隨著 CAD 及 CAE 技術的迅速發展，在設計塑膠射出成型產業中，模
流分析軟體的使用已經非常普遍。而模流分析前需要將欲分析的 CAD 模型
網格化，提供實體網格資料給分析軟體計算，因此實體網格資料的品質優劣，
會直接影響分析結果的準確性及分析時間的長短。若使用者需要較精確的
分析分析結果，需要利用手動的方式，將模型進行切割並搭建結構化網格，
其過程非常繁瑣且複雜。為了輔助結構化網格的搭建，本實驗室開發「特徵
辨識」及「特徵分解」相關技術，辨識出薄殼元件 CAD 模型中的各種孔特
徵、面屬性及凸起特徵，並藉由這些特徵資料將 CAD 模型分解為數個可以
搭建結構化網格的規則區塊。本研究將接續先前開發的演算法，繼續發展凸
起特徵辨識及分解技術，以提高辨識及分解完整度，並期望將此技術應用於
更廣泛的 CAD 模型上。首先，為了能同時辨識出內部及外部凸起特徵，有
對演算法中的資料結構進行整理，以達成此目的。接續為開發對稱凸起特徵
辨識及分解技術。在辨識的部分，改善原先第一型對稱凸起特徵的演算法，
並接續開發第二型及第三型對稱凸起特徵辨識。接著在分解的部分，將產生
每一種對稱凸起特徵的主要特徵區塊，並建立其接合特徵區塊及中間層特
徵區塊，得到完整的對稱凸起特徵分解結果。

關鍵字：凸起特徵辨識，凸起特徵分解，對稱凸起特徵，結構化網格

With the rapid development of CAD and CAE technology, the use of mold
flow analysis software has become very common in the design of the plastic
injection molding industry. Before the mold flow analysis, the CAD model needs
to be converted into solid meshes. Then, the solid mesh data will be employed to
the solver for calculation. Therefore, the quality of the solid mesh data can affect
the accuracy of the analysis results and the efficiency of the computational time.
If users need more accurate analysis results, they need to manually separate a
model and build structured meshes, which is a very tedious and complicated
process. To assist the construction of structured meshes, our laboratory develops
the technology of "feature recognition" and "feature decomposition". Thus, we
can recognize hole features, face types, and protrusion features for thin shell CAD
model, and decompose the CAD model into several regular blocks by which
structured meshes can be built. This research will continue the previously
developed algorithm, continue to develop protrusion recognition and
decomposition technology to improve the recognition and decomposition
integrity and expect this technology can deal with more kinds of CAD models.
First, to recognize the inner and outer protrusion at the same time, the data
structure in the algorithm needs to be organized to achieve this purpose. Secondly,
this study is to develop the recognition and decomposition technology of
symmetric protrusion. In the recognition part, this study is to improve the original
algorithm of the first type of symmetric protrusion, and continue to develop the
second and third types of symmetric protrusion recognition. Then in the
decomposition part, the main feature blocks of each symmetric protrusions are
generated, and connect feature block and the mid-layer feature blocks are
established. Finally, the complete symmetric protrusion decomposition results are
obtained.

Keywords: Protrusion recognition, Protrusion decomposition, Symmetric
protrusion, Structured mesh

[1] Moldex3D BLM, Website: https://www.moldex3d.com/products/software
/moldex3d/blm/, Accessed 11 June 2022.
[2] Rhinoceros, Website: http://www.rhino3d.com, Accessed 11 June 2022.
[3] Y. Li, W. Wang, X. Liu and Y. Ma, “Definition and recognition of rib features in aircraft structural part,” International Journal of Computer Integrated Manufacturing, Vol. 27, No. 1, pp. 1-19, 2014.
[4] H. L. Lockett and M. D. Guenov, “Graph-based feature recognition for injection moulding based on a mid-surface approach,” Computer-Aided Design, Vol 37, No. 2, pp. 251-262, 2005.
[5] Y. Woo and S. H. Kim, “Protrusion recognition from solid model using orthogonal bounding factor,” Journal of Mechanical Science and Technology, Vol 28, No. 5, pp. 1759-1764, 2014.
[6] L. Sun, C. M. Tierney, C. G. Armstrong and T. T. Robinson. “Automatic decomposition of complex thin walled CAD models for hexahedral dominant meshing,” Procedia Engineering, Vol. 163, pp. 225-237, 2016.
[7] H. Zhu, Y. Shao, Y. Liu and J. Zhao “Automatic hierarchical mid-surface abstraction of thin-walled model based on rib decomposition,” Advances in Engineering Software, Vol. 97, pp. 60-71, 2016.
[8] J. Y. Lai, P. P. Song, A. S. Hsiao, Y. C. Tsai and C. H. Hsu, “Recognition and classification of protrusion features on thin-wall parts for mold flow analysis,” Engineering with Computers, pp.1-22, 2019.
[9] Y. Lu, R. Gadh and T. J. Tautges, “Feature based hex meshing methodology: feature recognition and volume decomposition,” Computer-Aided Design, Vol. 33, No. 3, pp. 221-232, 2001.
[10] B. Jüttler, M. Kapl, D. M. Nguyen, Q. Pan and M. Pauley, “Isogeometric segmentation: The case of contractible solids without non-convex edges,” Computer-Aided Design, Vol, 57, pp. 74-90, 2014.
[11] Y. Zhang, X. Luo and Y. Zhao, “An approach to the automatic recognition of boolean decomposition loops for swept volume decomposition,” 25th International Meshing Roundtable, Washington DC, U.S.A., pp. 26-30, 2016.
[12] H. Wu and S. Gao, “Automatic swept volume decomposition based on sweep directions extraction for hexahedral meshing,” Procedia Engineering, Vol. 82, pp. 136-148, 2014.
[13] B. C. Kim and D. Mun, “Feature-based simplification of boundary representation models using sequential iterative volume decomposition,” Computers & Graphics, Vol. 38, pp. 97-107, 2014
[14] D. R. White, S. Saigal and S. J. Owen, “CCSweep: automatic decomposition of multi-sweep volumes,” Engineering with computers, Vol. 20, No. 3, pp. 222-236, 2004.
[15] 宋培溥，「應用於模流分析之薄殼元件CAD模型特徵辨識與分解技術發展」，國立中央大學，博士論文，民國110年。
[16] 王培懿，「CAD模型混接面、虛擬環及孔洞辨識技術發展」，國立中央大學，碩士論文，民國111年。
[17] Pradiktio Putrayudanto, “Face types recognition for thin-shell cad models,” Master Thesis, National Central University, 2021.
[18] 黃羿中，「薄殼CAD模型之內外面類型辨識技術發展」，國立中央大學，碩士論文，民國111年。
[19] 方聲翔，「自動化分解薄殼CAD模型殼本體之增強演算法發展」，國立中央大學，碩士論文，民國111年。
[20] Moldex3D, Website: https://www.moldex3d.com/, Accessed 11 June 2022.
[21] 王明暄，「應用於模流分析之 CAD 模型特徵辨識與實體網格品 質提升之研究」，國立中央大學，博士論文，民國106年。
[22] 張耀中，「Boss特徵之結構化實體網格自動化建構技術發展」，國立中央大學，碩士論文，民國109年。