Learning and Reasoning Techniques for Automatic Feature Recognition from CAD Models
Abstract
During the product development, Automatic Feature Recognition (AFR) techniques are an important tool for achieving a true integration of design and manufacturing stages. In particular, AFR systems offer capabilities for the identification in Computer-Aided Design (CAD) models of high-level geometrical entities, features that are semantically significant for manufacturing operations. However, the recognition performances of most of the existing AFR systems are limited to the requirements of specific manufacturing applications. This paper presents a new AFR method that facilitates the deployment of such systems in different application domains. In particular, the method provides a formal reasoning mechanism that combines the advantages of inductive and deductive techniques for feature recognition from Boundary Representation (B-Rep) part models. The proposed AFR method is implemented within a prototype feature recognition system and its capabilities are verified on a benchmarking part.
Keywords: Feature recognition, CAD/CAM integration
By: E.B. Brousseau(a), S.S. Dimov(a), R.M. Setchi(b)
(a): The Manufacturing Engineering Centre, Cardiff University, Cardiff, CF24 3AA, UK
(b): Cardiff School of Engineering, Cardiff University, Cardiff, CF24 3AA, UK
login or register to download the paper.
Hi Emanual and colleagues,
I have two questions related to applications of Feature Recognition and CAD formats.
Question No.1: Related to surface and solid modeling as well as CAD formats
Feature recognition (FR) is commonly used for two main purposes: (1) to recognize the parametric features of the CAD models for re-building the design based on changing the parametric features; and (2) Combining with Group Technology coding and classification for CAPP (Computer Aided Process Planning).
Currently, there are many CAD/CAM/CNC packages applying FR in CAD and CAPP. The typical ones are FeatureWorks®, and embedded modules in UG NX for CAD/CAPP/CAM integration in which Machining Knowledge Database, Tools and Machine Information, and Machining Feature Recognition are used for CAPP, and finally outputting CNC programs.
These Software allows using of most commonly used CAD interfaces, including STEP, IGES®, SAT (ACIS®), VDAFS (VDA) and Parasolid® files in which both Boundary Representation (B-Rep) and Constructive Solid Geometry (CSG) schemes are used, especially today most of the CAD/CAM packages provide hybrid surface-solid modeling capabilities.
Your research methodology is based on the Boundary Representation (B-Rep) inside STEP format. The question is as follows:
“Do you think that your proposed algorithms can be applied to all Surface and Solid kernels, not only relying on the CAD format such as STEP?”
Question No.2: Related to the application areas, the first question is
Question 2.1: “ what is the area that your proposed algorithms can provide the most benefit: CAD or CAPP?”
Question 2.2: “ “For the coming works, are you planning to apply the proposed algorithms to specific applications?”“
Many thanks in advance to your reply
Le, Cardiff University, Wales, UK
Hello Le,
Thank you very much for your comments and interest in this research.
Question No1: "Do you think that your proposed algorithms can be applied to all Surface and Solid kernels, not only relying on the CAD format such as STEP"
The proposed method has been implemented with STEP but in theory it is not tied up to this format only. What is important is to be able to work with CAD data stored using the B-REP scheme. In particular, in this work it was important to determine how features could be described using B-REP entities. For this, the feature concept proposed by Sakurai and Gossard (c.f. reference 6) was adopted in order to define a feature as a single face or a set of contiguous faces.
Question 2.1: "What is the area your proposed algorithms can provide the most benefit: CAD or CAPP?"
This is a little bit more tricky to answer as the implementation was done in a CAPP scenario only. What I can say however is that since the top level of the taxonomy adopted in this study classifies features as either depressions or protrusions, it may be difficult in CAD to include always design features in one or the other of those classes.
Question 2.2: "For the coming works, are you planning to apply the proposed algorithms to specific applications?"
Not at the moment, but it would be very interesting to do so.
Emmanuel
Hi Emanual and Colleagues,
You may have this paper already. However, I found this is a good review about Manufacturing Feature Recognition for your and the ones who concern.
Especially, it focusses on the three of the major algorithmic approaches for feasture recognition:
(i) Graph-based algorithms,
(ii) Volumetric Decomposition Techniques, and
(iii) Hint-based geometric reasoning.
One of the author, Dr. Mike Pratt, was actively involved in the development of the STEP standard, as leader of the ISO TC184/SC4.
Please CLICK HERE to download and read the paper.
Le
Dear Le,
Thank you for this information. I came across this paper during my study. It is indeed a good review.
Emmanuel
Bonjour Emmanuel!
La Coupe Mondiale, c'est vraiment dommage!
Just saw the results on the BBC website.
Never mind. Try again in 2010!
However, something else you might like to try sooner.
How about Afify's RULES-6?
Best wishes.
DTP.