Preliminary Validation of an Automatic Modal Identification Methodology for Structural Health Monitoring of Historical Buildings

Automatic structural modal parameter estimation is a key aspect for a continuous health monitoring process. In the case of existing buildings and historical constructions, the implementation of an automatic process for structural health monitoring (SHM) is important since it provides qualitative information of the structural behavior through time. For parametric system identification, this task is faced with the automatic interpretation of stabilization diagrams and the elimination of the mathematical or spurious modes. The present paper shows a fully automated modal identification methodology based on the application of four processing stages: (i) digital signal pre-processing of the recorded data and the application of the Data-Driven Stochastic Subspace Identification method to obtain modal parameters; (ii) automatic analysis of the stabilization diagram with the application of soft\hard validation criteria and the use of hierarchical clustering approach; (iii) automatic estimation of the modal parameters (frequencies, damping and modal shape) of the structure and (iv) automatic modal tracking of the data with a time-windows technique. This paper presents the adopted methodology to automatically track the variation of the modal parameters of earthen systems and a particular application in a sixteenth century adobe church. The developed methodology shows good results for the task of cleaning of stabilization diagrams which then allows the accurate automatic estimation of the modal parameters.