Detection and Localisation of Damage on Industrially Produced Concrete Slabs Through Time- and Frequency-Domain Approaches

The objective of this work is to address the problem of damage detection in civil engineering structures using non-destructive techniques and dynamic measurements. To this purpose, time- or frequency-domain methods are used for the diagnostics. It consists in practical output-only techniques as Stochastic Subspace Identification (SSI) for modal identification or Enhanced Principal Component Analysis (EPCA) for detecting the presence of damage. The use of the Hankel matrix instead of the observation matrix improves effectively the robustness of these methods. Damage localization is based on Frequency Response Functions (FRFs) and sensitivity analysis of PCA results.
The efficiency of the above-mentioned methods has been demonstrated in earlier studies mainly on numerical models and small-scale laboratory experiments [3, 4]. It was also tested successfully on industrial examples to perform machine condition monitoring using a reduced set of sensors [2]. In this work, the investigation is performed on precast prestressed and non-prestressed concrete slabs. Successive damages were artificially introduced in the slabs by loading heavy weights and by cutting steel wires, which induced cracks in the structure. The examples show the consequences of the considered techniques for damage identification. The results that are very different between prestressed and non-prestressed slabs may be used as input for the condition control of this kind of structures.