Automatic Spot Welding Feature Recognition From STEP Data

Automatic feature recognition is the process of extracting significant information out of a CAD model without any human intervention. It is an important task between CAD and CAPP, which itself is a bridge between CAD and CAM. Automated feature recognition is therefore a pivotal component in the integration of CAD/CAPP/CAM. Research in feature recognition has generated substantial interest; however, to date the majority of feature recognition techniques have focused on extracting machining features. This work is an attempt to develop a framework for automatic feature recognition of spot welding from a neutral file STEP (AP 203). This study explains the software development in Python language for feature recognition that works as an intermediary between CAD and CAPP. A three-dimensional assembly with multiple spot welds was created in Creo 3.0. The three main features of spot welding are: (1) Total number of spot welds; (2) Coordinates and diameter of each spot weld; and (3) The plane of weld. Rule-based technique is employed for recognizing these features. The computer program has been implemented and successfully tested on several components. An illustrative case study is presented demonstrating the validity of the method.