Contents lists available at ScienceDirect Tunnelling and Underground Space Technology journal homepage: www.elsevier.com/locate/tust A coupled 3D laser scanning and digital image correlation system for geometry acquisition and deformation monitoring of a railway tunnel Behzad V. Farahani a,b, ⁎ , Francisco Barros b , Pedro J. Sousa a , Pedro P. Cacciari c , Paulo J. Tavares b , Marcos M. Futai c , Pedro Moreira b a FEUP – Faculty of Engineering, University of Porto, R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal b INEGI – Institute of Science and Innovation in Mechanical and Industrial Engineering, R. Dr. Roberto Frias., N 400, s/n, 4200-465 Porto, Portugal c EPUSP – Polytechnic School of the University of São Paulo, São Paulo State, Brazil ARTICLEINFO Keywords: Railway rock tunnel 3D Laser scanning system Geometry acquisition NDT SHM DIC ABSTRACT In this study, a 40 m-long section of a shallow railway rock tunnel so-called “Monte Seco tunnel” located at Vitoria Minas Railway in Brazil is investigated. In the real environment, the tunnel has been scanned by a 3D TerrestrialLaserScanning(TLS)instrumentationanditsgeometrywasreconstructedonapointcloud.Duetothe complexgeometryoftheirregularrockfacecreatedbydrillandblast,theobtainedgeometrylackedanumberof regions (occlusions). In this study, a small-size model of the obtained point cloud was built through additive manufacturing and submitted to laboratory tests. A scaled demonstrator was developed for the acquisition of the 3D tunnel model profle with a Laser Scanning System (LSS), comprising a camera and a circular laser that scans the entire tunnel surface. After 3D geometry acquisition, the tunnel model was compressively loaded by im- posing a displacement from the exterior wall and the deformation was monitored by a 3D Digital Image cor- relation (DIC) system setup adapted to the rail structure. Promising results have been accomplished and the achieved tunnel’s point cloud verifed the geometrical characteristics with minimum occlusions. Owing to ob- tained successful results on the geometry documentation, a real-scale 3D LSS model is presented for operation in a real environment. 1. Introduction A large number of ageing civil infrastructures, such as railway tunnels, are progressively deteriorating and in prompt necessity of in- spection, damage assessment and repair due to senescence, environ- mental factors, over loading, operation and insufcient maintenance or deferred repairs. The inspection requirements are more demanding in underground transportation tunnels including a number of tunnels operating for more than half a century, which already present large evidences of corrosion, whereas there are some collapse paradigms (Balaguer and Victores, 2010). The signifcance of appropriate inspection methodologies to address potential problems before they reach a critical status and the infra- structure becomes inoperative, can’t be overstated, in view of the im- plications of shutting down some of these infrastructure, even for short time spans. Detecting and characterising these defects is therefore de- cisiveforefective predictive maintenance (Menendezetal.,2018;Mori et al., 2002; McCann and Forde, 2001). Manual and visual approaches to tunnel inspection may demand personnel to access hazardous environments has triggered the search for automatic inspection operations, in order to minimize human in- tervention. Generally, a system for automated inspection of railway tunnels involves three main phases: image acquisition, processing, and result exploitation (Medina et al., 2017). The use of automated struc- tures in the inspection feld had been a cooperative research area, and several studies (Balaguer and Victores, 2010; Protopapadakis et al., 2016; Loupos, et al., 2014) reviewed the advantages of the use of au- tonomous platforms in railway tunnels. These systems have the po- tential to fulfl the inspection process with objective outcomes and high profciency. They also enhance safety by performing inspection in dangerous environments, relieving the involved workmanship (Monteroetal.,2015).Consequently,manualandvisualinspectionsare being substituted with more precise systems using mechanical, elec- tronic and automated systems and processing data provided by cam- eras, lasers and other sort of sensing devices (Menendez et al., 2018; Montero et al., 2015; Balaguer et al., 2014). https://doi.org/10.1016/j.tust.2019.102995 Received 15 April 2019; Received in revised form 29 May 2019; Accepted 8 June 2019 ⁎ Corresponding author at: FEUP- Faculty of Engineering, University of Porto, R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal. E-mail address: behzad.farahani@fe.up.pt (B.V. Farahani). URL: https://orcid.org/0000-0001-8901-9493 (B.V. Farahani). Tunnelling and Underground Space Technology 91 (2019) 102995 0886-7798/ © 2019 Elsevier Ltd. All rights reserved. T