Automatic part localization in a CNC machine coordinate system by means of 3D scans. Issue 5 (November 2015)

• Record Type: Journal Article
• Authors: Srinivasan, Harshad; Harrysson, Ola; Wysk, Richard
• Is Part Of: International journal of advanced manufacturing technology. Volume 81:Issue 5(2015)
• Page Start: 1127
• Page End: 1138
• Publication Date: 2015-11
• Subjects: 3D Scanning -- Finish machining -- Localization -- Registration -- Fiducial feature; Manufacturing processes -- Periodicals; Production engineering -- Periodicals; 670.427
• DOI: 10.1007/s00170-015-7178-z
• Languages: English
• ISSNs: 0268-3768

Abstract

Abstract Due to the rough nature of surfaces produced by current additive metal processes, a deviation between the nominal and actual pose of the part arises when fixturing in an NC machine. Modern 3D scanning systems are capable of generating high density measurements that may be used to localize the part and compensate for these errors; but require that scan data be transformed from the scanner to the CNC coordinate systems. In this work, precisely located fiducial features are automatically detected in the scan data and used to compute the pose of the scanner. This information is used to register each scan to the CNC machine space, and thereby create a model of the workpiece as-built and as-mounted in the machine, for use in process planning and localization. An implementation of the system was built and tested against a machined part that was offset to simulate uncertain position and orientation. The initial performance evaluation of the system indicates that it is capable of successfully registering surfaces to the machine coordinate system with accuracy in line with the scanner performance specification.