Optical Sensing of 3-D Contours for Online Control in Incremental Profile Forming
Loading...
Date
2019-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Ohio State University
Abstract
Metal forming processes are facing demands for increased accuracy and geometric complexity of manufactured parts, even in small-lot production. However, metal forming processes are inherently subject to uncertainties in raw material characteristics as well as errors in the models of material behavior and of the manufacturing processes used for process planning. The impact of these uncertainties and model errors can be reduced through implementation of closed loop control of metal forming processes, based on online sensing of process variables of importance.
One of metal forming processes that benefit from such closed loop control is the incremental profile forming (IPF) process for tubular structures, recently developed at the Technical University of Dortmund, Germany. The desired geometry of the tubular cross-section is generated by indenters placed around the circumference, and commanded to execute the desired motions. Control of the indenter motions is not adequate to ensure the desired cross-sectional geometry after completion of the forming operation, since the latter depends also on the recovery of elastic deformation across the tubular structure following the removal of the forming load. Control of machine motions controls only the local geometry under the forming load.
For IPF process closed loop control, on-line sensing of profile geometry is therefore essential to improve the final part geometry. To develop robust closed loop control of IPF, non-contact sensing of part geometry is needed. This research evaluates the feasibility of optical sensing of 3-D contours on manufactured surfaces, and investigates the robustness of the measurement and the speed of data acquisition of the optical sensor for the IPF application. In addition, issues related to improving the part geometry based on optical sensing are considered.
A literature survey and determination of the sensing requirements for the IPF application was done, and indicated that the laser triangulation optical sensor best fulfilled the requirements. Profile geometry data on tubes was gathered using the sensor and issues related to sensor use under practical conditions of the IPF application were identified. Optical sensor data was compared with more accurate contact sensors, for the purpose of evaluating accuracy of the optical sensor. Sensor measurements were accurate to within 0.1mm, and data could be obtained at 40Hz. Sensing of the entire cross-section geometry requires multiple sensors that need to be used in a coordinated fashion, and issues in integrating the multiple sensor outputs were identified. Inaccuracy caused by the bending of the tube occurred during the IPF process continues to be a significant issue, which needs to be analyzed further. Future work related to the use of multiple sensors and options for eliminating the bending of the tube is also identified.