| Cold-formed steel members are prone to initial geometric imperfection due to their lightweight characteristics and the influence of uneven production and processing technology,transportation,storage,and assembly factors.Geometric imperfection have a certain impact on the ultimate strength,stiffness,and post-buckling behavior of the members,which in turn reduces the stability of the structure.However,traditional imperfection measurement techniques have low efficiency and accuracy,making it difficult to obtain a large number of high-precision geometric information samples of the members and truly quantify the impact of geometric imperfection on the mechanical performance of the members.In response to these problems,this paper takes cold-formed C-shaped steel members as the research object and conducts the following research to draw conclusions:This study proposes a machine vision inspection technology for cold-formed C-shaped steel members based on high-precision 3D laser,aiming to address the complexity and low accuracy of geometric imperfection measurement.The technology is supported by intelligent geometric information processing algorithms,including data structure reconstruction and geometric feature recognition algorithms.Firstly,the point cloud of the member is processed for inner and outer layer segmentation,obtaining the target point cloud data.Then,the point cloud data structure is reconstructed based on point cloud sorting algorithm.Finally,the feature recognition algorithm is utilized to divide the section point cloud of the member into geometric features,which can be used for geometric imperfection research and high-precision structural analysis.The intelligent geometric imperfection detection algorithm is based on the full component imperfection classification,recognition,and extraction under digital driving,and reveals the feature rules of accurate and complete geometric imperfection data.The following conclusions were drawn through analysis: the largest global imperfection often occurs in the middle position of the component,which conforms to the general rule;the local imperfection did not show a clear pattern in time domain analysis,but the distortional imperfection on both sides of the flange roughly showed symmetrical changes.By classifying and statistically analyzing the imperfection amplitudes along the length direction of the component,a imperfection probability table for three commonly used C-shaped steel members: C180-70-20,C200-70-20,and C280-70-20,was obtained,providing key reference data for random geometric imperfection simulation.Cold-formed steel members are highly sensitive to geometric shape and geometric imperfection.In order to accurately predict the impact of imperfection on the mechanical properties of the members,a high-precision finite element model must be established.Based on machine vision detection technology,a high-precision geometric model of the members was obtained by scanning the members in all directions for numerical simulation.At the same time,material tests and axial compression verification tests were designed for cold-formed steel members to obtain material parameters,real post-buckling geometric shapes,ultimate bearing capacity,and deformation phenomena,providing key verification data for high-precision numerical simulation methods of cold-formed steel members.A high-precision numerical simulation method based on digital twins was proposed to model and parameterize the reconstructed and feature-recognized point cloud data,and to obtain a 1:1 twin finite element model.The accuracy of the numerical simulation method was validated using the above axial compression test results,and a random imperfection model was established using the validated simulation parameters.In this study,traditional imperfection simulation methods and one-dimensional spectral analysis were used to simulate random imperfection,and different random imperfection simulation methods were studied using the high-precision finite element model as a tool to analyze the effects of random imperfection on structural analysis.When using the one-dimensional spectral method for random imperfection simulation,the amplitude value with an energy percentage threshold of80% was extracted and coupled with random phase within the range of for analyzing the effect of random imperfection on the load-bearing capacity of the component.It is worth noting that although geometric imperfection did not exhibit obvious patterns in the time domain,a certain pattern was found in the frequency domain analysis,which can be used for modeling and analyzing section imperfection.Through analysis,the conclusion was drawn that the one-dimensional spectral method not only considers the randomness of amplitude in the frequency domain but also considers the randomness of shape,which can better predict the ultimate bearing capacity and deformation mode of the component,while traditional imperfection simulation methods are more conservative. |
PDF Full Text Request