PSI - Issue 44

Federico Germano et al. / Procedia Structural Integrity 44 (2023) 902–909 F. Germano et al./ Structural Integrity Procedia 00 (2022) 000 – 000

904

3

and wind data have to be correlated with structural response (typically lateral acceleration or displacement). The exact synchronization of the meteo station with the response data is required to properly address input/output correlation. • Traffic pattern is one of the main load concerns for a bridge. The synchronization of Weight in Motion sensor data is both used as stochastic information for the mean load of the bridge, correlated with his response in terms of acceleration/displacement/stress, as well as deterministic information where load data and speed data of the vehicle are in synch with camera images and the response of the bridge in one or several specific structural point (typically, the displacement of the bridge center over time, correlated to the specific vehicle passage of an oversized load. Load-induced displacement lines can be generated only with perfect synchronous data between load characteristic, position over time and istantaneous structural response. This is even more relevant for train passages, in railroad bridges, because the traffic train pattern is profoundly changed in the last 20 yrs (and there are major signs that will again dramatically change in the near future) and there is special need to understand if actual pattern is still inside design parameters. In railroad bridges, the passage of train can be measured with piezo or strain sensors, and with proper analysis it is possible to achieve train number, wagon number for each train, loading distribution, response of the structure. Also in this case there is a requirement for a measurement system able to cope and condition different sensors and synchronize everything both for load data and response. • Temperature excitation is causing variation in structure displacement. For some structures, like towers, long bridges, especially in areas with high thermal excursion, this condition is the main load acting on the artifact. Temperature data, displacement data and acceleration data are often required together, as displacement can vary in amplitude and direction due to thermal effect and modal behaviour can also significantly change due to difference in stiffness. The transition pattern between one temperature to another (day/night) is also very interesting where the change in stiffness can cause some resonance frequency to be coupled with other excitation data, like wind. This requires the acquisition of sometimes hundreds of temperature sensors, both air and concrete temperature, properly conditioned, synchronized with response data that very often are more quick in frequency (named, accelerometers). Correlation and model updating between simulation and test is one of the main reasons to acquire measurement data. The model, correlated, can be used for qualitative prediction of the structure, as well as helping the insight of the data acquired through monitoring system. The techniques used to gather experimental dynamic information are different because aimed to different goals. Classic experimental modal analysis (EMA) are based upon FRF, typically A/F, and requires measurement of both excitation and response. Excitation must be unique, so the response of the structure has to be ideally 100% coherent only with the force applied and measured: all the response is due to the measured excitation. This condition can be approximated using a prominent force source (coming for instance from a VibroDyna), measured through proper load cell, limiting the input of other forces into the structure (for instance, for a bridge, stopping the traffic through the time of test and choosing a day without other environmental forces acting on the structure) and check that the force level is properly exciting the structure through analysis of the COH function. EMA techniques are used when there is possibility to control the status of the structure (no traffic – no other loads), requires less stringent specification for response sensors (it is possible to raise the force level to get a better response), can give clean data in mere minutes after the measurement start and it will be easier for a simulation engineer to 1.2. Correlation with simulation model, SHM techniques