Steel bridges in coastal environments may experience pitting corrosion in the early stages of corrosion,and the appearance of corrosion pits can easily exacerbate the local stress concentration and the deterioration of struc-tural bearing capacity,which affecting the safety and reliability of steel bridge service.Q345qD bridge structural steel was taken as the research object,the samples were designed and a 64 h full immersion corrosion test was con-ducted,the surface morphology point cloud data of the corrosion samples was collected through the KathMatic laser microscopy,and a high-precision method was proposed for extracting corrosion pits and their geometric parameters based on point cloud data.The statistical filtering algorithms was used to smooth the point cloud data,the Random Sample Consensus(RANSAC)algorithm was used for planar segmentation of corrosion pit point clouds,the Density-Based Spatial Clustering of Applications with Noise(DBSCAN)algorithm was used to cluster the corro-sion pits and obtain the key data of corrosion pits.The Alpha Shape algorithm and Graham algorithm were used for corrosion pit extraction and labeling.The extraction results show that the Alpha Shape algorithm can accurately obtain the depth and the surface area of the corrosion pits.Compared with Graham algorithm,the accuracy of extracted pit surface area using Alpha Shape algorithm is overall improved by 22.39%.The statistical analysis was conducted on the depth and surface area of corrosion pits,the distribution empirical functions were fitted,and the distribution hypothesis tests were done.The results show that at a significant level of α=0.05,the corrosion pit depth of Q345qD steel samples at the pitting stage conforms to the Gumbel,Logistic and Weibull distribution,and the correlations is 99.10%,96.23%,and 99.02%,respectively,while the corrosion pit surface area follows the Logistic distribution,with a correlation of 97.02%.This method can provide reference for the research of pitting corrosion distribution of Q345qD steel in the same environment and the decay of mechanical properties of Q345qD steel under random pitting corrosion.
NETL
NSTL
万方数据
