Automatic Identification and Geometrical Modeling of Steel Rivets of Historical Structures from Lidar Data

Abstract

Riveting is a joining technique that was widely used in iron and steel structures from the mid-nineteenth to the early twentieth century. Nowadays, many of these riveted structures are still in service around the world, and in many instances, also constitute an important part of the built cultural heritage. The maintenance and conservation of this type of construction is a crucial task for which the HBIM (Historic Building Information Modeling) methodology has recently gained increased attention, postulating itself as the ideal tool for tracking control and conservation-related actions. In the process of data collection and 3D modeling of the structure, the rivets are an important part to be taken into account in the structural safety assessment and health monitoring over time. Any structure of this typology typically presents thousands of rivets, so its measurement and subsequent 3D geometrical modeling is a laborious task and a source of possible errors. Accordingly, this work presents a novel methodology that allows the automatic identification and 3D modeling of rivets in iron and steel structures from Lidar data. The proposed methodology has been tested with both laboratory specimens and a full-scale real bridge.