Asap Optimization

台灣ASAP使用者會議 Taiwan ASAP User Conference

ASAP在非成像系統的優化方法 ASAP : Optimization tips for Non-Image Optics

歐崇仁 [email protected] 台中 修平技術學院 電機工程系 0921182917 中華民國九

八年四月三 日 新竹煙波飯店

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

1. 2. 3. 4.

1

Optimization in ASAP Syntax $ITER How? GUI and Merit Function Two Case Study

• 感謝 BRO Research, RayTrain Optics 非常歡迎您可以將此投影片電子檔提供給需要的人員作為 非營利的研究參考用 但前提為請勿移除本電子檔的密碼 此外本講義內容目前撰寫光學教學書籍中 敬請您協助維 護合作單位相關的權益 謝謝您 C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

2

1

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

3

Last Year

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

4

2

太陽於天球上運行軌跡 冬至

春、秋分

北緯 0o (赤道)

夏至

冬至

春、秋分 夏至

北緯 45o

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

春、秋分

夏至

北緯 90o (北極)

5

Non Imaging Optimizations are still not easy

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

6

3

However 1. ASAP Provide More than 1000 INR Files for your optimization Applications – they are reliable and free 2. Translation between Existing INR File for Optimization Purpose is simple (but not necessary easy)

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

7

Optimization in ASAP Much Longer history than other software

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

8

4

ABERRATIONS MINIMIZE …

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

9

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

10

5

Rule1 : PIXELS + AVERAGE Improved the S/N Ratio

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

11

2nd Syntax of $ITER SVD – Singular Value Decomposition

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

12

6

Example : $ITER FOC 10 50 0.05 Æ (50-10)*0.05=2

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

13

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

14

7

/ (ITER 400, FOC*=60)

(ITER 400, FOC*=30)

/ (ITER 100, FOC*=60)

(ITER 100, FOC*=30)

15

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

You need to take care on the warnings messages!!

(ITER 100, FOC*=45)

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

16

8

NSTEPS=0

NSTEPS=1

NSTEPS=100

NSTEPS=400

/

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

17

Optimization GUI and Number of Variables Problems ASAP 2008

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

18

9

19

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

/ C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

20

10

/ C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

21

Rule2 : Using no more than 2 important variables for one optimization

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

22

11

核能研究所

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

23

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

24

12

GET (ASAP Command)

25

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

FILE ASAP2009_7.INR

V1=3.0

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

V1=4.57

26

13

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

27

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

28

14

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

29

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

30

15

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

31

Rule 3 : Using the Powerful Optical Data Base for MF, also test the MF in EXCEL

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

32

16

Summary

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

33

1. Calibration is required (SEED) 2. Building the Robust Merit Function 3. Number of Rays and S/N Ratio of Merit Function (Noise Level) 4. Number of Important Parameters 5. Brute Force is unacceptable!

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

34

17

1. SPOT & SPREAD 2. EXTREMES 3. GET Æ Solar Cells 4. DISPLAY 5. TABLE Æ Uniformity

6. $grab

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

35

LED/OLED Layer Case Started : C.R.Ou International Display Workshop, Japan, 2003

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

36

18

OLED發光效率

η ext = γ bal • γ st • q •η coupling γ bal

γ st q

稱為電荷平衡因子(charge balance factor)， 稱為singlet-triplet ratio,q 稱為發光效率(Luminescent Efficiency)， 稱為發光效率(Luminescent Efficiency)，

η coupling 稱為輸出耦合效率(optical output coupling efficiency)

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

37

輸出耦合效率展開

η coupling = η coupling ( nITO , norg , t ITO , torg )

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

38

19

模擬要點 • 本文使用光學軟體ASAP撰寫巨集來進行模擬

• 每一次的實驗產生的資料為2500個資料點 • 每一點若追跡光線10,000條，則於PIII-800約為2.6hr

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

39

實驗安排

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

40

20

分析策略 1. 實驗(E1,E2,E3)可以研究在折射率有變化的情 況下厚度變化導致的效應 2. 實驗(E6,E7,E8)可以探討厚度與折射率同時變 異的效果 3. 實驗(B0,B1,B2,B3)可以探討光線數目的效應 4. 實驗(E2,E8)可以探討 ITO 層折射率的效應 5. 實驗(E2,E7)可以探討ORG層折射率的效應

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

41

Results (1)

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

42

21

Results(2) • 由(E2/E7)發現當厚度有變化時,且nORG =1.7，nITO於 1.55~1.65之間的變化得到的光耦合效率差異會減少。 • 由(E2/E8)發現當厚度有變化時, nITO =1.6， nORG於 1.65~1.75之間的變化不會對於光耦合效率有顯著的 影響 • 由(E1,E2,E3)得知厚度變化對於OLED耦合效率在沒有 其他效率的考量下(如吸收或多重反射)沒有影響

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

43

Results(3) ASAP or EXCEL

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

44

22

For this case… • ITO層的折射率變化影響最小，ORG層的折射率影響 較大。 • 製程導致的厚度變異影響可以忽略 (不考慮色偏，多 重反射與吸收機制) • 發光耦合效率平均約為33%，文獻估計值27%應偏低 • ASAP協助進行二階分析與ANOVA計算

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

45

Reflector Case Started : C.R.Ou International Display LED Conference, Korea, 2007

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

46

23

SCENARIO 2 1. No particular USERAPOD 2. No scattering on surface

47

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

BENDING & FOCUSING

~ 40 %

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

EFL<2mm

48

24

BENDING & FOCUSING cont.

49

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

DIMENSIONS H* H * /W y = = W − L 1− L /W 1− x H = kH * (k = 2 for example)

s * ≡ Tanθ =

s = 2s

*

z≡

R W

0 ≤ L ≤ W ,0 ≤ H

y

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

50

25

DIMENSIONS cont. For a group of fixed R and H (s, z) can represent a set of reflector design within the fixed (R,H) space

W = R/z

H* y= W

R/H Ratio as third parameter

x = 1−

y s

0≤s≤∞ 0≤ z≤∞

L = xW

0 ≤ z ≤1 51

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

No Energy

No Energy

Θ=90°

Θ=0°

sÆ ∞

s= 0

No scattering

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

52

26

RESULTS Company A,B,C

Keep in mind : what the Merit Function is ?

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

53

More….FEM MODELS

Manufacturing issues : Statistics analysis on the LED

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

54

27

Conclusions You need BOTH good tool, and a lot of adaptive toolbox

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

55

1. Reduced Number of Important Parameters is the only way to success (no more than 2 at one time are suggest for reliable design) 2. Rays@Color = 5~10 times Rays@Illumination 3. Using ASAP Samples files – They are reliable! 4. Files are compatible and can work in ASCII Mode - 絕對不要小看這個問題

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

56

28

Thank you, Questions?

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

57

C.R.Ou : Hsiuping Institute of Technology, Taichung, Taiwan

58

29