臺灣博碩士論文加值系統

隨著科技發展，光纖收發器朝向小體積高功率的方向發展，對於系統設備商而言熱
管理的壓力不斷提高，因此提供簡化的數學模型將能有效的減少其開發成本。
本研究針對 QSFP(Quad Small form-Factor Pluggable, QSFP)的熱測試載具(Thermal
Test Vehicle, TTV)進行風洞實驗，再透過風洞實驗驗證運用 DELPHI(DEvelopment of
Libraries of PHysical models for an Integrated design environment)方法學概念生成的
CTM(Compact Thermal Model, CTM)，熱測試載具共 7 個熱源，總功率19.729W。研究
內容包括針對熱測試載具進行風洞實驗以及透過商業模擬軟體Simcenter FloTHERM建
立計算流體力學(Computational Fluid Dynamics, CFD)模型，再透過與實驗完成驗證的
CFD模型生成熱測試載具的CTM。CTM的生成包括設計熱阻網絡、使用田口法建立邊
界條件集以及運用MATLAB來進行CTM熱阻值配置的最佳化。
生成 CTM 後將進行誤差評估，首先探討 CTM 的邊界條件獨立性(Boundary
Condition Independence, BCI)，在所有代表不同環境的邊界條件集中CTM與DTM(Detail
Thermal Model)的誤差均小於 10%，說明 CTM 具有良好的適用性。接著將 CTM 輸入
FloTHERM，模擬風洞實驗的條件來和CFD模型進行誤差評估，結果發現熱源溫度最大
誤差來到13.55%，CTM與風洞實驗的熱源溫度最大誤差則來到10.52%，可見本研究生
成的CTM雖然有一定的準確性，但仍然有改進空間。未來可使用本研究提供的方法針
對實際的光纖收發器開發CTM，但若要應用於系統級的開發則還要進一步改善誤差。

With the advancement of technology, fiber optical transceivers are moving towards smaller
sizes and higher power, placing increasing thermal management pressure on system equipment
vendors. Therefore, providing simplified mathematical models can effectively reduce their
development costs.
This study focuses on conducting wind tunnel experiments on the Thermal Test Vehicle
(TTV) of the Quad Small form-Factor Pluggable (QSFP), and verifies the Compact Thermal
Model (CTM) generated using the concept of DELPHI(DEvelopment of Libraries of PHysical
models for an Integrated design environment) methodology. The TTV includes 7 heat sources
with total power of 19.729W. The research involves wind tunnel experiments on the TTV and
uses the commercial simulation software Simcenter FloTHERM to creat TTV’s Computational
Fluid Dynamics (CFD) models. With the CFD model validated by the experiment, we generate
a CTM for the TTV. The CTM generation process includes designing thermal resistance
networks, using the Taguchi method to generate boundary condition sets, and optimizing CTM
thermal resistance configurations using MATLAB.
After generating the CTM, error assessment is conducted. Boundary Condition
Independence (BCI) of the CTM is first discussed, errors between CTM and DTM(Detail
Thermal Model) are all less than 10% across all boundary condition sets representing different
environments, demonstrating the good applicability of the CTM. Subsequently, the CTM is
input into FloTHERM to simulate wind tunnel experiment conditions and evaluate errors
against CFD models. Results show maximum errors in heat sources temperatures are less than
13.55%, while the maximum error between CTM and wind tunnel experiment heat sources
temperatures is 10.52%. This indicates that while the CTM generated in this study has a certain
level of accuracy, there is still room for improvement.
The methods provided in this study can be applied to develop CTMs for actual optical
ii
transceivers in the future. For system-level development applications, the accuracy of CTM
waits further improvements.

目錄
摘要 .......................................................................................................................................................... i
Abstract ................................................................................................................................................... ii
目錄 ........................................................................................................................................................ iv
圖目錄 .................................................................................................................................................... vi
表目錄 .................................................................................................................................................. viii
符號表 ..................................................................................................................................................... x
1.1研究動機 ....................................................................................................................................... 1
1.2 文獻回顧 ...................................................................................................................................... 2
1.2.1 光纖收發器 ........................................................................................................................... 2
1.2.2 CTM ....................................................................................................................................... 5
1.3 研究目的 ...................................................................................................................................... 7
第二章、研究方法 ................................................................................................................................. 8
2.1 風洞實驗 ...................................................................................................................................... 9
2.1.1熱測試載具 ............................................................................................................................ 9
2.1.2風洞系統及實驗設備 .......................................................................................................... 10
2.1.3實驗設計 .............................................................................................................................. 14
2.1.4 實驗流程 ............................................................................................................................. 16
2.2 模擬軟體 .................................................................................................................................... 17
2.2.1 Simcenter FloTHERM .......................................................................................................... 17
2.3 CFD模型 .................................................................................................................................... 18
2.3.1 CAD圖檔簡化 ..................................................................................................................... 18
2.3.2 CFD模型模擬條件 ............................................................................................................. 20
2.3.3 CFD模型網格獨立性測試 ................................................................................................. 21
2.3.4 CFD模型接觸熱阻 ............................................................................................................. 23
2.4 DTM ............................................................................................................................................ 24
2.5 CTM ............................................................................................................................................ 25
2.5.1 選擇節點 ............................................................................................................................. 26

v

2.5.2 建構熱阻網絡 ..................................................................................................................... 30
2.5.3 邊界條件集 ......................................................................................................................... 32
2.5.4 目標函數與最佳化 ............................................................................................................. 37
2.5.5 CTM應用於FloTHERM .................................................................................................... 38
2.5.6 各模型網格數量比較 ......................................................................................................... 40
第三章、結果與討論 ............................................................................................................................ 42
3.1風洞實驗結果 .............................................................................................................................. 42
3.2 CFD模型結果 ............................................................................................................................. 44
3.3 CTM熱阻值 ................................................................................................................................ 45
3.4 訓練用邊界條件集模擬結果 ..................................................................................................... 47
3.5 測試用邊界條件集模擬結果 ..................................................................................................... 53
3.5.1 自然對流邊界條件集模擬結果 ......................................................................................... 54
3.5.2 強制對流邊界條件集模擬結果 ......................................................................................... 57
3.5.3 混合對流邊界條件集模擬結果 ......................................................................................... 62
3.6 CTM與CFD模型誤差評估 ...................................................................................................... 68
3.7 CTM與風洞實驗誤差評估 ........................................................................................................ 72
第四章、結論與未來展望 .................................................................................................................... 75
4.1結論 .............................................................................................................................................. 75
4.2 未來展望 ..................................................................................................................................... 77
參考文獻 ................................................................................................................................................ 78
附件 ........................................................................................................................................................ 81
A1 目標函數和能量方程式求解器MALTAB程式碼 .................................................................. 81
A1-1 目標函數： ......................................................................................................................... 81
A1-2 最佳化： ............................................................................................................................. 83
A1-3 輸出結果： ......................................................................................................................... 84

1. Lukasik, S., Why the Arpanet Was Built, IEEE Annals of the History of Computing, vol. 33, no. 3, pp.4-21, 2011.
2. Forouzan, B.A., Data Communications and Networking(4th ed.), McGraw Hill, 2012.
3. Chow, W. W., Koch, S.W., & Sargent, M., Semiconductor-laser physics, Springer Science & Business Media, 2012.
4. SFF Committee, Gigabit Interface Converter (GBIC), 1999.
5. SFF Committee, SFP (Small Form factor Pluggable) Transceiver, 2001.
6. Peng Z., Guiming H., Liwu Z., 1000BASE-T SFP, Proceedings of the 5th Electronics Packaging Technology Conference , IEEE, 2003.
7. SFF Committee, QSFP (Quad Small Formfactor Pluggable) Transceiver, 2006.
8. QSFP-DD MSA, QSFP-DD Specification for QSFP Double Density 8x pluggable transceiver, Rev1.0, 2016.
9. QSFP-DD MSA, QSFP-DD/QSFP-DD800/QSFP112 Hardware Specification for QSFP Double Density 8x and QSFP 4x pluggable transceivers, Rev6.01, 2021.
10. QSFP-DD MSA, QSFP-DD/QSFP-DD800/QSFP-DD1600 Hardware Specification for QSFP Double Density 8x pluggable transceivers, Rev7.0, 2023.
11. A. Romero and S. Kipp, Cooling 8×100GbE switch blades with high power optical modules, 13th InterSociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems, pp. 1327-1333, 2012.
12. Mack, B., and Graham, T., Thermal specifications for pluggable optics modules, 2016 32nd Thermal Measurement, Modeling & Management Symposium (SEMI-THERM). IEEE, 2016.
13. Samtec, Inc., Thermal Design of QSFP-DD Cages and Heatsinks For High Power High Density Applications, 2018.
14. Dogruoz, B., Giovanni, G., Nowell, M., Nering, R., Tsai, A., Aranyosi, A., Maki, J., Ali, H., Kapuscinski, C., Shah, V., Sommers S., Daou, F., Daou, H., Best, B., Cheng, N., Optimizing QSFP-DD Systems to Achieve at Least 25 Watt Thermal Port Performance, QSFP-DD MSA, 2021.
15. Chen A., Kocsis S., Ajersch P., Mack B., Giobbio G., Jacques J., Nowell M., Little T., Park W., Yang P., Aranyosi A., Le V., Maki J., Daou H., Farajalla S., Ali H., Sommers S., Kuzhikkali R., Sammon K., Enabling QSFP-DD1600 Ecosystem With Performance-Driven Thermal Innovations, QSFP-DD MSA, 2023.
16. Raghupathy A.P., Boundary-Condition-Independent Reduced-Order Modeling for Thermal Analysis of Complex Electronics Packages, University of Cincinnati, 2009.
17. JEDEC Standard JESD15-4, DELPHI Compact Thermal Model Guideline, 2008.
18. Bar-Cohen, A., Elperin, T., Eliasi, R., θjc Characterization of Chip Packages- Justification, Limitations, and Future, IEEE Transactions on components, hybrids, and manufacturing technology, vol. 12, pp.724-731, 1989.
19. Shidore, S., & Lee, T. Y. T., A comparative study of the performance of compact model topologies and their implementation in CFD for a plastic ball grid array package, J. Electron. Packag., vol. 123, no. 13, pp. 232-237, 2001.
20. JEDEC Standard JESD15-3, Two-Resistor Compact Thermal Model Guideline, 2008.
21. Lasance C. J. M., Vinke H., Rosten H., Thermal Characterization of Electronic Devices with Boundary Condition Independent Compact Models, IEEE Transactions on Components, Packaging, and Manufacturing Technology-Part A, vol. 18, no. 4, pp. 723-731, 1995.
22. Aranyosi, A., Ortega, A., Evans, J., Tarter, T., Pursel, J., & Radhakrishnan, J., Development of compact thermal models for advanced electronic packaging: Methodology and experimental validation for a single-chip CPGA package, ITHERM 2000. The Seventh Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (Cat. No. 00CH37069) (Vol. 1, pp. 225-232). IEEE, 2000.
23. 陳霈祺，光纖收發器緊湊熱模型，碩士論文，國立中央大學，民國112年。
24. 陳彥瑋，運用田口法於光纖收發器之散熱分析，碩士論文，國立中央大學，民國111年。
25. Murshed, S. M. S., Introductory Chapter: Electronics Cooling—An Overview, Electronics Cooling, pp.1-11, 2016.