MS17.5 - Structural Health Monitoring

The field of structural health monitoring (SHM) is based on collecting data, and according to that, useful information about the monitored structure is obtained. Therefore, it is essential that the signals from the sensors used in the monitoring campaign can be trusted. However, some of these sensors will sometimes fail or will, at least, pick up excessive noise that might influence the quality of the information extracted from the data. In this paper, a new technique of sensor fault detection is illustrated that is based on constraints between multiple output signals from the monitored structures.

The dynamic identification of modal parameters plays a fundamental role in structural health monitoring: mode shapes, frequencies and damping ratios can be exploited to assess the current state of a structure, used for damage detection or in numerical model validation. In recent years, vibration-based methods have become a popular solution for the state of health estimation of strategic civil infrastructures such as bridges: in particular, free vibration tests represent a fast and economic method only requiring the temporary installation of a limited number of sensors on the structure. This contribution presents an automatic procedure for the identification of modal properties of existing bridges exploiting their free decay responses: each mode’s contribution is adaptively extracted from free vibration tests data using the Variational Mode Decomposition after a proper tuning procedure and a noise-robust area-based approach is exploited to identify modal damping ratios. The method is preliminarily validated on synthetic multi-modal signals showing excellent results even in case of closely spaced modes. The performance of the proposed approach is also tested for two real existing structures: the first case is the identification of modal parameters for a prestressed concrete girder bridge deck; the results are compared with the ones provided by a finite element model of the structure. The second case-study deals with the dynamic identification of cables in a stay-cabled bridge: modal damping ratios and frequencies are compared to estimates from ambient vibration data analyzed with alternative techniques. The identified frequencies are related to cable stresses and are exploited to critically study the effect of relaxation phenomenon on the cables.

The improvements in the development of reliable and low-cost sensors, capable of measuring different structural response quantities (e.g. acceleration, strain, etc.) have attracted extensive research in the field of Structural Health Monitoring (SHM) over the last decades. However, despite the advancements in the field, SHM has not been yet translated to extensive full-scale applications on real structures and infrastructures. In parallel, in the last few years, new paradigms such as the Internet of Things (IoT) have gained a lot of attention in both industry and academia. Fundamental enabling technology for the IoT framework is the fifth generation (5G) network, a solution conceived to satisfy the increasing demand of mobile data traffic and meet the stringent requirements of the emerging IoT applications. Among the many emerging applications that can potentially rely on a 5G communication network and exploit IoT solutions, great interest has emerged in SHM applications on critical structures and infrastructures, since unfortunate recent events have drawn primary attention to the challenges related to public roads safety. Starting from this premises, the present paper describes the pilot application of a new wireless accelerometric prototype to an ongoing structural monitoring activity over a very common kind of prestressed concrete bridge in Italy. The infrastructure under investigation is located in a small town near Genoa, Northern Italy. It is a 200 m long beam bridge with 27 m span. It has been equipped with a monitoring system including environmental and structural response sensors. A meteorological station is installed on a lighting column recording wind speed, atmospheric pressure, temperature and humidity with a sampling frequency of 1 Hz. Structural response is recorded by six triaxial MEMS accelerometers, placed on the outer beams of the first two spans, characterized by 100 Hz sampling rate. For accelerometric data acquisition and transmission, in this work we propose a low-cost, versatile and self-sustaining hardware prototype based on the Narrow-Band IoT (NB-IoT) technology. Each sensor is connected to a Raspberry Pi 4 (RPi4) for real-time edge processing of the collected signals. The RPi4s are equipped with a 5G module, allowing the system to connect the remote nodes by exploiting the cellular network infrastructure through NB-IoT technology without the need of ad-hoc connectivity. Data from the sensors are recorded continuously since October 2022. Identification of the dynamic properties of the structure is carried out through output-only operational modal analysis (OMA) techniques. Results are compared with the outcomes of the tests performed in a preliminary phase of the experimental activity, when classical cabled accelerometers were used. In addition, the variability of bridge’s dynamic characteristics is investigated in near real-time for different environmental and operational conditions, representing the first step of a potential SHM application for the active maintenance of the infrastructure. The research presented in this paper has been developed in the framework of the project ”5G SMART ROADS PER GENOVA” (5GSMARTG), partially funded by the Italian Ministry of Economic Development (MiSE).

Mass timber is a wood product category that includes various alternatives. One of them is a veneer-based product known as Mass Ply Panels (MPP), which was recently introduced and certified per ANSI-PRG 320. A full-scale three-story mass timber building structure, which is 12.2 m by 12.2 m in plan and 9.1 m tall, was constructed and tested at Oregon State Uni-versity to demonstrate the potential of MPP in building application. The building structure comprises MPP diaphragms and walls, and laminated veneer lumber (LVL) beams and col-umns. Two dynamic tests were performed to characterize the dynamic properties of the structure. First, an implosion of a stadium within 300 m of the building location was used as the main excitation source, during which bi-directional horizontal acceleration data were collected for approximately 10 seconds. Second, an ambient vibration test was conducted to collect horizontal acceleration data for one hour. In both tests, sixteen accelerometers were used to record the structure's response. Due to the nature of the tests, the modal features were extracted using an output-only method and compared with the estimates from a finite element model. Lessons learned can be used to inform future modeling efforts of a mass tim-ber building to be tested on the Experimental Facility at the University of California, San Diego's high-performance outdoor shake table (LHPOST) and future codes.