Rehabilitation assessment of a centenary steel bridge based on modal analysis

Modal testing through ambient vibration holds unique advantages over other experimental techniques, including static load tests, for the measurement of global parameters that characterize the bridges' behaviour, among which are the fastness of execution, low cost, non-compulsory restriction of the traffic during the test, and the ability of simultaneous evaluation of different directions. Moreover, quantities that decisively influence the structural safety, such as damping for the fatigue resistance in old steel bridges, can only be measured through appropriate dynamic testing. On the other hand, the use of this non-destructive testing technique to assess the changes in rehabilitated structures and to enable the evaluation of the actual effectiveness of different rehabilitation strategies emerge as extremely important, since the current knowledge in this field is short and therefore demands an improvement in order to achieve more cost-effective designs without lowering the required safety levels. This paper presents the ambient vibration tests conducted on a centenary through-truss steel bridge before and after its rehabilitation, with the purpose of evaluating the changes in its dynamic properties as a result of the adopted strengthening strategy. The implemented testing program, experimental setup, data processing and modal identification technique are described. Field data validated numerical models for the analysis of the structural changes produced by the strengthening process are presented. Significant conclusions were drawn by comparing the experimental and numerical results between the pre and post-rehabilitation conditions, namely in what concerns the vibration level experienced by the structure, its stiffness variation and suitability of the adopted modelling methodology. Furthermore, an unexpected coupled behaviour of the simply supported spans that constitute the bridge, translated by a global vibration mode, could even be detected and validated. (C) 2013 Elsevier Ltd. All rights reserved.