Side-Surface-Form Analysis of Arc Additive Manufacturing Based on Laser Vision Sensing System

A laser vision sensing three-dimensional (3D) reconstruction system has been designed to collect the surface-profile-depth point cloud information of the welded side surfaces under different welding conditions and to solve the problem of surface-form detection of the side surfaces obtained by using multi-layer single-channel arc additive manufacturing. Point cloud processing algorithms, such as RANSAC (Random Sample Consensus) and KNN (K-Nearest Neighbors), arc used to extract the 3D point cloud of a deposition layer. We analyze the interlayer distribution of the multi-layer single-channel weld deposition, quantify a side-surface roughness of the deposition layer, and investigate the influence of the distance between the end of the welding wire and the plate on the 3D forming of the deposition layer. The results denote that the laser vision sensing system can accurately determine the surface-forming condition of the side surfaces obtained by using arc additive manufacturing. The 3D point cloud algorithm is used for 3D reconstruction and feature extraction of the side surfaces obtained by using arc additive manufacturing to visually describe and quantify the 3D forming features of the deposition layer, providing a novel method for performing surface-form inspection and quantitative analysis in case of arc additive manufacturing.