A comprehensive methodology for residual capacity evaluation of damaged steel tubular columns

Steel tubular sections are extensively used in modern-day construction due to their superior structural efficiency and aesthetic appearance. The thin-walled tubular sections are often subjected to accidental loading due to heavy object impact, collision, or improper handling during transportation. Due to induced geometric defects by unanticipated loads, the load-carrying capacity of the section decreases. This study proposes a methodology to estimate the residual capacity of damaged circular and square tubular specimens in axial compression. Initially, the specimen's surface coordinates were obtained in the form of a dense point cloud using a 3D digital image correlation (DIC) setup. Using a linear interpolation function based on Delaunay triangulation, the dense point cloud was regularized into a set of sparsely spaced nodes that would yield computationally efficient finite element (FE) analysis. The interpolation introduced errors at the locations of sharp curvature on the surface. A systematic smoothing (filtering) approach was proposed to minimize the effect of the interpolation error. The proposed methodology was validated against compression tests on five damaged and three as-built specimens, which consisted of circular and square tubular sections. The average residual strength estimated was found to be 93% of the experimental observations.