A novel method and modelling technique for determining the initial geometric imperfection of steel members using 3D scanning. (March 2023)

• Record Type: Journal Article
• Authors: Xu, Dongdong; Wang, Yuanqing; Liu, Xiaoling; Chen, Boshan; Bu, Yidu
• Is Part Of: Structures. Volume 49(2023)
• Page Start: 855
• Page End: 874
• Publication Date: 2023-03
• Subjects: Initial geometric imperfection -- Data processing algorithm -- Modelling technique -- 3D laser scanner; Structural engineering -- Periodicals; 624.1
• DOI: 10.1016/j.istruc.2023.01.136
• Languages: English
• ISSNs: 2352-0124

Abstract

Abstract: Initial geometric imperfections significantly affect the strength and stability of steel members. The traditional modelling method only considers the amplitude of imperfection, which may lead to unreliable results. A novel method and modelling technique for obtaining the full-filled initial geometric imperfection of steel members was proposed in this study. A handled 3D laser scanner was employed to generate 3D point-cloud model, and a computer software of data processing algorithm (3D Pr.) was developed to convert the 3D point-cloud model to actual digital geometric model. After that, such geometric models can be directly imported into general FE software. Compared with the traditional modelling method, a FE model with full-filled initial geometric imperfection can be developed automatically, and the overall and local initial geometric imperfections can be extracted separately. Finally, a total of 27 test results reported by the authors previously were compared against the results determined from two modelling methods. Upon comparison, it was found that the results determined from the new modelling method were close to those values obtained from the experiments (only 1.75% difference), while the traditional modelling method was found to be conservative by 4.11%. The calculation efficiency of new modelling method can be improved by 9.09%, compared to traditional modelling method. It indicated that the new modelling method proposed in this study can closely predict the initial geometric imperfection of steel members.