An integrated seismic monitoring system for a full-scale embedded retaining wall

Over the last 50 years, data from laboratory tests and post-earthquake reconnaissance have been used to gain knowledge about the dynamic and seismic behavior of geotechnical structures and to improve analysis and design procedures. The scarcity of reconnaissance data has pointed out the need for full-scale and near full-scale tests for research purposes in earthquake engineering. Structural health monitoring (SHM) systems have been applied to different kinds of structures, but although static control of displacements and pressures is quite common, dynamic monitoring is fairly limited in geotechnical engineering. In the present paper, an integrated structural and geotechnical monitoring program aimed at filling this knowledge gap is described with reference to a real flexible retaining wall. The objective of the research is to establish a combination of sensors, numerical analyses, and data processing procedures to turn the monitored retaining wall into a smart geotechnical structure. Attention is herein focused on two aspects related to the implementation of the monitoring system and the dynamic identification of the soilstructure system. Specifically, the paper describes an example of seismic monitoring system for full-scale flexible retaining walls based on sensors embedded in the reinforced concrete piles. Installation details are given together with a description of the monitoring system architecture in its current stage of implementation. Issues and requirements for the measurement chain are discussed, taking into account possible installation drawbacks (i.e., shocks) and the opportunity to monitor the response under operational conditions. Furthermore, a novel and successful application of operational modal analysis (OMA) to such a complex geotechnical system is reported. It allows for the identification of the fundamental modes of the soil-wall system in operation (but before the erection of a building on the excavated side). Implementation of the monitoring system prototype is still in progress, but some interesting results have already been outlined.