Exploring the Performance of Low-Cost Part Scanning in Wire Arc Additive Manufacturing

Wire arc additive manufacturing (WAAM) is a metal additive process which utilizes traditional welding equipment in conjunction with robotic manipulators to create metal parts for a variety of applications. The WAAM process requires material deposition which involves rapid material melting and solidification. Heating and cooling cycles are extensive and produce internal material stresses in turn creating distortion. Distortion can become extensive and commonly produces nonconforming parts, reducing yield, and increasing manufacturing time and cost. To address these challenges, 3D scanning can be employed as an in-situ monitoring technique to detect distortion as it develops by gathering accurate scans in-between layers. Surface measurements using 3D scanning are desired with uncertainties below 1 mm. Existing scanning solutions already exist, but are expensive and difficult to co-locate in welding cells where excessive operating temperatures, weld spatter, and airborne soot often occur. Therefore, a low-cost solution is necessary. Combining an inexpensive depth sensor, a robotic manipulator already utilized for WAAM processes and data processing scripts creates a promising solution for low-cost 3D scanning. Open3D integration is utilized alongside aligned image frames with robot position data. Performance experiments were conducted to access the system’s capability and limitations using test artifacts of different shapes and sizes. Robot scan velocity was varied to test effects on surface reconstruction. Experiments scanning test artifacts shows the low-cost system is capable of producing meshes accurate to 1 mm. Test artifacts also demonstrated limitations of the system when evaluating parts with low-texture and sharp edges being prone to poor scan quality. Experiments were conducted during WAAM processes and demonstrate the ability of the system to be implemented in-situ to monitor part geometry.