Laser Scanning-Based Diagnostics In The Structural Assessment Of Historic Wrought Iron Bridges

Abstract

This paper introduces a workflow to create the geometric documents for conducting finite element based structural assessment of wrought iron bridges using laser scanning data as the input dataset. First, a methodology for identifying actual cross-sections of the bridge components based on a point cloud obtained from a terrestrial laser scanner (TLS) is presented. Next, a non-parametric regression kernel density estimation is employed to determine the overall bridge dimensions to populate a computation model by projecting the position of the web and/or flange surface of the cross-section (appearing as local maximum peaks of a probability density shape). The process is demonstrated with respect to the previously undocumented Guinness Bridge in Dublin, Ireland to determine the bridge’s behaviour. The successful generation of this model proves that TLS can surpass other common techniques (e.g. UAV-based images) for acquiring the bridge geometry necessary for reconstructing accurate member cross-sections and overall bridge dimensions, regarding quantity and quality of the data points, and timing. The finite element analysis showed that the bridge currently satisfies both strength and serviceability requirements under self-weight, but would be unlikely to support a new slab and a modern pedestrian load level as per current code requirements for re-opening the bridge.