scatterer on top of the micro structure ( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

Percen

t pow

er (%

)


Texture : Prism with mirrorbottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

percen

t pow

er(%

)


Texture: Prism with lambertian scattering


44

There are three different kinds of lines in the Fig. 4.12.: The blue line with the diamond dots represents the results for the model with only prisms. The red line with square dots gives information about the results for the prisms model with mirror. The green line with triangular dots represents the results for the prims model with Lambertian scatterer. The x-axis represent the bottom angle of the prisms, from o0 to o90 (bottom angle o0 means there is no texture.), whereas the y-axis represents the in-coupling efficiency. It is clear from Fig. 4.12 that, without micro-structure prisms, the highest efficiency is almost 55% for bottom angle o0 . After micro-structures such as prims and mirrors are added to the model, the model with the highest efficiency appears to be the prism models with mirror. The efficiency at o30 is 67% which is larger than 55%, so the prism model with mirror on top of the micro-structure is helpful to raise the in-coupling efficiency with an optimum at bottom angle of

o30 .

Fig. 4.12 The curve for how much light power will be kept inside the light guide, when prisms are used. The three different colour lines represent the results for different reflectors on top.

In the simulations, we replaced the prisms by pyramids and performed simulations under the same conditions. The results are shown in Fig. 4.13 (a) for the sphere model with nothing on top, (b) for the sphere model with a mirror on top, and (c) for the sphere model with a Lambertian scatterer on top. They are quite similar to the ones in prism model. The models of the pyramids or the pyramids with mirror on top do increase the in-coupling efficiency, while the efficiency of a pyramid with a Lamberian scatterer changes slightly. The same formula (4.2) is used again, and the similar results can be found: The results depicted in Fig. 4.14 show that the sphere models of pyramids with a mirror on top (red line with square dots) increase the in-coupling efficiency. The highest point is 67% approximately at o30 . Other two kinds of pyramid models (green line with triangle point and blue line with diamond) seem not to improve the efficiency. To summarize, we find that the in-coupling efficiency will be increased by adding a micro-structure, prisms or pyramids, in combination with a mirror for the models where the light source is not in optical contact and on the opposite side of the light guide.

0102030405060708090

100

0 20 40 60 80

Efficiency (%

)

prism bottom angle(degree)


prism: no zone

prism: mirror

prism: lambertian scattering

45

Fig. 4.13 (a) The curves of the percentages of power varying with angles from 0V to 180V for 10 different pyramid bottom


Fig. 4.13 (b) The curves of the percentages of power varying with angles from 0V to 180V for 10 different pyramid bottom

angles and a mirror on top of the micro structure ( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

Percen

t pow

er(%

)


Texture: Pyramidbottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

Percen

t pow

er(%

)


Texture: Pyramid with mirror


46

Fig. 4.13 (c) The curves of the percentages of power varying with angles from 0V to 180V for 10 different pyramid bottom

angles and a Lambertian scatterer on top of the micro‐structure ( ooooooooo 89,80,70,60,50,40,30,20,10 , no texture).

Fig. 4.14 The curve for how much light power will be kept inside the light guide, when pyramids are used. The three different color lines represent the results for different reflectors on top.

4.2.4 Conclusion: To summarize, in this section, a simple sphere model in Light-Tools has been built to verify if the solutions in Chapter 3 are useful. The sphere model easily allowed us to understand why certain models improve efficiency and others do not. Two different positions for the micro-structure on the light guide are simulated in Light-Tools by the sphere model. The results clearly show that the sphere model in Fig. 4.4 (a) and in Fig. 4.4 (c) are helpful for holding more power,

0

10

20

30

40

50

60

70

80

90

100

3 15 27 39 51 63 75 87 99 111 123 135 147 159 171

Percen

t pow

er(%

)


Texture: Pyramid with Lambertian scatteringbottom angle: 1 degreebottom angle: 10 degreebottom angle: 20 degreebottom angle: 30 degreebottom angle: 40 degreebottom angle: 50 degreebottom angle: 60 degreebottom angle: 70 degreebottom angle: 80 degreebottom angle: 89 degreeno texture

0102030405060708090100

0 20 40 60 80

Efficiency (%

)

pyramid bottom angle(degree)


pyramid: no zone

pyramid: mirror

pyramid:lambertian scattering

47

but the model in Fig. 4.4 (b) hardly helps to improve the in-coupling efficiencies. The former models will be simulated by more accurate light guide based models again to get the optimized values of the parameter under more real condition which is important for making test sample. However, due to the disadvantages of the sphere models, some models with a mirror or Lambertian scatterer on the side of the light guide that is opposite to the light source in group 2 and group 4 or other models with micro-structure on the same side as light source in group 1 and group 3 will be simulated in the light guide models as well. Based on the previous discussion, the models that will be simulated in the light guide are those listed below:

• In prism model: Method A1, Method B1, Method C1, Method G1, Method H1, Method I1, Method K1.

• In pyramid model: Method A2, Method B2, Method C2, Method G2, Method H2, Method I2, Method K2.

4.3 Simulation result and discussion: Light guide model In this section, fourteen methods selected out of those listed in chapter 4.2.4 will be simulated in more accurate light guide based models which are closer to reality. Method A1 and Method A2 will be simulated together since the only difference between them is the micro-structure used. The same holds for Method B1 and B2, Method C1 and C2, Method G1 and G2, Method H1 and H2, Method I1 and I2, and Method K1 and K2. 4.3.1 Method A1 (Prism) &A2 (Pyramid): Light source is in optical contact with the light-guide. The micro-structure is on the same side of the light guide as the light source. The parameters that needed to be optimized are:

The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prisms or pyramids are located in.

After models are set up and simulated in Light-Tools, 100,000 light rays are traced. The results are (unit: mm)

model texture opt_length opt_width opt_angle efficiency

A1 prism 1 1 31 0.41 A2 pyramid 1 1 37 0.41

48

Fig. 4.15 The light guide model A1 and A2 in Light‐Tools.

4.3.2 Method B1 (Prism) & B2 (Pyramid): Light source is in optical contact with the light-guide. The micro-structure is on the same side of the light guide as the light source. A contacted mirror is on the opposite side of the light guide.

Fig. 4.16 The light guide model B1 and B2 in Light‐Tools. The mirror(grey) is in the centre of light guide.

The parameters that needed to be optimized are: The bottom angles of the prisms or pyramids.

49

The width and length of the texture zone where the prisms or pyramids are located in. The width and length of the mirror.


model texture opt_length opt_width opt_mirror_length

op_mirror_width opt_angle efficiency

B1 prism with mirror

1 1 0.5 0.5 26 0.32

B2 pyramid with mirror

1 1 0.1 0.1 41 0.41

4.3.3 Method C1 (Prism) & C2 (Pyramid): Light source is in optical contact with the light-guide. The micro-structure is on the same side of the light guide as the light source. A contacted Lambertian scatterer is on the opposite side of the light guide. The simulation models in Light-Tools are the same as B1 and B2 in Fig. 4.16, but the Lambertian scatterer is used instead of a mirror. The parameters that needed to be optimized are:

The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prism or pyramids are located in. The width and length of the Lambertian scatterer.

After models set up and simulated in Light-Tools, 100,000 light rays are traced. The results are (unit: mm)



C1 prism with scattering 1 1 1 1 31 0.55

C2 pyramid with

scattering 1 1 1 1 29 0.54

4.3.4 Method G1 (Prism) & G2 (Pyramid): Light source is not in optical contact with the light-guide. The micro-structure is on the same side of the light guide as the light source. The parameters that need to be optimized are:

The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prism or pyramids are located in.

50

Fig. 4.17 The light guide model G1 and G2 in Light‐Tools.


model texture opt_length opt_width opt_angle efficiency

G1 prism 1.00 1.00 43 0.31 G2 pyramid 1 1 46 0.45

4.3.5 Method H1 (Prism) & H2 (Pyramid): Light source is not in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. A contacted mirror is on the opposite side of the light guide.

Fig. 4.18 The light guide model H1 and H2 in Light‐Tools. The mirror (grey) is in the centre of the light guide.

51

The parameters that needed to be optimized are:

The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prisms or pyramids are located in. The width and length of the mirror.




H1 prism with mirror 1 1 1 1 44 0.36

H2 pyramid with mirorr

1 1 0.8 0.8 80 0.45

4.3.6 Method I1 (Prism) & I2 (Pyramid): Light source is not in optical contact with the light-guide. The micro-structure is on the same side of light guide as light source. A contacted Lambertian scattering is on the opposite side of the light guide. The simulation models in Light-Tools are the same as H1 and H2 in Fig. 4.18, but a Lambertian scatterer is used instead of a mirror. The parameters that needed to be optimized are:

The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prisms or pyramids are located in. The width and length of the Lambertian scatterer.




I1 prism with scattering 1 1 1.1 1.1 36 0.60

I2 pyramid with

scattering 1 1 1 1 30 0.55

4.3.7 Method K1 (Prism) &K2 (Pyramid): Light source is not in optical contact with the light-guide. The micro-structure is on the opposite side of light guide as light source. A contacted mirror is located on the surface of the prisms or pyramids.

52

Fig. 4.19 The light guide model K1 and K2 in Light‐Tools. The bottom of the image shows the micro‐structure is on the opposite of the light guide.

The parameters that needed to be optimized are: The bottom angles of the prisms or pyramids. The width and length of the texture zone where the prisms or pyramids are located in. The width and length of mirror.

After models are set up and simulated in Light-Tools, 100,000 light rays are traced. The results are (unit: mm).

model texture opt_length opt_width opt_mirro

r_length op_mirror_

width opt_angle efficiency

K1 prism with mirror 1 1.2 1 1.2 24 0.45

K2 prism with mirror 1 1 1 1 24 0.42

4.3.8 Discussion Fig. 4.20 gives an overview of light guide model results.

53

The results of model B2&B2, H1&H2 and K1&K2 which not only have the micro-structure but also have a mirror on the side of the light guide opposite to the light source, show that adding a mirror is not useful for increasing the in-coupling efficiency. For the models with a Lambertian scatterier, they do increase the efficiency (i.e. the percentage of the power held inside light guide), which can be found in the results of model C1&C2 and I1&I2. When a Lambertian scatterer is used, the prism structure is better than the pyramid structure. Also, non-optical contact cases have higher efficiency than optical contact cases. It is easy to find that the highest efficiency is 60% for Model I1 in which the light source is not in optical contact with the light guide and the mirror is on the opposite side of the light guide, while the lowest is 30% for Model G1. Comparing the efficiency of K1 in the light guide model, 45%, with that in sphere model, 66%, the former one is less. The reasons are listed as below:

• The different reflectance used in the surfaces of light guide: 100% in sphere model and 90% in light guide model. The other 10% in the light guide model is absorbed.

• The different absorption used in the bottom surface of the light sources: 0% in the sphere model and 40% in the light guide model.

Due to above reasons, it is reasonable that all the efficiencies obtained with the light guide models are somewhat lower than those obtain for the sphere models. The optimized prism bottom angle in model G1 is o43 , which matches the value, around o40 , calculated for sphere model 1. A similar result holds for model G2: The bottom angle is o46 in model G2, which in sphere model is nearly o50 . A similar result also can be found in model K1 and model K2. These results prove that the sphere model is useful as a simple model to roughly sort out which are the useful solutions in chapter 3 before using a more complex and computational demanding light guide model.

Mode

Mode

Mode

Mode

el A1 (Prism) &

el B1 (Prism) &

el C1 (Prism) &

el G1 (Prism) &

F

0,00

0,10

0,20

0,30

0,40

0,50

0,60

0,70

0,80

0,90

1,00

Efficiency

&A2 (Pyramid)

& B2 (Pyramid)

& C2 (Pyramid)

& G2 (Pyramid)

ig. 4.20 Overv

0,41

0,32

0,5

In‐coup

)

)

d)

iew of results f

5

0,310,36

0

pling effi

54

for 14 light guid

0,60

0,450,41

ciency in

Model H1 (Pri

Model I1 (Pris

Model K1 (Pris

de models with

1 0,41

0,540,

n Light gu

ism) & H2 (Pyr

sm) & I2 (Pyram

sm) &K2 (Pyra

h prisms or pyr

45 0,45

0,55

uide mod

ramid)

mid)

amid).

amids.

0,42

del

A1B1C1G1H1I1K1A2B2C2G2H2I2K2

4.4

WhefficFirsefficeffic[2], reflemmprevthe efficSecselemacthe on tConwithstruOnlintothinfilmWhbacksurfinsicritiincr

Thethe

4 Mod ich modelsciency, but stly, the apciency in tciency in prand 55.7%

ector [1]. Inm thick lighvious paper light guide

cient. ondly, whe

ection rule chine a micsame time h

the oppositensidering thh white met

uctures, we dly models so a prototypn film from m is located

en the lightk to the lighface of the de the lightical angle, reased.

erefore, the light source

els sele

s will be ialso on con

ppropriate this projectrevious way

% in a 0.2 mn this reportht guide whr [1], the in-e, so we ex

ether or nohere. Fromro-structurehave sufficie side. he very smatal, in previodecided notsimilar to mpe by the woAvery Denon the side

t rays hit thht source aLEDs, thes

t guide. Theand be kep

Fig. 4.21

models wite will be sel

ction

implementenvenience foones need t with otheys is 45% in

mm thin light, the higheshile the oth-coupling efxpect our m

ot a methodm a manuface with a smient position

all improvemous paper at to make a pmodel K1, sorkshop. Fo

nnison (the re of the lighhe surface olong the sase rays willey might hitpt inside th

1 The prototyp

th retro-reflelected to be

55

ed in workor manufact

to be sorer methods n a 0.3 mmht guide in st in-couplinhers are allfficiency is

model I1 as

d can be imcturing poi

mall area (1×n accuracy

ment, only and the comprototype osee Fig. 4.2or the retro-reader is refht guide oppof the retro-ame directiol be diffuset the top surhe light gu

e with Retrore

ective Film implement

kshop depenturing. rted out by

mentionedm thin light g

the first geng efficienc

l below 55.inversely p discussed

mplementedint of view

1× mm2) onrelative to L

8% when mplicated pro

f model I1.20, could re-reflective fferred to chposite to th

-reflective fon as the ined and get arface with a

uide. Then,

flective Film (w

on the sideted into a pr

nds not on

y comparind in chapterguide in theeometrical ocy is 60% fo.7%. Accor

proportionalin this rep

d easily is , it is not

n the surfaceLambertian

compared toduction pro

elatively simfilm we use

hapter 2 for he light sourfilm, the refcident lighta second chan angle thathe in-cou

white square).

e of the lighrototype.

nly on incr

ng the in-cr 1.2. The e wave-optioptics way, for model I1rding to stu to the thickort to be th

also an imeasy to eme of PMMA

n scatterer o

to the resulocess of the

mply impleed a retro-remore detailrce (see Figflected rays t. Due to thhance to prat is larger t

upling effici

ht guide opp

rease in

coupling highest

ical way using a

1 for 0.2 udies in kness of he most

mportant mboss or A and at r mirror

lts, 52% e micro-

emented eflective ls.) This g. 4.21).

will go he rough ropagate than the iency is

posite to

Ch

In thwill

5.1Theligh

Fig. the

In pto pof tlighsourcircPMmig

hapte

he previousl explain the

1 Test‐e prototype ht source and

5.1 The picturRetroreflectiv

prototype, Pprevent the Tthe in-couplht guide. Thrce is locat

cular PMMAMA light g

ght be kept i

er 5 M

s chapter, we structure o

‐modulewill be build Retrorefle

re of prototypeve film glued on

PMMA lighTIR happenling efficien

he details arted in the cA light guidguide will einside the li

Meas

we have alreof prototype

e introdlt of three pective Film

e: the transparn the black pap

ht guide is mning in the sncy will bere illustratedcenter of thde, in whichescape, comght guide b

56

surem

ady selectede and discus

ductionparts as showith black

rent disc in theper. In the bot

made to be aside surfacee decreased d in Fig. 5.2

he light guidh all the in-mparing to y TIR.

ment

d the modelss the exper

own in Fig. paper.

e middle is the tom of light gu

a circular ds of light guby trappin

2. It is veryde. The lef

-coupled ligthe right p

ts in l

l to be implriment resul

5.1: PMMA

PMMA light guide is the light

disc with a duide. Otherwng in-coupliy important ft picture ofght ray hittinpicture in w

lab

lemented. Hlt.

A light guid

guide. On the tt source with h

diameter ofwise the tesing light inthat the LE

f Fig. 5.2 sng the boun

which the li

Here we

de, LED

top of it is heat sink.

f 40 mm st results side the

ED light shows a ndary of ight ray

57

Fig. 5.2 The blue square is the LED light source and blue lines are the light emitted from the LEDs. The purple line represents the side surface of light guide. Left: the circular light guide. All the light hitting the side surface will escape except Fresnel reflection. Right: square light guide. Some light rays hitting the side surface with incident angle larger than critical angle will be trapped inside the light guide.

The LED light source used is a commercial LUXEON Rebel LEDs made by Philips and glued on an Aluminum cylinder as depicted in Fig. 5.3. The blue square is the heat sink with LEDs and three pillars on the top surface are used to fix the light guide. The diameter of heat sink is 30 mm. The Fig. 5.4 shows the LEDs structure using Leica microscope. The dark dots on the LEDs die are the contacts and they do not emit light at all.

Fig. 5.3 The draft of LED light source: the LED (blue square) is mounted on the Aluminum cylinder.

Fig. 5.4 Leica microscopic pictures of LEDs die.

+

30 mm

Retrinve

Thediamsprecoun

BetwThe

5.2

roreflectionestigated

1. OmniC2. High In3. Metaliz

e setup of bmeter of fieading out nted as in-c

Fig. 5.5

ween light e refractive i

2 Integ

n Film is p

Cube Series ntensity Prized Series W

black plasticlm is 25 mfrom the to

coupled ligh

5 Scheme of Re

guide and Lindex of it i

grating s

produced by

White T-11smatic GradWhite 4500

c with retromm. The fuop side of ht.

etroreflection f

LED light sis 1.5 at 25

sphere i

Fig. 5.6

58

y AVERY

1500 de Series W.

oreflection ifunction of

the light gu

film layout. Th

source, optiCo for 589

introdu

6 Integrating s

DENNISO

White T-6500

s illustratedblack plas

uide. Other

his will be adde

ical oil is u9.3 nm.

ction

phere in lab.

ON. Three k

0

d in the Figtic is to abrwise these

ed on the top o

used to ensu

kinds of fi

g. 5.5, in whbsorb all thlight rays

of light guide.

ure optical

ilms are

hich the he light will be

contact.

59

The integrating sphere [14], showed in Fig. 5.6, is used to measure the in-coupling efficiency. The integrating sphere is an empty ball with a high-reflecting diffuser coating on the inner surface with a photon detector connected at one of its port. The light emitting object will be located in the center of the sphere, and emitted light will hit the inner surface and bounced several times until it hits the detector surface. In this way the total flux of light emitted by the object can be accurately measured.

5.3 Test results 5.3.1 Linear Response of LEDs In order to give a stable working condition for LEDs in the following experiments, the linear response of LEDs is measured here and result is showed in Fig. 5.7.

Fig. 5.7 LED I-V curve.

Compared with Fig. 2.2 in Chapter 2, the threshold voltage of this kind of LEDs is 2.544 V. After that the current increases fast. Also, the emitting power is tested by using integrating sphere and the related software. The Power-Current Curve is made below:

050

100150200250300

0 0,5 1 1,5 2 2,5 3 3,5

I (mA)

V (v)

LED I‐V Curve

60

Fig. 5.8 LED L-I curve.

The LEDs emitted light is almost linear to the applied current. However, beside the current applied on the LEDs devices, the temperature of LEDs when it is working will also have impact on the LEDs working performances. Therefore, the V-Time Curve is measured in the laboratory, too, which is illustrated in Fig. 5.9:

Fig. 5.9 LED voltage -Time curve.

We can see from the figure that LEDs voltage starts from 2.8944 V, and then decreases to 2.89 V, the lowest point, at 1 hour later. However, after it is lighted on 1.5 hours, the curve fluctuates around 2.8912 V and finally, increases to 2.8921 V after 15 hours. The deviation of LEDs voltage is calculated as below:

0

0,05

0,1

0,15

0,2

0,25

0,3

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280

L (w

)

I (mA)

LED Luminace Flux‐I Curve

2,889

2,89

2,891

2,892

2,893

2,894

2,895

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Volta

ge (V

)

Time (hours)

LED Voltage ‐Time Curve

It isSo tfollo

5.3Fig.

Fig. fromretroretro

Fig. Seco

s easy to finthe LED ligowing meas

.2 Tes. 5.10. and F

5.10 Three com the LEDs. Soreflection filmoreflection film

5.11 The prond row is side

mmax −=

VV

D

nd that the dght source surements.

st‐prototypFig. 5.11.sh

onfiguration uSecond configm. Third con

m.

ototypes measuview of protot

10max

min×

−V

deviation is can be trea

pe experimhows three c

used for the mguration (B) mfiguration (C)

ured using thetype and top vi

61

2894.2%00 =

very small ated as a sta

mental resconfiguratio

measurement. Fmeasures the ) measures th

e labsphere (boiew of Retrore

8944.89.244

×−

during 16 hable light so

sult ons measure

First configuratotal light flu

he total light

ottom): Top leflection film on

1.0%100 =

hours, whichource, and

ed in the lab

ation (A) measux coupled in

flux coupled

eft is model Bn the black pa

%5

h can be necan be use

boratory:

sures the total nto light guidd into light gu

B. Top right is per, respective

eglected. d in the

light flux e without uide with

model C. ely.

Thrand calc

Aftethe showlighThelighFilmin 1threComa 7%case

ee kinds of1 mm. Th

culated by:

er luminanc0.25 mm,

wed in Fight guide, 46e Retroreflehtguide thickm 1 is 37% mm thick

ee films. mparing to t% increase ie, respective

f light guidhe current

ce flux pow0.5 mm an. 5.12. With.9% for thective Film iknesses andin 0.25 mmlight guide.

the sample in 0.25 mm ely.

Fig. 5

es with diffapplied on

wer is measund 1 mm lhout the fil

e 0.5 mm this helpful fo

d Film 1 is tm thick light. It is also o

without Recase, 6.2%

5.12 The efficie

62

fferent thickn LEDs is

ured with lalight guide lm, the effi

hick light gufor increasinthe most ust guide, 53.

obvious that

troreflectiv% increase in

encies of differ

knesses are 20 mA. T

absphere, thare calcula

iciencies areuide and 54ng in-coupliseful one of.1% in 0.5 mt Film 3 has

e Film (Blun 0.5 mm ca

rent Retrorefle

measured: The in-coup

he in-coupliated using e 30% for .3% for theing efficienf all 3 Filmsmm thick lis the lowest

ue Diamondase and 2.8%

ection Film.

0.25 mm, pling effici

ing efficien(5.1) and tthe 0.25 m

e 1 mm lighncy for all ths. The efficiight guide at efficiency

d), Film 1 re% increase i

0.5 mm iency is

ncies for they are

mm thick ht guide. he three iency of and 57.1% in these

esults in in 1 mm

63

5.4 Conclusion We can see from the results in chapter 5.3 that Retroreflective Film really has an effect on increasing in-coupling efficiency and the Film 1 shows the best performance. In previous study [1], a similar experiment was conducted using white reflecting metal instead of the retroreflection film (referring to the figure 6-11: method A), in which the results are 42% for the 0.25 mm thick light guide, 60.7% for the 0.5 mm thick light guide and 71.4% for the 1 mm thick light guide. However, our results are all lower than that in previous study: 5% in 0.25 mm light guide case, 7.6% in 0.5 mm light guide case, and 14.3% in 1 mm light guide case. The difference is firstly caused by different scatterers used in model: in this paper the Retroflective Film from DELTE with 60% reflectivity [15] along the incident direction, compared with Sol-Gel coating with 96% reflectivity used in previous paper [1]. Besides the difference between scatterers, the temperature is also one of the factors affecting efficiency. Since there are 4 samples in total to be measured for each thickness, the LED is applied with 20 mA current for a long time. The temperature of PMMA is very high, which affects the shape of light guide and the centre of the disc is bent. The shape bending in the centre will lead to the error that light source is not in optical contact with light guide, and consequently the efficiency will decrease. Therefore during the experiment, keeping the temperature of LEDs stable is important in order to achieve the high efficiency.

64

Chapter 6 Results and conclusion

6.1 Conclusions In this project we investigate methods to efficiently couple light from top-emitting LEDs into thin light-guides. More specificly, we investigate the possibilities of using micro-structuring of the light-guide close to the LEDs to improve the in-coupling. The micro-structures considered are prisms and pyramids. We also look into using retro-reflection films for improving the in-coupling efficiency. The basic geometry consists of a top-emitting LEDs positioned close to a thin light guide. Close to the LEDs, the light-guide is micro-structured; the micro-structures considered are prisms and pyramids. The micro-structures can be on the same side of light-guide as the LEDs or on the opposite side. When located on the opposite side, the micro-structures are coated with a specular reflecting layer or with a diffuse reflecting layer. The LEDs can either be in optical contact with the light-guide or not in optical contact. In this manner, different geometries are possible. We have investigated the in-coupling efficiency of all these geometries in a systematic way by means of ray-tracing simulations with the ray-tracing software package Light-Tools. For all of the geometries, a model in Light-Tools has been made. For each of the models, we have determined the in-coupling efficiency. We took a two-step approach. In order to ascertain which geometry is most likely to show a high in-coupling efficiency, we decided to first build in Light-Tools a simple model of an LED in proximity to a micro-structured light-guide. In the next step, we made a more elaborate and more accurate model of the most promising geometries. The simple model is based on locating an LED close to a semi-sphere with a micro-structure in between the LEDs and semi-sphere. This model allows us to understand more easily why certain models improve efficiency while others do not. The results apparently show that the sphere models in Fig. 4.4 (a) and in Fig. 4.4 (c) are helpful for holding more power. On the other hand, the model shown in Fig. 4.4 (b) hardly helps to improve the in-coupling efficiency. Resulting from the outcome of this simple model, we selected fourteen geometries out of those listed in Chapter 4.2.4 and used a full model to more accurately model the LEDs, light guide and micro-structure. The results of the simulations show that, for micro-structured located on the light-guide on the side opposite to the LEDs, a Lambertian scatterer is more useful than a mirror. Also, in general, the prisms structures perform better than the pyramid structures. The highest efficiency is 60% for Model I1 in which the light source is not in optical contact with the light guide and the mirror is on the opposite side of the light guide. Considering the very small in-coupling efficiency improvement which is only 8%, compared to 52% the result in previous paper and complicated production process of micro-structure based models, we decided not to make a prototype of model I1.

65

Due to the special optical property of Avery Dennison Retroreflection Film, prototypes are decided to use Retroreflection Film instead of Lambertian scatterer. A prototype was built and made of three parts: a PMMA light-guide, an LED light source and Retroreflection Film. A piece of black paper was used to absorb rays that originate from locations other than the exit of the light-guide. The I-V curve and the L-I curve of LEDs are measured at first to find the appropriate working current. We made sure the measurements were performed after the LEDs voltage (for obtaining a fixed current) was stabilized. Three kinds of light guides with different thicknesses are measured: 0.25 mm, 0.5 mm and 1 mm. The results in chapter 5.3 show that a Retroreflection Film really has an effect on increasing in-coupling efficiency. We tested several Retroreflection Films. The best in-coupling efficiencies measured are 37% for a 0.25 mm thick light guide, 53.1% for a 0.5 mm thick light guide and 57.1% for a 1 mm thick light guide. Compared to the sample without Retroreflective Film, the best Retroreflection Film results in a 7% increase for the 0.25 mm case, 6.2% increase for the 0.5 mm case and 2.8% increase for the 1 mm case, respectively.

6.2 Future Work So far, the project on using micro-structure to couple more light into light guide has finished. However there are still some problems left for further study: Firstly, a thorough study should focus on measuring the optical properties of Retroreflection Films as well as developing a model in Light-Tools. Given that the Retroreflected light will spread out actually in a cone, how much light will be reflected back exactly along the direction of the incident ray should be carefully analyzed as a function of the angle of the incident ray. Secondly, although not in our control, the reflection of the LEDs die should be improved. After the light is emitted from the top surface of the LEDs into the light guide, some light rays still have a chance to be reflected back to the LEDs die and be reflected back by the bottom surface of LEDs. If the back of the LEDs die has a large reflectance, more energy will be reflected back into the light guide again. Other methods can also be studied in the future: 1) other types of micro-structures could be used. Here we only tried two geometric structures in this project, prism and pyramid. There are still several other structures being widely used in the surface luminance uniformity, which may also help for light in-coupling. 2) The position of the micro-structure could be relocated. In our project, it is located on the top or bottom surface of the light guide. Instead of this, the micro-structures can be located inside the light guide. 3) Although the best performing geometry (model I1) has 60% in-coupling efficiency, it is not easy to manufacture in the workshop due to its sophisticated procedure of production The problem is in making the patterns: it is difficult to emboss the patterns in such small size (1 x 1 mm2), which means a more careful embossing process is required.

66

Appendix 1

Instruction to simulation software: Light-Tools

Light-Tools [16] is optical modeling software where users can build, view, modify and analyze optical systems on computer before implementation, which is similar to CAD programs. However, according to requirements of optical design and engineering, Light -Tools has more function than CAD, for instance, extended numerical accuracy and prefessional ray tracing software.

1.1 Interface Elements Light-Tools is windows-based software which consists of some other windows and controls, as illustrated in Fig. A1.1. There are three navigation windows and output window is fixed inside the main window, but if needed, they can be separated or closed from main window. Menu bar: Light-Tools has some common menus on the Menu bar: File, Edit, View, Window, Help and so on, which will change depending on the active window.

System Navigator: Uses tree structures to show all the names and relationships between objects built in 3D design view window. In this window, you can view all the objects and modify them by clicking the plus and minus symbols. Preferences Navigator: Provides a way to Preferences dialog box where users can not only restore the default values but also set and save usual and view preferences. Windows Navigator: Gives a view of all the open windows which you are working with. 3D Design View Window: Provides the whole view of user designed model. In the top and left side, there are many options which user can use to set, modify and edit the structure and property of the model. Properties dialog box: this box is just a window whose content depends on things user selected. Console window: Gives a view of command operated by user and error message fed back to user.

67

2 Fig. A1.1 Light‐Tools working interface.

1.2 Immersion manager In Light-Tools, all the space are assumed surrounded by air, so without specially defined optical contact between two close optical surface there is always an air gap, which may lead to TIR for incident rays. In Light-Tools, user can use Contact command or immersion manager to avoid this problem. Contact command deletes the air gap, makes them as optical contact and removes the effects of the air gap. User can click the button on the left side of the 3D Design View and follow the steps in the Fig. A1.2.

Fig. A1.2 Process of making optical contact using button on the left side of 3D design View window.

However, contact command can only be used in lens. In the case of this project, since two objects are designed to be optical contact, the immersion manager should be used, which allows ray transmitting through components that are totally or partially in the other material or objects, instead of air. The components are including:

• X prisms • Clad fibers • LEDs • Nested volume sources

Thefigu

Fig.

In telem

1.3In thneedthe selemerof e

pres“Re

e immersionure. (See Fig

A1.3 Three diff

the Fig. A1ments in win

3 Optimhis project,ds to be usedesigned s

ected or delrit function.every variab

ss the greenesults” and a

n manager g. A1.3)

ferent ways to

1.4 there is ndows into

mization because med to help osystem. Theted by clic In the righ

ble can be c

n button all the curve

can be acc

access to immnavigator wi

a dialog bother eleme

Fig. A1.4 I

n multiple paraorganize ande dialog bock the “>”

ht side of thcontrolled o

to start thes for every

68

cessed in s

mersion managendow, and from

box of imments.

Immersion man

ameters werd control thox is showand “<” bue box, the mone by one.

he simulatioy parameter

everal way

er: from the 3Dm the main me

mersion man

nager windows

re used, thehe designed wed in Fig. utton in the min and ma. After set a

on. Then tencan be foun

ys illustrated

D design view wenu.

nager. User

s.

e Parametervariables aA1.5. The

middle of tax value andall the para

n results cannd in “Analy

d in the fo

window, from th

r can imme

r Sensitivityand performe variables the box, as d the increaameters in th

n be checkeysis”.

ollowing

he system

erse any

y Utility mance of

can be well as

ase steps his box,

ed in the

69

Fig. A1.5 Main box of Parameter Sensitivity Utility.

1.4 Finding help Same as other popular design software, Light-Tools has two kinds of help system: Contents and Index, and What’s This?. The former one is illustrated as below (Fig A1.6). In the left side of the dialog box, the keyword could be typed in to find the related topics. The other one is What’s This?, which can be found in the help option. When user clicks on a menu toolbar button or command palette with this question mark, the explanation of the menu or button is showed.

Fig. A1.6 One help system: Contents and Index.

70

Appendix 2 Introduction to Latitude-Longitude coordinates

(a) (b)

Fig. A2.1 Latitude –Longitude coordinates showed in three dimensions (3D) (a) and two dimensions (2D) (b).

In the Fig. A2.1, the Latitude-Longitude coordinates are illustrated in three dimensions (a) and two dimensions (b). As showed in Fig. A2.1 (a) θ is for the latitude angle which presents the angle from the positive direction of z axis to the point on the 3D globe, and φ is for the longitude angle which, on the x-y plane, defines the rotation angle around z axis. Therefore, each point on the 3D globe can be specified by these two coordinates θ andφ . Meanwhile, Latitude-Longitude coordinates also can be showed by 2D raster chart (see Fig. A2.1 (b)), in which each circle marked with different θ value describes latitude angle and on the every circle, the number of φ stands for longitude angle. Latitude-Longitude coordinate will be used in analyzing chart view of far field receiver in Light-Tools.

71

References 1. D. S. Canelles, M.P.C. Krijn, G.E. Onac, Light in-coupling in thin light-guides.

Technical Note PR-TN 2010/00534.

2. M. Li, H. Cornelissen, Injection of LED Light into Thin Lightguides using Dielectric Multilayers Technical Note PR-TN 2010/00307.

3. Julio Chaves, Introduction to Nonimaging Optics (CRC, 1st ed., 2008).

4. Hecht. Eugene, Optics (Addison Wesley, San Francisco, 2002).

5. S. Desmond Smith, Optoelectronic Devices (Prentice Hall, 1995).

6. Schubert, E. Fred, Light-emitting diodes (Cambridge University Press, 2004).

7. Door S. M. Sze, Kwok Kwok Ng, Physics of semiconductor devices (Wiley-Interscience, 2007).

8. http://en.wikipedia.org/wiki/Diffuse_reflection.

9. Whelan. T, Goff. J, Polymethyl Methacrylate (PMMA). British plastics & rubber, p. 41-43, July-Aug. (1988).

10. P. Avakian, W. Y. Hsu, P. Meakin, H. L. Snyder, Optical absorption of perdeuterated PMMA and influence of water. Journal of Polyer Science: Polymer Physics Edition, 22(9), p. 1607-1613 (1984).

11. B. Crist, M. E. Marhic, G. Raviv, and M. Epstein. Optical absorption in polymer glasses by laser calorimetry. Journal of Applied Physics, 51(2), p. 1160-1162 (1980).

12. DELTA technical note -101.

13. Philips Lumileds. Luxeon Rebel Es datasheet. (Philips Lumiled, Eindhoven, 2009).

14. Labsphere. A guide to integrating sphere: Theory and applications. (CRC, 2008).

15. Avery Dennison OmniCube T-11500 Prismatic Reflecive Film datasheet. Avery Dennison.

16. Light-Tools handbook.

using micro structures to light into thin light guides462939/fulltext01.pdf1.1 project background...

Documents