EACS 2012 – 5 th European Conference on Structural Control Genoa, Italy – 18-20 June 2012 Paper No. # 204 1 * Corresponding author Automatic Identification of Modal Parameters: Application to a Reinforced Concrete Arch Bridge Filippo UBERTINI*, A. Luigi MATERAZZI University of Perugia, Department of Civil and Environmental Engineering 93 Via G. Duranti, 06125 Perugia, Italy ubertini@unipg.it, materazzi@unipg.it Carmelo GENTILE Politecnico di Milano, Department of Structural Engineering 32 Piazza Leonardo da Vinci, 20133 Milano, Italy gentile@stru.polimi.it Fabio PELLICCIA Province of Perugia 21c Via Palermo, 06124 Perugia, Italy fabio.pelliccia@provincia.perugia.it ABSTRACT The paper presents the results of the experimental modal analysis of a reinforced concrete arch bridge, about 70 m long, crossing the Tiber river close to the city of Umbertide in central Italy. The dynamic tests were performed in operational conditions in February 2012 and represented the first known experimental survey carried out on the global dynamic characteristics of the bridge since its completion which approximately dates back to the early 50’s. Baseline modal parameters of the structure were extracted through an automated procedure recently proposed by the authors. Eight sufficiently long consecutive data sets were considered for this purpose and variations of modal parameters estimates among the different data sets were investigated. Keywords: ambient vibration testing, operational modal analysis, automatic system identification, structural health monitoring. 1 INTRODUCTION Structural health monitoring (SHM) based on dynamic measurements is a very active field, especially in bridge engineering, which main idea is to monitor the health state of a structure by looking at its dynamic response. Currently, it is widely recognized that the potential of SHM is mainly associated to the possibility of performing a continuous acquisition of the structural response and a real-time processing of a large amount of data [1-5]. Hence, a significant increment of interest has arisen, in the last few years, towards the development of stand-alone automatic system identification (SI) techniques that are able to process a large amount of data without any user interaction during their operation. The authors have recently developed a similar SI procedure [5] in the framework of subspace methods [6,7].