Please use this identifier to cite or link to this item: https://doi.org/10.1016/S0010-4485(00)00138-X
Title: A hybrid method for recognition of undercut features from moulded parts
Authors: Ye, X.G. 
Fuh, J.Y.H. 
Lee, K.S. 
Keywords: Feature recognition
Mould design
Undercut feature
Issue Date: Dec-2001
Source: Ye, X.G., Fuh, J.Y.H., Lee, K.S. (2001-12). A hybrid method for recognition of undercut features from moulded parts. CAD Computer Aided Design 33 (14) : 1023-1034. ScholarBank@NUS Repository. https://doi.org/10.1016/S0010-4485(00)00138-X
Abstract: Products manufactured by moulding processes, such as injection moulding or die casting, are called moulded parts. This paper describes a hybrid method to recognise undercut features from moulded parts with planar, quadric and free-from surfaces. The hybrid method takes advantage of graph-based and hint-based approaches. Various undercut features, including interacting undercut features, are defined by extended attributed face-edge graphs (EAFEG). Unlike conventional graph-based methods, which recognise features by graph matching, the new approach recognises the undercut features by searching the cut-sets of subgraphs. Face properties and parting lines are used as hints to guide the search of cut-sets. To recognise undercut features from parts with free-form surfaces, a convex-hull algorithm is used to determine the face properties (positive, negative and horizontal). The case study shows that the proposed method can recognise various undercut features successfully. © 2001 Elsevier Science Ltd. All rights reserved.
Source Title: CAD Computer Aided Design
URI: http://scholarbank.nus.edu.sg/handle/10635/54272
ISSN: 00104485
DOI: 10.1016/S0010-4485(00)00138-X
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

64
checked on Dec 13, 2017

WEB OF SCIENCETM
Citations

50
checked on Nov 12, 2017

Page view(s)

28
checked on Dec 9, 2017

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.