Application of non-invasive seismic techniques for the characterisation of a gravity concrete dam.

The use of non-invasive investigation methods is becoming frequent as a support for the monitoring of water reservoirs and for the management of dam safety. In this context, geophysical techniques are much less invasive than geotechnical tests and allow to obtain two-dimensional or three-dimensional representations of the mechanical parameters of an engineering structure. A case of application of constrained inversion and joint interpretations of non-invasive seismic techniques is discussed, in order to monitor the Dirillo gravity concrete dam, in Sicily. To investigate the foundation soil, a stratigraphic borehole and a vertical seismic profile, carried out in the same hole, were used to constrain the inversion of a MASW. This latter allowed to generate the starting model for a seismic refraction tomography. The results of the seismic surveys on the ground surrounding the dam allowed a detailed seismic characterisation the foundation soil. Furthermore, for the characterisation of the concrete of the dam, a vertical seismic profile was acquired, using a vertical inspection duct in the dam body, and three seismic refraction tomographies were performed placing sources and receivers on the two opposite walls of the dam. Also in this case the tomographic inversions were constrained by the P wave velocity values obtained from the vertical seismic profile. 2D representations of P wave velocity and the trend with depth of S wave velocity and Poisson ratio allowed an accurate reconstruction of the state of the concrete conservation of the dam. The high velocity values of the P and S waves, together with normal values of the Poisson ratio, suggest that the concrete is still of good quality and local decreases in the P wave velocity are due to moisture and infiltrations. The results obtained confirm the efficacy of seismic refraction tomography, if appropriately constrained, for the characterisation and validation of concrete in large-scale infrastructures that are difficultly investigable using direct methods. [ABSTRACT FROM AUTHOR]