Desenvolvimento de um sistema para monitoramento automático de integridade estrutural

Abstract

Recent research has used methods based on raw vibration data analysis as a basis for Structural Health Monitoring (SHM). In some of them, the concept of Symbolic Data Object (SDO) has been used and spread. A remarkable and advantageous feature of using SDOs is that they have a great capacity to compress raw data, that is, dynamic signals collected directly from the structure, while maintaining the essence of the original information. Furthermore, methods based on the use of SDOs generally have a much lower computational cost compared to those based on the use of modal parameters, since they avoid the modal identification process. Thus, the present work presents a methodology for performing SHM, implemented in a software called ReTiA-SHM (Real-Time Automated Structural Health Monitoring). To verify the functionality and accuracy in detecting structural changes, data from four dynamic monitoring of structures that suffered some kind of alteration or exceptional force were used. The four experimental applications refer to: i) a steel frame in Canada; ii) a masonry tower in Italy; iii) a road bridge in France and iv) a railway viaduct in France. The results obtained showed that the structural changes or exceptional forces were successfully identified, except for some minor progressive damage to the steel frame. In some cases, it was observed that an adaptation period is necessary – when some false positive alarms occur – so that the software automatically configures the parameters to sound an alarm only in situations when some change of greater intensity is found in the signal.