近年來市場的需求，造成電子零件往輕薄與多功能化演進，其中扇出型面板級封裝(Fan-Out Panel Level Packaging, FO-PLP)是能符合市場需求的選項之一。面板級封裝在製程中會有溫度的升降溫，因而使封裝產生翹曲現象。其中列出兩種產生翹曲的原因，一種是面板級封裝結構是由多種材料組成，而不同材料間的熱膨脹係數不同，製程中升降溫會導致膨脹收縮程度不同而造成翹曲;另一種是封裝環氧樹脂(Epoxy Molding Compound)在升溫製程中會發生固化反應導致體積收縮產生翹曲。若是在此階段產生之翹曲量過大則會造成後續製程的困難，故本研究將會專注於使用有限元素法(Finite Element Method)預測翹曲值。
本研究使用有限元素法(Finite Element Method, FEM)來進行建立三維模型，並且在環氧模壓樹脂材料使用經實驗驗證過之等效熱膨脹係數(Equivalent CTE)，此舉為簡化該材料在固化反應產生的體積收縮等化學效應，最後成功建立扇出型面板級封裝之模型。
FEM相較於實驗雖然可以節省大量的時間與成本，但常因人為疏失導致結果不同，並且在改變結構尺寸下也必須重新建立模型。因此本研究為了減少上述情況，藉由極度隨機樹(Extremely Randomized Trees) 集成學習演算法來預測不同封裝尺度下的翹曲量，該演算法特點為對於大規模數據具有運算快速的特點。首先會透過有限單元法建立出不同幾何尺寸之扇出型面板級封裝模型的訓練數據庫，接著機器學習透過這些訓練數據庫來學習，機器學習在完成學習後即可快速的預測出隨便一個幾何尺寸面板級封裝的翹曲值。
然而從眾多的文獻中可以發現扇出型面板級封裝的翹曲是一個非對稱的形狀，影響此結果可能有眾多原因，例如，封裝環氧樹脂厚度不均勻，封裝環氧樹脂厚度不同會有不同的收縮率，當有一側收縮率不同即會影響到其他地方的變形行為，因而有機率產生不對稱的翹曲，另外在模壓成型(Compression Molding)過程中，封裝環氧樹脂是透過接觸來進行加熱固化，而封裝環氧樹脂在固化過程當中會產生體積收縮，進而產生翹曲並影響接觸面積，因此加熱溫度會有所差異，進而造成封裝環氧樹脂的固化反應不相同。
因此本研究將深入探討造成非對稱翹曲的原因，例如製程的影響等。同時透過有限元素法建立三維模型，來去模擬製程差異產生的非對稱翹曲。

In recent years, the demand of the market has resulted in the evolution of thin and multi-functional electronic components. Fan-Out Panel Level Packaging (FO-PLP) is one of the options that can meet the market demand. Panel-level packaging will have temperature rise and fall during the process, which will cause warpage. There are two reasons for warpage. One is that the panel-level packaging structure is composed of a variety of materials, and the coefficients of thermal expansion of different materials are different. The temperature rise and fall during the process will lead to different degrees of expansion and contraction, resulting in warpage; the other is Epoxy Molding Compound will undergo a curing reaction during the heating process, resulting in volume shrinkage and warpage. If the amount of warpage generated at this stage is too large, it will cause difficulties in the subsequent process, so this research will focus on using the Finite Element Method to predict the warpage value.
In this study, the finite element method was used to build a 3D model, and the experimentally verified Equivalent CTE was used in the epoxy molding compound. This is done to simplify chemical effects such as volume shrinkage of the material during the curing reaction. Finally, the model of the fan-out panel-level package was successfully established.
Although FEM can save a lot of time and cost compared with experiments, the results are often different due to human error, and the model must be re-established when the structure size is changed. Therefore, in order to reduce the above situation, this study uses the Extremely Randomized Trees ensemble learning algorithm to predict the amount of warpage at different packaging scales. The algorithm is characterized by its fast operation on large-scale data. First, a training database for fan-out panel-level packaging models of different geometric sizes will be established through the finite element method, and then machine learning will learn through these training databases. Machine learning can quickly predict the warpage value of a panel-level package of any geometric size after completing the learning.
However, from many literatures, it can be found that the warpage of the fan-out panel level package is an asymmetric shape. There may be many reasons for affecting this result. For example, the thickness of the package epoxy resin is not uniform, and the thickness of the package epoxy resin will be different. There are different shrinkage rates. When the shrinkage rate of one side is different, it will affect the deformation behavior of other places, so there is a chance to produce asymmetric warpage. In addition, in the process of compression molding, the encapsulation epoxy resin is passed through the contact During the curing process, the encapsulated epoxy resin will shrink in volume, which will cause warpage and affect the contact area. Therefore, the heating temperature will be different, and the curing reaction of the encapsulated epoxy resin will be different.
Therefore, this study further explores the causes of asymmetric warpage, such as the influence of the process. At the same time, a three-dimensional model is established by the finite element method to simulate asymmetric warpage caused by process differences.

摘要 …………………………………………………………………I
Abstract ……………………………………………………………….III
目錄 ……………………………………………………………….VI
圖目錄………………………………………………………………..IX
表目錄 ……………………………………………………………..XIV
第一章 緒論 1
1.1 簡介 1
1.2 文獻回顧 2
1.3 扇出型晶圓級與面板級封裝製程 15
1.4 研究動機與目標 17
第二章 基礎理論 19
2.1 有限元素法理論 19
2.1.1 有限元素法之線彈性理論 19
2.2 有限元素接觸理論 23
2.2.1 懲罰函數法 24
2.2.2 拉格朗日乘子法 25
2.2.3 增廣拉格朗日乘子法 25
2.3 翹曲現象 26
2.3.1 熱膨脹係數不匹配 26
2.3.2 固化反應 27
2.4 P-V-T-C方程式 27
2.5 等效熱膨脹係數法 29
2.6 模壓成型 33
2.6.1 環氧模壓樹脂厚度不均勻 33
2.6.2 環氧模壓樹脂固化程度不均勻 33
2.7 機器學習 33
2.7.1 人工神經網路 35
2.7.2 梯度下降法 37
2.7.3 學習速率常數 39
2.7.4 反向誤差傳播法 42
2.7.5 激活函數 44
2.7.6 CART決策樹 46
2.7.7 集成學習 48
2.7.8 極度隨機樹 48
2.8 機器學習優化和評估 49
2.8.1 資料前處理 49
2.8.2 交叉驗證 51
第三章 有限元素模型建立 53
3.1 扇出型面板級封裝模型 53
3.2 扇出型面板級封裝模型尺寸與材料系數 54
3.3 扇出型面板級模型邊界條件與負載設定 56
3.4 扇出型面板級翹曲量分析 56
3.5 載板脫離(Debonding)翹曲量分析 57
第四章 機器學習 60
4.1 扇出型面板級模型數據庫之特徵選取 62
4.2 扇出型面板級模型數據庫之建立 63
4.3 極度隨機樹回歸模型之超參數 65
4.4 極度隨機樹回歸模型之資料前處理比較 69
4.5 極度隨機樹回歸模型之超參數設定與預測結果 71
4.6 預測結果與比較 74
第五章 非對稱翹曲 78
5.1 扇出型面板級封裝模型之尺寸與材料參數設定 79
5.2 環氧模壓樹脂厚度與固化程度不均勻之模擬設定與探討 81
5.3 結果 93
第六章 結論與建議未來工作 97
參考文獻 99

[1] C.-Y. Chou, T.-Y. Hung, S.-Y. Yang, M.-C. Yew, W.-K. Yang, and K.-N. Chiang, "Solder joint and trace line failure simulation and experimental validation of fan-out type wafer level packaging subjected to drop impact," Microelectronics Reliability, vol. 48, no. 8-9, pp. 1149-1154, 2008.
[2] M. C. Yew, M. Tsai, D. C. Hu, W. K. Yang, and K. N. Chiang, "Reliability analysis of a novel fan‐out type WLP," Soldering & surface mount technology, vol. 21, no. 3, pp. 30-38, 2009.
[3] P. Wang, Y. Lee, C. Lee, H. Chang, and K. Chiang, "Solder joint reliability assessment and pad size studies of FO-WLP with glass substrate," IEEE Transactions on Device and Materials Reliability, vol. 21, no. 1, pp. 96-101, 2021.
[4] H.-W. Liu et al., "Warpage characterization of panel fan-out (P-FO) package," in 2014 IEEE 64th Electronic Components and Technology Conference (ECTC), 2014: IEEE, pp. 1750-1754.
[5] H. J. Borchardt and F. Daniels, "The application of differential thermal analysis to the study of reaction kinetics1," Journal of the American Chemical Society, vol. 79, no. 1, pp. 41-46, 1957.
[6] K. E. J. Barrett, "Determination of rates of thermal decomposition of polymerization initiators with a differential scanning calorimeter," Journal of Applied Polymer Science, vol. 11, no. 9, pp. 1617-1626, 1967.
[7] K. Horie, H. Hiura, M. Sawada, I. Mita, and H. Kambe, "Calorimetric investigation of polymerization reactions. III. Curing reaction of epoxides with amines," Journal of Polymer Science Part A‐1: Polymer Chemistry, vol. 8, no. 6, pp. 1357-1372, 1970.
[8] M. Kamal and S. Sourour, "Kinetics and thermal characterization of thermoset cure," Polymer Engineering & Science, vol. 13, no. 1, pp. 59-64, 1973.
[9] M. R. Kamal and M. Ryan, "The behavior of thermosetting compounds in injection molding cavities," Polymer Engineering & Science, vol. 20, no. 13, pp. 859-867, 1980.
[10] S.-J. Hwang and Y.-S. Chang, "PVTC equation for epoxy molding compound," IEEE Transactions on components and packaging technologies, vol. 29, no. 1, pp. 112-117, 2006.
[11] S.-S. Deng, S.-J. Hwang, H.-H. Lee, D.-Y. Huang, Y.-R. Chen, and G.-S. Shen, "Simulation and experiments of fan-out wafer level package during encapsulation," in 2009 4th International Microsystems, Packaging, Assembly and Circuits Technology Conference, 2009: IEEE, pp. 48-51.
[12] K. Oota and M. Saka, "Cure shrinkage analysis of epoxy molding compound," Polymer Engineering & Science, vol. 41, no. 8, pp. 1373-1379, 2001.
[13] J. W. Kong, J.-K. Kim, and M. M. Yuen, "Warpage in plastic packages: Effects of process conditions, geometry and materials," IEEE transactions on electronics packaging manufacturing, vol. 26, no. 3, pp. 245-252, 2003.
[14] T. Lin et al., "Warpage simulation and experiment for panel level fan-out package," in 2016 IEEE CPMT Symposium Japan (ICSJ), 2016: IEEE, pp. 129-131.
[15] F. Hou et al., "Experimental verification and optimization analysis of warpage for panel-level fan-out package," IEEE Transactions on Components, Packaging and Manufacturing Technology, vol. 7, no. 10, pp. 1721-1728, 2017.
[16] M. Su et al., "Warpage simulation and experimental verification for 320 mm× 320 mm panel level fan-out packaging based on die-first process," Microelectronics Reliability, vol. 83, pp. 29-38, 2018.
[17] D. V. Sorono, J. Lin, C. T. Chong, S. C. Chong, and S. R. Vempati, "Study on mold flow during compression molding for embedded wafer level package (EMWLP) with multiple chips," in 2012 IEEE 14th Electronics Packaging Technology Conference (EPTC), 2012: IEEE, pp. 336-341.
[18] J. McCarthy, M. L. Minsky, N. Rochester, and C. E. Shannon, "A proposal for the dartmouth summer research project on artificial intelligence, august 31, 1955," AI magazine, vol. 27, no. 4, pp. 12-12, 2006.
[19] S. Aich, K. Younga, K. L. Hui, A. A. Al-Absi, and M. Sain, "A nonlinear decision tree based classification approach to predict the Parkinson's disease using different feature sets of voice data," in 2018 20th International Conference on Advanced Communication Technology (ICACT), 2018: IEEE, pp. 638-642.
[20] D. Wu, C. Jennings, J. Terpenny, R. X. Gao, and S. Kumara, "A comparative study on machine learning algorithms for smart manufacturing: tool wear prediction using random forests," Journal of Manufacturing Science and Engineering, vol. 139, no. 7, 2017.
[21] T. Mikolov, M. Karafiát, L. Burget, J. Cernocký, and S. Khudanpur, "Recurrent neural network based language model," in Interspeech, 2010, vol. 2, no. 3: Makuhari, pp. 1045-1048.
[22] R. A. Fisher, "The use of multiple measurements in taxonomic problems," Annals of eugenics, vol. 7, no. 2, pp. 179-188, 1936.
[23] D. R. Cox, "The regression analysis of binary sequences," Journal of the Royal Statistical Society: Series B (Methodological), vol. 20, no. 2, pp. 215-232, 1958.
[24] F. Rosenblatt, "The perceptron: a probabilistic model for information storage and organization in the brain," Psychological review, vol. 65, no. 6, p. 386, 1958.
[25] T. Cover and P. Hart, "Nearest neighbor pattern classification," IEEE transactions on information theory, vol. 13, no. 1, pp. 21-27, 1967.
[26] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, "Learning representations by back-propagating errors," nature, vol. 323, no. 6088, pp. 533-536, 1986.
[27] J. R. Quinlan, "Induction of decision trees," Machine learning, vol. 1, no. 1, pp. 81-106, 1986.
[28] T. K. Ho, "Random decision forests," in Proceedings of 3rd international conference on document analysis and recognition, 1995, vol. 1: IEEE, pp. 278-282.
[29] P. Geurts, D. Ernst, and L. Wehenkel, "Extremely randomized trees," Machine learning, vol. 63, no. 1, pp. 3-42, 2006.
[30] W. S. McCulloch and W. Pitts, "A logical calculus of the ideas immanent in nervous activity," The bulletin of mathematical biophysics, vol. 5, no. 4, pp. 115-133, 1943.
[31] D. H. Hubel and T. N. Wiesel, "Receptive fields, binocular interaction and functional architecture in the cat's visual cortex," The Journal of physiology, vol. 160, no. 1, p. 106, 1962.
[32] K. Fukushima and S. Miyake, "Neocognitron: A self-organizing neural network model for a mechanism of visual pattern recognition," in Competition and cooperation in neural nets: Springer, 1982, pp. 267-285.
[33] A. Waibel, T. Hanazawa, G. Hinton, K. Shikano, and K. J. Lang, "Phoneme recognition using time-delay neural networks," IEEE transactions on acoustics, speech, and signal processing, vol. 37, no. 3, pp. 328-339, 1989.
[34] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, 1998.
[35] A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," Advances in neural information processing systems, vol. 25, 2012.
[36] B. E. Boser, I. M. Guyon, and V. N. Vapnik, "A training algorithm for optimal margin classifiers," in Proceedings of the fifth annual workshop on Computational learning theory, 1992, pp. 144-152.
[37] C. Cortes and V. Vapnik, "Support-vector networks," Machine learning, vol. 20, no. 3, pp. 273-297, 1995.
[38] P. Chou, K. Chiang, and S. Y. Liang, "Reliability assessment of wafer level package using artificial neural network regression model," Journal of Mechanics, vol. 35, no. 6, pp. 829-837, 2019.
[39] H. Kuo, B. Lai, and K. Chiang, "Study on Reliability Assessment of Wafer Level Package Using Design-on-Simulation with Support Vector Regression Techniques," in 2020 15th International Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), 2020: IEEE, pp. 192-195.
[40] H. Hsiao and K. Chiang, "AI-assisted reliability life prediction model for wafer-level packaging using the random forest method," Journal of Mechanics, vol. 37, pp. 28-36, 2021.
[41] Q.-H. Su and K.-N. Chiang, "Predicting Wafer-Level Package Reliability Life Using Mixed Supervised and Unsupervised Machine Learning Algorithms," Materials, vol. 15, no. 11, p. 3897, 2022.
[42] T. Wang, Y. Chen, M. Qiao, and H. Snoussi, "A fast and robust convolutional neural network-based defect detection model in product quality control," The International Journal of Advanced Manufacturing Technology, vol. 94, no. 9, pp. 3465-3471, 2018.
[43] F. Yin, H. Mao, L. Hua, W. Guo, and M. Shu, "Back propagation neural network modeling for warpage prediction and optimization of plastic products during injection molding," Materials & design, vol. 32, no. 4, pp. 1844-1850, 2011.
[44] S. Taghizadeh, A. Ozdemir, and O. Uluer, "Warpage prediction in plastic injection molded part using artificial neural network," 2013.
[45] C. Selvanayagam, P. L. T. Duong, R. Mandal, and N. Raghavan, "Machine learning approach to improve accuracy of warpage simulations," in 2019 IEEE 69th Electronic Components and Technology Conference (ECTC), 2019: IEEE, pp. 834-841.
[46] S. Liu, S. Panigrahy, and K. Chiang, "Prediction of fan-out panel level warpage using neural network model with edge detection enhancement," in 2020 IEEE 70th Electronic Components and Technology Conference (ECTC), 2020: IEEE, pp. 1626-1631.
[47] T. Braun et al., "Material and process trends for moving from FOWLP to FOPLP," in 2015 IEEE 17th Electronics Packaging and Technology Conference (EPTC), 2015: IEEE, pp. 1-6.
[48] K.-J. Bathe, Finite element procedures. Klaus-Jurgen Bathe, 2006.
[49] C. Tsai, S. Liu, and K. Chiang, "Warpage analysis of fan-out panel-level packaging using equivalent CTE," IEEE Transactions on Device and Materials Reliability, vol. 20, no. 1, pp. 51-57, 2019.
[50] M. Sadeghinia, K. Jansen, and L. Ernst, "Characterization and modeling the thermo-mechanical cure-dependent properties of epoxy molding compound," International Journal of Adhesion and Adhesives, vol. 32, pp. 82-88, 2012.
[51] H. Cheng, K. Chiang, and M. Lee, "An effective approach for three-dimensional finite element analysis of ball grid array typed packages," 1998.
[52] H. Chang, B. Chen, and K. Chiang, "The effect of data distribution in Ensemble Learning Algorithms on WLCSP reliability Prediction," in 2021 16th International Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), 2021: IEEE, pp. 60-63.
[53] B. Wang, G. Huang, and K. Chiang, "PSO-based Modified Convolution Neural Network on Fan-Out Panel Level Package Prediction," in 2021 16th International Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), 2021: IEEE, pp. 56-59.
[54] G. Yang et al., "A Quantitative Model to Understand the Effect of Gravity on the Warpage of Fan-Out Panel-Level Packaging," IEEE Transactions on Components, Packaging and Manufacturing Technology, vol. 11, no. 11, pp. 2022-2030, 2021.
[55] D. Ecoiffier, P. Vernhes, R. Ooi, T. Braun, M. Dreissigacker, and H. Fu, "Experimental Study of Panel Level Packaging Warpage," in 2020 IEEE 8th Electronics System-Integration Technology Conference (ESTC), 2020: IEEE, pp. 1-4.
[56] C.-C. Lee et al., "Comprehensive Investigation on Warpage Management of FOPLP with Multi Embedded Ring Designs," in 2019 IEEE 69th Electronic Components and Technology Conference (ECTC), 2019: IEEE, pp. 1413-1418.