Recognition and classification of protrusion features on thin-wall parts for mold flow analysis

• Published in: Engineering with computers Vol. 37; no. 2; pp. 833 - 854
• Main Authors: Lai, Jiing-Yih, Song, Pei-Pu, Hsiao, An-Sheng, Tsai, Yao-Chen, Hsu, Chia-Hsiang
• Format: Journal Article
• Language: English
• Published: London Springer London 01.04.2021; Springer Nature B.V
• Subjects: Algorithms; CAE) and Design; Calculus of Variations and Optimal Control; Optimization; Classical Mechanics; Classification; Columns (structural); Computer Science; Computer-Aided Engineering (CAD; Control; Extrusion; Feature recognition; Injection molding; Math. Applications in Chemistry; Mathematical and Computational Engineering; Molds; Original Article; Systems Theory; Tubes; B-rep model; Thin-wall part; Feature recognition; Recognition of inner and outer faces; Protrusion recognition
• ISSN: 0177-0667; 1435-5663
• DOI: 10.1007/s00366-019-00859-1

Thin-wall plastic parts exist in many products and are frequently manufactured by injection molding. The inner surface of a thin-wall part has many functional and structural features, typically divided into depressions and protrusions. While recognition of protrusion features is important in mold flow analysis, this recognition is difficult owing to the complexity and variety of the designed shapes. The purpose of this study was to develop a method for the detection and classification of protrusion features on thin-wall plastic parts. In the proposed algorithm, the inner and outer faces of a part model were detected first. Auxiliary faces, including translational, wall, and bottom faces, were recognized next. With auxiliary faces available, protrusion faces can thus be recognized and grouped in accordance with their neighboring relationship. A feature classification algorithm was finally implemented to classify five types of protrusions, namely tubes, ribs, columns, polygon blocks, and irregular extrusions. A detailed description of the procedures in each step of the proposed algorithm is provided. Twelve CAD models and analysis results are also presented to demonstrate the feasibility of the proposed protrusion recognition algorithm.