隨著電子產品朝向輕量化與高性能發展，IC 晶片亦趨於微小化，使IC
封裝設計面臨愈發複雜的挑戰。在封裝設計階段，模流分析可作為評估設
計與優化製程的重要工具，而在進行模流分析前，必須先將CAD 模型轉
換為實體網格，並利用求解器(Solver)進行計算。六面體網格(Hexahedron)
因具備良好的計算穩定性與精確性，被視為品質最佳的實體網格，但其自
動建構難度高，多依靠人工處理。由於六面體網格可由四邊形表面網格構
成，因此發展四邊形表面網格的自動建構技術，已成為未來重要的研究方
向。本研究旨在發展混合結構化與非結構化的四邊形網格自動建構方法，
透過將IC CAD 模型劃分為多個區域，依區域形狀與需求自動建構適當的
網格尺寸與類型，以提升模流分析的精確度與效率。為強化方法穩定性與
建構品質，本研究針對自動區域劃分流程進行優化，特別針對不合理的狹
長區域進行分類與合併處理，避免產生長寬比過大的低品質網格。同時，
導入高效率的Half-edge 資料結構，取代原先節點為主的儲存方式，提升
非結構化網格的建構效率。由結果顯示，原始方法及全非結構化網格各有
優缺點，而本方法能在整體網格品質與模流分析效率上取得較佳平衡，證
實本研究提出之優化策略具備良好的可行性。

As electronic products pursue lighter and higher-performance designs, IC
chips are becoming more compact, increasing the complexity of IC packaging. In
the design phase, mold flow analysis is essential for evaluating and optimizing the
packaging process. This requires converting CAD models into solid meshes
before simulation. Among mesh types, hexahedral meshes are known for their
high stability and accuracy, but they are difficult to construct automatically and
often rely on manual work. Since hexahedral meshes can be built from
quadrilateral surface meshes, developing automated quadrilateral meshing
techniques is a key research direction. This study proposes an automatic method
for generating hybrid structured and unstructured quadrilateral meshes. The IC
CAD model is divided into multiple regions, and mesh types and sizes are
determined automatically based on regional geometry and analysis needs, aiming
to improve simulation accuracy and meshing efficiency. To enhance the quality of
meshing, the region partitioning process is optimized by identifying and merging
elongated regions, which typically lead to low-quality elements with large aspect
ratios. Additionally, a Half-edge data structure is introduced to replace the
traditional node-based storage, significantly improving the construction
efficiency of unstructured meshes. Simulation results show that while existing
methods have their advantages, the proposed hybrid approach offers a better tradeoff
between mesh quality and computational performance, demonstrating the
practicality and effectiveness of this method.

[1] H. J. Fogg, C. G. Armstrong and T. T. Robinson, “Automatic generation of multiblock decompositions of surfaces.”, International Journal for Numerical Methods in Engineering, Vol. 101, No. 13, pp. 965-991, 2015
[2] H. J. Fogg, C. G. Armstrong and T. T. Robinson, “Multi-block decomposition using cross-fields.”, Proceedings of Adaptive Modelling and Simulation, Lisbon, pp. 254-267, 2013.
[3] J. Jezdimirovic, A. Chemin and J. F. Remacle, “Multi-block decomposition and meshing of 2D domain using Ginzburg-Landau PDE.”, Proceedings of the 28th International Meshing Roundtable, pp. 402-418, 2019.
[4] N. Kowalski, F. Ledoux and P. Frey, “A PDE based approach to multidomain partitioning and quadrilateral meshing.”, Proceedings of the 21st international meshing roundtable, pp. 137-154, 2013.
[5] H. J. Fogg, C. G. Armstrong and T. T. Robinson, “Enhanced medial-axis-based block-structured meshing in 2-D.”, Computer-Aided Design, Vol. 72, pp. 87-101, 2016.
[6] L. Sun, C. G. Armstrong, T. T. Robinson and D. Papadimitrakis, “Quadrilateral multiblock decomposition via auxiliary subdivision.”, Journal of Computational Design and Engineering, Vol. 8, No. 3, pp. 871-893, 2021.
[7] Y. Zhang, Y. Jia, S. S. Wang and M. Altinakar, “Composite structured mesh generation with automatic domain decomposition in complex geometries.”, Engineering Applications of Computational Fluid Mechanics, Vol. 7, No. 1, pp. 90-102, 2013.
[8] A. A. Rushdi, S. A. Mitchell, A. H. Mahmoud, C. C. Bajaj and M. S. Ebeida, “All-quad meshing without cleanup.”, Computer-Aided Design, Vol. 85, pp. 83-98, 2017.
[9] Z. Ali, J. Tyacke, P. G. Tucker and S. Shahpar, “Block topology generation for structured multi-block meshing with hierarchical geometry handling.’’, Procedia Engineering, Vol. 163, pp. 212-224, 2016.
[10] J. Rantakokko, “Partitioning strategies for structured multiblock grids.”, Parallel Computing, Vol. 26, No. 12, pp. 1661-1680, 2000.
[11] 戴宇辰，「混合結構化與非結構化四邊形網格之自動化區域劃分方法發展」，國立中央大學碩士論文，2023.
[12] 吳奕翰，「IC CAD 模型之混合結構化與非結構化四邊形網格自動建構技術發展」，國立中央大學碩士論文，2024.
[13] J. Y. Lai, P. Putrayudanto, Q. Wang, J. Huang, P. P. Song and Y. C. Tsai, “Development of automatic quadratic meshing technique with unequal spreading intervals on outer and inner contours for IC CAD models.”, System Innovation for an Artificial Intelligence Era, CRC Press, pp. 11-18, 2024.
[14] J. Y. Lai, J. S. Su, S. C. Tseng, H. J. Liou, S. J. Jheng, J. Huang, P. P. Song and Y. C. Tsai, “Quality Improvement for an Automatic Quadratic Mesh Generation Method for IC CAD Models.”, 2024 International Conference on Machining, Materials and Mechanical Technologies, Phan Thiet, Viet Nam, Sep. 11-15, 2024.
[15] 陳定輝，「應用於非結構化四邊形網格建構之輪廓撒點與網格品質改善技術發展」，國立中央大學碩士論文，2023.
[16] 王齊，「四邊形網格自動建構之多尺寸輪廓撒點技術研究」，國立中央大學碩士論文，2024.
[17] J. Docampo-Sánchez and R. Haimes, “A regularization approach for automatic quad mesh generation.”, 28th International Meshing Roundtable, Zenodo, 2020.
[18] C. G. Armstrong, H. J. Fogg, C. M. Tierney and T. T. Robinson, “Common themes in multi-block structured quad/hex mesh generation.”, Procedia Engineering, Vol. 124, pp. 70-82, 2015.
[19] K. Weiler, “Edge-based data structures for solid modeling in curved-surface environments.”, IEEE Computer Graphics and Applications, Vol. 5, No. 1, pp. 21-40, Jan. 1985.
[20] Y. Zhang, X. Luo and J. Jia, “A compact face-based topological data structure for triangle mesh representation.”, Computer-Aided Design and Applications, Vol. 16, No. 3, pp. 539-557, 2019.
[21] S. Salinas-Fernández, N. Hitschfeld-Kahler and R. Carrasco, “GPolylla: Fully GPU-accelerated polygonal mesh generator.”, arXiv preprint, arXiv:2403.14723, 2024.
[22] Y. Mu, Q. Yang, J. Li and X. Meng, “Unstructured cut-cell generation for complex geological modeling.”, Electronics, Vol. 14, No. 2, p. 332, 2025.
[23] P. Block, G. Boller, C. DeWolf, J. Pauli and W. Kaufmann, “Spanning trees and half-edge mesh construction in the complete graphic statics method for frames.”, Proceedings of the IASS 2024 Symposium: Redefining the Art of Structural Design, Zurich, Switzerland, Aug. 26-30, 2024.
[24] H. Brönnimann, “Designing and implementing a general-purpose halfedge data structure.”, Algorithm Engineering: 5th International Workshop, Proceedings of the 5th International Workshop on Algorithm Engineering, pp. 51-66, Århus, Denmark, August 28-31, 2001.
[25] V. B. Sunil and S. S. Pande, “Automatic recognition of features from freeform surface CAD models.”, Computer-Aided Design, Vol. 40, No. 4, pp. 502-517, 2008.
[26] 蔡敬崙，「四邊形網格自動建構之網格品質改善研究」，國立中央大學碩士論文，2024.
[27] B-rep Data Structure. Avaliable: https://developer.rhino3d.com/guides/cpp/brep-data-structure. [Accessed Sep. 12, 2023].
[28] Rhinoceros. Avaliable: https://www.rhino3d.com. [Accessed Sep 12, 2023].
[29] OpenNURBS. Avaliable: https://www.rhino3d.com/tw/features/developer/opennurbs. [Accessed Sep. 12, 2023].
[30] 蘇敬勛，「表面網格自動建構之網格品質優化方法發展」，國立中央大學碩士論文，2025.
[31] 鄭憲杰，「應用於IC CAD模型之非結構化四邊形網格自動建構與測試模組研究」，國立中央大學碩士論文，2024.
[32] 劉弘堅，「四邊形網格自動建構之輪廓撒點方法與四邊形網格轉三角形網格之研究」，國立中央大學碩士論文，2024.