Finite Element Modeling of Geometric Imperfections using Point Cloud Scans

Abstract

Finite element modeling, widely used for structural analysis, traditionally depends on idealized geometries that assume perfect fabrication or standardized imperfection profiles. While such assumptions are acceptable in early‐stage design or theoretical studies, they often fall short in capturing the true behavior of as‐built structures, particularly when small geometric deviations lead to significant local effects. This limitation is especially apparent in thin‐walled structural members such as profiled steel decks or angular specimens, where local imperfections may influence buckling modes, stress concentrations, and overall capacity under load. To address this challenge, point cloud technology has emerged as a powerful solution capable of capturing high‐resolution three‐dimensional representations of physical objects. Generated through laser scanning or photogrammetric techniques, point clouds provide a dense array of spatial data points that accurately reflect the true geometry of structural components. When integrated into finite element modeling platforms such as ABAQUS, point cloud‐derived geometries enable more accurate and representative simulations. The objective of this study is to develop and validate a comprehensive workflow for integrating point cloud data into finite element models in ABAQUS, with a focus on profiled steel deck and angle specimens. The workflow involves 3D scanning of specimens to capture as‐built geometry, preprocessing the point cloud to remove noise, and converting the data into mesh‐compatible formats. The resulting geometry is imported into ABAQUS to create high‐fidelity models that reflect true geometry.