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

905

4

correlate his model to cleaner data. On the other side, EMA cannot be a practical way to monitor several structures (like bridges) over time, as there’s no possibility to apply such technique consistently when the bridge is open to traffic. On the other side, OMA techniques are very much applicable on civil infrastructures for monitoring purposes, as they can rely upo n “natural” excitation of the structure to address his modal behaviour. In this technique, the main function to be calculated upon acceleration signals are the crosspectra (CS) realized between one or more acceleration point (acting as a “reference”) and s everal acceleration responses. The resulting function is dimensionally an acceleration squared, and given that the measurement condition is coming from natural excitation (presumably lower than imposed force like in the EMA) it requires more stringent requirement over sensors (accelerometers) in response. Acceleration sensors needs to have a low noise floor, to being able to detect small acceleration signals born from small excitation forces detecting it correctly from the noise floor. Also, acceleration sensors have to ensure a good phase response in the frequency range of interest (normally well below 30 Hz for civil structures), without generating an oscillating phase response that couldn’t allow the test analyst to detect properly frequency, damping and modeshapes from a stabilization diagram. Moreover, temperature variation should not mask the acceleration signal or -else- alter his shape and create mistakes in the determination of any modal parameter. A measurement system should also allow the possibility to register the acceleration time histories for hours, without any constraint, as measuring over hours is the key to first-level strategy to remove noise from signal: make many averages. One of the most promising techniques used for structural health monitoring purposes is evolution monitoring of the MAC matrix overtime. MAC matrix is a far more rich environment to check overtime for differences in respect of mode frequency or other single-parameter technique. Bonifacio, Barbosa and Cury ran damage detection methods in a numerical study, based on several method including MAC evolution overtime, changing tempreratures, noise levels and excitation type (random/impulsive) on the structure. The result is indicating robustness of the MAC indication against temperature variation, noise or type of excitation, false positive indication are less common to happen. On the measure architecture side, that means having tens (or more!) acceleration signals for each structure to be acquired all together with very low (close to zero) phase error, for hours; regardless of the quantity of data generated, the system should continuously work. Also, as any structure is different from the other, the system should allow the possibility to use a wide variety of analysis tools, so cannot be a “closed” system but should openly communicate with other software or algorithms. This is even more important when the engineers face the effort of monitor many structures, each one different, in a wide territory, and needs to have a tool of asset management to trace issues, setup maintenance plans, … for the whole installed base. In this last case, it is of upmost importance the capability of the monitoring system to properly connect and feed data to the asset management tool. New techniques, based on innovative optical fiber measurement for the deformation shape of the bridge, needs also capability to be acquired in synch with the rest of the instrumentation. Optical fiber can catch the shape of a bridge and come back with a signal giving the deformation pattern, to be put in synch with the loads acting each time.

1.3. Requirements target setting Summing up the main requirements discussed up to now, a pattern emerges:

• The need to synchronize several type of sensors, quasi-static and especially dynamic, without phase issues;