The Recognition of Machining Features in Turned Components
The work reported in this thesis describes a computer assisted system for recognising machining features in turned components with a view to automating the operation planning activity. The system uses object-oriented and artificial intelligence techniques for recognising complex turned components ca...
Saved in:
Main Author: | |
---|---|
Format: | Dissertation |
Language: | English |
Published: |
ProQuest Dissertations & Theses
01-01-1996
|
Subjects: | |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | The work reported in this thesis describes a computer assisted system for recognising machining features in turned components with a view to automating the operation planning activity. The system uses object-oriented and artificial intelligence techniques for recognising complex turned components capable of being machined on mill-turn centres.The modules implemented are part of TECHTURN which is a technologically-oriented computer-aided process planning system for turned parts. Two main modules have been developed: a feature-based input module and a feature recognition module. The feature recognition module, in turn, consists of a face detector, a feature volume builder and an expert system for recognising manufacturing features.Firstly, a 2D outline profile of the turned part is input and rotationally swept by 360° degrees about the Z-axis to generate the 3D B_Rep solid model representation of the workpiece. A predefined feature library allows both axi- and non- axisymmetric features to be added to the part model. The features available in the library include grooves, keyways, cross holes and flats. Features can also be added by using conventional Boolean operations. A graphical user interface has been implemented to support the functions of the input module and to facilitate the visualisation of the component on the screen. Work has also been undertaken to provide TECHTURN with a loop-based data structure capable of representing 3D solids in B_Rep form. In addition to that, the data structure is able to represent 2D parts in wireframe form.Since there is always a need for feature recognition, a novel 3D method referred to as, the equilibrium method, is used for recognising the non-axisymmetric depressions, protrusions and flats. A 2D version of this method is used to recognise axisymmetric grooves/recesses in the component. The equilibrium method is general in nature and can, therefore, be applied to any turned component represented in B_Rep form. The recognition procedure takes into account the existence of the blank, be it a bar, casting or forging. Research has also been carried out to expand the equilibrium method to cater for feature-feature interactions. A classification of the different types of interaction is provided.In the second phase of the feature recognition module, the corresponding feature volume is determined. Finally, in the third phase, an expert system module gives a manufacturing meaning to the volume by giving the feature a name and associating it with a suitable machining process.Examples of realistic industrial components are provided throughout, along with a discussion about the limitations and the benefits of the system. These points are finally emphasised in the conclusions and suggestions for future work are drawn. |
---|---|
ISBN: | 9798426844490 |